Final Technical Report Contr act n° QLK5 - 2001 2001 - 0061 00619 9 – Apr il 1st , 2002 2002 - Marc Marc h 31st , 2005 OPTIMISATION OF SCREENING AND CLEANING TECHNOLOGY TO CONTROL DEINKING PULP CLEANLINESS
JACKSTÄDT GmbH
INSTYTUT CELULOZOWO-PAPIERNICZY PULP & PAPER RESEARCH INSTITUTE
Centre Technique du Papier (CTP), France
Project Coordinator
Advanced Fibre Technologies Oy (AFT), Finland Jackstädt GmbH (Avery Dennison), Germany Instytut Celulozowo-Papierniczy (ICP), Poland Papiertechnische Papiertechnische Stifftung (PTS), Germany Laboratoire des Ecoulements Géophysiques et Industriels (LEGI), France
Octob er 2005 2005
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Contract details details and partnership
Title: Optimisation of Screening and Cleaning Technology to Control Deinking Pulp Cleanliness Acronym: S C R E E N C L E A N Type of contract: Shared-cost RTD action Thematic priority: 1.1.1 / 5.3.2
Total project cost 1 619 988 €
Contract number
Duration
EU contribution
QLK5-2001-00619
36 months
809 986 €
Commencement date
Period covered by final report
st
April 1 , 2002 Project coordinator
1 April 2002 – 31 March 2005 CENTRE TECHNIQUE DU PAPIER
Name: François JULIEN SAINT AMAND
Address: BP 251, Domaine Universitaire, 38044 Grenoble Cedex 9, France
Telephone +33 4 76 15 40 25
E-mail address
[email protected]
Telefax + 33 4 76 15 40 16
Key words: Paper Recycling, Deinking, Screening, Cleaning, Stickies World wide web address: www.webCTP.com List of participants Aut hor s Centre Technique du Papier (CTP), France – Coordinator François Julien Saint Amand, Bernard Perrin, Thierry Delagoutte Advanced Fibre Technologies Oy (AFT), Finland – Contractor (AC to CTP, ICP, PTS) Robert Robert Gooding, Antti Huovinen Huovinen Jackstädt GmbH (Avery Dennison), Germany – Contractor (AC to CTP, ICP, PTS) Peter Heederik, Andreas Pahl, Wolfgang Haar Instytut Celulozowo-Papierniczy (ICP), Poland – Contractor Henryk Gonera, Jozef Jozef Dabrowsk i, Tomasz Mik Papiertechnische Stifftung (PTS), Germany – Contractor Lutz Hamann, Oliver Cordier Institut National Polytechnique de Grenoble (INPG) / Laboratoire des Ecoulements Géophysiques et Industriels (LEGI), France – Contractor (AC to CTP) Dariusz Asendry ch, Mich el Favre-Ma Favre-Marinet rinet
2
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Table of Content 1.
Introduction................................................................................................................... Introduction....................................................................................................................................... .................... 5 1.1. Background Background ............................................................................................ ............................................................................................................................... ................................... 5 1.1.1. Recovered paper recycling................................................................................................ 5 1.1.2. Deinking.................................................................................... Deinking ............................................................................................................................. ......................................... 6 1.1.3. Stickies............................................................................................................................... Stickies............................................................................................ ................................... 8 1.2.
2.
3.
Objectives .............................................................................................. ............................................................................................................................... ................................. 10
Project developmen developmentt and partnership partnership ............................................................................................ 11 2.1.
Project structure...................................................................................................................... 11
2.2.
Development Development of of the programme programme ............................................................................................. 13
Material and and methods..................................................................................................................... methods..................................................................................................................... 13 3.1. Equipment and raw materials ................................................................................................. 13 3.1.1. Research means means and pilot equipment equipment ............................................................................. 13 3.1.2. Paper raw materials......................................................................................................... 14 3.1.3. Adhesive raw raw materials.................................................................................................... materials.................................................................................................... 14 3.2. Stickies control methods ............................................................................................. ......................................................................................................... ............ 16 3.2.1. Laboratory Laboratory screening methods (macro-stickies).............................................................. (macro-stickies).............................................................. 16 3.2.2. Stickies size and shape shape analysis analysis ..................................................................................... 17 3.2.3. Extraction methods methods (micro-stickies)................................................................................. (micro-stickies)................................................................................. 19 3.3. Adhesive rheological rheological properties.............................................................................................. properties.............................................................................................. 20 3.3.1. Low-speed elongation elongation tests............................................................................................. 20 3.3.2. High-speed compression tests ........................................................................................ 21
4.
Results and discussions................................................................................................................. discussions................................................................................................................. 23 4.1. Pulping ............................................................................................... .................................................................................................................................... ..................................... 23 4.1.1. Background and and objectives.............................................................................................. objectives.............................................................................................. 23 4.1.2. Study of of basic pulping parameters parameters .................................................................................. 24 4.1.3. Comparison of drum and batch batch pulpers pulpers .......................................................................... 32 4.1.4. Development Development of a new pulping technology...................................................................... technology...................................................................... 35 4.1.5. Mill trials ............................................................................................ ........................................................................................................................... ............................... 45 4.1.6. Conclusions and perspectives perspectives ......................................................................................... 57 4.2. Pressure screening screening ..................................................................................... ................................................................................................................. ............................ 58 4.2.1. Background and and objectives.............................................................................................. objectives.............................................................................................. 58 4.2.2. Numerical simulation studies ........................................................................................... 59 4.2.2.1. Numerical model of pressure screening .......................................................................... 59 4.2.2.2. Numerical flow simulation ................................................................................................ 63 4.2.2.3. Particle deformation analysis ........................................................................................... 66 ..................................................................................................................... ........................ 69 4.2.2.4. Conclusions ............................................................................................. 4.2.3. Optimisation of stickies screening ................................................................................... 70 ............................................................................................................. .......... 70 4.2.3.1. Stickies extrusion ................................................................................................... 4.2.3.2. Optimisation of screen plate design ................................................................................ 77 ............................................................................................. 89 4.2.3.3. High-consistency screening ............................................................................................. 4.2.4. Simulation of screening screening systems ..................................................................................... 93 4.2.5. Conclusions and perspectives perspectives ......................................................................................... 98 4.3. Centrifugal cleaning ..................................................................................... ................................................................................................................ ........................... 99 4.3.1. Background and and objectives.............................................................................................. objectives.............................................................................................. 99 4.3.2. Stickies density ........................................................................................ .............................................................................................................. ...................... 100 4.3.3. Hydrocyclone cleaners .................................................................................................. 101 4.3.3.1. Stickies and cleaner operating parameters ................................................................... 101 4.3.3.2. Cleaner design parameters ........................................................................................... 103
3
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.3.3.3. Cleaning versus screening ............................................................................................ 106 4.3.4. 4.3.5.
Rotary cleaner.......................................................................... cleaner................................................................................................................ ...................................... 108 Conclusions and perspectives perspectives ....................................................................................... 111
4.4. Flotation ........................................................................................... ................................................................................................................................ ..................................... 113 4.4.1. Background and and objectives............................................................................................ objectives............................................................................................ 113 4.4.2. Basics of the flotation flotation process process ...................................................................................... ........................................................................................ 114 4.4.3. Deinking flotation – lab lab and pilot equipment.................................................................. equipment.................................................................. 116 4.4.4. Study of basic basic stickies flotation flotation parameters parameters (lab flotation) ............................................ 117 4.4.5. Pilot stickies flotation tests............................................................................................. 131 4.4.6. Conclusions and perspectives perspectives ....................................................................................... 135 4.5. Pressure filtration ............................................................................................... .................................................................................................................. ................... 137 4.5.1. Background and and objectives............................................................................................ objectives............................................................................................ 137 4.5.2. Basics of pressure pressure filtration process.............................................................................. 138 4.5.3. Pressure filtration – pilot equipment equipment .............................................................................. 139 4.5.4. Preparation Preparation of process process waters....................................................................................... 140 4.5.5. Pilot pressure pressure filtration tests........................................................................................... tests........................................................................................... 142 4.5.6. Conclusions and perspectives perspectives ....................................................................................... 148 5.
Conclusions.................................................................................................................................. Conclusions.................................................................................................................................. 149 5.1. Removal of stickies in deinking deinking lines .................................................................................... 149 5.1.1. Optimisation of pulping pulping to improve further further stickies removal removal .......................................... 149 5.1.2. Optimisation of of macro-stickies removal......................................................................... removal......................................................................... 152 ....................................................................................................................... 152 5.1.2.1. Screening ....................................................................................................................... ......................................................................................................................... 156 5.1.2.2. Cleaning ......................................................................................................................... ......................................................................................................................... ..................................... 157 5.1.2.3. Flotation .................................................................................... 5.1.2.4. Global macro-stickies removal process ......................................................................... 158 5.1.3. Optimisation of micro-stickies micro-stickies removal removal .......................................................................... 161 ......................................................................................................................... ..................................... 161 5.1.3.1. Flotation .................................................................................... ................................................................................................ ............ 162 5.1.3.2. Process water treatment .................................................................................... 5.1.4. Conclusion ...................................................................................... ..................................................................................................................... ............................... 162 5.2. Recycling friendly friendly adhesives................................................................................................. adhesives................................................................................................. 163 5.2.1. Pressure sensitive adhesives ........................................................................................ 163 5.2.1.1. What is a pressure sensitive adhesive? ........................................................................ 163 5.2.1.2. What determines the tack and the adhesion of a PSA? ................................................ 163 ....................................................................................................... .......... 165 5.2.1.3. Emulsion adhesives ............................................................................................. 5.2.1.4. Hot melt pressure sensitive adhesives .......................................................................... 169 5.2.2. Potential influence influence of adhesive components components on the separation separation of stickies ................... 170 5.2.3. Recommendations Recommendations to improve PSA’s ............................................................................ 171
6.
7.
8.
Exploitation and and dissemination dissemination of results results ..................................................................................... 172 6.1.
Exploitation................................................................................................................... Exploitation................................ ............................................................................................ ......... 172
6.2.
Dissemination..................................................................................................... Dissemination........................................................................................................................ ................... 173
Policy related benefits ..................................................................................... .................................................................................................................. ............................. 174 7.1.
Communities Communities added value and contribution contribution to EU policies................................................... 174
7.2.
Contribution to Community Community social objectives objectives ........................................................................ 176
Literature Literature cited.............................................................................................................................. cited.............................................................................................................................. 177
ANNEXE
List and copies of the publications publications resulting from the project
4
182
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
SCREENCLEAN PROJECT FINAL REPORT 1. Introduction 1.1.
Background
1.1.1. Recovered paper recycling Recycling can be considered as a very old tradition of the paper industry as the first papers were manufactured from old rags and recovered papers have always been a source of fibres raw materials. Today, the European CEPI countries represent about 28 % of the total paper and board production, i.e. 95 millions tons out of 340 million tons produced worldwide in 2003 [1]. The average worldwide recovered paper recycling rate is now close to 50% and even higher in Europe. Referring to the CEPI countries considered as the European Declaration on Paper Recovery was launched in 2000, it seems that the target of 56 % fixed for 2005 should effectively be reached as the recycling rate increased from about 50 % in 2000 to 53.9 % in 2003 [1]. This corresponds to an increase by 10 million tons of the recovered paper utilisation rate over this 5 years period. Most of the recovered papers and boards are still used to produce brown grade packaging papers and boards (figure 1) though a drastic increase of their use in the production of white paper grades has been observed over the two last decades through the development of the deinking process.
100 Case materials materials
90 80
) % ( e t a r n o i t a s i l i t U
Newsprint
70 60
s n d o r t r a a o C B
50
r e h t o , r s e g p n a i p p . p k a c r a p
& d l o h y r e a s i t u n o a H s s r e
40 30 20 Other graphic grades 10 0
Share of total paper & board p roduction (%) Mixe Mixed d grad grades es
Corr Corrug ugat ated ed and and kraf kraftt
Newsp Newspap aper ers s and and maga magazi zine nes s
High High grad grades es
Figure 1 : Recovered paper utilisation by sector within the CEPI countries in 2003 [1]
Wood-containing recovered papers, i.e. old newspapers (ONP) and magazines (OMG) are mainly used to produce newsprint, while wood-free recovered papers such as mixed office waste (MOW) are mainly used for the production of “tissue” household and sanitary papers. By contrast, the utilisation of deinked pulp in other graphic grades, such as super-calendered (SC) and light-weight coated (LWC) papers is currently limited to about 10 % on average in Europe, as shown in figure 1. Consequently these paper grades show the highest potential towards increased recycling in the European pulp and paper industry. A few mills already produce SC or LWC papers with up to 100 % DIP [2].
5
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
1.1.2. Deinking Deinking is indeed a way of recycling, which is used to produce high quality papers. White grade papers can be produced from post-consumer or post-industrial recovered papers. This means that the components which cause a reduction of brightness, mainly the inks, must be removed, but also that all the additives used during printing, converting and using the paper must also be removed. From the recycling point of view these additives are contaminants. They include various grades of adhesives (such as binding materials, labels, tapes), staples, plastic films, inks, varnishes, and all the components of the pulp which cannot be used to produce paper. In some cases mineral fillers and coating pigments must also be removed to produce paper grades such as tissue papers and to a lesser extent LWC papers.
Figure 2 : Typical deinking process for improved newsprint, SC and LWC [3]
A typical deinking line proposed for the production of improved newsprint, SC and LWC papers [3] is illustrated in figure 2. Such a deinking process with 2 process water loops or similar processes without washing has become very common in Europe for the production of newsprint and graphic papers [2]. The global deinking process includes a number of pulp treatment processes, i.e. pulping, dispersing, bleaching and refining in some cases, and particle separation processes, i.e. screening, centrifugal cleaning, froth flotation and washing [4], as well as pulp thickening processes with filters and presses and process water treatment processes, including pressure filtration and dissolved air flotation.
Pulping
Recovered papers are separated into individual fibres while inks and all the additives added to the paper during the printing and converting process are (or should be) detached from the fibres, during the pulping step. In the field of deinking, high-consistency batch pulpers or continuous drum pulpers are used to promote the action of the deinking chemicals (caustic soda, sodium silicate and soap) and bleaching chemicals (hydrogen peroxide) normally added in the pulper. Various studies were recently devoted to the pulping process and to the analysis of the defibering and ink detachment kinetics [5-8]. The optimisation of the pulping step is a prerequisite for the optimisation of the subsequent deinking process steps. Increasing pulping time, temperature and pH leads to excessive fragmentation and even re-deposition re-deposition of inks on the fibres, which is detrimental to their removal at the flotation step [5-8]. The effects of the pulping conditions on the stickies size, shape and physical-chemical dispersion had not been investigated as much as for the inks, but the pulping conditions were assumed to have a strong influence on the stickies as well, which motivated the relatively large research effort devoted to the optimisation of pulping in this project.
6
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Dispersion and bleaching
Hot dispersing with high-speed disperser of kneaders [9] is performed after pulp thickening at least once at the end of the first deinking loop. The main objective is to detach the inks still attached onto the fibres in order to remove them in the second deinking loop. Bleaching is normally associated with dispersion to make use of the high consistency and temperature to promote the chemical reactions. These processes also produce some effects on stickies, which were shown to become more roundshaped during kneading and easier to remove by subsequent screening, cleaning and flotation [10]. However, as stickies should be removed before, in the first deinking loop, the dispersing and bleaching steps appeared less relevant to the stickies issue.
Screening
Pressure screens are implemented after pulping (after the pulper screen or drum screening section) to remove the coarse contaminants. Screening is performed at high or medium consistency (2 to 4 %) with holes and/or with slots down to 0.20 mm. Low-consistency screening with typically 0.15 mm slots is then performed after flotation to complete the removal of stickies. The screening process, where the fibres have to pass the screen plate while the contaminants should be retained, has been extensively investigated through theoretical and experimental studies. The probability screening theory can basically be used for paper pulps as it applies to particles such as fibres and thin contaminants having at least one dimension smaller than the slot width [11-15]. The very complex flow conditions which govern the hydrodynamic particle separation phenomena were investigated experimentally and with the help of CFD simulation [16-21]. The effects of the screen operating and design parameters were investigated on pilot scale, extensively but mainly in the field of mechanical pulp for fractionation and for the removal of shives [22-27] as well as with flat-shaped model contaminants [28-30]. Stickies screening is more complex as soft stickies particles can be extruded through slots and more or less fragmented in pressure screens [31-32]. A large part of the research effort has therefore been devoted to the understanding and optimisation of stickies screening, and also because screening is the most effective technology to remove contaminants in the macro-stickies size range, as illustrated in figure 3.
Figure 3: Unit operation removal efficiencies versus particle size range [33]
Cleaning
High-density (heavy-weight) contaminants such as sand, glass and metal particles are removed with forward cleaners (hydrocyclones), generally in two steps, first at high-consistency to remove heavy particles and protect downstream equipment and then at low consistency (about 1%) to remove fine sand and protect fine slotted screen baskets from excessive wearing. Special low-density contaminant cleaners, including reverse, through-flow and rotary cleaners have been developed [34]. Stickies are known to be particularly difficult to remove because of their density close to neutral buoyancy. Comparative studies between different cleaners have shown that cleaners with high radial acceleration (small static cleaners) and with high residence time (rotary cleaner) are more effective [34-36].
7
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Flotation
The froth flotation process is the key technology to remove inks with limited fillers and fines losses. Hydrophobic inks are removed in a large particle size range as far a they are detached from the fibres, which requires normally a second post-flotation step implemented in the second deinking loop after hot-dispersing (figure 2). Macro-stickies are, on average compared to inks, less hydrophobic and also too large to be removed efficiently under conventional flotation conditions. The use of flotation to remove macro-stickies macro-stickies from screening reject streams at particularly low consistency has been reported in a deinking mill [37-38]. Micro-stickies should in principle be easier to remove according to figure 3. Considerable research work has been devoted to the physical-chemical aspects of the ink flotation process [39-44]. These aspects of stickies flotation were however not sufficiently understood regarding the stickies surface properties and the role of surfactants, which led to additional investigation on these aspects in the framework of this project.
Washing
The very small microscopic particles tend to follow the flow split in the pulp thickening and washing processes. Washing is a very effective deinking process to remove small inks, the equipment being specially designed to avoid the retention of such small particles by the fibres retained on the filtering element in order to increase their removal with the filtrate [4]. High wash-deinking efficiency is always associated with high solid losses as mineral fillers, pigments and micro-stickies are also removed with the ink. Washing is mainly used for tissue papers where almost complete deashing is required, as well as for the production of market DIP and LWC paper grades.
Process water treatments
Deinking process waters are normally clarified by dissolved air flotation (or micro-flotation) in order to remove the inks from the water circuits. The DAF process is however not selective as all suspended solids are removed together with the inks. Considerable research work has been devoted to the characterisation of the dissolved and colloidal materials in the process waters, their destabilisation leading to the formation of secondary stickies and the impacts on deposits and paper quality [45-48]. The presence of micro-stickies in deinking process waters has been reported and pressure filtration has been proposed to remove them selectively [49]. Pressure filtration is based on the same technique than pressure screening, except that the apertures (holes down to 0.1 mm diameter or ultra-fine slots) are so small that fibres are normally retained.
1.1.3. Stickies It is well known that, among the difficulties encountered encountered in the field of deinking to maintain or increase the paper quality, the development of various adhesive material to be found in the recovered papers is one of the most important problems, if not the most crucial in some mills. Such adhesives lead to numerous “stickies” problems including deposits on the paper machine, visual defects in the paper and problems in the printing machines due to residual sticky specks. Despite considerable progress in deinking technology technology the stickies problems are far f rom being solved: -
On one hand the amount of adhesive material in deinking furnish is currently growing even faster for various reasons such as the development of advertising inserts and product samples glued in magazines and newspapers.
-
On the other hand, the increasing use of deinking pulp in high quality graphic papers, a necessity to further increase the recycling rate up to 56 % set by the EC for 2005 (European Declaration on Paper Recovery), requires almost a complete removal of stickies to meet today’s quality standards of super calendered (SC) and light weight coated (LWC) papers.
Among the various adhesives and hot melt glues recovered with wood-containing wood-containing deinking furnish, newspapers (ONP) and magazines (OMG) from different sources including household collection, as well as with wood-free deinking furnish, mainly mixed office waste (MOW), the pressure sensitive adhesives (PSA) used for adhesive labels and tapes are of considerable concern.
8
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Indeed, due to their inherent physical properties, the solid adhesive particles, or “primary stickies” produced during the re-pulping of the recovered papers are very difficult to remove from the pulp even with the latest deinking technology. In addition some components of the adhesive material are soluble under conventional deinking conditions and contribute to increase the load of dissolved and colloidal material (DCM) in the process water, at the origin of the formation of “secondary stickies”. -
Adhesives are are soft and tend tend to be broken broken down into small particles particles during the pulping step. step.
-
Small particles are difficult to remove by pressure screening, screening, even with fine slots, since adhesives may be extruded through the slots.
-
Adhesives are also difficult to remove by centrifugal cleaning, since their density is generally generally very close to the density of the pulp.
-
Deinking flotation is not very effective in removing primary stickies since adhesives, especially water-based adhesive products, have normally no particular hydrophobic character.
-
Microscopic stickies and associated dissolved and colloidal material can be separated from the pulp by washing, and may be partly removed by process water treatments such as micro-flotation, but not selectively, i.e. at the expenses of increased rejects and chemical costs.
Figure 5 shows a rough outline of these processes prioritised according to application and classified according to pulp suspension and process water [50]. The outline also contains the currently conventional conventional classification of particle sizes and the associated definition of the three important stickies fractions into macro-stickies, micro-stickies and potential secondary stickies.
molecular 0,001
colloidal 0,01
finely dispersed
0,1
1 µm
10
coarsly dispersed 100
1000
fillers, fines, short fibers, long fibers salts, colours, lignin, toner, binder particles orig., resin, adhesive particles orig., fragments of binder- and adhesive films
n o i s n e p s u s k c o t s
screening cleaning deinking-flotation washing / thickening fixation particle-filtration iber recovery f iber
r e t a w s s e c o r p
mikroflotation (DAF) water cleaning
I fiber recov recovery ery
sedimentation water cleaning
I fiber recovery
membrane filtration rev. os osm. m. I nanofiltration I
ultrafiltration I
microfiltration
biological treatment, ozonisation, evaporation colloidal and dissolved material
(stock ock consistency) filtrable dispersed solids (st
mol./coll oidal subs substances tances (potential secondary st stickies) ickies) L. Hamann, P TS Heidenau
microstickies
macrostickies
creation of secondary stickies
Figure 5: Classification of stickies and separation processes in pulps and process waters [50]
9
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Indeed, there is no precise definition of the term "stickies" which is used to describe various materials and a lot of problems in the recycling and papermaking processes: Tacky particles (or particles able to develop tackiness in particular conditions) present in the paper and in deposits in various places (paper machine wires, drying cylinders, etc.), represent represent various forms of stickies. All these problematic problematic materials originate from various sources, mainly additives used during papermaking and converting. Nevertheless, adhesive products probably represent one of the main sources. According to the terminology terminology drawn up in a worksheet by the Zellcheming Committee of Experts on “Stickies/Recycling Criteria of Recovered Paper Utilisation” in 2003, stickies can be subdivided into macro-stickies, micro-stickies and potential secondary stickies, depending on their tendency to form deposits and on their screening behaviour under defined separation criteria. The dissolved and colloidal substances are termed a “potential to form secondary stickies”. -
Macro-stickies and micro-stickies are distinguished by their separation behaviour under standard testing conditions (such as the INGEDE method n°4), which is usually determined using laboratory screening with 0.10 mm slots in the case of deinking pulps. Stickies found in the screening residue are macro-stickies whereas stickies in the screening accepts are referred to as micro-stickies. Macro and micro-stickies are defined as filterable particles (suspended solids).
-
Primary stickies are introduced with the raw material and show an adhesive effect under standard testing conditions (e.g. INGEDE method n°4).
-
Secondary stickies are produced by physical-chemical processes during the recycling treatment and show an adhesive character under standard testing conditions.
-
Colloidal and dissolved substances are not referred to as micro-stickies. They are considered as potentially (secondary) sticky forming substances if they have or assume an adhesive character. At present there is as yet no sharp distinction in particle side between micro-stickies and potential potential secondary stickies. The particle size that can be separated using a filter sheet for pulp consistency determination, determination, usually about 5 µm, might in future become a possible alternative for definition [50].
It is now agreed that “recycling friendly adhesives” should, as far as possible, be designed in such a way to be removable as solid particles in the recycling process, and should not be soluble, since the potential “secondary stickies” problems are believed to be more difficult to manage.
1.2.
Objectives
The general objectives of the project was to develop new solutions to solve recycling problems and improve the quality of paper products from recovered papers. Practically, the following targets were set at the beginning of the project: -
To identify the most relevant adhesives materials, which have to be investigated to achieve the project objectives, i.e. widely used adhesives causing stickies problems in the field of deinking, and to develop new methods to improve the characterisation of the adhesive particles in the pulp.
-
To investigate the influence of all the pulping parameters on the size and shape distribution of the adhesives particles, and to develop new pulping conditions to promote their removal in the subsequent deinking steps, especially regarding their screening ability.
-
To study the mechanisms of pressure screening, with special emphasis placed on the behaviour of soft visco-elastic particles, in order to develop the understanding of the behaviour of stickies in screens, to optimise the screening conditions and to develop more efficient screening technology.
-
To investigate the centrifugal cleaning parameters, in order to optimise the stickies removal efficiency regarding the specific properties of pressure sensitive adhesives, and to evaluate the possibilities of new cleaning and rotation cleaning technology.
-
To evaluate the contribution of conventional deinking flotation, a process essentially used for the removal of inks, to the overall removal of stickies.
-
To investigate new ways for the selective removal of stickies from process waters, with special effort placed on the optimisation of pressure filtration.
10
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
-
To define the best strategies for the production of “stickies free” deinked pulp, and to establish guidelines for the implementation in current deinking lines.
-
To evaluate the impact of the rheological properties of the adhesives on their removal ability, in order to establish guidelines for the development of new recycling friendly adhesive products for the pulp and paper industry.
The project is focused on the optimisation of the removal of primary PSA stickies in the deinking lines, and more particularly on the macro-stickies as the objective is to remove as much as possible stickies before dispersion into microscopic particles or colloidal and dissolved components. The issues related to secondary stickies and more generally to colloids in process waters are out of the scope of this project, since extensive research, including the European project “Colloid control” [51], has already been devoted to this topic. Dissolved and colloidal stickies components were however considered for the relevant deinking process steps treated in this project.
2. Project development and partnership partnership 2.1.
Project structure
As the project is essentially devoted to the improvement improvement of the recycling techniques, which are the most effective in removing the pressure sensitive adhesives, a partnership has been elaborated to provide the basic scientific knowledge and the test facilities required for the study of these techniques. techniques. It includes three Paper Research Institutes (CTP, ICP and PTS) which were in charge of all the papermaking papermaking trials, one adhesive suppliers (Jackstädt, now Avery Dennison) for the supply of different types of adhesive materials, one screening equipment supplier (CAE, now AFT) for the design and the manufacture of screen plates, and one University (LEGI) for the numerical simulation of screening phenomena. A deinking mill was involved for the evaluation of a new pulping process on mill scale. The first part of the programme was devoted to the selection and characterisation of the adhesive material to be investigated and to the development of test methods. New analytical techniques to quantify the amount of adhesive in the pulp were further developed in the course of the project as well as new adhesive products to be tested (WP1). The second part is the most important part of the project since the key recycling technologies were investigated simultaneously by the different partners. This part was devoted to the optimisation of pulping to improve the size and shape of adhesive particles (WP2), to pressure screening (WP3) and centrifugal cleaning (WP4) to improve the removal of the adhesives and to deinking flotation (WP5) and pressure filtration (WP6) to further complete the stickies removal. Most of the research effort has been placed on the optimisation of pulping, screening and to a lesser extend cleaning, since the basic idea was to avoid as far as possible the fragmentation of the adhesives in the process. Screening is the most effective technology to remove impurities in the visible size range. Optimised pulping is a prerequisite to produce large adhesive particles removable by screening. Centrifugal cleaning under optimised conditions should be very effective for the removal of high-density or low-density adhesives. The efficiency of deinking flotation and the potential of pressure filtration were also investigated. Indeed, microscopic adhesive particles were expected to be produced, at least with some adhesives, even under optimised pulping conditions. Such microscopic particles can only be removed significantly from the pulp by flotation, as far as the adhesive material is sufficiently hydrophobic. Microscopic particles tend to accumulate in the process water, if no efficient means are provided to remove them. Pressure filtration was investigated regarding the possibility to remove stickies from the process water. In this second step of the project, the fundamental mechanical and hydrodynamic aspects, such as numerical flow simulation for the optimisation of screen plate design, were investigated first, to develop a better understanding of the specific behaviour of stickies with respect to the properties of the adhesive material, and further to develop new and improved technology.
11
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
In the third part of the project, all the results obtained in the previous steps were drawn together and analysed in order define optimised stickies removal strategies and guidelines for the development of new recycling friendly adhesive products (WP7). The general structure of the project and the interactions between the different work-packages as well as the related co-operations between the project partners are illustrated in figure 6.
WP1: Raw Materials and Methods CTP AFT - ADJ - ICP ICP - PTS 1.1: 1.1: Papers and adhesives 1.2: 1.2: Stickies control methods
WP2: Pulping ICP CTP
WP3: Screening CTP AFT - LEGI
WP4: Cleaning PTS CTP
WP5: Flotation PTS CTP
2.1: 2.1: Basic parameters
3.1: 3.1: CFD simulation
4.1: 4.1: Conv. cleaners
5.1: 5.1: Lab. flotation
2.2: 2.2: Drum pulper
3.2: 3.2: Screen plate design
4.2: 4.2: New cleaners
5.2: 5.2: Pilot flotation
2.3: 2.3: New technology
3.3: 3.3: Screening model
4.3: 4.3: Rotary cleaner
2.4: 2.4: Mill Trials
3.4: 3.4: High cons. screen WP6: Filtration PTS WP7: Stickies Removal Strategies CTP AFT - ADJ - ICP ICP - PTS
6.1: 6.1: Process waters 6.2: 6.2: Pilot tests
7.1: 7.1: Removal of stickies in deinking lines 7.2: 7.2: Recycling friendly adhesives
Figure 6: Project management management structure and partners involved in the different workpackages
The main contributions and responsibilities of the project partners were as follows: -
CTP CTP - Centre Technique du Papier (France) was responsible for the co-ordination of the project, for WP1 and WP7, which include most of the partners and for WP3 about pressure screening.
-
AFT AFT - Advanced Fiber Technologies, formerly CAE (Finland) was in charge of screening related work in WP1 and more particularly in WP3, as a subcontractor of CTP.
-
ADJ ADJ - Avery Dennison, Jackstädt GmbH (Germany) was in charge of adhesive product related work in WP1 and WP7, as a subcontractor of CTP, ICP and PTS.
-
ICP ICP - Instytut Celulozowo Papierniczy (Poland) was responsible for WP2 about pulping. A Polish deinking mill was involved in this workpackage, as a sub-contractor sub-contractor of ICP, for the mill trials.
-
PTS PTS - Papiertechnische Stiftung (Germany) was responsible for the workpackages WP4 about centrifugal cleaning, WP5 about flotation and WP6 about pressure filtration.
-
LEGI LEGI - Laboratoire des Ecoulements Géophysiques et Industriels (France) was in charge of the numerical simulation in WP3 about screening as CTP subcontractor.
The project structure in figure 6 will be followed in this document to report on materials and methods used in the project (WP1) and on the main project results (WP2 to WP7).
12
L SCREEN C CLEAN EAN
2.2.
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Development of the programme
As already mentioned mentioned the research research programme programme was focused focused on the removal of primary primary macro-stickies. macro-stickies. The basic idea was to optimise pulping and screening conditions in order to promote the production of macro-stickies which would be essentially removed by fine slot screening. Too small residual stickies particles to be removed by size with screens would then be removed by density with cleaner, by surface properties with flotation cells and finally by size with pressure filters for the stickies fraction to be found in thickening and washing process waters (see figures 3 to 5). The research effort initially planned was almost 50% on pulping and screening for the removal of macro-stickies, about 15% on cleaning and less than 15% on flotation and pressure filtration for the removal of micro-stickies. Other research projects carried out at CTP and PTS during this project showed that micro-stickies were responsible for the largest part of the deposits observed on paper machines [52, 53], which can be considered as the major stickies-related problem. In addition, it came out that the two reference adhesives were or became high or neutral density particles after pulping and that the work initially planned on the optimisation of low-density cleaning with the rotary cleaner at increased consistency was consequently no more relevant towards the final objectives of the project. These findings led to reduce the programme planned on cleaning and to increase the research effort planned on flotation, the most effective technology to remove micro-stickies. The initial research programme and the contributions of the project partners (figure 6) were essentially carried out as planned with half of the manpower input devoted to pulping and screening and with some relatively small changes decided during the project further to the development of knowledge. CTP transferred research effort from low-density cleaning to high-density cleaning, while PTS reduced the programme devoted to high-density cleaning and increased the research effort on flotation.
3. Material and methods 3.1.
Equipment and raw materials
3.1.1. Research means and pilot equipment Standard laboratory and various pilot equipment was used by the papermaking research institutes. Pilot equipment includes different types of pulpers at ICP and various pulping, screening, cleaning, flotation and pressure filtration equipment at CTP and PTS. Detailed descriptions of pilot equipment used in the different work-packages are reported in the corresponding sections. Figure 7 shows an example of such pilot equipment.
Figure 7 : Some pilot deinking equipment: ICP pulper, CTP pressure screen and PTS flotation cell
Research and production facilities were used at AFT and Jackstädt / Avery Dennison mainly for the supply of respectively test equipment (screen plates) and raw materials (adhesive labels). Advanced numerical simulation hardware and software were used at LEGI for the numerical simulations.
13
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
3.1.2. Paper raw materials
Background
Data about the average composition of deinking raw material in Europe has been collected in the framework of INGEDE projects [2]. Most of the recovered paper grades were ONP (Old NewsPapers) and OMG (Old MaGazines) generally mixed (household collection) or delivered separately, as well as MOW (Mixed Office Waste). The data collected from the participating mills gave: -
50 % ONP, 45 % OMG and less than 3% MOW for newsprint,
-
more OMG (about 60 %) for SC papers,
-
slightly less OMG and more MOW (about 5 %) for LWC papers
The main raw material for the production of wood-free market DIP is MOW.
Reference paper mixtures
These data and other sources were used to define typical European wood-free and wood-containing deinking raw material to be used (or simulated) by the partners in this project: -
Wood-containing furnish: a mixture of 50% newsprint (ONP) and 50 % magazines (OMG)
-
Wood-free furnish: mixed office waste (MOW)
Reference re-pulping conditions were based on the standard INGEDE deinking chemistry (table 1).
NaOH Sodium silicate Oleic acid H2O2
0.6 % (100%) 3 1.8 % (1.3-1.4 g/cm ) 0.8 % (extra pure) 0.7 % (100%)
Table 1: Deinking formulation (related to oven-dry paper) according to INGEDE method n°11
The option to use unprinted papers as model papers was selected in some cases as new stickies control methods based on the coloration of the adhesive were investigated, since ink from printed papers tends to cover the adhesive and thus to produce grey particles (see section 2). The use of unprinted papers also improved the evaluation of stickies in the case of handsheet image analysis methods, since the background is brighter and more even. Bleached chemical fibres (hardwood and softwood pulp mixtures) were also used for screening and cleaning tests, as far as the most relevant deinking raw material characteristics, such as fibre composition and ash content, were respected.
3.1.3. Adhesive raw materials
Background
Adhesives enter the paper recycling chain with deinking raw materials in the form of self adhesives labels, tapes and envelopes, as well as glues for advertiser samples in magazines and book bindings. Water-based adhesives play a major role as they account for about 80 % of the total PSA market for label applications in Europe, which represent a total consumption of 551,000 t pressure sensitive laminates, including [54]: -
267,000 t of face material (label paper)
average: 80 g/m²
-
67,000 t of adhesive
average: 20 g/m²
-
217,000 t of backing paper (75% glassine type)
average: 65 g/m²
14
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
These pressure sensitive adhesives are usually water-based acrylates modified with tackifier resins, the typical formulation of a water-based adhesive being as follows [54]: -
Acrylic polymer
(2-ethylhexylacrylate (2-ethylhexylacrylate or butylacrylate butylacrylate copolymers) copolymers)
60 to 100 %
-
Tackifier resin
(rosin acid, rosin ester, hydrocarbon hydrocarbon base resins)
0 to 40 %
Additives
(Ammonia, (Ammonia, defoamers, wetting wetting agents, agents, fungistatica)
1 to 5 %
-
Others categories of pressure sensitive adhesives include, besides water-based acrylic adhesives: -
Water-based rubber adhesives including Styrene Butadiene Rubber (SBR) and Styrene Isoprene Styrene (SIS) types, the later being the most common
-
Hot-melt based rubber adhesives
Among these rubber adhesives, hot-melts show a growing market and are becoming more common than solvent based PSA [55]. Further more detailed information on adhesives is reported in section 4.6.2.
Reference adhesives
For various reasons, including the time needed to come to a clear conclusion about the distribution of adhesive types in European deinking raw material (a prerequisite to define any standard adhesive mixture) and the difficulty to distinguish different types of adhesives mixed in the same pulp sample, it was decided to focus on the most common adhesives. Two reference adhesives were selected: -
Water-based Water-based acrylic adhesives:
ADJ Reference E 115
-
Hot-melt based rubber adhesives: ADJ Reference D 170
The compositions of the Avery Dennison Jackstädt (ADJ) reference adhesives are given in table 2. The E 115 adhesive is coated as dispersion in water and the D170 adhesive is coated as hot-melt.
Adhesive Component E 115
D 170
Product description
function
1
Copolyacrylate Copolyacry late
Backbone polymer
2
Terpenphenolic resin
Tackifier Tackifi er
3
Wetting agent
Wetting of adhesive when coating
4
Antifouling Antifouli ng agent
Avoids fungus formation
1
Styrene Isoprene block copolymer
Backbone polymer
2
Styrene Butadiene block copolymer
Backbone polymer
3
Hydrocarbon Hydrocarbon tackifying resins
Tackifiers
4
Polyethylene Polyethyl ene powder
Avoids migration
5
Antioxidant
Avoids thermal and UV degradation
Table 2: Composition of the E 115 and D170 reference adhesives
15
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Adhesive samples
The reference adhesives were supplied by ADJ to CTP, ICP and PTS in the following forms: 2
-
Standard labels with 20 g/m adhesive (in the form of rolls delivered on request) to be used as raw material in the deinking process optimisation tests (WP2 to WP5). Special samples with the adhesive films between 2 silicone release papers, were also delivered in order to produce adhesive particles for the stickies extrusion tests reported in section 2.4.
-
Adhesive material in the form of liquid dispersion (about 50 % dry content) for the E 115 115 adhesive and in the form of solid samples for the D170 adhesive, to be used for manufacturing the adhesive test samples required for the different rheological tests.
3.2.
Stickies control methods
The aim was to define common methods and to develop new methods for the efficiency assessments of the different deinking process steps in order to generate consistent input data for the optimisation of the stickies removal strategy and to compare the results obtained by the project partners. A clear need for improved and standardised stickies control methods also exists at European industry level. One important point considered in the development of improved and standardised stickies control methods was that the project needs were somewhat different from the objectives in European standardisation: -
In this project project, the pulp samples were generally over-contaminated with adhesives, and efforts in terms of control costs could be placed on complete and accurate characterisation of the stickies properties, especially particle size and shape.
-
Standard stickies control control methods methods are primarily intended for use in deinking deinking mills, which refers refers to low adhesive contents in the pulp and implies easy and low time consuming control procedures.
3.2.1. Laboratory screening methods (macro-stickies) (macro-stickies)
Background
The most common primary stickies control methods are based on the use of laboratory laboratory screening with fine slots to remove stickies and contaminants from the fibre suspension in order to be able to analyse the sticky particles with to various methods. -
“Macro-stickies” “Macro-stic kies” generally refer to particles retained on the laboratory screen plate
-
“Micro-stickies” “Micro-stickies” generally generally refer to particles particles passing passing the the slots of the laboratory screen plate
These definitions are widely accepted though their are somewhat confusing since macro-particles normally refer to particles in the visible size range while micro-particles refers to particle sizes under the visibility limit of the human eye, i.e. about 100 µm and less for high contrasted particles. With lab screens using slots widths around this size limit, it is clear that large flat-shaped particles in the visible size range cannot be completely removed if particle thickness is significantly lower than the slot width. Some preliminary tests, clearly confirmed that the effectiveness of the laboratory screening methods in extracting the “macro-stickies” from the pulp strongly depends on particle shape and that a large amount of adhesives in the visible size range passed the slots. Consequently, it was clear that lab screening accepts should also be controlled to analyse both visible and microscopic “micro-stickies”. Different lab screening conditions are used. The Haindl Fractionator, the Somerville screen and the Pulmac Masterscreen are the most common, though the Brecht-Holl and Weverk screens are also used. -
CTP used the Somerville screen equipped with 0.10 mm slots (0.08 mm slots before the project) for deinking pulps. The Weverk and Brecht-Holl screens were also available.
-
PTS used used the Pulmac Masterscreen Masterscreen instead of previously previously the the Haindl Haindl Fractionator, Fractionator, both devices devices being equipped with 0.10 mm slots for deinked pulps.
-
ICP used the Brecht-Holl Brecht-Hol l apparatus equipped with 0.10 mm slots (0.15 mm before the project).
16
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Standard methods
The INGEDE method n°4 for the analysis of macro-stickies [56], which recommends 0.10 mm slots for deinking pulps (or 0.15mm slots for long fibre recycled pulps), is the most commonly used in Europe. A modified method which allows allows to use 0.10 mm for all types of recycled pulps after enzyme digestion digestion has also been proposed [57]. Consequently, Consequently, it was decided to use 0.10 mm slots in this project for the macro-stickies, macro-stickies, keeping in mind that the micro-stickies should also be analysed. Concerning the equipment is was decided to use the Somerville screen, the Haindl Fractionator or the Brecht-Holl apparatus, and not the automatic Pulmac Masterscreen, which worked on a different way, i.e. with a rotor instead of a membrane to generate the pressure pulses. A prerequisite to obtain comparable comparable adhesive control results was first to agree on the slot width, i.e. 0.10 mm, and then to use the same detailed slot design. Laboratory screen plates were supplied by AFT to CTP, ICP and PTS (deliverable D2) to ensure slot accuracy and uniformity. The slots were milled. The internal control data at AFT gave the following slot characteristics: characteristics: -
Nominal slot width:
0.10 mm
measured slots: min. 0.10 mm, max. 0.11 mm
-
Nominal slot length:
45 mm
measured length: min. 44.6 mm, max. 44.8 mm
3.2.2. Stickies size and shape analysis
Standard methods
Among the key parameters which govern the efficiency of the stickies removal process steps, particle shape is almost as important as particle size. Macro-stickies from self-adhesive labels are initially flatshaped (adhesive film thickness of typically 20 µm) and can become long-shaped or round-shaped, depending on pulping and hot dispersing conditions. With usual laboratory control methods methods of macro-stickies, such as INGEDE method n°4, stickies size is measured after heating and pressing, pressing, which increases the apparent size of the particles. If stickies are analysed in handsheets the apparent size is also increased but to a lesser extent because of the lower temperature and pressure (handsheet formers using vacuum drying). These usual stickies control methods do not give any information about the thickness of the particles. With on-line measuring systems, such as the Simpatic on-line speck analyser (Techpap), the projected size of contrasted particles is measured directly in the pulp flow. Again, there is no information on the shape of the stickies shape especially with flat-shaped particles, since the measured size depends strongly on the orientation of the particle in the pulp flow.
Development of direct measuring methods
With usual laboratory stickies control methods, the size distribution of the adhesive particles after heating and pressing is measured by image analysis. Most of the image analysers include various functions, which can be used to evaluate the shape of objects (spread out stickies) in two dimensions. - 2D particle shape analysis: ICP developed a new 2D image analysis methods to evaluate the shape of unaltered adhesive particles in two dimensions. In this macro-stickies control method the adhesives are removed from the pulp by lab screening and analysed directly (without heating and pressing). The method also allowed to distinguish the two reference adhesives through colouration. To obtain coloured stickies in the pulp the adhesives labels were first stuck onto the coloured surface of paper which had been printed with one-colour toner in a xerox copying machine. The coloured macro-stickies recovered by lab screening were then picked up with a preparation needle and gently placed on a transparent adhesive tape in a preparation box (figure 8) for further analysis. The different stickies types were discriminated by colour image analysis. The system also allowed to determine particle particle size and shape f actor and to distinguish the black ink particles. Figure 8 shows some coloured long-shaped stickies and toner inks obtained according to this method.
17
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Figure 8: Method for the particle shape analysis and distinction by colour of two model stickies
This method gives a good evaluation of particle shape in the case of round or long-shaped particles, but does not allow to evaluate the thickness of the particles in the case of flat-shaped adhesives. - 3D particle shape analysis: The analysis of the 3 dimensions of a particle is in principle possible if the particle is observed under different angles. Image analysis methods based on the use of several video cameras could be developed to evaluate the shape of particles viewed in a pulp flow. The feasibility of such 3D on-line analysis of particle size and shape has been evaluated at CTP. Technical solutions were defined but showed to be too expensive to develop.
Development of indirect measuring methods
A stickies size and shape analysis method has been evaluated at CTP, based on the assumption that laboratory screens have a higher potential in separating particles according to their smallest dimension (particle thickness) than pressure screen, though the largest dimensions have also a significant effect (higher lab screening efficiency with large adhesive films compared to small films at given thickness). The size distribution of the adhesive particles in the accepts of laboratory screens gives an indication about the shape of the adhesive particles. Large particles passing the slots correspond to flat-shaped adhesives (with lower thickness than the s lot width). A small proportion of small particles in the screen accepts corresponds to round-shaped particle (with higher thickness or diameter than the slot width). The 3-stage laboratory screening method illustrated in figure 9 includes two steps. Step 1: Pulp samples are first treated with a 2-stage laboratory screening arrangement, the first stage being equipped with 0.15 mm slots and the second one with the standard 0.10 mm slots. The accepts of the second stage are collected on a 200 mesh wire screen. The fraction passing the wire screen should not contain micro-stickies in the visible size range. Indeed, the retention of small adhesive particles is much higher with a thickening device equipped with a wire screen with 75 µm wire opening, compared to a laboratory screen equipped with slots with 50 µm slot width. In this first step the fractions retained on the screen plates (R 0.15 and R 0.10) are analysed for the evaluation of stickies size and shape. The “water fraction” passing the wire screen can also be analysed to control small micro-stickies and evaluate the amount of these small micro-stickies, which should be found in mill process water circuits. Step 2: The pulp fraction passing the 0.10 mm slots is then submitted to enzyme treatment after thickening on the 200 mesh screen. The enzymatic stickies control method [57] has been developed at CTP in order to facilitate the use of finer slots (0.08 mm slots instead of usually 0.10 or 0.15 mm slots). The method enables the use of even finer slots, i.e. 0.05 mm slots. Enzyme digestion degrades the fibres and consequently reduces the amount of fibres retained on the screen. It improves particle count with the INGEDE method n°4 since residual fibres tend to cover stickies which should normally be counted. Both fractions retained and passing the slots (R 0.05 and P 0.05) are analysed.
18
L SCREEN C CLEAN EAN
0.15 mm
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
R 0.15
Step 1: 2 laboratory screens in cascade
0.10 mm
Step 2: laboratory screening after enzyme digestion
Enzyme digestion
R 0.10
200 mesh (75 µm) P 200
0.05 mm
R 0.05 P 0.05
Figure 9 : Three-stage laboratory screening method for indirect stickies size and shape assessment
In this indirect method for the evaluation of the size and shape of the stickies, the size analyses of the fractions retained on the screen plate (R 0.15, R 0.10 and R 0.05) can be performed by image analysis either on standard handsheets or with the INGEDE method. The last fraction passing the slots (P 0.05) has to be analysed on handsheets. The evaluation of the shape of the particles requires to establish, for each of the 3 fractions retained on the screen plates, correlation between particle size distributions measured directly on the particles (ICP method in figure 8) and after pressing and drying (INGEDE method n°4 or handsheet analysis). This methods however is very time consuming.
3.2.3. Extraction methods (micro-stickies) (micro-stickies) Micro-stickies in the form of finely dispersed adhesive particles and potential stickies in the form of colloidal and molecular mater contribute strongly to the stickies problems observed in mills [52, 53] Extraction methods are currently the only possibilities to control these categories of stickies as they are not separable by laboratory screening, even with very fine slots.
Conventional DCM Extraction Method
The conventional extraction method using dichloromethane (DCM) as the solvent is used by several research institutes to extract and quantify characteristic component of adhesive material. This method has gained widespread acceptance to evaluate micro-stickies and colloidal stickies in pulp samples, though the method is not selective, since other materials such as coating binders are extracted and quantified together with PSA material. CTP used and developed this method for targeted stickies analyses performed in this project. The main problem of this method at the present time is its accuracy. Indeed, the rate of extract is quantified by weight after total solvent evaporation. Unfortunately, the weight measured may be sometimes very low (some mg) which makes it difficult to get accurate result.
Development of a new DMF extraction method
PTS has developed a more targeted extraction method using DMF (dimethylformamide) which permits complete, selective measurements measurements of sticky loads by means of solid-liquid extraction. The description of the protocol and advantages of this new method have been reported (in the Annex 6 of report D1). The recognition rate of coating binders achievable with DMF extraction showed to be nearly 50 %, which is enough to clearly identify the individual binder loads of different pulps. The reproducibility of the extraction results was therefore regarded as excellent. The recognition rates of further potential stickies such as adhesives were also investigated to make sure that all stickies types present were detected by the measurements.
19
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The comparative results of the DCM and DMF extraction are shown in figure 10 in the case of binders. The binder content of the paper samples increased towards the sample comprised of 100 % offset papers. The DMF extraction results reflected the trend towards increasing binder contents, with a sufficient degree of detail or differentiation, while the results of the DCM extraction were significantly less satisfying. The absolute extract contents determined by DMF extraction were on average about twice as high as those obtained by DCM extraction, i.e. the binder detection by DMF extraction was significantly more complete.
100 % coated offset papers
growing content of coating binders
50% newsprint 30% coated gravure 20% coated offset
100 % coated gravure papers
Extraction in turbulent flow - DMF
Conventinal Soxhlet Extraction - DCM
100 % newsprint
0 ,0
0 ,5
1,0
1 ,5
2,0
2,5
3 ,0
3,5
Extract yield [ % ]
Figure 10: DCM/DMF extraction in comparison for the evaluation of coating binders
3.3.
Adhesive rheological properties
The analysis of the forces applied on stickies during pulping and screening indicated that adhesive particles should be submitted to high-speed deformation, namely during contacts with pulper or screen rotor elements, as rotor velocities are generally in the range of 10 to 20 m/s. In the case of screening, the analysis of possible single step extrusion of adhesives through slots showed that both strain and strain rate would be very high with particles much larger than the slots, if the particle effectively passes the slot during the screening phase. With integral calculation of the average strain at high deformation (ε = ln d0/d = ln d 0/w) and for a particle squeezed by a factor 3 in 10 ms, the strain rate ( ∂ε/∂t) would be 2 -1 of the order of 10 s . Results about the dynamic moduli, G’ and G”, of adhesive products have been reported but the measurements were performed in dry state at much lower strain rate [58]. The analysis of the dynamic behaviour of adhesives was necessary to provide input data for the numerical simulation simulation work performed at LEGI about stickies screening. The work was subcontracted to LEMTA (Laboratoire d’Energétique et de Mécanique Théorique et Apliquée, INPL/CNRS, Nancy, France) where expertise was available in this field and for the analysis of friction factors of soft materials [59]. The equipment at LEMTA had however to be adapted to the very soft adhesive materials to be tested. Preliminary low-speed elongation tests were done at CTP to provide some first simulation input data.
3.3.1. Low-speed elongation tests Preliminary tests were performed with paper testing equipment at elongation velocities up to 1 m/min, -1 which gave a maximum strain rate of 0.4 s in the narrow part of the adhesive test samples (40 mm length, 15 mm width and 2 mm thickness). The dimensions of the test samples were not exactly those defined for standard rubber testing procedure because of the difficulties encountered in preparing samples with very tacky material (figure 11).
20
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Adhesive sheets were produced from the two reference adhesive raw materials, cut in water and tested after soaking. Both adhesives recovered almost their initial length in a few minutes after elongation by a factor 10. Figure 11 shows pictures of an adhesive sample back to its initial position after a first elongation and submitted to a second elongation.
St an an da dar d r ub ub be ber te tes t s am am pl pl e
A dh dh es es iv iv e s am am pl pl e an d t en en si si le le t es es t (2
nd
elongation)
NF T 46-006 Type H1 test sample shape & dimensions (narrow portion) 33 mm length 6 mm width 2 mm thickness
Figure 11 : Low-speed adhesive elongation tests
The typical stress strain relations are shown in figure 12. Both adhesives showed a first elastic part followed by a drop of the modulus with lower strength at the second elongation, especially with the hot-melt rubber product. The first pass elastic modulus was about one order of magnitude higher with the hot-melt rubber than with the acrylic adhesive, which gave G ≈ 20 kPa at 1m/min after soaking in hot water. The elastic modulus showed to increase with increasing elongation velocity and to decrease with increasing temperature.
Water-based acrylic adhesive
F
E 115
Hot-melt rubber adhesive D 170 F
1st
1st 2nd
2nd
L-L0 / L0 0
1
5
L-L0 / L0
10
0
1
5
10
Figure 12: Typical low-speed elongation stress strain curves
3.3.2. High-speed compression tests The experimental equipment and method illustrated in figure 13 were developed to record stress strain relations applied on adhesive materials at high strain rates: -
a cylindrical cylindric al adhesive sample (18 mm diameter and 28 mm mm height) is placed on a bottom plate equipped with a load sensor,
-
a weight is dropped from different heights on the top plate above the adhesive sample,
-
the compression of the adhesive sample is measured with a laser recording the height of the bottom plate.
21
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Mass
V ma x = (2gh) 1/2
h
Sample
Lubricant
Laser
F(t)
δ (t)
Figure 13: High-speed adhesive compression tests
The initial strain rate are controlled by the drop height (0.15 to 0.9 m, e.g. 0.8 m gives a drop velocity -1 of 4 m/s and an initial strain rate of 150 s ) and the stress level by the weight. The samples are covered with talcum to allow adhesive material to slip between the plates and avoid barrelling of the cylinder during the compression. The recorded data (figure 14 left) are converted into stresses of Cauchy and true axial strains. 1
3
s s e r t S
0,8 . ) V ( t u0,6 p t u o r o s 0,4 n e S
Experimental Model 2
Load on bottom plate 1
Distance Distanc e to top plate (laser)
0,2
0
0 0
0, 05
0,1
0,15 Time (s)
0
0,1
0,2
0,3
0,4
0,5
Strain
Figure 14: Dynamic test: recorded load and compression and experimental results versus model
The dynamic compression test is distinguished from the usual quasi-static tensile test by the fact that the strain rate drastically decreases during compression, towards zero at the end of the tests. The residual stress due to the weight at the end of the test is considerably lower than the one due to the high deceleration of the weight at the impact. The strain level reached with adhesive material can be relatively significant, up to 1. Figure 14 (right) shows a typical example of stress strain curve obtained with the dynamic compression test, at non-constant strain rate. In principle, a large number of tests should be performed to determine, as a function of the strain rate, the stress strain relations, which are required for the simulation of particle extrusion. The development of a model describing the dynamic behaviour of viscoelastic material will enable to determine these relations from the experimental data. The theoretical curve plotted in figure 14 represents the estimated result obtained from a viscoelastic model with two relaxation times. Different relaxations times generally refer to the components of the generalised Maxwell model where infinite viscosity of one component gives the solid behaviour. The new theoretical model under development is based on the thermodynamics of irreversible processes produced by the stresses and strains applied to the polymers [60-61]. Indeed, polymer chains exhibit the highest disorder, i.e. maximum entropy, at steady relaxed state, while stresses tend to stabilise. The complex behaviour of adhesive material depends on the relaxation times spectrum which is related to the micro-structural reorganisations observed at different scales. Further analyses of adhesive rheological properties are in progress to determine more completely the dynamic behaviour of the adhesive material at high strain rates.
22
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4. Results and discussions 4.1.
Pulping
4.1.1. Background and objectives Pulping is the first step of the deinking process. Defibering, ink and adhesive detachment, and particle comminution in the pulper are achieved by forces imposed on the recovered paper through the action of the rotor. And therefore the pulping step strongly affects the quality of the pulp to be treated in the subsequent steps of the deinking process, and as such it is decisive for good deinking. Most of the research work about pulping has been devoted to the optimisation of the recovered paper re-pulping and deinking processes, namely in terms of deflaking and ink detachment kinetics [5-8]. However, and despite of its considerable role, little work has been done to characterise the pulping phenomena with respect to stickies and other contaminants besides inks. The contaminant size and shape distributions are also governed by the pulping conditions, as understood in terms of mechanical stresses and physical chemistry. The latter is determined by the deinking bath, with a possible admixtures, as well as by pulping temperature. The development of new optimised pulping conditions aimed at creating the quite large particles of the contaminants is a prerequisite for their efficient removal in the subsequent process step, especially fine slot screening. The objectives in the research programme devoted to pulping in this project were more particularly: 1. to study the influence of of basic pulping pulping parameters parameters on both stickies and inks, inks, 2.
to evaluate state of the art pulping technology, i.e. batch and drum pulpers,
3. to develop a new pulping technology, technology, which is based on the agglomeration agglomeration of stickies, and 4. to test, after the lab pilot pilot scale optimisation, optimisation, the the new pulping pulping technology technology on mill scale. Except for the second task, which was performed at CTP where relevant pilot pulping equipment was available, all the pulping studies were done at ICP. The scientific approach towards the development of a new pulping process was based on the experience gained by ICP with the agglomeration of printing ink particles, which was acquired in the time when newspapers were printed by letterpress with inks being dispersion of carbon black in dyed mineral oil, and drying process of the newsprint ink was mainly by absorption [62]. By setting proper conditions (hydrodynamic and physico-chemical) during pulping of newspapers it was possible to achieve agglomeration of such inks. Spherical-shaped particle were obtained after 20-45 minutes pulping and were large enough to be removed by cleaners. Further investigations investigations also allowed to agglomerate agglomerate successfully tiny particles of xerographic toner inks.
Figure 15 : Agglomerates removed from recovered newspaper re-pulped with deinking chemicals
23
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.1.2. Study of basic pulping parameters Deinking of pulp fibres is essentially a laundering or cleaning process in which the ink is considered considered to be the dirt. To dislodge the ink particles from the fibres, chemicals are used in the pulping step, along with heat and mechanical energy. The detached ink particles are then removed from the stock in subsequent steps of the deinking process. Both deinking lines and deinking technologies are designed to remove the ink particles, and therefore they are usually not suitable for efficient separation from the stock such specific contaminants as the sticky particles (stickies) are. The latter are created during the pulping step by comminution of various adhesive materials to be found in recovered papers, among them the pressure sensitive adhesives (PSA) used for adhesive labels and tapes, also as adhesive layers applied to attach samples to paper, are of considerable concern. The printing ink, which usually is fixed on paper in the form of small solid dots, consists of pigments and dyes, dispersed in vehicles or binders, to which some other ingredients are added. During the pulping step these dots behave like brittle solids and they are divided into smaller elements which are close in their size to the size of the pigment particles used in given ink. However, the layers of PSA materials applied on paper are such a continuous film of elastomers in which there is a lack of such frontiers as perceived in printing ink dots (between pigment particles and binders), so the comminution of the PSA layers is more chaotic in its character. Moreover, ink particles are usually more hydrophobic than sticky particles, and therefore the ink particles may be important constituent of the co-agglomerates with stickies, under favourable conditions of the pulping process. process. The ICP studies carried out in a frame of this project, were aimed at applying the forces imposed on recovered paper in the pulper not only for defibering, ink and adhesive detachment, and particle comminution, but also for agglomeration of adhesive impurities and for their co-agglomeration with ink (and other) particles. In such an approach the typical activities of the pulper are perceived as the first step which is necessarily required for successful accomplishment of such agglomeration and co-agglomeration processes in which ink particles, and perhaps some other particles too, may be useful components for such enlarging of the co-agglomerates, which is needed for their successful removal in the subsequent steps of the deinking process, especially in fine slot screening. The studies, in the work package devoted to the pulping technology (WP 2: Pulping), were aimed at combining together two different processes, namely: de-inking and de-sticking, therefore. According to the results previously gained by ICP, the agglomeration agglomeration of printing ink particles (and toner particles too) is promoted by adding to the pulping step such properly selected substances which are able to combine the particles together; however, such agglomeration is possible under the laminar regime flow of slurry in the pulper. A very distinctive feature of such flow is the continuity of external layer of the slurry in the pulper, and the laminar regime of flow occurs at different consistencies, depending on construction of the pulper and its rotor as well.
Studies on behaviour of different PSA labels during pulping
The mechanical strength of the self-adhesive layer is decisive for its susceptibility to comminution; however, in technology of the pressure sensitive adhesives (PSA) two different strengths are recognized: adhesive strength and cohesive strength. Experiments performed by ICP, clearly showed such a different behaviour of the same adhesive layer applied on different base materials (figure 16).
Figure 16: The PSA labels on plastic foils after pulping (45 minutes) with the model recovered paper
(the rectangles are equal to an initial surface of the label)
24
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Almost all studied labels on plastic foils (only the Acetat foil was an exception) survived the pulping of recovered paper without their dividing into smaller parts, and with an inconsiderable comminution of the adhesive layer. Therefore, the experiments proved that the forces applied to the self-adhesive layer during the pulping step were weaker than the adhesive strength between the adhesive layer and the plastic foil. However, the labels with a pressure sensitive adhesive applied on paper base are such a completely different case in which the paper base is disappearing during the pulping step, and this fact is quickly followed by comminution of the adhesive layer. In such a special case, which is of utmost significance for the industrial practice, the cohesive strength of the adhesive layer is decisive for its susceptibility to comminution. Size of the small granules, being created from the adhesive layer during flow of slurry in the pulper, is determined by the velocity gradient between the slurry layers in which ends of the adhesive layer are anchored, as well as by an initial length of the adhesive layer. Such a specific course of the phenomenon should not be expected during the plug flow of slurry in the pulper, therefore. In pulping experiments with non-printed wooden paper and with an admixture of the paper labels coated with the PSA layer, amounted to about 300 mg per kg of the oven dried pulp, the reduction of stickies’ size progressed during the pulping run under the laminar regime of flow, and with deinking chemistry according to the INGEDE method n°11. The acrylic adhesives were more susceptible to comminution than the hot-melt rubber adhesives. Particles of the latter, however, became spherical in shape during such pulping, so they could be easily separated on slotted screens. Nevertheless, under such conditions there was a lack of agglomeration of the sticky particles. Their quite hydrophilic character was an obstacle in the way of the agglomeration process. To that end an admixture of the agglomerant agglomerant is needed, and the presence of such contaminants contaminants in the slurry as ink or toner particles would be very helpful.
Effect of pulping intensity and addition of the agglomerants
In this project two PSA paper labels, both of the Avery Dennison Jackstädt (ADJ), were selected as the reference adhesives, namely: the water-based acrylic adhesive (denoted as E 115), and the hotmelt rubber adhesive (denoted as D 170). Before pulping experiments, the rubber adhesive (D 170) was stained yellow and and the acrylic adhesive (E 115) was stained cyan. After pulping experiments, the particles rejected on the slotted screen (with 0.10 mm slots) were collected on white filter paper. Together with the paper, and still in wet state, t hey were placed between two sheets of the transparent polyethylene foil. Such kind of sandwich was introduced into the scanner. In that procedure a detachment detachment of the particles from surface of filter paper is avoided, and the scanner glass is kept clean. The particles of both adhesives, as well as ink or toner particles, were finally identified during the computer-aided image analysis (using the Spec*Scan System, by Apogee System Inc. ), according to the grey scale value range ( GSV ) established for each kind of particles. Therefore the D 170 particles were detected within the GSV range of 220-240, and the E 115 particles were detected within the GSV range of 135-190. The black particles of ink or toner were detected for the GSV below 100. Such a method, which we will refer to as the ICP method, gives an insight into the size and shape of different particles, which are separately perceived in image analysis. Their shape was additionally characterised as an eccentricity (Ecc), (Ecc), according to the Spec*Scan System, by Apogee System Inc.; for perfectly round particle its Ecc=1, and the more oblong the particle is the higher its Ecc. Possibilities of such successful co-agglomeration co-agglomeration of the stickies with toner particles were shown in the pulping experiments carried out with wood-free xerographic paper (denoted as OWP) had been nonimpact printed on its entire surface (one side only) with black toner. Among the studied agglomerants, added in quantities from 1.2% to 1.5% (in relation to oven dried paper), 1-octadecanol (denoted OD) was quite efficient in promoting such co-agglomeration during the pulping run under the laminar o regime of flow; with deinking chemistry according to the INGEDE method n 11 (without peroxide, however). The ICP method made possible the visualisation of the agglomeration and coagglomeration processes during such pulping. An example is presented in figure 17, showing agglomerates agglomerates of the sticky particles and their co-agglomerates co-agglomerates with toner particles.
25
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Figure 17: Alterations in size and shape of the stained stickies and their co-agglomerates co-agglomerates with toner particles during the pulping step of recovered paper; macro-photograph of the particles rejected on the slotted screen (with 0.1 mm slots) as deposited on white filter paper: particles of the D 170 adhesive (yellow), particles of the E 115 adhesive (cyan), toner particles (black), and their co-agglomerates
26
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
From figure 17 it is clearly visible that the comminution of toner and adhesive particles is followed by their agglomeration and co-agglomeration, during the pulping step run under the laminar regime of flow and with an admixture of the agglomerant. Due to such co-agglomeration of adhesive particles with toner particles, the co-agglomerates were created with a prevailing effect of the black colour, and therefore with grey scale values below 100; it means they were classified as the toner agglomerates, in the computer aided image analysis, besides particles and agglomerates detected within the GSV ranges characteristic of both coloured adhesives. The area of toner agglomerates, it means their content, was much larger than contents of the sticky particles from both studied PSA materials. It is the proof that such successful co-agglomeration of toner particles with properly disintegrated debris of the PSA layer is possible during pulping run under the laminar regime of flow, and with an admixture of the agglomerant, i.e. such hydrophobic surface active agent which is able to combine together particles with different surface activities. All of the studied contaminants contaminants had practically the same mode value of their eccentricity coefficient (Ecc). It seems to be a proof that under conditions of the pulping step favouring agglomeration and co-agglomeration processes there is a strong tendency to create more homogenous particles of the contaminants. More intensive pulping, run at higher rotation speeds of the rotor as well as at higher pulp consistencies (to keep the laminar flow of the slurry), resulted in a violent growth in amount of the co-agglomerates created by toner and sticky particles. It may be understood, that the co-agglomeration process is promoted both by proper comminution of the particles and by better dispersion of the agglomerant, during such laminar pulping of recovered paper non-impact printed with toner, in the presence of the PSA materials. There is a need to adjust properly temperature during pulping to the melted point of the agglomerant.
An influence of the energy parameters and their joint action with physico-chemical parameters upon the alterations in size and shape of the particles
At the beginning of the ICP investigations, carried carried out in a frame of the WP 2 Pulping, it was assumed 2 that thorough knowledge of the average size A av [mm ] of the macro-stickies and their shape expressed as an eccentricity Ecc would be sufficient to characterise their properties. The above discussed experiments showed, however, that such values as the amount of macro-stickies S A 2 o [mm /kg] and their number S N [n /kg] used alone were not fully proper in investigating the phenomena connected with creating stickies and their common agglomeration and/or their co-agglomeration with other particles, while pulping recovered paper. In order to accurately describe the phenomena governing the creation of the stickies and their agglomerates and/or co-agglomerates, the knowledge not only of the average values of such properties but also of their distributions is required. Nevertheless, even in such pilot plant experiments there is no possible to analyse the entire population population of stickies, owing to the uncontrolled participation of sticky particles of the PSA materials in creating the deposits on pulper surfaces, see figure 18. Moreover, the statistical analysis is not used for the entire population of the stickies created during experiments, experiments, but only for a part of it, collected in 2 the screen (with 0.10 mm slots), i.e. without the stickies with a surface size smaller than 0.014 mm , which corresponds to the value of the lower threshold of size, as established for the particles counted in the computer-aided image analysis.
Figure 18 : Deposits of organic matters on pulper surfaces
27
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
2
In attempts to describe the size A [mm ] distribution of the macro-stickies, the parameters of the following distributions were taken into consideration, namely: normal distribution, lognormal distribution, and dislocated lognormal distribution. It would be much more correct to refer to the dislocated lognormal distribution to accurately describe the distribution of the macro-stickies’ properties. Additionally, a general evaluation of the comminution process of the adhesive layer, which mostly determines the size and shape of the stickies, also appeals to such a choice. The dislocated lognormal lognormal distribution has the following probability density function: 1/2
2
2
f x(x) = 1 / [(x-θ)σ(2π) ] * exp [-{ln(x- θ)-µ} /2σ ]; θ <
x < ∞, σ > 0
where: µ
σ θ
is the scale parameter is the shape parameter is the threshold (location) (location) parameter
Keeping in mind that in our experimental data the populations of stickies are seriously reduced in uncontrolled (deposits) and controlled (screening) ways, the attempts were made to fit a proper model of the distribution (with the Statistica program by StatSoft ) for the determined sizes of the macrostickies. Using graphical comparisons of the theoretical and empirical forms of the distribution functions, it was concluded that the model of the dislocated lognormal distribution, with the location 2 parameter value equal to the lower limit of the macro-stickies’ size A 0 = 0.014 mm , would be sufficient for our purposes. An example of such graphical comparisons is illustrated in figure 19, for one of the studied pulps. The normal distribution: Aav=0.098, s(A s(A)=0.168 The lognormal distribution: av ln(A ln( A)=1.1614, s[ln(A s[ln(A)]=-3.17297, θ=0 The dislocated lognormal distribution:
µ=-4.4194, σ=2.2637, θ=0.014
1,0
0,8
n o i t c n u 0,6 f n o i t u b i r t s i d l a c 0,4 i r i p m E
0,2
0,0 0, 0
0, 2
0, 4
0, 6
0, 8
1, 0
Theoretical distribution function
Figure 19: Comparisons between theoretical and empirical distribution functions for three different distribution functions of the macro-stickies’ size A , in the studied pulp denoted as HX–S1 30
28
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Assuming that the dislocated lognormal lognormal distribution properly describes describes the macro-stickies’ distribution, we simultaneously acknowledge acknowledge it as that one which is equivalent to a truncated lognormal distribution distribution of the prime lognormal distribution of stickies. Therefore, in further studies the dislocated lognormal distribution was used as better fitted to elaborate experimental data on the stickies’ properties. Such distribution is not mentioned in the bibliography on the topic, so such explanation that the dislocated lognormal lognormal distribution is better suited to solving the stickies’ problem appears to be an important result of the ICP investigations. Regarding the physical sense of the parameters of the dislocated lognormal distribution of the macro-stickies’ size, a meaning of the threshold (location) (location) parameter θ is explained above as the lower limit of A 0. The shape parameter σ means value of the standard deviation of the natural logarithm of the variable f(x) =x– θ. The scale parameter µ is the mean value of the natural logarithm of the variable f(x) =x– θ. In former descriptions of the Statistica program by StatSoft , the restriction µ>0 was specified for such distribution; and it was repeated in ICP presentations and reports. However, finally such restriction was eliminated. It is rather obvious that in the case of so small particles the logarithmic value of their size must be below zero. Generally speaking, the higher µ values the larger particles are in the studied population of macro-stickies and their agglomerates or coagglomerates, commonly detected by the INGEDE method n°4; and the smaller σ values the narrower limits are within which the particle sizes of such population are included. Such a completely different behaviour of the ink particles (from inked newspaper, denoted as ONP), in comparison with the toner particles (discussed above), was found in the pulping experiments done with ONP recovered papers and with an admixture of equal parts of the reference PSA paper labels (altogether from about 200 mg to about 300 mg of the ADJ adhesives per kg of the oven dried pulp), and with agglomerants, such as 1-octadecanol (OD) or lauric acid (AL). The ink particles from the inked newspaper combined together with the particles of both studied adhesives so vigorously, during such pulping under the laminar regime of flow, that only the black co-agglomerates (GSV<100) were detected in the computer aided image analysis of rejects from the slotted screen with 0.10 mm slots. The adhesive particles alone or their agglomerates were totally outside of the observation which had been possible in the above studies with the recovered paper non-impact printed with toner (denoted OWP). It is so valuable information about the significance of both the size of particles (of toner or ink) and its hydrophobic character for the co-agglomeration of the particles (of toner or ink) with adhesive particles. A size of the ink particles are on a level of a few micro-meters, whereas the toner particles are almost ten times larger. It may be concluded from the ICP experiments that smaller and more hydrophobic particles are strongly inclined towards agglomeration and/or co-agglomeration processes. And therefore it may be assumed that smaller and more hydrophobic particles of the PSA materials should suit better towards such processes, and generalising, the large and hydrophilic particles are hardly appropriate for such processes. Moreover, even quite large particles of the PSA materials but in a form of flakes, i.e. leaf-like shaped particles, could be extruded through the fine slotted screen. So such quite chaotic process of the comminution of the PSA materials during the pulping step should be controlled, also by an admixture of the properly selected agglomerant which would additionally be able to increase the hydrophobic character of particle surfaces in statu nascendi , i.e. at the moment of creating of the minute particles of the PSA materials during pulping. Their further co-agglomeration with ink (or other) particles will lead to oblong granules of the contaminants, at the end of the pulping step properly run. And therefore further experiments carried out at the ICP pilot plant were devoted to such directed new pulping technology aimed at possibly conscious creating of such quite large and granulated particles of the contaminants. For describing an influence of the energy consumed during the pulping step on a course of such phenomena as: comminution, agglomeration, and co-agglomeration of sticky particles; the following parameters were used: the specific energy per unit weight of pulp (E s in [Wh/kg]), and the power provided per unit volume of slurry (in [W/l]) called in short the specific power N v. Such energy parameters were changed in a set of the pulping experiments in different pulpers in the ICP pilot plant, also using different rotors (Shark, Barracuda, and Helical) and applying different rotation speeds of the rotor. In each case, however, the attempts were made to keep the laminar flow of the slurry by proper adjusting its consistency. In the presented studies such papers were used as recovered paper: woodfree xerographic paper non-impact printed with black toner (denoted OWP) and inked wooden newsprint (denoted ONP). Additionally, agglomerants were applied such as: 1-octadecanol (denoted OD), lauric acid (AL), and stearic acid (AS), to the pulping experiments in alkaline medium, according to the INGEDE method n°11. The conditions under which the particular experiments were carried out are specified in table 3, in which the results gained for pulping by 30 minutes are also presented.
29
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Table PE.1: Pulping conditions tested during studying an influence of the energy parameters on the properties of macro-stickies; at constant pulping time = 30 minutes s e p l b t e a s i r a g v n t i p n l e u d p n e e h p t e d f n o I s r e p t e e t s m g a r n a i p p l t u n p a e t l t u h s f e o R
HX SB
HX SB
HX S
HX S
HX H
HX H
B
B
H1
H2
H2
H1
H3
m [k g] v[ l] no t 0 [oC]
ONP AL 25, 0 376 300 44, 0
OW P OD 25, 3 300 300 44, 0
ONP OD 26 338 30 0 50,0
OW P AL 35, 0 375 300 47, 0
ONP AL 6 7, 5 743 600 5 7, 0
OW P OD 90 667 300 52, 0
OW P OD 30 ,0 20 5 30 0 35 ,0
ONP AL 30 ,0 22 8 30 0 37 ,0
ONP OD 14 ,5 20 3 20 0 45
OWP AS 18 ,4 17 8 20 0 60
ONP AL 17, 5 213 200 45, 0
OW P O OD D 16, 5 183 200 59, 0
OW P OD 21, 5 179 200 57, 0
Final temperature S t oc k c ons is tenc y Net power consumpt ion
tk[ oC] c [ g/ l] Nn[k W ]
49, 0 66, 6 11, 8
49, 0 84, 3 12, 4
55,0 76,8 10,9
50, 0 93, 3 11, 7
5 6, 0 9 0, 8 2 7, 2
64, 0 13 5, 0 27, 2
33 ,0 146, 0 1, 6
35 ,0 1 31,7 1, 6
54 71 ,6 3, 1
65 103, 1 7 ,1
54, 0 82, 3 7, 5
56, 0 90, 3 2,6
Specific power (per unit of volume)
Nv [W/l]
31,4
41,1
32,1
31,2
36,7
7,9
7,2
15,1
39,6
35,3
14,5
80,5
Specific energy (per unit of mass)
Es [Wh/kg]
235,5
244,1
208,9
166,9
201,8
151,1
27,2
27,2
105,2
192,1
214,3
80,0
334,3
Admixture of PSA a dhesives
Sm [mm /kg 13572
13411
13050
9694
10053
7540
11310
11310
15600
12293
12926
13709
10521
Kind of pulper
Kind of recovered paper Kind of agglomerant Rec ove red paper in pulper S l urry volum e Num ber of labels (29x 39 mm ) Initial temperature
2
Amount of macro-stickie s S A[m m2/kg] l r a i s - ; s o c o i e f Number of macro-stickie s S N [no/ kg] r i t s y c k t a s l a i c a i t a Scale pa rameter of log-norm. distr. a n M t s d t s a Shape para meter of log-norm. Distr
Key:
76, 0 120, 4 14, 4
6036
5268
2829
3725
3394
2618
3599
3490
7360
7034
3403
8548
7774
59040
45720
28920
17280
37380
12420
11700
16380
41820
59880
32760
61656
213060
-3, 7544
-3, 7787
-4, 4194
-3, 5482
-4, 5892
-3, 0252
-3, 7629
-4, 0987
-3,5846
-4, 6558
-3, 9897
-4,3295
1, 9892
2, 0507
2, 2637
2, 4249
2, 3128
2, 2242
2, 4579
2, 4204
2,1697
2, 2994
2, 1225
2,1109
HX S – hydrop ulper (V = 1000 l) with the Shark rotor HX SB – hydropulper (V = 1000 l) with the Shark and Barracuda rotors HX H – hy dropulper (V = 1000 l) with the Helico rotor Hi – hydropulper (V = 400 l) with the changeable revolutions (i – index of revolution number n)
i
n [rev./min.]
i=1 n=420 [rev/min.];i=2 n=640 [rev./min]; i=3 n=830 [rev./min.]
B – drum pulper
1
420
Kind of recovered paper:
ONP – old newspaper
OWP – wood-free xero-paper with toner
2 3
640 820
Kind of agglomerant:
AL – lauric acid
OD – octadecanol
AS – stearic acid
Table 3: Pulping conditions tested during studying an influence of the energy parameters on the
properties of macro-stickies, macro-stickies, at constant constant pumping time = 30 min. min. In the pulping experiments presented in table 3, the specific energy consumption E s was changed within the range from 27.2 to 244.1, in [Wh/kg]; the lower limit for the drum pulper (denoted B) and the upper one for the hydropulper equipped with Shark and Barracuda rotors (denoted HX SB). And the specific power N v, or energy dissipation during during pulping, was within the range from 7.2 to 41.1, in [W/l]; again the lower limit for the drum pulper and the upper one for the hydropulper equipped with Shark and Barracuda rotors. Such gentle conditions under which the pulping step is run in the drum pulper need short comments. In industry drum pulpers are used in deinking lines to improve efficiency of hydrogen peroxide (H 2O2) in bleaching the lignified fibres of recovered newspapers. At given dosage of the bleaching agent, the higher its concentration is as the higher stock consistency is in the pulper. And the higher concentration of hydrogen peroxide peroxide means the more efficient bleaching activity of this agent. And therefore the pulping of the recovered newspaper at higher consistencies, which are possible in drum pulping, is such an efficient way to maximising the bleaching effect; however, under gentle conditions applied in the drum pulping (perceived also in such low values of E s and of Nv, presented above) there is no possible to achieve all effects of the fully completed pulping which are required for good deinking. So the drum pulping is only the first step of the pulping process which, in the deinking lines applied drum pulpers, is followed by an additional treatment (e.g., in turboseparator), consuming much more energy, in which slurry is processed under severe conditions, to achieve all effects of the efficient pulping which are needed for good deinking. Such additional treatment of the slurry after the drum pulping was not applied in the ICP pulping experiments. They were aimed at investigating of such a wide spectrum of the conditions under which the pulping step is run in the industrial practice. It was possible due to possessing of so different pulpers in the ICP pilot plant, after many years of research & development activities on processing recovered papers. Because the distribution of the macro-stickies’ size is better described as the dislocated lognormal distribution, the attempt was made to find a relationship between the parameters of such distributions (i.e. the scale parameter µ and the shape parameter σ) and the energy parameters of the pulping process. The results gained in a stepwise method of the multiple regression analyses are collected in table 4, showing a lack of linear relationship between the scale parameter µ of the dislocated lognormal distribution of the macro-stickies’ size and the studied energy parameters of the pulping experiments in which two different recovered papers (with toner or printing ink) were pulped under different conditions, with an admixture of both reference PSA paper labels, and with an addition of three different agglomerants; in each case under the laminar regime of flow of the slurry in the pulper. Instead of so many variables in the ICP pulping experiments, it was proved, however, (see table 4) that the shape parameter σ, of the dislocated lognormal distribution of the macro-stickies’ size, depended in statistically significant way on both the energy dissipation N v during pulping and the amount of the PSA adhesives S m in recovered paper. And therefore further optimisation studies were planned for the rational selection of the most efficient agglomerants.
30
L SCREEN C CLEAN EAN
a)
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
average macro-ctickies size - µ, according to the dislocated lognormal distribution N=33 R=0 s(y)=0,53
i
Variable
0
b0
βi
s(βi)
bi
s(bi)
-3,31
0,65
b)
shape parameter σ ln(A)t ln(A)t - of the dislocated lognormal distribution of the macro-stickies size N=33 R=0,671 s(y)=0,19
i
Variables
0
b0
1
Nv [W/l] 2
3
Sm[mm /kg]
βi
s(βi)
bi
s(bi)
3,31
0,23
-0,64
0,14
-0,0128
0,0028
-0,47
0,14
-0,000055
0,000017
Table 4: Summary of the regression analysis of the µ and σ parameters parameters related to the
energy parameters of pulping recovered papers
Conclusion
Almost all studied PSA labels on plastic foils survive the pulping of recovered paper without their dividing into smaller parts, and with an inconsiderable comminution of their adhesive layer. In PSA paper labels, however, their paper base is quickly defibering during their pulping together with recovered paper, and it is followed by the comminution of their adhesive layer. Nevertheless, there is a lack of any agglomeration effects among the sticky particles without admixture of the agglomerant. Processes of agglomeration and co-agglomeration are efficiently promoted by an admixture of the agglomerant to the pulping step run under the laminar regime of flow. There is a possibility to combine together the sticky particles from the PSA materials with toner particles and especially with ink particles which are smaller than toner particles, to create oblong co- agglomerates needed for their successful removal in the subsequent steps of the deinking process, especially in fine slot screening. So such new pulping technology is possible which is aimed at serving (in the same time) to remove both contaminants from the slurry (in the subsequent steps of deinking process), namely: ink (or toner) particles and sticky particles (stickies). The dislocated lognormal distribution is better fitted to characterise the populations of stickies, as well as their agglomerates and co-agglomerates (with other particles). The parameters of that distribution, such as the scale parameter µ and the shape parameter σ, are useful in search for solutions towards the stickies’ problem abatement. The higher µ values the larger particles are in the studied population of macro-stickies and their agglomerates or co-agglomerates, commonly detected by the INGEDE method n°4; and the smaller σ values the narrower limits are within which the particle sizes of such population are included. The studied energy parameters, such as the specific energy consumption E s and the specific power N v (or power dissipation during pulping), are useful as the criteria of energy similarity for rational planning of the industrial trials, according to the results gained in the ICP pilot plant. However, quite chaotic process of the comminution of the PSA materials during the pulping step may be controlled not only by the energy parameters, during the pulping run under the laminar regime of flow, but additionally by an admixture of the properly selected agglomerant which would also be able to increase the hydrophobic character of particle surfaces in statu nascendi , i.e. at the moment of creating of the minute particles particles of the PSA materials during the pulping step. Optimisation studies are required for a rational selection of the agglomerant which will be able to combine together particles from typical contaminations present in recovered paper, such as ink and toner, with the sticky particles, and which will be cost efficient as well.
31
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.1.3. Comparison of drum and batch pulpers The drum pulper is generally considered as the reference technology for gentle pulping conditions (reduced fragmentation of the adhesives). Therefore, pilot trials were performed with a drum pulper (a slice of an industrial-sized drum) compared to a batch helical pulper which represents conventional technology in deinking grades (figure 20). The comparison of the influence of both pulping devices on the adhesive fragmentation was based on the assessment of: -
the stickies particle size distribution achieved in both pulpers,
-
the screening screening ability of of the resulting stickies under under pilot screening conditions with fine fine slots.
Figure 20: KADANT-LAMORT HELICO Batch Pulper (left) and CTP “Slice” Drum Pulper (right)
A lab helical pulper was also compared compared to the two pilot pulpers. The test were done with the reference reference adhesives (acrylic adhesive E115 and hot melt rubber adhesive D170 introduced at 0.3% in weight) using conventional conventional deinking chemistry (0.7% NaOH, 2% silicate, 0.7% H 202 and 0.75% calcium soap). Pulping temperature was 40°C and the other pulping conditions reported in table 5 were according to usual conditions for each type of pulper.
Pilot Pilot helic helical al pulpe pulperr
Pilot Pilot drum drum pulpe pulperr
Lab helica helicall pulpe pulperr
Raw material quantity (kg)
25
50
0,35
Duration (min)
15
25
15
13 to 15
16 to 20
10
475
11
475
Consistency Consistenc y (%) Rotation speed (rpm)
Table 5: Pilot and lab pulping conditions
The pilot screening was performed under the following conditions: -
High-consistency High-consis tency screen:
KADANT LAMORT CH3 (300 mm screen cylinder diameter)
-
Rotor type:
Foil rotor
-
Screen plate:
0.20 mm slots, MicroVortex design, 1.2 mm profile height
-
Pulp consistency:
2.5 %
-
Slot velocity:
about 1 m/s
-
System design:
Simulation of a 3-stages feed forward screening system. Typical final reject rates were between 1 and 2 % in weight.
32
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Pilot plant results: comparison drum versus helical pulper
The cumulative size distributions of stickies particles achieved during pulping are reported in figure 21. It can be seen from these results that drum pulping induces for both adhesives a shift of the stickies particles towards larger size compared to helical pulping.
Water-based acrylic adhesive E115
Hot-melt rubber adhesive D170
100
100 Helico
Helico
Drum
) 80 a e r a l a t o t f 60 o % ( a e r a 40 s e i k c i t S
) 80 a e r a l a t o t f 60 o % ( a e r a 40 s e i k c i t S
20
Drum
20
0
0
0 2 2 0 0 -
0 4 4 0 0 -
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 6 0 5 0 1 1 0 1 5 1 2 2 0 2 5 2 3 3 0 5 0 0 > 0 0 0 0 0 0 0 -
0 2 0 0 - 2
0 4 0 0 - 4
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 0 5 0 1 0 5 1 2 0 5 2 3 0 5 0 0 - 6 > 0 - 1 0 - 1 0 - 2 0 - 2 0 - 3 0 0 -
Figure 21: Cumulative stickies size distribution after pulping (drum versus helical pilot pulpers)
The results in figure 21 confirm the softer conditions of drum pulping (compared to helical pulping) which finally give rise to larger stickies particles, which should be easily removed during screening. This point has therefore been cheeked by the determination of the pressure screening cleanliness efficiencies achieved in both pulping conditions: -
Acrylic adhesive E115:
-38% with the drum pulper
-4% with the drum pulper
-
Hot-melt rubber adhesive D170:
17% with the drum pulper
65% with the drum pulper
The negative cleanliness efficiency achieved in both cases with the acrylic adhesive corresponds to a higher adhesive content in the screening accepts compared to the feed, which suggests that the stickies particles were very small (i.e. they pass easier the slots than the fibres) and may also have been fragmented during the 3 screening stages. The positive efficiency achieved in both cases with the other adhesive suggests that the adhesive particles were larger and/or that the hot-melt based adhesives had been less fragmented during screening. The higher screening efficiencies achieved with both adhesives after the drum pulper compared to the helical pulper are in line with the results in figure 21 showing larger stickies after the drum pulper. Figure 21 also shows somewhat larger stickies particles in the case of the acrylic adhesive compared to the hot-melt rubber adhesive, i.e. a higher fragmentation with the acrylic adhesive. This is consistent with the results reported in section 3.3.1 as the acrylic adhesive showed much lower strength than the hot-melt rubber adhesive. These results also suggest higher fragmentation of the acrylic adhesives during screening and thus a lower efficiency. Higher extrusion of the softest stickies through the slots contributed probably also to the low screening efficiencies observed with the acrylic adhesives, as reported hereafter in section 4.2. Stickies shape, a parameter which could not be measured, might also contribute to explain the higher screening efficiency achieved with the hot-melt rubber adhesives, since the particles were probably thicker and more round-shaped, as observed earlier at CTP in the case of low-consistency lab pulping tests done with other acrylic and rubber based (SBR) adhesives.
Comparison with lab helical pulping
Lab pulping is often used to evaluate the behaviour of adhesive products during recycling and determine their “recyclability”. Nevertheless, for scale reasons, the shear forces which take place in lab pulpers are certainly different from those generated in pilot or industrial pulpers. The consequence of this situation is a possible non representative fragmentation achieved during lab pulping which may induce wrong estimation of the adhesive behaviour, especially regarding screening ability.
33
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Therefore, the adhesive fragmentation achieved during lab helical pulping has been checked, with the E115 acrylic adhesive only, and compared to that one achieved during pilot drum and helical pulping. The results in figure 22 showed that lab helical pulper induced a much lower adhesive fragmentation than the pilot pulpers. This trend (which was expected) shows the difficulty to produce representative stickies fragmentation at lab scale.
100 a e r a l a t o t f o % ( a e r a s e i k c i t S
90
Helico
80
Drum
70
Lab Helico
60 50 40 30 20 10 0 0 2 2 0 0 -
0 4 4 0 0 -
0 0 0 0 0 0 0 0 0 0 0 0 0 6 6 0 5 0 5 0 1 1 0 1 1 2 2 0 2 2 3 0 5 0 > 5 0 0 0 0 0 0 0 0 0 -
Figure 22: Cumulative stickies size distribution after pulping (lab versus pilot pulpers)
Conclusion
The fragmentation of adhesives during pulping showed to depend on the type of pulper: The drum pulper induced lower adhesive fragmentation compared to the helical pulper and consequently enabled to achieve higher stickies removal efficiencies during subsequent screening with fine slots. Moreover, it was shown that the lab helico pulper cannot lead to adhesive fragmentation similar to that one achieved in the pilot pulper: Stickies fragmentation was much lower in the lab helical pulper, for both reference adhesives, which confirmed the difficulty to assess adhesive behaviour (overall their ability to be broken up during pulping) by simple lab tests. The pilot pulping trials showed clear benefits of drum pulpers, at least from the stickies point of view. However, as the results obtained with the pilot batch pulper were not consistent with those obtained on lab scale, it seems difficult to conclude about stickies fragmentation in batch pulpers on mill scale. Indeed, the drum pulper has generally been regarded as a reference in terms of gentle pulping action with consequently reduced fragmentation of stickies and contaminants, which are rejected in large pieces at the outlet of the drum. By contrast recent macro-stickies analyses performed in several deinking mills, showed a strong reduction of the average stickies surface area and an increase in the stickies number with drum pulpers compared to batch pulpers, which led to the conclusion that drum pulpers produced a stronger stickies fragmentation [38]. Consequently, Consequently, it seems difficult to draw clear conclusions about which pulping technology should lead to the lowest stickies fragmentation on the basis of the pilot trials performed in this project. Recent comparative tests of the drum and batch pulpers available at the pilot facilities of a major equipment supplier [38] did not allow to draw clear conclusion about stickies fragmentation as well. The drum pulper may still be considered as a relatively gentle pulping technology since no strong mechanical forces are exerted on the adhesives. By contrast, the helical rotor of the batch pulper may generate locally stronger mechanical forces, especially if the rotor velocity and/or the pulping time exceed the minimum conditions required for the defibering of recovered papers and for the optimised detachment, fragmentation and re-deposition of inks. Indeed, the batch pulper offers a larger set of parameters, including consistency, pulping time and rotor velocity and design, which can be optimised compared to the drum pulper. The batch pulper is also more adapted to the development of the new pulping technology technology at ICP, as the flow conditions (pulp rheology, rheology, velocity field) can be changed quite easily.
34
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.1.4. Development of a new pulping technology technology The ICP experiments proved that an admixture of the agglomerants to the deinking bath was successful for provoking and further progress of the agglomeration and co-agglomeration processes among the sticky particles and the particles of other contaminants present in recovered paper, mainly ink and toner particles, during pulping of the recovered paper run under the laminar regime of flow. However, the agglomerants should be properly selected for such a new pulping technology oriented to satisfy two such different processes as de-inking and de-sticking, and, moreover, the agglomerants should be cost efficient. And therefore the agglomerants should replace a part of oleic soap, the primary component of the traditional deinking bath.
A course of optimisation studies
The aim of the optimisation studies was to find the optimum balance among the component used for deinking and the agglomerants additionally added to control the comminution of the PSA layers and to enhance chances of successful agglomeration and co-agglomeration of the stickies created from the PSA layers; under the conditions properly chosen for favouring the agglomeration and coagglomeration processes. To achieve the aim, among the known statistical procedures, used for optimising a composition of the mixture according to the results of properly planned experiments, one procedure was selected which is called the procedure to maximize of the product. That procedure consists of two basic stages, namely: -
Searching for adequate models (i.e., prediction equations) aimed at approximation of the product characteristic as a function of the studied process parameters. In the studied case the product, i.e. deinked pulp, was characterised by such optical properties as R 457, k700 and by both parameters µ and σ of the dislocated lognormal distribution describing the properties of macro-stickies present in the deinked pulp. And the studied parameters of the process, i.e. the pulping step, were both a composition of the surface active agents and the conditions under which it was applied.
-
Specifying optimum values of the studied parameters of the process according to analysis of the function of the overall desirability of the product. In the studied case, specifying optimum values both for the composition of surface active agents and for the parameters characterising the conditions under which the pulping step was run.
The computer program Experimental Design (Industrial DOE) – Mixture Design and Triangular was used both for planning the experiments experiments and for statistical analyses of the Surfaces by the StatSoft was results gained in the experiments, devoted to study an influence of the studied parameters, i.e. the composition of selected surface active agents and the conditions during pulping step, on an efficiency of the removal of both ink particles and stickies. To assess the both processes (i.e., de-inking and desticking) following properties of the deinked pulp were studied: -
Brightness R457 after hyperwashing and bleaching; Absorption coefficient k700 after hyperwashing and bleaching; i.e., the scale parameter of the dislocated lognormal lognormal distribution of the macro-stickies’ size;
-
µ,
-
σ, i.e.,
the shape parameter of the dislocated lognormal lognormal distribution of the macro-stickies’ size.
And the integral simplex design was applied to determine approximation functions functions of the relationships relationships between the above properties of deinked pulps and compositions of the ternary mixtures of selected surface active agents. The starting models for the above defined properties of deinked pulps were elaborated from the analysis of the results gained in the experiments had been realised according to the so-called simplexcentroid design for a ternary mixture of the chemically pure surface active agents, with exactly defined physico-chemical properties. The components were selected in a former part of our studies, when their usefulness as the agglomerants had been experimentally confirmed. The agglomerants, together with oleic acid, are presented below: -
oleic acid (molecular mass 282.4 Daltons, melting point 16 °C);
-
stearic acid (molecular mass 284.5 Daltons, melting point 69.9 °C);
-
1-octadecanol (molecular mass 270 Daltons, melting point 58 °C).
35
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
According to the plan, the optimisation studies should involve commercial products, replacing the model surface active agents used in the former studies in a form of chemically pure reagents. And therefore the commercial products, with a similar chemical structure but lower melting points, were used in the studied slurries with purposefully lowered temperature during pulping step. An algorithm of the applied procedure is graphically presented in figure 23.
Figure 23: Three simplexes for three different (AT1, AT2, and AT3) compositions of surface active
agents, with oleic acid (AO) in each composition. Additionally other surface active agents were applied: 1-octadecanol 1-octadecanol (OD) and stearic acid (AS) in the AT1 composition, OD was replaced by di-stearate of ethylene glycol (E2) in the AT2 composition, and AS was substituted with ethoxylated technical stearine (S2) in the AT3 composition In each new simplex a part of the results from the former one was used, reducing labour demand of the research process. A selection of the new vertex of the triangle, done after analyses of the results of the former simplex, was arbitrary in its character. Oleic acid was applied in each composition of the selected surface active agents, because it is needed for the deinking process in an alkaline medium. Oleic acid content was changed within a range 0 – 1 (i.e. from 0% to 100%). However, in contents of these other surface active agents, used as the agglomerant, values of their upper level were limited, in an arbitrary way, applying for such cases the proper plans for areas with constrained designs. Applying the computer program program of regression analysis analysis for ternary mixtures, containing (among (among others) the significance analysis of regression coefficients, as well as of the adequacy of regression equations, such a model equation was fitted which described in the best way the relationship between experimental results and the composition of the selected surface active agents. The regression equations defined in the above way were the basis for establishing optimum values of the studied parameters parameters of the pulping step, i.e. the composition of surface active agents and the temperature of its use, according to the analysis of the overall desirability desirability function. However, individual properties of deinked pulps from recovered paper, such as optical properties of handsheets (R 457, k700) and the parameters of the dislocated lognormal distribution of macro-stickies (µ and σ), have various relationships with their approximated values (according to the regression equations) and with the desirability of the values. The relationship between predicted responses on one or more dependant variables and the desirability of responses is called the desirability function. In the computer program Statistica (by StatSoft ); ); used in these optimisation studies, a procedure developed by Derringer and Suich (1980) is applied for specifying that relationship. Their procedure involves transforming scores on each of the outcome variables into desirability scores that could range from 0.0 for unacceptable to 1.0 for very desirable. However, these optimisation studies were conducted within certain limitations such as a linear character of the desirability function, assigning to the desirability value of 0.5 to the arithmetic mean of the extreme values had been measured for the studied property. And the extreme values were assigning (respectively) (respectively) to the minimum and maximum values of the desirability desirability function. According to Derringer and Suich the overall desirability may be computed as the geometric mean of the individual desirabilities, and their procedure provides a straightforward way for transforming predicted values for multiple dependent variables into a single overall desirability score. The problem of simultaneously optimisation of several response variables then is reduced to selecting the levels of the predictor variables that maximize the overall desirability.
36
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The optimisation experiments were performed in the ICP pilot plant under the following conditions: -
Volume of fibrous slurry in the hydropulper H: 150 litres,
-
Model recovered paper: ONP, i.e., returns of not sold daily newspapers,
-
Source of the stickies: the PSA paper labels of the Avery Dennison Jackstädt (ADJ), namely ADJ water-based acrylic adhesive (E115), and ADJ hot-melt based rubber adhesive (D170). Both stuck to the xerographic paper that had been non-impact printed on its entire surface (one side only) with the full intensity of the toner (cyan for E115, and yellow for D179), and later cut down into the pieces: 39 mm x 29 mm. Hundred pieces of each kind were introduced into the pulper to every pulping experiment, i.e. altogether about 300 mg of the PSA layers per kg of the oven dried pulp;
-
Pulping time: a cycle of the pulping consisted of : 1) actual pulping during 30 minutes, at consistency 7-9% and at 660 rpm of the rotor, both assuring the laminar flow of the slurry; 2) additional pulping during 15 minutes, after dilution of the slurry with cold water in quantity needed to chill the slurry below the melting point of the most easily melted agglomerant (after the dilution consistency within the range 4.5-6.5%), with a reduction of rotor revolutions to 416 rpm, to keep the laminar regime of flow;
-
Deinking formulation: constant in the pulping experiments, according to the INGEDE method no. 11, however, without H 2O2, and without the constant percentage of chemicals defined by the method, but with quite constant concentrations of the chemicals, to keep constant physico3 chemical effects, namely within the ranges: NaOH 0.6 – 0.7 g/l, sodium silicate (density 1.4 g/cm ) 1.7 – 2.2 g/l, oleic acid 0.9 – 1.2 g/l;
-
Agglomerant Agglomerant admixture: the agglomerants, agglomerants, in quantities and ratios according to the experiment experiment design, were dissolved in oleic acid and the mixture was later emulsified in the alkaline deinking bath in the pulper;
-
Temperature: an initial temperature of the slurry in the pulper was a function of melting points of the applied agglomerants.
For some reasons hydrogen peroxide was not introduced to the slurry, mainly to avoid any uncontrolled decomposition of hydrogen peroxide during the pulping step. However, handsheets prepared from final pulps (after full cycle of the pulping during 45 minutes) were additionally hyperwashed and later bleached with hydrogen peroxide under constant conditions. The handsheets were used to determine the optical properties of the pulp, such as brightness R 457 and absorption coefficient k700. Owing to a lack of the instrument to testing at ICP the optical properties in the near infrared area, the absorption coefficient k 700 was determined instead of ERIC (i.e. Effective Residual Ink Concentration) recommended by the standard TAPPI T567, for λ > 950 nm. In comparisons of the optical properties of the handsheets after their hyperwashing alone and of the handsheets after their hyperwashing and after their additional bleaching, values of the absorption coefficient k 700 were practically not changed. This means that the k 700 value was only influenced by the ink content in the handsheets, not by chromophore groups of fibres. However, the parameters of the dislocated lognormal distribution of macro-stickies ( µ and σ) were studied for both deinked pulps, received after pulping during 30 minutes, denoted as µ30 and σ30, and received after the fully completed pulping (during 45 minutes), after dilution of the slurry with cold water in quantity needed to chill the slurry below the melting point of the most easily melted agglomerant, denoted as µ45 and σ45. Such dividing of the pulping cycle was aimed at finding a possible influence of the phase transition of agglomerants (from liquid to solid) on the sticky particles and ink particles as well.
Pilot plant results gained in the optimisation studies
In each set of the experiments, performed for three compositions of surface active agents (AT1, AT2, and AT3), the experiment design and the results gained in it were presented, together with regression analyses of the gained results, as well as with the optimisation of the given composition according to the overall desirability function. Such road to specify the optimal ratio among the studied surface active agents for the given composition is shown below, using experiments with final the AT3 composition as an example. The optimisation studies were finished with the studies on the AT3 composition thanks to such very positive results gained for the AT3 composition. composition.
37
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
In the final set of experiments (on the AT3 composition) the chemically pure stearic acid, a strongly hydrophobic hydrophobic component, was substituted with the product of ethoxylation of the technical stearine, with the chemical formulae: CH 3(CH2)16(CH2CH2O)n, where n=2; denoted as S2. The S2 component is not 0 as strongly hydrophobic as pure stearic acid is, and moreover, its melting point (40 C) is visibly lower than melting point of pure stearic acid (69.9 °C). It means that such a substitution could reduce the pulping temperature. Like in the former experiment design, with the AT2 composition, the upper limits of the agglomerant content were established in the AT3 composition, namely 75% for both S2 and E2 components. The experiment design for the AT3 composition and the results are presented in table 6.
N°
1 2 3 4 5 6
A ratio between the studied surface-active surface-active agents S2 AO E2 0,00 1,00 0,00 0,75 0,25 0,00 0,75 0,00 0,25 0,00 0,25 0,75 0,25 0,00 0,75 0,35 0,30 0,35
Parameters of the dislocated lognormal distribution of the stickies size σ30
µ30
σ45
µ45
2,2532 2, 2532 1,8347 1, 8347 2,2016 2, 2016 2,2648 2, 2648 2,3835 2, 3835 2,2332 2, 2332
-3,6015 -5,5399 -3,2347 -3,7236 -3,4059 -5,3109
2,4605 1,7895 2,2529 2,4061 2,2630 1,9156
-3,5601 -5,5200 -3,4372 -3,9427 -4,4321 -5,0859
Optical properties R457 61,9 61.6 55,5 60,7 59,3 61,9
k700 1,7 2.1 3,5 2,4 3,0 2,0
Table 6: Parameters of the stickies’ size distribution (µ and σ ) and optical optical properties (R 457 ) of 457 and k 700 700
deinked pulp after hyperwashing and bleaching; the data gained in realizing the planned experiments related to the AT3 composition of surface-active agents: S2 –ethoxylated technical technical stearine, AO – oleic acid, E2 – di-stearate of ethylene ethylene glycol glycol
According to the results of the regression regression analyses, given in tables 7 and 8, the value of each studied parameters has significantly been influenced by every component of the AT3 composition of surface active agents, and moreover, also by an interaction between components AO and S2. A lack of significant difference between values of the adequate regression coefficients for µ30 (pulping during 30 minutes) and µ45 (pulping during 45 minutes) may be understood as an evidence that the second part of pulping has no influence on that scale parameter; it means that the second part is without alteration in the mean macro-stickies’ size. However, different interactions among the components of the AT3 composition cause alteration in values of the shape parameter σ.
Denoting the variables 1 – x1 (S2) 2 – x2 (AO) 3 – x3 (E2) 4 – x1*x2 5 – x1*x3 Denoting the variables 1 – x1 (S2) 2 – x2 (AO) 3 – x3 (E2) 4 – x1*x2 5 – x1*x3
µ30
bi -2,92 -3,65 -3,80 -14,1
s(bi) 0,47 0,34 0,34 2,4
µ45
-l0,95 -4,9 -5,1 -5,2 -24,6 -24,6
+l0,95 -0,9 -2,2 -2,3 -3,6
bi -3,21 -3,50 -4,47 -12,2
s(bi) 0,41 0,29 0,29 2,1
σ30
-l0,95 -5,0 -4,8 -5,7 -21,3 -21,3
+l0,95 -1,4 -2,2 -3,2 -3,0
-l0,95 1,7 2,1 1,9 -5,9
+l0,95 2,8 2,9 2,7 -0,1
σ45
bi 1,729 2,246 2,238
s(bi) 0,061 0,047 0,061
-l0,95 1,46 2,04 1,97
+l0,95 1,99 2,45 2,50
1,60
0,30
0,3
2,9 2,9
bi 2,24 2,46 2,30 -2,99
s(bi) 0,13 0,09 0,09 0,67
Table 7: 7: Regression analyses of the relationship between parameters of the dislocated lognormal distribution of the macro-stickies’ size ( µ and σ) and the AT3 composition of surface active agents: S2 – ethoxylated technical stearine, AO – oleic acid, E2 – di-stearate of ethylene glycol
38
L SCREEN C CLEAN EAN
Denoting the variables 1 – x1 (S2) 2 – x2 (AO) 3 – x3 (E2) 4 – x1*x2
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
R457 bi 53,9 61,8 60,7 30,6
s(bi) 0,4 0,3 0,3 2,2
k700 -l0,95 52 60 59 21
+l0,95 56 63 62 40
bi 3,82 1,68 2,65 -6,56
s(bi) 0,11 0,07 0,08 0,57
-l0,95 3,3 1,3 2,3 -9,0
+l0,95 4,3 2,0 3,0 -4,1
Table 8: Regression analyses of the relationship between optical properties (R 457 and k700) of the deinked pulps (after hyperwashing and bleaching) and the AT3 composition of surface active agents: S2 – ethoxylated technical stearine, AO – oleic acid, E2 – di-stearate of ethylene glycol The alterations in values of the shape parameter σ are better visible in figure 24. In the first part of pulping the interactions of oleic acid (AO) with the E2 component result in an increase of the σ30 values, adding a convex shape to the response with a distinct decrease in direction to the S2 vertex of the triangle. However, during the additional pulping, after chilling the slurry, interactions of oleic acid (AO) with the S2 component become prevailing, causing a decrease of the σ45 values, and adding a concave shape to the response with minimum value of σ45 for equal parts of oleic acid (AO) and S2.
σ30
σ45
2,4 2,3 2,2 2,1 2 1,9 1,8
2,4 2,2 2 1,8 1,6
σ30=1,729x 1+2,246x2+2,238x 3+1,60x1*x3
σ45=2,24x1+2,46x 2+2,30x 3-2,99x 1*x2
2
2
R =0,9735
R =0,9523
Figure 24: The three-dimensional three-dimensional contour plot and as its projection on a two-dimensional plane
illustrating the regression relationship between the shape parameters ( σ 30 ) of the dislocated dislocated 30 and σ 45 45 lognormal distribution of the macro-stickies’ size and the AT3 composition of surface active agents
Such interactions of oleic acid (AO) with the S2 component result also in lowering the µ value, and therefore it may be assumed that the changes in both parameters ( µ and σ) of the dislocated lognormal distribution of the macro-stickies’ size are caused not by comminution of the PSA particles (decrease of µ) but rather by co-precipitation of the S2 particles together with ink particles on surfaces of the macro-stickies. In such co-agglomerates, created by the macro-stickies’ particles and ink particles, the sticky character of the macro-stickies may be lost. After such detackification those macro-stickies could be not detected by the INGEDE method n°4, which is utilising the adhesive properties of the sticky particles to provide the contrast to the specimen’s background which is required for image analysis. It means that the discussed decrease in macro-stickies’ size (regarding a decrease in the µ values) is rather doubtful.
39
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Interactions of components of the AT3 composition of surface active agents in shaping the optical properties R 457 and k700 (after hyperwashing & bleaching) of deinked pulps, according to the given regression models, are illustrated in figure 25, in a form of contour plots. It is clearly perceived that presence of oleic acid (AO), in the AT3 composition of surface active agents is of significance to improving both optical properties of the deinked pulps, i.e. to increase their R 457 values and to decrease their k 700 values.
R457
k700
E2
E2
0,00 1,00
0,00 1,00
0,25
0,25
0,75
0,50
0,50
0,50
0,75
0, 25
0, 50
0, 75
0,50
0,75
0,25
1,00 0, 00
0,75
0,00 1, 00
S2
AO
66 64 62 60 58 56 54
0,25
1,00 0, 00
0, 25
S2
R457=53,4*x1+61,8*x2+60,7*x3+30,6*x1*x2
0, 50
0, 75
0,00 1, 00 AO
3,5 3 2,5 2 1,5 1
k700=+3,82*x 1+1,68*x2+2,65*x 3-6,56*x1*x2
2
2
R =0,994
R =0,994
Figure 25: The contour plots, of the coordinates of the regression relationships for the optical properties (R 457 ) of deinked pulps pulps (after hyperwashing hyperwashing and and bleaching), bleaching), dependent dependent on a ratio 457 and k 700 700 among the surface active agents applied in the AT3 composition: S2 – ethoxylated technical stearine, AO – oleic acid, E2 - di-stearate of ethylene ethylene glycol
Results of the optimisation procedure for optimising the ratio among the surface active agents in the studied composition, regarding the specified desirability functions of both parameters ( µ and σ) of the dislocated lognormal lognormal distribution of the macro-stickies’ size and both optical properties (R 457 and k700) of the deinked pulp – are illustrated in figure 26. Like in former series of these experiments, the desirability functions were specified by assigning extreme values (the lowest and the highest) of the results gained for the AT3 composition to (respectively) a desirability value of 0.0 or a desirability value of 1.0; see the right column of graphs in figure 26. And in the lowest graphs of the overall desirability (see also figure 26), it is possible to find an optimum balance among surface active agents in the AT3 composition, namely: AO – oleic acid
0.90 (90%)
S2 – ethoxylated technical stearine
0.10 (10%) (10%)
E2 – di-stearate of ethylene glycol
0.00 ( 0%)
In this case the overall desirability of the product has a satisfactory value, equal to 0.668. And the predicted responses for such an optimum AT3 composition are presented below: σ45
-
2.18;
µ45
- -4.52;
95%
<1.8 – 2.5>
95% <-5.6 – -3.4>
R457 - 63.7;
95% <62.5 – 64.9>
k700 -
95% <1.0 – 1.6>
1.3;
40
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
S2
AO
E2
Desirability
3,50 2,54
2,46
0, ,5
2,18
1,
5 4
1,83
2,12 1,79
1,00
-3,39
1, -3,437
5 4
-4,52
,5
-4,479
0,
-5,520
-5,65 -9,00 64,9
60,0
63,7
0,
62,5
,5
1,
58,7 55,5
7 5 4
R
4,5 0, 0 0 7
1,6
k
1,3
3,5 ,5
2,6 1,
1,7
1,0
y t i l i b a r i s e D l l a r e v O
0,668
0 0,10
0,75
0
,90
0
0, 75
Figure 26: Final results of the optimisation optimisation studies on the AT3 composition composition of the surface active active
agents (S2 – ethoxylated technical stearine, AO – oleic acid, E2 – di-stearate of ethylene glycol), according to desirability profiles It should be pointed out that the 95% confidence intervals of the predicted responses are satisfactorily narrow for the optimum balance of ingredients (i.e. surface active agents in the AT3 composition) that optimises the overall desirability of the product, i.e. the parameters and properties presented above. In comparison with the optimum ratios specified for the formerly studied compositions (AT1 and AT2), a small percentage of the agglomerant, only 10% of ethoxylated technical stearine (S2) attracts attention in the optimum AT3 composition in which the rest (90%) is oleic acid. Perhaps, however, stearic acid, specified in larger quantities in the former optimum compositions (AT1 and AT2), was in a greater part present in the compositions as the stearate soap, after the saponification reaction with sodium hydroxide, and as such supported activities of oleic soap. Under such conditions the agglomerant role was played only by unsaponified part of stearic acid, or by that part of the stearate soap which was hydrolysed. However, ethoxylated technical stearine is not so susceptible to saponification, and therefore so small admixture was sufficient for its efficient activity as the agglomerant.
Recapitulation of the optimisation procedures
A comparison of the results gained in three series of the experiments, with the compositions compositions denoted as AT1, AT2, and AT3; is presented in table 9, with the following shortenings for surface active agents: AO – oleic acid, AS – stearic acid, OD – 1-octadecanol, 1-octadecanol, E2 – di-stearate of ethylene glycol, and S2 – ethoxylated technical stearine. stearine. It is clearly visible in the comparison of table 9 that such a satisfactory balance among the surface active agents, selected in a given composition, is needed to optimise the deinking process, as well as to control the macro-stickies’ size and to strengthen hydrophobic properties of the macro-stickies particles; during the pulping step run under the laminar regime of flow. Such desirable effects depend
41
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
on a balance among those ingredients (i.e. surface active agents) of a given composition and are also determined by the very nature of individual ingredients. From theoretical point of view, it may be assumed that such an optimum balance between surface active agents (of a given composition) is also a token of reaching (during the pulping step) such a specific hydrophilic-lipophilic balance (HLB) between those surface active agents, ink particles, and the PSA particles. The studied compositions of agglomerants agglomerants with oleic acid cannot exclude, of course, the possibility of selecting other useful agents. Moreover, the optimisation studies were carried out for equal parts of the selected (model) pressure sensitive adhesives, such as water-based acrylic adhesive (E 115) and hot-melt based rubber adhesive (D 170), both by Avery Dennison Jackstädt; however, in industrial deinking lines a wide variety of the pressure sensitive adhesives adhesives may be present.
Denoting the components
A symbol of the optimisation experiments experiments and and the optimum optimum composition composition of surface active agents (%) in the experiment AT1
AT2
AT3
67 33 0 -
36 30 34 -
90 0 10
AO AS OD E2 S2 Properties
Approx. value of the property
<95%> conf. interval
Approx. value of the property
<95%> conf. interval
Approx. value of the property
<95%> conf. interval
σ45
2,45
2,1 – 2,8
2,24
1,0-3,4
2,18
1,8-2,5
µ45
-3,60
- 4,2 - -3,0
-4,07
-9,1 - 1,0
-4,52
-5,6 - -3,4
R457
58,5
50-60
63
62,5-64,7
63,7
62,5-64,9
k700
3,0
0,5-5,4
1,3
0-4
1,3
1,0-1,6
0,041
0,029-0,064
0,031
0,015-2,732
0,025
0,017-0,048
Θ+
µ45
e 2 [mm ]
Table 9: 9: A comparison comparison of the final results results gained in the optimisation optimisation studies for three different different (AT1, AT2, and AT3) compositions of surface active agents
In table 9 quite wide confidence intervals (95%) are shown for approximated values of the parameters (σ45 and µ45) of the dislocated lognormal distribution of the macro-stickies’ size. Probably this is a result of the influence of such factors which are difficult to be fully controlled, for example: properties of recovered paper; or which are uncontrollable, for example: creating deposits on the pulper surfaces. Nevertheless, the results have been gained in the pilot plant installation working under conditions similar to the conditions in industrial deinking line. And therefore, the very fact of selecting the significant factors, from (allow us to say) ‘a large field of unwanted noise’ in the pulping step, is raising a reasonable hope for gaining similar results at an industrial level. The optimisation studies, discussed above, show clearly that the attempts to keep up the good optical properties of deinked pulps need additionally such a physico-chemical action on the PSA particles which promotes their proper comminution and later their co-agglomeration with ink particles. The very occurrence of that co-agglomeration is connected with strengthening hydrophobic properties of the sticky particles, after reaching a satisfactory state of their comminution. The last one, regarding the macro-stickies, can be minutely characterised by the parameters of the dislocated lognormal distribution. distribution. The ICP optimisation studies studies show that a satisfactory level of the optical properties properties (R 457 > 2 60% and k700 < 2 m /kg) may be achieved just thanks to applying the agglomerant which promote the comminution process process of the sticky particles, decreasing decreasing the µ parameter of the macro-stickies’ size, in comparison with the formerly gained results. For easier interpretation of changes in values of the µ parameter the values are added, in the last line of table 9, which specify corresponding values of the 2 macro-stickies’ size, expressed in [mm ], determined as white ‘traces’ (according to INGEDE method n°4) with the computer-aided image analysis. Such a proper comminution of the PSA particles during the pulping step, promoted by agglomerants, has significance also for optical properties of deinked pulp.
42
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
This does mean, however, that by applying (in the pulping step) such carefully selected mixture of the deinking agent with the agglomerant we do influence not only an ability of ink particles and sticky particles to co-agglomerate together but also we do modify those sticky particles to become susceptible of comminution, and a possible detackification of the sticky particles could also be taken into account. Such effects of the agglomerant admixture are probably connected with such a high affinity of the properly selected agglomerants with the stickies, resulting in changes in physico-chemical properties of the stickies. Such a mechanism may be supported by results of the experiment in which the ink 0 flotation test was applied; however, the studied surface active agents (at temperature about 70 C) were used instead of writing ink, and the ADJ pressure sensitive paper labels (applied in our studies, with D 170 and E 115 adhesives) were tested instead of writing paper. Results of the penetration time (in seconds) are gathered in table 10.
Adhesive reference reference Oleic acid (AO) Stearic acid (AS) Lauric acid (AL) 1-octadecanol (OD) Di-stearate Di-stearat e of ethylene glycol (E2) Ethoxylated Ethoxylat ed technical stearine (S2)
D 170
E 115
395 215 37 195 16 12
322 260 205 315 40 320
Table 10: A comparison of the penetration penetration times (in seconds), seconds), during the the flotation tests (at 70+/-5 70+/-5 °C),
in which melted agglomerants penetrated through the PSA papers with different adhesives It is proved, in such a simple experiment, that the agglomerants are able to such a quick penetration through a structure of the adhesive layer. For some of them the complete penetration is accomplished in a very short time, and this is a proof of their close affinity with the studied adhesives. And therefore those agglomerants are able to modify properties of the adhesive layer, especially to decrease its strength, and also to increase hydrophobic properties of surfaces of the particles from the PSA layer. Such a controlling influence on rather chaotic way of the comminution of the PSA layers is additionally strengthened by the pulping run under the laminar regime of flow of the slurry in the pulper. An increase of the hydrophobic character of the macro-stickies’ surfaces was frequently observed in the ICP studies as almost complete blackening of surfaces of the sticky particles by ink particles, after pulping with agglomerants, instead of the fact that the ADJ pressure sensitive papers with the dyed adhesive layers (D 170 as yellow and E 115 as cyan) had been introduced to the pulper. Nevertheless, such a specific behaviour is not detected for pulping with oleic acid as the only surface active agent in the deinking bath; it means without other surface active agents applied as agglomerants in the pulping step, also run under the laminar regime of flow. It may be understood, therefore, that the sticky particles with pronounced hydrophobic properties of their surfaces, thanks to the agglomerants, are easily co-agglomerating with black ink particles. And they should be efficiently removed in the subsequent steps of the deinking process, therefore. However, questions arise about the methodology of determination of the macro-stickies. Probably in a part of the co-agglomerates with ink particles the sticky particles are loosing their tackiness and later they are not detected in the o INGEDE method n 4. Such effects may be strengthened by the phase transition of the agglomerant, from liquid to solid. Generally speaking, in such optimisation studies two independent methods should be applied for determining the macro-stickies, and especially in such a new pulping technology which is aimed at combining together the contaminants present in recovered papers, the sticky contaminants and non-sticky ones, such an additional method would be very useful. Nonetheless, there is a lack of o an alternative to the INGEDE method n 4.
Essential principles of the new pulping technology
According to the results gained in the ICP studies on the pulping step, carried out in a frame of the ScreenClean Project, the new pulping technology can be proposed for the deinking process of recovered papers. Such a new pulping technology for the deinking process is aimed at proper preparing of the pulp to two different processes, namely to de-inking and to de-sticking, by the
43
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
agglomeration and co-agglomeration processes among the sticky particles and the particles of other contaminants present in recovered paper, mainly ink and toner particles, during pulping of the recovered paper run under the laminar regime of flow. This results in enlarged and granulated particles of the agglomerates and co-agglomerates which could be easily removed in subsequent steps of the deinking process, especially in fine slot screening. To that end the process of comminution of the PSA layers producing the sticky particles (stickies), which is rather chaotic, is controlled during the pulping step not only by its laminar regime of flow but also by the presence of the properly selected surface active (SA) agents which additionally increase hydrophobic character of the stickies’ surfaces, required for their successful agglomeration and/or co-agglomeration with other contaminants. Nevertheless, those SA agents do not disturb the very process of ink detachment from recovered paper. It was experimentally proved, that the selection requirements are met by such SA agents which are insoluble in water, and their melting points are lower than temperature of the fibrous slurry at an initial phase of pulping in the pulper; however, their melting points are higher than temperatures of slurries in subsequent steps of the deinking process. Moreover, an affinity of those SA agents to the PSA materials, manifesting itself in adsorbing a liquid form of those agents by these PSA materials, is advantageous for due alterations in susceptibility of these PSA materials to the very comminution process, and later for shaping by those agents the susceptibility of the sticky particles created in that process to agglomerate together or to co-agglomerate with ink particles. The substances which met the requirements are called here the agglomerants of the stickies or the agglomerants in short. Necessary condition for the efficient performance of the agglomerants is applying them in an emulsified form which is improving their penetration through the fibrous slurry in the pulper into surfaces of the PSA materials. In the new pulping technology, preparing the fibrous slurry for its deinking and de-sticking too, those agglomerants are emulsified in different ways, depending also on pH of that slurry. In a case of the classic deinking process, in an alkaline medium, emulsifying of those agglomerants is done by introducing them, in a melted form, or advantageously in a form of their solution in fatty acid, e.g. in oleic acid, to deinking bath buffered with sodium silicate and containing alkaline agents, such as sodium hydroxide, with admixture of soaps or other emulsifiers commonly used in the deinking bath, or without such admixtures. In that new pulping technology, all components of the deinking bath are fed during filling the pulper with process water, before or during introducing recovered paper, having the temperature equal or higher, advantageously higher at least about 5 °C, regarding the melting point of an easiest melted agglomerant, and also equal or slightly higher, regarding the melting point of an individual agglomerant of that mixture with a highest melting point, advantageously higher no more o than about 2 – 5 C. Feeding those components during highly intensive mixing in the pulper is enhancing the emulsifying process of those agglomerants and their uniform distribution in the fibrous slurry as well. In a case of the deinking process performed in a neutral medium, the feeding of those components to the pulper is the same as in that former, alkaline, case; however, those components should be melted together with an emulsifier or they are introduced in a melted form to process water in which an emulsifier is dissolved. There is also possible to apply those agglomerants agglomerants uniformly mixed with other components of the deinking bath. Such mixture, containing a certain amount of water, advantageously above 65%, may be prepared in different forms (as flakes, powder, granulated product, or paste), able to be completely dissolved during a short time of filling the pulper with process water. Such products may be prepared in situ, in the proper installation being a part of the deinking line, or it may be supplied by other producer. That new pulping technology overcomes the difficulties encountered in the field of deinking which are connected with the PSA materials found in growing quantities in recovered papers. Applying of the new pulping technology will result in increasing consumption of such deinked pulps in the manufacture of graphic papers or tissue, with maintaining or increasing the paper quality. That new pulping technology may be applied without any additional costs, for a part of the mass of the soaps consumed in former technologies is efficiently replaced by the agglomerants agglomerants used in that new technology. Patent application on the new pulping technology is registered in the Polish Patent Office under number P th 372730 dated February 10 , 2005. Such compositions of the chemicals for deinking process with the new pulping technology, according to the claims of that patent application, were also registered under th brand name ‘De-Stick-Ink’ in the Polish Patent Office: number Z-289769 dated January 8 , 2005.
44
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Conclusion
In the ICP optimisation studies the goal of this work done in a frame of the ScreenClean Project has been achieved, i.e. such a new pulping technology of recovered papers in the pulper has been elaborated which enables to exert controlling influence on properties of the sticky particles of PSA materials (stickies), such as their size and shape, as well as the hydrophobic character of their surfaces, both improving separation of the stickies during subsequent steps of the deinking process. Additionally, the formulated hypotheses have been confirmed on mechanisms mechanisms of the comminution comminution process of the PSA layers and of the co-agglomeration of stickies with ink particles. Nevertheless, improving both the method of identifying the stickies and the manner of presenting the results gained in their analyses, as well as refining the ways of interpretation of those results – require careful consideration. The sticky particles become non tacky in a part of the co-agglomerates of the stickies with ink (and other contaminant) particles. There is a need to increase sensitivity of their identification o by the INGEDE method n 4 and/or to develop such an alternative method which would be able to identify the detackified stickies in their co-agglomerates. co-agglomerates. Dispersion degree and shape of the PSA particles, as well as their susceptibility to co-agglomeration with ink (and other) particles, are under the common control of the laminar flow of slurry in the pulper and the admixture of the agglomerants. In properly selected ratio among the agglomerants and oleic acid, commonly used in the deinking process, it is possible to reconcile the requirements of high optical properties of the deinked pulp with proper preparation of the macro-stickies, and perhaps stickies in general, for their successful separation in subsequent steps of the deinking process. During the optimisation studies, for each ternary composition of agglomerants with oleic acid, the optimum ratio among the ingredients was specified that optimises the overall desirability of the product. The results gained for the studied compositions provide a sufficiently solid basis for the industrial verification of the studies, as well as for further investigations into such directed way towards solving problems problems with the t he sticky particles in which the pulping step is efficiently used for both deinking and desticking of secondary pulps from recovered papers.
4.1.5. Mill trials During identification works and initial studies a detailed description of the Krapkowice deinking line was prepared and a detailed ‘photograph’ of the deinking process in the Krapkowice mill was made, the latter regarding separation of the macro-stickies from the deinked pulp. To prepare such a detailed ‘photograph’ three series of trials were conducted in the Krapkowice deinking line in which the samples taken from different steps of the deinking process were carefully analysed. It should be 2 pointed out that the absolute values of both factors of the macro-stickies’ content - S A [mm /kg] and o SN [N /kg] - were within the same range as observed in the ICP pilot plant experiments on pulping with admixture of the model pressure sensitive papers. Such a high level of the macro-stickies’ content is achieved in the Krapkowice mill instead of careful attempts to select recovered papers. In the studied series only fine shavings and printer’s waste were used, i.e., wood-free papers. Moreover, the usefulness of the dislocated lognormal distribution in studies on the macro-stickies was also confirmed in the attempts to prepare a detailed ‘photograph’ of the Krapkowice deinking line, regarding separation separation of the macro-stickies in that line. The parameters parameters of the macro-stickies’ size distribution - µ and σ - were within the range observed in the ICP pilot scale experiments, showing also such a high reproducibility. And therefore, the dislocated lognormal distribution of the macro-stickies’ size gave a possibility to show a striking similarity between the phenomena of creating particles of the macrostickies in that industrial deinking line and in the ICP pilot plant installation. In figure 27 the parameters parameters of the dislocated lognormal lognormal distribution of the macro-stickies’ macro-stickies’ size are specified and verified by means of the Probability-Probability Probability-Probability diagram for the results gained in the K3 trial at the Krapkowice mill. Such usefulness of the dislocated lognormal distribution of the macro-stickies’ size, approved also in the industrial scale, gave a possibility to modify the cleanliness efficiency E C defined by the TAPPI Standard. Using the computer calculations, it was easily to calculate the cleanliness efficiency E C with applying the density functions of the dislocated lognormal distribution of the macro-stickies’ size, 2 x [mm ].
45
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
1,0
0,8 n o i t c n u f n 0,6 o i t u b i r t s i d l a 0,4 c i r i p m E
0,2
0,0 0,0
0,2
0,4
0,6
0,8
1,0
Theoretical Theoretical distri bution function function
Figure 27: The P-P diagram for the results gained in the K3 trial at Krapkowice; the dislocated
lognormal lognormal distribution of the macro-stickies’ size: θ =0.014, =0.014, µ=-3.86679, σ =1.9580 =1.9580 According to t o the results gained in those trials, performed performed to prepare a ‘photograph’ ‘photograph’ of the deinking process in the Krapkowice deinking line, a lack of efficiency in de-sticking of the deinked pulp was stated in that process, in that line designed for de-inking and mainly for de-ashing of the secondary pulp. And therefore that deinking process was not giving an effective protection against migration of the macro-stickies to the process of manufacturing the final paper product (tissue) in paper machine. Additionally, in such trials the preliminary preliminary planned attempts to some changes in the industrial pulper construction were recognised as premature, and therefore further works were concentrated on modifying the pulping parameters in industrial scale towards the optimum conditions elaborated in the pilot scale, on the basis of their technological technological similarity. The mill trials with the new pulping technology were performed at the Krapkowice deinking line within certain limitations, according to the agreement with the Krapkowice mill. It was possible to modify the pulping conditions, adjusting them to the laminar regime of flow in the pulper, as well as to add the chemicals with the agglomerant; both according to the planned design of the experiments. However, further steps of the deinking process were run in the routine way to achieve the aimed goals of the entire process of tissue manufacture in the Krapkowice mill, and therefore the ICP activities were limited only to taking samples and collecting some measurement results. So the ICP role was ‘active’ only in the pulping step, and ‘passive’ in further steps of the deinking process.
A course of the pulping trials in the Krapkowice deinking line and the results gained in them
The new pulping technology of recovered paper, in the deinking process, is based on applying the emulsion of highly hydrophobic surface active agents (agglomerants) in the pulping step run under such flow conditions which are advantageous for the agglomeration and granulation processes between sticky particles and ink particles, i.e. under the laminar regime of flow in the pulper. Because the Krapkowice deinking line is not equipped with installations for dosing individual components of the deinking bath, such a paste-like form of the agglomerant and other components was most useful for applying in that line. Manufacturing of that paste for the trials was ordered to a domestic producer of soap, who is a member of the Global Pollena group. The product called De-Stick-Ink 4.5 was prepared, according to the ICP specification, in quantity amounted to 2.5 tons, and the Krapkowice mill was supplied with this product packed in buckets (10 kg each) for making easily dosing of this paste to the industrial pulper. However, there was a need to resign from using the agglomerant with melting point about 40°C and to substitute it with another agglomerant having melting point above 50°C. For achieving the aimed goals of the new pulping technology, an empirical adjustment of workable parameters of the industrial pulper to the kind of recovered paper was required. To that end the experiment design was elaborated according to the statistical plan called: Mixed 2 and 3 Level Design
46
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
- Designs for Factors at 2 and 3 Levels . The experiment design and the actual conditions, under which the industrial experiments on the new pulping technology have been done, are presented in table 11.
Regulated parameters e Planned values of the parameters e h t h t t s f s Codes values Values in physical units f n o r o e t t e e m e X2 X3 X1 X2 X3 s e X1 c i r e m n e a e p h r Stock Stock DSI u x t a q e f p cons. vol. conc. o e ° S N
[-]
[-]
[-]
3
[%]
[m ]
3
[kg/m ]
Resulting parameters Actual parameter values Values in physical units X1
X2
X3
T
Stock cons.
Stock vol.
DSI conc.
Stock temp.
[%]
[m ]
3
3
[kg/m ]
o
[ C]
PH
[-]
Pn
DPn
Net power cons.
Net power dissip.
[kW]
[kW/m ]
3
1
1
+1
+1
-1
17,1 17,1
20,5
4,5
18,23
23,7
4,1
43,0
7,54
321,3
13,6
2
4
+1
+1
+1
17,1
20,5
9,4
17,90
20,4
9,5
42,0
10,80
264,8
13,0
3
8
+1
+1
-1
17,1 17,1
20,5
4,5
18,11
20,2
4,8
44,0
7,20
283,4
14,1
4
4*
+1
+1
+1
17,1
20,5
9,4
17,95
20,2
9,7
44,0
10,60
246,0
12,2
5
8*
-1
-1
-1
13,9 13,9
17,9
4,5
14,96
17,7
5,3
44,0
7,60
224,9
12,7
6
2
-1
-1
+1
13,9
17,9
9,4
14,84
18,4
10,2
45,5
10,98
192,4
10,5
7
6
-1
+1
-1
13,9
20,5
4,5
12,45
20,9
4,4
46,5
7,16
198,5
9,5
8
2*
-1
0
+1
13,9
19,2
9,4
13,93
18,8 18,8
9,9
49,0
11,05 11,05
269,1
14,3
9
6*
-1
-1
-1
13,9 13,9
17,9
4,5
14,17
16,0
5,0
48,0
8,44
208,6
13,0
10
5
-1
-1
+1
13,9
17,9
9,4
15,49
14,8
10,9
48,0
10,12
196,3
13,3
11
1*
-1
0
-1
13,9 13,9
19,2
4,5
12,39
18,7
4,2
48,5
7,03
263,3
14,1
12
5*
-1
-1
+1
13,9
17,9
9,4
12,94
18,1
8,7
48,0
9,63
197,3 197,3
10,9
13
3
+1
+1
-1
17,1
20,5
4,5
15,55
20,5 20,5
4,0
48,5
8,80
247,2 247,2
12,1 12,1
14
7
+1
0
+1
17,1
19,2
9,4
16,94
19,2
8,6
48,5
11,22
251,0
13,1
15
3*
+1
0
-1
17,1
19,2
4,5
16,75
18,6
4,4
49,0
9,29
266,3
14,4
16
7*
+1
0
+1
17,1
19,2
9,4
15,47
19,9
8,1
49,0
11,03
290,7
14,6
Table 11: Experiment design and the actual conditions of the industrial pulping trials The following independent variables, i.e. the parameters being regulated, were selected: - Stock consistency, - Stock volume, (DSI). - Concentration of the De-Stick-Ink (DSI). Both the stock consistency and the DSI concentration were changed on two levels of the value of 3 these parameters, namely: stock consistency as 13% and 17%, and DSI concentration as 4.5 kg/m 3 and 9.4 kg/m , counting on the commercial product. But the stock volume was changed on three levels 3 3 3 of the value of this parameter, namely: 17.9 m , 19.2 m , and 20.5 m . Both stock consistency and stock volume were changed within the ranges covering the laminar regime of flow observed in the pulper. To describe the laminar regime of stock flow in the pulper in detail the power consumption was registered with the analyser installed specifically for purposes of the mill trials, and the factors derived from that power consumption were calculated. In table 11 the actual values are also given of such parameters which were not regulated, such as: - Stock temperature, temperature, - Stock pH, - Power consumption, - Power dissipation in the unit volume of the slurry. The above listed parameters could significantly influence the results gained in the industrial pulping trials. For example, it was planned that temperature of the slurry in the pulper would be equal to 50°C, i.e. close to the melting point of the agglomerant used in the De-Stick-Ink 4.5 composition. However, temperature of recovered paper was below 0°C (according to the measurements: minus 7°C). This resulted in the stock temperature which was lower than the planned one. Differences in stock pH measurements were found, also caused by the same factors which altered the stock consistency (inaccurate dosing of water and recovered paper). Values of the power consumption, gathered in table 11, are calculated as an average from the period between the feeding of the pulper (with recovered paper and water) and the beginning of its emptying. Values of the power dissipation are computed
47
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
from the average power consumption and actual values of the stock volume. The factorial statistical analyses and the regression analyses were made for the results gained in the trials regarding the power consumption. The results gained in these analyses show that both the net power consumption (table 12) and the net power dissipation in unit stock volume (table 13) are much more influenced by the stock volume(X2) in the pulper (the power of 2) than by the stock consistency (X 1). In those tables the statistically significant relationships are printed with the bold blue type; the same way is applied in following tables with such statistical analyses.
Fact Facto or Constant X2 (Sto Stock vol vol.) 2 X2 (Sto (Stock ck vol. vol.)) X1*X2
Resu Result lt 179 -62 134 134 77
s(e s(e) 16 29 30 23
t(1 t(13) p -95,% 95,% 11,47 0,0000 145 -2 -2,12 ,12 0,05 ,05 -126 4,52 4,52 0,00 0,0006 06 70 3,41 3,41 0,00 0,005 5 28
+95,% 95,% bi 212,6 178,9 1,2 1,2 -31,1 1,1 197, 197,7 7 66,9 66,9 125, 125,9 9 38,5 38,5
s(bi) 16 15 15 11
-95,% +95,% 145 213 -63 -63 1 35 99 14 63
Table 12: Statistical analyses of the relationship between between the net power consumption and
the pulping parameters (R^2= (0.76721, MS=438) Fact Facto or Cons Consta tant nt X2 (Sto (Stock ck vol. vol.)) 2 X2 (Stock (Stock vol.) vol.) X1*X2
Resu Result lt 9,69 9,69 -5,6 -5,6 6,6 3,5
s(e s(e) 0,69 0,69 1,3 1,3 1,3 1,0
t(1 t(13) p -95,% 95,% +95,% 95,% bi 14,0 14,00 0 0,00 0,0000 00 8,2 8,2 11,2 11,2 9,69 9,69 -4,32 -4,32 0,00 0,0008 08 -8,4 -8,4 -2,8 -2,8 -2,8 -2,82 2 5,02 5,02 0,0002 0,0002 3,8 9,4 3,29 3,29 3,51 0,004 1,4 5,7 1,76
s(bi) 0,69 0,69 0,65 0,65 0,66 0,66 0,50
-95,% +95,% 8,2 8,2 11,2 11,2 -4, -4,2 2 -1,4 -1,4 1,9 4,7 0,7 2,8
Table 13 : Statistical analyses of the relationship between between the net power dissipation and
the pulping parameters (R^2=0.66243, MS=0.863) Fully completed results gained during the studies of the pulping process in the industrial trials, giving the possibility of more detailed analyses, are presented in table 14. The columns with the pulping parameters, taken from table 11, are in table 14 supplemented with the results characterising the pulp after that process, such as its brightness and the parameters characterising characterising the macro-stickies. Stock consist.
Stock volume
pH
Net power consump.
Net power dissip.
[%]
[m ]
[kg/m ]
R457
S A
SN
µ
σ
[ C]
[-]
[kW]
[kW/m ]
[%]
[mm /kg]
[no [no/kg] /kg]
[-] [-]
[-] [-]
18,23
23,7
4,1
43,0
7,54
321,3
13,6
65,4
9331
37440
-3,3685
2,2956
17,90
20,4
9,5
42,0
10,80
264,8
13,0
64,1
8452
21120
-3,6049
2,5304
18,11
20,2
4,8
44,0
7,20
283,4
14,1
59,5
9993
35040
-3,6285
2,3668
17,95
20,2
9,7
44,0
10,60
246,0
12,2
65,0
10306
50640
-3,1674
2,1313
14,96
17,7
5,3
44,0
7,60
224,9
12,7
69,6
1006
9480
-4,5762
2,2739
3
DSI concentr.
3
Stock temp.
o
3
Pulp brightness
Macro-stickies’ content
2
Parameters of the dislocated lognormal distribution of the macro-stickies’ macro-stickies’ size
14,84
18,4
10,2
45,5
10,98
192,4 192,4
10,5
59,9
15933
45600
-2,9265
2,2883
12,45
20,9
4,4
46,5
7,16
198,5
9,5
63,5
14007
29520
-3,3651
2,4391
13,93
18,8
9,9
49,0
11,05
269,1
14,3
66,2
11010
57000
-3,4760
2,2052
14,17
16,0
5,0
48,0
8,44
208,6
13,0
76,6
2661
8040
-3,8805
2,7957
15,49
14,8
10,9
48,0
10,12
196,3
13,3
62,5
6606
20040
-3,0263
2,2307
12,39
18,7
4,2
48,5
7,03
263,3
14,1
56,4
12305
25680
-3,8488
2,2413
12,94
18,1
8,7
48,0
9,63
197,3
10,9
59,6
6998
27300
-4,0507
2,3745
15,55
20,5
4,0
48,5
8,80
247,2
12,1
65,7
23248
55080
-2,8817
2,5997
16,94
19,2
8,6
48,5
11,22
251,0
13,1
63,4
4557
18060
-3,7452
2,2977
16,75
18,6
4,4
49,0
9,29
266,3
14,4
63,6
6536
18480
-3,6901
2,5761
15,47
19,9
8,1
49,0
11,03
290,7
14,6
64,8
4921
21840
-3,2359
2,3009
16,60
21,8
8,8
48,5
11,25
290,4
13,3
63,3
14643
33120
-2,5874
2,4504
Table 14: 14: Parameters of the pulping process and properties of the pulp after that process, its brightness and the parameters parameters characterising the macro-stickies present in that pulp
48
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
An influence of the new pulping technology on properties of the produced pulp, regarding the macro-stickies present in the pulp and the pulp brightness
Assessing the absolute values of the results gained it the pulping trials, it should be kept in mind that the mill trials were run under different conditions than the optimal ones elaborated in the pilot scale; regarding regarding especially the following issues: -
In the mill trials the water temperature was lower than the melting point of the agglomerant; moreover, there was no possible to apply the agglomerant with melting point about 40°C, selected in the optimisation studies.
-
Pulping time in the mill trials was about 50% shorter than in the pilot scale; by sheer coincidence, during the trials the deinking line had to supply the pulp for two paper machines.
-
Different adhesives were present in recovered papers processed during the mill trials. Large quantities of bookbinding hot-melt adhesives were detected, and the pressure sensitive adhesives materials on plastic foil base prevailed.
Instead of the above obstacles, such very positive results were gained in the pulping trials, consciously run according to the new pulping technology, regarding both pulp brightness and shaping such properties of the macro-stickies present in the pulp which are required for successful stickies removal in subsequent steps of the deinking process. The searching of relationships between the two groups of data, i.e. parameters of the pulping process and properties of the pulp after that process, shows the statistically significant influence (at p=0.05) on the scale parameter µ of the dislocated lognormal distribution of the macro-stickies’ size by two parameters of the pulping process, namely stock volume (X2), related to the power dissipation during the pulping run under the laminar regime of flow, as well as of the De-Stick-Ink concentration concentration (X 3); see table 15.
Fact Facto or Cons Consta tant nt X2 (Stock vol.) X3 (DSI conc.) X2*X3
Resu Result lt -3,5 -3,54 4 0,57 0,46 -0,3 0,35
s(e s(e) 0,10 0,10 0,24 0,20 0,24 ,24
t(1 t(13) -35, -35,38 38 2,38 2,28 -1,4 1,48
p 0,00 0,00 0,03 0,04 0,16 ,16
-95,% 95,% +95,% 95,% bi -3,8 -3,8 -3,3 -3,3 -3,5 -3,54 4 0,1 1,1 0,28 0,0 0,9 0,23 -0,9 0,9 0,2 0,2 -0,1 -0,18 8
s(bi) 0,10 0,10 0,12 0,10 0,1 0,12
-95,% +95,% -3,8 -3,8 -3,3 -3,3 0,0 0,5 0,0 0,4 -0,4 0,1 0,1
Table 15: Results of statistical analyses of the interrelationships between the µ parameter of the
macro-stickies’ size distribution and parameters of the pulping process (R^2= 0.44974, MS=0.162) The results, which are presented in table 15, have been received thanks to eliminating not so significant factors, according to the Pareto chart illustrated in figure 28. In the Pareto chart the factors are ordered in sequence of their significance in shaping values of the µ parameter of the macrostickies’ size distribution. The data illustrated in this Pareto chart are evidence that concentration of the (DSI), denoted X 3, is the most significant for alteration in the µ parameter. De-Stick-Ink (DSI),
x3
1,90
x2
1,10
x2*x3
-0,46
x2*x2
-0,36
x1*x3
x1
x1*x2
-0,25
0,11
-0,06
p=,05 Standardized Effects (Absolute Value)
Figure 28: Pareto chart of the standardized effects of the studied pulping parameters
(and their interactions) on shaping the µ parameter
49
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
A lack of statistically significant influence of the stock consistency (X1) should be pointed out. After limiting the number of factors to first three factors only, the Pareto chart is transformed into the version illustrated in figure 29.
x2
2,38
x3
2,28
x2*x3
-1,48
p=,05 Standardized Effects(Absolute Value)
Figure 29: Pareto chart of the standardized effects of the studied pulping parameters
(and their interactions) on shaping the µ parameter; only for first three factors In that version of the Pareto chart (see figure 29) a sequence of the factors was changed; however, an influence of the DSI concentration (X 3) was still statistically significant for shaping the µ parameter. The significance of the DSI concentration (X 3) and the stock volume in the pulper (X 2) for alteration in the µ parameter should be understood as the serious justification of the elaborated principles of the new pulping technology for the deinking process. The same may be said about the results of the statistical analyses of such interrelationships interrelationships in shaping σ parameter, presented in table 16. However, in this case the lower significance level (p=0.1) was applied which is also acceptable for industrial trails. Among the studied parameters, only the DSI concentration (X 3) was statistically significant for shaping the σ parameter. This may be perceived also in figure 30, presenting the Pareto chart for first three factors having an influence on the σ parameter. Fact Facto or Consta Constant nt X1 (Stock (Stock conc.) conc.) X2 (Stock vol.) X3 (DSI conc.)
Resu Result lt 2,384 2,384 0,10 -0,10 -0,155 -0,155
s(e s(e) 0,041 0,041 0,12 0,14 0,084
t(1 t(13) 57,89 57,89 0,90 -0,70 -1,83
p 0,0 0,4 0,5 0,1
-95,% 95,% 2,31 2,31 -0,1 -0,3 -0,30
+95,% 95,% 2,46 2,46 0,3 0,2 -0,01
bi 2,384 2,384 0,052 0,052 -0,050 -0,050 -0,077 -0,077
s(bi) 0,041 0,041 0,058 0,058 0,071 0,042
-95,% 2,31 2,31 -0,05 -0,05 -0,17 -0,15
+95,% 2,46 2,46 0,16 0,08 -0,00
Table 16: Results of statistical analyses of the interrelationships between the σ parameter parameter of the
macro-stickies’ size distribution and parameters of the pulping process (R^2=0.21836, MS=0.0279) MS=0.0279)
X3
-1,83
X1
X2
0,90
-0,70
p=,1 Standardized Effects (Absolute Value)
Figure 30: Pareto chart of the standardized effects of the studied pulping parameters
on shaping the σ parameter, parameter, only for three main factors 50
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Such statistical analyses showed a lack of any significant relationship between the macro-stickies’ 2 o content - S A [mm /kg] and SN [N /kg] – in pulps received according to the new pulping technology and the pulping parameters in the mill trials. Probably this was also a result of the random quantities of adhesives in the following batches of recovered papers introduced to the pulper. Additional and and such a very convincing argument argument for the new pulping pulping technology, technology, which is approving its principles, was presented by the results gained in statistical analyses of the interrelationships interrelationships between the pulp brightness after the pulping process and parameters of that process, which are gathered in table 17. They were received after gradual eliminating not so significant factors, according to the Pareto chart illustrated in figure 31. Fact Facto or Cons Consta tant nt X2*X3
Resu Result lt 64,3 64,36 6 5,2
s(e s(e) 0,98 0,98 2,30
t(1 t(13) 65,8 65,82 2 2,23
p 0,00 0,00 0,04
-95,% 95,% +95,% 95,% bi 62,3 62,3 66,4 66,4 64,3 64,36 6 0,2 10,2 2,6
s(bi) 0,98 0,98 1,2
-95,% +95,% 62,3 62,3 66,4 66,4 0,1 5,1
Table 16: Results of statistical analyses of the interrelationships between the pulp brightness after the
pulping process process (R 457 ) and parameters parameters of the pulping pulping process (R^2=0. (R^2=0. 24901, 24901, MS=15.93) 457
X2*X3
2,10
X2*X2
-1,50
X3
-1,22
X*X2
X1
X2
X1*X3
-1,14
0,96
0,92
-0,92
p=,05 Standardized Effects (Absolute Value)
Figure 31: Pareto chart of the standardized effects of the studied pulping parameters (and their interactions) on shaping the pulp brightness (R 457 ) after the pulping pulping process
As it is shown both in table 17 and in figure 31, the most significant for shaping the pulp brightness is an interaction between the stock volume in the pulper (which is related to the power dissipation) and the DSI concentration (X2 * X3). Such interaction, so advantageous for brightness of the pulp after the pulping, may be understood as the result of the agglomeration between the ink particles and probably also their co-agglomeration with sticky particles, during the pulping step run according to the new pulping technology which is properly preparing such contaminant particles to processes of their agglomeration and co-agglomeration. So the pulping trials at the Krapkowice deinking line proved the principles of the new pulping technology which enables to exert controlling influence on comminution of the adhesive layers to shape properties of the minute particles (stickies) created from such layers, especially the stickies’ size and the hydrophobic character of the stickies’ surfaces. Applying both the agglomerant and the laminar regime of flow during the pulping step, according to the new pulping technology, resulted in an increase of the stickies’ size, of their agglomerates, and of such their co-agglomerates with ink particles which were sufficiently tacky to be detected according to the INGEDE method n°4. Such effects are perceived in so valuable alteration both in pulp brightness and in parameters of the dislocated lognormal lognormal distribution of the macro-stickies’ macro-stickies’ size. The latter was perceived as an increase of the µ parameter (meaning the particle size increase) and a decrease of the σ parameter (meaning the narrower range of the particle size). The statistically significant influence on pulp brightness and on both parameters µ and σ by two parameters of the pulping process, namely stock volume (X 2) and the De-Stick-Ink concentration (X 3) or their interaction - was found in the pulping trials. Such two parameters of the pulping step are characteristic feature of the new pulping technology as the stock
51
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
volume in the pulping trials was related to the power dissipation during the pulping run under the laminar regime of flow, and the DSI concentration was measure of the agglomerant concentration. Moreover, the pulping trials proved usefulness of the parameters ( µ and σ) of the dislocated lognormal distribution of the macro-stickies’ size for characterising population of the macro-stickies and their agglomerates, as well as such their co-agglomerates with ink particles which are detected by the INGEDE method n°4. It should be pointed out that running of the pulping of recovered papers according to the new pulping technology have not caused any problems in industrial scale. And therefore the pulping trials at the Krapkowice deinking line proved that the pulping step may be consciously and without any difficulties run according to the new pulping technology to prepare the pulp in such a way which should enable not only the successful deinking of that pulp but also its efficient de-sticking during subsequent subsequent steps of the deinking process, especially in fine slot screening.
Separation of the macro-stickies in the deinking process after the pulping trials at the Krapkowice deinking line
To examine any possible influence of shaping by the new pulping technology such properties of the macro-stickies which are important for their efficient removal in the subsequent steps, there was a need to observe their fates in the entire deinking process. To that end more detailed information about flow intensities of fibrous slurry and process water was required. According to measurements of consistency and flow intensity in the selected places of that deinking line, the balances of slurry flows were elaborated (see figure 32), presented in a form of the Sankey diagrams, separately for each of three trials performed in the mill trials, and denoted: ‘De-Inking’, ‘De-Stick-Ink 1’, and ‘De-Stick-Ink 2’. In the trial denoted as ‘De-Inking’ (DI) the pulping of recovered paper was performed in the routine way practiced in the Krapkowice deinking line. However, in the trials denoted as ‘De-Stick-Ink 1’ (DSI 1) and ‘De-Stick-Ink 2’ (DSI 2) the pulping of recovered paper was run according to the experiment design presented in table 11. And therefore three Sankey diagrams in figure 32, – also denoted: DI, DSI 1, and DSI 2 – are showing a course of further processing of the pulps gained in each trial at the pulping step. Samples for elaborating the DSI 1 data were taken after introducing to the deinking process the pulps from first seven batches of the pulping step run according to the experiment design (see table 11), and samples for elaborating the DSI 2 data were taken after introducing to the deinking process the pulp th from the 15 batch of the pulping step run according to the experiment design.
Figure 32: Sankey diagrams with flow rates in the t he Krapkowice deinking line, separately shown for three mill trials, and denoted: ‘De-Inking’ (DI), ‘De-Stick-Ink 1’ (DSI 1), and ‘De-Stick-Ink 2’ (DSI 2)
52
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Once again it should be pointed out that the ICP attempts to evaluate further steps of the deinking process, which followed the pulping step, were limited only to registration of the actual parameters of these further steps, without, however, any attempts to modify their levels; according to the agreement with the Krapkowice mill. So the further steps of the deinking process were run in the routine way to achieve the aimed goals of the entire process of tissue manufacture, and this also clarifies the differences in some stream rates perceived in the Sankey diagrams presented in figure 32. The differences were taken into account in further analyses of the macro-stickies’ behaviour in different steps of the deinking process, during the mill trials. Moreover, as it was formerly stated, large quantities of bookbinding adhesives were detected in recovered papers processed in the Krapkowice deinking line during the mill trials. The bookbinding hot-melt adhesives are applied in liquid form (after their melting) in binding processes. After cooling down such adhesives transform into solid state loosing their tackiness. In such solid state, however, layers of the hot-melt bookbinding adhesives present in recovered paper may survive the pulping step thanks to their high cohesive strength. Nonetheless, comminution of the layers or larger particles of the bookbinding hot-melt adhesives may progress after the pulping step, in the subsequent step of the deinking process. Solid particles of the bookbinding hot-melt adhesives are non-sticky, nevertheless, under conditions applied in the INGEDE method n°4 such particles are detected as the macro-stickies. Most likely the minute particles of the bookbinding hot-melt adhesives present in deinked pulp can cause troubles during the process of making paper from the deinked pulp contaminated by such particles, especially at elevated temperature resulting in their stickiness, so they should be removed from the deinked pulp. Regarding the mill trials at the Krapkowice deinking line, however, such additional portion of the nonsticky particles (but detected as macro-stickies according to the INGEDE method n°4), not present in the pulp just after the pulping step but introduced to the pulp stream in the subsequent step of the deinking process, caused serious difficulties in evaluating an influence of the pulping step run according to the new pulping technology on the separation efficiency of the sticky particles in subsequent steps of the Krapkowice deinking line. For this reason, the issue of such specific stickies needed more thorough discussion. To examine the question closely, more detailed analyses were made regarding possible sources of the macro-stickies present in the pulp directed to the ADS 7 separator. In figure 33 the theoretical frequency distribution functions of the macro-stickies’ size are additionally shown for the stickies present in the pulp after the pulper diluted with process water (indexed as ‘+ wo’), and in that diluted pulp supplemented with the reject of light trash from the ADS 7 separator (indexed as ‘+ wo + rl’). Such additional frequency distribution functions were calculated from the stream balances and the macro-stickies’ contents in appropriate streams of the pulp. In some cases, however, analyses of some streams were consciously omitted, according to a growing need to limit number of the samples. And therefore the data characterising characterising the stream indexed ‘rh’ were applied instead of the data characterising a stream of the accepted pulp from the Diabolo DF3 separator. Both such specified theoretical frequency distribution functions are located much below the distribution functions for the pulps just before the ADS 7 separator (denoted as S NA); see figure 33. This may be understood that an increase of the macro-stickies’ content in that pulp stream directed to the ADS 7 separator is caused by comminution activities of the separators, such as the Poire separator and both Diabolo separators (DF3 and DT2), all of them working in similar way but being different in size.
53
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
De-Inking
De-Stick-Ink c k-Ink 1
De-Stick-Ink c k-Ink 2
1,2E5
1E5
SNA[mm2/kg]=14850 80000
] g k /
(x-0,014;-3,6969;2,6585) 6585) SNA =36180*ILognorm(x-0,014;-3,6969;2,
=109620*ILognorm(x-0,014;-4,8854;2,2704) 8 854;2,2704) SNA =109620*ILognorm(x-0,014;-4,
(x-0,014;-3,2973;2,0767) 0767) SNH =18540*ILognorm(x-0,014;-3,2973;2,
=29500*ILognorm(x-0,014;-3,3651;2, 3651;2,4391) 4 391) SNH =29500*ILognorm(x-0,014;-3,
(x-0,014;-4,7742;2,5424) 5 424) SNA =27060*ILognorm(x-0,014;-4,7742;2, (x-0,014;-3,6901;2,5761) 5 761) SNH =18480*ILognorm(x-0,014;-3,6901;2,
o
N 60000 [ N S
40000 SNA[mm2/kg]=10290
SNA[mm2/kg]=4130
+ wo + rl + wo
20000
SNH [mm2 /k /kg]=14 g]=1401 010 0
SNH [mm2 /k /kg]=35 g]=3540 40
0 0 ,0
0 ,1
0,2
0 ,3 X [mm2 ]
0,4
SNH [mm2 /k /kg]=65 g]=6536 36
0 ,5 0,0
0,1
0 ,2
0 ,3 X [mm2 ]
0 ,4
0 ,5 0 ,0
0 ,1
0,2
0 ,3
0 ,4
0 ,5
X [mm2 ]
Figure 33 : A comparison of the frequency distribution functions of the macro-stickies’ size in the pulp
after the pulper (S NH – in blue ) and in the pulp just before the ADS 7 separator (S NA - in red ), for the t he studied variants DI, DSI 1, and DSI 2. Experimentally determined total surface areas of the macrostickies, in [mm2 /kg], are also presented. The theoretical distribution functions of the macro-stickies present in the pulp from the pulper pulper diluted with process water, water, indexed as ‘+ wo’ (in green ), and in that diluted pulp supplemented with the rejects of light trash (from the ADS 7 separator), indexed as ‘+wo+rl’ (in violet ) - are added Neglecting a contribution made by each one of the separators, it was more significant to know what had been divided in them into minute particles which later were detected as macro-stickies in the INGEDE method n°4. During detailed inspections of the rejects from the separators considerable amounts of the PSA plastic foils were found in both rejects. In the reject from the Poire separator also plastic bands and plastic labels were found, with quite well survived adhesive layers; as it had been observed in the ICP pilot plant experiments with the PSA on plastic foils. However, the rejects from the Diabolo DF3 separator contained additionally considerable quantities of larger pieces derived from the bookbinding hot-melt adhesives used to combine together the spines of the folded paper sheets. Such larger pieces of the bookbinding hot-melts maintained almost the same size and shape as in the recovered paper introduced to the pulper; see the photograph shown in figure 34.
Figure 34 : A comparison between the piece of the bookbinding hot-melt adhesive, found in the reject
from the DF3 Diabolo separator, and the form of that adhesive combining together a spine of the folded paper sheets, present in recovered papers processed during the mill trials
54
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Most likely during the mechanical treatment in the DF3 separator a part of larger fragments of the bookbinding hot-melts is divided into minute particles which are turned back to the pulp stream. This additional additional admixture of such stickies, being non-sticky particles in the deinking line but detected as the macro-stickies by the INGEDE method n°4, results in a considerable increase of the macro-stickies’ content (denoted as S NA) in the pulp stream directed to the ADS 7 separator. It may be assumed therefore that such increase of the macro-stickies’ content in the pulp just before the screening step is caused by the mechanical comminution of the binding hot-melts in the DF3 separator. This posed such a serious obstacle to the detailed evaluation of an influence of the new pulping technology on the separation efficiency efficiency of the macro-stickies (derived from the PSA materials) in subsequent steps of the deinking process. Moreover, Moreover, it must be pointed out that such stickies st ickies (from the bookbinding hot-melts) were generated outside of the pulper, so without due interaction with the agglomerant added to the pulping step, according to the new pulping technology. Nevertheless, it was possible to detect some positive influences of the new pulping technology on separation of the sticky particles in subsequent steps of the Krapkowice deinking line. An extrusion of the flat-shaped macro-stickies through slots of the slotted screen manifested itself after the routine pulping (variant DI), resulting in higher maximal size of the macro-stickies (in the accepts after 2 2 screening in the ADS 7 separator), equal to 0.4 mm , in comparison with 0.32 mm , characteristic of the results gained after the new pulping technology technology (variants DSI 1 and DSI 2). This does show such a positive aspect of the new pulping technology in which, thanks to the processes of agglomeration and probably also co-agglomeration, the granulation of the sticky particles is progressing, creating oblong particles (granules) less susceptible to their comminution, as well as not proper for their extrusion through slots of the slotted screen. This is perceived in the blue lines (depicting the pulps accepted after screening in the ADS 7 separator) which are shorter for the variants run according to the new pulping technology (DSI 1 and DSI 2), in comparison with the variant run in routine way (DI); see figure 35.
De-Inking
De-Stick-Ink 1
De-Stick-Ink 2
400000
SNa=148320*ILognorm(x-0,014;-5,440;2,179) SNi=109620*ILognorm(x-0,014;-4,8854;2, (x-0,014;-4,8854;2,2704) 2704)
350000
SNrh=403140*ILognorm(x-0,014;-4, =403140*ILognorm(x-0,014;-4,1263;2,6079) 1 263;2,6079) SNrl= 52260*ILognorm(x-0,014;-4,4032;2,4838)
[mm m2/kg]=339490 SAr Arh h [m 300000
[mm m2/kg]=270640 SAr Arh h [m (x-0,014;-4,5513;2,2493) 2 493) SNa =60420*ILognorm(x-0,014;-4,5513;2,
250000
SNa=45360*ILognorm(x-0,014;-3,6208;2, (x-0,014;-3,6208;2,5287) 5287)
SNi =27060*ILognorm(x-0,014;-4,7742;2, (x-0,014;-4,7742;2,5424) 5424)
SNi=36180*ILognorm(x-0,014;-3, =36180*ILognorm(x-0,014;-3,6969;2,6585) 6 969;2,6585)
=412220*ILognorm(x-0,014;-4,169;2, 169;2,6194) 6 194) SNrh=412220*ILognorm(x-0,014;-4,
SNrh=491580*ILognorm(x-0,014;-4, =491580*ILognorm(x-0,014;-4,2482;2, 2482;2,6800) 6 800)
(x-0,014;-4,6033;2,6044) 6 044) SNrl =31080*ILognorm(x-0,014;-4,6033;2,
SNrl=47280*ILognorm(x-0,014;-3,6089;2, (x-0,014;-3,6089;2,5356) 5356)
] g k / o
N 200000 [ N
S
150000
SArh[mm2/kg]=338150
[mm m2/kg]=13640 SAa [m
[mm m2/kg]=14850 SAi [m 100000
[mm m2/kg]=11660 SAa [m
SAa [mm2/kg]=11600
[mm m2/kg]=8460 SArl [m
[mm [m m2/kg]=12256
SArl
50000
SArl [mm2/kg]=7610
[mm m2/kg]=10290 SAi [m 0
0,05
0,15
0,25 2
x [mm ]
[mm m2/kg]=4130 SAi [m 0,35
0,05
0,15
0,25 2
x [mm ]
0,35
0,05
0,15
0,25
0,35
2
x [mm ]
Figure 35 : A comparison between the frequency distributions of the macro-stickies in the pulps
processed in the ADS 7 separator, during the mill trials for the variants denoted: DI, DSI 1, and DSI 2. Macro-stickies in inlet pulps are denoted S Ai – in green green,, in accepted pulp as S Aa – in blue blue,, and in rejected pulps as S Arh (in red ) and as S Arl (in brown ). Experimentally determined total surface areas of the macro-stickies, in [mm2 /kg], are added
55
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Some positive results of the new pulping technology, regarding separation of the macro-stickies from the stock were found in the first stage of hydrocyclones, during the cleaning step. This is perceived in the violet lines depicting the cleanliness efficiency E C; see figure 36. De-Inking
De-Stick-Ink 1
De-Stick-Ink 2 100
240000
100% 90
220000
EC
[mm m2/kg]=54950 SAr [m
EC
EC
200000
80
80% 80% 180000
SNa=29220*ILognorm(x-0,014;-4, =29220*ILognorm(x-0,014;-4,8771;2,4805) 8 771;2,4805)
SNa=20940*ILognorm(x-0,014;-3, =20940*ILognorm(x-0,014;-3,3268;2,5463) 3 268;2,5463)
SNi=57360*ILognorm(x-0,014;-3,5531;2, (x-0,014;-3,5531;2,4639) 4 639)
SNi=57060*ILognorm(x-0,014;-3, =57060*ILognorm(x-0,014;-3,5766;2,45314) 5 766;2,45314)
70
SNr =151500*ILognorm(x-0,014;-3, =151500*ILognorm(x-0,014;-3,8331;2, 8331;2,534) 5 34)
SNr =141300*ILognorm(x-0,014;-4, =141300*ILognorm(x-0,014;-4,4464;2, 4464;2,4652) 4 652) 160000
60
SAr [mm2/kg]=19840
140000
[mm m2/kg]=34050 SAr [m
] g k /
o
N [
50
] % [ C
E
N120000
S
=40500*ILognorm(x-0,014;-4,1136;2,6475) 1 136;2,6475) SNa =40500*ILognorm(x-0,014;-4, =72120*ILognorm(x-0,014;-3,722;2, 722;2,4055) 4 055) SNi =72120*ILognorm(x-0,014;-3, =264540*ILognorm(x-0,014;-4,5316;2, 5316;2,6163) 6163) SNr =264540*ILognorm(x-0,014;-4,
100000
40
80000
[mm m2/kg]=16240 SAi [m
60000
30
[mm m2/kg]=14960 SAi [m
[mm m2/kg]=13130 SAi [m [mm m2/kg]=8660 SAa [m
40000
20
[mm m2/kg]=5940 SAa [m 10
[mm m2/kg]=4070 SAa [m
20000
0
0
0,05
0,15
0,25
0,35 2
X [mm ]
0,45
0,05
0,15
0,25
0,35 2
X [mm ]
0,45
0,05
0,15
0,25
0,35
0,45
2
X [mm ]
Figure 36 : A comparison between the frequency distributions of the macro-stickies in the pulps
processed in the first stage of the cleaning in hydrocyclones, during the mill trials for the variants denoted: DI, DSI 1, and DSI 2. Macro-stickies in inlet pulps are denoted S Ai – in green green,, in accepted pulp as S Aa – in blue blue,, and in rejected pulps as S Ar - in red . Experimentally determined total surface areas of the macro-stickies, in [mm2 /kg], are added. The graphs are supplemented supplemented with the cleanliness efficiency E C C – in violet According to the the data presented presented in figure 36, 36, the first stage of the cleaning cleaning in hydrocyclones hydrocyclones is such an efficient way to remove the macro-stickies from the deinked pulp. The cleanliness efficiency is above 80% within the entire range of the studied macro-stickies’ size. After stock preparation in the pulping step according to the new pulping technology (variants DSI 1 and DSI 2), however, that cleanliness efficiency was additionally improved to 85%, regarding separation of the macro-stickies smaller than 2 0.05 mm from the stock. However, there was a lack of possibility to study the separation of the macro-stickies from the stock in further stages of the cleaning step owing to their malfunction during the mill trials. According to the detailed detailed analyses analyses of the results gained gained in the mill trials, it may be concluded concluded that that such combination of deinking and de-sticking processes, in which the new pulping technology is applied, requires the properly adjusted management of accepts and rejects in some key points of the deinking line to separate efficiently also the macro-stickies, successfully using their modification made during the pulping step run according to the new pulping technology. In the Krapkowice deinking line such a satisfactory solution towards resolve the stickies problem could be the directing of the accepted pulp from the DF3 Diabolo separator to the pulper via the water pre-heater; however, the accepted pulp from the DT2 Diabolo separator should be introduced to the water circuit just before clarifying of process water in the micro-flotation step. And therefore analyses of the flow balances of the macrostickies in the entire deinking process process are needed, as well as the control system of the macro-stickies’ content and properties should be established in the key points of the process, to put into practice the new pulping technology. The mill trials have approved the technical possibilities of that new pulping technology to conscious shaping the macro-stickies’ properties in such a way which is advantageous for their further separation from the stock, in subsequent step of the deinking process.
56
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.1.6. Conclusions and perspectives Besides the comparative pilot pulping trials, which did not allow to conclude definitely about the advantages of the drum pulper compared to the batch pulper in terms of stickies fragmentation and subsequent removal with slotted screens, most of the research effort has been placed on the development of a new pulping technology by ICP, from the lab and pilot phases until the industrial scale tests in the Krapkowice deinking deinking mill. Indeed, some identification works and initial studies at the Krapkowice deinking line proved a lack of efficiency in de-sticking of the deinked pulp, because (like in other cases) both deinking lines and deinking technologies are designed to remove ink and toner particles, also excess of fillers, but they are usually not suitable for efficient separation from the stock such specific contaminants as the sticky particles (stickies) are. However, also in industrial practice the usefulness of the parameters ( µ and σ) of the dislocated lognormal distribution of the macro-stickies’ size for characterising populations of stickies was approved. This enabled calculating of the cleanliness efficiency with applying the density functions of that distribution of the macro-stickies’ size, for evaluating their separation in subsequent steps of the deinking process. The pulping trials on stock preparation in the pulper, at the Krapkowice deinking line, clearly demonstrated technical possibilities of the new pulping technology in exerting the controlling influence on properties of the sticky particles and on the detachment of ink particles, which manifested itself in increasing the µ parameter (an increase of the stickies’ size) and decreasing the σ parameter (narrower range of the stickies’ size), also in improving the pulp brightness after the pulping step. Such positive results were connected, in statistically significant way, with the parameters characteristic of the new pulping technology, related to the agglomerant concentration and to the power dissipation during the pulping step run under laminar regime of flow in the pulper. Moreover, the pulping trials at the Krapkowice deinking were consciously and successfully run according to the new pulping technology, technology, without any difficulties. In the second part of the mill trials at the Krapkowice deinking line, run in the routine way and without any possible alteration in the course of subsequent steps, the obstacles were identified in the way of exploiting fully the modified (by the new pulping technology) properties of contaminant particles in processes of their further separation from the stock. It was shown that the pulping step run according to the new pulping technology must be followed by the modified management of selected accepts and rejects in some steps of the deinking process, to satisfy both deinking of the stock and its de-sticking. And therefore to exploit fully such positive effects of the new pulping technology in a given deinking line, the systematic analysis of the stickies’ problem in that deinking line is required, applying the methods elaborated in the ScreenClean project. After such analyses the strategy should be worked out for applying the new pulping technology to the stickies’ problem abatement. Implementation of that strategy should be successful in existing deinking lines; however, further improvements may be expected in such modified deinking systems, consciously oriented towards removing not only ink particles (or toner and filler) but sticky particles too, according to the new pulping technology. And there is a need to cooperate with suppliers of the deinking line equipment, also with producers of the chemical additives for paper-making. A conceptual framework of such further activities to put into practice the new pulping technology directed towards both deinking and de-sticking, elaborated in the ScreenClean project, could be proposed by the coordinator of that project.
57
L SCREEN C CLEAN EAN
4.2.
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Pressure screening
4.2.1. Background and objectives Pressure screening is the best available technology to remove macro-stickies and contaminants. Fibres pass the slots while contraries should be retained on the screen plate. Screening is currently performed in deinking mills with slots down to 0.10 or 0.12 mm for respectively wood-free (MOW) and wood-containing (ONP/OMG) deinked pulp [31, 63-64]. Fine screening systems in European DIP lines are typically operated with 0.15 mm wedge wire slots in the low-consistency range [2]. The particular behaviour of stickies in screens is characterized by their ability to be extruded through the slots by the pressure generated by the rotor and the pulp flow, which is a consequence of the visco-elastic properties of the adhesive material. Soft deformable stickies particles which can be extruded through slots should be considered as “probability” contaminants (normally particles with at least one dimension smaller than the slot width) rather than “barrier” contaminants (normally particles with all dimensions larger than the slot width). The behaviour of such particles in pressure screens can be characterised, on average, by a particle passage ratio, which is defined by the ratio of the downstream to the upstream particle concentration in a screen cylinder section, according to the probability screening theory [11-15]. Basic studies were carried out over the last decade at several research institutes, as reviewed in [4], to develop the understanding of the screening process at the scales of both pulp “macro-flow” conditions around the rotor and unsteady “micro-flow” conditions and particle separation phenomena at the surface of the screen plate and through the slots (figure 37). Typically with foil rotors, the pressure pulse and duration are respectively about 5 to 50 kPa over 10 to 30 ms for the positive pressure pulse and 50 to 200 kPa over 1 to 5 ms for the negative pressure pulse.
Vp = Ave Average rage Vs+ Vs-
F
V x
P+
A Vs-
R
3 ms
Vs+
P-
20 ms
Figure 37: Schematic of the pressure screen (left) and pressure variations at the screen plate (right)
Investigations at CTP in this field were first performed with relevant probability contaminants, i.e. 2 flat-shaped particles (0.5 mm films) and long-shaped particles (shives) with lower thickness than the slot width. A major conclusion about these studies [4] was that all the parameters which improved the passage of the fibres through the slots, i.e. increasing slot width and the effective normalised slot velocity, also increased the passage of the tested probability contaminants. Increasing the passing velocity (calculated from the accept flow rate), reducing the slot friction factor with wedge wire design and increasing the intensity and duration of the positive pressure pulse created by the rotor were assumed to increase the effective slot velocity during the screening phase and thus to increase the probability for particles to be captured in the fluid exit layer and then to pass the slot. The effective normalised slot velocity was an extended definition of the normalised slot velocity [16], i.e. the ratio of the effective slot velocity to the tangential velocity at the surface of the screen cylinder. Increasing the thickness of the exit fluid layer taken from the flow at the top of the profiles and turning into the slot was assumed to enhance its ability to drive particles towards the slot inlet. Numerical simulation showed that this fluid layer thickness was roughly proportional to the normalised slot velocity and to the slot width [17]. The effects observed when changing rotor velocity were attributed to changes induced by the tangential fluid velocity in the exit layer thickness [4].
58
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Practically, probability probability screening efficiency, i.e. the selectivity of the separation between fibres and thin contaminants showed to be very difficult to improve by changing slot velocity and design or rotor velocity and design, at given profile design. Clear improvement of the screening efficiency was only achieved by reducing the height of the profiles, which was attributed to particle contacts with the inclined profile walls assumed to reject more efficiently large and stiff contaminants compared to flexible fibres. Indeed, the hydrodynamic screening mechanisms revealed through the investigations with thin flat and long-shaped particles should also apply to stickies at least until the particle has reached the slot inlet. Then the extrusion process should become decisive. The objectives in the research programme devoted to screening in this project were more particularly: 1.
to develop the understanding of the very complex hydrodynamic screening and stickies sticki es extrusion phenomena through numerical simulation,
2. to improve the the stickies removal removal efficiency, on the basis of pilot pilot low-consistency low-consistency fine-slot screening screening tests and with special focus on the optimisation of screen plate design, 3. to develop a model for the simulation of screening screening systems in mills, 4. and to investigate, investigate, on pilot pilot scale, new and and conventional conventional high-consistency high-consistency screening screening technology technology in order to remove stickies as early as possible in the deinking line. The numerical simulation work has been performed at LEGI while all the other tasks, including lab studies about the extrusion of stickies in cooperation with LEGI/ITM, have been performed at CTP. AFT was essentially essentially involved in the the pilot screening screening tests, providing providing test screen screen plates and and knowledge. knowledge.
4.2.2. Numerical simulation studies The principle of operation of pressure screens is to produce overpressure at the surface of cylindrical + + basket and so-called "feed" flow during the overpressure (P and flow through the slots at velocity Vs in figure 37) and to enforce the passage of paper fibres through very fine slots (100-200 µm wide) and to reject the contaminants, which are generally much larger in size. To avoid screen slots plugging the intensive negative pressure pulses are produced by means of a rotor equipped with either foils or blades. The experimental practice indicates, however, that soft and viscoelastic PSA particles are able to pass the screen slots even if they are 3-4 times larger than the slot width [65-66]. The effectiveness of the separation process is dependent on several factors among which the design of the screen plate and rotor shape are the most important ones. During recent years, a variety of different screen slot geometries were proposed, e.g. [18, 67], and introduced into industrial practice. The screening mechanism is still, however, far from fully understood and must be investigated using different methods. Since experimental studies are rather limited, the numerical modelling of the flow can provide the complementary information, which could be used for the optimisation of the process. The present research study was aimed at providing the insight into the behaviour (deformation and possible extrusion) of "sticky" particles at the screen slot in order to identify and quantify the parameters promoting the particle passage or rejection during screening process. In particular a number of parameters was tested characterising the flow kinematics in the screen, screen profile geometry as well as stickies size and material properties. properties.
4.2.2. 4.2.2.1. 1.
Numerical mod el of press ure scr eening
The optimisation of the contaminants screening process in pressure screen cannot be done without the use of powerful numerical methods for flow simulation (CFD - computational fluid dynamics). In order to built the relevant numerical model of the process the real features of the physical phenomena should be taken into consideration, consideration, i.e.: -
multi-phase character (water, paper fibres, contaminants),
-
three dimensional complex geometry with moving elements (rotor, screen plate surface),
-
turbulence,
-
rheological rheological (viscoelastic) properties of adhesive materials,
-
flow unsteadiness. unsteadiness.
59
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
As the numerical numerical modelling, modelling, taking into account all the features mentioned above, seems to be extremely challenging, it was decided to apply a simplified, the so-called decoupled approach, which means that the entire screening process was split into two steps: -
-
during the former step single-phase (water) flow was modelled (by means of Fluent commercial software) with the contaminants (treated as the rigid bodies) assumed to be at rest and located at the inlet to the slot, the latter step was aimed at deformation analysis of viscoelastic particles and was performed performed with the use of Ansys commercial code.
In order to couple both simulation steps the CFD solution was used as the boundary condition for the deformation analysis. Such a solution strategy allowed to apply simplified partial models and as a consequence to avoid problems related to the implementation of more complex (i.e. multiphase) physical models.
Numerical model of the flow
According to previous extensive studies of screening process [17-18] the flow in the pressure screen can be divided into two flow regions (figure 37): -
macro-flow - associated to the rotor, which can be assumed to be almost "solid rotation" motion (except for the close neighbourhood of rotor blades/foils),
-
micro-flow - in the vicinity of the screen surface where radial (slot) velocity component is of great importance.
It was shown in [18] that the macro-flow can be substituted with the relevant boundary conditions making possible the consideration of only micro-flow what in turn allows for huge reduction of the extent of computational domain. In the present study, the following additional assumptions were made: -
no curvature of the screen basket is taken into consideration,
-
flow is characterised only by centrifugal (tangential) and radial (slot) velocity components V t and Vs, respectively, with neglected axial velocity component,
-
pure water with no contaminants and no paper fibres is assumed as the flowing medium,
-
no flow unsteadiness unsteadiness resulting from foil passage is considered.
According to the simplifications above the two-dimensional model of the flow could be defined as a single waveform (the close vicinity of a single slot) with periodic boundaries ("inlet" and "outlet") as shown in figure 38. The flow is enforced by the prescribed, constant (time-independent) velocity components at the "feed" edge (simulating the macro-flow) in both tangential and slot directions. The tangential velocity component V t was assumed to be related to linear velocity of passing foils [18] while the slot velocity component Vs,f results from the "accepted" flow rate (passing the slot). There were no prescribed flow parameters at periodic boundaries.
Figure 38: Computational domain formulation
60
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Screen profile design
During the present simulation two different screen profile designs were considered in order to compare their performance performance in terms of stickies rejection: -
Microvortex Cobra (Kadant Lamort),
-
Wedge Wire (Advanced Fiber Technologies)
The geometry and the most important dimensions of both profile types are shown in figure 39. In order to make the comparison possible the slot width was kept constant and equal to w sl = 100µm what was also the case for t he experimental trials. The radius r c of the slot inlet edge should also be pointed out as it differs significantly for both designs, i.e. for WW profiles its value is appr. 200 µm while for MV-like type is much smaller and according to laboratory measurements it lies within the range 10-100 µm. For the present simulations value of 50 µm was chosen as the most representative. representative.
b
a
wsl = 100 µm
θ = 15° l = 0.8mm (r c)MV = 20÷90µm (r c)WW = 200µm
Figure 39: Screen profile designs analysed: (a) MicroVortex (MV) and (b) Wedge Wire (WW)
Sticky's geometry model
To start up the simulation, the proper choice of the particle model should have been made first. It was done on the basis of the experimental work [65-66] devoted to the study on the passage of "sticky" particles through the slot. The authors considered acrylate-based PSA particles, which were produced during laboratory pulping, performed, however, under standard industrial conditions. The image analysis of the particles allowed to find that the "sticky" particles population is dominated by the so-called "string-like" particles [66], for which one dimension (termed as length) was one order of magnitude greater than the two others (termed as width and thickness) with their average values equal to: -
length: 2mm
-
width: 0.3mm
-
thickness: 0.3mm
The above data allowed to define the particle model geometry as a cylinder enabling both for 2D and 3D simulations.
Figure 40: 3D computational geometry and particle orientation
With respect to real screening conditions, the following simplifications were made during the CFD simulation:
61
domain
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
-
particle becomes at rest at the inlet to the slot at the starting moment - the case corresponds to the experimental experimental conditions of [65-66] as well as to the t he experiment reported reported in section 4.2.3.1.
-
particle is positioned as shown in figure 40 with its axis of symmetry parallel to the slot inlet edges, which is the most-extrusion-promoting most-extrusion-promoting orientation,
-
particle is regarded as a rigid body, i.e. there is no particle deformation and in turn no feedback on the flow pattern.
Constitutive model of PSA material
According to the available technical literature, e.g. [65-66], and experimental experimental tests reported in section 4.2.3., pressure sensitive adhesives reveal viscoelastic material properties. As the experimental trials aimed at selection of the relevant viscoelastic model and determination of its parameters had not been completed before simulation work was started, the generalised Maxwell model has been chosen to represent the behaviour of PSA materials. The mechanical analog of generalised Maxwell model is represented by N Maxwell elements (dashpot simulating viscous behaviour and spring acting as an elastic element connected in series) and is shown schematically in figure 41. The constitutive equation relating stress τ •
and strain rate γ (dot over symbol denotes time differentiation) is given by the following convolution integral: t
•
τ(t ) = ∫ G(t − t' ) γ (t' )dt' 0
The material properties G (shear or bulk moduli) are expressed in integral form using the kernel function of the Maxwell elements as: 1
2
3
G( t ) =
N
∑G
j
exp( −t / λ j ) + G ∞
j =1
where λ j = η j/G j is the relaxation time and η j stands for viscosity. Except for the Maxwell elements the model includes an additional spring preventing the material against unlimited deformation (characterised by G ∞).
Figure 41 : The mechanical representation of
generalised Maxwell model
In the present numerical investigations the single-element Maxwell model was applied with the following set of parameters: parameters: -
shear modulus G = 0.07MPa
-
shear modulus at infinite time G ∞ = G / 10
-
relaxation time λ = 1s
which were estimated from the elongation tests described in section 3.3.1.
Numerical tools
The simulations were performed with the use of well recognised commercial software packages, i.e.: -
FLUENT - the CFD code with its pre-processor GAMBIT for geometry and mesh generation, ANSYS - software for structural structural mechanics analysis. analysis.
62
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The CFD simulation requires a high-quality numerical mesh which means that the cell size should be fine enough to match the smallest flow scales. In order to satisfy that requirement requirement the multi-step mesh generation and refinement procedure was applied (for details see report D3) which allowed to obtain mesh-independen mesh-independentt solution. The characteristics characteristics of the CFD solver used: -
finite-volume (FV) schemes for the discretization of the flow governing equations,
-
k-ε RNG turbulence model,
-
two-layer zonal model for near-wall treatment.
The second part of the simulation devoted to particle deformation was performed with the use of ANSYS, version 5.7 with the following following solver configuration: configuration: -
finite-element finite-element (FE) method for the discretisation of equations,
-
"sparse direct solver" used.
All the details of numerical numerical procedure procedure (constraints definition, contact definition, transient loading following the progressing deformation, etc.) can be found in the progress report D3.
4.2.2. 4.2.2.2. 2.
Numeric al flo w sim ulati on
Two-dimensional flow simulation - Single-profile domain
As the first f irst step the simulations for simplified two-dimensional case were carried out with no particle present in the computational domain. The computations were aimed at delivering the general information about the flow pattern and providing an insight into the physics of the screening process. As the flow kinematics in the pressure screen is of great importance for the screening efficiency the two "governing" velocities have been changed at the following levels: -
average slot velocity Vs = 1m/s, 4 m/s, 10m/s,
-
tangential velocity at "feed" edge (see figure 38) V t = 5m/s, 10 m/s, 20m/s
and the corresponding velocity ratio α = V s / Vt = 0.05 ÷ 2, which covers completely typical industrial conditions.
a
b stagnation point
Figure 42: Flow-field in the vicinity of the screen plate for MV-like design: velocity vectors (a)
and static pressure distribution (b) The general flow pattern for the velocity ratio α = 0.4 (Vs = 4m/s, Vt = 10m/s) is shown in figure 42a. Velocity vectors clearly reveal the presence of permanent cavity vortex rotating in clockwise direction and occupying nearly whole room between consecutive screen profiles. The vortex shape and dimensions dimensions as well as its kinematical characteristics characteristics are significantly affected by velocity ratio (see D3 for details). Besides the velocity field, the pressure distribution is of great importance in the present study, especially in terms of the analysis of large contaminants contaminants extrusion into the slot.
63
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
As it is shown in figure 42b, the neighbourhood neighbourhood of the slot inlet is characterised characterised by huge pressure gradients with the maximum pressure at stagnation point and the minimum (not considering the slot) at the vortex centre. The pressure again increases at the inlet to the slot, i.e. in the region of possible particles deformation. deformation. It should also be noted that the flow-field just above the profile crest starts to be uniform. As far as the screening efficiency is concerned concerned the so-called exit fluid layer thickness should be regarded as one of the key parameters [17]. That layer is defined as the flow region between two streamlines crossing the stagnation points (the points "A" and "B" in figure 43) and seems to be responsible for driving suspended particles (contaminants (contaminants as well as the fibres) into the slot. The simulations allowed to determine the exit layer thickness in the interesting regions, regions, i.e. over the profile crest and "under" the vortex as a function of velocity ratio α. More results characterising the vortical flow pattern in the screen vicinity can be found in the report D3.
B
A
Figure 43 : Streamline pattern in the vicinity of the slot inlet
Two-dimensional flow simulation - Multi-profile domain
As the second step the simulations simulations were conducted for the case case with particle located at the inlet inlet to the slot. As for 2D configuration it resulted in plugging the slot, the so-called multi-profile domain had to be considered, i.e. computational domain was extended to several profiles to enable the flow as shown in figure 44. The simulations were focused on the analysis of pressure distribution at the particle surface and allowed to find out that pressure does not depend on: -
particle size (the following radius values were analysed: r p = 60, 100, 140 µm),
-
circumferential circumferential position,
-
number of profiles (plugging intensity) providing that it is normalised by dynamic pressure inside the slot (based on the average slot velocity Vs).
slot plugged by the
particle
Figure 44: Geometry for the 2D free (unplugged) slots
multi-profile configuration
The above findings allow to treat pressure acting on "stickies" as the function of screening kinematics (slot and tangential velocities) only. Moreover, the correlations were found between pressure distributions for single- and multi-profile configurations allowing to use the results (pressure loss coefficients) of the former case to deduce pressures for the latter one.
64
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Three-dimensional flow simulation
The extension of the previous simulations to three-dimensional case required to define the particle length l and computational domain extent (depth d) along the "z" axis (see figure 39). The following values were chosen: -
computational domain depth d = 4mm,
-
three particle length values l = 0.8, 1.0 and 1.33mm,
-
with corresponding reduced particle lengths l/d = 0.2, 0.25 and 0.33.
a
b
Figure 45: Velocity vector field in the slot vicinity (a) Pressure distribution at the particle particle surface (b) The simulations allowed to find out that the flow-field in the screen for 3D configuration is in principle qualitatively similar to the corresponding ones obtained for 2D cases. As it can be easily noticed from velocity vector maps shown in figures 42a and 45a (both simulations performed for the same velocity ratio α=0.4) the flow within the screen cavity is dominated by the permanent vortex of the similar shape and dimensions. More details about the flow field pattern, especially 3D effects and the flow structure inside the slot as a function of particle length, can be found in the progress report D3. The main goal of flow simulation in the screen was to deliver necessary information about the pressure load at the sticky particle as it was further required for the deformation analysis. The general qualitative view of static pressure distribution at particle surface is presented presented in figure 45b. As it can be clearly seen the pressure distribution at side (cylindrical) surface can be divided into 2 regions: -
"top" - being exposed to rotating cavity vortex (marked in red),
-
"bottom" - corresponding to wake inside the slot (marked in blue)
both of nearly uniform pressure distribution and separated by contact edges between particle and slot inlet (yellow and green). The more detailed analysis of the results for different particle lengths revealed that indeed pressure at particle surface can be treated as a two-level function as the maximum scatter of pressure does not exceed 1% of average value. Qualitative pressure uniformity does not concern the cylinder "caps" (not shown in figure 45b) but it seems to be of minor importance as far as the particle deformation is concerned. Pressure uniformity at particle surface is of great practical importance as it allows to regard the mechanical loading as two-dimensional what enables in turn to simplify the deformation analysis and particle extrusion into the slot, leading to huge reduction of computational effort.
65
L SCREEN C CLEAN EAN
4.2.2. 4.2.2.3. 3.
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Part Part icl e defor matio n analysi s
The second step of the "decoupled" simulation of stickies screening process was aimed at delivering information about the behaviour of deformable (viscoelastic) PSA particles at the slot inlet and provide the passage/rejection statistics. The following parameters were identified as the influential ones on the deformation deformation and possible particle extrusion: -
screening pressure - particle load,
-
particle size,
-
screen profile geometry - Microvortex versus Wedge Wire,
-
material parameters parameters (elasticity, viscosity),
-
friction.
The simulations of particle deformation were conducted with the use of a number of simplifications among which the following assumptions are the most important: -
particle becomes in rest before pressure load is applied - inertia force is neglected,
-
no unsteadiness of loading is taken into account,
-
progressing deformation/extrusion does not influence the flow-field (pressure distribution),
-
particle axis is assumed to be parallel to the slot inlet edges, i.e. the particle orientation is the most-extrusion-promoting one, leading to critical extrusion parameters
Pressure load
The screening pressure level was assumed to be the most influential parameter on the stickies rejection efficiency. The results of simulations conducted for the MV-like profile are presented in figure 46. Due to rheological properties of adhesives the sticky particles are subjected to temporal load, so the extrusion time is also an important resulting parameter. As it can be seen from figure 46 the extrusion times grow rapidly for decreasing pressures and for certain critical values (depending on particle diameter) reach infinity what corresponds to no particle passage. Such a viscoelastic behaviour could be possible due to non-zero value of shear modulus at infinity G ∞. Two other observations should be given: -
-
particle size (diameter d normalised by slot width wsl) significantly influences the extrusion time, the durations of extrusion time are unexpectedly long when compared to typical duration of positive pressure pulse (10-30ms) in screens.
∞ 100
d/w sl = 2 d/w sl = 3
80
Microvortex G = 0.07MPa λ = 1s 1s
60
textr [s] 40
20
0 0
20
40
60
80
1 00
p [kPa] Figure 46: Effect of pressure on particle extrusion time
Especially the latter observation seems to be extremely important. In other words the single pressure pulse is not able to cause the passage of sticky particle and a number of pulses has to contribute to the "successful" extrusion. There is also a possibility that the reverse pressure pulse (shorter but more intensive) may flush out the particle. These preliminary results give a new qualitative contribution to the stickies passage mechanism and may suggest it as a multi-step process. It should be pointed out that the observation about long particle extrusion times was verified by the experimental studies reported in section 4.2.3.1.
66
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Screen profile design - Microvortex versus Wedge Wire
The influence of the screen profile design on the PSA particles passage was recognised as one of the most interesting point of the simulation. The comparative study was performed for MV and WW profiles for standard set of material properties, screening pressure p = 10kPa and particle diameter d = 200µm, i.e. double the slot width.
p=0 t=0
p=10kPa t=10s
p=10kPa t=20s
p=10kPa t=21s
load (pressure + time) → Figure 47: Influence of screen profile type on the evolution of particle deformation and
extrusion time: Microvortex (upper sequence) and Wedge Wire (lower sequence) Figure 47 presents the selected images of progressing deformation for both geometries and corresponding load configuration. It should be noted that during simulations much more intermediate deformation stages have been determined due to high nonlinear character of the phenomenon. As it can be easily deduced from the presented results Wedge Wire screen profile design promotes the extrusion of stickies. The exact (with respect to the temporal resolution of performed simulations) extrusion times for both profiles were: (textr )MV = 38.2s and (t extr )WW = 20.7s
100
Microvortex Wedge wire
80
60
textr [s]
p = 10kPa G = 0.07MPa λ = 1s
40
20
d / wsl 0 0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
Figure 48: The influence d of /particle wsl size on extrusion time for two screen profile designs
Nearly 2 times shorter extrusion time characterising the Wedge Wire profile is the most likely caused by its "roundness". Deformation of the particle during progressing extrusion is relatively uniform in entire particle volume, whereas MV design leads to "accumulation of deformation" in limited particle regions. Physically the particle squeezing into Microvortex profile is subjected to higher stresses, what results in higher material resistance. The non-uniformity of particle deformation in case of MV profile is especially evident for longer times (see figure 47).
67
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The comparison of MV and WW screen profiles performance performance in terms of stickies rejection efficiency is presented in figure 48 as a function of particle size. The already done observation about the rejectionpromoting shape of Micovortex profile is confirmed for all particle diameters analysed. On average the extrusion time characterising characterising the MV profile is roughly twice the Wedge wire one (textr )MV ≈ 2 (textr )WW That results should not be, however, unexpected as the Wedge Wire screen plates are recognised as providing reduced flow resistance and in turn enabling for increased paper pulp capacity. It seems that profile shape revealing low curvature leads to increased both paper fibres (carried by water) and contaminants passage ratios.
Slot inlet edge curvature
The conclusion resulting from MV - WW profiles comparison suggested that slot inlet edges curvature may have significant influence on stickies rejection efficiency. According to the results of laboratory analysis of screen profile prints the slot inlets of industrial MV screen plates are characterised with significantly different curvatures as the edge radius has been found to vary in the range: r c = 10÷60 (100) µm It should be noted that such a scatter of r c values mainly results from the technology of making slots which are either water jet milled or laser cut. In order to quantify the influence of slot inlet edge curvature on particle behaviour the simulations were conducted for different radius values keeping standard set of rheological properties. Figure 49 shows the evolution of deformed particle shapes (particle diameter twice the slot width) loaded with screening pressure p = 10kPa and running time for two radius values, i.e. r c = 20µm and 90 µm, which are nearly extreme with reference to standard value of 50µm. The images clearly indicate the importance of inlet edge curvature and its influence on significant variations of local stresses, which are related to the deformation. It should be noted that reduction of radius value from 90 µm down to 20 µm resulted in increased extrusion time by the factor exceeding 3.
t=0 p=0
t=0 p=10kPa
t=2s p=10kPa
t=5s p=10kPa
t=10s p=10kPa
t=20s p=10kPa
r c=20µm
r c=90µm
load (pressure + time) → Figure 49: Influence of slot inlet edge curvature on the evolution of particle deformation:
sharp edge (upper sequence) and rounded edge (lower sequence)
68
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
In order to generalise the above observation the calculations were also conducted for standard radius value r c= 50µm and for the larger particle size, i.e. d/wsl = 3. The extrusion time variation as a function of both parameters is presented in figure 50. It can be easily seen that the smaller the edge radius the greater the influence on extrusion time.
250
MV profile p = 10kPa
200
d/w sl = 2 d/w sl = 3
150
textr [s]
For r c values smaller than 50 µm the deformation concentrates in much more limited area what in turn leads to increased stresses and longer times to be relaxed.
textr [s]
100
50
0 0
20
40
60
80
100
m] rr c [[ µµm] Figure 50 : Extrusion time vs radius of slot inlet edge
Material PSA parameters
The simulations were completed with the analysis of the influence of PSA material properties properties on sticky particle deformation in the screen slot. As the generalised Maxwell model with single element was employed to simulate viscoelastic behaviour the shear modulus G and relaxation time λ were tested. Additionally the friction factor has also been taken into consideration although its value for the system steel-adhesive material was not known. Applying for the simulations standard set of key parameters (screening (screening pressure, particle size, slot inlet edge radius etc.) the observations below could be done. The increase of shear modulus leads to nearly linear growth of extrusion time, which for certain G level goes to infinity (no passage of particle occurs). It can be explained with non-zero shear modulus at infinite time G ∞ preventing the infinite elongation (feature of Maxwell model). The linear dependence of extruding time upon relaxation time in log-log coordinating system was found. The lower stickies rejection efficiency of Wedge Wire profiles was confirmed also in this case. The elastic-type friction was modelled (dynamic friction coefficient increases from zero in continuous way to the "working" level following realistic behaviour) at three levels of dynamic friction coefficient: -
cf = 0 (corresponding to no friction),
-
cf = 0.33 and 0.66 (covering typical range for steel-rubber system).
With the friction switched on the dependence of extrusion time versus particle size takes more pronounced non-linear character then with no friction included (see figure 48). For the greatest value of friction coefficient the screen plate resistance against extruding stickies particles can be increased at least twice.
4.2.2.4.
Conclusions
The simulation of sticky particle behaviour at the screen plate was conducted using decoupled approach, i.e. the flow field was firstly resolved using single-phase model (CFD simulation of water flow with particle treated as a rigid body) which was then followed by the analysis of particle deformation deformation with the load taken from the solution of preceding preceding simulation step. The CFD computations computations delivered the detailed flow-field description of the flow in the close vicinity of the screen plate (microflow) as a function of process key parameters (screen kinematics, sticky particle and profile geometries). As the most important conclusion of the CFD simulation the uniform pressure distribution at particle surface was found allowing for simplified two-dimensional two-dimensional analysis of sticky deformation.
69
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The simulation of PSA particle behaviour in the screen slot concerned viscoelastic constitutive model of adhesive material and particle load consisting of screening pressure and progressing time. The numerical tests took into consideration a number of influential parameters characterising screening process and sticky contaminants allowing to quantify their influence on particle passage through the slot. The following parameters turned out to oppose particle extrusion:
low screening pressure (optimal screen flow kinematics),
particle "hardness" - high elastic/shear moduli values,
long stress relaxation time (less pronounced pronounced viscous behaviour), behaviour),
screen profile design - MV-like more resistant than WW-likeslot inlet edge sharpness,
high friction coefficient.
The simulation was performed using the default values of material and friction parameters as they were not known in relevant time. The obtained results reveal, however, the consistency with experimental experimental observations made during visualisation tests being conducted simultaneously. The model used in current simulation assumed a number of simplifications, which should not, however, significantly influence the expected outcomes of the work. As the most important and influential simplification, simplification, the assumption about steady flow conditions was made which means that the motion of a rotor with foils passing periodically close to the screen plate surface was neglected. As the extrusion times are for most of cases analysed much longer (approximately 2-3 orders of magnitude) than pressure pulse period, the adequate particle deformation phenomenon should be simulated under unsteady pressure load.
4.2.3. Optimisation of stickies screening The extrusion of stickies particles through slots has been investigated experimentally in order to check the numerical simulation results. The studies were performed at CTP in cooperation with ITM/LEGI in the framework of two-month training periods of students from Technical University of Czestochowa. Stickies extrusion was studied on lab scale, through visualisation, first under steady pressure and later under unsteady pressure. Pilot plant screening tests were performed at CTP during the same period, in order to investigate the effects of key screening parameters parameters (related or not to stickies extrusion) and to optimise screen plate design in cooperation with AFT. The main results were published [69-71]. The optimisation of high-consistency screening has been studied at CTP on pilot scale, in cooperation with AFT for the construction construction of a rotary screen screen cylinder for for the experimental experimental pilot equipment. equipment.
4.2.3. 4.2.3.1. 1.
Stick ies extru sio n
Studies about experimental investigations on the passage of acrylic PSA particles through a fine slot in a single slot laboratory device using vacuum to extrude the particle through the slot have been reported at the beginning of the project [65-66]. The particle passage probability was shown to increase as the temperature and pressure drop increased and as the particle width, thickness or area decreased. Particle size and shape were controlled by changing the pulping conditions which produced mainly string-like particles. Slight changes in particle size and shape were observed after the extrusion of the particles through the slot. Long-shaped particles were considered as the most relevant to study the extrusion of mill PSA stickies. Indeed, flat-shaped stickies with lower thickness than the slot width (down to 20 or 30 µm, the thickness of wet adhesive films) are not relevant while spherical particles are typically found after dispersing, especially kneading, i.e. in the second deinking loop after fine screening. Consequently it was decided to use the cylinder as the model particle shape in the extrusion studies. Special rolls of adhesive films between two silicone release papers were used to manufacture cylindrical adhesive particles of different diameters (figure 51) with the two reference adhesives: adhesives: -
the water-based acrylic adhesive (ref E115)
-
the hot-melt based rubber adhesive (ref D170)
70
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Adhesive cylinders (rolled (rolled up water-based acrylic adhesive adhesive film) 0.9 mm diameter 1.5 mm diameter
Figure 51: Adhesive particles manufactured for the tests
The adhesive cylinders were measured and tested after soaking about one hour, which is particularly important with the acrylic adhesive which showed to absorb about 50% water.
Stickies extrusion under steady pressure
Experimental set up
The single slot channel screen (figure 52) was designed to simulate the flow conditions observed at the surface of screen plates and through the slots during the positive pressure pulse and to change and control easily slot width and design. The height of the transparent channel was 10 mm and the width 15 mm. The syringe shown in figure 52 was not used in the test procedure described later. It was intended to introduce stickies particles at the right level to drive them toward the slot inlet.
Figure 52: Single slot channel screen with camera A high-speed high-speed CCD camera was used at 100 or 200 i/s to visualise the particles in transmitted light, using continuous light source behind the transparent channel. Resolution was 2 or 8 µm depending on the magnification of the optics. Visualisation was limited to the extrusion of stickies through the slot, as a consequence of the shape of the PSA particles and procedure used for the tests. A procedure was defined to start the visualisation of the extrusion of the adhesive cylinder as it was correctly placed at the slot inlet and a continuous pressure was applied. Most of the tests were performed only to determine the pressure required to extrude the particle as a function of the adhesive and slot characteristics. Images were recorded under certain conditions to analyse deformation of the particles and to determine the extrusion time. Some tests were also performed at extrusion pressures far above the usual range of positive pressure pulses observed in screens, i.e. typically 20 to 100 kPa.
Effect of adhesive material parameters
A first test series was performed performed with 0.5 mm slots with “wedge wire like design” design” to study the adhesive material parameters. New slots with more accurate design were used later to study the effects of slot design parameters. Figure 53 shows the pressure required to extrude the adhesive particle versus the normalised particle size defined as the ratio (d/w) of the cylinder diameter to the slot width, for the reference adhesives at room temperature (22°C) and at typical temperature in deinking circuits (50°C). The extrusion pressure was significantly lower with the acrylic adhesive (E115) than with the hot-melt rubber adhesive (D170) at both temperatures, which suggests that the acrylic adhesive material is softer and that acrylic stickies should be more difficult to screen as larger particles are extruded through the slots for a given pressure. Increasing the temperature reduced the extrusion pressure,
71
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
especially with the acrylic adhesive, which suggests higher softening of the adhesive material. It is interesting to note that the extrusion pressure is roughly proportional to the particle “oversize” (d-w) though the extrusion phenomena is probably strongly non-linear. The results in figure 53 indicate that acrylic adhesive particles with 2 to 3 times larger diameter than the slot width were extruded through the slots in the temperature and extrusion pressure range of mill screening systems. 250
D170 22°C D170 50°C E115 22°C E115 50°C
200
] a P 150 k [ e r u s s 100 e r P
D170
E115
50
0.5 mm "WW" Slot 0 0
1
2
3
4
5
Normalised particle size (d/w)
Figure 53: Effects of adhesive material and temperature temperature Extrusion tests were performed after soaking the adhesive particles in cold and hot water at pH 10, in order to investigate the effect of caustic soda added to the pulper in deinking mills. The results showed a large decrease of the extrusion pressure when the acrylic adhesive particles were soaked at room temperature with caustic soda. At high temperature the decrease of the extrusion pressure was lower as soaking pH was increased. These results suggested that the penetration of water and/or the dissolution of adhesive components, assumed to be responsible for softening the adhesive particle, required higher pH at low temperature [69]. Soaking caustic soda produced the opposite effect with the hot-melt rubber adhesive, as a significant increase of the extrusion pressure, by about 50%, was observed as soaking pH was increased (pH 10) at both temperatures (22 and 50°C).
Effect of s lot design paramete parameters rs
Milled and wedge wire screen plates were compared compared to investigate the effect of slot inlet design which was supposed to influence significantly the extrusion process. Metal blocs shown in figure 54 were manufactured at scales 1 and 5 to improve the accuracy of the slot inlet radius. The dimensions in figure 54 were defined on the basis of 0.10 mm slots and chosen to cover the variations observed with commercial screen plates. Slot print measurements were performed on various screen plates of CTP’ experimental screen including Microvortex (Kadant Lamort) and Macroflow (AFT) designs. Slot type
Slot w r b l h scale (mm) (mm) (mm) (mm) (mm)
MS
1
0.10
0.05
0.10
0.8
1
WW
1
0.10
0.20
-
-
1
MS
5
0.50
0.25
0.50
4.0
5
WW
5
0.50
1.00
-
-
5
Milled slots (MS)
w
b
Wedge Wire slots (WW)
w
w
b
h r
r
h l
r
Figure 54: Design and dimensions of the slots used for the stickies extrusion tests
The stickies extrusion tests were first performed with the slots at scale 5 in order to improve the accuracy of the size and shape of the adhesive cylinders. The results in figure 55, which were obtained at 47°C after soaking the adhesives in tap water, confirmed the higher extrusion ability of the acrylic adhesive. The extrusion of both types of adhesive particles is clearly increased with the wedge wire slot compared to the milled slot, as expected when looking at the shape of the slot inlet section. In
72
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
short, the particle is submitted to lower stresses when extruded through one round edge with slightly converging profile walls (165° angle with the wedge wire profile in figure 54) than through two sharp edges (4 times lower radius with the milled compared to the wedge wire slot) with slot inlet location on a flat profile bottom.
250
10
D170
D170 - MS
D170 - MS
D170 - WW
200
D170 - WW
8
] s [ e m 6 i t n o i s 4 u r t x E
E115 - MS
] a P 150 k [ e r u s 100 s e r P
E115 - WW
E115
0.5 mm Slot Scale 5 P = 50 kPa
D170
E115 - MS E115 - WW E115
2
50
0.5 mm Scale 5 Slot 0
0 0
1
2
3
4
0
Normalised particle size (d/w)
1
2
3
4
Normalised particle size (d/w)
Figure 55: Effect of slot design on extrusion pressure pressure (left) and on extrusion time (right)
The time needed for the adhesive particles to pass completely the slot has been recorded under the same conditions, except for the pressure which was 50 kPa for all the tests. The results in figure 55 showed a much higher extrusion time with the milled compared to the wedge wire slot, but only with the acrylic adhesive. The higher trend of the hot-melt rubber adhesive to stick at the slot walls was responsible for the very high extrusion time measured with wedge wire design. In fact, the extrusion time is over-estimated as it was recorded when the whole adhesive cylinder had passed the slot, after detachment of all the cylinder parts stuck on the slot outlet wall. The sharp inlet edges and the higher depth of the milled slot are responsible for the longer extrusion observed with the acrylic adhesive. Indeed, the most important conclusion about the results in figure 55 is that the extrusion time is about two order of magnitude higher than the duration of the screening phase, i.e. a few seconds compared to typically 20 ms, which gives new insight in the stickies extrusion mechanisms in pressure screens. This means that adhesive particles with significantly larger thickness than the slot width cannot pass the slot during one positive pressure pulse and that the passage of such particles through the slots takes most probably place over several rotor revolutions in a multi-step extrusion process. One should however keep in mind that the experimental procedure does not take the inertia of the particle approaching the slot into account, which means that extrusion times in figure 55 are over-estimated compared to the real situation, especially at low extrusion time. All the experimental experimental stickies extrusion results showed to be consistent with the numerical simulation results reported in section 4.2.2, especially the linear relations between the particle oversize (d-w) and the extrusion pressure or time [22] which were observed both experimentally and numerically. However, further investigations were considered as necessary to analyses the stickies extrusion phenomena under industrial conditions, i.e. under short pressure pulse as observed in screens.
Stickies extrusion under unsteady pressure
Experimental set up
The experimental set up and test procedure are illustrated in figure 56. The visualisation cell includes a single slot of 15 mm length between two metal blocs and two glass plates with the light source placed on one side and the camera on the opposite side. A high-speed CCD camera operated at 4000 images per second was used to record the extrusion phenomena. The pressure pulse is produced by a very simple device using a weight dropped on a flexible pipe at the cell inlet. Different pressure pulses were obtained by changing the weight and the drop height after preliminary tests had been performed to determine the most suitable length and diameter of the flexible pipe.
73
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
H
V1
V0
Adjustable weight and drop height H Introduction of a stickies stic kies parti particle cle V2 Flexible pipe
V3
V4
P
Figure 56: Stickies extrusion visualisation cell, experimental set up and test procedure The experimental test procedure was as follows: -
the adhesive cylinder was introduced in the visualisation cell through the valve V2,
-
water was circulated at very low flow rate until the particle was correctly placed at the slot inlet (V2 and V3 closed, V4 open, V0 and V1 slightly open),
-
the weight was dropped on the flexible pipe and the recording of both images and pressure pulse were started at the same time.
A piezoelectric piezoelectric pressure transducer transducer was used used to record the the pressure pulse. pulse. Due to difficulties difficulties to trigger off the camera with the signal of the pressure transducer, a mechanical solution was used to detect the beginning of the pressure pulse. A small stickies ball was stuck on the front glass plate (outer wall) and connected with a thin thread to the area where the weight hits the flexible pipe. As the weight was dropped, the impact started the pressure pulse and removed the stickies ball from the view field of the camera at the same time, which allowed to detect easier the phenomena in the case of very low particle extrusion (see figure 58). The duration of the pressure pulse was about 15 ms in all cases, with peaks between 50 and 250 kPa
Results
The extrusion of the adhesive particle during the pressure pulse was characterised by a simplified “stickies slot penetration ratio” defined defined by h.w / π.(d/2)² according to the drawing in figure 57, the penetration distance h being measured on the video images (figure 58). The series of images shown in figure 58 to illustrate the extrusion of the adhesive cylinders were taken from the video films at different times every 10 ms, the pressure pulse starting between the first and the second images (visual signal visible on the right of the first images, especially on the second line images: the small stickies ball is still at rest on the first picture has started to move on the second picture and is out of the view field on the 3rd picture). The images in figure 58 58 correspond to screening conditions where the extrusion of stickies is promoted, i.e. with soft adhesive material (soaked acrylic PSA at high temperature and pH) and at much higher and longer positive pressure peak than observed with typical foil rotors, in order to show some extrusion with relatively large stickies: -
The pictures in the first line show a small adhesive particle (d/w = 1.3) which passed completely the slot during the pressure pulse. The third image shows that one end on the adhesive cylinder passed the slot before the other end.
-
The second lines pictures show a much larger particle (d/w = 3.5) where about half of the particle was extruded through the slot during the pressure pulse at a two-times higher pressure (150 kPa). The three last images and the subsequent ones showed that one end of the adhesive cylinder was released from the slot during a small negative pulse (reverse flow) recorded after the positive
74
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
pressure pulse, while the other end remained jammed in the slot. In such a case, a second similar pressure pulse would most probably make the particle pass completely through the slot. -
The third line shows pictures of a very large particle (d/w = 5.5) which is practically not extruded through the slot.
-
The pictures in the fourth line show the extrusion of a slightly smaller particle under a two-time higher pressure pulse. About half of the adhesive cylinder passed the slot inlet and remained completely jammed in the slot. A second pressure pulse should lead to the complete passage of the particle.
d h
w
Figure 57: Definition of stickies slot penetration ratio
WW - d/w 1.3 - 80kPa - Na NaO OH - 50°C
WW - d/w 3.5 - 150kPa - NaOH - 50°C
WW - d/w 5.5 - 110kPa - Na NaO OH - 50°C
WW - d/w 5.0 - 200kPa - Na NaO OH - 50°C
Time T
T + 10 ms
T + 20 ms
T + 30 ms
T + 40 ms
T + 50 ms
Figure 58: Records of stickies extrusion at different times under various conditions The effects of stickies extrusion parameters, including temperature, adhesive soaking and slot type, were mainly studied with the acrylic adhesive and with the slots at scale 5, to improve the precision. Typical results are illustrated in figure 59, where the stickies slot penetration ratio is given as a function of the normalised stickies size (d/w) for different pressure pulse levels, i.e. about 50, 100 and 200 kPa positive pressure peaks. Some values above 100% are presented, which indicate that the whole adhesive particle or one end of the cylinder passed the slot.
75
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Figure 59 shows the stickies extrusion results obtained with the wedge wire slots at scale 5 using tap water at 20°C. Experimental curves can be drawn, despite significant dispersion of the experimental points, which are attributed to the low accuracy of the stickies size for the small particles (variations in the stickies cylinder diameter, precision of the measurements) and to the low precision of the slot penetration distance at low stickies slot penetration ratio. The maximum normalised diameter of the stickies passing completely the slot during the pressure pulse is estimated to 1.1 to 1.9 for a pressure peak of respectively 50 to 200 kPa. The size range of adhesive particles showing a maximum slot penetration ratio of 30% is about 1.7 to 2.1 for a pressure peak range of 50 to 100 kPa. The evaluation of the size range of particles with lower maximum slot penetration (for example if only 10 % of the particle volume penetrates the slot) seems less relevant since one can expect that such particles which reach the slot inlet during the positive pressure pulse will be removed easily during the negative pressure pulse. Such particles should then finally be rejected, even if subjected to several subsequent partial extrusion phenomena, as there is no large particle deformation (for example line 3, figure 58). On the other hand, is the particle penetration ratio becomes higher than about 30% (for example lines 2 and 3, figure 58) the probability for the particle to be finally extruded after several pressure pulses also becomes higher, as already mentioned.
10,000 WW5 - H20 - 20°C . o 1,000 i t a r n o i t a r 0,100 t e n e p t o l S 0,010
100 % 30 % 10 % t e o l m s u e l o h v t e g l n c i i s t r s a a P p
50 kPa 100 kPa 200 kPa
0,001 0
1
2
3
4
5
6
7
Normalised stickies size (d/w)
Figure 59: Stickies slot penetration with WW scale 5 slots at 20°C, pH 7 The tests performed with the slots at scale 1 showed consistent results with those obtained at scale 5, based on the observation of the particles which passed completely the slot or not [71]. The main results obtained at scale 5 are summarised in table 17, which shows the influence of the tested parameters parameters (temperature, soaking pH, slot design). The higher efficiency of milled slots in terms of reduced stickies extrusion is only observed for complete extrusion of the particle through the slot, which was attributed to the fact that a significantly lower part of the stickies cylinder surface area should be exposed to the pressure pulse with the milled slot compared to the wedge wire slot.
Normalised stickies diameter versus slot penetration ratio for the different test conditions
Pressure pulse (kPa)
WW 5 H2O 20°C
WW 5 H2O 50°C
WW 5 NaOH 50°C
MS 5 NaOH 50°C
Curves at 30 % stickies slot penetration ratio (e.g. figure S23)
50 100
1.7 2.1
2.3 2.6
2.6 3.4
2.6 3.4
Curves at 100 % stickies slot penetration ratio (e.g. figure S23)
50 200
1.1 1.9
1.8 2.8
1.9 3.3
1.6 2.8
Stickies passing completely the slot
50
1.3
1.8
1.9
1.5
Table 17: Effect of screening parameters parameters on the extrusion of stickies cylinders. Synthesis [71]
76
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The results in table 18 show the effects of the main screening parameters responsible for the stickies extrusion phenomena. These results were obtained at very short pressure pulse of about 15 ms, which correspond to the typical duration of the positive pressure generated during the screening phase, as illustrated in figure 37 in the case of a foil rotor. In fact, there is a pressure peak in front of the foils, which is much shorter and higher than the positive pressure observed during the screening phase after the foils. In addition, stickies particles which are captured in the exit layer do not reach the slot inlet at the beginning of the positive pressure pulse, but statistically rather during the second half of the screening phase since such stickies particles have to move back towards the slot inlet in the exit layer located under the vortex. Consequently, one should rather consider the duration of the positive pressure peak observed before the foils instead of the duration of the screening phase to evaluate stickies extrusion phenomena in low-consistency pressure screens. The values in table 18 of the normalised size of the stickies which should pass the slot during a single pressure pulse were calculated on the basis of the experimental results [71], assuming that particle extrusion is determined determined by the product of the duration by the intensity of the extrusion pressure. With this hypothesis, a positive pressure pulse of about 50 kPa over 3 ms (typical pressure peak in front of foils in lowconsistency pressure screens) is then equivalent to a lower pressure pulse of only 10 kPa over 15 ms (typically the positive pressure pulse during the screening phase after the foils). Evaluation of single-step stickies extrusion for the different test conditions
Pressure peak (kPa)
Pulse duration (s)
WW 5 H2O 20°C
WW 5 H2O 50°C
WW 5 NaOH 50°C
MS 5 NaOH 50°C
Normalised diameter of stickies passing the slot
10 50
15 3
1.06
1.16
1.18
1.10
Table 18: Effect of screening parameters parameters on the extrusion of stickies cylinders.
Extrapolation of the experimental results, including figure 59 and table 17 [71] According to the results in table 18, the maximum diameter of stickies cylinders which can be completely extruded through the slots during one single typical foil rotor pressure peak should only be 6 to 18 % larger than the width of the slots. The hypotheses used to establish this conclusion are, of course, very simplified since several phenomena such as particle inertia, visco-elastic behaviour of the adhesive material (time dependant elasticity modulus) and particle slip conditions (boundary water layer between the stickies and the slot walls) are not taken into account, but allow to quantify the extrusion phenomena under industrial conditions, which are very difficult to simulate experimentally. A main conclusion of the data in table 18 is that stickies particles having a diameter larger larger than about 1.2 times the slot width cannot be completely extruded through the slot during one single pressure pulse generated by the passage of the foils, and that significantly larger stickies which finally pass the screen plate (as observed in mills), should be extruded during several rotor revolutions. Multi-step extrusion assumes that the particle is not removed from the slot inlet during the reverse pulse, which probably requires significant particle slot penetration ratio, or that a part of the particle, the thin end of typical “string-like” stickies, passed completely the slot during the first pressure pulse.
4.2.3. 4.2.3.2. 2.
Optim isati on of scr een plate desig n
The screening tests aiming at determining the best screen plate design for the removal of PSA stickies under low-consistency screening conditions were performed at CTP on pilot scale, in cooperation with AFT for the design design and manufacture manufacture of special screen screen cylinders for the experimental experimental pressure pressure screen.
Raw materials, equipment and methods
The stickies-containing pulps used for the screening tests were prepared as follows: 2
-
Acrylic or hot-melt rubber adhesive labels (20 g/m adhesive) were stuck onto fresh newspaper to produce dark particle particles, s, as ink is absorbed on the adhesive film.
-
The adhesive labels (newspaper / adhesive / back paper complex) were added at a ratio of 2% adhesive material, to a mixture of 50% newspapers / 50% magazines and re-pulped in a lowconsistency pulper for 30 minutes at 55°C, with standard INGEDE deinking chemistry.
77
L SCREEN C CLEAN EAN
-
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The adhesive-containing adhesive-cont aining pulp was then mixed with a mixture of bleached chemical pulps (50% hardwood, 50% softwood) to enhance the contrast between the adhesive particles and the handsheets, which led to about 0.1 % adhesives in the final blend.
The pilot screening tests were performed with the experimental screen designed to simulate a slice of industrial screens (figure 60). This slice screen design facilitates the investigations of the influence of some of the most relevant screening parameters, since flow conditions, pulp consistency and composition are roughly constant over the width of the screen slice, and the reject flow rate is kept large. In the case of full size screen baskets, the pulp consistency and long fibre content increase from the inlet to the reject section, and the flow conditions are not constant. The screen cylinder diameter of the experimental slice screen is 500 mm. The pressure in the screen is maintained above 150 kPa to avoid cavitation and the reject flow rate is normally 50 % to improve the precision of the determination of the particle passage ratios.
FEED
Cf CAMERA
L A S E R
ROTOR
Cr REJECT
Cu Cd Ca
ACCEPT
Figure 60: Experimental “slice screen” – 500 mm screen cylinder diameter
The ability of a particle to pass the screen plate under given screening conditions can be characterised by its passage ratio, P = Cd/Cu, which is calculated from the consistencies measured upstream Cu and downstream Cd the screen plate. According to this approach [11-15] the characteristics of a full size screen can be assessed using the mixed-flow model (case of perfect mixing in both radial and axial direction, which is close to the flow observed in short screen cylinders with an open rotor) or the plug-flow model (case of no axial mixing, close to the flow observed with long screen cylinders with a closed rotor). The mixed-flow model applies to the slice screen design equipped with foil rotor, which allows the fibre and contaminant passage ratios to be determined in order to provide input for the simulation and optimisation of screening systems, as reported in section 4.2.4. Practically the analysis of the feed and accept pulp samples allows to determine, on this basis: -
the distribution distribution of the fibre passage ratios ratios as a function of of fibre length length according according to fibre length length analyses performed with the MorFi (Techpap) fiber analyser
-
the distribution distribution of the stickies passage passage ratios ratios as a function function of stickies size (area size classes) classes) according to stickies spot analyses
The screening results were also reported in terms of average pulp passage ratio and cleanliness efficiencies evaluated evaluated from the total stickies spot area in the accept and feed pulps. The method used to control the stickies in the pulp samples were: -
the INGEDE method n°4 for the analysis of some screening conditions, conditions,
-
image analysis analysis on handsheets handsheets for all all the tests tests conditions, conditions, as the the method is more accurate (since more particles were counted) and less time consuming.
78
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Influence of temperature and extrusion pressure
The objective in this first test series was to determine determine the influence of the screen operating parameters parameters assumed to have an effect on the extrusion of stickies, namely the passing velocity which influences the average extrusion pressure, the rotor velocity which influences the maximum extrusion pressure during the positive pressure pulse (roughly proportional to the square of the rotor velocity), the screening temperature which influences the rheological properties of the adhesive material, and the slot inlet design as shown in the previous section. The following screening conditions conditions were tested: -
Type of stickies stickie s
Hot-melt rubber adhesive D170
-
Screen plate
0.15 mm slots, wedge wire, 1.2 mm contour height
-
Rotor
4 foils, 15 and 23 m/s
-
Passing velocity
1 and 3 m/s
-
Consistency
1.3 %
-
Temperature
23, 36, 51 and 69 °C
The hot-melt rubber adhesive was chosen instead of the more common acrylic adhesive because it was suspected that temperature would have more influence on the rheological properties of such a “hot-melt” adhesive material. The wedge wire screen plate design with high contours was chosen as it was supposed to be more sensitive to stickies extrusion compared compared to low contours and milled slots. The effects of temperature and slot passing velocity on the pulp passage ratio are shown in figure 61. A large increase of the pulp passage ratio with the passing velocity, as well as a low influence of the rotor velocity were observed which is usual under similar screening conditions [4-69]. The slight decrease of the pulp passage ratio observed as the temperature is increased is attributed to the reduction of the fibre to liquid friction (less fibre entrainment in the exit layer) and to the increase of the fibre to slot wall friction (more solid to solid contacts) caused by the viscosity decrease (by about 60% between 20 and 70°C). This observation is consistent with recently reported results showing a clear increase of the fibre passage ratio as CMC was added to increase strongly the fluid viscosity [72].
1,0
0,8 o i t a r e 0,6 g a s s a p 0,4 p l u P 0,2
Averag Average e 1 m/s slot velocity
D 170 0.15 mm WW 1.2
Averag Average e 3 m/s slot velocity
0,0 10
20
30
40
50
60
70
80
Temperature (°C)
Figure 61: Effect of temperature and slot velocity on fibre passage
The influence of temperature on the stickies removal efficiency (figure 62) is much higher, with the stickies content in the accepts being much lower for large stickies (spots > 0.5 mm) at low temperature, which suggests that stickies extrusion is much higher at high temperature. The results in figure 62 also indicate a certain fragmentation of the stickies particles during the screening tests, which were performed in recirculation on the feed chest. In particular, less large stickies were counted at the inlet at the beginning of the test series (tests at 23°C) than at the end of the test series (tests at 69°C).
79
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
20 D 170 0.15 mm WW 1.2
) g / ² 15 m m ( t n e t n 10 o c s e i k c i t 5 S
Feed - 23°C Feed - 69°C Acc ept - 23°C Acc ept - 69°C
0 0
1
2
3
4
5
Stickies spot size by INGEDE method n°4 (mm)
Figure 62: Effect of temperature on stickies removal efficiency
The cleanliness efficiency based on the INGEDE method n°4 decreases from 87% to 72% for the 2 stickies producing spots larger than 0.02 mm and from 93% to 83% for the stickies producing spots 2 2 between 1 mm and 5 mm , as the screening temperature was increased from 23°C to 69°C. These efficiency drops are exaggerated since they correspond to relatively large spot size classes, because of the fragmentation of the large stickies, the average stickies particle size in the feed pulp is lower for the tests at high temperature. temperature. The cleanliness efficiencies in the figures 63 and 64 were determined according to feed and accept pulp handsheet image analyses. The size classes do not correspond to those in figure 17 as the stickies spread out more with the INGEDE method n°4 (heating at 105°C after vacuum drying) than in handsheets (dried at 94°C under vacuum). In addition, only the macro-stickies retained on 0.10 mm lab slots are controlled with the INGEDE method n°4, which means that thin stickies (which spread out less than thick ones) are only taken into account with the handsheet image analysis method. The drawback with this method is that small thin stickies are difficult to count or may be counted as several very small particles if they are embedded in the thickness of the handsheet. Consequently the efficiencies obtained in the small size classes were not relevant and were removed from the figures. It should also be kept in mind that the screening tests were performed with the slice screen at 50% reject flow rate (to provide input for further screening system simulation studies), which means that the accept pulp is only 40 to 50% of the feed pulp (cf. figure 61). The stickies removal efficiency becomes much lower as the final reject rate is decreased down to about 1% fibre losses, especially at low efficiency level. Consequently the low cleanliness efficiencies obtained with the stickies in the small 2 size classes (< 0.15 mm ) are not only wrong but also not relevant since such stickies would practically not be removed in multistage screening systems. Figure 64 shows the effect of the velocity of the rotor (Vr) on the stickies removal efficiency, at different passing velocities (Vp). No large influence is observed at low passing velocity, while an increase of the efficiency is observed as the rotor velocity is increased at high passing velocity. The results in figure 64 are quite surprising since rotors with low positive pressure pulses are generally recommended to reduce the extrusion pressure applied on the stickies and thus to improve their removal efficiency. Indeed, the positive pressure pulse is relatively low with the tested foil rotor even at a high rotor velocity, i.e. about 7 to 15 kPa between 15 and 23 m/s [18]. At a high passing velocity, the variation of the extrusion pressure (i.e. the positive pressure pulse plus the average pressure drop caused by the passing velocity) explains that the reduction of the rotor velocity is more positive (no negative effect) at low passing velocity. On the other hand, the increase of the velocity of the rotor increases, roughly proportionally, with the velocity of the pulp at the surface of the screen plate, if the equilibrium tangential velocity is reached. This is the case with the slice screen where only the central part of the contoured screen cylinder cylinder in slotted, but generally not the case in the feed section of industrial screens where the pulp is fed at much lower tangential velocity. This means that dynamic phenomena such as
80
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
particle slip and particle contacts with the inclined contour wall could have more impact on the efficiency than stickies extrusion phenomena. phenomena.
100 ) % ( y c n e i c i f f e l a v o m e r s e i k c i t S
80 Avera Average ge 23°C 60
Avera Average ge 36°C Avera Average ge 51°C
40
Avera Average ge 69°C 20 D 170 - 0.15mm WW 1.2 0 0
1
2
3
Stickies size size mea sured on handshee handshee ts (mm)
Figure 63: Effect of temperature on stickies removal efficiency
100 ) % ( y c n e i c i f f e l a v o m e r s e i k c i t S
80 Vp 1m/s - Vr 15m/s 60
Vp 1m/s - Vr 23m/s Vp 3m/s - Vr 15m/s
40
Vp 3m/s - Vr 23m/s 20 D 170 - 0.15mm WW 1.2 0 0
1
2
3
Stickies size size mea sured on handshee handshee ts (mm)
Figure 64: Effect of rotor velocity on stickies removal efficiency
Influence of screen plate and slot design
The objective in this test series was to determine the influence of the screen plate parameters and of the parameters interacting with screen plate design, i.e. slot inlet design, contour height, passing velocity and rotor velocity. The tests were limited to the most usual situation regarding the type of stickies and the screening temperature, which corresponds to soft adhesive material and should consequently include stickies extrusion phenomena. The following screening conditions were tested: -
Type of stickies
Water-based Water-based acrylic adhesive E115
-
Screen plate
0.15 mm slots with 3 different contour / slot designs (figure 65)
-
Rotor
4 foils, 15 and 23 m/s
-
Passing velocity
1 and 3 m/s
-
Consistency
1.3 %
-
Temperature
42 °C
81
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The three screen plates compared in this second test series are illustrated in figure 65. The pitch was about the same for the 3 screen plates (3.2 mm), which gave about equal contour angle for the two high-contour screen plates. -
The wedge wire screen plate with high contours (1.2 mm) was a conventional conventio nal wedge wire screen cylinder with welded wires.
-
The milled screen plate with high contours (1.2 mm) was made up up of U-shaped rings with milled contours and laser cut slots.
-
The wedge wire screen plate with low contours (0.6 mm) was a constructed construct ed screen plate, with a more accurate slot width compared to the conventional wedge wire screen plate.
H = 1.2 mm
H = 1.2 mm
H = 0.6 mm
Milled Milled scre screen en plate plate - MS 1.2
WW 1.2 1.2 - Wedge Wedge wire screen screen plates plates - WW 0.6 0.6
Figure 65: References and design of the screen plates tested in the second test series
The results in figure 66 confirmed the trends already observed, i.e. a relatively low effect of rotor velocity on the pulp passage ratio and improved fibre passage as passing velocity is increased and as well as with wedge wire screen plates, which was attributed to the higher effective slot velocity and to the favourable slot inlet design [4,67]. A reduction of the contours reduced the pulp passage ratio, which was attributed to a higher probability for the fibres to be rejected through contacts with the lowangle contour walls, especially at low passing velocity as the vortex generated between between the contours is elongated elongated in the main flow direction [17] with the exit layer turning sharper back to the slot inlet.
1,0
0,8 o i t a r e 0,6 g a s s a p 0,4 p l u P 0,2
0.15mm WW 1.2 - Vr 15m/s 0.15mm WW 1.2 - Vr 23m/s 0.15mm WW 0.6 - Vr 15m/s 0.15mm WW 0.6 - Vr 23m/s 0.15mm MS 1.2 - Vr 15m/s 0.15mm MS 1.2 - Vr 23m/s
E 115 43 °C 0,0 0
1
2 Passing velocity (m/s)
3
4
Figure 66: Effect of screen plate and rotor velocity on fibre passage
The efficiency of the screen plates in removing stickies is shown in figure 67, where the cleanliness efficiency is plotted against the stickies spot size according to INGEDE method analyses, which were performed for the screening trials at low passing velocity only.
82
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
100 ) % ( y c n e i c i f f e s s e n i l n a e l C
80 E 115
Vp = 1m/s
60 0.15mm WW 0.6 - Vr 15m/s 0.15mm WW 0.6 - Vr 23m/s
40
0.15mm MS 1.2 - Vr 15m/s 0.15mm MS 1.2 - Vr 23m/s
20
0.15mm WW 1.2 - Vr 15m/s 0.15mm WW 1.2 - Vr 23m/s
0 0
1
2
3
Stickies spot size by INGEDE method n°4 (mm)
Figure 67: Effect of screen plate and rotor velocity on efficiency
The curves in figure 68 correspond to the average values, at 15 and 23 m/s rotor velocity, of the efficiencies given in figure 67. They represent the distributions of the stickies passage ratio Pk, which were calculated directly from the average efficiencies (mixed-flow model), and of the stickies to pulp passage ratio β = Pk/Pf. As indicated in section 4.2.4, this β ratio represents the selectivity of the separation of the contaminants with respect to the fibres, and would allow one to calculate screening system efficiencies in the case of homogeneous homogeneous stickies and fibres in different size or length classes.
0,8 Pk/Pf - 0.15mm WW 1.2 Pk/Pf - 0.15mm MS 1.2
s o 0,6 i t a r e g a s 0,4 s a p s e i k c i t 0,2 S
Pk/Pf - 0.15mm WW 0.6 Pk - 0.15mm WW 1.2 Pk - 0.15mm MS 1.2 Pk - 0.15mm WW 0.6 E 115
Vp = 1m/s
0,0 0
1
2
3
Stickies spot size by INGEDE method n°4 (mm)
Figure 68: Effect of screen plate on stickies passage ratios
The results in figure 67 and the Pk values in figure 68 indicate a higher cleanliness efficiency of the milled slots at given contour height or angle. The β ratios indicate that a slightly higher screening selectivity is finally obtained with the milled slots despite lower fibre passage ratio. The stickies to pulp passage ratios ( β = Pk/Pf) in figure 69 were calculated from the average cleanliness efficiencies determined by handsheet image analysis for the tests at 1 and 3 m/s passing velocity. The best results, i.e. the lowest β ratios, were always obtained with the low-contour wedge wire screen plate. The comparison of the two high-contour screen plates showed higher efficiency with the milled than with the wedge wire screen plate, at least at low passing velocity as in figure 68, the results being rather unclear at high passing velocity.
83
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
1,6 0.15mm WW 1.2 - Vp 3m/s
f 1,4 P / k P 1,2 o i t a r 1,0 e g a 0,8 s s a p 0,6 s e i k 0,4 c i t S 0,2
0.15mm MS 1.2 - Vp 3m/s 0.15mm WW 0.6 - Vp 3m/s 0.15mm WW 1.2 - Vp 1m/s 0.15mm MS 1.2 - Vp 1m/s 0.15mm WW 0.6 - Vp 1m/s E 115 43°C
0,0 0
1
2
3
Stickies spot spot size o n ha ndshee ndshee ts (mm) (mm)
Figure 69: Effect of screen plate on stickies removal at Vp = 3m/s
Indeed, the experimental points in figure 69 with the high-contour wedge wire screen plate are average values of the results obtained at the 2 rotor velocities, where the test at 23 m/s rotor velocity has been the last one of the test series. At this step, there were very few stickies left in the size 2 classes above 2 mm , so that the high efficiencies (low β ratios) found for these stickies are not accurate. This means that the better results reported in figure 69 for the large stickies with the highcontour wedge wire screen plate compared to the milled screen plate could be wrong. If not, i.e. if the wedge wire screen plate was effectively better than the milled one, one explanation could be that the reverse pulse was less effective in removing stickies located at the slot inlet, especially at high passing velocity, which reduces the efficiency of the reverse pulse. The best results in terms of separation selectivity between stickies and fibres were clearly obtained with the low-contour screen plate compared to the high-contour screen plates, where slightly higher efficiency was achieved with the milled screen plate, most probably because of less stickies extrusion than with the wedge wire design. A main conclusion of this second pilot test series is that the extrusion of PSA stickies particles is not the only phenomena to consider in the optimisation of stickies screening conditions, conditions, as it is suggested that hydrodynamic phenomena, which should occur at the surface of the screen plate before stickies even reach the slot inlet, should be considered first. The higher efficiency of low-contour screen screen plates, which was already observed with small model films and shives [67], has been attributed to the higher effectiveness of low-angle contours in rejecting contaminant as they approach the screen plate and hit and slip over the inclined contour walls. The relatively lower influence of stickies extrusion phenomena phenomena on the screening efficiency under the tested conditions was confirmed by the low influence of the velocity of the rotor (a foil rotor with quite low positive pressure pulse) pulse) on the efficiency.
Optimisation of screen plate contour design
The results obtained in the previous test series suggested that further research should focus on the optimisation of the design of the contours of the screen plate, i.e. contour height, angle and related pitch. The previous stickies screening trials also showed that the test procedure should be improved and the number of tests increased in order to improve the precision of the results since slight differences between different screen plates will probably have to be determined for the optimisation of contour design. The stickies preparation procedure in this third test series was the same as previously, except that a new batch of bleached kraft pulp (50 % softwood, 50% hardwood) was prepared for each screen plate, by mixing each time the same amount of stickies-contaminated pulp taken from a large quantity prepared for the whole test series. The stickies were prepared as previously by re-pulping adhesive labels with standard deinking raw material and chemistry in a small low-consistency pilot pulper (5% consistency, 100 l capacity), but the size and shape of the stickies appeared to be different when looking at the handsheets:
84
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
-
the E115 acrylic adhesive labels used for the second test series were about 8 month old, and showed an average particle thickness (calculated from the total stickies weight and the total stickies area measured in hansdheets) of 38 µm.
-
the E115 adhesive labels used for the third test series was a mixture of 75 % fresh labels produced 2 week before and 25 % of the previous labels which were 14 month old. The average stickies thickness in handsheets was 51 µm.
Additional analyses performed on the fresh and old adhesive labels re-pulped re-pulped separately, confirmed the significantly lower average stickies thickness when re-pulping of the labels was performed with aged adhesive material compared to fresh adhesive. The following screening conditions were tested: -
Type of stickies
Water-based Water-based acrylic adhesive E115
-
Screen plate
0.15 mm slots with 5 different contour height and/or angle (table 19)
-
Rotor
4 foils, 15 and 23 m/s
-
Passing velocity
1, 2 and 3 m/s
-
Consistency
1.2 %
-
Temperature
about 42 to 45 °C
Screen plate references
Slot width (mm)
Contour height (mm)
Pitch (mm)
Contour angle (°)
06 32 13° WW
0.149
0.6
3.35
13.2
06 26 17° WW
0.153
0.6
2.75
17.3
09 32 20° WW
0.150
0.9
3.35
19.6
09 26 25° WW
0.153
0.9
2.75
24.8
12 29 25° WW
0.154
1.2
3.05
25
Table 19: References and characteristics of the tested screen plates.
The five screen plates are compared in figure 70 in terms of pulp passage ratio achieved on average at 15 and 23 m/s rotor velocity. Figure 71 confirms the trends observed during the previous tests series: A slightly higher pulp passage ratio was achieved with all the screen plates as rotor velocity was reduced. The comparison of the screen plates in figure 70 revealed an increase of the fibre passage as the wire width was reduced down to 2.6 mm compared to 3.2 mm, at 0.6 and 0.9 mm contour height. The relatively high pulp passage ratio of the wedge wire screen plate with 1.2 mm contours could be due to both the intermediate wire width (2.9 mm) and the higher contours, which were shown to improve slightly the fibre passage ratio in the frame of previous studies [67]. The broader slot width distribution of the 12 29 25° WW screen plate (with welded wires) compared to all the other screen plates (constructed) could also contribute to explain this higher pulp passage ratio. The positive effect of the reduction of the pitch on the fibre passage is assumed to be due to the increase of the pulp flow passing the screen plate per unit area, which increases the quantity of fibres accumulated during the screening phase for a given instantaneous fibre passage ratio and consequently the consistency of the pulp at the feed side of the screen plate at the end of the screening phase. As the instantaneous fibre passage ratio is defined as the instantaneous (or local) feed to accept pulp consistency ratio, this hypothesis would explain the higher average pulp passage ratio observed with the 2.6 mm wires. One should however note that these experimental results about the influence of wire width and their analysis correspond to low/medium consistency pulp (1.2 % fibre consistency with 50 % softwood, which should be equivalent to 1.5 to 2 % consistency with deinking pulp). At high screening consistency, the faster increase of the consistency assumed to take place during the screening phase as pitch is reduced, might lead to a too high consistency increase and thus to fibre flocculation, which might reduce the average pulp passage ratio, instead of increasing it, or even plug the screen, especially at low rotor velocity.
85
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
1,0
0,8
o i t a r e 0,6 g a s s a 0,4 p p l u P
06 06 09 09 12
0,2
32 26 32 26 29
13° W W 17° W W 20° W W 25° W W 25° W W
0.15 mm slots (av. Vr 15-23 m/s)
0,0 0
1
2
3
4
Passing velocity (m/s)
Figure 70: Effect of contour design and slot velocity on fibre passage
1,0
0,8
o i t a r e 0,6 g a s s a 0,4 p p l u P
15 m/s rotor velocity
0,2
0.15 mm slots (av. test series 3)
23 m/s rotor velocity 0,0 0
1
2
3
4
Passing velocity (m/s)
Figure 71: Average effect of rotor velocity on fibre passage
100
) % ( y c n e i c i f f e s s e n i l n a e l C
80
60
06 06 09 09 12
40
20
32 26 32 26 29
13° WW 17° WW 20° WW 25° WW 25° WW
0.15 mm slots (av. Vr 15-23 m/s)
0 0
1
2
3
Passing velocity (m/s)
Figure 72: Cleanliness efficiencies of the different screen plates
86
4
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The cleanliness efficiencies observed in the accept pulp (less than 50 % of the feed pulp) with the different screen plates are compared in figure 72, showing the results obtained with the stickies in the 2 most sensitive size classes, i.e. 0.3 to 1.0 mm measured on handsheets. A decrease of the height and angle of the contours improved the cleanliness efficiency, which confirmed the high efficiency increase already observed in previous studies with flat-shaped model contaminants [67] as the height of the contours was reduced when comparing the 06 32 13° WW to the 12 29 25° WW screen plate. The cleanliness efficiencies achieved with the different screen plates are shown in figure 73 as a function of the pulp passage ratio in order to compare the results at equal reject rate. The two screen plates with the lowest contours (0.6 mm) gave the best results. A reduction of the contour angle down to 13° reduced the stickies passage ratio (improved their retention) which was attribute to the higher probability of the particles to be rejected through contacts with the inclined contour walls as the angle was reduced [67]. The reduction of the contour angle also reduced the passage of the fibres in about the same proportions, which was attributed to a slower fibre concentration during the screening phase rather than to the effects of fibre contacts with the contour walls, as fibres are flexible and contacts should occur with only one end of the fibre [67]. These phenomena could explain the similar screening selectivity observed with the two low-contour screen plates. The results in figure 73 indicate that, with the low-contour (0.6 mm) screen plates, a further reduction of the contour angle from 17 to 13° does not improve the screening selectivity. By contrast, with the medium-contour (0.9 mm) screen plates the reduction of contour angle, from 25 to 20° improved the screening selectivity, suggesting that the reduction of the angle had more effect on rejecting stickies through contacts with the inclined contour walls, compared to the effect of pitch increase on the fibre passage ratio reduction. Finally the screen plate with the highest contours (1.2 mm) showed slightly lower screening selectivity than the one with lower contour height (0.9 mm) and equal angle (25°).
100
) % ( y c n e i c i f f e s s e n i l n a e l C
80
60
06 06 09 09 12
40
20
32 26 32 26 29
13° WW 17° WW 20° WW 25° WW 25° WW
0.15 mm slots (av. Vr 15-23 m/s)
0 0,0
0,2
0,4
0,6
0,8
1,0
Pulp passage ratio
Figure 73: Selectivity curves of the different screen plates
The velocity of the rotor showed no large influence on the screening selectivity as observed in previous studies with model flat-shaped contaminants [67] and illustrated in figure 74 showing the average selectivity curves obtained with the five screen plates at 15 and 23 m/s rotor velocity. The lower rotor velocity gave, on average, slightly better results but only at low passing velocity. The results in figure 73 show that the two low-contour screen plates achieved about the same efficiency at given pulp passage ratio, i.e. at given reject thickening factor, corresponding to lower slot velocity with the 06 26 17° compared to the 06 32 13° screen plate. However, since the 06 26 17° screen plate offers about 20 % higher open area, both screen plates should have similar screening capacities at given stickies removal efficiency.
87
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
100
) % ( y c n e i c i f f e s s e n i l n a e l C
80
60
40
15 m/s rotor velocity
20
23 m/s rotor velocity
0.15 mm slots (av. test series 3)
0 0,0
0,2
0,4
0,6
0,8
1,0
Pulp passage ratio
Figure 74: Average effect of rotor velocity on cleanliness efficiency
Indeed, the comparison of the different screen plates should be further analysed with the help of the simulation model developed to predict screening system efficiencies (see section 4.2.4). The model however requires the passage ratio distributions of the stickies (figure 68) assuming constant stickies shape in a given size class, as well as the fibre passage ratio distributions, i.e. the fractionation effect to predict the concentration of the long fibres in the rejects. Simplified calculations can be established to compare the efficiency of multistage screening systems equipped with the different screen plates. The results in figure 75 were obtained on the basis of the experimental results in figure 73, by using the plug-flow model (representing quite well screening systems with several stages and all the accepts fed forward) and for a final reject rate of 1 %. The calculations in this figure are only valid for the stickies in the size range considered in figure 71, i.e. 2 stickies forming specks between 0.3 and 1 mm area, and assuming that all these stickies have the same shape and rheological properties. It is also assumed that all the fibres have the same passage ratio as the average pulp passage ratio. Clearly these hypotheses are not realistic, but the fact that both the thick stickies and the long fibres concentrate in the rejects more or less compensates the wrong assumptions.
100
) % ( y c n e i c i f f e l a v o m e r m e t s y S
Plug-flow model 1% reject rate
80
0.15 mm slots (av. Vr 15-23 m/s)
60
06 06 09 09 12
40
20
32 26 32 26 29
13° W W 17° W W 20° W W 25° W W 25° W W
0 0,0
0,2
0,4
0,6
0,8
Pulp passage ratio
Figure 75: Simulation of the efficiency of screening systems
88
1,0
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Under these simplifying hypotheses, the curves in figure 75 look similar to those in figure 73 and lead to the same conclusions when comparing the different screen plates. The main difference is the lower efficiency level, which was logically obtained as the final reject of the screening system was reduced down to 1 % fibre losses. Only the screen plates with low contours (0.6 mm) and/or low angle (< 20°) should be able to achieve stickies removal efficiencies above 50 %, if operated at low passing velocity, under the hypothesis in figure 75. Finally, it should be keep in mind that the shape of the stickies influences strongly the screening efficiency, as illustrated in table 20 where the cleanliness efficiencies measured on accept pulp 2 handsheets for the stickies in the 0.4 to 1.0 mm size class are compared at the same screening conditions between the different tests series. Table 20 also gives the average thickness of the adhesive spots measured in the handsheets.
Screen plate
12 29 25° WW
06 32 13° WW
Test series - Adhesive Av. spot spot thickn thickness ess
1 - D170 Old 52 µm
2 - E115 Old 38 µm
3 - E115 Fresh 51 µm
2 - E115 Old 38 µm
3 - E115 Fresh 51 µm
Efficiency Efficie ncy - Vp 1 m/s
76 %
20 %
60 %
78 %
92 %
Efficiency Efficie ncy - Vp 3 m/s
56 %
10 %
37 %
20 %
68 % 2
Table 20: Cleanliness efficiency measured on accepts handsheets handsheets for the 0.4 - 1mm stickies
The best efficiencies were obtained with the hot-melt rubber adhesive D170 and the lowest with the acrylic adhesive E115 produced from old labels. The higher efficiency obtained with the acrylic adhesive in the third test series with fresh re-pulped adhesive labels is attributed to the higher thickness of the stickies with the fresh adhesives, compared to the old labels which produced more st flat-shaped particles. The fact that the efficiencies obtained with the hot-melt adhesive (1 test series) rd were higher than those obtained with the fresh acrylic adhesive (3 test series) could be due to the lower extrusion ability of the hot-melt adhesive D170 compared to the acrylic adhesive E115, since both adhesive particles gave the same average spot thickness in the handsheets. However, the higher trend of the hot-melt adhesive to spread out during handsheet drying suggests that the average stickies thickness calculated from the total stickies spot area is more underestimated with the hot-melt adhesive compared to the acrylic one. Previous studies have effectively shown that the hot-melt adhesive films showed higher trend to stick onto itself during re-pulping and to form thicker particles compared to the acrylic adhesive. Consequently, it is not possible to conclude clearly that the higher screening ability observed with the hot-melt adhesive particles during the pilot screening tests is due to their lower extrusion ability which was observed at lab scale compared to the acrylic adhesives.
4.2.3. 4.2.3.3. 3.
High-con sis tency scr eening
The tests aiming at determining the best screening conditions for the removal of PSA stickies in the high-consistency range, i.e. 2 to 4 %, were performed at CTP on pilot scale. The design and the manufacture of a screen cylinder pressure screen were done by CTP in cooperation with AFT.
Background
Referring to typical deinking processes (figure 2) and to European mills [2], low-consistency screening is generally performed with 0.15 mm slots (used in previous section to optimise screen plate design). By contrast, The first pressure screening step implemented after the pulper screening step with relatively large holes (screening zone of a drum pulper or secondary pulper after a high-consistency batch pulpers equipped with typically 5 to 6 mm diameter holes) is normally performed at relatively high consistency, with holes and/or with slots under the following conditions [2]: -
1.2 to 2.0 mm diameter holes at 3 to 4 % consistency
-
0.20 to 0.25 mm slots at 2.5 to 3.5 % consistency
89
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The strategy proposed in the SWP 3.4 “High consistency screening“ research programme to improve deinking pulp cleanliness cleanliness is based, for a large part, on the increase of the stickies removal efficiency at the first fine screening step, i.e. preferably in the high consistency range. Typically, special rotors are used at high velocity to avoid fibre flocculation, which may lead to the fragmentation of some soft adhesives, as a result of the high shear rates. If stickies particles are fragmented fragmented in pressure screens, the stickies removal efficiency will decrease at the first stage and even more at the secondary stages, with consequently a significant decrease decrease of the efficiency of the whole screening system. In addition, the extrusion of stickies through slots should be minimized, despite the conclusions of the work reported in the previous section, which showed that screen plate design had more influence on the screening efficiency than parameters known to affect the extrusion pressure, the best results being obtained with low-contour screen plates. These results suggested that contour height and associated hydrodynamic particle slip phenomena over screen plate contours may have more impact on the screening efficiency than slot width and associated stickies extrusion, at least in the low-consistency range, where the pilot tests were performed. Conclusions about the optimisation of coarse screening in the high-consistency range might be different since rotors producing higher pressure pulse are normally used at this step and because concentrated fibres and flocs probably transfer particle separation phenomena to the slot inlet and should then contribute to increase the extrusion of stickies. The research work performed in this field included mill analyses to quantify the fragmentation of stickies at the high-consistency screening steps in deinking lines as well as pilot screening tests, with both conventional and new high-consistency screening technology. A new rotating basket technology has been investigated for slot screening and evaluated on the experimental slice screen, as planned. The idea was to protect the soft adhesives from high shear by placing the rotor at the accept side of the basket, and to generate the required pulp de-flocculation at the feed side of the screen plate.
Evaluation of high-consistency screening steps in a deinking mill
A European deinking mill producing copy paper from wood-free deinking raw material was chosen as the mill offered the possibility to check two high-consistency screening steps, respectively equipped with small holes and fine slots, regarding the usual holes and slots size ranges: -
High-consistency screening with holes: 3-stage screening system at 4.6 % feed consistency st nd (first stage), 1.2 mm holes at the 1 and 2 stages (accepts fed forward) and open final stage rd screen equipped with 1.8 mm holes (accepts to 3 slot screening stage inlet)
-
High-consistency screening with slots: 4-stage screening system at 2.8 % feed consistency (first stage) with 0.20 mm slots and with accepts fed forward at all stages
The results of the analyses of the pulps sampled during the first day over the whole holes and slots high-consistency high-consistency screening system showed that: -
32 % of the macro-stickies were removed
-
47 % of the macro-stickies were left in the accepts of the system
-
25 % of the macro-stickies were missing and had thus probably been fragmented in smaller stickies particles (micro-stickies not counted by the INGEDE method n°4)
The analyses of the pulps sampled during the second day at the first stage of the slot screening step nd showed a higher stickies load (probably due to more stickies in the raw material the 2 day) and no significant stickies fragmentation. In addition the last slot screening stage showed a high efficiency of about 82 %. Both results were not surprising since macro-stickies which passed the hole-screening step are normally more resistant to fragmentation (weak stickies should have been fragmented at the first screening step) and stickies which concentrate in screening rejects are easier to remove. The mill sample analyses revealed a clear by relatively limited fragmentation of the macro-stickies at the high-consistency screening steps. The sampling conditions unfortunately did not allow to identify precisely where macro-stickies fragmentation took place, but the results suggested that the first highconsistency screening steps should be responsible for this stickies fragmentation, fragmentation, as expected.
90
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Design and construction of a rotary screen cylinder screen prototype
The prototype of pressure screen with rotary screen cylinder has been designed on the basis of the experimental screen described in figure 60, which has been modified to investigate this technology, as shown in figure 76, where the internal static parts are marked in blue and the rotating parts in red. The characteristics of the two experimental screen versions are compared below. -
Basic version:
Static screen cylinder Rotor
Modified version:
500 mm diameter 4 foils
Rotary screen cylinder 400 mm diameter Static “rotor” 4 static foils
FEED
23 m/s max velocity 22 m/s max velocity
FEED
REJECT
ACCEPT
ACCEPT
REJECT
Figure 76: Modification of the experimental “slice screen”
Version with static screen cylinder cylinder (left) – Version with rotary screen cylinder (right)
Profile A
Direction of rotation
Figure 77: Experimental “slice screen” with rotary screen cylinder
Static foils (left) – Rotary screen cylinder (right)
91
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The central slotted part (100 mm width) of the rotary screen cylinder (300 mm width) was constructed and designed by AFT (figure 77 left). The slotted width was 80 mm. The slots were milled and had the following design called “Profile A”: -
Slot width:
0.15 mm nominal, 0.18 mm on average (measured with gauges)
-
Contour height:
0.9 mm
-
Pitch:
3.2 mm
-
Contour angle:
35°
A conventional conventional hydrodynamic hydrodynamic sealing device was implemented implemented in the front side of the prototype in order to avoid as much as possible by-pass flows between screening accepts and rejects. The “static foils” have been designed in such a way to be able to change their shape and location in order to produce suitable pressure pulses. Indeed, there are no available (published) data about static foils, though rotary screen basket screens are proposed by some equipment suppliers for coarse screening with holes. Figure 77 (left) shows the dimensions of the static foils for two combinations of pieces of equipment to produce different pressure pulses. Other combinations are also possible and the gap between the static foils and the rotating screen basket can be adjusted independently by changing the intermediate plates (several plates are available with thickness between 2 and 9 mm). Preliminary tests performed with the two static foil configurations shown in figure 77 showed no large difference on the screening capacity, in terms of maximum flow rate before plugging. All the tests were done with the upper design in figure 77 which achieved slightly higher screening capacity. The first screening tests showed a surprisingly high pulp passage ratio, even at high consistency and low rotor velocity. This result appeared to be “too good” and no explanation was found except some high by-pass flow at the hydrodynamic sealing. In such a case, the expected effects of the screening parameters (pulp consistency, screen cylinder and slot velocity) would be completely overshadowed by variations of the by-pass flow. Changing for example the velocity of the screen cylinder modifies strongly the flow and the pressure distribution around the rotary screen cylinder and consequently the by-pass flow. In addition, the prototype achieved practically no stickies removal, which confirmed the hypothesis that a strong by-pass flow should take place through the hydrodynamic sealing from the reject area to the accepts, and consequently pollute the accept pulp. Various changes, including the implementation of radial plates inside the rotary screen cylinder to reduce the vortex or feeding the screen through the reject outlet to replace by-pass of reject pulp by less contaminated feed pulp, were tested (details can be found in the report D15) but did not enable to operate the prototype with sufficiently low by-pass flow at the hydrodynamic sealing, and consequently to conclude on the rotary screen cylinder screen technology.
Evaluation of stickies fragmentation in screens
The objectives were to compare the rotary and static screen cylinder technologies in order to asses stickies fragmentation in the high-shear zone in the gap between the rotor and the screen cylinder, taking into account that there is no gap on the feed side with rotary screen cylinder technology. Both screens were equipped with 0.15 mm milled slots (profile A, 0.9 mm contour height, 3.2 mm pitch and respectively 0.18 and 0.17 mm measured slot width with the rotary and the static screen cylinders). The gap between the foils and the screen cylinder was about 4 mm for both screen versions. Stickies fragmentation was evaluated by handsheet image analysis in order to measure smaller particles than with the INGEDE method n°4. The tests were performed with a bleached softwood/hardwood kraft pulp mixture containing about 1% acrylic adhesives (E115). The adhesives labels were stuck onto fresh newspaper to produce dark particles and re-pulped in the pilot drum pulper with newspapers during a relatively low pulping time (15 min at 40°C with standard INGEDE deinking chemistry) to get relatively large stickies particles. The adhesive containing pulp was then mixed with the bleached chemical pulp. The screening tests were performed in recirculation on the feed chest (pulp samples were taken on the feed) with the accepts closed and open, in order to asses the effect of stickies particles passing the slots or partly jammed at the slot inlet (long-shaped stickies) on their fragmentation. fragmentation.
92
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
One fresh batch of stickies-containing pulp was used per type of screen and the tests started with the closed accepts, under the flowing conditions: -
Duration of the test
1 hour = about 30 average passages
-
Pulp consistency
3.2 to 3.5 %
-
Temperature
44 to 54°C (end of the test)
-
Passing velocity
1 m/s with open accepts (Vp = 0 means closed accepts)
-
Rotor velocity
20 m/s
The results in figure 78 showed a clear fragmentation of the stickies during the high-consistency screening tests but the effect was relatively limited since the average stickies size was only reduced by about 20% after 30 passes in the screen.
Static Screen Screen Cylinder Vp = 1 m/s
Static Screen Cylinder Cylinder Vp = 0 100
100
80
80
60
60
40
40 t= 0 30 min 60 min
20
60 min 90 min 120 min
20
0
0 0
0, 5
1 1, 5 2 2, 5 Stickies particle size (mm)
0
0, 5
1 1,5 2 2, 5 Stickies particle size (mm)
Figure 78: Stickies particle size distribution during conventional high-consistency screening tests
These results were obtained with weak stickies, i.e. acrylic adhesives produced under alkaline pulping conditions with gentle equipment (drum pulper) and screened at relatively high temperature (50°C). High-consistency High-consistency screening conditions with higher rotor velocity, i.e. up to 25 m/s instead of 20 m/s, or even at higher temperature (about 60°C seems to be a maximum in the first deinlking loop) might have led to higher stickies fragmentation. fragmentation.
4.2.4. Simulation of screening systems The objectives in this sub-workpackage about the development of a screening model were: -
to develop simulation tools and a model for the simulation of multistage screening systems, which is based on the probability screening theory,
-
to optimise screening systems in terms of stickies removal efficiency and fibre losses, on the basis of input data generated by the pilot screening tests (section 4.2.3), in terms fibre and stickies particle passage ratio distributions, according to the probability screening screening theory,
-
and finally to use the simulation tools and experimental input data in order to define the best strategies for the removal of stickies in deinking lines (section 5.1).
The latest developments of the screening model and an example of simulations of screening systems established with input data from earlier pilot screening tests with flat-shaped model contaminants, are presented in this section. The complete results of the first simulations can be found in the report D9. The applications to the optimisation of stickies removal in screening systems combining cleaning and flotation are reported in section 5.1.
93
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Screening model and simulation tools
Among the different possibilities to develop models for the optimisation of screening systems in mills, such as empirical models based on pulps consistency, freeness or global cleanliness measurements, the approach based on the particle passage ratio concept based on to the probability screening theory seemed the most promising as it relies on the characterisation of the physical particle separation phenomena. Both fractionation screening (to separate long/coarse fibres or shives from the short or good fibre fraction) and contaminant screening (to remove various contraries from recycled pulps) were then treated on the same basis since the evaluation of the fibres losses in contaminant screening systems requires the characterisation of the pulp fractionation at the different screening stages. The average behaviour of particles (fibres or contaminants of given characteristics) in an elementary height or “slice” of a screen basket, can be characterised by a passage ratio, which includes the average particle passage probabilities in the two directions through the screen plate in order to take the pulsation flow into account [12]. The passage ratio is given by:
P = Cd/Cu, where Cd is the downstream consistency and Cu the upstream consistency, with the hypothesis of perfect radial mixing on the feed and accept side of a screen cylinder (figure 79 left). The passage ratio P is is equal to the accept to reject consistency ratio Ca/Cr for for the mixed-flow model assuming also perfect mixing in the axial direction. The plug-flow model assumes no axial mixing and corresponds to single mixed-flow elementary screens in cascade, with feed forward second stage accepts (figure 79 right). The performances of screens can be predicted according to these flow models assuming no interactions between the suspended particles and constant particle passage ratios P x over the screen plate, the subscript x identifying the type of particle [12-15]. The effects of the reverse flow during the negative pressure pulse [17] can be included in the passage ratio P , which is in fact an average resulting particle passing probability, when defining instantaneous passing probabilities during the screening and reverse flow phases [73]. Standard nomenclature (figure 79) is used in the equation below, where Rv and and Rw are are the reject rates by volume and by weight and Si , Sa and Sr the the shives or contaminant contaminant contents in the inlet, accept and reject.
Inlet
Ci, Si Qi Mi = Qi.Ci
Upstream
Cu Cd
Reject
F
F
Cr, Sr Qr = Qi.Rv Mr = Mi.Rm A
Ca, Sa Downstream Qa Accept Ma = Qa.Ca
A R
R
Figure 79: General screening symbols (left) – Mixed-flow and plug-flow models (right) The equations giving the reject thickening factor ( T = Cr/Ci ) are only valid for homogeneous particles:
T = 1/(P - Rv P + Rv) T = Rv (P-1)
for the mixed-flow model for the plug-flow model
In the hypothesis of homogeneous fibres with a passage ratio P F F and contaminants with a passage ratio P K , the ratio is simply related to the standard efficiency definiti ons and to the Nelson definitions = P K / P F β = K F screening quotient Q, which found widespread use in pulp screening:
E R R = Rw Sr / Si E C C = 1 – Sa / Si Q = E C / (E C + Rw – E C Rw) = E C / E R R = 1 – Sa / Sr E R – β Rw) Rw) or Q = 1 - β R = Rw / (Rw + β – β E R R = Rw
94
Reject (removal) efficiency Cleanliness efficiency for the mixed-flow model for the plug-flow model
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
In real situation, the passage ratio distributions of the contaminants and pulp components have to be included in the calculations to establish the models of screening systems. The first step in the elaboration of the screening model is to establish experimentally the particle passage ratio distribution as a function of the screening conditions in a pressure screen cross section and for the relevant categories of the pulp components and contaminants, at least: -
the fibres in several length classes and the fines, fillers and shives,
-
the contaminants contaminants in in several several particle size classes classes and, if possible, possible, in several several classes classes of particle shape and viscoelasticity in given “size” classe.
The separation of pulp components is essentially a matter of probability separation of long-shaped particles, with some barrier separation for the removal of shives. The separation of the contaminants involves probability screening of thin particles and barrier screening of round shaped particles with possible extrusion through the slots, and includes more or less particle size and shape alteration. The fibre passage ratio distribution has to be evaluated first as a function of the fibre length to evaluate fractionation efficiency and fibre losses [24-27] and then according to fibre coarseness at least with mechanical fibres. Recent studies showed that the relation between the passage ratio P (l) and the fibre length l could be approximated by a negative exponential function [74]: P( l )
= exp− (l / λ )
β
where λ was shown to increase approximately linearly with the aperture diameter with β = 1 in the case of smooth-hole fractionation, while the best fit value was given by β = 0.5 in the case of contoured slots (used for stickies screening). screening). The evaluation of the contaminant passage ratio distribution is more difficult under real mill situation since there exists no equipment to measure the size and shape of the contaminants in recycled pulps (contrary to the situation with the fibres where various devices are available to characterise the fibres). Therefore, relevant model contaminants were used in this first study (adhesive particles in this project) to evaluate the contaminant passage ratio as a function of the screening conditions. The second step in the elaboration of the screening model is to evaluate the particle consistencies in the screen accepts and rejects for all the categories of particles, on the basis of the particle passage ratio distribution and as a function of the particle distribution in the screen inlet, of the reject flow rate and of the internal flow models. These include the plug-flow and mixed-flow models as well as other intermediate models to characterise the bulk flow inside a screen. Screening system configuration (cascade, feed forward and other systems), reject dilution (internal or external) and discontinuous operation of final stage screens are then relatively easy to include in the model with adapted software. Circuit simulation expertise and tools have been developed [75] on the basis of the commercial simulation software GII (Gensym). These tools enable the dynamic simulation of the water and particle flows in any stock preparation system or water circuit. Changing the equipment, the circuit design, the screening parameters or the pulp composition (including contaminants) is very easy once the models have been established. Typically, the output data are solid and liquid flows, particles consistencies, efficiencies and losses, as well as responses responses to input changes taking the capacities into account. An interesting interesting aspect in the development development of of screening systems systems simulation is that that the implementation implementation of other separation techniques such as cleaning and flotation can be included in the model to evaluate potential benefits and optimise the location the equipment in the system, typically on screening reject compared to the main stream. High-density cleaners, used in some cases on screening rejects to remove sand and protect screen cylinders from excessive wearing, should also remove PSA stickies. The low-density rotary cleaner offers the possibility to extract low-density stickies from screening rejects without long fibre losses [67,76]. Flotation cells have been implemented implemented on screening rejects to remove selectively hydrophobic stickies [37,38]. The simulations of screening systems performed on the basis of the experimental data gained with the pilot stickies screening, cleaning and flotation tests include high-density cleaning (since the reference PSA stickies were neutral or high-density contaminants contaminants as reported in section 4.3.2) or flotation of screening rejects (see section 5.1.2).
95
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Example of simulations of screening systems
The two 3-stage screening systems shown in figure 80 have been simulated: -
System A is a conventional conventional 3-stage 3-stage cascade system with feed-forward feed-forward second stage accepts and and feedback third stage accepts
-
System B is a 3-stage cascade system with special series feed-forward arrangement at the second and third stages, as described in [77]
Figure 80 – Simulated screening screening systems (systems A and B)
Example of screen display of CTP’s simulation platform The simulations were performed to illustrate the influence of profile height and the screening conditions were chosen among those where the most complete data were first available, i.e. with MicroVortex screen plates (see figure 65), the foil rotor and bleached chemical fibres. Table 21 gives the particle passage ratios used to simulate four different screening conditions: -
0.15 mm MV slots, high profiles (1.2 mm) and 0.19 mm MV slots, low profiles (0.6 mm)
-
Low passing velocity (0.7 m/s) and high passing velocity (2 m/s)
-
Bleached chemical pulp mixture, 50% softwood / 50% hardwood, low consistency consistenc y (1 to 2%), with 20% fillers (P = 1) and 10% fines (postulated passage ratio: P = 0.9)
-
Hard spheres with postulated passage ratios
-
Soft spheres spheres with with postulated postulated passage ratios evaluated from previous previous stickies screening screening tests [78] [78]
-
Films with the passage ratios measured for the 0.5mm² PE films [79]
96
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Screening conditions
Screen plate
Pulp (50% BSK / 50% BHK)
0.15mmMV/1.2
0.19mmMV/0.6
Passing velocity (m/s)
0.7
2
0.7
2
Lambda Beta
3.2 0.56
8.2 0.63
3.5 0.56
10 0.69
Con Contam taminant ants
%
%
Shape ape
Dia Diam/a m/area
Typ Type
Mixture 1: 80% films
5
20
Spheres
0.15mm
Hard
0
0
1
1
5
20
Spheres
0.20mm
Hard
0
0
0
0
Mixture 2: 20% films
5
20
Spheres
0.30mm
Soft (PSA)
0.12
0.23
0.23
0.6
5
20
Spheres
0.40mm
Soft (PSA)
0.04
0.06
0.12
0.23
80
20
Films
0.5mm²
PE
0.14
0.27
0.022
0.066
Table 21 – Hypothesis on particle passage ratios used for the screening system simulations
The hard and soft spheres were treated as barrier contaminants. With the soft particles the passage ratios were evaluated from previous stickies screening tests performed with the 0.15mmMV/1.2 slots. The effect of contour height was not taken into account in the postulated passage ratios of spheres. The fibre passage ratio distribution curves characterised by λ and β were used to calculate the fibre consistency distributions in length. Fibre coarseness functions were determined for the softwood and hardwood fibres [80] and included in the simulation in order to evaluate the real reject consistencies (in weight) and predict correctly the fibre losses. The cleanliness efficiencies and the total losses of the screening systems were calculated for the two contaminant mixtures and for the soft spheres only.
100 s s 90 o l m 80 e t s 70 y s g 60 n i n e 50 e r c S 40 r o y 30 c n e 20 i c i f f E 10
100 s s 90 o l m 80 e t s 70 y s g 60 n i n e 50 e r c S 40 r o y 30 c n e 20 i c i f f E 10
Ec (%) Mixture 80% films Ec (%) Mixture 20% films Ec (%) Soft contaminants Rw (Kg/t) Screening loss
Ec (%) Mixture Mixture 80% films Ec (%) Mixture Mixture 20% films Ec (%) Soft contaminants Rw (Kg/t) Screening loss
0
0 0
5
10
15
0
20
5
10
15
20
Single Single scre en reject flow rate (%)
Single screen reject flow rate (%)
Figure 81: Screening system A: 0.15mmMV/1.2 slots (left) versus 0.19mmMV/0.6 slots (right)
An example of simulation results obtained with system A (mixed flow model at 0.7 m/s slot velocity) is given in figure 81 showing the effect of the reject flow rate (supposed to be equal at each stage) on the total system losses (Kg/t) and on the cleanliness efficiencies calculated for the two contaminants mixtures and for the soft particles. Compared to the 0.15mm high-profile slots (figure 81 left), the 0.19mm low-profile slots (figure 81 right) showed about the same losses and a higher efficiency with the contaminant mixture containing 80% films, which was consistent with experimental results [78 ]. At low reject rate, the efficiency tended towards the fraction of barrier contaminants contaminants (P = 1), i.e. 40 and 10% with the 0.15mm slots and 20 and 5% with 0.19mm slots, for the mixtures 1 and 2 respectively. The simulated curves with system B looked very similar to those with system A, except that the range of system losses was about two times higher and the efficiency level slightly higher for a given single screen reject flow rate. The efficiencies of both systems were compared at the equal losses of 2%. System B was slightly better with plug flow only and the screening efficiency was improved as: -
the internal internal flow in the screens was closer to the plug-flow model than to the mixed-flow mixed-flow model, model,
-
the screens were operated at low passing velocity
97
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.2.5. Conclusions and perspectives The experimental and numerical simulation studies about the extrusion of adhesive particles through slots, both under steady and unsteady pressure, have contributed to develop the knowledge about the specific behaviour of stickies in slot screening systems. Stickies particles up to 3 times the slot width, in the case of soft PSA stickies such as acrylic adhesives after soaking, can be extruded through the slots under screening temperature and maximum extrusion pressure observed in mills, as far as the high pressure is applied continuously, which is not the case. In real situation the pressure pulses generated by the rotor ahead of the foils are very short, so that only stickies particles up to 1.2 times the slot width should be able to pass the slot during a single pressure pulse. Larger particles, namely string-like stickies, might however be able to pass the slots, according to some multi-step extrusion process, which should take place over several rotor revolutions. Concerning Concerning the optimisation of low-consistency screening, the pilot screening confirmed the effects of the main parameters tested on lab scale, e.g. higher efficiency at reduced temperature and with the hot-melt rubber adhesive compared the softer soaked acrylic adhesive material. Concerning the optimisation of screen plate design, better results where effectively obtained with a milled screen plate compared to a wedge wire screen plate with similar contour design, which was attributed to less stickies extrusion. However, the main conclusion of the comparative tests with other wedge wire screen plates with different contours was that a reduction of the height or angle of the contours led to much higher efficiency gains than using milled screen plates. This means that particle contacts and slip phenomena over the inclined contour walls have more impact than particle extrusion through slots, as these phenomena contribute to remove the stickies before they even reach the slot inlet. Practically, when comparing only screen plates with equal slot width (0.15 mm), the best results were obtained with the lowest contours, i.e. 0.6 mm height and 13 to 17° angle, while a reduction of the contour angle from 25° to 20° improved the efficiency with the medium contour height of 0.9 mm. Finally, it should be kept in mind that the screening tests were performed performed at relatively low consistency, using a conventional low-consistency foil rotor. Conclusions might be different about coarse screening in the high-consistency range (a key screening step as soft stickies should be removed as early as possible in the recycling line to minimise their fragmentation) since rotors producing higher pressure pulse are normally used at this step and because concentrated fibres and flocs probably transfer particle separation phenomena to the slot inlet and should contribute to increase stickies extrusion. Concerning high-consistency screening, there have been some technical problems with the pilot screen prototype where a rotary screen cylinder and specially designed static foils were implemented. Indeed, it has not yet been possible to conclude about the advantages of the rotary screen cylinder technology compared to the conventional technology with rotor and static screen cylinder, because of excessive by-pass flows at the hydrodynamic hydrodynamic sealing. The expected benefit was a lower fragmentation fragmentation of the stickies in order to remove them, as large particles, as early as possible in the deinking line. Stickies fragmentation during high-consistency screening was however evaluated in the framework of both deinking mill sample analyses and pilot plant trials with the weakest PSA stickies. The results were consistent and showed a clear but however limited fragmentation of stickies. The pilot screening tests performed with the conventional screening technology also allowed to evaluate the influence of some major high-consistency screening parameters. Future work should concentrate on high-consistency screening, in order to confirm the first findings and to develop the understanding of the particle separation and stickies extrusion phenomena at high fibre consistencies. The rotary screen cylinder technology should also be further investigated, investigated, as far as the technical problems can be solved, and because the basic idea of reducing the high-shear zones to minimise the fragmentation of stickies is still regarded as promising. Future work on pressure screening will continue, after this project, not only at CTP, but also at ITM (Institute of Thermal Machinery, Technical University of Czestochowa) through numerical simulation studies which were engaged further to the cooperation developed between LEGI and CTP. Numerical simulation should allow to further develop the understanding of the stickies extrusion phenomena, though the problem is very difficult, namely because of the complex models to be used for the rheological properties of adhesive materials and because particle slip conditions are fairly unknown.
98
L SCREEN C CLEAN EAN
4.3.
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Centrifugal cleaning
4.3.1. Background and objectives Centrifugal cleaning used to be a common and effective technique to remove high-density as well as low-density macro-stickies in recycled pulps. Cleaning has however become much less competitive with respect to pressure screening due to recent progress in slot screening. In addition, stickies from pressure sensitive adhesives adhesives are known to be very difficult to remove because of their density close to the pulp density. Most of the PSA stickies in deinking raw material are high-density particles and should thus be removed by high-density cleaners. By contrasts, low-density stickies from hot-melt glues to be removed by low-density cleaners, are more common in the field of packaging. Progress achieved in centrifugal cleaning since cleaners became, in the 1950’s, commonly used in the pulp and paper industry has mainly concentrated on low-density cleaning, while high-density cleaners still have about the same design as the conventional hydrocyclone invented in 1891. Figure C1 illustrates the different types of low-density cleaners developed for the removal of low-density contaminants, i.e. particles with lower density than the liquid phase of the pulp suspension, which tend to float to the surface under gravity acceleration (g) or to migrate to the vortex core under the high centripetal acceleration ( γ) in the cleaner [36]. The first reported applications of low-density cleaning were combination cleaners [81] and reverse cleaners [82], which were installed in 1969 in deinking mills. The through-flow cleaner [83] and the rotary cleaner [34, 84] have been developed later and marketed in the early 1980’s. Combination cleaners, or core bleed cleaners, are conventional highdensity cleaners equipped with an additional low-density contaminant outlet in the vortex core. These combined versions of conventional cleaners are normally operated at a higher feed pressure and provide additional removal of air and low-density contaminants, with low additional equipment costs. However, the efficiency is limited with low-density contaminants since they have to cross the accept streamlines to be removed in the air core [34]. Under these conditions, the time available for lowdensity contaminant cleaning cleaning in the upward accept flow is extremely short.
Combination Cleaner LR
Reverse Cleaner
A
F
Through-flow Cleaner
LR
F
F
HR LR
Rotary Cleaner
A A
LR
Figure 82 : Low-density cleaners. Working principle of the core-bleed cleaner, reverse cleaner,
through-flow cleaner and rotary cleaner (F: feed, A: accepts, HR/LR: heavy/light rejects). The design criteria are the same for the optimisation of the efficiency of reverse and forward cleaners. A simplified cleaning index (I = γ.T/D) was proposed [85] and shown to give a rough prediction of the efficiency level of conventional cleaners [35-36]. In this cleaning index, the cleaner head diameter (D) is directly related to the migration distance which a particle has to cover with respect to the fluid to escape from the accept flow, during the residence time (T) and with a radial slip velocity proportional to the radial acceleration ( γ). According to the Stokes law, the radial slip velocity U = (d²/18 µ).(ρs - ρ).γ is primarily determined by the fluid to particle density difference ( ρs - ρ) and depends also strongly on the particle size (d) as far as the laminar flow regime applies, which is generally the case in cleaners with small stickies having small radial slip velocity.
99
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Practically, small head-diameter forward cleaners showed to achieve the highest efficiencies with small contaminants, as far as they were correctly designed [35], which led to test some high-density cleaners with small diameter (65 to 130 mm) at CTP. By contrast, other studies [86] as well as some preliminary preliminary performed at PTS with stickies, revealed a better efficiency with a large-diameter large-diameter (270 mm) cleaner having ceramic walls compared to other cleaners with synthetic walls. The objectives in the research programme devoted to cleaning in this project were more particularly: 1. to determine the the influence of basic basic centrifugal cleaning cleaning parameters, parameters, including including stickes density and and ways to alter it, in order to optimise the stickies removal efficiency with conventional conventional cleaners, 2.
to investigate the influence of cleaner design parameters, including cleaner size and wall material, in order to determine the best forward cleaning technology to remove stickies,
3. and, initially, to investigate investigate the possibilities possibilities of the rotary rotary cleaner to remove remove low-density low-density stickies at increased consistency, consistency, in order to remove such stickies as early as possible in the deinking line. Further to the findings in this project, research effort was transferred from low-density to high-density cleaning, since the reference adhesives did not lead to low-density stickies after re-pulping. Additional Additional forward cleaning tests were performed after the addition of talcum in order to increase the density of the stickies by the absorption of mineral particles and consequently to improve the efficiency as reported in a recent paper [87] about mill trials where improved macro-stickies removal was observed when talcum was added before the cleaners.
4.3.2. Stickies density The first point in the elaboration of a cleaning test programme is to evaluate the density range of the particles to remove, in order to determine the type of cleaning equipment which should be tested, i.e. high-density or low-density cleaners. Tests were performed to measure the density of the reference adhesives (table 22). The method chosen was to introduce small pieces of the adhesive material in water and then to increase or decrease the density of the liquid by adding respectively either salts or alcohol until the particles neither sink nor float. The density of the adhesive particles was evaluated by measuring the density of the fluid mixture which maximised the proportion of neutral buoyancy particles. The particle densities were determined directly on the adhesive material and after soaking several hours in water. The soaked particles were then surface dried (with a blotter), weighted, oven dried and weighted again in order to determine the amount of water absorbed during soaking. The particle densities reported in table 22 for the pure adhesive material and after soaking are consistent with the measured water absorption. The water-based acrylic adhesive is a high-density material and absorbs much more water (as expected) than the hot-melt rubber adhesive, which showed to be low-density material.
Type of adhesive
Adhesive 3 density (g/cm )
Density after Absorbed water water 3 soaking (g/cm ) (%)
Density after 3 pulping (g/cm )
Water-based Water-based acrylic E115
1.03
1.02
50
1.05
Hot-melt rubber D170
0.96
0.96
<8
1.00
Table 22: Density and water absorption of the reference adhesive materials
However, as the observation of the adhesives during the cleaning tests (mainly sinking particles) as well as the cleaning results (next section) revealed an increase of the density of the re-pulped re-pulped hot-melt rubber adhesives, it was then decided to perform additional density analyses on the macro-adhesive particles, i.e. the particles retained by laboratory screening with the 0.10 mm slots. The principle was to determine the ash content to assess the amount of mineral pigments adsorbed on the adhesives in the pulper. The density of the macro-stickies was then determined by calculation (table 22) assuming 3 a mineral pigment density of 2.7 g/cm . These results of the evaluation of the real density of the stickies after pulping suggest that no pure centrifugal separation should be observed with the neutral
100
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
buoyancy hot-melt rubber adhesive. By contrast, the adsorption of mineral pigments on the acrylic adhesive should be very positive as the particle to fluid density difference is about doubled, which should double the centrifugal centrifugal slip velocity of small adhesive particles according to the Stokes law [36]. It is important to note that the density increase caused by the adsorption of mineral pigments, and probably to a much lower extent by the absorption of inks, depends on the surface properties of the materials, the physical chemistry of the process waters and the initial thickness of the adhesive films. PSA contaminants from other sources than adhesive labels and tapes might have higher thickness than 20 µm and consequently be less sensitive to density changes caused by adsorbed pigments.
4.3.3. Hydrocyclone cleaners The cleaning tests were performed with industrial equipment, including high-density forward cleaners with different head diameters (270, 130 and 65 mm) and a low-density through-flow cleaner (100 mm head diameter). Some of these cleaners are shown in figure 83.
- Forward Forward Cleaner Cleaner 270 mm head diameter (smooth ceramic walls)
- Forward Forward Cleaner Cleaner 130 mm head diameter (spiral cone)
- Through-flow Through-flow Cleaner 100 mm head diameter
Figure 83: Forward cleaners at PTS (1) and CTP (2) and through-flow cleaner at CTP (3)
The effects of the basic centrifugal cleaning parameters were studied at PTS with the two reference adhesives. PTS also investigated means to selectively increase the density of stickies by adsorbing mineral adsorbents in order to improve their removal with cleaners. Four different cleaners were tested at CTP with the two reference adhesives to investigate the influence of cleaner design parameters and finally to compare the possibilities of centrifugal cleaning compared to fine slot screening. screening.
4.3.3. 4.3.3.1. 1.
Stick ies and cleaner operati ng parameters
The high-density cleaning tests at PTS were performed with the large 270 mm head diameter cleaner, which showed to be effective according to preliminary tests. The INGEDE method n°4 was used to evaluate the stickies removal efficiencies by particle size classes. The effects of the main cleaner operating parameters were tested in relatively large ranges, i.e. between 0.5 to 1% for the pulp consistency, 100 to 200 kPa for the pressure drop, 35 to 55 °C for the temperature and 2 to 12 % for the reject flow rate. The stickies were produced by re-pulping the reference adhesive labels with copy paper in a small (100 l) low-consistency pilot pulper. The macro-stickies content in total INGEDE spot area was generally 5 to 20 % higher in the accepts compared to the feed, with both reference adhesives, which suggested a slight agglomeration of the stickies in the cleaner. None of the tested parameters showed large and clear effect on the stickies removal efficiency, as illustrated in figure 84 showing the effect of pulp consistency.
101
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
efficiency of macrostickies removal in cleaners - variation stock consistency (E115) (E115) 125 stock consistency 1,0 % 100
stock consistency 0,7 % stock consistency 0,5%
75 50 ] % [ y 25 c n e i c i 0 f f e r e n a -25 e l c
1 00 00...150
150...200
200...500
500...1000
1000...1500
1500...2000
2000...5000 5000...50000
-50 -75 -100 -125 particle size [µm]
Figure 84: Effect of consistency on macro-stickies removal efficiency (270 mm cleaner)
The increase of the temperature up to 55°C allowed to reach, with the acrylic adhesive, the highest average macro-stickies macro-stickies reduction by 15%, which must be regarded as low. By contrast, higher stickies removal efficiencies efficiencies were reported in mills, which were attributed to the mineral fillers [38, 87]. Additional cleaning tests were performed performed after the addition of mineral particles under the conditions which should, as much as possible, increase the density of the particles by the adsorption of a heavy 3 layer of mineral particles. The density of minerals range between about 2.7 g/cm for kaolin and 3 calcium carbonate and 4.2 g/cm for titanium dioxide [88-89]. The adsorbents that come into question include minerals which are commonly used in the paper industry, i.e. kaolin, calcium carbonate, talc and zeolite. These minerals are characterised and differentiated mainly by the following parameters: average particle size, specific surface area, density, surface energy and charge [87, 90]. The zeta potential of the minerals is frequently highly dependent on the pH. The average particle size is usually less than 1 µm. A low average particle size, a high specific surface area which should be hydrophobic, hydrophobic, if possible, and a charge character matched to the anionic charge of the finely dispersed stickies are generally considered to be advantageous for the desired complete encapsulation of the stickies. There is already proof that talc and resin particles in particular exhibit a high affinity [88-91]. This effect is selectively used to mask the stickies in pulp preparation. The influence of the adsorption of mineral particles on the final stickies particle density increase is higher, at given layer thickness and density, for small stickies, but the effect in terms of radial particle slip velocity increase is lower with small particles as a consequence of the Stokes law [92]. The mineral adsorbents were added to the stickies-containing pulp in two different ways: -
adding a hydrophobic hydrophobic mineral (APIPHOB) into the the chest, i.e. mixing mixing it into the diluted stock,
-
adding talc directly directly into the high-consisten high-consistency cy stock in the pulper during recovered paper pulping.
The results showed a reduction in the macrostickies area by approx. 25% with talc. A volume-related reject rate of 5.5% was measured in this case. This means that the macro-stickies reduction was significantly higher than the reject rate. This trial was the only trial in which a reduction in the macrostickies load in fact occurred. The adsorption of talc in the pulper with high turbulence and high stock consistency produced much better results than adding the mineral to the low-consistency stock in the chest. In the trials conducted with APIPHOB, the macrostickies area at least did not increase, as was the case in the trials without mineral adsorbents. At least agglomeration of adhesive particles in the cleaner was apparently prevented.
102
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
efficiency of macrostickies removal in cleaners - addition of minerals (E115) 125 APIPHOB/chest 100
talkum/pulper
75 50 ] % [ y 25 c n e i c i 0 f f e r e n a -25 e l c -50
10 10 0. 0. .... 15 15 0
1 50 50 .... .2 .2 00 00
2 00 00 .... .5 .5 00 00
5 00 00 .... .1 .1 00 00 0
1 00 00 0. 0. .... 15 15 00 00
1 50 50 0. 0. .... 20 20 00 00
2 00 00 0. 0. .... 50 50 00 00
5 00 00 0. 0. .... 50 50 00 00 0
-75 -100 -125 particle size [µm]
Figure 85: Effect of mineral adsorbents on macro-stickies macro-stickies removal efficiency (270 mm cleaner)
The results in figure 85 demonstrated that effects became visible in particular in the case of larger particles. When talc was used, a marked macro-stickies reduction was achieved in particular in the upper three size classes. As these size classes (1 to 5 mm) were precisely those where most of the macro-stickies were found after pulping, the effect of the addition of talc was significant on the global stickies reduction (about 25%) after cleaning.
4.3.3. 4.3.3.2. 2.
Cleaner Cleaner design parameters
Two test series were performed at CTP under different stickies cleaning conditions. The aim of the first test series was to compare high-density and low-density cleaners with the two reference adhesives. The second test series was focused on the optimisation of the removal of the acrylic adhesives with a new cleaner prototype. The stickies removal efficiencies were determined on handsheets in order to take all the stickies in the visible particle size range into account.
First cleaning test series
The reference adhesive labels were stuck onto fresh newsprint to produce dark adhesive particles and re-pulped with a mixture of 50 % newspapers and 50 % magazines in a high-consistency lab pulper, at 50°C for 20 minutes with standard INGEDE deinking chemistry. The adhesive containing pulp was then mixed with re-pulped copy paper to make the adhesive particles easier to count in handsheets, since the unprinted copy paper had much higher brightness than the printed newspapers and magazines. The adhesive content of the pulp was about 2 % in the pulper and 0.2 % in the cleaners. Both high-density and low-density low-density cleaners were selected for the tests with respectively the acrylic and the hot-melt rubber adhesives, as a consequence of the adhesive density measurements: -
A small-size small-siz e and a medium-size forward cleaner with respectively 80 and 130 mm head diameter.
-
A small size through-flow cleaner with 100 mm head diameter.
However, as the observation of the re-pulped hot-melt rubber adhesives (low-density) showed that only few particles were floating while most of the particles were sinking, it was decided to test both the forward cleaners and the low-density through-flow cleaner with this adhesive, and to limit the tests to low consistency conditions as poor efficiency was expected.
103
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The tests were performed at 0.5 and 1 % consistency with the acrylic adhesives (forward cleaners) and at 0.5 % consistency only with the hot-melt rubber adhesives (three cleaners). All the cleaners were tested at about 50 °C, at 100 and 200 kPa pressure drop and at 6 and 12 % reject flow rate (Rv). The maximum pressure drop tested with the through-flow cleaner was 180 kPa. With the forward cleaners, reject thickening showed to decrease as reject flow rate or consistency were increased and as pressure drop was decreased, as usually observed [4]. The 80 mm diameter cleaner showed a higher reject thickening (C Rejects/CFeed = 2.0-2.3 at 100 kPa and 2.5-3.3 at 200 kPa) than the 130 mm cleaner (1.5-2.0 at 100 kPa and 1.7-2.6 at 200 kPa), which is commonly observed with small cleaners compared to medium-size cleaners. In addition, the 130 mm cleaner was equipped with a spiral cone designed to reduce reject thickening. With the through-flow cleaner, the light reject outlet is located at the tip of the cleaner in the central zone of the annular accept outlet in such a way to collect the light particles and the air concentrated in the vortex core, while the fibres are centrifuged. A decrease of the reject thickening factor was observed as reject flow rate was decreased and/or pressure drop was increased, as normally observed with low-density cleaners. The cleaning systems in deinking mills are multistage cascades systems, with a number of stages which depends on the reject rate. Typically cleaning systems with the tested cleaners would be: -
High-density cleaner
130 mm head diameter
Rw = 10 to 23 %
3 to 4 stages
-
High-density cleaner
80 mm head diameter
Rw = 13 to 36 %
4 to 5 stages
-
Low-density cleaner
100 mm head diameter
Rw = 0.7 to 3 %
2 to 3 stages
The average stickies removal efficiencies (cleanliness efficiency defined by E c = 1 – (S A / SF), where S is the surface area of adhesive particles per gram pulp measured in the feed and accept pulps), were very low, as observed at PTS with the 270 mm head-diameter high-density cleaner. Further analyses were therefore limited to the tests at low-consistency (0.5%) and high pressure drop (200 or 180 kPa), which showed slightly better efficiency. The results are summarised in figure 86 showing the average efficiency curves at 6 and 12% reject flow rate.
Water-based acrylic adhesives (E115)
Hot-melt based rubber adhesives (D170)
Efficiency (%) / Particle size (mm)
Efficiency (%) / Particle size (mm)
30
20
20
10
10
0
0
-10
-10
HD 130mm
-20
HD 80mm
-20
HD 130mm HD 80mm
-30
-30
LD 100mm
-40 0, 0
0, 5
1, 0
1, 5
2, 0
0, 0
0, 5
1, 0
1, 5
2, 0
Figure 86: Average cleaning efficiency with the E115 and D170 adhesives versus particle size
With the acrylic adhesive particles, the efficiency is very low but in general positive, as expected with high-density particles close to neutral buoyancy (adhesive material density 1.02 after soaking). The efficiency increases with particle size (according to the Stokes’ law) and seems to reach an optimum before decreasing as particle size is further increased. The decrease of the efficiency after the optimum is attributed to centripetal forces caused by contacts between the particles and the cleaner wall and to the shear-induced particle spin (lift forces), especially in the reject area.
104
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Similar curves showing a maximum fibre reject thickening factor for a given fibre length were observed in the frame of centrifugal fractionation studies and attributed to the competition between centrifugal and centripetal forces [93]. It was observed that very long fibres were not rejected, especially with the cleaners having small reject outlet tip. Increasing shear rate at the cleaner wall and reducing the diameter in the reject area should increase the centripetal forces and thus reduce the removal efficiency of large particles with low buoyancy, typically flat or long-shaped particles as well as particles with density close to the liquid phase density. The shift towards higher particle sizes of the optimum efficiency observed with the 130 mm cleaner compared to the 80 mm cleaner is probably due to a reduction of the centripetal forces caused by shear and particle wall contacts, since the 130 mm cleaner is equipped with spiral cone. The slightly higher efficiency achieved by the 80 mm cleaner with the small adhesives is attributed to the higher centrifugal separation potential generally observed with small low-capacity cleaners [4, 35]. It is interesting to note that the efficiency curves achieved with the larger 270 mm head-diameter cleaner (figures 84 and 85) do not show the efficiency drop observed for large stickies with the smaller cleaners, especially with heavier stickies (figure 85), which is consistent with the fact that the shear forces in the reject area should be lower with larger reject tip. With the hot-melt based rubber adhesives D170, the average removal efficiency is very poor with the low-density through-flow cleaner and even negative with the high-density cleaners, which suggests that the adhesives are low-density particles with very close to neutral buoyancy. A more detailed analysis of the curves in figure 86 (right) shows that the efficiency is very low but positive with the small particles whose separation is mainly governed by the Stokes’ law. Consequently, the average density of these adhesives must be very close to the fluid density, with probably a certain density distribution, i.e. some low-density particles giving a positive efficiency with the through-flow cleaner and some high-density particles giving a positive efficiency with the forward cleaners. With increasing particle size, the negative decreasing efficiency observed with the forward cleaners and the positive increasing efficiency observed with the through-flow cleaner are both explained by the increase of the centripetal forces on the adhesives, since wall contacts and lift forces increase with particle size. It is interesting to note that the decrease of the efficiency of the forward cleaners starts at lower particle size with the 80 mm cleaner compared to the 130 mm cleaner (reduced lift forces) and with the hotmelt rubber adhesive (neutral buoyancy) compared to the acrylic adhesive, which confirms the proposed analysis about the competition of centrifugal and centripetal forces.
Second cleaning test series
Stickies cleaning tests were performed with the acrylic adhesive only, under the same stickies and pulp preparation conditions than in the first test series, except that the adhesive containing pulp was mixed to bleached kraft pulp (50 % softwood, 50% hardwood, which was used for the screening tests) instead of re-pulped copy paper. The medium-size high-density cleaner tested previously was used as a reference cleaner (typical cleaning conditions observed in deinking mills), and compared to a new, very small-size, high-density cleaner prototype, delivered by the same equipment supplier: 3
-
Reference cleaner: 130 mm head diameter, 12 to 17 m /h capacity (commercial cleaner)
-
New cleaner:
3
65 mm head diameter, 4 to 5 m /h capacity (prototype) (prototype)
The two cleaners were tested at the same cleaning conditions, i.e. 0.7 % pulp consistency at 50°C, 100 and 200 kPa pressure drop and 6 and 12 % reject flow rate. Reject thickening showed to decrease as the reject flow rate was increased and the pressure drop was decreased, as usually observed with forward cleaners. The new 65 mm cleaner showed higher reject thickening than the 130 mm cleaner, which is commonly observed with small cleaners compared to medium-size cleaners. In addition, the 130 mm cleaner was equipped with spiral cone designed to reduce reject thickening, while the 65 mm cleaner was equipped with a conventional smooth cone. The cleaning efficiencies in figure 87 revealed a higher efficiency of the new 65 mm cleaner prototype for the small stickies while the 130 mm reference cleaner achieved better results with large stickies. Cleanliness efficiency was higher at 12 % reject flow rate compared to 6 %, as observed usually. The efficiencies at 200 kPa pressure drop were generally higher than at 100 kPa pressure drop as usually, except for the large stickies particles where the efficiency was lower at the higher pressure drop. The
105
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
analysis proposed to explain these results is as previously the competition between centrifugal and centripetal forces. The efficiency first increases with particle size, according to the Stokes’ law and reaches an optimum before decreasing as particle size is further increased (except for one curve). The efficiency decrease after the optimum is attributed to centripetal forces caused by particle contacts with the cleaner wall and by the shear-induced shear-induced particle spin (lift forces), especially in the reject area.
Cleaner 65 mm - Efficiency (%) / Stickies size (mm)
Cleaner 130 mm - Efficiency (%) / Stickies size (mm)
80
80
70
70
60
60
50
50
40
40
30
30
20
20
10
10
0
0 0,0
0,5
1,0
1,5
2,0
0, 0
0,5
1,0
1, 5
dP =1 =100k Pa Pa Rv Rv= 6% 6%
dP =1 =100k Pa Pa Rv Rv= 12 12%
dP =1 =100k Pa Pa Rv Rv= 6% 6%
dP =1 =100k Pa Pa Rv Rv= 12 12%
dP =2 =200k Pa Pa Rv Rv= 6% 6%
dP =2 =200k Pa Pa Rv Rv= 12 12%
dP =2 =200k Pa Pa Rv Rv= 6% 6%
dP =2 =200k Pa Pa Rv Rv= 12 12%
2,0
Figure 87: Cleanliness efficiencies versus stickies spot size in handsheets (E115 acrylic adhesive)
New 65 mm head diameter cleaner (left side) – Reference 130 head diameter cleaner (right side)
The shift towards higher particle sizes of the optimum efficiency observed with the 130 mm cleaner compared to the 65 mm cleaner is probably due to a reduction of the centripetal forces caused by shear and particle wall contacts, since the 130 mm cleaner is equipped with spiral cone. At the lower pressure drop, no efficiency drop was observed with large stickies, probably because of the higher effectiveness of the spiral cone in rejecting large stickies, which is also consistent with lower wall shear normally observed at lower pressure drop. The higher efficiency achieved by the new 65 mm cleaner with the small adhesives is attributed to the higher centrifugal separation potential generally observed with small low-capacity cleaners.
4.3.3. 4.3.3.3. 3.
Cleanin Cleanin g versu s scr eening
Since the stickies containing pulp used for the pilot cleaning tests was the same pulp than that used for the pilot screening tests for the optimisation of screen plate design (section 4.2.3.2), it became possible to establish rigorously, and with the most relevant PSA stickies (the water-based acrylic adhesive E115), the comparison between the best screening and cleaning conditions which had been defined in the framework of this project: -
Centrifugal cleaning with the new 65 mm head diameter cleaner prototype
-
Pressure screening with optimised 0.15 mm slots, i.e. with with low low contours contours
The goal was to establish the comparison between screening and cleaning, first of all at equal reject rate and, as far as possible, at low reject rate, since the pilot test results were established with single stage equipment, while multistage screening and cleaning systems are used in mills. The first step was to evaluate the stickies removal efficiency curves to be achieved by optimised slot screening at the same reject rate than at the best cleaning conditions, i.e. at 19% average reject rate (instead of more than 50% with the experimental slice screen, operated at 50 % reject flow rate).
106
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Simplified simulations were done on the basis of the formulae reported in section 4.2.4, for the best screen plates (0.15 mm wedge wire slots, 0.6 mm contour height, 13 and 17° contour angle average) at low passing velocity (1 m/s). The following hypotheses hypotheses were used: -
Same pulp passage ratio, which was determined at a first screening stage, for all the reject treatment stages. This means that reject rates will be underestimated, underestimated, since long fibres with lower passage ratio concentrate in the rejects.
-
Experimentally Experimentally determined determined stickies passage passage ratio distribution, distribution, assuming that all all stickies in a given particle size class have constant passage ratio in the reject stages, which is not the case as thick stickies should concentrate in the rejects compared to flat-shaped stickies producing the same spot size in handsheets. This means that stickies reject efficiencies will be underestimated
Handsheet analyses St ic k ies s iz e c las s es St ic k ies mean spot s iz e
(mm²) 0.010.01-0. 0.020.02020.02-0. 0.040.04040.04-0. 0.08 08 0.080.08-0. 0.150.15150.15-0. 0.300.30300.30-0. 0.400.04400.04-0. 0.60 60 0.600.60-1. 1.0 0 1.0-2 1.0-2.0 .0 (mm) 0, 119 0,168 0, 238 0, 331 0, 461 0, 589 0, 700 0, 880 1, 189
1,778
2.02.0-5. 5.0 0
Test EC AA1 AA2 AA3 AA4 AA5 AA6
0.15 MF 06 32 13° Vr (m/s ) Vp (m/ s ) 23 1 23 2 23 3 15 1 15 2 15 3 Averag Average e 1 m/s
Pf 0,54 0,72 0,86 0,58 0,78 0,93 0,56
0,119 -11 -23 -2 -5 -4 1 -8
0,168 -2 -18 8 5 -3 1 2
0, 238 9 6 17 17 2 22 13
0, 331 11 -22 13 12 -14 5 11
0, 461 32 17 24 39 10 14 35
0, 589 55 38 37 71 44 7 63
0,700 87 64 56 82 57 45 84
0, 880 97 94 89 97 92 69 97
1, 189 100 98 99 97 100 94 99
1, 778 100 100 100 100 100 100 100
Test Ec CC1 CC2 CC3 CC4 CC5 CC6
0.15 MF 06 26 17° Vr (m/s ) Vp (m/ s ) 23 1 23 2 23 3 15 1 15 2 15 3 Averag Average e 1 m/s
Pf 0, 63 0, 83 0, 94 0, 71 0, 91 0, 96 0,67
0,119 -6 -17 -17 -13 2 7 -10
0,168 -1 1 -9 -2 11 -5 -2
0, 238 16 14 -6 -1 8 9 8
0, 331 -7 -15 -4 3 -8 -1 -2
0, 461 32 13 16 17 15 8 25
0, 589 55 46 17 71 27 25 63
0,700 77 55 43 73 44 15 75
0, 880 95 88 76 96 74 62 95
1, 189 99 91 97 97 98 90 98
1, 778 100 100 100 100 100 100 100
0, 69
-9
0
10
5
30
63
80
96
98
100
Av Ec 0632/0626
1 m/ s
Pk: Stick ies passage ratio Pf: Pulp passage ratio Ec: Cleanliness Cleanliness efficiency efficiency Mixed-flow model: Pk = Pf (1 - Ec /(Ec + Rv (1 - Ec) / (Pf- PfRv + Rv))) Mixed-flow Mix ed-flow model: Er = Rw / (Rw + b – b Rw) Rw) Ec = 1 - (1-Er)/(1-Rw) Plug-flow model: Er = Rw ^ β Ec = 1 - (1-Er)/(1-Rw) Average MF 0632 / 0626 St ic k ies mean spot s iz e Pk 1 m/ s Ec Mix Rw 19% 1 m/ s Ec Plug Rw 19% 1 m/ s Ec a v. Rw 19% 1 m/ s
Pf 0, 69 0, 69 0, 69 0, 69
0,119 0, 80 -3 -5 -4
0,168 0, 69 0 0 0
0, 238 0,57 4 7 5
0, 331 0, 63 2 3 2
Er: Reject efficiency efficiency
0, 461 0, 40 12 24 18
0, 589 0,18 35 57 46
0, 700 0, 09 56 76 66
Reject rate rate β: Pk/P f Rw: Reject
0, 880 0, 02 89 95 92
1, 189 0, 01 95 98 96
Cleaner 65 mm head diameter St ic k ies s iz e c las s es (mm²) 0.010.01-0. 0.020.02020.02-0. 0.040.04040.04-0. 0.08 08 0.080.08-0. 0.150.15150.15-0. 0.300.30300.30-0. 0.400.04400.04-0. 0.60 60 0.600.60-1. 1.0 0 1.0-2 1.0-2.0 .0 St ic k ies mean spot s iz e (mm) 0, 119 0,168 0, 238 0, 331 0, 461 0, 589 0, 700 0, 880 1, 189 100 k Pa Rv 6% Rw = 17 57 40 37 40 37 35 41 36 37 100 k Pa Rv 12% Rw = 30 62 50 40 55 51 56 58 48 50 200 k Pa Rv 6% Rw = 22 62 44 37 50 43 49 48 38 26 200 k Pa Rv 12% Rw = 46 61 55 54 58 56 70 65 56 49 Ave ra ge Rv 6% Rw = 19 60 42 37 45 40 42 45 37 32
1,778 0, 00 100 100 100
2.02.0-5. 5.0 0
1,778 18 52 6 26 12
Table 23 : Experimental stickies screening and cleaning results and calculations to compare
optimised screening (0.15 mm slots) to cleaning (65 mm cleaner) at equal reject rate (19%)
Table 23 gives the experimental results and calculated data in several stickies sizes classes. The tables included from top to bottom in table 23 give respectively the following data:
107
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
-
Experimentally Experimental ly determined cleanliness efficiency efficienc y values for respectively respectivel y the 0.15 MF 06 32 13° and the 0.15 MF 06 26 17° screen plates at different rotor and passing velocities (Vr and Vp).
-
The average average values of of the efficiencies of the two screen screen plates, plates, at 1 m/s passing velocity.
-
The formulae formulae used for the calculations calculations of the stickies passage passage ratios ratios in the different different stickies stickies size classes on the basis of the experimental results (the slice screen follows the mixed flow model) and for the calculation of the cleanliness efficiencies with the mixed-flow and plug-flow models.
-
The cleanliness cleanliness efficiencies calculated at 19% reject rate in the different different stickies stickies size classes with the two flow models and the average values of an intermediate model.
-
The cleanliness efficiencies achieved with the new 65 mm head diameter cleaner prototype under different cleaning conditions and on average at 150 kPa pressure drop and 19 % reject rate.
100
) % ( y c n e i c i f f e s s e n i l n a e l C
100
) % ( y c n e i c i f f e s s e n i l n a e l C
80 Screening 0.15mm slots low profile
60
Cleaning 65 mm diam. cleaner
40
20
0
80
60
40
20
0 0
1 2 Stickies spot size in handsheets (mm)
3
0, 1
1, 0 Stickies spot size in handsheets (mm)
10, 0
Figure 88 – Comparison of screening and cleaning efficiency curves with acrylic PSA stickies at 19 % reject rate (low-contour 0.15 mm wedge wire slots at 1 m/s passing velocity compared to a 65 mm head-diameter head-diameter cleaner at 100 kPa pressure drop and 0.7 % consistency)
The screening and cleaning efficiency curves achieved with acrylic stickies under these conditions are compared in figure 88, in arithmetic and logarithmic scales. The main conclusion is that high-efficiency high-efficiency small forward cleaners can become more efficient than screens operated at low passing velocity with optimised 0.15 mm slotted screen plates, for stickies spot sizes of less than 0.6 mm, while screens are clearly more efficient for the larger stickies.
4.3.4. Rotary cleaner
Background
The rotary cleaner offers, compared to other low-density cleaning (figure 82) the possibility to work at increased pulp consistency consistency while keeping a high removal efficiency of low-density contaminants, as far as the production is not increased [34, 84]. The reject rate of rotary cleaners is normally extremely low in one single stage, because of the large fibre centrifugation volume and time available in the central zone of the rotating body. The very high efficiency of the rotary cleaner is due to the high residence time and acceleration in the separation zone. However, as both reference adhesives were shown to be or to become high-density stickies particles after re-pulping, it was decided to reduce the programme initially planned at CTP with the rotary cleaner. The agglomeration of stickies in low consistency equipment was reported recently. It was observed in flotation deinking equipment [94] and in rotary cleaners used in a folding boxboard mill [95]. The results reported in section 4.3.3.1 also suggested a slight agglomeration agglomeration of stickies to take place in the large forward ceramic cleaner tested at PTS. The clear stickies agglomeration agglomeration observed with the rotary cleaners in the recycling mill opens several questions about the stickies agglomeration agglomeration mechanisms in
108
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
cleaners and more particularly in rotary cleaners. The main idea was that high-density micro-stickies and small macro-stickies which would agglomerate in the rotary cleaner could then easily be removed by subsequent slot screening, especially in the case of the rotary cleaner which can be operated at increased consistency consistency and would avoid too high dilution at the screening step. Theoretical aspects about stickies collision rates and possible stickies agglomeration mechanisms in cleaners have been proposed recently [67]. The analysis was based on gradient collision of spherical high-density stickies particles, which concentrated in the boundary layer at the cleaner wall. The initial concentration of stickies in the deinking raw material was evaluated to up to 0.1% in weight in the upper hypothesis based on the amount of PSA label stock recycled in the European mills [54, 96]. Assuming then that more than 90% of the adhesive material had been removed at the first screening steps the adhesive content of the pulp at the cleaning process step should be in the range of 10 ppm. Figure 89 (left) gives the number of adhesive particles per gram pulp as a function of the particle size for different adhesive contents in the pulp.
Number of particles / g pulp
Number of collisions of one particle with another
10 000
100
3 000
30
1 000
10
0.1 % 100 ppm
300
3
10 ppm
100
1
1 ppm
30
Hypothesis 2 (hydrocyclone)
0,3
0.1 ppm
10
Hypothesis 1 (rotary cleaner)
0,1
3
0,03
1 10
20
30
50
100
200
300
0,01 0,1
0,2
0,5
Particle diameter (µm)
1
2
5
10
20
50
100
Adhesive content in the pulp (ppm)
Figure 89: Theoretical values of adhesive particle content (left) and collision rates in cleaners (right)
The much higher theoretical number of collision which should take place in the rotary cleaner, i.e. several collisions for about 10 ppm adhesives compared to few collisions in the case of hydrocyclones hydrocyclones as shown in figure 89 (right), was mainly due to the higher residence time and stickies concentration factor in the boundary layer (about 250 assuming a boundary layer thickness of 200 µm). These theoretical analyses and the results observed in a board recycling mill [95] led to decide upon further investigation about possible agglomeration agglomeration of PSA stickies in the rotary cleaner.
Evaluation of stickies agglomeration in the rotary cleaner
Pulp samples were collected in French recycling and deinking mills at the inlet and outlets of rotary cleaners (Kadant-Lamort Gyroclean, model GYT shown in figure 90). The pulp samples were analysed to establish the pulp components and stickies balances in order to determine if significant stickies agglomeration takes place or not. The following stickies analyses were performed on the pulp samples: -
Macro-stickies according to INGEDE method n°4 (particles retained by laboratory screening with 0.10 mm slots and producing tacky spots according to the procedure)
-
Micro-stickies and dissolved & colloidal stickies defined as the DCM extractible materials of the whole pulp minus the DCM extractible materials of the fibre fraction retained after hyper-washing of the pulp sample on a 100 µm wire screen
Macro-stickies and DCM extractible materials of the fibre fraction do however not correspond to the same stickies since larger particles pass the 0.10 mm lab slot compared to the 100 µm wire screen.
109
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
OCC recycling mill
The mill produced fluting from French OCC. The Gyroclean was implemented on a long-fibre fraction. The fractionation screen was equipped with slots and implemented after the pulping and coarse screening steps. This means that most of the macro-contaminants, which passed coarse screening with 1.8 mm holes, were fed to the Gyroclean. The results of the analyses are reported in table 24. -
The reject flow rate of the Gyroclean was low (2.6 %) and the solid losses extremely low (0.05 %) and without any fibres.
-
The macro-stickies were practically not removed at the rejects (less than 1%) though the rejects were much dirtier than the feed pulp and their volume concentration in the rejects was lower than at the feed, which suggested that the macro-stickies were essentially high-density particles. The balance showed significantly less macro-stickies at the outlet compared to the inlet.
-
The micro-stickies and dissolved & colloidal stickies concentration were about two times higher in the rejects compared to feed/accept. The balance showed much more micro and dissolved & colloidal stickies at the outlet compared to the inlet.
Sampling point 3
Feed
Accept
Reject
Flow rate
m /h
580
665
15
Suspended solids
g/l BDT/h
6.20 3.60
6.36 3.60
0.13 0.002
Macro-stickies (INGEDE method n°4)
mm²/g m²/h
86.2 310
62.6 225
757 1.5
26 15.3
42 23.7
69 1.04
Micro/dissolved/colloidal Micro/dissolved/colloidal stickies mg/l (DCM whole whole pulp - DCM fibre fraction) fraction) kg/h
Table 24: Pulp flows and stickies balances at the rotary cleaner of the OCC recycling mill
The results in table 24 indicated that some contaminants, which were measured as macro-stickies (tacky particles > 0.10 mm lab slots) had been fragmented and transformed in contaminants measured as micro-stickies and dissolved & colloidal particles. One explanation could be that Styrofoam particles are submitted to the high pressure (about 1 MPa) and pressure fluctuations generated in the rotating drum of the Gyroclean and might then be “exploded” into microscopic particles, which produce fine fraction of DCM extractible materials. Such exploded Styrofoam balls become small high-density particles (polystyrene), (polystyrene), which are accepted (more micro-stickies in the accepts, as shown in table 24) while intact or less fragmented Styrofoam balls are removed at the rejects, together with “stronger” contaminants contaminants such as hot-melt glues and polyethylene films, as illustrated in figure 90.
5 mm
Hot-melt glue
PE film
Styrofoam
Figure 90: Rotary cleaner rejects of the OCC recycling recycling mill
110
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Deinking mill
The mill produced deinked pulp from wood-free raw material. The Gyroclean was implemented after flotation and followed by high-density cleaning and slot screening, the accepts of the last-stage screen being re-circulated at the inlet of the cleaning/screening system. The pulp sample analyses and stickies balances showed the following results. -
The reject flow rate of the Gyroclean was relatively higher (7 %) but the solid losses were still low (1.1 %) with very low fibres losses (0.05 %)
-
The macro-stickies were poorly removed (about 6%), the rejects were much dirtier than the feed pulp and the macro-stickies concentration in the rejects was lower than at the feed, which suggested that these stickies were essentially high-density particles, as in the OCC recycling mill. The balance also showed significantly less macro-stickies macro-stickies at the outlet compared to the inlet.
-
The DCM extractible materials are about constant for the fibre and dissolved & colloidal fractions, while more micro-stickies were found at the outlet compared to the inlet, as observed in the OCC recycling mill.
Conclusion
The analyses of the pulp samples taken at the inlet and outlets of rotary cleaners in two different mills, an OCC recycling mill and a wood-free deinking line, did not reveal any agglomeration of PSA stickies. By contrast, results have been published which showed some stickies agglomeration in the rotary cleaners implemented on different lines of a packaging paper mill. Theoretical analyses also showed that, if physical-chemical physical-chemical conditions are favourable to the agglomeration of stickies, the agglomeration agglomeration rate should be higher in a rotary cleaner compared to a conventional cleaner. It is consequently concluded that the agglomeration of PSA stickies, should first be managed through the addition of agglomeration chemistry, as developed in WP2 with the new pulping process. The mill samples analyses also suggested that most of the macro-stickies were high-density high-density particles, which cannot be removed by a low-density cleaner. A minor part of the macro-stickies (the low-density fraction) was removed at the rotary cleaner rejects and the stickies balances indicated that some macro-stickies macro-stickies were fragmented into micro-stickies. The disintegration, disintegration, under high pressure variations, of Styrofoam balls into microscopic particles might contribute to explain the phenomena at least in the case of the OCC recycling mill.
4.3.5. Conclusions and perspectives Extensive pilot cleaning tests were performed with various cleaners, including high-density forward cleaners with different head diameters (270, 130 and 65 mm) and a low-density through-flow cleaner (100 mm head diameter), to evaluate their efficiency with the two reference adhesives, a high-density 3 3 (1.03 g/cm ) water-based acrylic adhesive and a low-density (0.96 g/cm ) hot-melt rubber adhesive. The pulping conditions of the adhesive labels changed significantly the density of the adhesive particles by soaking and by the adsorption of mineral pigments as the adhesive labels were stuck onto 3 newspapers. The final in-situ density of the adhesives was then increased by 0.03 to 0.04 g/cm , which was positive for the acrylic adhesives but led to neutral buoyancy particles and very poor cleaning efficiency with the hot-melt rubber adhesive. The stickies removal efficiencies were consequently very poor on average though some cleaner operating parameters such as increasing the pulp temperature improved slightly the efficiency. There was potential to improve the separation efficiency of stickies in cleaners by selectively increasing the density difference by adding mineral adsorbents such as talc. Adding adsorbents to the stock and mixing them in the pulper subsequently led to a macro-stickies reduction in the cleaner by about 25% in the case of the large cleaner, which removed mainly the larger macro-stickies. The addition of fresh "unused" mineral seemed necessary for effective adsorption of the minerals to the stickies. The fillers already present in recovered paper did not produce a positive result on macro-stickies separation.
111
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The analysis of the efficiency curves observed with small and medium sized cleaners revealed a competition between centrifugal separation and shear-induced centripetal separation for the large stickies particles. The results obtained with these cleaners confirmed the poor average PSA stickies removal efficiency, but revealed a potential to remove stickies in the optimum particle size range. A new small-capacity small-capacity cleaner prototype (65 mm head diameter) was finally tested. The new cleaner achieved higher efficiency than optimised screening with 0.15 mm slots, for small stickies producing spots under about 0.6 mm in handsheets. Efficient high-density cleaners offer thus possibilities to remove some additional macro-stickies and micro-stickies, which cannot be removed by fine slot screening. Such cleaners could also be implemented in screening systems, namely on screening rejects, in order to reduce the cleaning costs. The effectiveness of low-density cleaning in deinking lines seems to be restricted to the selective removal of some hot-melt PSA and glues (e.g. from advertising inserts) as re-pulped adhesive labels and tapes should essentially produce high-density stickies particles. The dynamic flow conditions in rotary cleaners should normally promote the agglomeration of high-density stickies at the cleaner wall. Macro and micro-stickies balances performed in a deinking and a packaging paper recycling mill did however not reveal any agglomeration of stickies, which might have been used to remove them by subsequent subsequent fine slot screening. Further investigations about stickies cleaning should focus on the physical chemical aspects in order to develop the understanding PSA stickies behaviour such as the adsorption of mineral fillers to increase their density and improve the removal of smaller stickies with high-density cleaners or the agglomeration of micro-stickies in cleaners in order to enable the removal of the stickies agglomerates by subsequent slot screening.
112
L SCREEN C CLEAN EAN
4.4.
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Flotation
4.4.1. Background and objectives If printed recovered paper is to be used as a raw material for producing graphic or sanitary papers, the printing inks must be removed to the greatest extent possible in the recovered paper treatment plant. There are essentially two processes – flotation deinking and wash deinking – that have gained acceptance for ink removal in industrial practice. Both methods aim at increasing the brightness of the accept, enhancing the cleanliness cleanliness of the pulp and more recently reducing stickies as well [97]. In flotation deinking, the relatively small ink particles are rendered hydrophobic by suitable collector chemicals and agglomerated to a floatable size so that they can attach to finely dispersed air bubbles to form a foam, which is then removed from the suspension. Using flotation deinking for sticky separation is especially important, since the sticky particles inherently have a hydrophobic surface in most cases. This property can be put to ideal use in flotation deinking. The main topic in the Screenclean project was the stickies reduction not the ink removal. The following fundamental objectives were to be met in WP 5.1 of this project.
SWP 5.1 – Laboratory flotation tests
-
Use of the VOITH laboratory laboratory flotation cell with a useful storage volume of 25l
-
Studies on how of the following parameters impact the separation action: stock consistency, type of stickies, raw materials, process water and amount of soap
-
Use of special hydrophobic hydrophobic mineral for possible better stickies separation
-
Studies on the impact of the flotation time and the associated flotation air volume
•
Macro-stickies
-
Studies on the impact of the particle size of the stickies on the separation action (PTS Heidenau)
•
Micro-stickies and colloidal potential stickies
-
Studies on the separation ability of microstickies in lab and mill samples (PTS Heidenau)
-
Studies on the separation separation ability of colloidal potential stickies in model samples (CTP Grenoble) -
To develop a better understanding of the mechanisms governing the flotation of colloidal stickies in order to increase their removal efficiency.
-
To identify and test at the laboratory scale inorganic collector systems which can be used as an alternative to calcium soaps.
The results of the studies are intended to work out the maximum attainable level of separation of deinking flotation with respect to several stickies fractions. The necessary conditions are to be worked out as well.
SWP 5.2 – Pilot flotation tests
This task was aimed at studying the transferability into industrial practice of the results of the laboratory laboratory tests on macrostickies macrostickies flotation carried out in SWP 5.1. For this purpose, selected tests from SWP 5.1 were implemented on a pilot scale in the PTS pilot plant, using a mobile VOITH-Eco-Cell VOITH-Eco-Cell with a filling volume of 700l.
113
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.4.2. Basics of the flotation process
Introduction
The stock suspension contains fibres, fillers, ink particles and the stickies as the solid phases. It also contains aqueous chemical solutions and the gas phase in the form of air bubbles. Together, they form a multi-component system. The characteristic feature of flotation as a separation process is that the solids to be separated attach themselves to air bubbles. The separation behaviour is determined by the properties of the aggregates formed. Selective attachment of the substances to be separated to air bubbles is important, and this presupposes that the surfaces are or have been rendered hydrophobic. Basically, the flotation process can be broken down into three fundamental steps [98]: -
Aeration of the the stock suspension suspension and creation creation of the air bubble bubble size spectrum
-
Mixing the air bubbles with the stock suspension and attachment attachment of the hydrophobic substances substances to the air bubbles
-
Separation and removal of the air-bubble complex from the pulp suspension
Chemistry
The entire process depends on how effectively the ink particles are removed from the fibres and how stable the particles are dispersed in the suspension. Removing the ink from fibres is the prerequisite for flotation. Deinking by flotation (flotation deinking) is carried out throughout the world based on a more or less uniform method. The following substances are used in particular (figure 91) [97]: Dosing Dosing point point Chemical use
pulping
1st flotation
dispersing
sodium hydroxide
X
X
hydrogen peroxide
X
X
sodium silicate
X
X
oleic acid
X
X
dispersant
X
X
2nd flotation
Figure 91: Common deinking chemicals [1]
Basic parameters
Fundamental variables that influence the results of flotation are listed below (figure 92). In general • stock consistency • stock composition (fibers, ink, fillers, fillers, adhe sives) • air/stock ratio (flotation time) • siz size e of the the bubbles bubbles • flotation chemistry • fiber loss • construct construction ion of the the cell cell and the air-injector
Concerning stickies • hydrophobic surface of the particles • particle size distribution • macro- or microsti microstickies ckies (important (import ant for measurement) • propensit propensity y to floccul flocculate ate • temperature, pH
Figure 92: Parameters affecting flotation [99-104]
114
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The bubbles in modern facilities are formed by self-aspirating or pressurised air injectors or step diffusers. Mixing the aerated suspension causes the reagents to be distributed and is also responsible for the collisions between the solids and air bubbles as a prerequisite for bubble adhesion. In contemporary flotation facilities, the stock consistency plays an ever-greater role in the inlet. For reasons of capacity, the systems are often hydraulically overshot, i.e. the inlet stock consistency is far above the recommended value of 1.0%. 1.0%. Stock consistency, however, is essential for the formation of air bubble-particle agglomerates and for floating these agglomerates to the surface of the suspension.
Arrangement of flotation in stock preparation
The process layout at the left in the diagram in figure 93 is the standard setup for the production of recovered fibre based mass printing papers. Other circuit arrangements and applications have been tested within the scope of research projects [105-110]. Standard in DIP-Lines
Pulping
FineScreening
PreScreening
Fine-Screening 2nd Stage
Flotation FineScreening
Research in a paper mill Research mill (PM Kymmene,Voith AG): Stickies flotation during screening
Dissolved Air Flotation
Fine-Screening 3rd Stage Fine-Screening 4th Stage
Thickening
Flotation
Disperger Flotation
Research in lab scale (PTS, Institute of paper science scie nce Washin Washington) gton):: Stickies flotation in packaging papers
Dissolved Air Flotation
Thickening
FineScreening
Flotation
Figure 93: Deinking flotation step in the t he stock preparation system The recovered paper can be slushed either in a HC pulper or in a drum. The deinking chemicals are normally added during this process step. This is followed by pre-cleaning by HC cleaners and than by pressure screens with holes for particles separation. In some cases paper mills are using now slotted screens in this step. This is then followed by the first flotation stage for removing printing inks and then by a fine screening stage with slotted screen baskets. Recovered paper treatment plants for the production of graphic papers normally have a disperger after the fine cleaning stage to reduce optical inhomogeneities. In many cases, this process step, in which the stock has been dewatered to as much as 30% consistency, is also used f or bleaching [97]. The requirements on optical properties and cleanliness of the RCF pulp for improved grades make additional treatment stages necessary. For instance, after disperger bleaching a second flotation stage becomes necessary and then, after thickening and dewatering, a second disperger/homogenisation stage as well as a final bleaching step with dithionite [97]. The quality of the final DIP is a function of the RCF pulp that is used as well as the treatment technology and mode of plant operation. Depending on the application, important quality properties include the following: static and dynamic strength dewatering behaviour, porosity specific volume brightness and optical homogeneity ash content stickies content (micro-stickies and macro-stickies)
115
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.4.3. Deinking flotation flotation – lab and pilot equipment equipment
Lab flotation
The stock suspension was pulped for 30 min at 15-17% stock consistency and 40 °C. Pulping devices were the HOBART lab pulper or a small pilot device. A commercially available VOITH laboratory flotation cell was used for the flotation trials.
3-ways valve (trialstopcleaning
parameters consistenc stency: y: 1,0 +/- 0,8% • consi volume ume 22. 22...2 ..25l 5l • vol • air flow: 10 l/min • typical fiber loss: 8...12% • max. temperature: 65 °C
Air flow control
Foam flow Accept valve
Rotor with air injection
Figure 94: Lab flotation Equipment
Pilot flotation
The trials were conducted by means of a mobile ECO flotation cell of the company VOITH (figure 95). The cell may be operated both independently as a single unit and in the bypass of an industrial flotation system. The foam overflow is adjusted via the filling level of the cell.
Figure 95: Pilot flotation equipment (ECO cell) The system was charged with an overall suspension volume of 2200 l, 700l of which went into the flotation cell, and 1300l into the large-volume chest. The stock suspension was kept in circulation throughout throughout the flotation process. The feed pump produced a volumetric flow rate of approx. 700 l/min. Part of the volume flow was re-circulated into the supply tank via a bypass, i.e. the actual volume flow through the cell amounted to approx. 450 l/min. The entire stock volume of 2000 l was pumped through the flotation cell within approx. 4.5 min. This means the stock suspension passed the cell 5 times in the overall flotation period of 22.5 min. A serial connection of 5 flotation cells is currently stateof-the-art in paper mills.
116
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.4.4. Study of basic stickies stickies flotation flotation parameters parameters (lab flotation) 4.4.4. 4.4.4.1 1 Macro-sti cki es
Work programme
The tables below summarise the work programme of the flotation trials concerning macrostickies in lab scale. Two different pressure sensitive adhesives were studied. Both adhesives (E115 and D170) were labels that had a normal adhesive grammage of 20 g/m². All trials were conducted containing an adhesive share of 0,5% based on o.d. fibres. The adhesives had been applied as labels to the newspapers before pulping. All trials were conducted with clean water, if there were no special trial description. The following fundamental parameters were varied during flotation:
trial no. VV1+VV2 V1...V6 V7...V14 V15a, V16 V15b, V17, V18
description pre tests stock composition stock consistency macrostickies particle size f atty acid
variables stock with/without deinking chemicals DIP stock, adhesiv e, dilution water ONP, OMG, ONP/OMG* stock consistency, dilution water ONP/OMG* pulping time ONP/OMG* amount f atty acid ONP/OMG* with/witout deinking chemicals and hydrophobic hydrophobic f illers ONP/OMG*
V19...V24 hydrophobic fillers * mixture 50% ONP / 50% OMG or 50% D170 / 50% E115
adhesives D170 D170, E115 D170/E115* D170/E115* D170/E115*
s oc oc consistency inlet flotation 1,0 1,0 0,5...1,8 1,0 1,0
D170/E115*
1,0
Figure 96: Work programme – lab flotation macro-stickies The following deinking chemicals were used in the trials (standard as per INGEDE Method 11), if there were no special trial description:
• • • •
caustic soda sodium silicate hydrogen peroxide fatty acid
0.6 % (in relation to suspended solids) 1.8 % 0.7 % 0.8 %
Trial objective was also achieving further improvements in flotation efficiency by means of a special highly hydrophobic mineral. The hydrophobic mineral surface attaches to the already hydrophobic macrostickies, thus increasing their hydrophobicity. This was expected to improve the attachment of air bubbles to the stickies during flotation. Main characteristics of the mineral [53]:
• • • •
supplied as: density: particle size: pH:
white powder (hydrophobic mineral mixture) 2.74 g/cm³ 50% < 10µm; 90% < 50 µm 8,0...9,5 (at a concentration of 1g/l in water)
Dosing:
• • • •
dosing quantity: reaction time: consistency: consistenc y: pH (stock suspension): suspension):
0.3% based on o.d. fibres (to be optimised by the mill trial) at least 15 min minimum 4% (should be high in general) > 7.0
The mineral was added directly in the pulper in all trials. Because the macro-stickies content used in the project (0.5%) was higher than the common practical level, the dosing quantity of hydrophobic fillers was increased accordingly. accordingly. After pulping the water water was added to the stock stock suspension to dilute dilute it to a consistency of 1.0% and the suspension was then floated in the 25l flotation cell. The total flotation time was 15 minutes, and samples of the flotation foam were taken every three minutes.
117
L SCREEN C CLEAN EAN
Main results
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Varied Varied stock c ompositi on, chemicals and and dilut ion water
The results of flotation are compiled in Figure 97. The pulps that were used included market DIP, old newsprint (ONP) and old magazines (OMG). In addition, one trial was conducted without any deinking chemicals at all. Clean tap water and process water from a deinking plant were used prior to flotation to dilute the stock consistency from approx. 16.7% after pulping to 1.0%. The results gave evidence that macro-stickies separation separation efficiencies of more than 70% were 70% were normally achieved when clean water was used. This applied both to the market DIP as well as to the ONP and OMG. A mixture of ONP and OMG in a 1:1 ratio produces a separation efficiency somewhat lower than approx. 50%. The separation efficiency attained in the trial that used process water from a paper mill instead of clean water was much lower. A total separation efficiency of approx. 10% was 10% was achieved in this case. The action of the flotation chemicals is apparently greatly impaired by the colloidal and molecular contents of the used process water. The trial that involved using no deinking chemicals whatsoever showed no reduction in the macro-stickies area. lab flotation (25l VOITH cell) - varied stock and adhesives 100 absolute reduction
90
reduction by balance
80 ] % [ n o i t c u d e r a e r a s e i k c i t s o r c a m
70 60 50 40 30 20 10 D170
D170
D170
D170
D170
D170
E115
DIP (clean water)
ONP (clean water)
OMG (clean water)
OMG+ONP (clean water)
OMG+ONP (process water)
ONP (no chemicals)
ONP (clean water)
0 -10 -20
Figure 97 : Flotation, macro sticky reduction total and by balance (varied stock composition) composition) Figure 98 compares the macro sticky separation efficiency values in the important particle size classes for clean water on one hand and for process water on the other. Two important aspects become evident in this graph. On one hand, smaller macro-stickies are floated generally much better than large macro-stickies. macro-stickies. The smaller macro-stickies macro-stickies were separated efficiently in both cases – with clean water and with process water. On the other hand, the flotation results in the separation of bigger macrostickies when process water is used to dilute t he stock prior to flotation are generally poorer than when clean water is used for this purpose. One reason could be the interaction of the deinking chemicals with the ingredients of the used process water. In that case the chemicals will work not only at the stickies like in the trials with clean water. Another reason could be a different amount of calcium ions (water hardness) in the process water in comparison to the clean water in PTS. If the calcium concentration is too low, not enough fatty soaps will be built by reaction between fatty acid and calcium.
118
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
lab flotation (25l VOITH cell) - varied stock and adhesives 100 average - clean water 90
proces water
80 ] % [ n o i t c u d e r a e r a s e i k c i t s o r c a m
70 60
2
R = 0,86
50 40 30 20 10 0 -10
200...500
500...1000
1000...1500
1500...2000
2000...5000
-20 -30 -40 macrostickies particle size [µm]
Figure 98: Flotation D170 and E115, flotation efficiency acc. to size classes
Flotation – varied stock consis tency
One of the most important parameters in deinking flotation is the stock consistency in the inlet. The guideline here is about 1.0%. 1.0%. As was already mentioned in the introduction, deinking plants in paper mills are in many cases currently operated far above the limits for which they were originally designed. It is impossible to shorten the dwell time excessively, since sooner or later there will insufficient particles moving into the foam. This is why in most cases the stock consistency is increased in the inlet in order to force more pulp through the flotation system (throughput). (throughput). During this project, the inlet consistency was varied through a wide range to ascertain what impact this would have on the results of flotation with respect to the macro-stickies. lab flotation (25l VOITH cell) - varied stock consistency 70 absolute reduction
65
reduction by balance
60 ] 55 % [ 50 n o i t 45 c u d e r 40 a e r 35 a s e 30 i k c i t 25 s o r c 20 a m 15 10 5 0 0,50
1,00
1,25
1,50
stock consistency [%]
1,80
0,50 (process water)
Figure 99: Flotation, macro sticky reduction total and by balance (varied consistency)
119
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Figure 99 displays the dependency between absolute macro-stickies separation and pulp consistency. A stock consistency of 0.5% resulted 0.5% resulted in a separation efficiency of 50%, 50%, whereas a higher consistency of 1.8% only 1.8% only separated out 10% of 10% of the macro-stickies. The trial with the lowest stock consistency was also carried out using the process water from a paper mill. The macro-stickies separation efficiency in this case amounted to only about 17% 17% compared to about 50% when 50% when clean water was used. If the separation of macro-stickies is calculated using the exact stock balance, then higher separation efficiency values result on the whole when compared to the absolute values (figure 102). The reason is the loss of fibres and fillers during flotation as this is not included in the absolute values. The results of flotation were standardised to an inlet stock consistency of 1.0% in figure 100. The separation efficiency at 1.0% inlet 1.0% inlet stock consistency was then set at 100% 100% and the other separation efficiency values were calculated accordingly. lab flotation (25l VOITH cell) - varied stock consistency 150 trend macrostickies reduction (1,0% stock consistency = 100%)
140 130 120 ] % [ 110 n o i t 100 c u d 90 e r a 80 e r a 70 s e i k 60 c i t s 50 o r c a 40 m 30
2
R = 0,97
20 10 0 0,50
1,00
1,25
1,50
1,80
stock consistency [%]
Figure 100: 100: Flotation, macro sticky reduction trend, standardised (varied consistency) One of the most important aspects in the flotation of macro-stickies is the efficiency of flotation in the various particle size classes. Flotation is intended to separate as many small macro-stickies as possible with a particle size < 1000 µm. It is virtually impossible to separate these particles using screening technology in paper mills. Figure 101 shows the flotation efficiency in the important particle size classes as a function of stock consistency. The best separation efficiency by far amounted to approx. 70% and 70% and was achieved with the smallest macro-stickies. In this case, the stock consistency had only a minor impact on the result of flotation. Approximately Approximately 40% of 40% of the macro-stickies were separated in the next size class from 500 to 1000 µm. As the stock consistency increased, increased, the separation efficiency fell noticeably. This means that with increasing stock consistency and thus greater obstruction to the upward motion of the particles, less and only smaller macro-stickies are floated. Incidentally, the effect of discharging smaller hydrophobic particles using air bubbles is precisely the mechanism of action that serves to separate out dirt specks and fillers in deinking flotation. These particles are normally less than 200 µm in size, thus placing them in the range of smaller macro-stickies. All in all, this study establishes that deinking flotation is principally effective for smaller macro-stickies. macro-stickies. This is an ideal way of complementing complementing the separation of larger macro-stickies macro-stickies in pressure screens.
120
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
lab flotation (25l VOITH cell) - varied stock and adhesives 90 80 70 60
] 50 % [ 40 n o i t 30 c u d 20 e r 10 a e r 0 a s -10 e i k -20 c i t s -30 o r c -40 a m-50
stock consistency 0,50% 1,00%
1,25%
-60 -70
1,50% 1,80%
-80 -90 200...500
500...1000
1000 ...1500
1500...2000
macrostickies macrostickies particle size [µm]
Figure 101: Flotation, macro sticky reduction in size classes (varied consistency)
Flotation – testing testing w ith a real real process st ock sus pension
In addition to the trials conducted with the model stock suspension the laboratory cell was also used to float a stock suspension that originated from a tissue mill. The suspension had a stock consistency of approx. 1.0% and was taken immediately upstream of the flotation stage in the paper mill. So we had a direct comparison of the mill and lab flotation cells. The result of macrostickies separation in figure 102 documents the very good concurrence between the results obtained with the laboratory flotation cell and mill flotation facilities. The total efficiency of 25% achieved in the mill facility is not particularly good. The best result of flotation amounted to approx. 30-50% and 30-50% and was achieved in the two smallest particle size classes. The separation efficiency of the medium-sized macro-stickies was negative in the mill and lab flotation. Since the mass of the medium-sized macro-stickies in the mill suspension does not matter as much as that of the small macro-stickies, there was still a positive overall efficiency for the mill suspension. macrostickies flotation - comparision mill and lab results (same stock suspension) 60 paper mill
lab trial (V14)
40 20 ] % [ n 0 o i t c u d -20 e r a e -40 r a s e i k -60 c i t s o r -80 c a m -100 -120 -140 100...500
500...1000
1000...1500
1500...2000
to t a l
macrostickies particle size [µm]
Figure 102: flotation efficiency in size classes (stock suspension from paper mill)
121
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Flotation – The influence of amounts of fatty acid
The fatty acid or the soap that forms as a result of the reaction with hardness-producing hardness-producing substances substances in the water are creating the preconditions for attachment of the particles to the air bubbles and ultimately the discharge of the stickies together with the flotation foam. The test results show a general tendency for the flotation efficiency to depend on the amount of fatty acid used. Halving the fatty acid dosage to 0.4% as compared to the standard formulation led to a significant decrease in flotation efficiency from approx. 46% to 32%. An increase to 1.2% fatty acid produced only a slightly better macrostickies flotation than the standard formulation. formulation. lab flotation (25l VOITH cell) - variation amount fatty acid 100 fatty acid 0,4% fatty acid 0,8% fatty acid 1,2%
90 80 ] 70 % [ 60 n o i 50 t c u d 40 e r a 30 e r a 20 s e i k 10 c i t s 0 o r c -10 a m
-20 -30 -40 -50 100...500
500...1000
1000...1500
1500...2000
2000...2500
total
macrostickies particle size [µm]
Figure 103: Efficiency for macrostickies size classes (variation of the amount of fatty acid)
Ad di ti onal on al u se o f a s pec ial hy drop dr opho hobi bi c m in eral
As expected, nearly no macro-stickies macro-stickies separation was achieved without deinking chemicals and minerals (black column). Adding 0.4% extra minerals brought a minor increase in flotation efficiency to approx. 8%. Adding different amounts of hydrophobic minerals together with conventional deinking chemistry failed to produce any improvement in flotation efficiency when using clean water. lab flotation (25l VOITH cell) - variation deinking chemicals/addition special hydrophobic fillers 70 65
total macrostickies separation
60
macrostickies separation by balance
55 50 45
] % [ 40 y c 35 n e i c 30 i f f e 25 n o i 20 t a t o 15 l f
10 5 0 -5 -10 -15 standard/no
no/0.4%
no/no
standard/0.4%
standard/0.8%
standard/1.6%
amount of deinking chemicals/hydrophobic fillers
Figure 104: Flotation, total macrostickies reduction (variation of hydrophobic hydrophobic fillers)
122
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Visualization Visualization of f lotation results
In the pictures 105 a+b there can be seen the prepared samples for macro-stickies measurement. The differences between inlet and accept of the flotation are significant – especially in the amount of the smaller macro-stickies. In the accept mostly bigger macro-stickies have been found because the smaller macro-stickies have been removed during flotation very effectively.
Figure 105a: Flotation – inlet
Figure 105b: Flotation – accept after 12min
The high amount of smaller macro-stickies in the foam will be visualized in pictures 106 a+b. The majority of the macro-stickies separation separation took place in the first 3 minutes of the flotation. Especially the smallest macro-stickies have been removed very fast. In the next 3 minutes also some medium macro-stickies macro-stickies have been removed from the stock suspension.
Figure 106a: Flotation – foam after 3 min
123
Figure 106b: Flotation – foam after 6 min
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.4.4.2 Micro-stickies Per definition, micro-stickies are particles sized between approx. 5 and 100 µm. This corresponds to the approximate size range of ink particle agglomerates (dirt specks), whose separation efficiency by deinking flotation is very high. For this reason and because of the preferential flotation of smaller stickies observed in the test series, deinking flotation was expected to be suitable for removing microstickies from stock suspensions as well. The flotation efficiency for micro-stickies separation was determined directly by comparative measurements of mill samples taken from the inlet and outlet of industrial flotation stages. The microstickies content was determined by dimethyl formamide extractions extractions of stock consistency samples. Different types of micro-stickies are generated by adhesive fragmentation, the detachment of printing inks or disintegration of paper coatings in the pulper. Micro-stickies can also be agglomerates formed by desired or undesired adhesive, binder and pigment reactions. These agglomerates are frequently referred to as secondary stickies, and their formation is promoted by flocculant addition. All these types of micro-stickies can be found in the stock suspensions of each paper mill. Micro-stickies analyses based on dimethyl formamide extraction do not differentiate between the various different forms of microstickies because the extraction is usually limited to adhesives and binders.
deinking flotation - reduction of microstickies in paper mills 5,0 4,5 4,0
100 95 90 85 80 75 70 65 60 55 50 45
inlet accept reduction
] % [ ) 3,5 s e i k c 3,0 i t s o r c 2,5 i m ( t c a 2,0 r t x e F 1,5 M D
40 35 30 25 20 15 10 5 0
1,0 0,5 0,0 mill A I preflotation
mill A II preflotation
mill A III preflotation
mill B preflotation
mill B postflotation
] % [ s e i k c i t s o r c i m n o i t c u d e r
mill mill C flotati flotation on mill mill D flotati flotation on
Figure 107: Micro-stickies reduction by deinking flotation in real stock suspensions suspensions Figure 107 shows the micro-stickies loads measured in the deinking flotation inlets and accepts of four different paper mills. The incoming micro-stickies loadings were in the range between 1,0 and 3,0%. Paper mill A achieved a reduction of approx. 30-50% (measured at three different times: I-III). In mill B, the loadings and reduction levels of the post-flotation stage were, as expected, significantly lower than those of the upstream pre-flotation. The deinking flotation stages of mills C and D achieved a microstickies removal of approx. 40-50%. This results in an average micro-stickies reduction by deinking flotation of approx. 50%. 50%.
124
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.4. 4.4.4. 4.3 3 Colloidal potential stic kies
Work programme
Part I – Clean Clean water / No fibr es
In order to understand better the potential use of flotation to remove stickies, this part of the study was focused on the stickies fraction supposed to be the most difficult to remove from the pulp suspension: hydro-dispersible hydro-dispersible µ-stickies. The particle size of these colloidal stickies is normally lower than 5 µm. Two model PVA and acrylic adhesives were chosen to perform flotation experiments. In a first time we worked in the absence of fibres and without drying the adhesives. PVA and acrylic emulsions were directly diluted in water to obtain stable and well-dispersed micro-stickies suspension whose floatability was tested in the presence and in the absence of collectors. These basic glue/water/collector systems were progressively upgraded during the next phases of this study in order to match with real industrial systems and the study of these model systems was coupled with that one of industrial pulp suspensions. Hydro dispersible adhesives (PVA-81085, and acrylic based E115 aqueous slurries at about 48% consistency) were dispersed in deionised water at 0.5 g/L concentration. The stickies suspensions were then floated using a Voith Delta 25 lab cell under the conditions given in tables 25 to 27. In addition to the conventional sodium oleate, two inorganic collectors were also tested, viz. talc and Apiphob. The talc used used was a commercial talc (Luzenac (Luzenac 133P) for paper paper coatings formulation. formulation. Despite an extremely fine grinding (< 5µm), this talc was delivered in powder and it was chosen for the absence of dispersing agents generally used to stabilise talc slurries. Apiphob (from Api Paper Chemicals) is a dolomite mineral modified with a surface treatment which confers to carbonate particles an extremely hydrophobic behaviour. Also Apiphob was used as delivered (i.e. a finely grinded powder). In order to test the sensitivity of inorganic collectors to the presence of surface active contaminants, after that an optimal collector concentration was defined (viz. 0.5 g/L), flotation experiments were performed at increasing non-ionic surfactant concentrations (nonyl-phenol ethoxylated) using the chemical dosages given in table 4. The surface tension (after 15 s relaxation) and the turbidity of the stickies suspensions were measured before and after flotation in order to evaluate the stickies removal efficiencies under the tested conditions. Flotation conditions Time (min) 7
Temperature (°C) Consistency Consiste ncy (%) ~45 1
++
Ca (mg/L)
Air flow (L/min) 6.5 - 7
150
Air ratio (%) 200
Table 25: Experimental conditions used to floate a 0.5 g/L PVA sticky suspension. Flotation condition 1 2 3 4 5 6 7
Acrylic sticky (g/L) 0.5 0.5 0.5 0.5 0.5 0.5 0.5
++
Ca (CaCl2) (g/L) 0.15 0.15 0.15 0.15 0.15 0.15 0.15
NaOH (mg/L) 2.6 2.6 2.6 2.6 2.6 2.6 2.6
pH
Talc (g/L) 0 0.1 0.2 0.5 0 0 0
Apiphob (g/L) 0 0 0 0 0.1 0.2 0.5
Table 26: Chemicals dosage used during PVA-81085 stickies flotation.
125
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
++
Flotation condition
Acrylic sticky (g/L)
Ca (CaCl2) (g/L)
NaOH (mg/L)
1
0.5
0.15
2
0.5
3
pH
Talc (g/L)
Apiphob (g/L)
NaOl (g/L)
2.6
0
0
0
0.15
2.6
0.1
0
0
0.5
0.15
2.6
0.2
0
0
4
0.5
0.15
2.6
0.5
0
0
5
0.5
0.15
2.6
0
0.1
0
6
0.5
0.15
2.6
0
0.2
0
7
0.5
0.15
2.6
0
0.5
0
8
0.5
0.15
2.6
0
0
0.1
9
0.5
0.15
2.6
0
0
0.2
Table 27: Chemicals dosage used during Acrylic-E115 stickies flotation. ++
Flotation condition
Sticky dosage (g/L)
Ca (CaCl2) (g/L)
NaOH (mg/L)
1
0.5
0.15
2
0.5
3
0.5
pH
Collector (g/L)
Surfactant (g/L)
2.6
0.5
0.003
0.15
2.6
0.5
0.006 (cmc)
0.15
2.6
0.5
0.017
Table 28: Non-ionic surfactant concentrations and chemical dosage used when testing
the stickies removal efficiency in the presence of a surface-active contaminant. Collector = talc or Apiphob, Sticky = PVA or acrylic stickies.
Part II – Process w aters / No fibres
The studies in the previous part had demonstrated that micro-stickies of different natures and physicalchemical characteristics might be efficiently removed by flotation when performed in simple environment: clear water without any contaminants and fibrous material brought by the process. Consequently, a question arose: what would be the situation when the micro-stickies are dispersed in process water, in other words, what will be the impact of the numerous contaminants contained in this kind of water on the stickies flotation efficiency? The PVA and acrylic dispersions were used as model µ-stickies. Except for one trial, the adhesive dispersions were directly added at a known rate into the process water studied without prior drying, which is still a difference compared to the industrial conditions where stickies are “re-dispersed”. “re-dispersed”. All process waters were previously filtrated in order to reach suspended solids concentrations as low as possible. Indeed, only the dissolved and colloidal fraction was of interest for the study. A dispersion of about 0.5 g/l was obtained by mixing the adhesive dispersion in the considered process water. The flotation tests were performed with a Voith Delta 25 lab cell during 10 minutes under the same conditions than previously. No chemicals were added.
126
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Main results
Part I – Clean Clean water / No fibr es
The plots of the turbidity versus PVA and acrylic stickies concentration, figure 108, displayed a rather linear shape, thus indicating that in the range of tested concentrations, mutual interactions between polymer spheres (and coalescence) were negligible. The different slopes observed for the acrylic and -1 the PVA, viz 2833 and 1333 NTU/gL respectively, were associated with the different particle sizes of PVA and acrylic µ-stickies. The particle size distribution of polymer colloids determined by dynamic light scattering, figure 109, showed that both acrylic and PVA emulsions had a monomodal size distribution with intensity peaks at 262 and 1201 nm, respectively. The size of acrylic µ-stickies, about 4.6 times smaller than PVA, was therefore responsible for an increase in the solid/water interfaces available for light scattering and in the stickies suspension turbidity.
1000
50
PVA (81085)
PVA (81085)
45
Acrylic (E (E 115) 115) 800
Acrylic (E115)
40
) U T 600 N ( y t i d i b r 400 u T
) % ( y t i s n e t n I
200
35 30 25 20 15 10 5 0
0 0.00
0.10
0.20
0.30
0.40
1
Concentration Concentration (g/L)
10
100
1000
10000
Diamete Diamete r (nm)
Figure 108 108: Turbidity of PVA and acrylic hydro
Figure 109 : Particle size distribution of PVA and
dispersible sticky suspensions plotted as a function of concentration. concentration.
acrylic stickies dispersed in deionised water.
Flotation in th e absence absence of non-ionic sur factant
The low turbidity of the floated suspension shown in figure 110 – 2.6 NTU corresponding to 0.004 g/L 2+ of PVA – indicated that PVA stickies dispersed in deionised water (150 mg/L Ca , pH ~8) were efficiently removed by flotation. The removal efficiency calculated by using turbidity measurement was about 99 %. %. The addition of Apiphob as stickies collector did not induce any further decrease in the turbidity of the floated suspension. This behaviour was interpreted as reflecting the complete elimination by flotation of the collector and probably a decrease in the residual concentration of PVA stickies. The use of talc as collector displayed an increase in the turbidity of the floated suspension when the talc dosage was increased, as shown in figure 110. The measurement of the chemical oxygen demand, figure 111, showed that although an increase in turbidity, the amount of organic compounds (i.e. PVA) in the floated suspension decreased when increasing the talc dosage (talc was supposed not to interfere with the COD measurements). The increase in turbidity was therefore due to the presence of residual un-floatable talc particles.
127
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
90
16 14
Talc Luzenac 133P
85
Apiphob
80
12
) U T 10 N ( y 8 t i d i 6 b r u T 4
) L / g m ( D O C
Talc Luzenac 133P
75 70 65 60 55 50
2
45
0
40
0. 0
0. 1
0. 2
0. 3
0. 4
0.0
0. 5
0.1
0.2
0.3
0.4
0.5
Concentration Concentration (g/L)
Concentration (g/L)
Figure 110 : Turbidity of PVA sticky suspensions
Figure 111 : Chemical oxygen demand (COD) of
after flotation in the presence of increasing talc and Apiphob concentrations.
PVA suspensions after flotation when using talc as collector.
The flotation of acrylic stickies in the presence of increasing concentration of inorganic collectors, figure 112, did not induce any increase in the turbidity of the floated suspension, thus showing, in first approximation, that talc and Apiphob did not improve the floatability of acrylic stickies. Moreover, a large increase in the floated suspension turbidity was observed when using sodium oleate as collector.
200
Talc Luzenac 133P
180
Apiph ob NaOL (Serfax)
160 ) U T N ( y t i d i b r u T
140 120 100 80 60 40 20 0 0.0
0.1
0.2
0.3
0.4
0.5
Concentration (g/L)
Figure 112: Turbidity of acrylic sticky suspensions after flotation
in the presence of talc, Apiphob and sodium oleate.
Flotation in th e presence presence of non-ionic surfactants
To test the floatability of inorganic collectors and model stickies in the presence of surface-active contaminants, flotation experiments were performed using a fixed collector dosage (0.5 g/L) and increasing the concentration of a model non-ionic contaminant, namely nonyl phenol ethoxylate. Figure 114 shows that the presence of the non-ionic surfactant did not affect the floatability of PVAcollector (both talc and Apiphob) systems, the turbidity of the floated suspension remained remained constant for all the tested contamination levels. A slight increase in the turbidity of the acrylic-collector (both talc and Apiphob) systems was observed when increasing the contaminant concentration.
128
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
75
60 ) m / N m ( s 5 1 n o i s n e t e c a f r u S
70
50
) U T 40 N ( y t i d 30 i b r u T 20
65
60
55
10 0
50 0.00
0.10
0.20
0.30
0.40
0.E +00
0.50
1. E-05
2.E -05
3. E -05
4. E -05
8EO concentration (mol/L)
Concentration (g/L)
Tal c Luzenac 133P - In
Tal c Luzenac 133P - Out
Talc FO
Apiphob FO
Apiphob - In
Apiphob - Out
Talc FO
Apiphob FO
NaOl (Serfax) - In
NaOL (Serfax) - Out
Figure 113 113: Surface tension of acrylic sticky
Figure 114 : Turbidity of acrylic (full ( full symbols) and
suspension before (In) and after (Out) flotation at different collector concentration. concentration.
PVA (white symbols) stickies suspensions after flotation in the presence presence of talc and Apiphob at fixed concentration (0.5 g/L) and increasing concentration of non-ionic surfactant.
The comparison of the removal efficiency of PVA and acrylic stickies calculated from turbidity measurements, 99% 99% and 95% 95% respectively, showed that the higher turbidity of the floated acrylic suspension was due in part to the smaller size of acrylic µ-stickies and in part to the slightly lower floatability of these colloids compared to PVA. Probably the destabilisation of µ-stickies and their 2+ flocculation due to the addition of CaCl 2 (to adjust the Ca concentration to 150 mg/L), can explain the generally very high floatability. Another possible reason for the very high flotation efficiency is the complete absence of fibrous material material – that t hat means the stock consistency during flotation was nearly 0.
Part II – Process w aters / No fibres
Clean water
The results of the flotation in demineralised water confirmed that in a clean environment, both types of adhesive might be rather efficiently removed by flotation. Nevertheless, it must be pointed out that in these particular conditions (pure water, without any calcium addition), the removal efficiencies were lower than in the presence of 150 mg/l of calcium. This is especially true for the case of the acrylic based adhesive which only achieved 60 to 65 removal efficiency in the pure water versus 95 % in the presence of calcium ions. This would tend to confirm the role of calcium ions in the micro-stickies flotation mechanism.
Mechanical Mechanical pulp mil l proc ess water
The analyses showed that the process water was rather loaded in contaminants, which could interfere with the flotation process: COD was high which indicated the presence of organic compounds liable to be oxidised. The conductivity level, although not so high, indicated the presence of soluble ions in the medium, and overall, the surface tension was rather low (65 versus 72.8 mN/m for pure water). However, contrary to what was expected, the floatability of the micro-stickies was still high in this contaminated medium. The results obtained with the two types of adhesives were nevertheless not entirely similar.
129
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Recycli Recycli ng packaging mill process water
The results showed that this water was highly contaminated in soluble organic materials as indicated by the level of COD (4.5 g O2/l). However, in spite of this high soluble organic materials contamination, contamination, the level water surface tension remained quite high (68 mN/m). This would indicate that the main part of the soluble compounds does not present any surface activity. Conductivity was also very high which showed a high level of free ions dissolved in the medium. As in the case of the mechanical pulp mill process water, no detrimental detrimental effect (in terms of flotation efficiency) was observed when the micro-stickies were dispersed in this contaminated contaminated medium. On the contrary, in the case of the acrylic adhesives, very higher removal efficiency was reached in this condition, compared to the case of the demineralised water. In addition, it is interesting to compare these results with those obtained in the case of the mechanical pulp mill process water. Indeed, conductivity of packaging mill process water was much higher than that one of the mechanical pulp mill (3140 versus 517 µs/cm). Consequently, one may assume that the calcium ion concentration in the packaging process water was higher. This could explain the higher micro-stickies removal efficiency achieved in the case of the packaging mill process water (95 %) compared to the efficiency achieved with the mechanical pulp mill process water (80 %).
Deinking Deinking m ill proc ess water
First of all this kind of water also contained a rather high load in soluble compounds which could interfere with the flotation: high COD level (lower however than in the case of the packaging mill), high conductivity and high content in surface active substances as shown by the low surface tension achieved in that case (60 mN/m). In this condition, the two adhesives behaved behaved totally differently: on one hand, the PVA based µ-stickies was not negatively affected by the contamination present in this process water: still very high removal efficiency was achieved (more than 95 %). On the other hand, acrylic based µ-stickies were, here, significantly affected by the nature of the contaminants contained in this water. This is a totally different situation compared to the case of the previous process waters studied which had a positive impact on the stickies floatability (compared to the reference trial performed in demineralised water).
Summary
Figure 115 summarises the results obtained in the different flotation conditions determined by the various process water used for the dispersion of the two model adhesives studied.
120 ) 100 % ( y c 80 n e i c i f f 60 e l a v 40 o m e R 20
PVAc Acrylic Acryli c
0 Dem. Wate Waterr
Dem. Wate Water r Pack. Pack. Wate Waterr Deink. ink. Wate Waterr + Ca++
Mech Mech.. Pu Pulp wat.
Figure 115 115: removal efficiencies of PVA and acrylic based stickies when dispersed
in different water qualities
130
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.4.5. Pilot stickies flotation tests
Work programme
The flotation of macro-stickies was studied in two trials in the PTS pilot plant. Both trials were conducted with a fibre mixture of 50% ONP and 50% OMG containing 0,5% adhesives (dispersion adhesive E115 and hot-melt D170) based on o.-d. fibres, and the standard deinking formulation of INGEDE. The consistency in the flotation inlet amounted amounted to 0,85% and 0,85% and 1,35%, respectively. The figure below lists all important trial parameters. parameters. trial number sample
fiber stock mass [g] o.d. mass [g] a.d. adhesiv e mass [g]
dilution water pulping
pulpe lper (1 (10% co consiste isten ncy, cy, 30mi 30min n, 40°C 40°C ave avera rag ge)
chemicals (inlet during pulping)
V T1 VT 2 ONP/OMG ONP/OMG 20000 30000 21000 31500 E115+D170 E115+D170 100 150 normal normal water water normal normal water water tech techn nicu icum tech techn nicu icum 2 pulpers 3 pulpers
flotation: stock f low [l/min] temperature [°C] pH
450 39 8,9
450 39 9,1
NaOH (100%ig) [%] sodium silicate [%] peroxide (100%ig) [%] f atty acid (100%ig) [%] NaOH [g] sodium silicate [g] peroxide [g] f atty acid [g]
0,6 1,8 0,7 0,8 120 360 140 160
0,6 1,8 0,7 0,8 180 540 210 240
0,85
1,35
stock consistency (aim) inlet flotation [%]
Figure 116: Work programme – WP 5.2 The focal points of the measuring programme were macro-stickies determination and a plausibility check of the stock flows. Foam was collected over a flotation period of 4.5 min, which was the time required for one pass of the overall suspension volume (2200 l) through the flotation cell. VT 1 and VT2 0,85 % and 1.35% inlet f lotation accept f lotation f oam 0 - 4,5 min f oam 4,5 - 9,0 min f oam 9,0 - 13,5 min f oam 13,5 - 18,0 min f oam 18,0 - 22,5 min
temp. [°C] x x
pH [-] x x
dilution water [l]
x x x x x
suspended stock mass solids [%] [ g] x x x x x x x x x x x x x x
ash [%] x x
total loss [%]
fiber loss [%]
x
x
macrostickys DMF-Extract [mm²/kg] [%] x x x x x x x x x x x x x x
Figure 117: Measuring programme – WP 5.2 As for the previous macro-stickies flotation tests, the macro-stickies area was evaluated in detail, i.e. for different particle size classes.
131
L SCREEN C CLEAN EAN
Main results
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Separa Separation tion of macro-stickies
In the flotation tests evaluated here, clean water was used for dilution. As a result the stickies reduction achieved was higher than in industrial flotation systems using process waters loaded with various detrimental detrimental substances (cf. report D10) for dilution. The basic mechanisms of stickies flotation are reflected very well by the tests, however. Figure 118 shows the overall macro-stickies reduction achieved by the pilot and laboratory flotation tests. The reduction by balance, i.e. based on the actual stock balance, takes into account also the stock losses during flotation as well. The macro-stickies reduction was in the range between 50 and 70%. Raising the inlet consistency of the flotation cell from 0,85% to 1,35% led to a relative decrease in reduction efficiency by about 25%. The same tendency, i.e. a lower macro-stickies reduction at higher consistencies, had been observed in the laboratory tests (cf. progress report D10). The pilot test results were similar to the lab test results at 1.0 % inlet consistency. This underlines the generally good comparability of the pilot and lab cell results. pilot flotation (700l VOITH eco-cell) - in comparision to lab flotation 100 absolute reduction reduction by balance
90 ] 80 % [ n 70 o i t c u d 60 e r a e 50 r a s e i k 40 c i t s o r 30 c a m
20 10 0 pilot flotation ( stock consistency consistency inlet 0,85%)
pilot flotation (stock consistency inlet 1,35%)
lab flotation (stock consistency inlet 1,00%)
trial
Figure 118: Pilot flotation, separation of macrostickies Both the laboratory and pilot flotation processes were most efficient for smaller macro-stickies, as shown in figure 119. Approx. 80% of the macro-stickies ≤ 500µm and approx. 60% ≤ 1000µm could be floated. The share of macro-stickies larger than 1500µm could only be reduced by approx. 35%.
132
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
pilot flotation (700l VOITH eco-cell) - in comparision to lab flotation 100 pilot flotation ( stock consistency i nlet 0,85%)
90
pilot flotation ( stock consistency i nlet 1,35%) lab flotation ( stock consistency inlet 1,00%)
80 70 ] % [ y c n e i c i f f e n o i t a t o l f
60 50 40 30 20 10 0 -10 -20 100...500
500...1000
1000...1500
1500...2000
total
macrostickies particle size [µm]
Figure 119: Pilot flotation, macro-stickies separation for different size classes The relative shares of the individual passes in the overall macro-stickies reduction are shown in figure 120. At the lower consistency, most macro-stickies were removed in the first pass. The process run at higher consistency was more balanced, with macro-stickies being removed by all five passes. However, the first pass was most efficient here as well. pilot flotation (700l VOITH eco-cell) - partly macrostickies reduction per flotation step (each 4.5 min) 100 95
pilot flotation (stock consistency inlet 0,85%)
90
pilot flotation (stock consistency inlet 1,35%)
85 80 ] 75 % [ n 70 o i t c 65 u d 60 e r 55 a e 50 r a s 45 e i k 40 c i t 35 s o r 30 c a m 25 20 15 10 5 0
4,5 mi n
9,0 min
13,5 min
18,0 min
22,5 min
flotation time [min]
Figure 120: Pilot flotation, macro-stickies separation at different flotation times
Separa Separation tion of mi cro-stickies
In addition to the macro-stickies analyses, the total extract was measured by extracting the filter cakes from the consistency measurements with dimethyl formamide. The reduction of micro-stickies sized between 5 and 100 µm may be estimated from the extract. Because deinking flotation was most efficient at removing smaller macro-stickies, it was expected to remove micro-stickies as well. The micro-stickies in the ONP/OMG used originated mainly from binders or binder agglomerates of printing inks and paper coatings.
133
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
pilot flotation (700l VOITH eco-cell) - microstickies load vs. flotation time 20 pilot flotation (stock consistency inlet 0,85%) 18
pilot flotation (stock consistency inlet 1,35%)
16 14 ] % [ 12 t c a r t x 10 e F M 8 D
6 4 2 0 Inlet
Accept
foam 0 - 4,5 min
foam 4,5 - 9,0 min
foam 9,0 - 13,5 foam 13,5 - 18,0 foam 18,0 - 22,5 min min min
Figure 121: Pilot flotation, total amount of extract As can be seen from figure 121, 121, the extractable content of the samples samples taken from the flotation accept accept is significantly lower than that of the inlet samples. The good removal efficiency of deinking flotation for extractables is also reflected by the significantly increased extract contents of the flotation foams. The total extracts from the filter cakes of the consistency measurements contained three different types of stickies: extract from macro-stickies + extract from micro-stickies + extract from fibrous material total extract The macro-stickies share can be estimated from the known adhesive inputs (0,5% E115/D170) and the reduction achieved by flotation. Another 0,05% macro-stickies were introduced by recovered paper (approx. 5000 mm²/kg). Experience shows that approx. 0,7% of the extract introduced by recovered paper originates originates from fibrous f ibrous material. Based on these data, the micro-stickies contents of the flotation inlet and accept were calculated (table 29). The micro-stickies reduction of the pilot flotation was estimated at approx. 80%. 80%. pilot flotation (stock consistency inlet 0,85%) total ex tract m acrostickies (f rom E115/D170) macrostickies (from ONP/OMG) from fibers (PTS experience) microstickies microstickies (from ONP/OMG) reduction macrostickies reduction microstickies pilot flotation (stock consistency inlet 1,35%) total extract m acrostickies (f rom E115/D170) macrostickies (from ONP/OMG) from fibers (PTS experience) microstickies microstickies (from ONP/OMG) reduction macrostickies reduction microstickies
m easured m easured PTS experience experience PTS experience experience
calulated
m easured m easured PTS experience experience PTS experience experience
calulated
inlet 2,35 0,50 0,05 0,70 1,15
accept 1,15 0,15 0,02 0,70 0,30 70 74
inlet 2,85 0,50 0,05 0,70 1,65
accept 1,20 0,25 0,02 0,70 0,25 50 85
Table 29: Pilot flotation, calculation of micro-stickies
134
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.4.6. Conclusions and perspectives
Lab flotation
Model macro-stickies + clean water
-
Using model stickies with an inlet stock consistency of 1.0% and standard deinking chemicals, chemicals, the total separation efficiency achieved amounted to approx. 45% based 45% based on the absolute values and approx. 60% based 60% based on the stock balance.
-
In the case of smaller macro-stickies (<1000 µm, INGEDE Method N°4), separation efficiency values of 70 - 80% 80% were achieved on average using a 25l VOITH laboratory flotation cell and a flotation time of 12 minutes. Hence, flotation always tends to separate out smaller macro-stickies. macro-stickies.
-
No macro-stickies reduction was possible without the use of flotation chemicals.
-
The stock consistency in the flotation cell inlet had a significant impact on the results of flotation. Increasing the stock consistency from 1.0 to 1.25 1.25 caused the separation efficiency of macrostickies to fall by 20%. 20%. A further increase in stock consistency to 1.5% 1.5% reduced the separation efficiency by a total of 40%.
-
For the stock suspension under test, adding a special hydrophobic mineral to the standard deinking chemicals formulation brought no increase in flotation efficiency.
-
The flotation result depends on the amount of fatty acid used. A 50% reduction in the standard dosing amount reduced the removal efficiency by approx. 25%. A 50% increase in fatty acid dosage resulted in a slightly improved macro-stickies flotation whilst increasing the total pulp loss.
-
When using real mill water instead of clean water for sample dilution the flotation efficiency was lowered significantly.
Stock suspension f rom a paper paper mill
-
An absolute value for macrostickies separation of approx. 25% was 25% was measured during the flotation of a mill stock suspension from a tissue mill in the laboratory flotation cell. Precisely this macrostickies separation efficiency was also achieved in the mill flotation plant.
-
A study of macro-stickies macro-stickies separation in the flotation plant of a newsprint paper mills that was carried out within the scope of system analysis produced a macro-stickies reduction by approx. 40-50% in 40-50% in the flotation stage.
-
Smaller macro-stickies are preferably floated even in mill suspensions.
Micro-stickies / industr ial stock suspension
-
The analysis of the deinking flotation stages of four different paper mills showed an average reduction in micro-stickies load by approx. 50%. 50%.
Model Model col loidal potential stic kies / clean and and process w ater ater / no f ibres
-
Model micro-stickies were efficiently removed by flotation even without the addition of collectors, PVA model stickies displayed a slightly better floatability than acrylic stickies. The total removal efficiency was more than 90%.
-
The PVA based stickies floatability was not significantly affected by the type of process water. Very high flotation efficiency were always found (90 to 95 % ) whatever the process water used for the preparation of the dispersion. On the contrary, the floatability of acrylic based stickies was influenced by the nature of the process water.
135
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Pilot Flotation
The results of the deinking flotation trials using a pilot Voith-Eco cell can be summarised as follows:
Model macro-stickies / clean water
-
The macro-stickies reduction was in the range between 50 and 50 and 70%. 70%. Raising the consistency in the flotation inlet from 0,85% to 1,35% caused an approx. 25 % relative decrease in removal efficiency.
-
The preferred flotation of smaller macro-stickies sized < 1000 µm (according to INGEDE Method N° 4) was observed in both trials.
-
The pilot and lab flotation cells achieved similar results in terms of macro-stickies reduction.
-
Micro-stickies / clean water A micro-stickies reduction reduction of of approx. 80% was was calculated for for the pilot trials. trials.
Overall conclusions
-
Deinking flotation is generally suitable for separating macro-stickies. Total macro-stickies separation efficiency values of approx. 40-50% 40-50% can be achieved in well functioning flotation plants.
-
Deinking flotation mainly separates smaller macro-stickies and micro-stickies with a high efficiency that can be as much as 70%. 70%.
The macro- and micro-stickies reduction clearly exceeds the fibre losses and even the filler reduction. In industrial practice, the fibre losses will be even lower due to the secondary flotation stage, i.e. approx. 2-3 % in a 5-step plant. The use of mill process water instead of clean water in the lab/pilot trials can result in a considerably lower stickies reduction, however. however. A 50% lower removal efficiency than in the tests is quite probable. On the whole, the stickies reduction by deinking flotation is a very useful addition to pressurised screening. Deinking flotation removes preferably preferably smaller macro-stickies, which can only be eliminated to a limited extent by the screens. The possibility of micro-stickies removal by deinking flotation is a very valuable result. Apart from thickeners, no other method has been available so far for removing these particles from stock suspensions. Moreover, thickening stages discharge the micro-stickies into the process water so that they are reintroduced into the suspension by dilution.
136
L SCREEN C CLEAN EAN
4.5.
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Pressure filtration
4.5.1. Background and objectives Both the separation of contaminants from stock suspensions and the cleaning of water circulations have gained in importance during the past few years. Mill circuits narrowing or even closure causes a variety of different substances to accumulate in the water circulation. Paper fibres, fibre fragments, fines and fillers are not the problem in this context. The negative effects of closing water loops are due to the gradual increase in concentration of non-paper substances such as sticky particles, inks and dyes, micro-organisms, heavy metals, salts and fatty acids. The addition of NaOH to release printing ink during the recycling of used newspapers and magazines normally forces the pH value that prevails during the re-pulping of the recovered paper far above neutral. This high pH value then causes even greater fragmentation and re-dispersion of adhesives. The result is greater accumulation of nonabsorbable detrimental substances in the mill loops. In the past few years, finely dispersed sticky particles, better known as micro-stickies, have become the focus of interest among papermakers. As stock preparations and paper machine loops become more and more concentrated, most of these sticky particles pass into the filtrates. These filtrates or white waters are reused as dilution water in stock preparation and in the paper machine. Against this background, there is an elementary need for internal cleaning of the mill loops in order to irreversibly eliminate micro stickies from the stock suspension and avoid gradual concentrations of contaminants. A new, potentially suitable process for cleaning water loops in paper mills is the pressure filtration process. This is a modified screening operation that takes place in a pressure screen using specialpurpose screen cylinders. The separating criteria is the particle size of the substances to be separated that are contained in the water loop in relation to the hole or slot width of the screen cylinder [111].
Work Package WP 6 of this project is aimed at the following important objectives: objectives: SWP 6.1 – Preparation of process waters -
Studying the composition of mill process waters with respect to suspended substances, stickies, pulp fractions and dissolved and colloidal substances
-
Conducting pilot trials to create a practice-oriented model water
-
Selecting the process water with the best suitability for the trials
SWP 6.2 – Pilot pressure filtration -
Using an industrial pressure filter
-
Varying the operating parameters of the pressure filter
-
Using flocculants to selectively flocculate finely dispersed and colloidal substances substances in the process water
-
Evaluating the possibility of separating micro stickies from the process water
-
Evaluating the separation action with respect to the fibre fractions and fines/fillers
Pointing out the possibilities and limitations of pressure filtration in the treatment of process waters
137
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.5.2. Basics of pressure pressure filtration process The pressure filter has evolved as a further development of the known process of pressure screening. The pressure screen serves to screen the pulp suspension, whereas the pressure filter is especially designed to separate particles from process waters. The mode of operation of pressure screens and pressure filters is basically the same. The pulp suspension suspension or process water flows into the interior of the screen cylinder. It passes through the screen cylinder in a centrifugal direction towards the outer housing of the screen/filter. The reject, i.e. the particles retained by the screen cylinder, are removed at the bottom of the unit. The principal different between pressure screens and pressure filters is the aim of the process. Pressure screens are intended to separate only coarsely dispersed contaminants and shives. This in turn means that almost all paper fibres that can be used for papermaking are passed into the accepts. In the case of the pressure filter, on the other hand, all or as many coarsely and finely dispersed substances as possible, i.e. the paper fibres, are to be separated from the process water. The design of the cylinders of pressure screens and pressure filters differs basically in the width of the holes or slots in the cylinders and in the design of t he surface profile of the cylinders, as shown in table 30 and figure 122. Another distinguishing feature of pressure screens and pressure filters is the way in which the rotor is constructed. Pressure screen
Pressure filter
2.4 mm hole
200 µm hole
1.3 mm hole
100 µm hole
0.25 mm slot
100 µm slot
0.20 mm slot
80 µm slot
0.15 mm slot
60 µm slot
0.10 mm slot
40 µm slot
Table 30: Holes and slots sizes in pressure screens compared to pressure filters
pressure scre pressure screening ening - aggre aggressiv ssive e - bi big g pr prof ofil ile e he heig ight ht
pressure scre pressure screening ening - ge gent ntle le - sma small ll pro profil file e hei height ght
pressure filt pressure filtrati ration on - no profile
Figure 122: Design of the screen cylinder inlet angle
138
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.5.3. Pressure filtration – pilot equipment The pressure filter employed to filter the process water at the PTS pilot facility is shown in figure 123. The water flow is re-circulated until the operation has stabilised itself and is then divided into accepts and rejects as soon as the trials actually begin.
PDCI A PDCIA
M
-
accept reject Figure 123: Pressure filter, device and process design Figure 124 illustrates the bump rotor and the slotted screen cylinder of the pressure filter. The figure clearly illustrates the large number and symmetrical shape of the bumps on the rotor as well as the completely level surface of the screen cylinder.
scre sc reen en ba bask sket et wi with th sl slot ots s
rotor Figure 124 : Pressure filter, rotor r otor and cylinder design
139
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.5.4. Preparation of process waters
Work programme
The following tests were conducted to basically characterise paper mill process waters for producing mass printing papers and for pre-selecting the process water suitable for testing: a) Procurement of 5 different process waters from paper mills and measurement of the following parameters:
• • • • • •
stock consistency composition of fractions (McNett fractionation) fractionation) macro sticky content (INGEDE Method No. 4) micro sticky content (extracting the filter cake from the stock consistency measurement using dimethyl formamide) COD content of anionic trash (polyelectrolyte titration)
b) Trials to prepare prepare a practice-oriented practice-oriented process process water water at the PTS pilot facility facility
• • • • •
repulping deinked pulp + 1% adhesive mixture D170/E115 in the pulper dilution to approx. 1% stock consistency case 1: thickening the stock suspension using a drum filter case 2: thickening the stock suspension using a screw press filtrate measurements, measurements, refer to section a) above
Based on the results of measurement, the most suitable process water was chosen for the rest of the filtration trials.
Main results
It was necessary to prepare a practice-oriented process water in the pilot facility before the pressure filtration trials could be conducted. The choice of the water that was ultimately used for the trials in SWP 6.2 was intended to model the composition of the industrial process waters used in practice. The stock consistency (suspended solids) in the process waters from the paper mills for producing printing paper ranged between 0.5 and 2.9 g/l. A range especially from 1 to 2 g/l was found more frequently. Studies of more than 30 paper mills in various PTS projects confirm that process waters are commonly loaded with from 1 to 2 (3) g/l of suspended solids. Compared to the mill water, the model water from the drum filter showed a very low content of suspended solids amounting to only 0.2 g/l. Most of the suspended solids in the model stock suspension were retained on the screen of the drum filter. The model water produced by the screw press had a stock suspension of 3.1 g/l and was thus slightly above the usual range measured in paper mills. Measurement of the macro stickies demonstrated an unrealistically high load in the model waters compared to the mill waters. The high adhesive load in the model suspension was chosen with a view to the fact that part of the adhesive disintegrates into micro stickies during re-pulping.
140
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
pre tests - pressure filtration 8 7,2 7
] 6 % [ s e i k 5 c i t s o r c 4 i m t c a r t x 3 e F M D 2
2,4
2,6
2,4
2,1
1 0,1
0,1
0,1
0 Deinking Mill 1 - Loop I
Deinking Mill 1 - Loop II
Deinking Mill 2 - Loop I
Deinking Mill 2 - Loop II
Deinking Mill 2 - Loop III
model stock suspension
model water - model water filtrate drum filtrate screw filter press
Figure 125 : Preliminary tests, micro stickies The level of the micro-stickies load in the mill water normally ranged between approx. 2 to 3%. This load was not nearly achieved in the model waters. It proved to be impossible under the chosen repulping conditions to satisfactorily produce finely dispersed micro stickies using both adhesives E115 and D170 on a pilot scale. The adhesives did not disintegrate to the extent desired and existed almost entirely in the form of c oarsely dispersed dispersed macro stickies. pre tests - pressure filtration 14,0
12,0
11,6
] c a m d 10,0 a d y l o P n 1 8,0 0 0 , 0 l m 6,0 [ h s a r t c i n 4,0 o i n a
10,4
4,7
2,5
2,0 0,3
0,4
Deinking Mill 2 - Loop III
model stock suspension
0,4
0,3
0,0 Deinking Mill 1 - Loop I
Deinking Mill 1 - Loop II
Deinking Mill 2 - Loop I
Deinking Mill 2 - Loop II
model water - model water filtrate drum filtrate screw filter press
Figure 126 : Preliminary tests, anionic trash The charge character of the dissolved matter in the mill and model waters differed considerably. Whereas the mill waters exhibited a very high anionic substance load, the load in the model waters was rather low. The contents of anionic substances in the water are of immense importance for the flocculation of the dissolved matter with well-chosen cationic chemicals.
141
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The composition of the mill waters was examined using a standardised Bauer McNett classifier. The fibre content (long and short fibres) in the analysed mill waters amounted to a mere 2-11%. The largest portion of the suspended dissolved matter consisted of fibre fines and fillers. The share of fibres in the model water from the filtrate screw press amounted to 35% and was thus substantially higher. So to conclude: model waters from the pilot plant cannot be used because: filtrate drum filter
filtrate screw press
- quantities of suspended substances are too low - not enough micro stickies - low level of anionic trash - too many macro stickies - not enough micro stickies - high fibre content - low level of anionic trash
use of real process water from a paper mill for the trials: concentration concentration on
- micro stickies reduction - fibre fractionation - macro stickies reduction - particle flocculation, if possible
In order to be able to conduct the pressure filtration trials, approx. 2000 litres of process water were obtained from a paper mill that produces printing papers. This water was referred to as "deinking mill 1, loop II" in the preliminary trials.
4.5.5. Pilot pressure filtration tests
Work programme
All trials were conducted using the process water from a paper mill. The following fundamental fundamental pressure filter operating parameters were varied:
• • • • • •
screen cylinder reject rate rotor speed volumetric inlet flow rate inlet stock consistency addition of flocculant to flocculate finely dispersed and colloidal substances
The inlet/accept pressure difference was not varied. The value resulted from the mode of operation and amounted to 0.2 ± 0.05 bar. A preliminary preliminary trial was carried out to flocculate the process water and very good flocculation was achieved. After a sedimentation time of approx. 5 min during which the flocculated dissolved matter was allowed to settle, the residual turbidity of the water amounted to approx. 1% of the original level. All filtration trials were conducted conducted at the PTS pilot facility using a conventional conventional industrial pressure pressure filter. The trials were evaluated on the basis of the following analytical measurements: pressure filtration inlet reject accept
stock co consistency [%] x x x
turbidity [NTU] x x x
evaporation residue
[%] x x x
ash content [%] x x x
Table 31: Analytical measurements
142
microstickies [%] x x x
stock fractions fiberlab x x x
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Main results
Changing screen basket
During the first trial series, the screen cylinder of the pressure filter was varied. A 100 µm perforated cylinder was used as were slotted cylinders with slot widths of 60, 40 and 20 µm. The volumetric reject rate amounted to 10%. Rotor speed was 15/s and the associated circumferential speed of the rotor amounted to 14 m/s. pressure filtration - variation screen basket reject rate (volume): 10%, rotor speed: 15/s 4,0 inlet
accept
reject
3,5
3,0 ] l / g [ 2,5 y c n e t s 2,0 i s n o c k c 1,5 o t s
1,0
0,5
0,0 100 µm O
60 µm //
40 µm //
20 µm //
screen basket
Figure 127 : Pressure filtration, variation screen cylinder – stock consistency Figure 127 illustrates the fundamental mode of operation of the pressure filter. The stock consistency in the inlet, accepts and rejects does not differ significantly as is the case in other filters (disk filters, drum filters, screw presses). The difference in stock consistency between the inlet and accepts is very small. The rejects do show a progressive concentration of suspended dissolved matter, since larger particles tend to accumulate here. The composition of the suspended dissolved matter was determined in order to be able to evaluate the separation efficiency of the pressure filter in detail. Measurement was carried out using an automated FIBERLAB fibre dimensions analyzer. The following important fibre fractions were identified in the results of measurement: Title
Particle Size
fibres (long fibres + short fibres)
> 500 µm
fines
200...500 µm
fillers + small fines
< 200 µm
The results of the fractionation of the suspended dissolved matter have shown that they mainly contained fillers and very small fines particles. The total content of this stock fraction amounted to approx. 83%. 83%. The fibre fines in the particle size range of 200 – 500 µm amounted to approx. 15%. 15%. The content of paper fibres in the process water merely amounted to approx. 2%. 2%.
143
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
pressure filtration stock fractions at inlet 100 90 80 70 ] % [ 60 t h g i e 50 w e v i t a 40 l e r
30 20 10 0 fibers (> 500µm)
fines (200..500µm)
small fines + fillers (< 200µm)
stock fractions (fiberlab)
Figure 128 : Pressure filtration, variation in screen cylinder – stock fractions inlet
Figure 129 clearly shows that it is mainly the paper fibres (> 500 µm) that are separated in the pressure filter. The separation efficiency in the accepts from the 100 µm perforated screen exceeded 85%. A separation efficiency of approx. 50% using the perforated screen was also achieved in the case of the fibre fines (200…500 µm). This means that the pressure filter basically operates like a fibre fractionator. fractionator. The perforated screen was ineffective with the specifically largest fraction < 200 µm.
pressure filtration - variation screen basket reject rate (volume): 10%, rotor speed: 15/s 140 130 ] 120 % [ ) 110 % 0 0 100 1 = t 90 e l n i ( t 80 p e c 70 c a n 60 i s n 50 o i t c a 40 r f k c 30 o t s 20
fill filler ers s+ sm small all fine fines s (<20 (<200 0µm) µm)
fine fines s (<50 (<500 0µm) µm)
fibers (> 500µm)
trend fillers + small fines
10 0 100 µm O
60 µm //
40 µm //
20 µm //
screen basket
Figure 129 : Pressure filtration, variation in screen cylinder – stock fractions accept
144
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The efficiency of slotted wire wedge screens was much lower with respect to paper fibres, especially with fibre fines, than that of the perforated screen. screen. Owing to their cylindrical shape, the paper fibres are able to pass through the slots much more easily than through the holes. In the case of perforated screens, the hole diameter as the separating element is effective in two dimensions (x and y direction), whereas the slot width as the separating element is effective in only one dimension (see Figure 130). Compared to the perforated screen, slotted wire wedge screens were also able to separate out a small portion of the fraction <200 µm. As was expected, this separation action intensified as the slot width narrowed.
l = 500µm, d = 30 µm
l = 500µm, d = 30 µm
100µm hole
60µm slot
d= 80 µm
d= 80 µm
Figure 130: Pressure filtration, principle of filtration of fibres and spheres In order to determine the micro-stickies content, the filter cake from the stock consistency determination was extracted using dimethyl formamide. No substantial micro-stickies reduction was found in the accepts or rejects from the pressure filter during the trials. This is also quite plausible because, by virtue of their particle size <200 µm, most of the micro-stickies will be found in the "small fines and fillers" fraction. As already mentioned mentioned above, the pressure filter functions almost like a valve with this stock fraction, splitting the micro-stickies up into accepts and rejects according to the set reject rate. pressure filtration - variation screen basket reject rate (volume): 10% 10 inlet
accept
reject
9 8 7 ] % 6 [ t c a r t x 5 e F M 4 D
3 2 1 0 100 µm O
60 µm //
40 µm //
20 µm //
screen basket
Figure 131 : Pressure filtration, variation in screen cylinder – micro stickies
145
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Effect of reject rate
A distinct trend in the rejects became obvious in the curves of the stock consistencies consistencies when a 100µm perforated screen was used. A decline in the reject rate, i.e. comparatively aggressive filtration, resulted in greater thickening of the rejects. This trend could not be observed when a 40µm slotted wire wedge screen was used. pressure filtration - variation reject rate screen basket: 100µm O, rotor speed: 15/s 150 fi llll er ers+small fi fi ne nes (< (<200µm) fibers (> 500µm) trend fines
140 ] 130 % [ 120 ) % 0 110 0 1 = 100 t e l n 90 i ( t p 80 e c c 70 a n i s 60 n o i 50 t c a r 40 f k c o 30 t s 20
f in ines (< (<500 µm µm) trend fibers
10 0 20
15
10
5
reject rate (volume) [%]
Figure 132 : Pressure filtration, variation in reject rate – stock fractions f ractions 100 µm O
Examination of the individual stock fractions produced a distinct trend when the 100µm perforated screen was used. Both pulp separation and fines separation deteriorated as the reject rate was decreased. More aggressive filtration at a lower reject rate forced the fibres to pass through into the accepts. Fibre separation with the perforated screen amounted to 90% and more in all cases and was considered to be very good. In addition, about 50% of the fibre fines in the particle size range 200…500µm were able to be separated using the perforated screen. No filtration action was found in the case of fines and fillers <200 µm. Varying the reject rate using the 40µm slotted wire wedge screen did not produce any clear trend. No significant filtration action could be found with either of the screens in the case of the micro sticky load. The distribution of the micro sticky load like in a valve based on the set reject amount is illustrated in Figure 133. The mass flow of the micro stickies in the accepts was almost consistent in all trials with the volumetric reject rate that was set in each case and thus corresponded to the range of the diagonal line depicted in the graph.
146
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
pressure filtration - variation reject rate screen basket: 100 µm O + 40µm //, rotor speed: 15/s 30 100 µm O
] % [ 25 ) t e l n i o t n 20 o i t a l e r n i ( 15 t c e j e r n i s 10 e i k c i t s o r c 5 i m
40 µm //
0 0
5
10
15
20
25
30
reject rate (volume) [%]
Figure 133 : Pressure filtration, variation in reject rate – micro-stickies flow rate
Ad di ti on of fl oc cu lant lan t
Precipitants or flocculants can be used to selectively agglomerate finely dispersed and even colloidal substances from process waters. In a preliminary test a suitable flocculant had been added under high turbulence (agitator). After a sedimentation time of 10 minutes, the flocculant had reduced the turbidity in the process water by more than 99%. During filtration in the pressure filter there was almost no reduction in turbidity when a flocculant was used. The strong turbulence in the chest, pump, pipes and pressure filter that was present during the filtration trial prevented the formation of flocs. pressure filtration - addition of flocculant screen basket: 100µm O, rotor speed: 15/s 4000 without flflocculant
with flflocculant
3500
3000
] 2500 U T N [ y t i 2000 d i b r u t 1500
1000
500
0 20
15
10 reject rate (volume) [%]
Figure 134: Pressure filtration, 100 µm O – addition of flocculant
147
5
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
4.5.6. Conclusions and perspectives Figure 135 compiles the most important results of the pressure filtration trials. The pressure filter separation efficiency values that were achieved were broken down according to perforated screen and slotted wire wedge screen and compared with possible separation efficiency valves achieved with dissolved air flotation (DAF). DAF or micro-flotation is the method most frequently used for treating process water.
separation of fibers (> 500µm) separation of fines (200...500µm) separation of small fines + fillers (200...500µm) reduction of big macrostickies (>500 µm) reduction of small macrostickies (<500 µm) reduction of microstickies
pressure filtration 100µm O 20...60µm // + +/+/+ + + +/-
DAF + + + + + +
Figure 135 135: Pressure filtration - summary (+/good, +/-/medium, +/-/medium, -/bad efficiency)
The key trials results are as follows: Mode of operation
•
The pressure filter invariably operates more like a fractionator and not so much like a conventional filter that forms a filter cake. Only a slight concentration effect was detectable in the rejects.
•
The mode of operation with respect to reject rate, hydraulic load and rotor speed had only a minor effect on the results of filtration. The design of the screen cylinder (perforated or slotted wire wedge) in particular was significant.
•
It was not possible to flocculate the substances in the process water owing to the high turbulence in the water while the pressure filter was in operation.
Separation efficiency
•
The perforated screen produced much better results with respect to the fibre and fines separation (> 200 µm) than did the slotted screen.
•
The pressure filter showed almost no separation efficiency in the case of fines/fillers (<200 µm), the largest specific particle fraction in the process waters from paper mills. The particles were split up into accepts and rejects like in a valve approximately approximately according to the reject flow rate.
•
Micro-stickies by definition have a particle size <100 µm. Consequently, it was not possible to separate micro-stickies in a pressure filter. The micro-stickies were also split up into accepts and rejects according to the volumetric reject rate.
• •
Almost all macro-stickies macro-stickies present present in the process process water were were retained in the rejects. The pressure filter can be employed as a fibre recovery system. The high hydraulic load of the pressure filter is advantageous. At the same time, however, macro-stickies accumulate in the rejects.
The pressure filter is employed for fibre recovery in the process water, perforated screens being recommended for this purpose. As far as fibre recovery is concerned, the pressure filter is in direct competition with conventional disk filters and drum filters. One advantage of the pressure filter is certainly its simple operation without sensitive filter media and without the frequently necessary dosage of additional fibrous materials such as that required in disk filters.
148
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
5. Conclusions 5.1.
Removal of stickies in deinking lines
The objective of the work, which was planned, within SWP7.1, in the last workpackage of this project, was indeed to draw conclusions about how to optimise the stickies removal stragtegy in deinking lines, on the basis of the findings developped in the previous workpackages where the optimisation of the stickies treatments and separation techniques had been investigated at the different process steps. The conclusions developped below are based on the synthesis of these findings and on additional simulation results about screening systems including cleaning or flotation. Details about the simulation conditions can be found in the report D19 “Optimisation of the removal of stickies in deinking lines”. The conclusions about possibilities to improve the removal of stickies in deinking lines are based on the state of the art deinking technology described described in section 1.1.
5.1.1. Optimisation of pulping to improve further further stickies stickies removal removal The development of new optimised pulping conditions in order to produce large and as far as possible relatively thick adhesive particles, is a prerequisite for an efficient removal of the stickies in the subsequent process steps, especially fine slot screening. The current trends have been to reduce the fragmentation of the contaminants, by using gentle pulping technology, such as the drum pulper which was considered as a reference in this respect. The approach in this project was to study existing pulping technology and to develop new technology in order to minimize the fragmentation of adhesives and to promote their agglomeration in such a way to remove them almost completely by fine slot screening in the form of macro-stickies, while avoiding as much as possible the production of micro-stickies and dissolved and colloidal stickies components.
Optimisation of pulping conditions to minimise stickies fragmentation
Pilot tests were performed at CTP to compare the drum pulper (a slice of an industrial-sized drum) to the batch pulper (a pilot-sized Helico pulper). The fragmentation of adhesives during pulping showed to depend on the type of pulper: The drum pulper induced lower adhesive fragmentation compared to the helical pulper and led consequently to higher stickies removal by subsequent screening with fine slots, as described in the progress report D6. Moreover, it was shown that the lab Helico pulper cannot lead to adhesive fragmentation similar to that one achieved in the pilot pulper: Stickies fragmentation was much lower in the lab helical pulper, for both reference adhesives, adhesives, which confirmed the difficulty to assess adhesive behaviour (overall their ability to be broken up during pulping) by simple lab tests. The pilot pulping trials showed clear benefits of drum pulpers, at least from the stickies point of view. However, as the tests with the pilot batch pulper were only performed at conventional non-optimised pulping time (and also because the results were not consistent with those obtained on lab scale), it seems difficult to conclude about the fragmentation of stickies in batch pulpers on mill scale under optimised pulping conditions. Indeed, the drum pulper has generally been regarded as a reference in terms of gentle pulping action with consequently reduced fragmentation fragmentation of stickies and contaminants, which are rejected in large pieces at the outlet of the drum. By contrast recent macro-stickies analyses performed in several deinking mills, showed a strong reduction of the average stickies surface area and an increase in the stickies number with drum pulpers compared to batch pulpers, which led to the conclusion that drum pulpers produced a stronger stickies fragmentation [38]. Consequently, Consequently, it seems difficult to draw clear conclusions about which pulping technology should lead to the lowest stickies fragmentation on the basis of the pilot trials performed in this project. Recent comparative tests of the drum and batch pulpers available at the pilot facilities of a major equipment supplier [38] did not allow to draw clear conclusion about stickies fragmentation as well. The drum pulper may still be considered as a relatively gentle pulping technology since no strong mechanical forces are exerted on the adhesives. By contrast, the helical rotor of the batch pulper may generate locally stronger mechanical forces, especially if the rotor velocity and/or the pulping time exceed the minimum conditions required for the defibering of recovered papers and for the optimised detachment,
149
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
fragmentation and re-deposition of inks. Indeed, the batch pulper offers a larger set of parameters, including consistency, pulping time and rotor velocity and design, which can be optimised compared to the drum pulper. The batch pulper is also more adapted to development of the new pulping technology at ICP, as the flow conditions (pulp rheology and velocity field) can be changed quite easily.
Development a new pulping process for the agglomeration of stickies
The initial idea at ICP was to develop new pulping conditions to agglomerate the adhesives in order to produce oblong shaped and less deformable particles to be removed easily by slot screening as well as more hydrophobic particles to be removed by flotation. The idea of co-agglomeration of ink and adhesive particles was developed later by ICP through the optimisation of the pulping conditions and chemistry, which led to smaller and darker stickies particles covered with inks. The flotation process should then easily remove such macro/micro-stickies with hydrophobic character. The new ICP pulping process has been developed first on lab scale to investigate the effects of the basic pulping parameters and later on pilot scale. The changes in the stickies particle size distribution during pulping are analysed on the basis of recent theories of particle comminuting and agglomeration processes. They are characterised by a dislocated lognormal particle size distribution, which were shown to fit best the experimental data. Practically, an increase of the scale parameter (µ) of the distribution indicates larger particles, i.e. a positive effect, while a decrease of the shape parameter ( σ) indicates a reduction of the particle size range (positive). Proper combinations combinations of surface active agents which are insoluble in water and have a melting point lower than the pulping temperature are used to promote comminuting of the adhesives in the pulper and subsequently their agglomeration / coagglomeration with inks. Furthermore, in the subsequent process steps, where the temperature should be lower than the melting point of these hydrophobic “agglomerants”, their phase transition (from liquid to solid) is additionally fixing the effects of such agglomeration/co-ag agglomeration/co-agglomeration glomeration processes. processes. The new process first requires producing small adhesive particles in order to be able to agglomerate them more strongly with the bridging chemistry (particle densification). To produce such effects in the pulper both proper flow with laminar motion and admixture of the agglomerants are needed. The latter impart a strongly hydrophobic hydrophobic character to surfaces of the stickies, significant also for their removal by flotation. A patent application on the new pulping technology has been registered in the Polish Patent Office th under the number P 372730 dated February 10 , 2005. The special compositions of the chemicals to be used for deinking processes with the new pulping technology, according according to the claims of that patent application, were also registered under brand name De-Stick-Ink in the Polish Patent Office under th number Z-289769 dated January 8 , 2005. The new pulping process was finally tested at the Metsä Tissue Krapkowice mill (Poland). The mill tests were performed over several days. The proper chemical admixture was added to the pulper and the changes in the pulp properties were monitored along the deinking process, with special focus on the deinking and “de-sticking” effects. Some promising results were obtained, though the optimised De-Stick-Ink chemicals could not be tested for some reasons. The results gained in systematic investigations realized under industrial conditions, in the Krapkowice deinking line, according to the statistical factorial design of experiments, experiments, approved the possibilities to exert consciously the controlling influence on properties of the sticky particles and on the detachment of ink particles, during the pulping of recovered papers run according to the new pulping technology. It was shown that both parameters (µ and σ) of the dislocated lognormal distribution of the macro-stickies size, and the pulp brightness as well, were influenced in the statistically significant way by the power dissipation and by the concentration in the slurry of De-Stick-Ink (DSI); the agent elaborated earlier to deinking and to agglomeration of the sticky particles, according to the new pulping technology. Running of the pulping step according to the new pulping technology has not caused any problems in the industrial scale. However, further steps of the deinking process were run in the routine way during the mill trials, and the activities were limited only to taking samples and collecting some measurement results characterising particular steps of the deinking process to detect such positive influence of the new pulping technology on separating the stickies in subsequent deinking process steps, as well as to identify the obstacles in the way of fully utilising so valuable alterations in properties of contaminant particles made during the pulping step consciously run according to the new pulping technology. As it was found, additional additional quantities of the macro-stickies macro-stickies were generated during the mill trials, however, outside the pulping of recovered papers, in further steps of the deinking process, especially during the processes of screening and separating the contaminants. It was the bookbinding hot-melt
150
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
adhesive that was the source of such additional stickies. Larger pieces of the bookbinding hot-melts, which survived the pulping process, were divided into minute particles in the different screening steps (pulper screen and coarse screens). This additional admixture of such stickies, as non-sticky particles o in the slurry but detected as the macro-stickies by the INGEDE method n 4, resulted in a considerable increase of the macro-stickies content in the pulp stream, making serious difficulties in clearer evaluating an influence of the new pulping technology on separation of the macro-stickies from the pulp having been deinked. Moreover, such additional stickies from the bookbinding hot-melts were generated outside the pulper, so without due interaction with the agglomerant added to the pulping step, according to the new pulping technology. Nevertheless, it was possible to detect some positive influences of the new pulping technology on separation of the sticky particles in subsequent steps of the Krapkowice deinking line. An extrusion of the flat-shaped macro-stickies through slots of the slotted screen manifested itself after the routine pulping, resulting in higher maximal size of the macro2 2 stickies (in screening accepts) equal to 0.4 mm , in comparison with 0.32 mm , characteristic of the results gained after the new pulping technology. This does show such a positive aspect of the new pulping technology in which, thanks to the processes of agglomeration and probably also coagglomeration, the granulation of the sticky particles is progressing, creating oblong particles (granules) (granules) less susceptible to their comminution, comminution, as well as not proper for their extrusion through slots of the slotted screen. Such granules of the sticky particles (perhaps also with other contaminant particles) were more efficiently separated in the first stage of hydrocyclones, hydrocyclones, during the cleaning step. After stock preparation preparation according to t o the new pulping technology, t echnology, the cleanliness efficiency in those hydrocyclones was additionally improved to 85% regarding separation of the macro-stickies smaller 2 than 0.05 mm from the pulp. However, there was a lack of such a positive influence of the new pulping technology on separating of the macro-stickies in the flotation step. Besides the additional stickies (created after the pulper from bookbinding hot-melts) complicating the evaluation, this was also caused by the very flotation process run in the Krapkowice deinking line in the way aimed mainly at stock de-ashing, i.e. removing finely dispersed filler particles (in size about of the one order smaller than the macro-stickies size) from the pulp by their flocculation with minutely dispersed air bubbles. Despite the positive influences of stock preparation according to the new pulping technology on separation separation efficiencies of the macro-stickies, perceived during the screening and cleaning processes, it was no possible to show such obvious superiority of the new pulping technology over the routine way of stock preparation in the pulper, expressed in terms of lower content of the macro-stickies in the pulp after the fully completed deinking process. Generally speaking, however, such combination of deinking and de-sticking processes, in which the new pulping technology is applied, requires the properly adjusted management of accepts and rejects in some key points of the deinking line. In any case, however, analyses of the flow balances of the macro-stickies in the entire deinking process are needed, as well as the control system of the macro-stickies content and properties should be established in the key points of the process, to put into practice the new pulping technology. This is possible in existing deinking lines. However, the importance of proper design of the deinking line in helping to remove not only ink particles (and toner or fillers) but also sticky particles should be pointed out. To that end the detailed analyses of the macro-stickies in the mill trials, discussed above in short, may also be useful and sufficiently solid basis for such new designs of the deinking line aimed at combining together deinking and de-sticking, with applying the new pulping technology.
Conclusion and perspectives
The comparative pilot tests performed with conventional pulping technology confirmed the gentle action of the drum pulper compared to the batch pulper, since larger stickies particles were produced, which should then be easier to remove by subsequent slot screening. However, as the batch pulper was only tested under conventional conditions and showed, according to further research performed outside this project, the possibility to be operated efficiently at lower pulping time, it is considered that drum pulpers should not bring clear advantages regarding PSA stickies compared to batch pulper operated under optimised conditions. This conclusion is in line with a recent paper reporting large mill experience, which suggested even lower stickies fragmentation with batch compared to drum pulpers. Beside the optimisation of conventional pulping technology based on concept to keep stickies large in order to remove then almost completely by screening, the new pulping process developed at ICP offers other possibilities to improve the global removal of stickies in deinking lines. It was shown that the pulping step run according to this new pulping technology, almost without any additional costs, must be followed by the modified management of selected accepts and rejects in some steps of the
151
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
deinking process, to satisfy both deinking of the stock and its de-sticking. And therefore to exploit fully such positive effects of the new pulping technology in a given deinking line, the systematic analysis of the stickies problem in that deinking line was required, applying the methods elaborated in this project. After such analyses the strategy should be worked out for applying the new pulping technology technology to the t he stickies problem abatement. Implementation of that strategy should be successful in existing deinking lines, though further improvements may be expected in such modified deinking systems, consciously oriented towards removing not only ink particles (or toner and filler) but sticky particles too, according to the new pulping technology. There is a need to cooperate with suppliers of deinking line equipment, and also with producers of chemical additives for papermaking. A conceptual framework of such further activities to put into practice the new pulping technology directed towards both deinking and de-sticking, is under evaluation together with an equipment supplier, ICP and CTP.
5.1.2. Optimisation of macro-stickies removal The different techniques available to remove macro-stickies, include pressure screening, centrifugal cleaning and froth flotation as well as combined systems.
5.1.2.1.
Screening
Experimental results
The main conclusions of the pilot screening tests were the following: -
Coarse as well as fine screening should be performed, as far as possible, at low temperature in order to minimise the fragmentation of soft PSA stickies (from acrylic adhesives) and their extrusion through the slots.
-
Screen cylinders with low contours (0.6 mm) should be used, at least at the fine screening step (0.15 mm wedge wire slots), and the passing velocity should be kept relatively low (about 1 m/s) in order to improve the stickies removal efficiency while keeping screening costs and solid losses within reasonable limits.
Simulation of screening systems
The two 3-stage screening systems shown in figure 136 have been simulated: -
System A is a conventional conventional 3-stage 3-stage cascade system with feed-forward feed-forward second stage accepts and and feedback third stage accepts
-
System B is a 3-stage cascade system with special series feed-forward arrangement at the second and third stages, as described in [77]
Screening system A
Screening system B
Figure 136 : Simulated screening systems
152
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The four following screening conditions were compared for the two screening systems, with the same screening conditions conditions at each stage: -
2 wedge wire screen plates:
0.15 mm slots with low contours of 0.6 mm (06 32 13° WW) 0.15 mm slots with high contours of 1.2 mm (12 29 25° WW)
-
2 passing velocities:
1 m/s (low passing velocity) and 2.5 m/s (upper limit in mills)
The consistency was not taken into account as it was shown that its influence was low under 1.5 % consistency, i.e. in the low-consistency screening screening range. The dilution flow rates were set in such a way to have a feed consistency of 1% at each stage. The reject flow rates were varied between 2 % and 20 % in order to achieve low final reject rates in all cases, keeping in mind that the lower values of 2 or 4% reject flow rate are not realistic (a fourth stage should be added in real situation). The input data required for the simulations simulations are the characteristics of the pulp and stickies at the inlet of the screening systems and fibre and stickies passage ratio distribution. The simulations were done for one type of stickies containing pulp. The characteristics of the fibres were taken from mill samples analyses. Figure 137 left shows the fibre length distributions of deinking pulps taken from two deinking lines of a newsprint mill, respectively in the first loop at the inlet of the medium-consistency slot-screening step and in the second loop at the low-consistency screening inlet. The fibre length distributions are given in total length of fibres in each size class of 200 µm. The input data used for the simulations were based on these curves (after smoothening). The fibre coarseness function was given by wl = 0.067*l + 0.026, where the fibre coarseness (w l) is given in mg/m and the fibre length (l) in mm. The pulp was assumed to contain 65 % of these fibres, 20 % fillers with a passage ratio of 1 and 15 % fines with a passage ratio of 0.9, in all cases.
8
Stickies size distribution (handsheet image analysis) Newsprint DIP line 1 - MC
100
Newsprint DIP line 3 - LC
6
Pilot tests series n°3 p l u p g / s e 10 i k c i t s ² m m
i l . ) % ( s 4 e r b i f i n 2
1
0 0
1
2
3
4
5
0, 0
6
0, 5
1, 0
1, 5
2, 0
Stickies particle size (mm)
Fibre length (mm)
Figure 137 137: Fibre length distributions in number and in length for newsprint deinking pulps pulps (left)
and stickies size distribution of pilot tests and hypothesis used for the simulations (right) The stickies size distribution used for the simulation was that of the pilot stickies screening tests performed for the optimisation of screen plate design. The stickies “sizes” corresponded to the spot sizes measured in handsheets. Figure 137 right shows the experimental data (in mm² stickies spots in the different stickies sizes classes) and the “smoothened” stickies size distribution, i.e. an exponential function (linear distribution in logarithmic scale) which should be most relevant as there are normally much more small stickies than large stickies particles. As described in section 4.2.4, 4.2.4, the behaviours of fibres or stickies stickies under given screening screening conditions are characterised by passage ratio distributions, which are determined experimentally. For the fibres, the passage passage ratio distributions distributions are given given by P (l) = exp - (l/ λ) β with β = 0.6 according to the experimental data. However, since the pilot stickies screening tests were performed with a mixture of bleached chemical fibres (50% softwood and 50% hardwood), it was not possible to use directly the
153
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
experimental λ and β values. The procedure used to evaluate the λ and β values of the deinking pulp used for the simulations was the following: -
Calculation (with the help of the simulation tools) of the λ value which for β = 0.6 gives the global experimental pulp passage ratio given in table 32, the calculation being based on the mixed-flow model with 50 % reject flow rate and with the fibre length distribution and fibre coarseness function of the bleached chemical pulp mixture, according to the experimental screening conditions.
-
Use of this λ value (with β = 0.6) in the simulations, which assumes that a bleached chemical fibre and a recycled fibre have the same passage ratio for a given fibre length (in real situation, the TMP fibres should have a lower passage ratio as they are stiffer, while the recycled chemical fibres should have a higher passage ratio as they are supposed to be more flexible).
Type of screen plate / Passing velocity
1 m/s
2 m/s
3 m/s
0.15 mm slots low contours 06 32 13° WW
0.56
0.75
0.90
0.15 mm slots high contours 12 29 25° WW
0.71
0.86
0.95
Table 32 : Experimental pulp passage ratios used for the simulation of fibres in screening systems For the stickies, the simulation input data were based on the experimental pilot screening tests results. Average values of the results obtained at 2 and 3 m/s passing velocity were used to evaluate the stickies passage ratio distributions at 2.5 m/s. This allowed increasing the precision as well as the relevance of the input data as passing velocities as high as 3 m/s are very unlikely to be used in mills. The input data used for the simulation were slightly modified in such a way to avoid passage ratios above 1 (experimental values can be slightly higher than 1) and assuming the same passage ratio decreases decreases for the large stickies particles having a passage ratio of less than 0.1, which corresponds corresponds to using the same slopes of the experimental curves drawn in logarithmic scale, as shown in figure 138.
10,000
10,000 Pk - 0.15 WW 0632 - 1m/s
Pk - 0.15 WW 1229 - 1m/s
Pk - 0.15 WW 0632 - 2.5m/s
s o 1,000 i t a r e g a s s 0,100 a p s e i k c i t 0,010 S
Pk - 0.15 WW 1229 - 2.5 m/s
s o 1,000 i t a r e g a s 0,100 s a p s e i k c i t 0,010 S
0,001
0,001 0
1 2 Stickies spot size in handsheets (mm)
3
0
1
2
3
Stickies spot size in handsheets (mm)
Figure 138 138: Experimental stickies passage ratio distributions used for the simulations
A number of simulations were performed performed for different internal screen flow models, i.e. the mixed-flow and the plug-flow model as described in section 4.2.4, as well as in the hypothesis of clear water and process water dilution of the screening rejects. Process water consistency was assumed to be 0.3 g/l, including 0.2 g/l mineral fillers and 0.1 g/l fines and the particle passage ratios were assumed to be constant and equal to 1 and 0.9 for respectively the fillers and the fines. The detailed results of the different simulations can be found in the progress report D19. Some of the results obtained with the mixed-flow model are illustrated in figure 139 where the solid losses and the removal efficiency of the stickies larger than 0.2 mm are given as a function of the single screening stage reject flow rate and in figure 140 where the two screening systems can be compared at equal solid losses.
154
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Compared to the screening system A, the system B removed significantly more stickies at given reject flow rate but generated about two times higher fibre losses than the conventional screening system A, since there were two screens in series which produced rejects at the second and third stages.
Screening system A
Screening system A - Stickies > 0.2 mm 100
10 0.15 0.15 0.15 0.15
8 ) % ( s e 6 s s o l
WW WW WW WW
0.6 0.6 1.2 1.2
-
1.0 2.5 1.0 2.5
m/s m/s m/s m/s
90 80 ) % ( y c n e i c i f f E
m e 4 t s y S 2
70
0.15 0.15 0.15 0.15
WW WW WW WW
4
6
0.6 0.6 1.2 1.2
-
1.0 2.5 1.0 2.5
m/s m/s m/s m/s
60 50 40 30 20 10
0
0 0
2
4
6
8
10 12 14 16 18 20 22
0
Single screen re ject flow ra te (%) (%)
2
8
10 12 14 16 18 20 22
Single screen reject flow rate (%)
Figure 139 : Solid losses and efficiency of systems A (process water dilution, mixed-flow model)
The comparison of the different screening conditions and screening systems at given reject rates is illustrated in figure 140 showing the simulation results limited to more realistic reject flow rates (8, 10, 12, 14, 18 and 20%) as the lower rejects flow rates (4 and 6%) are normally not used in mills.
Screening system A - Stickies > 0.2 mm
Screening system B - Stickies > 0.2 mm
100
100 0.15 0.15 0.15 0.15
90 80 ) % ( y c n e i c i f f E
70
WW WW WW WW
0.6 0.6 1.2 1.2
-
1.0 2.5 1.0 2.5
m/s m/s m/s m/s
80 ) % ( y c n e i c i f f E
60 50 40 30
70
0
0 4
5
Screening system losses (%)
1.0 2.5 1.0 2.5
m/s m/s m/s m/s
30
10 3
-
40
10 2
0.6 0.6 1.2 1.2
50
20
1
WW WW WW WW
60
20
0
0.15 0.15 0.15 0.15
90
0
1
2
3
4
5
Screening system losses (%)
Figure 140: Efficiency vs. solid losses of systems A and B (Rv = 8 to 20%, process water, mixed flow)
The simulations clearly showed that the best results should be achieved first by the use of low-contour low-contour screen plates (0.15 mm wedge wire slots with 0.6 mm contour height) and then by a reduction of the passing velocity (1 m/s). System B generated about two-times more rejects at given reject flow rate (for example a minimum of about 2 % losses compared to 1 % with system A to get the best results at the minimum reject flow rate of 8 % with the low-contour screen plates at low passing velocity) but achieved only slightly better stickies removal efficiency at given solid losses (about 60 % compared to 58 % efficiency with system A for 2 % solid losses under the best screening conditions). conditions).
155
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
By contrast, the simulations done under the hypothesis of plug-flow model (cf. progress report D19) showed that system B was more efficient than system A at given final losses. One explanation could be that system A, with 3rd-stage screening screening accepts fed back, re-circulates re-circulates the fraction of contaminants accepted at this stage and produces some kind of mixed-flow type behaviour at the secondary stages. Consequently, system A would not fully benefit from the efficiency increase produced when the single screens are changed from mixed-flow to plug-flow type, which could explain that system B, with no contaminant re-circulation, became more efficient with the plug-flow model. Further simulations showed that plug-flow type screens achieved higher efficiencies than mixed-flow type screens at given system losses, but that it was not possible to reach acceptable losses with a 3-stage screening system for the best screening conditions. In real situation plug-flow type screens are uncommon as even screens with closed rotor always include some internal mixing. Such screens should behave according to intermediate plug/mixed-flow models, which means that reject thickening and losses (as well as efficiency) should be lower. A fourth screening stage should be added. To sum up, the simulation results based on the pilot screening test results led to the conclusion that the best stickies removal efficiencies should be achieved: -
first by the use of low-contour screen plates (0.15 mm wedge wire slots, 0.6 mm mm contour height)
-
then by by a reduction of the passing velocity (1 (1 m/s), at the expense of higher higher screening screening costs,
-
with screens showing some plug-flow behaviour, at the expense of higher reject thickening,
-
and finally finally with more complex reject screening screening systems, also associated to higher higher screening screening costs.
5.1.2.2.
Cleaning
Extensive pilot cleaning tests were performed with various cleaners, including high-density forward cleaners with different head diameters (270, 130 and 65 mm) and a low-density through-flow cleaner (100 mm head diameter), to evaluate their efficiency with the two reference adhesives, a high-density 3 3 water-based acrylic adhesive (1.03 g/cm ) and a low-density hot-melt rubber adhesive (0.96 g/cm ). The pulping conditions of the adhesive labels changed significantly the density of the adhesive particles by soaking and by the adsorption of mineral pigments as the adhesive labels were stuck onto 3 newspapers. The final in-situ density of the adhesives was then increased by 0.03 to 0.04 g/cm , which was positive for the acrylic adhesives but led to neutral buoyancy particles and very poor cleaning efficiency with the hot-melt rubber adhesive. The stickies removal efficiencies were consequently very poor on average though increasing the pulp temperature improved slightly the efficiency. There was potential to further improve the efficiency by increasing stickies density by adding mineral adsorbents such as talc. Adding adsorbents to the stock and mixing them in the pulper led to a macro-stickies reduction in the cleaner by about 25% for the large cleaner, which removed mainly the larger macro-stickies. The addition of fresh "unused" mineral seemed necessary necessary for effective adsorption of the minerals to the stickies. The fillers already present in recovered paper did not produce a positive result on macro-stickies separation. The analysis of the efficiency curves observed with small and medium sized cleaners revealed a competition between centrifugal separation and shear-induced centripetal separation for the large stickies particles. The results obtained with these cleaners confirmed the poor average PSA stickies removal efficiency, but revealed a potential to remove stickies in the optimum particle size range. A new small-capacity small-capacity cleaner prototype (65 mm head diameter) was finally tested. The new cleaner achieved higher efficiency than optimised screening with 0.15 mm slots, for small stickies producing spots under about 0.6 mm in handsheets (see figure 88). The effectiveness of low-density cleaning in deinking seems to be restricted to the selective removal of some hot-melt PSA and glues (e.g. for advertising inserts) as re-pulped adhesive labels and tapes produce mainly high-density stickies. The flow conditions in rotary cleaners should normally promote the agglomeration of high-density stickies at the cleaner wall. Macro and micro-stickies balances in a deinking and a packaging paper recycling mill did however not reveal any agglomeration of stickies, which might have been used to remove them by subsequent fine slot screening.
156
L SCREEN C CLEAN EAN
5.1.2.3.
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Flotation
Laboratory Laboratory flotation tests were first performed with model macro-stickies and clean water: -
Using model stickies with an inlet inlet stock consistency of 1.0% and standard deinking chemicals, the total separation efficiency achieved amounted to approx. 45% based on the absolute values and approx. 60% based on the stock balance.
-
In the case of smaller macro-stickies (<1000 µm, INGEDE Method n°4), separation efficiency values of 70 - 80% were achieved on average using a 25l VOITH laboratory flotation cell and a flotation time of 12 minutes. Hence, flotation always tends to separate out smaller macro-stickies. macro-stickies.
-
No macro-stickies reduction was possible without the use of flotation flotation chemicals.
-
The stock stock consistency consistency in the flotation flotation cell inlet had had a significant impact on on the results of flotation. Increasing the stock consistency from 1.0 to 1.25 caused the separation efficiency of macrostickies to fall by 20%. A further increase in stock consistency to 1.5% reduced the separation efficiency by a total of 40%.
-
For the stock suspension under test, adding a special hydrophobic mineral to the standard deinking chemicals formulation brought no increase in flotation efficiency.
-
The flotation flotatio n result depends on the amount of fatty acid used. A 50% reduction in the standard dosing amount reduced the removal efficiency by approx. 25%. A 50% increase in fatty acid dosage resulted in a slightly improved macro-stickies flotation whilst increasing the total losses.
Laboratory flotation trials were also performed with a stock suspension from a deinking mill: -
An absolute value for macro-stickies macro-stick ies separation of approx. 25% was measured during the flotation of a mill stock suspension from a tissue mill in the laboratory flotation cell. Precisely this macrostickies separation efficiency was also achieved in the mill flotation plant.
-
A study of macro-stickies macro-stickies separation separation in the flotation plant of of a newsprint newsprint paper mill that that was carried out within the scope of system analysis produced a macro-stickies reduction by approx. 40-50% in the flotation stage.
-
Smaller macro-stickies are preferably floated even in mill suspensions.
The pilot deinking flotation trials were performed with a Voith-Eco cell and with model macro-stickies and clean water. The results can be summarised as follows: -
The macro-stickies macro-stick ies reduction was in the range between 50 and 70%. Raising the consistency consistenc y in the flotation inlet from 0.85% to 1.35% caused an approx. 25 % relative decrease in removal efficiency (see figure 119).
-
The preferred flotation of smaller macro-stickies sized < 1000 µm (INGEDE Method n°4) was observed in both trials.
-
The pilot pilot and and lab flotation cells achieved achieved similar similar results in terms of of macro-stickies macro-stickies reduction.
The conclusions are as follows: -
Deinking flotation flotation is generally generally suitable suitable for separating separating macro-stickies. macro-stickies. Total Total macro-stickies macro-stickies separation efficiency values of approx. 40-50% can be achieved in well functioning flotation plants.
-
Deinking flotation mainly separates smaller macro-stickies macro-stickies and micro-stickies micro-stickies with a high high efficiency efficiency that can be as much as 70%.
-
The macromacro- and and micro-stickies micro-stickies reduction reduction clearly clearly exceeds exceeds the fibre losses losses and even the the filler reduction. In industrial practice, the fibre losses will be even lower due to the secondary flotation stage, i.e. approx. 2-3 % in a 5-step plant.
-
The use use of mill process process water instead of clean water could result result in a considerably considerably lower lower stickies stickies reduction, however. however. A 50% lower removal efficiency than in the tests is quite probable.
-
On the whole, the stickies reduction by flotation flotation is a very useful useful addition addition to pressure screening. Deinking flotation removes preferably smaller macro-stickies, which can only be eliminated to a limited extent by the screens.
157
L SCREEN C CLEAN EAN
5.1.2. 5.1.2.4. 4.
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Global macro -stic kies remov al pro cess
Macro-stickies removal systems based on screening with additional cleaning or flotation on screening rejects have been simulated (figure 141) in order to evaluate the impact of reject cleaning or flotation on the overall stickies removal efficiency and the total system losses. The idea was to improve the selectivity of the separation between fibres and macro-stickies at the last screening stages where coarse fibres difficult to screen are concentrated, with the assumption that there are enough highdensity (or low-density) stickies which can be removed by high-density (or low-density cleaning) and/or hydrophobic hydrophobic stickies which can be removed by flotation.
Combi ombin ned Scre Screen enin ing g & Clean leanin ing g syst system em “C” “C”
Comb Combin ined ed Scre Screen eniing & Flota lotattion ion syst system em “F” “F”
Figure 141 : Simulated combined screening & cleaning and screening & flotation systems
Simulation of combined screening and cleaning system
The simulated system included two stages of high-density cleaners on the accepts of the two last stages of a 4-stage screening system, as shown in figure 141. High-density cleaners were chosen as the most common reference PSA led to high-density particles. The input data for the cleaners were those of the pilot cleaning tests (figure 88), which gave the best results: -
Small cleaner:
65 mm head diameter
-
Pressure drop:
150 kPa
-
Reject flow rate:
6%
-
Feed Feed consistency:
0.8 % at the 1 stage
-
Fibre reject thickening factor:
2.5 at the 1 stage
-
Fillers reject thickening factor:
1.2 at the 1 stage
-
Fines reject thickening factor:
1 at the 1 stage
-
Efficiency on stickies:
fig. 88 at the 1 stage
st
nd
0.4 % at the 2 stage
st
3.5 at the 2 stage
st
1.5 at the 2 stage
st
st
nd nd
nd
1 at the 2 stage nd
double value (max 90%) at 2 stage
The simulation results showed as previously a higher efficiency with the plug-flow model compared to th the mixed-flow model, with still no possibility to reach acceptable losses despite the 4 stage. The systems B (3 stages) and C (4 stages with cleaners on the 2 last stages) are compared in figure 142. The implementation of cleaners on the accepts of the two last stages together with the changes of the rd th arrangement of the last stage screens, i.e. in two stages (3 and 4 stage) instead of in series at the rd 3 stage, improved the overall stickies removal efficiency by about 10%, for example from 60 to 66% efficiency with the best screening conditions at equal solid losses of 2%. In addition, the 4-stage system with the cleaners allowed to reach these 2% losses with about 14% reject flow rate at each stage instead of 8% with the 3-stage system, which improves the runability of the screens.
158
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Screening system B - Stickies > 0.2 mm
Screening system C - Stickies > 0.2 mm
100
100 0.15 0.15 0.15 0.15
90 80 ) % ( y c n e i c i f f E
70
WW WW WW WW
0.6 0.6 1.2 1.2
-
1.0 2.5 1.0 2.5
m/s m/s m/s m/s
90 80 ) % ( y c n e i c i f f E
60 50 40 30
70 60 50 40 30
20
20
10
10
0
0 0
1
2
3
4
5
0.15 0.15 0.15 0.15 0
1
2
WW WW WW WW 3
0.6 0.6 1.2 1.2
-
1.0 2.5 1.0 2.5
m/s m/s m/s m/s
4
5
Screening system losses (%)
Screening system losses (%)
Figure 142 : Efficiency vs. solid losses (Rv = 8 to 20%, process water, mixed flow, stickies > 0.2 mm)
Combined screening and flotation system
The simulated system included one-stage flotation cells implemented on the accepts of the two last stages of a 4-stage screening system, as shown in figure 141 (right). Two different flotation conditions were considered, i.e. flotation at usual fibre consistency and special low-consistency flotation as reported for the existing mill case [37-38]. The input data were the following: -
Flotation cells: Feed consistency: Reject flow rate: Fibre losses: Fillers losses: Fines losses: Efficiency on stickies:
based on the pilot tests reported in section 4.4.3 0.8 % for standard flotation 0.4 % for low-consistency low-consistency flotation 2% 2% 20 % 5% Er = 70 – 35 x for x ≤ 1.8 then Er = 2 % for standard flotation Er = 55 – 35 x f or x ≤ 1.4 then Er = 2 % for LC flotation
These efficiency formulae where where x stands for to the stickies spot size (in mm) evaluated in handsheets are based on the experimental data in figure 119, obtained with a pilot flotation cell using clean water.
100 ) % ( y c n e c i c i f f E l a v o m e R
90
Pilot Flotation C=0.85%
80
Pilot Flotation C=1.35%
70
Lab Flotation C=1.00%
60 50 40 30 20 10 0 0
1
2
3
Stickies spot size INGEDE n°4 (mm)
Figure 143 : Experimental stickies flotation data and hypotheses used for the simulation:
low efficiency line for standard flotation and high-efficiency high-efficiency line for LC stickies flotation
159
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The influence of stickies size was supposed to be linear as shown in figure 143, which includes the experimental results of figure 119. It was also assumed that the efficiency would be lower in mills where process waters are contaminated with chemicals. A further shift towards lower efficiencies was then applied as the experimental stickies removal efficiencies referred to the INGEDE method n°4 while the simulations referred to stickies spot surface area in handsheets. This means that the input data used to simulate the flotation cells were not as rigorous as those used for the screens, thought the hypotheses led to flotation efficiencies which were consistent with mill results [38]. The results in figure 142 showed that the implementation of efficient stickies flotation conditions, i.e. low-consistency flotation, on the last stage screening accepts led to about the same final efficiency than with efficient cleaners (right curves in figures 142 and 144), while conventional flotation conditions were less effective in removing macro-stickies from screening rejects.
Screening system F - Stickies > 0.2 mm
Screening system F - Stickies > 0.2 mm 100
100 Mixed flow model - Flotation
90
80
80 ) % ( y c n e i c i f f E
Mixed-flow Mix ed-flow - LC Flotat ion
90
70
) % ( y c n e i c i f f E
60 50 40 30
0.15 0.15 0.15 0.15
20 10
WW WW WW WW
0.6 0.6 1.2 1.2
-
1.0 2.5 1.0 2.5
m/s m/s m/s m/s
70 60 50 40 30
0.15 0.15 0.15 0.15
20 10
WW WW WW WW
0.6 0.6 1.2 1.2
-
1.0 2.5 1.0 2.5
m/s m/s m/s m/s
0
0 0
1
2
3
4
5
0
1
2
3
4
5
Screening system losses (%)
Screening system losses (%)
Figure 144 : Efficiency vs. solid losses (Rv = 8 to 20%, process water, mixed flow, stickies > 0.2 mm)
The efficiency gain achieved with low-consistency flotation compared to cleaning of the screening rejects showed to be higher when the stickies between 0.1 and 0.2 mm size were included in the global stickies removal efficiency evaluation since flotation removes more efficiently the smaller stickies (figure 143 versus figure 88).
Conclusion
The highest macro-stickies removal efficiency gains should clearly be achieved with the optimisation of the screening conditions, i.e.: -
first by the use of low-contour screen plates (0.15 mm wedge wire slots, 0.6 mm mm contour height)
-
then by by a reduction of the passing velocity (1 (1 m/s), at the expense of higher higher screening screening costs,
-
with screens showing some plug-flow behaviour, at the expense of higher reject thickening,
-
and finally finally with more complex reject screening screening systems, also associated to higher higher screening screening costs.
The implementation of cleaners or flotation cells on screening rejects streams should be regarded as an interesting option to improve further the macro-stickies removal efficiency without increasing the fibre losses, especially regarding regarding the coarse and long fibres which are difficult to screen. Finally, it is believed that the optimisation of the first high-consistency screening steps in terms minimised macro-stickies fragmentation and improved removal efficiency is probably the corner stone towards a most complete removal of the PSA stickies in deinking lines.
160
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Further macro-stickies removal efficiency gains could be expected on the main stream process steps, with the optimisation of the cleaning conditions and more particularly with the optimisation of the physical-chemical flotation conditions (a priority research area), keeping in mind that the flotation process should primarily be optimised to remove selectively the inks while minimising the solid losses. Flotation is indeed regarded as the most promising technique towards further improvement of removal of the smaller stickies, i.e. the very smallest macro-stickies and the micro-stickies.
5.1.3. Optimisation of micro-stickies removal Micro-stickies are defined as particles with size between 5 to 100 µm. The amount of micro-stickies (measured by extraction method) in the stock suspension is mostly more than 10 times higher than the macro-stickies content. For that reason and because of a low grammage of the paper produced in DIP mills, also micro-stickies can have a significant effect in building up deposits, even if the particle size is somewhat lower in comparison to the macro-stickies. Figure 145 illustrates the characteristics microstickies should generally have in deinking lines.
A
B
C
50 µm
particle from adhesive layer
pigment + binder binder from printingink
pigment+ binder from papercoating
adhesiveor adhesiveor binder binder pigmentfrom pigment from printingink printingink pigmentfrom pigment from papercoating cluster pigments+ adhesive+ binder
Figure 145: Kind of micro-stickies in deinked pulp
5.1.3.1.
Flotation
Within this project the micro-stickies content was measured by solvent extraction methods, mainly with Dimethylformamide (DMF) used at PTS. When extracting the filter cake from stock consistency measurement all several kinds of micro-stickies were included in the measurement. Especially binders from printing inks and coated broke are a main part of the total extract.
Lab flotation
In lab flotation there were made no micro-stickies research within the project, but some basis trials in the field of colloidal potential stickies were performed at CTP. These particles are per definition smaller than the micro-stickies and had a particle size lower than 5 µm. The trials were performed with clean water and with process water without the presence of fibres. Model colloidal adhesive particles were efficiently removed by flotation even without the addition of collectors; PVA model stickies displayed a slightly better floatability than acrylic ones. Total removal efficiency was more than 90%. PVA based stickies floatability was not significantly affected by the type of process water used. Very high flotation efficiencies were always found (90 to 95 %) whatever the process water used for the preparation of the dispersion. On the contrary, the floatability of acrylic based stickies was influenced by the nature of the process water.
161
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Pilot Flotation
During the deinking flotation trials using a pilot Voith-Eco cell there have been used a lot of coated and printed paper (newspapers and magazines) as fibre source. After pulping the stock suspension included a big amount of coated broke and printing ink particles which partially consist of binders. A micro-stickies reduction of approximately approximately 80 % was calculated for the pilot trials when using clean water for dilution. So the micro-stickies reduction was nearly in the same range as the reduction of macro-stickies in the smallest particle size class during flotation. Especially the selective removal of the printing inks and binders has lowered the micro-stickies content after flotation significantly.
Mill flotation
In addition to the pilot results, the real micro-stickies removal was measured in several mills at the pre-flotation pre-flotation and post-flotation steps (see figure 107). The analysis of the deinking flotation stages of four different paper mills showed an average reduction in micro-stickies load by approx. 50%. The best removal efficiency was at 70%. Both flotation steps, pre-flotation and post-flotation showed a good removal efficiency concerning extractable substances.
Conclusion
The good possibility of micro-stickies removal by deinking flotation is a very valuable result. Apart from thickeners, no other method has been available so far for removing these particles from stock suspensions. Moreover, thickening stages discharge the micro-stickies into the process water so that they are reintroduced into the suspension by dilution.
5.1.3. 5.1.3.2. 2.
Process water treatment
Extensive tests were performed with a pilot pressure filter equipped with micro-holes (100 µm) or with ultra-fine slots (20 to 60 µm). The trials have shown that pressure filtration is unable to remove finely dispersed micro-stickies from the water circuits. The flow of micro-stickies in the screen was only like in a valve in relation to the volume related reject flow (see figure 133). Dissolved air flotation (DAF) or micro-flotation, micro-flotation, the technique most frequently used for treating process waters in deinking lines, showed to be significantly more efficient in micro-stickies removal than pressure filtration (see figure 135). The pressure filter is employed for fibre recovery in the process water, perforated screens being recommended for this purpose. As far as fibre recovery is concerned, the pressure filter is in direct competition with conventional disk filters and drum filters. One advantage of the pressure filter is certainly its simple operation without sensitive filter media and without the frequently necessary dosage of additional fibrous materials such as that required in disk filters.
5.1.4. Conclusion Minimising the fragmentation of stickies at the first deinking process step, during the re-pulping of the adhesives with conventional pulping technology, i.e. with drum pulpers or optimised batch pulpers, is clearly a prerequisite for successful subsequent stickies removal at the slot screening steps. Besides conventional pulping technology, the new “De-Stick-Ink” pulping technology developed at ICP showed promising results on pilot and mill scale, but needs further optimisation. Stickies fragmentation should further be minimised during the first screening steps (by similar means such a reducing mechanical forces or temperature) in order to remove them as early as possible in the deinking line. Optimised fine slot screening conditions were defined, which reduce stickies extrusion (low temperature) and improve the removal of PSA stickies, the best results being achieved with lowcontour screen plates at relatively low slot velocity.
162
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Cleaning showed much lower possibilities to remove macro-stickies compared to fine slot screening because of the close to neutral density of the PSA particles, though small-diameter cleaners were able to remove some small macro-stickies macro-stickies difficult to remove by screening. screening. Flotation removed only small macro-stickies and showed good possibilities to remove micro-stickies, which is regarded as a most valuable result since no other method is available to remove selectively such detrimental particles from the stock suspension and the process waters. Pressure filtration was effective in recovering fibres from process waters but showed no possibility to remove micro-stickies. Knowledge about the individual deinking process steps involved in the formation and removal of PSA stickies has been clearly developed in the framework of this project and allowed to improve the global understanding of the stickies behaviour in deinking lines.
5.2.
Recycling friendly adhesives
One of the objectives to be achieved in this project further to the scientific and technical cooperation developed between the adhesive supplier Jackstädt / Avery Dennison and the other project partners was to establish guidelines guidelines for the development of new recycling friendly adhesives. Indeed, products based on Pressure Sensitive Adhesives (PSA) are used for many other applications. Examples are self-adhesive labels, medical and functional tapes and self-adhesive graphics. These products are widely present in all kinds of market segments, such as the pharmaceutical, the industrial, the automotive, the graphic and the consumer Market. Pressure sensitive labels are of main concern for causing problems during the paper recycling process, because they are produced in volumes of billions of square meters of which a part will end up in the paper recycling process. General descriptions of pressure sensitive adhesives, followed by a more in depth review of emulsion and hot melt adhesives, are given in this section on the basis of internal Avery Dennison knowledge (more information can be found in [112], which gives a good overview of PSA technology), before to conclude with recommendations to improve pressure sensitive adhesives in such a way to facilitate the paper recycling process.
5.2.1. Pressure sensitive adhesives 5.2.1. 5.2.1.1. 1.
What is a press ure sensi tiv e adhesiv e?
Usually a PSA consists of a face material of paper or film, a pressure sensitive adhesive and a liner or backing. The liner is removed during the application phase. The adhesive can be based upon a large number of polymers: polyacrylates, natural and synthetic rubbers like styrene butadiene, isoprene, and silicone rubbers. A tackifying resin is often added to improve the performance. Label adhesives are in general based on emulsions or hot-melts. A typical characteristic of a PSA is that it is permanently “tacky” and that it forms a bond with a surface by pressure. Depending on the type of adhesive and the applied force, the binding force, “the peel adhesion”, builds up to a constant level (figure 146 left)
5.2.1. 5.2.1.2. 2.
What determi nes the tack and and the adhesion of a PSA?
Tack
The so-called “tack” is of high importance for label applications. The tack is the initial “grab” of an adhesive to a substrate. In physical terms the tack is the initial adhesion at very short contact time. A relatively high tack is needed to make sure that a label stays on a substrate (e.g. bottles) during dispensing in a high-speed industrial industrial label line. Labels will not stay on a substrate if the tack is too low.
Peel-adhesion
The (peel) adhesion is the binding force after prolonged time. It is also of high importance, because a PSA should have sufficient adhesion to keep a label on the substrate during use. The tack and adhesion are determined by the rheology and the surface tension (polarity) of the adhesive.
163
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Rheology of the adhesive
The rheology determines determines how the adhesive flows over the substrate. An important characteristic is the Glass Transition Temperature Tg. This is the temperature at which the polymer chains start to move if the adhesive is heated. During the heating process the polymer goes through different states. From a glass (brittle) phase at low temperature, through a transition phase into a rubber phase (figure 146 right). The mid point temperature in the transition state is defined as the Glass Transition Temperature Tg. PSA’s usually have Tg’s in the range of -60, -20 °C for label applications. applications.
14 12 e 10 c r o f 8 n o i s 6 e h d a 4 2 0 0
2
4
6
8
10
12
time
Figure 146 : Development of adhesive adhesion force and typical PSA rheological properties
A high Tg will result in a low flow and high internal strength of the adhesive (resistance to shear forces). A low Tg will result in a high flow, which result in a high tack, low application temperature and lower peel adhesion (figure 147, left).
80
80 70
70
60
peel
50
e c r 40 o f
60
peel
50
tack
e c r 40 o f
tack shear
shear
30
30 20
20
10
10 0
0 0
5
10
15
20
0
25
glass transition temperature
5
10
15
20
25
Polarity
Figure 147 : Relations between peel, tack and shear forces and adhesive Tg (left) and polarity (right)
Surface tension
The surface tension determines if the adhesive is able to flow over the substrate. The surface tension should be lower than the surface tension of the substrate, otherwise there is no flow. (In analogy with wetting of glass with water. The surface tension of the water is too high to wet glass. Wetting only occurs if a soap, a surfactant, is added to the water to reduce the surface tension). A high surface tension of an adhesive results in a poor wetting of low energy substrates like polyethylene, polyethylene, which ends up in a low peel adhesion. Low surface energy adhesives exhibit a high peel adhesion to low energy surfaces, because the wetting of the substrate by the adhesive is better. A high polarity of an adhesive improves the internal strength of the adhesive. This is due to polar-polar interactions. The adhesive is consequently (improved cohesion) more resistant to shear forces (figure 147 right).
164
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Molecular weight
The molecular weight is also a factor, which has an influence on the adhesion. A high molecular weight results in a low flow and consequently in a low tack. A high molecular weight results in a high internal strength of the adhesive. The internal strength is an important parameter, because it determines the degree of bleeding of the adhesive. Adhesive bleeding is a phenomenon that occurs over time. In the most extreme situation the adhesive will bleed out of a roll. In this case it will contaminate conversion equipment like a printer or dispenser. The above shows that a label adhesive should meet a number of requirements to make sure that it fulfils the demands of the end use. These demands limit the freedom to develop an adhesive that is easy to remove during the paper recycling process. But there is more, because a large number of additives are added to hot melts and emulsion adhesives. These additives are needed to improve the adhesive properties and to facilitate the process ability of the adhesive during the adhesive manufacturing and coating process. In the next chapter more information is provided about the adhesive manufacturing process and the additives that are generally used for emulsion and hot melt adhesives.
5.2.1. 5.2.1.3. 3.
Emuls ion adhesiv es
The emulsion PSA as used in the Screen Clean study belongs to the group of water based emulsion polyacrylate (acrylic) adhesives. adhesives. Emulsion acrylic PSA’s mostly contain tackifying resins to improve the adhesion to all kind of substrates. Tackifying resins can be selected from either the group of rosin, which originates from pine trees or from hydrocarbon resins, which consists of polyterpenes and synthetic hydrocarbon resins. Emulsion PSA’s also contain many additives to keep the emulsion stable: emulsifiers, defoamers, thickening agents, additives to prevent microbiological decomposition, cross-linkers etc. Traces of non-reacted monomer are also present. To produce a PSA construction consisting of a front, adhesive and a backing paper, the water-based emulsion is coated onto a siliconized backing paper. After drying and laminating to an appropriate face material, the PSA construction is ready for further conversion like printing, die cutting and dispensing of the label onto a substrate. Dry coat weight ~20 g/m². The total composition of the PSA construction has an influence on the paper recycling process. Each component component will have its own, often unknown, contribution.
The composition of an acrylic adhesive
In principle the design of an acrylic ink, paint, lacquer or pressure sensitive adhesive is based upon the same technology. Optimization of the performance for the end use takes mainly place by varying the polymer composition and molecular weight. The basis of the composition of an acrylic PSA is the acrylic ester:
H
H
H
hydrogen => acrylic
CH3
methyl acrylate
C2H5
ethyl acrylate
C=O
C4H9
butyl acrylate 2 ethyl hexyl acrylate
O–R
CH3 - CH - C4H8 I C2H5
C = C H
R can consist of many monomers, like:
165
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The selection of the ester monomers determines the glass transition temperature Tg, which is the temperature temperature at which the glass like matrix changes and begins to flow. The Tg is an important factor for the performance of a PSA. Polymers based on a single monomer do not make good pressure sensitive adhesives and therefore different monomers are used. Typical PSA monomers and their homopolymer Tg’s (°C) are listed below: Methyl Acrylate
15
Ethyl Acrylate
-15
n-Butyl Acrylate
-45
2-Ethyl Hexyl Acrylate
-60
Methyl Methacrylate
105
Butyl Methacrylate
20
Styrene
100
Acrylonitrile
105
Vinyl Acetate
28
The overall glass transition temperature of the resulting polymer is a function of the monomer building blocks and can be easily calculated with the Fox equation: 1 — = Tg
w1 w2 w3 —— + —— + —— + Tg1 Tg2 Tg3
.....
w = weight fraction of monomers, Tgn = glass transition transiti on temperature of polymers Compared to inks and paints, pressure sensitive adhesives are soft. Usually they have a Tg in the range of –60°C to –20°C. Commonly used polymers for pressure sensitive adhesives are 2-ethylhxylacrylate, butylacrylate, ethylacrylate and acrylic acid. The adhesive as used in the study is of the butylacrylate type.
The polymerization process
The pressure sensitive polymer can be produced by either solution- or by emulsion- polymerization. For emulsions the emulsion polymerization process process is generally used. A typical formulation consists of the raw materials listed in the table below. The raw materials monomer
60-65%
water
rest
surfactant
1.5-3.0%
initiator
0.3%
buffer
0.1%
coating package
0.3%
biocide
0.1%
The free radical polymerization reaction takes place in so-called micelles, which are formed by the surfactant molecules. The monomer is finely dispersed in small droplets in the water phase. The droplets act as reservoir and monomer migrates from the droplets into the micelles in which the polymerization takes place. The growing polymer particles adsorb the dissolved surfactant, which ultimately leads to the disappearance disappearance of micellar surfactant.
166
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The free radical polymerization process Growing macroradical
Monomer
Monomer droplet
Growing particle
Non nucleated micelle Free surfactant Initiator
Figure 148 : The free fr ee radical polymerization process
Three steps are of importance: initiation, propagation and termination (figure 147).
Initiation
I–I => I· + I· I· => M
®
(·) is a radical
I–M·
I
is Initiator
M
is Monomer
Propagation
I –M –M· + M => I-M-M· I-M-M· + M => I–M–M–M· I–M–M–M· + M => . . . . . . . . => Polymer ·
Termination
Polymer · + Polymer · => Polymer–polymeR
and/or
=> Polymer + Polymer
(combination) (disproportionation)
After the polymerisation polymerisation process the pH of the emulsion is acidic (pH < 6), because of the presence of the acrylic acid. To improve the stability the pH is increased to a level between 6.5 and 7.5. This is done by the addition of NH 3.
Tackifying resins
PSA’s require a balance of three main properties: peel adhesion, cohesion and tack. The tack of a pure acrylic adhesive for paper label application is too low and to improve this, tackifying resins are added. (Range 15 – 35 % dry/dry). The most commonly used resins for tackification are rosin derivatives, polyterpene resins and hydrocarbon resins:
167
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
1) Wood Rosin and its Derivatives Three types are used: gum rosin, wood rosin and tall oil rosin. All generated from the pine tree. Rosin resins are not polymers, but a blend of different molecules. Rosins are not stable, because of the high degree of unsaturation. This can be improved by disproportionation and hydrogenation. 2) Polyterpene Resins Alpha-Pinene Alpha-Pinene and Beta-Pinene and d-Limonene. d-Limonene. Fractional distillation of the turpentine yields terpene monomers, from which polyterpene resins can be polymerised. Expensive resins and not commonly used for emulsion adhesives 3) Hydrocarbon Resins There are three major types of hydrocarbon resins: C5
- aliphatic resins
C9
- aromatic resins
DCPD
- cycloaliphatic resins
In general Gum and Tall Oil Rosin resins and C9 and DCPD Hydrocarbon resins are used for emulsion adhesives. Tackifiers for emulsion adhesives are finely dispersed in water. Like for an emulsion acrylic, surfactants are added to stabilize the dispersion. A range of anionic and non-ionic surfactants is used to improve the stability. Like emulsion adhesives the pH is 6.5 to 7.5
Coating of the wet adhesive
The manufacturing process of a PSA construction consists of coating, drying and laminating. The adhesive is ready to coat after blending of the polymer with tackier(s) and other additives. In general the wet adhesive is coated onto a silicone backing paper (release liner). A number of coating methods are available like bar-, roll-, blade-, slot die- and curtain-coating. Each coating method requires a specific surface tension and rheological behaviour of the wet adhesive. Therefore surfactants, thickeners and defoamer are added. In particular surfactants are important to improve the mechanical stability of the emulsion and to reduce the surface tension to wet the low energy surface of the silicone liner. Drying is carried out in ventilated ovens. Dry adhesive coat weight ~20 g/m². The emulsion emulsion adhesive adhesive coating process
Figure 149 : The coating process of emulsion adhesives adhesives
During the drying process, heat and mass transfer take place simultaneously. As a result the water evaporates and the % solids of the adhesive goes up. The evaporation process can take place via several mechanisms. 1) Diffusion Diffusi on in a homogeneous adhesive mass 2) Capillary flow, if capillaries capillari es are developed during drying 3) Flow caused by shrinkage and pressure gradients In case of emulsion adhesives it is believed that the mass transfer mechanism is mainly driven by diffusion and capillary flow.
168
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The film formation process is an important phase in the drying process. In this phase the individual emulsion and tackifier resin particles ‘melt’ together into one phase. The film formation process can be divided into six stages: 1) Evaporation of water. The bulk of the water evaporates and the some particles reach their critical particle distance 2) Particles near near the surface surface reach reach their critical critical particle distance and and flocculate 3) All particles are densely packed 4) The water that that is present in the the voids between between the particles particles evaporates and and the particles particles start to deform 5) The last water diffuses to the surface 6) A homogeneous film is formed The pH of the emulsion plays an important role for stabilization of the emulsion prior and during coating and during the film formation process. The optimal pH of an emulsion adhesive is 6.5 – 7.5. During the drying process the NH 3 evaporates and the emulsion becomes instable, which facilitates 3 the film forming process. Density of the dried adhesive: adhesive: 0.95 – 1.0 kg/dm .
5.2.1. 5.2.1.4. 4.
Hot melt press ure sensi tiv e adhesives
A different class of pressure sensitive adhesives is hot melts. Pressure sensitive adhesives are high viscosity visco-elastic materials. Unlike in solvent based or water born (emulsion) adhesives, adhesives, hot melts do not require a vehicle to reduce the viscosity to the level needed for the coating process. Reduction of viscosity is achieved by heating and melting the product. Typically the coating temperatures ranges from 140 to 180 °C. Solidification of the coated film is achieved by cooling the laminate back to ambient temperature. The invention of thermoplastic block co polymeric rubbers in the early 1960’s made this possible.
The composition of a hot melt
Rubber based PSA’s contain three main types of components: block copolymers, tackifying resins, and plasticizers. In addition small quantities of anti-oxidants and in some cases mineral fillers are present. Block copolymers are almost exclusively of the Styrene-Isoprene and Styrene-Butadiene types. Tackifying resins can be based on rosin-esters, polyterpenes or mineral oil hydrocarbons. Plasticizers are usually lower molecular weight hydrocarbon resins or oils. The components are mixed in sigma blade mixers in batch mode or in twin-screw extruders in continuous mode. Coating is done by roll, die and extrusion coating. Dry coat weight ~20 g/m². Typical Formulation:
Rubber
~ 30 %
Tackifying Resin
~ 30 %
Plasticizer
~ 30 %
Anti-Oxidant
~ 0.5 %
Fillers
~5%
Density of adhesive: 0.95 – 1.0 kg/dm
3
Hot melt characteristics
PSA’s require a balance between three key properties: adhesion, cohesion and tack. Rubber based hot melts, unlike acrylic adhesives, do not have chemical crosslink’s. Cohesion is the result of physical crosslink’s caused by the incompatibility of Styrene and Isoprene and Butadiene as illustrated in figure 150 left.
169
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Polystyrene forms solid domains in a rubbery or sticky matrix. The Polystyrene domains melt above 120 °C, causing the viscosity to drop by a factor of 1000 or so. As in acrylic PSA’s the Tg is an important important characteristic characteristic of the hot melt PSA. The Tg of of the adhesive is determined by the Tg of the rubber phase. The Fox equation as explained under acrylic emulsion holds also in this case. The overall Tg of the adhesive is the result of the weight fractions of the low Tg rubber materials, Isoprene (- 56 °C), Butadiene (- 80 °C), high Tg tackifying tackif ying resins (+ 50 °C), and plasticizers with a Tg’s close to the overall adhesive Tg (-45 to -20 °C). Blends of rubber with resins may have the right Tg but are typically to “hard” to have PSA properties. The role of plasticizers is to soften the blend to the right hardness. Tackifying resins used in hot melt adhesives typically have higher softening points then the ones used in acrylics. Chemically the same classes of resins are used however. -
Rosin based resins
-
Polyterpene resins
-
Hydrocarbon resins
Hydrocarbon aliphatic C5 resins, which are not compatible with acrylics, are widely used in hot melt adhesives, as they are very compatible with isoprene rubber.
Hot melt adhesive properties
By nature of the thermoplastic rubbers and hydrocarbon resins and plasticizers used, hot melt adhesives generally have much more affinity to a-polar substrates such as Polyethylene. Tack and adhesion levels to those substrates are higher as compared to acrylic adhesives. Hot melts by their hydrophobic nature are much more water-resistant than emulsion acrylics.
5.2.2. Potential influence of adhesive components on the separation separation of stickies stickies
Hot melts versus acrylic emulsions
A hot melt adhesive can be considered considered as solid islands in a sticky sea, kept together by physical crosslinks as pictured above. Emulsion adhesives however consist of internally, chemically, crosslinked particles. The individual particles can still be recognized in a dried adhesive film as shown in the SEM picture in figure 150 right.
Figure 150: Structure of hot-melt rubber (left) and acrylic emulsion adhesive (SEM picture: right)
The hydrophobic hydrophobic character of a hot melt in combination with the physical crosslink’s makes it likely that hot melts will have a tendency to agglomerate in a repulping and de-inking process. Emulsion acrylics by their more hydrophilic and a particle structure will have a far larger tendency to redisperse. Redispersing Redispersing will be facilitated by the presence of surfactants in emulsion adhesives adhesives
170
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Molecular composition
The selection of monomers will have an impact on the absorption of water. The more polar the more affinity for water, which results in gelling of the polymer. The degree of chemical or physical crosslink’s will reduce the speed of gelling.
Molecular weight
The molecular weight will have an influence. Although also depending on the polarity of the adhesive, an adhesive of relatively low molecular weight will tend to absorb water pretty quickly and will turn into a gel type phase. For high molecular weight this takes more time. The level of molecular weight should also have an influence on the average particle size of adhesive particles in the repulping process. It is expected that a high molecular weight will result in a larger particle size. The strength of physical and chemical cross-links will also play a role here.
Additives
Surfactants are most likely the most important adhesive additives that influence the stickies separation process. As outlined before, surfactants play a role in the polymerization process. But that is not the only function. Surfactants play also an important role in preventing the soft polymer agglomerates to coagulate during the polymerization process and during storage. During the coating process they are essential, because they improve the mechanical stability and lower the surface energy to a level that is needed for wetting of the silicone coated backing paper. In general anionic and nonionic surfactants are used. Anionic surfactants stabilize the polymer particle by creating a charged surface. Nonionic surfactants create a hydrated layer on the particles. Surfactants are generally added in a 1.5 – 3 % (w/w) range and are still present in the dried adhesive and consequently have an influence on the recycling process. They might a) facilitate the speed of water absorption by the adhesive and b) might play a role in stabilizing of the re-dispersed adhesive particles. Playing with the pH of the paper slurry pH could be of help to improve separation of adhesive particles. Another option might be to add additives to reduce the effect of surfactants.
5.2.3. Recommendations to improve PSA’s Some thoughts to improve the separation of stickies during the repulping process are listed below. These ideas are not based upon experimental results, but on what could be done if we look at the adhesive composition. composition. Experimental work needs to be done to assess the feasibility.
•
Make use of hot melt instead of emulsion adhesives.
•
Increase/decrease density of adhesive. Current label adhesives have a density that is close to water, which makes it hard to separate stickies by gravity or centrifugal forces. A larger difference would facilitate the separation efficiency.
•
Increase molecular weight of the adhesive. This might result in larger stickies particles, which are easier to separate.
•
Decrease level of surfactants in adhesive.
•
Make use of surfactants that are ineffective in paper repulping process, e.g. pH or temperature sensitive systems, or add chemicals to the repulper that in effects the adhesive surfactants.
•
Develop adhesive that does not absorb water.
•
Develop adhesive that does not get soft at repulper temperature.
•
Make use of water stable paper to which the adhesive sticks during the repulping process. Paper and adhesive could be relatively easy removed from the pulp stream.
In theory some of these thoughts could be helpful, but one should bear in mind that most of the adhesive modifications will have an impact on the adhesive properties and add costs. This could be the showstopper. At the moment there is no real market pull to develop recycling friendly adhesives. But that could change in the course of time.
171
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
6. Exploitation and dissemination of results results 6.1.
Exploitation
The findings developed in the framework of this project are currently exploited and should be further exploited by the partners and their clients, in different ways depending on the partners status, i.e. the research institutes (CTP, ICP, PTS and LEGI), the suppliers to the pulp & paper industry (AFT and Jackstädt / Avery Dennison Europe) and the European pulp & paper industry. One patent application about project findings, i.e. a new pulping technology, has been registered by th ICP in the Polish Patent Office under the number P 372730 dated February 10 , 2005. As agreed at the beginning of the project, the ownership of this patent lies within the invention organisation, ICP. The special compositions of the chemicals to be used for deinking processes with the new pulping technology, according to the claims of that patent application, were also registered under brand name th De-Stick-Ink in the Polish Patent Office under number Z-289769 dated January 8 , 2005.
Exploitation through the research institutes
CTP, ICP, PTS and LEGI are using the results of the project for the development of their expertise to be exploited and disseminated under their own policy. This applies to practically all the knowledge developed in the different work-packages, which is currently exploited and will continue to be exploited in framework research projects (public and/or private funded projects, multi-client projects), private contracts, consultancy to deinking mills and education activities. Special actions have been undertaken towards further development and industrial exploitation of the new patented pulping technology developed by ICP.
Exploitation through the pulp & paper industry
The large dissemination of the project results to the European paper mills and Companies should lead to the exploitation of the several project results by deinking mills. Project results could be exploited by the recycling mills, either indirectly as far as new and less detrimental adhesive products will be recycled with the recovered papers, or directly by the implementation of new strategies and processes for the optimisation of the removal of adhesives in deinking lines. The industrial exploitation of project results should include the application in mills of the findings developed in the following areas: -
Pulping: further development and/or adaptation to mill cases of the new patented process
-
Screening: implementation of the screening conditions, which were shown to achieve the best stickies removal efficiencies. These include the optimisation of the operating conditions of the existing screens, the rearrangement of the equipment in the screening systems including the implementation of new equipment and the choice of most efficient screen plates. Access to the simulation of screening systems to evaluate the impact of such changes.
-
Cleaning: information about the possibilities and limits of cleaners could be used to optimise the cleaner operating conditions and make decisions regarding new equipment.
-
Flotation: the high potential of flotation in removing the smallest macro-stickies as well as microstickies (mainly responsible for deposit problems on the paper machine) has clearly been demonstrated in this project and will draw the attention of the deinking line operators on a better control of the physical-chemical conditions in this process step, in order to improve the removal of stickies et reduce deposits.
In addition the dissemination of the project results to the R&D centers of the large European paper companies will enhance bilateral cooperation with the research institutes and contribute to long-term implementation of the project results in mills, as research will continue in most of the areas treated in this project. These include further research on the flotation process, to continue to develop the basic knowledge gained about stickies flotation phenomena and the role of tensides.
172
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Exploitation through the equipment supplier
The basic knowledge developed in the field of screening (including numerical simulation) and the experimental results gained on pilot scale are exploited by AFT to provide expertise and optimised screening equipment to the paper mills. As the same models are used for the simulation of screening systems both at CTP (PS2000) and at AFT (SimAudit), the input data gained during the project can also be used by AFT to expertise and optimise screening systems for their clients.
Exploitation through the adhesive suppliers
The cooperation between the paper research institutes and the R&D laboratories at Jackstädt and Avery Dennison Europe has has been very fruitful in this project and contributed, contributed, besides the exchange of scientific knowledge, to develop a better mutual understanding of the difficulties induced on one side by the adhesive product quality requirements for the paper converters and on the other side by the recycling of adhesives with the recovered papers by the paper producers. New more recycling friendly pressure sensitive adhesive products were not developed in the course of the project, but background information has been made available to a major adhesive supplier for the development of such adhesives, keeping in mind that paper converters are not ready to accept higher costs for more recycling friendly adhesive products.
Exploitation of the new (patented) pulping technology
The new pulping technology developed by ICP has been tested on industrial scale in the Krapkowice deinking mill and showed promising results, but the some industrial constraints did not yet allows to conclude definitely definitely about the new process and strategy to remove both stickies and inks in the form of chemically and mechanically agglomerated particles. Further improvements may be expected and there is a need to cooperate with suppliers of deinking equipment and chemical additives before the new process can be implemented in deinking lines on larger time and scale. Cooperation with Kadant-Lamort (France), the equipment supplier who provided the deinking line at the Krapkowice mill, has been discussed in the framework of a special partners meeting organised by the coordinator and ICP before the final project meeting. All the data (pulp sample analyses) from the mill trials, which where performed at the very end of the project, were not available at this meeting. Discussions and exchange of information’s are in progress. A side application application of the new stickies stickies agglomeration agglomeration process, which which consists in applying applying the process process to the treatment of screening rejects in order to agglomerate the stickies into large particles and improve consequently their removal at the last screening stage, was envisaged as suggested in section 4.1.5. In order, to promote the exploitation of the new process, it has been agreed, together with ICP and after the approval of the CTP/CTPi members, that this approach will be included in an on-going CTP project devoted to the reduction of deinking and recycling rejects.
6.2.
Dissemination
A widespread dissemination dissemination of of the project results towards towards the pulp & paper paper industry has been been ensured in the course of the project, essentially through the pulp & paper research institutes. At CTP periodic Technical Meetings with the French papermakers papermakers are organised every six months. In addition, the participation of the CTPi members (international CTP structure to include papermakers from mills outside France) to these Technical Meetings has extended the dissemination of the project results to a large part of the European companies (R & D centers) and mills operating deinking lines. The dissemination of project results was also done in the framework of the Research Forum organised each year at CTP for the CTP/CTPi members.
173
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
PTS had published the results of the Screenclean project periodically in the PTS R&D-Forum “Recovered paper technology”. The forum takes place 2 times a year and is the meeting point of PTS experts in stock preparation with our industrial partners from paper mills, chemical suppliers and machine industry. The main purpose of the R&D-Forum is the direct dissemination of actual research results to the industry. On the other side the industry can give their expertise to bring the research projects to the best suitable result . ICP has been ensuring the dissemination of project results to mills and paper research institutes in countries of Central Europe. Information about that new pulping technology, elaborated in the th ScreenClean project, was thoroughly presented in Poland during the 15 International Papermaking Conference PROGRESS ‘05, Wroclaw, September 28-30, 2005. AFT and Jackstädt / Avery Dennison have disseminated project results to the Pulp & Paper industry through direct contacts with their clients as well as through their participation in conferences. conferences. Project results were also disseminated to papermakers through training sessions, seminars and conference organised together by CTP and PTS. These include: th
-
6 CTP/PTS Advanced Training Course on Deinking, Grenoble, March 18-20, 2003
-
1 CTP/PTS Training Course on Paper & Board Recycling, Grenoble, March 25-26, 2003
-
11 PTS-CTP Deinking Symposium, Leipzig, 27-30 April 2004
-
7 PTS Stickies Seminar, Dresden, Dresden, April 5-6, 2005
-
7 CTP/PTS Advanced Training Course on Deinking, Grenoble, May 31/June 1-2, 2005
-
2 CTP/PTS Training Course on Paper & Board Recycling, Grenoble, May 31/June 1-2, 2005
st
th
th th
nd
LEGI ensured the dissemination of project results also outside the pulp & paper industry in the form of papers published in scientific magazines and presented to seminars in the field of fluid mechanics. The contribution in this project of fundamental research institutes in the field of fluid mechanics, i.e. LEGI (CNRS, Grenoble), LEMTA (CNRS, Nancy) and ITM (University of Czestochowa, Poland) has opened the scientific research field and led to the development of further fruitful cooperation outside and after the duration of the project. The list of the papers published in scientific magazines and technical revues and presented in open conference and seminars is enclosed in the annexe. The public reports have been made available on the web, at CTP’ web site ( www.webCTP.com www.webCTP.com)) where this final report will also be placed. A presentation of the ScreenClean project project will be given during the next meetings of COST Action E48 “The Limits of Paper Recycling” to be held in Brussels on November 28-30, 2005 and this final report will be placed on the COST Action E48 web site (www.cost-e48.net ( www.cost-e48.net)) in order to ensure a wide dissemination dissemination and easy access to the project results over the coming years.
7. Policy related benefits 7.1.
Communities added value and contribution to EU policies
The pulp and paper industry, as well as its allied and downstream downstream industries, is based on the utilisation of sustainable resources, wood and recovered fibres, available all over Europe. Virgin fibre supply is strong in Nordic countries whereas the recycled fibre is the dominant raw material in Central Europe. This has led to a certain, though not equally pronounced specialisation of the production range in the various regions or countries. Both effects result in a very intensive exchange of raw materials and finished paper and board products within Europe.
174
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
The pulp and paper industry manufactures products which meet essential demands of the individuals and the society in important areas, such as printed information, protective packaging and hygiene. Such paper grades can all be produced on the basis of recycled fibres, as far as the recycled pulps meet the quality requirements of the final product. This is more particularly challenging in the field of deinking where the highest potential growth of the recycling rate relies in the increased use of deinked pulps for the manufacture of higher printing paper grades, such as SC and LWC papers. The pulp and paper industry is very international with a significant amount of trade both within the EU and across its boarders. This means that the market is a target for competition from certain areas of the world with abundant raw material source and low production costs (Far East, South America) as well as from North America from time to time. The European pulp and paper industry needs to develop significantly if it is to retain its market positions. Ways to achieve this is through greater concentration on cost-effective production, increase of basic knowledge and higher value products. Solving the stickies problem is clearly of great importance for the paper recycling industry in Europe. Indeed, the presence of residual adhesives in the deinked pulp still causes a lot of problems on the paper machine and during paper printing and converting. This reduces the amount of deinked pulp which can be incorporated in high quality graphic papers, and consequently limits the increase of the recycling rate in Europe. To fulfil this global objective of the ScreenClean project would be of great importance for the increase of the cost effectiveness and product quality and would help to produce high-quality papers with higher recycled fibre content. This would contribute to increase even more the sustainability of the entire papermaking process. More selective separation technology and strategies to remove stickies should lead to a reduction of fibres losses and to more cost effective deinking. This would further enhance the eco-efficiency and the sustainability of the entire papermaking process. The main objective of the ScreenClean project was to develop new solutions to the stickies problems in deinking mills, which are far from being completely solved despite the recent progress in deinking technology and the efforts engaged towards recycling friendly adhesives, by the following approaches: -
The development of the understanding of the basic phenomena involved in the production and removal of stickies in the recycling process.
-
The development of new and improved deinking technology for the removal of stickies.
-
The elaboration of guidelines for the development of new recycling friendly adhesive products for the paper chain.
The understanding of the basic phenomena involved in the production and removal of stickies has clearly been improved in the framework of this project. This applies among others, to the development of the basic knowledge about stickies screening phenomena, as shown by the large number of papers published about about the project findings in this area, to a better understanding of the possibilities and limits of the centrifugal cleaning and pressure filtration techniques in removing stickies and to the opening of an important research field on the flotation of stickies. Deinking flotation was, indeed, shown to be a most crucial technology towards solving completely the stickies deposit problems on paper machines, since it is was shown to be most promising and only process, process, which can be used to remove selectively the micro-stickies, without increasing the solid losses in the deinking process. The objective of the development of new and improved deinking technology for the removal of stickies was globally met in that extent that optimised operating conditions have been established and were disseminated to the paper recycling and suppliers industry. The development of new technology refers particularly to the development and mill scale evaluation of the new pulping technology, for which a patent application has been filed and concrete actions have been engaged towards the exploitation of the new technology, as reported in the previous section. New more recycling friendly pressure sensitive adhesive products were not developed in the course of the project, but background information has been made available and guidelines were established for development of such adhesives, which might be produced in the future at acceptable costs in such a way to widespread their use by the paper converters.
175
L SCREEN C CLEAN EAN
7.2.
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
Contribution to Community social objectives
The contribution of the project to improving paper recycling technology, a societal and industrial issue of ever increasing importance, addresses the European Community objectives as defined in: -
Key action 5 "Sustainable agriculture, fisheries and forestry, and integrated development of rural areas including mountain areas"
-
Section 5.3.2 "Strategies for the sustainable and multipurpose utilisation of forest resources; the forestry wood chain"
-
"Environmentally friendly and efficient processes recycling technologies and improved value added products".
By enhancing recycling and promoting the use of higher amount of recycled fibres in graphic papers this project finding should contribute to improve the environment, the management of waste and consequently consequently the quality of life of citizens in Europe. If recycled fibre inputs are raised, waste volumes will diminish and save landfill. In addition, increased use of recycled fibres in a more cost effective way helps to protect natural resources in terms of lower consumption of non-renewable energy sources and/or greenwood as compared to primary fibre manufacture. The project result should also become usefull for the central European countries who are quickly developing their trade with the European Union. This results in serious and continuous growth of imports, either paper itself or other goods in paper packaging, and therefore a growing increase of highly converted papers in the waste paper collected in central Europe is progressing and its proper reuse in papermaking becomes limited with a harm to environment. More generally, improving the cost-effectiveness of the recycling process and the quality of the final product will contribute to the competitiveness competitiveness of the European paper industry. Finally, the excellent cooperation between the project partners should be regarded as a very positive outcome of the project. By bringing together experts from different part of Europe with their specific knowledge and experience, the synergy developed between the industrial partners, the applied and the fundamental research institutes, further increased the mass of European research in the field of stickies and deinking. Cooperation was more particularly extended towards Central Europe countries and Universities, and will continue after the completion of this project.
176
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
8. Literature cited 1.
Ringmann, J., Lombard, B. : “The recycling industry – European perspective, facts and Figures”, th 7 CTP / PTS Advanced Training Course on Paper & Board Deinking Technology, Paper n°2, st nd Grenoble, May 31 – June 2 , 2005.
2.
Julien Saint Amand, F., Le Ny, C. : “DIP fractionation and fibre upgrading”, INGEDE project 81 01 th CTP, 12 INGEDE Symposium, proceedings, 30 January, 2003.
3.
th .: “Recovered fibres for improved newsprint”, SC and LWC papers”, 27 EUCEPA Selder, H .:
Conference - Crossing the Millennium Frontier, Emerging Technical and Scientific Challenges: proceedings: 219-224, Grenoble, France, 11-14 October 1999. 4.
th preparation – Part 2: Particle separation processes”, 12 Julien Saint Amand, F. : “Review: Stock preparation
Fundamental Research Symposium, proceedings: 81-191, Oxford, 17-21 September 2001. 5.
Fabry, B., Roux, J.C., Carré, B.: “Characterisation of friction during pulping: an interesting tool to th
achieve good deinking”, Proceedings of the 5 PAPTAC Research Forum on Recycling: 1-6, o Ottawa, 28-30 Sept.1 999 & Journal of Pulp and Paper Science, vol.27, n 8: 284-288, August 2001. 6.
Ben, Y., Dorris, G.: “Irreversible ink redeposition during repulping: part II: ONP/OMG furnishes”, th
Proceedings of the 5 PAPTAC Research Forum on Recycling: 7-13, Ottawa, 28-30 Sept.1999. 7.
.: “Kinetic model of ink detachment in the repulper”, Proceedings Bennington, C.P.J., Wang, M-H .: th of the 5 PAPTAC Research Forum on Recycling: 15-21, Ottawa, 28-30 September 1999.
8.
Fabry, B., Carré, B.: “Comparison between different type of pulper devoted to deinking processes”, TAPPI 2002 Pulping Conference, San Diego, 8-10 September 2002.
9.
Galland, G., Carré, B., Cochaux, A., Vernac, Y., Julien Saint Amand, F .: “Dispersion and kneading”, kneading”, Paper Recycling Challenge, Challenge, vol. III : Process Technology, chapter 9: 131-149, Editors: M.R. Doshi & J.M. Dyer, 1998.
10. Julien Saint Amand, F., Perrin, B, De Luca, P. : “Stickies Removal Strategy”, Progress in Paper Recycling, vol. 7, n° 4 : 39-53, August 1998. 11. Steenberg, B. : “Principles of screening system design. Studies in screening theory I”, Svensk Papperstidning, vol.56, n°20: 771-778, October 1953. 12. Kubat, J., Steenberg, B. : “Screening at low particle concentrations. Theory of screening III”, Svensk Papperstidning, vol.58, n°9: 319-324, May 1955. 13. Wahren, D.: “Fundamentals of suspension screening”, Svensk Papperstidning, n°18: n°18: 539-546, 1979. 14. Nelson, G.L.: “The screening quotient: a better index for screening performance”, TAPPI Journal, vol.64, n°5: 133-134, May 1981. 15. Gooding, R.W., Kerekes, R.J.: “Derivation of performance performance equations for solid-solid screens”, The Canadian Journal of Chemical Engineering, v. 67: 801-805, October 1989. 16. Kumar, A., Gooding, R.W., Kerekes, R.J.: “Factors controlling the passage of fibres through slots”, TAPPI Journal, vol.81, n°5: 247-254, 247-254, May 1998. 17. Grégoire, G., Favre-Marinet, M., Julien Saint Amand, F., Serres, A., Fernandez de Grado, A. : “Modelling of the Productive and Reverse Flow at the Screen Surface”, COST Workshop “Current Research on Multi-Phase Flows in Paper making”, Paris, 12-13 November 1998. 18. Grégoire, G.: “Study and numerical simulation of the flow in an experimental pressure screen”, Ph.D. Thesis, Institut National Polytechnique de Grenoble, Grenoble, January 2000. 19. Gooding, R.W., Kerekes, R.J., Salcudean, M .: .: “The flow resitance of slotted apertures in pulp th screens”, 12 Fundamental Research Symposium, proceedings: 287-338, Oxford, 17-21 September 2001. 20. Grégoire, G., Favre-Marinet, M., Julien Saint Amand, F. : “Modeling of turbulent fluid flow over a rough wall without or with suction”, Journal of Fluids Engineering, vol.125, n°4: 636-642, July 2003 21. Pinon, V., Gooding, R.W., Olson, J.A.: “Measurement of pressure pulse from a solid core rotor”, Tappi Journal 2 (10): 9, 2003. 22. Wakelin, R.F., Corson, S.R.: “TMP long fibre fractionation with pressure screens”, International Mechanical Pulping Conference, proceedings: 257-265, 1995.
177
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
23. Niinimäki, J., Dahl, O., Kuopanportti, H., Ämmälä, A .: “A comparison of pressure screen baskets with different slot widths and profile heights”, Paperi ja Puu, vol.80, n°8: 601-605, 1998. 24. Olson, J., Roberts, N., Allison, B., Gooding, R.W.: “Fibre length fractionation caused by pulp screening”, Journal of Pulp and Paper Science , vol.24, n°12: 393-397, December 1998. 25. Weckroth, R., Grundström, K.J.: “Effect on screening performance with different screen cylinder aperture design”, African Pulp & Paper Week, Durban, 17-19 October 2000. 26. Julien Saint Amand, F., Perrin, B .: “Fundamental aspects of mechanical pulp screening”, International Mechanical Pulping Conference, preprints: 387-406, Helsinki, June 4-8, 2001. 27. Gooding, R., Olson, J., Roberts, N .: .: “Parameters for assessing fibre fractionation and their nd application to screen rotor effects”, 22 International Mechanical Pulping Conference, Proceedings: 407-424, Helsinki, June 4-7, 2001. 28. Julien Saint Amand, F., Perrin, B. : "Screening : Experimental Approach and Modelling", TAPPI Pulping Conference Proceedings: 1019-1031, Montreal, 25-29 October 1998. 29. Julien Saint Amand, F., Perrin, B.: "Fundamentals of Screening : Effect of Rotor Design and Fibre Properties", TAPPI Pulping Conference Proceedings : Orlando, 31 October – 3 November 1999. 30. Julien Saint Amand, F., Perrin, B. : "Fundamentals of Screening: Effect of Screen Plate Design", TAPPI Pulping / Process & Product Quality Conference, Boston, 5-9 November 2000. 31. Schabel, S, Respondek P. : “Screening: fundamental aspects of stickies removal”, Wochenblatt für. Papierfabrikation, vol.125, n°16: 736-739, end August 1997. 32. Heise, H., Schabel, S., Bangji, C., Lorenz, K. : “Deformation and disintegration physics of stickies th in pressure screens”, 5 Research Forum on Recycling, Ottawa, September 28-30, 1999. 33. McCool, M.A., Silveri, L.: "Removal of specks and non-dispersed ink from a deinking furnish", TAPPI Journal, 70(11) : 75-79, November 1987. 34. Julien Saint Amand, F., Bernard, E. Lamort, P .: "Entwicklung eines rotierenden Cleaners für hocheffizientz Leichtschmutzentfernung", 2nd PTS Deinking Symposium, München, 12-14 Februar 1985, und Wochenblatt für Papierfabrikation, 113 (20) : 779-784, Oktober 1985. 35. Julien Saint Amand, F., Perrin, B., Bernard, E. : "Modellierung und Dimensionierung von Cleanern. Eine vergleichende Untersuchung zur Abscheidung von Schmutzpunkten aus th deinktem Altpapierstoff", 4 PTS Deinking Symposium, München, 3-6 April 1990, und Wochenblatt für Papierfabrikation 8 : 295-302,1992. 36. Julien Saint Amand, F. : “Principles and Technology of Cleaning”, Paper Recycling Challenge, vol. III, Process Technology, 1998. 37. Heise, O., Kemper, M., Wiese, H., Krauthauf, E. : “Removal of residual stickies at Haindl Paper using new flotation technology”, TAPPI Journal, vol.83, n°3: 73-79, March 2000. 38. Engert, P., Haveri, M.: “Praxiserfahrungen bei der Reduzierung von Stickys in der Altpa Altpapie piera rauf ufbe bere reitu itung ng graphischer graphischer Papiere“, Wochenblatt Wochenblatt für Papierfabrikation, Papierfabrikation, vol.132, n°19: 1162-1167, 2004. 39. Galland, G., Bernard, E., Sauret, G. : "Aspect physico-chimique du désencrage", Revue ATIP, Vol. 31 n°10 : 374-378, Decembre 1977. 40. Larson, A., Stenius, P., Ström, G. : "Zur Oberflächenchemie des Deinking Prozesses", Wochenblatt für Papierfabrikation, Papierfabrik ation, Vol. 110, n°14: 502-506, 1982. 41. Putz, H.J. , Schaffrath, H.J., Göttsching, L. : "Deinking of oil and water-born printing inks - a new st flotation deinking model", 1 Research Forum on Recycling, proceedings: 183-190, Toronto, 1991. 42. Schwinger, K., Hanecker, E .: .: “Zur Flotationsneigung verschiedener Faserstoffe im Deinkingprozeß”, Wochenblatt für Papierfabrikation, vol.119, n° 21, 1991. 43. Santos, A., Carré, B., Roring, A.: "Contribution to a better understanding of the basic mechanisms involved in the pulping and flotation of offset ink particles", TAPPI Recycling Symposium Proceeding : 339-347, 1996. 44. Beneventi, D., Carré, B., Gandini, A. : “Role of surfactant structure on surface and foaming properties”, Colloids and Surfaces A, 189 (65), 2001. 45. Strauß, J., Großmann, H.: “Kreislaufwasserreinigung unter besonderer Berücksichtigung der Entfernung klebender Verunreinigungen”, Wochenblatt für Papierfabrikation, vol. 125, n° 9, 1997.
178
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
46. Carré B., Brun, J., Galland G. : "Incidence of destabilisation of pulp suspension on the deposition of secondary stickies”, Pulp and Paper Canada, 99(7): 75-79 1998. 47. Gonera H., Dabrowski J.: “A study of the method reducing dissolved and colloidal materials in process water by recovery of chemicals after re-pulping step of wastepaper deinking”, COST Action E1 Final Final Conference, Conference, Las Palmas de de Gran Canaria, Canaria, 24-26 24-26 November November 1998. 48. Zhang, X., Beatson, R.P., Cai, Y.J., Saddler, J.N.: "Accumulation of specific Dissolved and Colloidal Substances during white water recycling affects paper properties", Journal of Pulp and Paper Science, vol.25, n°6, June 1999. 49. Gassmann. H.: “Mechanical filtration of circuit water by pressure filters”, Paper Technology: 3340, March 2000. 50. Hamann, L., Strauss, J. : “Disturbing potential of pressure-sensitive adhesive and packaging th adhesives”, 10 PTS-CTP Deinking Symposium, proc.: paper 25, Munich, 23-26 April 2002. 51. European Project FAIR 98 3893: “Colloid control: Characterisation Characteris ation and control of colloids in paper mills recycling wastepaper”, final report, July 1998 - October 2001. 52. Delagoutte, T., Brun, J., Galland, G. : “Drying section deposits: identification of their origin”, IPE International Symposium - New Technological Developments in paper recycling, Valencia, Spain, June 12-13, 2003. 53. Hamann, L., Strauss, J. : “Stickies: definitions, causes and control options”, Wochenblatt für Papierfabrikation, vol. 131, n°11-12: 652-663, June 2003. 54. Leppänen, A. : “Views on Recyclability of Pressure Sensitive Adhesive Label in Europe”, TAPPI Recycling Symposium, proceedings: 385-389, Washington, 6-8 March 2000. 55. Onusseit, H .: .: “Physikalisch-chemische Eigenschaften Eigenschaften von Klebstofffilmen und deren Relevanz im rd Papierrecyclingproze Papierrecyclingprozess”, ss”, 3 PTS Stickies Symposium, proc., Munich, November 13-14, 2000. 56. INGEDE Method 4: "Analysis "Analysis of Macro Stickies in Deinked Deinked Pulp (DIP)", INGEDE PR 12/99, 12/99, www.ingede.de . 57. Delagoutte, T., Laurent, A. : "Modified method for quantification of primary stickies in recycled pulp", Progress in Paper Recycling, vol. 10, n°4, August 2001. 58. O’Connor, A.E., Macosko, C.W .: .: “Melt versus solvent coating: structure and properties of blockcopolymer-based pressure-sensitive adhesives”, Journal of Applied Polymer Science, Vol.86: 3355-3367, 2002. 59. Sahi, M., Rahouadj, R., Herbach, R., Choulier, D. : “Influence of the Viscoplasticity on the Ring Test Interpretation”., Journal of Materials Processing Technology, 58, 286-292, 1996. 60. Rahouadj, R., Cunat, Ch .: “A Nonlinear Viscoelastic Model Based on Fluctuating Modes”., Handbook of Materials Behavior Models, Academic Press, section 2.6, 107-116, 2001. 61. Cunat, Ch.: “The DNLR approach and Relaxation Phenomena”, Part I, Historical account and DNLR formalism., Mech. Time-Depend. Mater., vol.5, n°1, 39-65, 2001. 62. Gonera H., Marcinkowski T., Dabrowski J.: “An investigation of office wastepaper using image analysis”, SPP-TAPPI Symposium “East-European Paper Recycling”, Warsaw, 1-2 October 1997. 63. Serres, A.: “ID2/ID3 - Un nouveau bond dans les procédés d’épuration”, 53 ATIP, Session 7, 7, Bordeaux, Bordeaux, 17-19 octobre octobre 2000. 2000.
ème
Congrès Annuel
64. Griffin, A., Witczak, C., Knoke, T .: .: “Optimizing pressure screen parameters in a deinking plant using mixed office waste”, 2002 TAPPI Fall Conference & Trade Fair Proceedings, San Diego, November 2002. 65. Flanagan, J., Venditti, R., Jameel, H., Wilson, N., Weaver, N., Lucas, B. : “Passage of pressure sensitive adhesives through a slot”, Progress in Paper Recycling, vol. 11, n° 3: 17-23, May 2002. 66. Lucas, B.E., Venditti, R.A., Jameel, H. : “Factors affecting the passage of pressure sensitive adhesive particles through a slot”, 2002 TAPPI Fall Conference & Trade Fair, CD Proceedings, San Diego, TAPPI Press, November 2002. 67. Julien Saint Amand, F.: “Optimization of stickies removal in screens and cleaners”, Recent Advances in Paper Recycling – Stickies, Chapter 4: 78-125, Editor: Mahendra Mahendra R. Doshi, Appleton, July July 2002.
179
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
68. Gassmann, H.: “Mechanical separation of stickies – The challenge of the millennium”, TAPPSA Journal, 2002. 69. Julien Saint Amand, F., Wojciechowski, G., Asendrych, D., Favre-Marinet, M., Rahouadj, R., Skali-Lami, S.: “Screening: Fundamental studies on the extrusion of stickies through slots”, Revue ATIP, Vol. 58, n°1: 6-18, Février/Mars 2004. 70. Julien Saint Amand, F., Perrin, B., Gooding, R., Huovinen, A. : “Optimisation of screen plate design for the removal of stickies from deinking pulps”, Revue ATIP, Vol. 58, n°4, Août/Septembre Août/Septembre 2004. 71. Julien Saint Amand, F., Perrin, B., Frach, D., Asendrych, D. : “Visualisation of stickies extrusion th th through slots in pressure screens”, PTS Stickies Seminar, Paper n°5, Dresden, April 5 -6 , 2005. 72. Paul, S.T., Duffy, G.G., Chen, X.D. : “Viscosity control as a new way to improve pressure screen performance”, TAPPI Journal , vol. 83, n° 9, September 2000. 73. Julien Saint Amand, F., Perrin, B .: “Screening: State of the art and future: to understand th mechanisms and innovations better in order to improve quality and increase productivity”, 54 ATIP Annual Congress, Congress, proceedings, proceedings, Grenoble, Grenoble, 9-11 9-11 October October 2001. 74. Olson, J.A.: “Fibre length fractionation caused by pulp screening, slotted screen plates”, 86 Annual Meeting Meeting PAPTAC, proceeding proceedings: s: 21-28, February February 2000. 2000.
th
75. Carré, B., Ruiz, J., Ottenio, P., Brun, J.: “Optimisation of deinking water circuit design by th modelling”, 9 PTS/CTP Deinking Symposium, proceedings: paper 36, 9-12 May 2000. th
76. Lascar, A., Fejoz, R. : “Concept expertise and field experience in fine slot screening”, 8 PTSCTP Deinking Symposium, proceedings: Paper n°15, Munich, May 1998. 77. Rienecker, R .: .: “Sortierung von Altpapierstoff zur Herstellung von graphischen Papieren”, Wochenblatt für Papierfabrikation, n° 23/24 : 1149-1159, 1997. 78. Julien Saint Amand, F., Perrin, B. : “Fragmentation and removal of PSA particles in screens and rd cleaners”, 3 PTS Stickies Symposium, proceedings, Munich, November 13-14, 2000. 79. Julien Saint Amand, F., Perrin, B. : “Basic parameters affecting screening efficiency and fibre th loss”, 9 PTS-CTP Deinking Symposium, proceedings: paper 26, 9-12 May 2000. 80. Julien Saint Amand, F., Perrin, B .: “Optimisation of fractionation of chemical pulps with screens. th Pilot tests and modelling”, 7 CTP-EFPG Wood Chemistry and Pulp Technology Forum, Paper 8, th st February 28 – March 1 , 2002. 81. Christensen, L.: “Bird Triclean operational experiences with reprocessed paper”, TAPPI Secondary Fibre Conference Proceedings: 97-101, San Francisco, 15-17 September 1971. 82. Maves, K.L.: “Removal of contaminant by reverse cleaning”, TAPPI Secondary Fibre Conference Proceedings, 20-23 September 1976. 83. Chivrall, G.B.: "The Beloit Uniflow Cleaner, a novel concept in hydrocyclone technology", SPCI World Pulp and Paper week Proceedings: 266-271, Stockholm 10-13 April 1984. 84. Serres, A., Julien Saint Amand, F .: .: "Entwicklung des Leichtschmutzcleaners Gyroclean GYT", Wochenblatt fur Papierfabrikation, vol.118, n° 23/24: 1059, December 1990. 85. Julien Saint Amand, F.: "Epuration : limites techniques et économiques des technologies existantes. Perspectives nouvelles", 48e Congrès ATIP, Grenoble, 5-7 novembre 1985. 86. Gassmann, H.: “Avoidance of stickies by mechanical separation in screens and cleaners”, 51 Congrès de L’ATIP, session 3, 27-29 October 1998.
ème
87. Biza, P., Kaiser, P., Gaksch, E.: “Verbesserter Austrag von Stickies durch den Einsatz von Talkum“, 10th PTS-CTP Deinking Symposium, proc.: Paper n°31, Munich, April 2002. 88. Biza, P.: “Talc - A modern solution for pitch and stickies control”, Paper Technology, n°4: 22-24, 2001. 89. Küchler, A.: “Korngrößenbestimmung von Füllstoffen und Streichpigmenten“, PTS-Seminar Füllstofftechnik, Füllstofftechnik, Proceedings, Proceedings, 1998. 90. Völkel, H., Weigl, J .: .: “Die Bedeutung der Dispergierung von Füllstoffen am Beispiel SC-Papier“, PTS-Seminar Füllstofftechnik, Proceedings, 1998.
180
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
91. Weigl, J., Grenz, R .: .: “Einsatz von Spezialfüllstoffen zur Kreislaufentlastung“, PTS-Seminar Füllstofftechnik, Füllstofftechnik, Proceedings, Proceedings, 1998. 92. Julien Saint Amand, F., Perrin, B .: “Possibilities and limits of cleaners in removing PSA stickies”, KRICT-KTAPPI RTM 2004, Daejon, Korea, 21 & 22 June 2004. 93. Julien Saint Amand, F., Perrin, B. : “Characterisation and simulation of fibre separation in screens and cleaners”, International Mechanical Pulping Conference, Proc.: Québec, June 2-5, 2003. 94. Heise, O., Cao, B., Schabel, S.: “A novel application of Tappi 277 to determine macro stickies disintegration and agglomeration in the recycling process”, TAPPI Recycling Symposium, Proceedings: 631-644, Washington D.C., 6-8 March, 2000. 95. Stemmer, M .: .: “Experience of sticky reduction in a white-line-chipboard mill”, 2nd CTP-PTS Packaging Paper & Board Recycling Symposium, proceedings: paper n°19, Grenoble, 27-29 November 2001. 96. Krauthauf, E.A. : “Europe looks to the US Postal Service PSA project with great expectations”, TAPPI Recycling Symposium, proceedings: 375-383, Washington, 6-8 March 2000. 97. Hanecker, E.: „Verbesserte Abscheidung hydrophober feinteiliger Verunreinigungen in der zweiten Flotationsstufe der Altpapier-Deinking-Anlagen“, PTS-Forschungsbericht, (1998) 98. Britz H.: „Flotationsdeinking – eine Schlüsseltechnologie für Weiße und Sauberkeit“, Das Papier, Nr. 10 (1997), S. 514-519 99. Geistbeck M., Wiese H.: „Abscheidung von Stickies in der Flotation“, Wochenblatt für Papierfabrikation, Nr. 16 (1997) 100. Brunthaler J. ,Kemper M.: Papierfabrikation, Nr. 1/2 (2000)
„Praxiserfahrungen
mit
der
EcoCell“,
Wochenblatt
für
101. Nerez R., Johnson D., Thompson E.: “Laboratory repulping and flotation studies of three pressure sensitive adhesives”, Progress in paper recycling, May 1997 102. Nelson N., Hsu P.: “Effectiveness of inonically charged chemicals as flotation aids in stickies removal, during mixed office wastepaper recycling”, Tappi Recycling Symposium, conference book, New Orleans 1997 103. Klein R.,Schwarze D., Großmann H.: „Beitrag zur Bewertung des Einflusses der Luftblasengröße auf das Deinkingergebnis“, Wochenblatt für Papierfabrikation, Nr. 21 (1994) 104. Stratton R.: “The flotation of sticky contaminants from recycled fiber streams”, Progress in paper recycling; No. 4 (1992) 105. McKinney R.: “A better insight could help flotation technology take of”, Pulp and paper international, Nr. 6 (1998) 106. Voosen F., Voosen L.: “Identification and reduction of stickies and stickies related contaminants in ONP/OMG recycled newsprint”, Tappi Pulping conference, conference book, Atlanta 1997 107. Hornfeck K., Nellessen B.: „Verbesserung des Flotations-Deinking-Prozesses durch Optimierung der Rejektflotation“, Wochenblatt für Papierfabrikation, Nr. 17 (2000) 108. Li B., Hipolit K., Longhini D.: “Removal of Stickies and electrostatic inks using flotation process”, Tappi Recycling Symposium, conference book, 1996 109. Ling T.: “The effects of surface properties on stickies removal by flotation”, Pulp & Paper Canada, 95:12 (1994) 110. Doshi M., Dyer J.: “Removal of wax and stickies from OCC by froth flotation”, Tappi Recycling Symposium, Washington, March 2000 111. Gabl H., Waupotitsch M., Hertl E.: “Increasing the yield of DIP-Production by means of the th CLEANFLOAT system”, 11 PTS CTP Deinking Symposium, Munich: PTS 2004 112. Donatas Satas: “Handbook of Pressure Sensitive Adhesive Technology”, Warwick, Rhode Island
181
L SCREEN C CLEAN EAN
Final Technical Report – CTP - AFT - ADJ - ICP - PTS - LEGI – October 2005
ANNEXE List and copies of the publications resulting from the project 1.
Gonera, H., Dabrowski, J., Mik, T.: “Pulping as a key step in the efficient removal of stickies th
during deinking process”, 15 International Papermaking Conference Progress’ 05, Wroclaw, September 28-30, 2005 2. Asendrych D., Favre-Marinet Favre-Marinet M., Julien Saint Amand F. : "Decoupled approach to the modelling of contaminants removal from recycled paper", Workshop on Multiphase Flows Simulation, Experiment & Application, Dresden, May-June, 2005. 3.
Julien Saint Amand, F., Perrin, B., Frach, D., Asendrych, D., Gooding, R., Huovinen, A. : “Stickies screening: Study of stickies extrusion through slots and optimisation of screen plate th design”, 6 KRICT-KTAPPI RTM 2005, Daejon, Korea, 18-19 April 2005. th
4. Hamann, L.: “The role of Deinking-Flotation and Micro-Flotation Micro-Flotation for stickies reduction”, reduction”, 7 PTS Stickies Seminar, Paper n°11, Dresden, April 5-6, 2005. 5.
Julien Saint Amand, F., Perrin, B., Frach, D., Asendrych, D. : “Visualisation of stickies th
extrusion through slots in pressure screens”, 7 PTS Stickies Seminar, Paper n°5, Dresden, April 5-6, 2005. 2005. 6. Schmid, W.-H.: “Combatting and avoiding stickies in waste paper processing”, IPW International Paperworld, paper n° 5, 2005. 7. Asendrych D., Favre-Mariner M., Julien Saint Amand F. : "Modelling of the separation process th of the adhesive materials from the paper pulp" (in Polish), Proceedings of the IV Conference "Symulacja 2004", Rydzyna, 18-20 October 2004. 8. Asendrych D., Favre-Mariner M., Julien Saint Amand F. : "Modelling of the flow through the th pressure screen" (in Polish), Proceedings of the XVI Polish Biennial Conference on Fluid Mechanics, Warsaw-Waplewo, 20-23 September 2004. 9.
Julien Saint Amand, F., Perrin, B., Gooding, R., Huovinen, A. : “Optimisation of screen plate design for the removal of stickies from deinking pulps”, Revue ATIP, Vol. 58, n°4, Août/Septembre Août/Septembre 2004.
10. Julien Saint Amand, F., Perrin, B. : “Possibilities and limits of cleaners in removing PSA stickies”, KRICT-KTAPPI RTM 2004, Daejon, Korea, 21 & 22 June 2004.
11. Julien Saint Amand, F., Perrin, B., Gooding, R., Huovinen, A., Asendrych, D., Favre-Marinet, Favre-Marinet, M. : th
“Optimisation of screen plate design for the removal of stickies in deinking pulps”, 11 PTSCTP Deinking Symposium, paper n°3, Leipzig, 27-30 April 2004 12. Julien Saint Amand, F., Wojciechowski, G., Asendrych, D., Favre-Marinet, M., Rahouadj, R., Skali-Lami, S.: “Screening: Fundamental studies on the extrusion of stickies through slots”, Revue ATIP, Vol. 58, n°1: 6-18, Février/Mars 2004.
182