Analysis of Sand Transportability in Pipelines

ANALYSIS OF SAND TRANSPORTABILITY IN PIPELINES

STUDY REPORT

07/2010 07/201 0

Author

Laras Wuri Dianningrum

Check and Verify Patria Indrayana Indrayana FO/AMB/MTH

LEMBAR PENGESAHAN

Menerangkan bahwa :

Laras Wuri Dianningrum 13007075 Teknik Kimia Fakultas Teknologi Industri Institut Teknologi Bandung

Telah menyelesaikan,

Program On the Job Training Di Departemen FO/AMB/MTH TOTAL E&P INDONESIE I NDONESIE East Kalimantan District, Balikpapan

Telah disetujui dan disahkan Di Balikpapan, tanggal 30 Juli 2010

Pembimbing

Patria Indrayana

Head of HRD Department

Bayu Parmadi

ii

TABLE OF CONTENTS

LEMBAR PENGESAHAN ........................................................................................................................... ii TABLE OF CONTENTS.............................................................................................................................. iii LIST OF TABLES........................................................................................................................................ v LIST OF FIGURES .................................................................................................................................... vii CHAPTER I INTRODUCTION ..................................................................................................................... 1 1.1

Background of Study ................................................... .......................... ................................................... .................................................... ..................................... ........... 1

1.2

Objectives ................................................. ....................... .................................................... ................................................... .................................................... ........................... .. 1

1.3

Methodology ................................................. ....................... .................................................... ................................................... .................................................. ......................... 2

1.4

References .................................................... .......................... .................................................... ................................................... .................................................. ......................... 2

CHAPTER II BEKAPAI OVERVIEW ............................................................................................................. 5 CHAPTER III LITERATURE STUDY ............................................................................................................. 8 3.1

Multiphase Flow in Pipeline .................................................... .......................... ................................................... .................................................. ......................... 8

3.1.1 Multiphase Flow Properties ................................................. ....................... ................................................... .................................................. ......................... 8 3.1.2 Flow Regimes Determination in Multiphase Flow Flow (Gas and Liquid Liquid System) .......................... 10 3.1.3 Experimental Correlation in Horizontal Horizontal Pipe .................................................. ........................ ................................................ ...................... 12 3.1.4 Empirical Correlation in Vertical Pipe ........................... ................................................... ......................... ............................... ..... 13 3.1.5 Beggs and Brill Correlation...................................... .................................................... .......................... ................................... ......... 17 3.2

Sand Transportability in Pipe .................................................. ........................ ................................................... ................................................ ....................... 21

3.3

Critical Flow Velocity in Sand Transport ................................................... .......................... ................................................... ............................... ..... 26

3.3.1 Horizontal Pipe .................................................... .......................... .................................................... ................................................... ....................................... .............. 26 3.3.2 Vertical Pipe .......................................................................................................................... 27

CHAPTER IV BEKAPAI OBSERVATION .................................................................................................... 29 4.1

Bekapai Production Network Configuration and Gas Lift.......................... ............................... .......................... ..... 29

4 .2

Well Head Data and Maximal Deliverable Potential in Bekapai ............................................... ......................... ...................... 30

4.3

Deposit Particle Analysis .................................................... .......................... .................................................... ................................................... ........................... 30

CHAPTER V BASIC CALCULATION FOR FLOW REGIME PREDICTION (COMPARISON OF METHOD) ......... ......... 33 5.1

Empirical Correlation(Mandhane, Aziz et al . versus Beggs & Brill) .................... ...................... 33

5.2

OLGA versus Beggs & Brill .................................................. ........................ .................................................... ................................................... ........................... 33

CHAPTER VI RESULTS AND DISCUSSION ................................................................................................ 38

iii

6.1

Analysis of Sand Behavior in Correlation with Flow Regime ................................................. ....................... ............................... ..... 40

6.1.1 Experimental Correlation Correlation (Mandhane, Aziz et al . versus Beggs Begg s & Brill) .................................. ......................... ......... 41 6.1.1.1 Horizontal Pipe ......................................................................................................... 41 6.1.1.2 Vertical Pipe/Upflow Risers .................................................... .......................... .................................................... ................................... ......... 46 6.1.2 OLGA versus Beggs & Brill....................... .................................................... .......................... ................................................... ........................... 49 6.1.2.1 Oil-Gas Flow ............................................................................................................. 51 6.1.2.1.1 8” BK-BP1................................................................................................. 51 6.1.2.1.2 12” BB-BP1............................................................................................... 54 6.1.2.1.3 6” BF-BL ................................................................................................... 57 6.1.2.1.4 6” BH-BG .................................................................................................. 60 6.1.2.1.5 12” BL-BA ................................................................................................. 63 6.1.2.2 Water-Gas Flow ......................................................................................................... 65 6.1.2.2.1 12” BL-BA.................................................................................................. 66 6.1.2.2.2 6” BH-BG................................................................................................... 68 6.1.2.2.3 6 ” BF-BL .................................................................................................... 71 6.1.2.2.4 6” BJ-BB .................................................................................................... 73 6.1.2.2.5 8 ” BK-BP1 ................................................................................................. 76 6.1.2.2.6 12” BB-BP1 ............................................................................................... 78 6.1.3 Main Finding ......................................................................................................................... 80 6.1.3.1 Experimental Correlation (Mandhane, Aziz et al . versus Beggs Begg s & Brill) ........................ 80 6.1.3.2 OLGA versus Beggs & Brill .................................................... .......................... ................................................... ....................................... .............. 80 6.2

Analysis of Sand Settling Condition ................................................. ....................... ................................................... ....................................... .............. 82

6.2.1 Horizontal Pipe .................................................... .......................... .................................................... ................................................... ....................................... .............. 83 6.2.2 Vertical Pipe .................................................... .......................... ................................................... ................................................... ............................................ .................. 88 6.2.3 Main Finding ................................................... ......................... ................................................... ................................................... ............................................ .................. 89

CHAPTER VII CONCLUSIONS AND RECOMMENDATIONS ....................................................................... 91 7.1

Conclusions .................................................. ........................ .................................................... ................................................... ................................................ ....................... 91

7.2

Recommendations ........................... ................................................... ......................... .................................................... ................................... ......... 91

iv

LIST OF TABLES

TABLE 2.1 Pipelines and wellheads in Bekapai area................................................. ....................... .................................................... ................................. ....... 6 TABLE 3.1 Multiphase flow correlations ................................................................................................. 17 TABLE 4.1 Deposit particles from bekapai area ...................................................................................... 31 TABLE 5.1 Average pressures and temperatures in Bekapai pipelines .................................................... 33 TABLE 5.2 Pipeline geometry data ......................................................................................................... 35 TABLE 5.3 Oil composition in OLGA ........................................................................................................ 35 TABLE 5.4 Gas composition in OLGA ...................................................................................................... 36 TABLE 5.5 Flow Properties in each Bekapai pipeline ........................... .................................................... .......................... .......................... 36 TABLE 6.1 Flow regime s of Bekapai pipelines from Mandhane’s map..................................................... 43 TABLE 6.2 Horizontal flow regimes in Bekapai pipelines by Beggs & Brill correlation (revised) ............... 43 TABLE 6.3 Horizontal flow regimes in Bekapai pipelines by Beggs & Brill correlation (1973) ................... 45 TABLE 6.4 Flow regimes of vertical Bekapai pipelines based on Azi z and Beggs & Brill co rrelation.......... 48 TABLE 6.5 GWR, GOR, and water cut values of Bekapai pipelines .................................................. ........................ ................................... ......... 48 TABLE 6.6 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in oil-

gas flow (8” BK-BP1) ............................................................................................................. 53 TABLE 6.7 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in oilgas flow (12” BB-BP1) ........................................................................................................... 56 TABLE 6.8 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in oilgas flow (6” BF-BL) ............................................................................................................... 59 TABLE 6.9 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in oilgas flow (6” BH-BG) .............................................................................................................. 62 TABLE 6.10 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in oilgas flow (12” BL-BA) ............................................................................................................. 65 TABLE 6.11 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in water-gas flow (12” BL-BA) ................................................................................................... 68 TABLE 6.12 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in water-gas flow (6” BH-BG) .................................................................................................... 70 TABLE 6.13 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in water-gas flow (6” B F-BL) ..................................................................................................... 73 TABLE 6.14 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in water-gas flow (6” B J-BB) ..................................................................................................... 75

v

TABLE 6.15 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in water-gas flow (8” BK-BP1) ................................................................................................. 78 TABLE 6.16 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in water-gas flow (12” BB-BP1) ............................................................................................... 80 TABLE 6.17 Salama versus Bekapai case ................................................................................................. 82 TABLE 6.18 Flow cri tical velocity in several Bekapai pipelines using Salama equation........... .................. 84 TABLE 6.19 Actual mixture velocity in vertical Bekapai pipeline for each particle ................. .................. 88

vi

LIST OF FIGURES

FIGURE 2.1 Bekapai pipeline pipeline system system ........................... .................................................... .......................... .................................................. ........................ 6 FIGURE 3.1 Experimental correlation catagories .................................................................................... 11 FIGURE 3.2 Mandhane’s map ................................................................................................................ 12 FIGURE 3.3 Regime Reg ime characteristics in horizontal pipe ................................................... ......................... ................................................... ........................... 13 FIGURE 3.4 Multiphase flow regime in vertical vertical pipe .................................................... .......................... ................................................... ........................... 14 FIGURE 3.5 Duns and Ros flow regime map ........................................................................................... 15 FIGURE 3.6 Aziz et al . map ..................................................................................................................... 16 FIGURE 3.7 Flow pattern in slurry flow................................................................................................... 22 FIGURE 3.8 Multiphase flow regime consist of liquid, gas and solid........................................................ 22 FIGURE 3.9 Schematic sand behaviors in slug with low gas superficial velocity................. velocity............................. ....................... ........... 23 FIGURE 3.10 Sand behaviors in smooth stratified regime..................... regime................................. ....................... .................... ..................... .................... ........ 24 FIGURE 3.11 Sand dune formation behaviors ......................................................................................... 24 FIGURE 3.12 Sand behaviors in stratified-wavy regime .......................................................................... 25 FIGURE 3.13 Sand behaviors in plug regime ........................................................................................... 25 FIGURE 3.14 Sand behaviors in slug regime ........................................................................................... 26 FIGURE 3.15 FL value vs. particle diameter, concentration as parameter ................................................ 26 FIGURE 4.1 Particles sieve analysis......................................................................................................... 30 FIGURE 5.1 OLGA model view for gas-water case ................................................................................... 37 FIGURE 6.1 Factors affeted sand transportation in pipeline ................................................................... 38 FIGURE 6.2 Flow F low regime determination used in this analysis ........................... ....................................... ......................... .............. 40 FIGURE 6.3 Mandhane’s map of Bekapai pipelines ................................................................................ 42 FIGURE 6.4 Beggs & Brill map (1973) of Bekapai pipelines ..................................................................... 44 FIGURE 6.5 Aziz et al. map of Bekapai pipelines ..................................................................................... 47 th

FIGURE 6.6 Flow regime, holdup, and fluid velocity at 50 section in 8”BK-BP1 (oil-gas flow) ................ 51 FIGURE 6.7 Flow regime, holdup, and fluid velocity at riser bottom in 8”BK-BP1 (oil-gas ( oil-gas flow) ............... 51 FIGURE 6.8 Flow regime, holdup, and fluid velocity at pipe outlet in 8”BK-BP1 (oil-gas flow) ................. 52 FIGURE 6.9 Flow regime, holdup, and fluid velocity at 50 th section in 12”BB-BP1 (oil-gas flow) .............. 54 FIGURE 6.10 Flow regime, holdup, and fluid velocity at riser bottom in 12”BB-BP1 (oil-gas flow) ........... 54 FIGURE 6.11 Flow regime, holdup, and fluid velocity at pipe outlet in 12”BB-BP1 (oil-gas flow) ............ . 55 th

FIGURE 6.12 Flow regime, holdup, and fluid velocity at 50 section in 6”BF-BL (oil-gas flow) ................. 57

vii

FIGURE 6.13 Flow regime, holdup, and fluid velocity at riser bottom in 6”BF-BL (oil-gas flow) ............... 57 FIGURE 6.14 Flow regime, holdup, and fluid velocity at pipe outlet in 6” BF-BL (oil-gas flow) ................. 58 th

FIGURE 6.15 Flow regime, holdup, and fluid velocity at 50 section in 6”BH-BG (oil-gas flow) ............... ...... ......... 60 FIGURE 6.16 Flow regime, holdup, and fluid velocity at riser bottom in 6”BH-BG (oil-gas flow) .............. 60 FIGURE 6.17 Flow regime, holdup, and fluid velocity at pipe outlet in 6”BH-BG (oil-gas flow) .. .............. 61 th

FIGURE 6.18 Flow regime, holdup, and fluid velocity at 50 section in 12”BL-BA (oil-gas flow) .............. ......... ..... 63 FIGURE 6.19 Flow regime, holdup, and fluid velocity at riser bottom in 12”BL-BA (oil-gas (oil-g as flow) ............. 63 FIGURE 6.20 Flow regime, holdup, and fluid velocity at pipe outlet in 12”BL-BA (oil-gas flow) .......... ..... 64 th

FIGURE 6.21 Flow regime, holdup, and fluid velocity at 50 section in 12”BL-BA (water-gas flow) ......... 66 FIGURE 6.22 Flow regime, holdup, and fluid velocity at riser bottom in 12”BL-BA (water-gas flow) ........ 66 FIGURE 6.23 Flow regime, holdup, and fluid velocity at pipe outlet in 12”BL-BA (water-gas flow) .......... 67 th

FIGURE 6.24 Flow regime, holdup, and fluid velocity at 50 section in 6”BH-BG (water-gas flow) .......... 68 FIGURE 6.25 Flow regime, holdup, and fluid velocity at riser bottom in 6”BH-BG (water-gas flow) ......... 69 FIGURE 6.26 Flow regime, holdup, and fluid velocity at pipe outlet in 6”BH-BG (water-gas flow) ........... 69 th

FIGURE 6.27 Flow regime, holdup, and fluid velocity at 50 section in6”BF-BL (water-gas flow) ............ 71 FIGURE 6.28 Flow regime, holdup, and fluid velocity at riser bottom in 6 ”BF-BL (water-gas flow) .......... 71 FIGURE 6.29 Flow regime, holdup, and fluid velocity at pipe outlet in 6”BF-BL (water-gas flow)............. flow)........ ..... 72 th

FIGURE 6.30 Flow regime, holdup, and fluid velocity at 50 section in 6”BJ-BB (water-gas flow)............ flow)....... ..... 73 FIGURE 6.31 Flow regime, holdup, and fluid velocity at riser bottom in 6”BJ-BB (water-gas flow) .......... 74 FIGURE 6.32 Flow regime, holdup, and fluid velocity at pipe outlet in 6”BJ-BB (water-gas flow)............. 74 th

FIGURE 6.33 Flow regime, holdup, and fluid velocity at 50 section in 8“BK-BP1 (water-gas flow) ......... 76 FIGURE 6.34 Flow regime, holdup, and fluid velocity at riser bottom in 8”BK-BP1 (water-gas flow)........ flow)....... . 77 FIGURE 6.35 Flow regime, holdup, and fluid velocity at pipe outlet in 8”BK-BP1 (water-gas flow) .......... 77 th

FIGURE 6.36 Flow regime, holdup, and fluid velocity at 50 section in 12”BB-BP1 (water-gas flow) ....... 78 FIGURE 6.37 Flow regime, holdup, and fluid velocity at riser bottom in 12”BB-BP1 (water-gas flow)...... flow)..... . 79 FIGURE 6.38 Flow regime, holdup, and fluid velocity at pipe outlet in 12”BB-BP1 (water-gas flow) ........ 79 FIGURE 6.39 Critical velocity profiles in 6” BJ -BB, BF-BL, and BH-BG .................................................. ........................ ............................... ..... 88 FIGURE 6.40 Range of critical velocity in several Bekapai pipelines based on particle diameter ............. 88

viii

ix

Analysis of Sand Transportability in Pipelines

CHAPTER I INTRODUCTION 1.1

Background of Study In recent years, sand behavior along oil and gas pipelines is one of the major problems as a consequence of sand production. Once sand is detached, it follows the fluid stream through the perforations and into the well. Phenomena such as sand deposition can lead to partial or complete blockage of flowlines, enhanced enhanced pipe bottom corrosion, corrosion, and trapping trapping of pigs. These failures failures can cause unexpected downtime and risk to equipment as well as personnel.

Bekapai production network includes several pipelines located under the sea to connect each platform with Bekapai production platform (BP1). Sand particles are investigated due to corrosion enhanced caused by bacteria in two Bekapai pipeline’s surface. Indirectly, they have supported the existence of bacteria by creating a layer that protects bacteria from corrosion inhibitor released. This layer is called sand bed that comes from the sand settling along pipe. When multiphase flow in pipe reaches below its critical value, solid particles carried by flow begin to settle and form sand bed in the bottom.

Therefore, sand control management which consists of an accurate study of the parameters such as flow rates of gas and oil, flow patterns, pressure drop, geometry and inclination design of pipelines, etc. is required in order to develop better understanding of the problem (e.g. sand behavior with fluid flow inside the pipeline). It must be done to overcome the lack of information available about sand behavior in flow, especially the relationship between flow regime and sand settling condition. However, these things are closely related in determining determining sand transportation, in order to prevent the early sand accumulation before it has an impact on the pipeline’s pi peline’s performance and overall systems.

1.2

Objectives This present study is going to investigate the sand behavior in Bekapai pipelines by finding the flow critical velocity to keep sand particles moving along the pipe and its relationship with flow regimes as multiphase flow. The other parameters influenced the phenomena such as holdup, liquid and gas velocities, inclination and sand properties (diameter and density) are al so observed in general.

1


1.3

Methodology This study was performed in the frame of 2 months on the job training traini ng using following foll owing methods:

1.4



Literature studies



OLGA training



Cases studies

References This study was performed using following references and information:

A. From internal of TOTAL E&P INDONESIE 1. Bekapai IP Inspection I nspection Summary Report 2007 2. ST-SNP-08-002 RVP Simulation During Senipah-Peciko Local Control Network Modification 3. Bekapai Wellhead Platform Operating Manual 4. Peciko Pigging Instruction Summary (revision) 5. PRODEM Section No. V, “Fluid Flow in Pipes” th

6. Bekapai Potential (Status: June 24 , 2010) th

7. Bekapai Production Network Configuration and Gas Lift ( Status: August 25 ,2006) th

8. Bekapai Production Test Summary (Status: June 24 , 2010) 9. Deposit BG-3 LS 241105 (A) 10. Deposit of ex pigging BKP to SNP_051006 (B) 11. Sieve Analysis BL 14 12. Sieve Analysis BL-6_03 May 2009 (C) 13. Sieve analysis_BK 2 S 18052009 14. Sieve analysis_BL-10LS_29.05.09 15. Sieve analysis_V-100 & 120 (LP Separator) 16. DKE/PRO Method Section, “Introduction to Multiphase Flow” by Bamban g Yudhistira and Zaki Hatmanda 17. Oil and Gas Processing Plant Design and Operation Training Course, DGEP/SCR/ED/ECP, March nd

nd

22 – April 2 , 2004

2


B. From external source (books, journals, articles, etc.) 1. Aggour, M. A.; Al-Yousef, H. Y.; Al-Muraikhi, A. J., “Vertical Multiphase Flow Correlations for High Production Rates and Large Tubulars ”, SPE Production & Facilities, 1996. 2. Anselmi, Ruth; Baumeister, Alberto J.; Marquez, Katiuska C., “Review of Methods and Correlations for the Analysis of Transport Lines with Multiphase Flow ”, XVIII Gas Convention, th

AVPG, Caracas, Venezuela, May 27 – 29 , 2008. 3. Beggs, H. Dale; Brill, James P., “ A Study of Two Phase Flow in Inclined Pipes”, Journal of Petroleum Technology, 1973, pp. 607-617. 4. Boriyantoro, Niels H.; Adewumi, Michael A.,” An Integrated Single-Phase/Two-Phase Hydrodynamic Model for Predicting The fluid Flow B ehavior of Gas Condensate in Pipelines ” 5. Bremer, Jeff, "Pipeline " Pipeline Flow of Settling Slurries", Slurries ", Sinclair Knight Merz, 2008. 6. Brennen, C.E. 2005 . Fundamentals of Multiphase Flow . UK: Cambridge U niversity Press. 7. Campbell, John M. 2004. Gas Conditioning and Processing Vol.1. Vol.1. Oklahoma, USA: John M. Campbell Company. 8. Chang, Yvonne S.H.; Ganesan, T; Lau, K. K.,” Comparison Comparison between Empirical Correlation and Computational Fluid Dynamics Simulation for the Pressure Gradient of Multiphase Flow ” , World Congress on Engineering 2008 Vol.1, 2008. 9. Chen, R. C., “ Analysis of Homogeneous Slurry Pipe Flow”, Journal of Marine Science and Technology Vol.2 No. 1, pp. 37-45. 10. Chien, Sze-Foo, “Settling Velocity of Irregularly Shaped Particles ”, SPE Drilling and Completion, 1994. 11. Danielson, Thomas J., ”Sand Transport Modeling in Multiphase Pipelines ”, OTC 18691, 2007. 12. Doan, Q. T.; Doan, L. T.; Ali, S. M. Farouq; Oguztoreli, M.,”Sand Deposition Inside a Horizontal Well – A Simulation Approach”, Journal of Canadian Petroleum Techology, Vol. 30, No. 10, 2000. 13. Escobedo, Joel; Mansoori, G. Ali., “Surface Tension Prediction of Liquid Mixture ”, AlChE Journal, Vol. 44, No. 10 1998, pp.2324-2332. 14. Gas Processors Suppliers Association. 2004. Engineering Data Book 12th Edition. Tulsa: Gas Processors Suppliers Association. 15. Gorji, M.; Rostamian, M., “ Analyzing the Influences of Different Parameters on Terminal Deposit in Hydrate Slurry ”, ”, International Journal of Dynamics of Fluids Vol. 2 No.1 2006, pp. 99-109. 16. Hameed, Abdul, “Pipeline Pulsing Flow of Slurries”, Open Dissertation and Theses, 1983.

3


17. Jimenez, Jose A.; Madse, Ole S.," S.," A Simple Formula to Estimate Settling Velocity of Natural Sediments", Sediments", ASCE 0733-950X, 2003, 129:2 (70). 18. Kovacs, Laszlo; Varadi, Standor, "Two " Two Phase Flow in the Vertical Pipeline of Air Lift ", ", Periodica Polytechnica ser. Mech. Eng. Vol. 43, no. no. 1, 1999, pp. 3 –18. 19. Lahiri, S.K.;Glasser, Benjamin J., "Minimize " Minimize Power Consumption in Slurry Transport ”, ”, Hydrocarbon Processing, 2008. 20. Lee, M. S.; Matousek, V.; Chung, C. K.; Lee, Y. N., ”Pipe Size Effect on Hydraulic Transport of Jumoonjin Sand-Experiments in a Dredging Test Loop ”, Terra et Aqua No.99, 2005. 21. Liss, Elizabeth, D.; Conway, Stephen L.; Zega, James A.; Glasser, Benjamin J., " Segregation of Powders during Gravity Flow Through Vertical Pipes ", Pharmaceutical Technology, 2004. 22. Maurer Engineering Inc., "Multiphase Flow Production Model, Theory and User’s Manual ", DEA 67, Phase 1, 1994. 23. McLaury, B. S.; Shirazi, S. A., “Generalization of API RP 14E for Erosive Service in Multiphase Production”, SPE 56812, 1999. 24. Rao, Bharath, “Multiphase Flow Models Range of Applicability” , CTES, L.C. Tech Note, 1998. 25. Ruano, Angel Perez, “Sand Transportation in Horizontal and Near Horizontal Multiphase Pipelines”,M.Sc. Thesis, Carnfield University, 2008. 26. Salama, Mamdouh M., “Sand Production Management ”, ”, OTC Proceedings, 1998. 27. Salama, Mamdouh M., “Influence of Sand Production on Design and Operating of Piping Systems”, Corrosion 2000 Paper No. 80, 2000. 28. Sutton, Robert P., “ An Improved Model for Water-Hydrocarbon Surface Tension at Reservoir Conditions”, SPE 124968, 2009. nd

29. Taitel, Yehuda, “Flow Pattern Transition in Two Phase Flow” , 2

Annual Meeting of the

Institute of Multifluid Science and Technology, 1999. 30. Tronvoll, J.; Dusseault, M.B.; Santarelli, F. J., "The " The Tools of Sand Management ", ", Society of Petroleum Engineers Inc., 2001. 31. Yuan, Hong; Zhou, Desheng, “Evaluation of Two Phase Flow Correlation and Mechanist ic Models for Pipelines at Horizontal and Inclined Upward Upward Flow” , SPE 120281, 2009.

32. http://www.unisanet.unisa.edu.au/Resources/10809/Mine%20Ventilation%20and%20Fluid% 20Flow%20Applications/Fluid%20Applications/Slurry%20Flow.pdf 33. http://www.csupomona.edu/~tknguyen/che435/Notes/P4-fluidized.pdf 34. http://sti.srs.gov/fulltext/tr2000263/tr2000263.html

4


CHAPTER II BEKAPAI OVERVIEW The Bekapai Field lies offshore about 15 km from the mouth of the Mahakam River delta in East Kalimantan, Indonesia. In partnership with Pertamina and Inpex, Total Indonesie has operated the field since production began in 1974. The field itself is in relatively calm water of 35 m depth and extends over an area approximately 3 x 6 km. In 2004, this field produced 2,600 BOPD oil, 10 MMSCFD associated gas, and 8,250 BWPD water. In the recent update (June 2010), Bekapai still has potential to deliver 6,361 STBD of oil, and 18.8 MMSCFD gas .

Figure 2.1 Bekapai pipeline system

There are several manifold well head platforms in this field: BA, BB, BE, BF, BG, BH, BJ, BK, and BL. The Central Complex consists of the set of: a well-head platform, named BA, jacket with 9 slots, a production platform, named BP, a living quarter platform named BQ, and a remote flare on a tripod, with an additional tripod intermediate platform. The well heads are in low pressure (LP) condition. They consist of some wells that three of them are g as lift sources (BJ-4-SS, BF-1-SS, and BL-10-LS) and gas lift wells (BJ-3-LS, BA-9-LS, BL-7-LS, BG-1-LS, and BF-2-SS).

5


The five different platforms in the central complex are interconnected by bridges. The general arrangement is in a East-West direction so that the prevailing wind is perpendicular and provides the best natural ventilation and so that the risks of gas cloud propagation and liquid spillage at sea are minimised, with the living quarters platform LQ upwind the platforms handling hydrocarbons. The central complex is permanently manned, with a maximum POB of 72. It is fitted with three boat landings, on the South sides of BA and BP, and on the East side of LQ, and a helideck (without any stand-by or refuelling facility) on LQ. BA is served by the control and safety systems, and the utilities of the central complex.

Table 2.1 Pipelines and wellheads in Bekapai area

Pipeline

Connected Well

Diameter (inches)

BK-BP1

BK-2-SS

8

BJ-BB

BJ-4-SS

6

BB-BP1

BB-6-LS, BB-9-LS, BJ-4-SS

12

BE-BA

BE-3-SS

6

BF-BL

BF-1-SS, BF-2-LS

6

BH-BG

BH-1-SS, BH-1-LS, BH-3-S

6

BG-BL

BG-6-S, BG-10-S

6

BG-BL

BG-6-S, BG-10-S (oil-water only)

12

BL-BA

BG-6-S, BG-10-S (gas only)

12

BL-BA

BG-6-S, BG-10-S, BL-1-S, BL-6-S,

6

BL-9-S, BL-14-S (oil-water only)

Bekapai production platform (BP1) collected the oil and gas from satellites. In this platform, water is separated and then disposed to sea. Gas and oil mixture are separated, they go then to compression

and pumping and mixed thoroughly before sent to Senipah by 12” multiphase sea line.

Detail of the process consists of three main steps: separation, oil pumping, and gas compression. Incoming LP well effluent from MWP is received by two separators (V 100 and V 120). V 100 acts as flow dampener only. Since the gas outlet is being closed, oil and gas leaves this vessel through oil outlet line. Then the second separator (V 120) will make a further separation to split the oil, gas, and water stream. Gas released from this vessel is compressed into HP level by turbine driven two-stage

6


centrifugal compressor (K 3020 and K 3050). Besides, oil is also pumped by series of booster (MP 210-220-230) and transfer pumps (MP 240, 250, 260) before mixed with compressed gas and delivered to Senipah terminal.

Produced water obtained from V-120 is treated in Oily Water Treatment Unit before being discharged to the sea. Bekapai OWTU i s equipped with two skimmer tanks operating i n series (T 3800 and T 3810). A cyclone (F 3850) is used to enhance oil removal of skimmer tank (T 3800) water discharge and can be used for direct cleaning of separator (V 120) water effluent. Final oil removal takes place in a floatator, named Wemco depurator (V 3870) which can reduce oil content to less than 50 ppm and the water is finally disposed to the sea.

7


CHAPTER III LITERATURE STUDY

3.1 Multiphase Flow in Pipeline The most commonly employed method of transporting fluid from one point to another is to force the fluid to flow through a piping system. Pipe of circular section is most frequently used because that shape offers not only greater structural strength, but also greater cross sectional area per unit of wall surface than other shape. Pipe always refer to a closed conduit of circular section and constant internal diameter.

The same thing occurred in oil and gas transportation. The flow is classified as multiphase flow which generally located in the part of the installations between the reservoir and the process units. Multiphase flow are first found in wells, whether production be carried out through the tubing or through the annulus. There is also multiphase flow in the flow lines transferring the production from the wellheads to the primary separator or the test separator. Multiphase flow may also occur in plant piping downstream of control valves or through heat exchanger tubes where condensation or vaporization is achieved (Prodem V).

Multiphase flow is defined as flow in which several phases are present. The phases which can be in presence in multiphase flow are: gas, oil or condensate, free water, methanol, glycols, additives such as corrosion inhibitors dissolved in water, soli ds (sand, clay).

3.1.1 Multiphase Flow Properties 

Liquid mixture density (Campbell, 2004) For determining liquid mixture density, the below equation is used.

Where

   

    =

= mol fraction of each component

= molar volume of each component = molar volume of the mixture

8




Liquid mixture viscosity (Campbell, 2004) For determining liquid mixture viscosity, the below equation is used.

Where



   

     1/3

=

3

( )

= mol fraction of each component = component viscosity

= viscosity of mixture in centipoise

Liquid mixture surface tension (Sutton, 2007) For determining liquid mixture surface tension, the below equation i s used.

  −   1.58

=

Where



   

= water density

+ 1.76 1.76

4

0.3125

= oil density = liquid mixture surface tension = reduced temperature

Gas density Compressibility factor (Z) for determine the non ideal gas is gained via S. Robertson method:

       −        −  −− =

/

= 0.12 0.1219 19

=

2

0.638

7.76 .76 +

14.75

= 0.3 + 0.44 0.441 1

Where

 

=

= reduced pressure

+

2

(

)

= reduced temperature

Then the actual density of gas can be found from the following equation:

Where

 

      = ( )(

= compressibility factor = universal gas constant = absolute pressure = absolute temperature

)/(

)

9




= relative molecular weight

3.1.2 Flow Regimes Determination in Multiphase Flow (Gas and Liquid Liquid System) The determination of the expected flow regime allows the proper selection of correlations or mechanistic model for calculating the pressure gradient and liquid hold-up. In addition, for operating purpose it is important to know which type of flow regime is predicted at various locations of the pipeline and obviously at the outlet. Phenomena such as erosion, corrosion and vibration depend on the flow regime.

This object has been studied in wide range of fields and applied in many sectors especially in oil and gas production. This is not an easy task, however, many researchers must find the exact correlation to relate among not less than 11 parameters that affect flow regimes:

   

[m/s] (it is customary to use the superficial velocity instead

a) The liquid superficial velocity, the flow rate). b) The gas superficial velocity,

             

c) Liquid density, d) Gas density,

[

Gas viscosity,

g) Pipe diameter,

].

/

[

].

/

e) Liquid viscosity, f)

[ / ].

. ].

[

[

. ].

[ ].

h) Acceleration of gravity,

/

].

[N/m].

i)

Surface tension,

j)

Pipe roughness, e [m].

k) Pipe inclination,

[

(Taitel, 1999) .

Theoretically, the method used for the prediction of flow pattern can be classified with respect to two categories: 

Experimental correlations

The first approach for the prediction of flow patterns is based on experimental data that are plotted on a flow pattern map. The earliest flow regime map is attributed to Baker (1954). Many more have since been suggested for horizontal, vertical and inclined pi pes. Then they are divided into three main catagories based on the basic assumptions and methods (Figure 3.8).

10


Experimental correlation

Catagory A

Catagory B

(No slippage and no flow pattern consideration)

(Slippage considered, no flow pattern consideration)

(Slippage and flow pattern consideration)

Pettmann&Carpenter,B axendel&Thomas,Fanch er&Brown

Hagedorn&Brown,Gray, Asheim

Dun&Ros,Orkiszewski,A ziz,etc

Catagory C

Figure 3.1 Experimental correlation catagories



Mechanistic model

In this procedure one should identify the dominant physical phenomena that cause a specific transition. Then the physical phenomena are formulated mathematically and transition lines are calculated and can be presented as an algebraic relation or with respect to dimensionless coordinates. It still needs correlation and closure law for input some parameters to solve the momentum balance equation. However, there is no guarantee that this method leads always to correct results, but the results based on this method then extrapolation to different conditions is much safer than those based solely on experimental correlation (Taitel, 1999).

The mechanistic model developments are divided into three categories: a.

st

Comprehensive Models (1 generation)

This model priors a separate prediction of flow pattern and pressure gradient prediction, for example: Taitel & Dukler Flow pattern and Xiao et a.l (Taitel & Dukler modification). nd

b. Unified Models (2 generation) Different from the previous one, this model is considered to consist only one prediction for determining flow pattern & pressure gradient. For example: TUFFP unified model (Zhang et al.). c.

Integrated Unified Model of Heat Transfer and Fluid Flow

11


This is somewhat called “future generation” of multiphase flow modeling and until this day the experiments and current studies are still performed.

So far those methods that had been explained are limited to the steady state flow condition. The problem occurred when they need to be applied in real situation on field which is preferably transient one. The mechanistic models for this case are developed by many universities and companies like SINTEF, IFE, IFP, University of Tulsa, etc. Software like OLGA and TACITE are widely known among the practices to solve determination of f low regime in transient flow.

3.1.3 Experimental Correlation Correlation in Horizontal Pipe Pipe The Taitel & Dukler (1976) flow model seems the most accurate one, even if its accuracy is decreasing for large pipeline diameters. The Taitel & Dukler approach is based on a combination of theoretical considerations of classical fluid mechanics. But it is more difficult to solve in manual calculation, so that this model required. required. Other map commonly used was developed by Gregory, Aziz, and Mandhane for horizontal flow. It has accuracy about 70% approximately and has considered the liquid hold up and pressure drop determination.

Figure 3.2 Mandhane’s map

12


The characteristic of each regime explained ex plained as follows:

Dispersed Flow

Segregated Flow

Intermittent Flow

Stratified

Bubble

(high gas-liquid ratio, medium gas flow rate, the fraction of each section is remain constant)

(Small gas -liquid ratio, continuous phase: liquid, very low slip velocity)

Annular

Mist/Spray (Very high gas flow rate, very high gas-liquid ratio, continuous continuous phase: g as)

(very high gas-liquid ratio, high gas flow rate, annular film on the wall is thickened at the bottom of pipe)

Slug (medium gas-liquid ratio, high liquid flow rate)

Plug (more transition regime between stratified wavy and slug flow/annular flow, derived from stratidied wavy)

Figure 3.3 Regime characteristics in horizontal pipe

The boundaries between the various flow patterns in a flow pattern map occur because a regime becomes unstable as the boundary (effect of shear force) is approached and growth of this instability causes transition to another flow pattern.

The other side, there are other serious difficulties with most of the existing literature on flow pattern maps, such Taitel- Duckler’s. One of the basic fluid mechanical problems is that these maps are often dimensional and therefore apply only to the specific pipe sizes and fluids employed by the investigator. Also there may be several possible flow patterns whose occurence may depend on the initial conditions, specifically on the manner in which the multiphase flow is generated (Brennen, 2005).

3.1.4 Empirical Correlation in Vertical Pipe In particular, horizontal flow regime maps must not be used for vertical flow, and vertical flow regime maps must not be used for horizontal flow. In vertical flow the force gravity opposes the dynamic forces. This result in slippage therefore it exhibits some different characteristics than horizontal flow and may be more complicated.

13


The gas-liquid of multiphase flow in vertical pipe are determined as follows: a. Bubble Flow The gas phase is distributed in the form of bubbles immersed in a continuous liqui d phase. b. Bubble - Liquid Slug Flow As the concentration of bubbles grows by the presence of a higher quantity of gas, bubbles group or coalesce into one whose di ameter approaches the pipe diameter. c.

Transition flow, Liquid Slug – Annular With greater flow rate, the bubbles formed in the bubble flow collapse, resulting in a sparkling and disorderly flow of gas through the liquid that is displaced to the wall of the channel.

d. Annular - Bubble Flow The flow takes the form of a relatively thick liquid film on the pipe wall, along with a substantial amount of liquid carried by the gas flowing in the center of the channel. e. Annular flow The liquid film is formed on the wall of the tube with a central part formed by gas (Anselmi, dkk., 2008).

Figure 3.4 Multiphase flow regimes in vertical pipe

Duns and Ros developed correlation for vertical flow of gas and liquid mixtures in wells. This correlation is valid for a wide range of oil and gas mixtures and flow regimes. Although the correlation is intended for using with dry oil/gas mixtures, it can also be applicable to wet mixtures with a suitable correction. For water contents less than 10%, the Duns-Ros correlation (with a correction factor) has been reported to work well in the bubble, slug (plug), and froth regions. The pressure profile prediction performance of the Duns & Ros method is outlined below in relation to the several flow variables considered:

14




Tubing Size. In general, the pressure drop is seen to be over predicted for a range of tubing diameters between 1 and 3 inches.



Oil Gravity. Good predictions of the pressure profile are obtained for broad range of oil gravities (13-56 °API).



Gas-Liquid Ratio (GLR). The pressure drop is over predicted for a wide range of GLR. The errors become especially large (> 20%) for GLR greater than 5000.



Water-Cut. The Duns-Ros model is not applicable for multiphase flow mixtures of oil, water, and gas. However, the correlation can be used with a suitable correction factor as mentioned above (Rao, 1998).

Figure 3.5 Duns and Ros flow regime map (N = Liquid Velocity Number, RN = Gas Velocity Number based on Eaton Correlation)

In Region I, at low gas numbers and high liquid numbers, one encounters a liquid with gas bubbles in it, as long as the gas-oil ratio is relatively low and the flowing pressure gradient primarily is the static head plus liquid friction loss.

For superficial liquid velocities less than 0,4 m/s (1,3 ft/s), increased gas flow causes the bubbles to combine and form plugs. As gas flow increases further these plugs collapse and form slugs. In these regions wall friction is rather negligible.

15


If Vsl is still less than 0,4 m/s but Vsg is about 15 m/s, or greater, the slug flow of Region II changes to mist flow in Region III.At this point the gas becomes the continuous phase with the liquid in droplet form and as film along the wall. In Region III wall friction is a major factor in pressure loss.

Froth flow which occurs across the lines of Regions I and II occurs at high liquid velocities, Duns and Ros expect it to occur when Vsl is greater then 1,6 m/s. At such rates no plug or slug flow was observed. No set flow pattern can be discerned ( Campbell, 2004).

The other vertical regime map is presented by Aziz et al . This map can be seen below.

Figure 3.6 Aziz et al . map

For manual calculation, Aziz is slightly more accurate than Duns and Ros due to the regime boundaries and calculation steps. This method is similar with Mandhane et.al because only based on superficial velocity of gas and liquid except it has been corrected for the fluid property by applying dimensionless numbers.

16


The coordinates used in the Aziz vertical map are:

                      =

=

0.333

=

0.25

=

Where

         ,,

,

and

are dimensionless number

o

= interfacial tension of air and water at 60 F o

= air density at 60 F and 14.7 psia

3.1.5 Beggs and Brill Brill Correlation In fact, Beggs and Brill (1973) correlation is one of many correlations used to predict the pressure loss in multiphase flow. Each multiphase correlation makes its own particular modifications to the hydrostatic pressure difference and the friction pressure loss calculations, in order to make them applicable to multiphase situations. The range of applicability of the multiphase flow models is dependent on several factors such as, tubing size or diameter, oil gravity, gas-liquid ratio , and two-

phase flow with or without water-cut. The effect of every factor on estimating the pressure profile in a well is discussed separately for all the multiphase models considered. A reasonably good performance of the multiphase flow models is considered to have a relative error (between the measured and predicted values of the pressure profile) less than or equal to 20% (Rao, 1998). 19 98).

In general, all multiphase correlations are essentially two phases (gas-liquid) and not three phases (gas, water, liquid). Accordingly, the oil and water phases are combined, and treated as a pseudo single liquid phase, while gas is considered a separate phase. The Beggs & Brill correlation is developed for tubing strings in inclined wells and pipelines for hilly terrain. This correlation resulted from experiments using air and water as test fluids over a wide range of parameters.

17


Table 3.1 Multiphase flow correlations Correlation

Notes

Vertical Upward Flow Duns & Ros

Good in mist and bubble flow regi ons.

Angel-Welchon-Ross

Applicable for high flow areas and annulus flow. Recommended for high volume wells and low gas/ oil ratios

Hagedorn & Brown

Best available pressure drop correlation for vertical up ward flow Most accurate for angles of inclination greater than 70 degrees

Orkiszewski

Result reliable for high gas/oil ratios Most accurate for angles of inclination greater than 70 degrees

Aziz

Generally slightly overpredicts pressure drop; other correlation tend to underpredict. This fact can be used to bracket bra cket the solution. Most accurate for angles of inclination greater than 70 degrees

Beggs & Brill

Good for all angles of inclination. Predicts the most consistent results for wide ranges of conditions.

Gray

Specifically designed for condensate wells (high gas/oil ratios) Recommended ranges: velocity< 15 m/s

Horizontal Flow Lockhart-Martinelli

Widely used in the chemical industry. ind ustry. Applicable for annular and annular mist fl ow regimes if flow pattern is known a priori. Do not use for large pipes Generally overpredicts pressure drop

Eaton

Do not use for diameters<50 mm [2 in] i n] Do not use for very high or very low liquid holdup. Underpredicts holdup for Hl<0.1. Works well for 0.1
Dukler

Good for horizontal flow Tends to underpredict pressure drop and holdup Recommended by API for wet gas lines

Beggs&Brill

Use the no-slip option for low holdup Underpredicts holdup

Inclined Flow Mukherjee-Brill

Recommended for hilly terrain pipelines New correlation based heavily on in si tu flow pattern Only available model that calculates flow patterns for all flow configurations and uses

18


Beggs and Brill model has been identified to be applicable in this study as it exhibits several characteristics that set it apart from the other multiphase flow models:

a) Slippage between phases is taken into account Due to the two different densities and viscosities involved in the flow, the lighter phase tends to travel faster than the heavier one – termed as slippage. This leads to larger liquid hold-up in practice than would be predicted by treating the mi xture as a homogeneous one.

b) Flow pattern consideration Depending on the velocity and composition of the mixture, the flow behavior changes considerably, so that different flow patterns emerge. Depending upon the flow pattern established, the hold-up and friction factor correlations are determined.

c) Flow angle consideration This model deals with flows at angles other than those in the vertical upwards direction (Chang et al ., ., 2008).

A little different from another correlation, Beggs and Brill need early determination of flow regime to calculate pressure drop. This part is used in this report for mechanistic models. These can be classified as three types of regimes: segregated flows, in which the two phases are for the most part separate; intermittent flows, in which gas and liquid are alternating; and distributive flows, in which one phase is dispersed in the other phase.



Segregated flow is further classified as being stratified smooth, stratified wavy (ripple flow), or annular. At higher gas rates, the interface becomes wavy, and stratified wavy flow results. Annular flow occurs at high gas rates and relatively high liquid rates and consists of an annulus of liquid coating the wall of the pipe and a central core of gas flow, with liquid droplets entrained in the gas.



The intermittent flow regimes are slug flow and plug (also called elongated bubble) flow. Slug flow consists of large liquid slugs alternating with high-velocity bubbles of gas that fill almost the entire pipe. In plug flow, large gas bubbles flow along the top of the pipe.



Distributive flow regimes described in the literature include bubble, mist, and froth flow.

19


Beggs & Brill determine the flow regime and liquid hold up as follow:

  −  −    −   1

2

0.302

= 316

2.4684

= 0.0009 0.0009252 252 3

= 0.1

4

= 0.52 .52

1.4516 6.738

L1, L 2, L 3, L4 are dimensionless numbers where can be determined if C L is known. Theoretically, Cl is input volume fraction of liquid that defined as the ratio of liquid superficial velocity and mixture velocity. The other dimensionless number that used to determine flow regime is Fraude number. It may be written as

   =

2

/

Based on above equations, the flow regimes are classified into four areas: 1. Segregated flow

2.Intermittent 2. Intermittent Flow

      ≥     ≤     ≤  ≥    ≤  ≥  ≥   ≥    

0.01

< 0.01 .01

<

1

0.01

<

2

< 0.4

0.04

3.Distributed 3. Distributed Flow

3

3

<

1

<

4

< 0.4

1

0.4

>

4

4.Transition 4. Transition Flow (not as an actual regime, only presents the existence of regi me boundaries)

0.01

≤     ≤ 2

<

3

20


Once the flow type has been determined then the liquid holdup can be calculated. Beggs and Brill divided the liquid holdup calculation into two parts. First the liquid holdup for horizontal flow, E L(0), is determined, and then the result is modified for inclined flow. E L(0) must be ≥ C L and therefore when EL(0) is smaller than C L, EL(0) is assigned a value of C L. There is a separate calculation of liquid holdup (EL(0)) for each flow type.

1. Segregated flow

 

2. Intermittent Flow

3. Distributed Flow

(0) =

(0) =

0

Where

0.4846

0.0868

0.845

0.5351

0.0173

            −−     −  0 =

4. Transition Flow

    

0.98

=

=

0

3

3

1.065

0.5824

0.0609

+

0

=1

3

If all of the steps have been completed, then the liquid hold up can be determined. This information will be used to find the actual liquid and gas velocity along the pipeline.

3.2 Sand Transportability Transportability in Pipe Sand transportation in multiphase pipelines depends on several factors. Some of these factors are:

Flow regime, hold up, fluid properties (such as viscosity), inclination of the pipe in hilly terrains, particle size distribution, relation between the superficial velocity of the liquid phase (Vsl) and the superficial velocity of the gas phase (Vsg), pipeline diameter, friction factors, etc. For example, a change of inclination implies a change in the flow pattern, and therefore, a change in the sand transportation and sand behavior. The same happens for the viscosity. If the viscosity of the liquid phase changes, the energy distribution of the gas and liquid phases is going to change, conditioning the geometrical distribution of both phases in the pi peline, which means changing the flow pattern.

21


Angelsen in 1989 found that sand transport in horizontal pipelines has four main patterns depending on the fluid flow rate (Salama, 1998). Basically, four flow regimes can be identified for the solid-liquid slurry flow in horizontal pipe; those are saltation static bed (sand bed), saltation moving bed

(moving dunes), heterogeneous flow (scouring), and homogeneous flow (dispersed) (Chen, 1994).

Figure 3.7 Flow pattern in slurry flow

However, sand transportation in pipe concludes of more complex fluid composition: water, oil, and gas. The flow regime determined before only told about the gas-liquid phase distribution, so that the behavior and flow pattern map is a combination from gas-liquid and slurry flow.

Figure 3.8 Multiphase flow regimes consist of liquid, gas and solid

In annular flow, the sand particles can be transported in the water firm and in the gas core. In this flow regime, since the velocities are high, the main concern is not the sand accumulation but the erosion rate produced by the aggressive sand particl es movement.

22


In low hold up wavy flow, the liquid is transported in a thin film on the bottom of the line, where the sand concentration may be high, enhancing the creation of a settled sand bed.

In plug flow, gas pockets move along the top of the pipe having little effect upon the solid behavior. As long as the gas velocity is increased, the gas pocket gets depth and the fluctuating velocities affect the sand transportation similar to described at next in slug flow. Under this flow regime, for upwardly inclined pipes, it can be seen that either the sand is transported in the plug body and in the film region, or the sand particles settle in the gas plug zone (film region), and are only transported into the plug body, or clusters of collided sand particles are formed, moving backwards in the gas pocket (film region) and only moving forwards in the liquid plug body.

In slug flow, the sand particles behavior is complicated since the solids may be settled during the film region and transported in the slug body; the sand movement is always intermittent and gas pockets moving along the pipe have high effect upon the solid behavior. There can be a large diameter effect as the depth of the film varies and shields the pipe bottom from the turbulence of the slug. Moreover, the slug frequency is an important factor in sand transportation (Ruano, 2008).

Figure 3.9 Schematic sand behaviors in slug with low gas superficial velocity

Ruano in 2008 came with his observation about sand behavior in multiphase horizontal and near o

horizontal (+5 ) pipelines for his magister thesis. He tried to find the correlation between sand behavior and flow pattern and vice versa. Flow regime analysis is conducted through the measured hold up by capacitance instrumentation, for its comparison with the visual observation, and a relation between flow pattern and sand transportation is pointed up. The real sand transportation in multiphase oil pipelines is studied here by using water/air flows which contain different loads of

23


sand, by means of conducting sand settling experiments in the 4” (0.1 m) facility loop of Process and System Engineering Department at Cranfield University, for a liquid superficial velocity interval from 0.55 to 0.15 m/s, for a gas superficial velocity range from 2.5 to 0.02m/s and for three sand production rates: 0.04275, 0.57 and 1.425kg/m3.

It has been found that the sand transportation tr ansportation strongly depends upon the flow regime and, however, upon each and every parameter which affects the flow pattern, such as inclination or, even, sand production. It has been seen that the flow regime observed mainly depends upon the inclination, o

showing big differences between horizontal and near horizontal (+5 ). Therefore, the sand behavior observed in horizontal pipe is completely different that in the upwardly inclined pipe.

First, Ruano identified the flow regime without any sand load to study how sand concentration affects the flow pattern. For a certain value of Vsl, Vsg values are varied until all the regimes are concluded. Then he replied those methods with different sand production rates; a recorded video from the bottom of the pipe is conducted. a.

Smooth Stratified Flow 

No obvious sand particles movements in li quid film zone



Sand settled in the bottom, sand dune formation i n higher sand concentration

Figure 3.10 Sand behaviors in smooth stratified regime

Figure 3.11 Sand dune formation behaviors

24


b. Stratified-Wavy Flow 

Formation of the big waves



Sand are seen to settle along the flow direction



Enough energy for sand to be transported

Figure 3.12 Sand behaviors in stratified-wavy regime c.

Plug Flow 

High value of Vsl and low Vsg



Sand is not transported to barely move sliding in a plug body and settle in the film zone



Sand is encountered to be rolling or creeping

Figure 3.13 Sand behaviors in plug regime d. Slug Flow 

Facilitate sand transportation



As soon as the turbulent energy reaches the sand settled on the pipe well, the sand wil l be carried and lifted into the slug body

25


Figure 3.14 Sand behaviors in slug regime

3.3

Critical Flow Velocity in Sand Transport

3.3.1 Horizontal Pipe The critical flow velocity, Vc, is defined as the minimum velocity demarcating flows in which the solids (sand particles) form a bed at the bottom of the pipe from fully suspended flows. It is also referred to as the minimum-carrying or limiting-deposition velocity. Below this velocity, solids will settle out and slurry flow cannot be maintained.

Below a critical velocity, sand will drop out of the carrier fluid and form a stable, stationary sand bed. As the sand bed builds over time, the fluid above the bed is forced into a smaller cross-sectional area, causing the fluid velocity to increase. When the velocity reaches a critical value, sand is transported in a thin layer along the top of the sand bed. A steady-state is reached, such that the sand eroded from the top of the bed is replaced by new sand production from upstream. At higher velocities, the sand bed begins to break up into a series of slow-moving dunes, with sand particles transported from the upstream to the downstream side of the dune. As the flow velocity increases still further, the dunes break up entirely, and the sand forms a moving bed along the bottom of the pipe. At liquid velocities above the critical sand-carrying velocity, the sand is fully entrained in the fluid phase, and potentially entrained into the gas phase in multiphase flow (Danielson, 2007).



There are some theories from such as Durand-Condolios (1952) and Newitt et al . (1955) that used to calculate Vc. Durand-Condolios classified the flow of slurries slurries according to particle particle size. Newitt suggested that it also depends on the density of material, the mean velocity, and the pipe diameter. He also derived deri ved the evaluation of the energy losses due to flow of the fluid and solid particles.

26


    − =

[2

(

1)]0,5

Where D = pipe diameter; F L is a numerical constant depending on solids concentration and particle size and is determined from Figure 3.15 ; (s-1) is equivalent to

− 

also = SGsoIids- 1.

Sand Concentration

Figure 3.15 FL value vs. particle diameter, concentration as parameter



Salama, M. M. (1998) combine the correlation developed by Oroskar and Turian (1980) with predictions made by the DNV Carroline software for predict sand-settling in both single and two-phase flows.

    ∆     0,55

0,53

0,17

=

Where

0,09

0,47

Vm : mininum mixture flow velocity velocity to avoid sand settling, m/s VsL : ratio between liquid superficial velocity and mixture velocity D : pipe diameter, m

3.3.2 Vertical Pipe

∆   

: density difference between sands and liquid, kg/m : liquid density, kg/m

3

3 2

: kinematic viscosity, m /s

To determine sand settling velocity in sand transportability through vertical pipe, a model developed by Chien (1994) can be used.

27


     = 120

1 + 0.07 .0727

    −     −  2

1

1

Where Vm : miminum mixture flow velocity velocity to avoid sand settling, m/s

   

: effective viscosity at various shear rates, Pa.s

d : particle diameter, m : particle density, kg/m : liquid density, kg/m

3

3

The above models do not account for the impact of condensate and added chemicals on sand behavior and sand settling predictions. It is, however, expected that the above equation can lead to conservative results because oil wetted sand should be expected to settle at a lower velocity than the water wetted sand (Salama, 1998).

28

29 Analysis of Sand T ransportability ransportability in Pipelines

CHAPTER IV BEKAPAI OBSERVATION 4.1

Bekapai Production Network Configuration and Gas Lift


4.2 Pipeline Data and Fluid Fluid Volumetric Flow Rate

Pipeline

P(bar)

T (oC)

Q Oil (STBD)

Q Gas (MSCFD)

Q Water (BWPD)

8 inch BK-BP1

10

60

1

960

68

6 inch BJ-BB

56

60

0

1302

1

12 inch BB-BP1

10

60

339

1608

2152

6 inch BF-BL

11

60

175

1712

1177

6 inch BH-BG

13

60

422

1239

478

12 inch BL-BA

10

60

5011

9540

4263

30


4.2 Pipeline Data and Fluid Fluid Volumetric Flow Rate

4.3

Pipeline

P(bar)

T (oC)

Q Oil (STBD)

Q Gas (MSCFD)

Q Water (BWPD)

8 inch BK-BP1

10

60

1

960

68

6 inch BJ-BB

56

60

0

1302

1

12 inch BB-BP1

10

60

339

1608

2152

6 inch BF-BL

11

60

175

1712

1177

6 inch BH-BG

13

60

422

1239

478

12 inch BL-BA

10

60

5011

9540

4263

Particle Sieve Analysis 50 45 40 35 t 30 h g i e 25 W % 20

Particle B Particle C Particle D

15

Particle E

10

Particle F

5

Particle G

0

Particle H

Diameter (mm)

*Particle A has no sieve analysis

Figure 4.1 Particles sieve analysis

There are eight samples of sand particles that have been found in most area in Bekapai wells and separators (V 100 and V 120) since 2006 until 2009. For this analysis, those particles are used to identify the flow critical velocity in Bekapai pipelines.

30


CHAPTER V BASIC CALCULATION FOR FLOW REGIME PREDICTION (COMPARISON OF METHOD)

5.1

Empirical Correlation (Mandhane, Aziz et al . versus Beggs & Brill) This calculation has been applied in Microsoft Excel 2007 for all pipelines in Bekapai (except 6” BE -

BA, 6” BL-BA, 6” BG-BL, 12” BG -BL because the flow are in single phase). 

Objectives a. Determine the flow regime for each pipeline b. Comparative analysis between methods that used in determining flow regime



Variations Variation used in this calculation is pipe geometry (horizontal or vertical).



Assumptions a. Steady state flow b. There is not an inter-phase mass or energy transfer c.

Temperature and pressure are constants along pipeline

Table 5.1 Average pressures and temperatures in Bekapai pipelines Pressure (bar) 10

Average Temperature o ( C) 60

6 inch BJ-BB

56

60

12 inch BB-BP1

10

60

6 inch BF-BL

11

60

6 inch BH-BG

13

60

12 inch BL-BA

10

60

Pipeline 8 inch BK-BP1

5.2

OLGA versus Beggs & Brill The calculation has been applied for some case studies in Microsoft Excel 2007 and OLGA. OLGA was originally developed as a dynamic one dimensional modified two fluid model for two-phase hydrocarbon flow in pipelines and pipeline networks, with processing equipment included. Later, a

33


water option was included which treats water as a separate liquid phase. OLGA was developed by IFE in 1983 for the Norwegian State Oil Company, Statoil.

This comparison analysis is based from the main background to know how accurate the prediction made by mechanistic model such OLGA compared with Beggs & Brill, which is really helpful to solve kind of situations such as lack of appropriate data when a new well or pipeline is being designed, “Industry Standard” correlations correlations do not fit the available test d ata for some or all wells, different correlations are used to match similar wells, or the same correlation yield incomparable results in different application.



Objectives The objective of this analysis is to compare flow regime, actual liquid and gas velocity, and also holdup results between OLGA and Beggs & Brill model at Bekapai pipelines.



Variations a. Pipe geometry (horizontal or vertical) b. Flow mixture (gas-water or gas-oil)



Assumptions a. Steady state flow b. There is not an inter-phase mass or energy transfer c.

Temperature and pressure are constant along pipeline

First, the representative pipeline models for this analysis are created. In most cases, the wellhead located under sea level and linked to the platform with subsea pipeline. Therefore, the geometry of the pipeline includes horizontal line and riser in order to reach the production deck located above sea level. level. The horizontal lengths follows real conditions in Bekapai, with riser are assumed 35 m height. Pi peline is divided into 100 horizontal and 10 vertical sections.

In real situation, the physical structure of pipe would follow the seabed contour. Moreover, the flow regime performance is really sensitive to the inclination angle; defined as angle between pipeline and the ground. In these following cases, inclined angles factor along the pipe are ignored.

34


Table 5.2 Pipeline geometry data

Pipeline

d (inch)

Length (m)

Wall thickness (mm)

BB-BP

12

1660

9.52

BH-BG

6

1900

9.52

BF-BL

6

1000

9.52

BJ-BB

6

850

9.52

BK-BP

8

1900

9.52

BL-BA

6

1530

9.52

In order to determine the fluid regime in multiphase flow, the input and output data are identified. Fluid data like composition and phase mixture are created in another program (PVT SIM) because OLGA is not applicable to build its own fluid data source. For case studies applied, the gas and oil composition can be seen below.

Table 5.3 Oil composition in OLGA

Components C6 C7 C8 C9

Mol fraction 0.33 0.12 0.005 0.04

Molecular weight 84 96 107 121

0.14 0.16 0.07 0.005 0.1 0.03

134 147 161 175 190 206

C10 C11 C12 C13 C14 C15



778 789 800 811 822 832

Table 5.4 Gas composition in OLGA Components C1

Molecular weight 16.043

3

(kg/m ) 685 722 745 764

Yi 0.75

C2

30.07

0.21

C3

4.097

0.04

35


OLGA provides an option to activate “NO SLIP” indicator. If it is turned off, slip between phases is calculated. In the other words, the actual liquid velocity between gas and liquid phase are become different each other (like the real situation). The other option is “STEADYSTATE”, indicator used to establish the steady state condition instead of transient.

o

Heat transfer is neglected for simplicity, and the wall temperature is assumed constant in 60 C. The other parameters like pressure, GOR (gas oil ratio), GWR (gas water ratio), pipe diameter, and standard gas flow rate follows data in Bekapai so that the model remains as close as possible to the actual circumstances. Nevertheless, in order to analyze the effect of fluid properties through flow regime, there are two variations in fluid flow applied: gas-water and gas-oil flow.

Table 5.5 Flow properties in each Bekapai pipeline

Pressure (bar) 10

GWR 3 3 (m /m ) 277.82

GOR (scf/STB) 720676.32

Gas std flow rate (Mscfd) 960

6 inch BJ-BB

56

5220.60

-

1302

12 inch BB-BP1

10

14.76

4742.69

1608

6 inch BF-BL

11

26.07

9781.97

1712

6 inch BH-BG

13

39.15

2953.69

1239

12 inch BL-BA

10

44.22

1903.88

9540


36


Figure 5.1 OLGA model view for gas-water case

It must be noticed that the output in OLGA are presented in two different ways, in TRENDPLOT and PROFILEPLOT. TRENDPLOT shows the behavior of variables versus time in constant position

(called “section”). In the other hand, PROFILEPLOT shows the variable profile along the pipe in certain range of time. So the results are not constant toward two variables, location and time.

37


CHAPTER VI RESULTS AND DISCUSSION

According to the previous explanation about the theory and calculation used in this report, it was a great challenge to make an analysis and summary about many parameters concluded in sand transportability behavior. In the beginning, it was only to find the relation between flow regime and sand settling, but the phenomena are not as simple as ever thought. Multiphase flow regime, in fact, has not been really understood by researchers until now, since most of multiphase models are based on only two phase, single liquid and gas. Many assumptions are made to generalize its application in complex flow like in oil-gas industry, but the real problem are too many parameters have been identified without sufficient correlation made.

In such a complex situation engineers avoid the mathematical difficulties by resorting to experimental methods and develop "correlation" for engineering application. These correlations are based on experimental results but, when the number of parameters that control the flow pattern is large, than even this basic problem has its difficulties.

Holdup

Flow regime

Further analyzed in this report Pipeline properties (D,roughness)

Fluid properties

Sand Transportation in Pipeline

ᵨ,

(Vsl,Vsg, σ,µ)

(critical flow velocity and sand behavior) Inclination Particle properties

(θ)

ᵨ

(Dp, ) Figure 6.1 Factors affected sand transportation in pipeline

38


To overcame this problem, this report only focused on the relationship between flow regime and sand transportation, especially sand behavior along pipeline. The other parameters such as inclination, hold up, etc. are discussed in general due to Bekapai flow as long as they have connection with flow regime determination.

A comparative study has been chosen to determine multiphase flow regime in Bekapai pipelines, due to various models that have been found so far. Inclination is the main issue in experimental/empirical correlation because of only one map accepted for a certain inclination angle. This is why the experimental correlation is not used in practice. There are several empirical used in this analysis, such as Taitel & Dukler and Duns & Ros map which widely used today, but they failed to describe flow regime because of the difficulty level applied in manual calculation and unclear boundaries between each regime. The others are not really accurate and not considered the slippage between phases.

Therefore, a different approach is introduced by another experimental correlation such as Duns & Ros, Hagedorn & Brown, Orkiszewski, and Beggs & Brill. They are based on experiments and used commonly to determine pressure drop in multiphase flow. From those ones, Beggs and Brill was chosen in this analysis because some advantages like liquid hold up and slippage consideration, relatively easy to use, and applicable in all inclination. But somehow it has some limitations in the application that explained below: 1. It has an increasing error if GLR (gas liquid ratio) above 5,000. 2. The experimental investigation was conducted for tubing size between 1 and 1.5 in. Any further increase in tubing size tends to result in an over prediction in the pressure loss. 3. The accuracy has been tested only for water-gas flow. Hence, it can be concluded that all models that explained above are not recommended to use in a different situation from which the experiment was done.

Then OLGA comes as one dimensional model of multiphase flow that capable to determine the sand behavior included its flow regime. OLGA has many improvements and makes multiphase flow analysis becoming easier to apply in industry. It can be used to make a prediction of oil-gas-flow behavior along pipeline in steady state or transient condition, something that have never investigated before.

39


A comparative study between these models is further investigated in this report. A block diagram in the following page shows the general mechanism of flow regime determination.

Flow regime

Sand Behavior

Mechanistic Model

Experimental Correlation

(all inclinations) inclinations)

Horizontal Pipe

Vertical Pipe

All inclinations

(Mandhane Map)

(Aziz Map)

(Beggs &Brill)

OLGA

Figure 6.2 Flow regime determination used in this analysis

Sometimes a very careful choice must be considered due to the percentage differences between the results. Even when the value is large, it means nothing related to the validity because the main

focus of this this analysis is “how closed”, closed”, not “how accurate”. accurate”.

In this chapter, chapter, the results results of

calculation described in chapter V will be analyzed further. This chapter will be divided into two sections: analysis of sand behavior and the flow critical velocity in vertical and horizontal pipe.

6.1

Analysis of Sand Behavior Behavior in Correlation with Flow Regime Essentially, flow regime defined as the physical distribution of the phases in flow, especially the distribution of energy. It has great effect to the sand transportation because the energy of flow has able to move sand and avoid its settlement as long as the velocity of mixture is not achieve the flow critical velocity. Every pattern occurred in flow has certain characteristics which depend on the superficial velocity of liquid and gas.

40


It is not an easy task to predict the sand behavior behavior from the flow regime and vice versa. Ruano (2008) has analyzed this subject comprehensively in his thesis and found that in each regime occurred, sand settling phenomena are possibly happened based on the rate of sand production and flow velocity. According to this, the only information gathered is about the sand behavior, not the sand settling condition. Thus, the main factor used to decide whether sand particles are settled or not is still the flow critical velocity.

It is important to note that all flow regime model used here are able to explain only two until three phases consist of single liquid-gas or gas-water-oil types. It will be described in the next section.

6.1.1 Experimental Correlation (Mandhane, Aziz Aziz et al. versus Beggs & Brill) The main reason to choose Beggs & Brill than the the other map is its simplicity. simplicity. In this model the dimensionless number equations are used to substitute the boundaries between each flow regime. Consequently, the flow regime map is not required anymore. In application, this method is preferable although the whole calculation is more difficult than empirical correlation.

6.1.1.1 Horizontal Pipe There are some maps that can be used to determine the flow pattern in horizontal pipe, but

Mandhane’s map is the one that widely used. This method is reported to give an overall accuracy of about 70% when compared to the full data bank on which it is based (6000 flow pattern observations).

41


Figure 6.3 Mandhane’s map for Bekapai pipelines

Blue node locations show the regime for each pipeline flow. This method is quite easy because it is only based on superficial velocity of gas and liquid. As summary, the regime for each pipelines are represented in table below.

Table 6.1 Flow regimes of Bekapai pipelines predicted using Mandhane’s map

Gas superficial velocity (ft/s) 3.64

Gas-oil superficial velocity (ft/s) 0.0134

Regime Stratified

6 inch BJ-BB

1.41

0.0003

Stratified

12 inch BB-BP1

2.71

0.2126

Stratified

6 inch BF-BL

10.48

0.4616

Slug

6 inch BH-BG

6.39

0.3074

Stratified

12 inch BL-BA

16.09

0.7916

Slug


42


The Mandhane map given in Fig. 6.3 was developed for horizontal lines flowing air and water at near atmospheric pressure. Inclinations in the range of 0.1-1.0 degrees can cause substantial regime boundary movement. With an assumption that Bekapai pipelines are straight horizontal in geometry (riser is not included), results above can be accepted. Besides, flow regime boundary adjustment has been observed due to fluid pressure, pipe diameter, and surface tension in this method. Because of three parameters above are assumed constant in these cases, the remaining problem is how if these results are being compared with Beggs & Brill correlation.

Table 6.2 Horizontal flow regimes i n Bekapai pipelines predicted using by Beggs & Brill correlation (revised)

Pipeline

Regime

8 inch BK-BP1

segregated

6 inch BJ-BB

segregated

12 inch BB-BP1

segregated

6 inch BF-BL

unknown

6 inch BH-BG

unknown

12 inch BL-BA

unknown

For 8”BK-BP1, 6”BJ-BB and 12”BB-BP1 pipelines, the flow regimes are matches with Mandhane so it can be concluded that the flow regimes for those pipelines are segregated/stratified. Segregated Segregated

includes annular and stratified in Beggs & Brill’s terms, so explicitly it can be said that the regimes are stratified, according to the Mandhane’s results. Beggs & Brill correlation, same with Mandhane, is based on water-air flow in early investigated. For 6”BF-BL, 6”BH-BG and 12”BL-BA, it does not show any information about the regime. One only reason is one or more requirements used to determine flow regime are out of boundaries.

Nevertheless, the Beggs & Brill correlation originally based from Beggs & Brill map. When a

problem like “undefined regime” happened, it is better to ensure the results using this map. One disadvantage of this model is the uncertainty of regime location related to the others. It does not give any information about how close or how far the flow from the other regimes or relative

43


position between flows. This map below illustrated more clearly about some information that are not provided by Beggs & Brill correlation.

Figure 6.4 Beggs & Brill map (1973) of Bekapai pipelines

Flow in 12”BL-BA, 6”BF-BL and 6” BH -BG are showed in transition and near transition regime. Their positions are quite far from other three ( 6”BJ-BB, 8”BK-BP1 and 12”BB-BP1). However, these results show different prediction from Beggs & Brill correlation that has been revised. With assumption that there is nothing wrong in calculation, it should be corrected once more to find another comparator.

Using the correlation which was published in 1973, Bekapai flow regime can be seen in Table 6.3. The results are same with Beggs Beg gs & Brill map in original line.

44


Table 6.3 Horizontal flow regime in Bekapai pipelines by Beggs & Brill correlation (1973)

Pipeline

Regime

8 inch BK-BP1

Segregated

6 inch BJ-BB

Segregated

12 inch BB-BP1

Segregated

6 inch BF-BL

Segregated

6 inch BH-BG

Segregated

12 inch BL-BA

Segregated

Finally, all flow regimes in Bekapai pipelines are determined. These results seem reasonable according to fluid velocity data. As known earlier, fluid velocity values (see Appendix A) between BB-BP1, BF-BL, and BH-BG are not very different each other. Only 12”BL-BA has high rate of gas 3

(tenth times higher than 6”BH-BG; 269,501 m /d). Hence, its gas velocity only approximately 5.15 m/s, a little bit larger than the others (0.43, 0.89, 1.12, 2.04, 3.34 m/s). Logically, it was not usual for Beggs & Brill to fail predicting the flow regime in this range. In order to make things clear, for all next cases the correlation from the origin paper wi ll be used.

According to sand behavior in each pipeline, stratified flow is occurred in relatively low liquid and gas velocity, so that the sand particles have consistent behavior in this regime. From wellheads, sand concentration in liquid phase tends to be higher than gas phase because the gravity factor. Liquid phase remains in the bottom and there is only little mass transfer between gas and liquid phase. In this situation, whether sand particles will be carried away or settled along the pipe depends on the liquid velocity. If the velocity is lower than the flow critical one, the sand will settle and in higher concentration, they will form sand dunes. But, the other hand, sand will be carried away by flow and there will be no sand accumulation in 8”BK-BP1, 6”BJ-BB, and 12”BBBP1.

Refer to Mandhane’s method, slug flow are occurred in 6” BF-BL and 12”BL-BA. Theoretically, slug regime is avoided in field because it introduces a flow rate and pressure intermittency that may be troublesome to process control, in example the flow can change from near 100% liquid to 100% vapor. High liquid rates may fill separators separators causing process trips due to high level. level. High vapor

45


rates can lead to flaring or temporarily overload compressors causing trips due to compressor instability and/or high pressure.

Nevertheless, the main focus in this section is the sand behavior in this regime which become more complex because of slug phenomenon. Mixing zone that occurred is very effective to move sand particles in the bottom. If pipe diameter is smaller, slug body can reach the bottom of pipe and wipe the sand dunes into it. Slug frequency is the important factor for sand transportation in this regime. In general, sand has m uch less possibility to settle in slug flow than stratified one.

6.1.1.2 Vertical Pipe/Upflow Risers In vertical flow, gravity is a main force in the flow behavior. The less dense fluid will flow up faster than the dense liquid and create create swirling patterns much like a milk shake mixer. mixer. The dense liquid will tend to flow downwards giving rise to what is defined as liquid holdup. For vertical flow, the stratified flow regime cannot exist as there is no preferred direction for the liquid to settle. An empirical flow regime map developed by Aziz et al . for vertical upward flow is shown in chapter 3. The coordinates for this flow map are the same as for the Mandhane map in Fig. 6.3 except that fluid property corrections are used.

46


Figure 6.5 Aziz et al . map of Bekapai pipelines

Until now, vertical map that can be used in various inclination angles does not not exist. Aziz et al . has made some correction in his method but in some situations it can make large error. Aggour et.al. (1996) from Saudi Aramco proved that this method provides better predictions for lower GLR values and higher water cuts (water volume fraction in oil/water mixture). In general, Aziz et al . tends a good precision for larger tubing sizes and may be greatly improved by implementing

Orkiszewski’s flow pattern transition cr iteria.

In Bekapai cases, slug flow dominates all upflow risers, while Beggs & Brill provides different predictions. According to Beggs & Brill, in vertical flow, the regimes are same as horizontal

47


because they follow same calculation rules. Thus, information about the results has been explained in horizontal section.

Table 6.4 Flow regimes of vertical Bekapai pipelines based on Aziz and Beggs & B rill correlation

Regime Pipeline

Aziz

Beggs & Brill

8 inch BK-BP1

Slug

Segregated

6 inch BJ-BB

Slug

Segregated

12 inch BB-BP1

Slug

Segregated

6 inch BF-BL

Slug

Segregated

6 inch BH-BG

Slug

Segregated

12 inch BL-BA

Slug

Segregated

As can be seen from Table 5.8, Aziz et al . and Beggs & Brill methods are not in agreement each other to determine vertical flow in Bekapai pipelines so there must be chosen between both of them. It may be not the main focus in this report, but when there is something like this happen in field; engineers are encouraged to find the best choice. If there is nothing wrong with calculation, Beggs & Brill has found to be better than Aziz et al . in accuracy with average percentage error about 6.72% (based on the present 414 data sets that cover a wide range of tubing size, production rate, water cut, and GLR,[Aggour, 1996]). Aziz et al. only achieved 15.5 % approximately. There are some cases like BK-BP1 and BJ-BB where GWR values are too high compared with Aziz et al. effective range or the water cut are too low (BL-BA and BH-BG). More details for GWR, GOR, and water cut values of each pipeline can be seen in Table 5.9.

Table 6.5 GWR, GOR, and water cut values of each Bekapai pipelines


GWR (m3/m3) 277.82

GOR (scf/STB) 720676.32

6 inch BJ-BB

5220.60

-

Water cut 0.98 1.00

12 inch BB-BP1

14.76

4742.69

0.86

6 inch BF-BL

26.07

9781.97

0.87

6 inch BH-BG

39.15

2935.69

0.53

12 inch BL-BA

44.22

1903.88

0.46

48


Granular materials like sand particles are known to show complex dynamical behavior, such as convection, size segregation, bubbling, standing waves, etc. in vertical pipe. Sand is impossible to settle under this condition, but much more effective to increase the pressure drop and erosion rates, especially for annular (segregated) flow in Bekapai pipelines. Annular flow exists at high superficial gas velocity and low superficial liquid velocity. The gas flows in the core region at high velocity and the liquid flows as a thin annular film around inside the pipe wall and partially in the form of liquid droplets entrained in the gas core. The droplet entrainment entrainment from a liquid film by a streaming gas flow is of considerable importance because the same mechanism that causes liquid droplets to be entrained can cause sand particles also to be entrained and contribute to the erosion/corrosion process in BK-BP1, BJ-BB, and BB-BP1.

6.1.2 OLGA versus Begs & Brill This section will explain further about flow regime in Bekapai pipelines by two different categories: mechanistic models (OLGA) and experimental correlation (Beggs & Brill). Beggs & Brill has been proved as the most accurate correlation for pressure drop prediction (Aggour et. al., al., 1996), while OLGA has known widely as multiphase flow simulator used in many fields. A little different with previous section, the fluid actual velocity and liquid holdup will be discussed since they have close relation with flow regime determination. determination. Sand behavior in each pipeline will not be explained related to the objectives of this analysis.

In fact, flow regime detected in pipe at a certain time and location should be different with another situation. So far, experimental correlation and Beggs & Brill have not yet considered effect of two dimensional parameters, time period and location. For example, sand bed formation for long period can cause smaller cross-sectional area for oil-gas flow and increase the flow velocity. It could change the flow pattern actually. Along pipeline, there will be a different concentration of sand, so that the flow regimes in pipe section are vary according to the location.

OLGA seems to pay attention more to those parameters. Besides, it has successfully generalized the flow pattern of horizontal and vertical flow into only four: stratified (1), annular (2), slug (3),

and dispersed (4) . BB has three different patterns (segregated, intermittent, and distributed) that also used for all inclined angles. The most difficult problem to solve in this case is how to compare

49


such flow regime, holdup, and fluid velocity profiles which strongly depend on period and physical parameters in OLGA (dynamic model) with Beggs & Brill correlation.

To overcome this problem, there are three sections which become the main concern of this th

th

analysis: one in horizontal section (50 section), in the bottom of riser (101 section), and in the th

pipe outlet (110 section). The parameters observed are the pressure, liquid volumetric flow rate, flow regime, hold up, liquid and gas actual velocity. The profiles will be investigated in 48 hours for every 10 minutes. Start from these, the comparative study with Beggs & Brill can be studied further, especially to predict the sand behavior.

In order to simplify the cases, water-oil-gas flow is divided into two types of flow: oil-gas and water-gas flow. Two phases flow phenomenon has been studied in many papers from various fields. In Bekapai, the flow is more complex; three phases flow (gas-oil-water) include solid or other deposits. This time, these are purely comparative studies between mechanistic model with experimental correlation with an assumption if there are only two phases exist (e.g. gas-oil or gaswater case). GOR values still follows the real condition in Bekapai.

Liquid holdup is that fraction of a pipe segment which is occupied by liquid. An estimation of liquid holdup is vital to analyzing two-phase flow systems because the liquid holdup not only determines the cross sectional area available for gas flow, but also determines the liquid inventory in the line. This is also associated with sand behavior and estimation of slug size. It is important to be noticed that liquid holdup is not the same as inlet liquid content in this case. If both values are similar, the method relies on the assumption that the gas and liquid travel through the pipe at the same velocity (no slip occurred between the phase). Beggs & Brill has considered the slippage in its correlation, while OLGA has provided alternatives to facilitate the requirements.

50


6.1.2.1 Oil-Gas Flow 6.1.2.1.1

8” BK-BP1

th

Figure 6.6 Flow regime, holdup, and fluid velocity at 50 section in 8” BK-BP1 (oil-gas flow)

Figure 6.7 Flow regime, holdup, and fluid velocity at riser bottom in BK-BP1 (oil-gas flow)

51


Figure 6.8 Flow regime, holdup, and fluid velocity at pipe outlet in 8” BK-BP1 (oil-gas flow) th

a. 50 Section (Horizontal Line) In horizontal line, OLGA illustrated the flow regime clearly in stratified flow. The pressure fluctuates (increases for approximately 0.5 bar) because there is bulk of oil that accumulate in o

riser bottom (effect of 90 elbow). It makes the liquid holdup increasing until close to 1,0. Then the pressure becomes very high to carry away the oil. The holdup range is between 0.07 and 0.95. Actual liquid and gas velocity are unstable according to slug formation.

b. 101th Section (Riser Bottom) Flow regimes are varies between stratified, annular, slug, and even dispersed in riser bottom. This is the first section of vertical pipe, and oil as heavy liquid become easier to accumulate here before it flows back. At first, the liquid velocity is too small to carry away oil through vertical section (<0.10 m/s). For a certain period ( ±50,000 s), gas is forced to flow in smaller cross sectional area until the way is totally blocked by oil (holdup values are between 0,0 until 1,0). Gas actual velocity reaches the highest value at this time. It can achieve 5.5 m/s at highest peak.

c. 110th Section (Pipe Outlet) Annular flow occurs within 48 hours at outlet pipe, except at T = 47,467 s and 132,670 s when it becomes slug. Slug is caused by phenomenon that has been described early. When oil blockage

52


occurs, gas velocity in this section decreases drastically and reaches its minimum value. Then the pressure along pipe becomes very high to anticipate the trapped gas. As consequences, gas and

liquid velocity increases sharply before it starts to move back to lower values because pressure’s falling down.

Table 6.6 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in oil-gas flow (8” BK-BP1)

Flow Regime Holdup Actual Liquid Velocity

Actual Gas Velocity

50th section (horizontal line)

Beggs & Brill (horizontal)

Stratified

Segregated

0.070-0.078 (fluctuating) Too low (fluctuating, closer to zero) 0.75-2.25 m/s (fluctuating)

0.01 0.01 m/s

1.12 m/s

101th section (riser bottom) Stratified, annular, slug, dispersed

110th section (pipe outlet) Mostly annular, slug, dispersed

Beggs & Brill (vertical)

0-1 (slug)

0-0.15 (slug)

0.02

Too low (assumed zero), except in slug regime (reach 1.5 m/s (-3.6)-5.5 m/s (back flow)

0-(-1.3) m/s

-1.3-(4) m/s (back flow)

Segregated

0 m/s

1.13 m/s

Since OLGA has dynamic value, it is quite difficult to compare the results with Beggs & Brill. Moreover, some results are too low and fluctuating because oil accumulation in horizontal line. Oil accumulation is possibly occurred because there is not enough pressure drop and pipe oversize. In the other hand, OLGA and Beggs & Brill still show similar results in flow regime determination and most of Beggs & Brill predictions have been included in OLGA results.

53


6.1.2.1.2

12” BB-BP1

th

Figure 6.9 Flow regime, holdup, and fluid velocity at 50 section in 12” BB-BP1 (oil-gas flow)

Figure 6.10 Flow regime, holdup, and fluid velocity at riser bottom in 12”BB-BP1 (oil-gas flow)

54


Figure 6.11 Flow r egime, holdup, and fluid velocity at pipe outlet in 12”BB-BP1 (oil-gas flow)

a. 50th Section (Horizontal Line) th It can be seen in Figure 6.8 that stratified regime is occurred in 50 section although the velocity of gas are varies from -0.79 until 2.25 m/s. Oil accumulation is still possible to make the liquid holdup reach value of 0.5. It also causes actual velocity of gas to fluctuate and reach its highest level.

b. 101th Section (Riser Bottom) There is such a complex phenomenon occurred in BB-BP1 since it has slug regime. The first slugdispersed is identified in 22,203 s, caused by accumulation of oil in the bottom of r iser. The holdup in this section increases drastically from 0.19 until 0.723 before it becomes slug. The pressure also rises up until 11.94 bar and becomes fluctuating. Oil is carried away after 26,400 s approximately. Then the pressure is stable at 11.04 bar before the next slug regime occurs.

Flow regimes change between stratified, slug, dispersed , and annular. Liquid velocity fluctuates when flow become slug and dispersed. Period of each flow regime depends on values of liquid and gas velocity. The negative value of liquid velocity causes the dispersed regime while gas velocity is high. There is situation when gas velocity reach 8 m/s and flow back at 4 m/s. Gas and oil create a serious turbulence that affect the outlet product of pipe.

55


c. 110th Section (Pipe Outlet) In general, outlet section has annular regime, but there is one time when it becomes slug and dispersed. It is very important to be noticed that this kind of situation should be avoided in real situation. High pressure and high volumetric flow of oil should cause some problems of stability and separator performance. For both regimes, liquid and gas velocity are very high (4.05 and 3.65 m/s).

Table 6.7 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in oil gas flow(12” BB-BP1) th 50 section Beggs & Brill 101th section 110th section Beggs & Brill (horizontal line) (horizontal) (riser bottom) (pipe outlet) (vertical)

Flow Regime Holdup Actual Liquid Velocity

Actual Gas Velocity

Stratified

Segregated

0.2-0.52 (fluctuating)

0.12

-0.79-2.25 m/s (fluctuating, closer to zero) 0.8-2.9 m/s (fluctuating)

0.08 m/s

0.93 m/s

Stratified, annular, slug, dispersed

Mostly annular, slug, dispersed

0-1 (slug)

0-0.25 (slug)

Too low (assumed zero), except in slug regime (reach 1.5 m/s (-4)-8 m/s (back flow)

0-(-1.5) m/s

-1.3-2.2 m/s (back flow)

Segregated 0. 15

0.14 m/s

2.18 m/s


56


6.1.2.1.3

6” BF-BL

Figure 6.12 Flow r egime, holdup, and fluid velocity at pipe outlet in 6” BF-BL (oil-gas flow)

Figure 6.13 Flow regime, holdup, and fluid velocity at riser bottom in 6” BF-BL (oil-gas flow)

57


Figure 6.14 Flow r egime, holdup, and fluid velocity at pipe outlet in 6” BF-BL (oil-gas flow)

a. 50th Section (Horizontal Line) Interaction between gas and oil phase occur in horizontal section. Although OLGA predict the flow regime is stratified, but it is not impossible for both phase to collide each other. This is supported by backflow of oil in the riser and horizontal section. Its velocity is quite high (>0.17 m/s). At the situation, gas also experiences flow back as consequences of liquid turbulences. In the other hand, 1,177 bpd of oil keep moving into the pipeline and make horizontal liquid holdup increases slowly from 0.17 until 0.18. Within 48 hours there is only a little amount of liquid can reach pipe outlet (liquid velocity = 0.001 m/s). It comes from water droplets carried by gas phase along the riser.

b. 101th and 110th Section (Riser) At the very beginning (T = 0 s), the pressure is large enough to fill the riser with oil (holdup = 1). The phenomena are actually same with BH-BG and BL-BA. Then liquid holdup will reach the average value lower than 0.1. The liquid amount becomes very small in the riser.

58


Table 6.8 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in oil-gas flow (6” BF-BL) th 50 section Beggs & Brill 101th section 110th section Beggs & Brill (horizontal line) (horizontal) (riser bottom) (pipe outlet) (vertical)

Flow Regime Holdup Actual Liquid Velocity Actual Gas Velocity

Stratified 0.17-0.18 (fluctuating) Too low (fluctuating, closer to zero) Too low (fluctuating,closer to zero)

Segregated

Annular

Annular

Segregated

0.06

0

0

0. 1

0.28 m/s

Negative (back flow)

0.001 m/s

0.18 m/s

0.002 m/s

0.001 m/s

3.63 m/s

3.78 m/s

Since OLGA has dynamic value, it is quite difficult to compare the results with Beggs & Brill. Moreover, some results are too low and fluctuating because oil accumulation in horizontal line. Oil accumulation is possibly occurred because there is not enough pressure drop and pip e oversize. In the other hand, OLGA and Beggs & Brill still show similar results in flow regime determination and most of Beggs & Brill predictions have been included in OLGA results.

59


6.1.2.1.4

6” BH-BG

th

Figure 6.15 Flow regime, holdup, and fluid velocity at 50 section in 6” BH-BG (oil-gas flow)

Figure 6.16 Flow regime, holdup, and fluid velocity at riser bottom in 6” BH-BG (oil-gas flow)

60

61


Figure 6.17 Flow regime, holdup, and fluid velocity at pipe outlet in 6” BH-BG (oil-gas flow) th

a. 50 Section (Horizontal Line) Outlet pressure in the outlet of 6 ”BH-BG is assumed constant at 13 bar. OLGA simulates the pressure inlet is 15.22 bar and create very high holdup in riser (1.0) and horizontal section (0.54). The actual velocities of gas and liquid in T= 0 s are very low (close to zero). Then the horizontal pressure fall down to 13.49 bar, makes oil back flowing down the riser. The holdup in the 50

th

section becomes 0.4.

Turbulences occurred in horizontal section, creates little waves with constant frequency. As consequences, liquid and gas velocities are unstable (their values changes between positive and negative values). It means gas flow is also influenced by the oil back flow.

b. 101th and 110th Section (Riser) In general, the velocities of mixture are very low (0.00-0.02 m/s) because the pressure drop between both of horizontal nodes are not high (only 0.01 bar). Oil is trapped in the horizontal section and gas still moves forward very slowly because of turbulence. In the riser bottom, oil is identified to flow back and has negative velocity. As consequences, there is only a little gas flow at the outlet section. There is no oil remaining.


Flow regimes are identified as stratified at 50

th

section and annular at the riser. Based on the

explanation above, those results are the most suitable to describe the phenomena.

Table 6.9 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill (6 ” BHBG) th th th 50 section Beggs & 101 section 110 section Beggs & (horizontal line) Brill (riser bottom) (pipe outlet) Brill (horizontal) (vertical)

Flow Regime Holdup

Stratified

Segregated

Annular

Annular

Segregated

±0.4, fluctuating

0.15

0

0

0.21

Actual Liquid Velocity Actual Gas Velocity

Too low (fluctuating, closer to zero) Too low (fluctuating,closer to zero)

0.28 m/s


Too low (fluctuating, closer to zero)

0.2 m/s

Too low, close to zero


2.01 m/s

2.16 m/s


62


6.1.2.1.5

12” BL-BA

th

Figure 6.18 Flow regime, holdup, and fluid velocity at 50 section in 12” BL-BA (oil-gas flow)

Figure 6.19 Flow regime, holdup, and fluid velocity at riser bottom in BL-BA (oil-gas flow)

63


Figure 6.20 Flow r egime, holdup, and fluid velocity at pipe outlet in 12”BL-BA (oil-gas flow) At time (T) = 0 s, 12”BL-BA pipeline has already fulfilled by water with horizontal and vertical hold up for 0.83 and 1. The pressure is about 12.14 bar in the inlet before it falls down to 10.13 bar within 160 s. Thus, oil will flow back into the horizontal section from riser and fulfills the area of pipe. This phenomenon create negative value of liquid actual velocity (-0.06 m/s) and also for gas. Hence, gas and oil is a little bit different. Oil will stay in horizontal pipe, make waves with constant frequency between two directions of flow while gas will continue to flow through the pipeline and the riser with average velocity 0.003 m/s.

th

a. 50 Section (Horizontal Line) Flow regime in horizontal pipe is stratified since most of oil are trapped, only a little amount of it is carried away by gas so that in the outlet pipe there always be liquid with actual velocity for 0.002 m/s (same with gas). Besides, the liquid holdup in the outlet always zero, different with the horizontal section which has liquid holdup approximately for 0.33. This value is increasing through time until reach 0.39 within 48 hours.

b. 101th and 110th Section (Riser) Annular regimes are occurred in outlet and in the bottom of the riser. The phenomenon is simply understood as follows: oil is flowing back through pipe wall while gas which carry little oil droplets

64


is flowing to the outlet. It is the reason why liquid velocities in the bottom of riser are always negative.

Table 6.10 Flow regime, holdup, and fluid velocity comparisons between OLGA OLGA and Beggs & Brill in oil-gas flow(12” BL-BA) th th th 50 section Beggs & 101 section 110 section Beggs & (horizontal line) Brill (riser bottom) (pipe outlet) Brill (horizontal) (vertical)

Flow Regime Holdup


Stratified ±0.35-0.4 (fluctuating)

Segregated

Too low (fluctuating, closer to zero) Too low (fluctuating,closer to zero)

0.92 m/s

0.12

5.8 m/s

Annular

Annular

Segregated

0

0

0.19


-2.5 -0 m/s (back flow)

0.68 m/s



6.08 m/s

Since OLGA has dynamic value, it is quite difficult to compare the results with Beggs & Brill. Moreover, some results are too low and fluctuating because oil accumulation in horizontal line. Oil accumulation is possibly occurred because there is not enough pressure drop and p ipe oversize. In the other hand, OLGA and Beggs & Brill still show similar results in flow regime determination and most of Beggs & Brill predictions have been included in OLGA results.

6.1.2.2 Water-Gas Flow Water and gas mixture has different characteristics (density, viscosity, and surface tension) from oil-gas in pipe flow. Basically, they will affect the sand behavior and flow pattern, such as viscosity that related to the energy distribution and fluid velocity. Hence, the main concern in this analysis is still comparative study between OLGA and Beggs & Brill, whether they still fit each other or not in water-gas horizontal flow.

65


6.1.2.2.1 12” BL-BA

th

Figure 6.21 Flow regime, holdup, and fluid velocity at 50 section in 12” BL-BA (water-gas flow)

Figure 6.22 Flow regime, holdup, and fluid velocity at riser bottom in 12” BL-BA (water-gas flow)

66


Figure 6.23 Flow regime, holdup, and fluid velocity at pipe outlet in BL-BA (water-gas flow) Water content in BL-BA is higher than any other Bekapai pipelines which are studied in this report. Moreover, gas volumetric flow is also very high (9,540 bpd). Liquid holdups in horizontal and vertical line are about 0.27 and 0.75. Then the values become low because the pressure decreases until 10.03 bar from 12.71 bar. Water in the riser falls down and creates turbulences with gas phase which flows in the opposite direction.

th

a. 50 Section (Horizontal Line) In stratified regime (horizontal line), water and gas are continuous phase. Since the liquid velocity is only 0.02 m/s, the liquid holdup increases slowly from 0.26 until 0.3 within 48 hours. It means gas can always reach the top of the riser with averag e velocity of 0.01 m/s.

th

th

b. 101 and 110 Section (Riser) The pressure drop now is only about 0.006 bar between 50

th

section and bottom riser. In this

situation gas and liquid velocities are fluctuating. The gas velocity is relatively too low to move water from horizontal area. Finally, liquid holdup in the riser is zero, but there is still little water droplets that come out from pipe. Liquid and gas velocity in this section are same ( ±0.01 m/s).

67


Table 6.11 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Bri ll in water-gas flow(12” BL-BA) th 50 section Beggs & 101th section 110th section Beggs & (horizontal line) Brill (riser bottom) (pipe outlet) Brill (horizontal) (vertical)

Flow Regime Holdup


Stratified ±0.35-0.4 (fluctuating)

Segregated 0.12 m/s

Too low, fluctuating,back flow Too low (fluctuating,closer to zero)

0.92 m/s

5.58 m/s

Annular

Annular

Segregated

0

0.01

1.443

0.015 m/s fluctuating -0.00448, (fluctuating)


0.015

0.68 m/s

0.01 m/s

From Table 6.6, it can be concluded that Beggs & Brill has the same flow regime with OLGA although the other values are not match. The possible reasons to explain its difference includes: the geometry of pipe, source location, and different principal between OLGA and Beggs & Brill to determine holdup and velocity.

6.1.2.2.2 6” BH-BG

th

Figure 6.24 Flow regime, holdup, and fluid velocity at 50 section in 6” BH-BG (water-gas flow)

68


Figure 6.25 Flow regime, holdup, and fluid velocity at riser bottom in 6” BH-BG (water-gas flow)

Figure 6.26 Flow regime, holdup, and fluid velocity at pipe outlet in 6” BH-BG (water-gas flow) th

a. 50 Section (Horizontal Line) The phenomenon is exactly the same with 12 ”BL-BA. The flow at 50

th

section is predicted as

stratified flow, while the other two sections are annular. Holdup at T = 0 s in horizontal and riser are 0.29 and 0.95. Before water can flow through the pipe outlet, the pressure moves from 16.56

69


bar to 13.03 bar. It makes lower pressure drop and lower fluid velocity (only 0.003 m/s average for gas and zero for liquid).

b. 101th and 110th Section (Riser) At the bottom of riser, water velocity is in negative value, means that water still flows back. In this section, holdup is zero, same with pipe outlet. Instead of water, gas has positive value in velocity even the velocity is very low (0.001 m/s). Gas and water has the same velocity at 0.001 m/s at the outlet. Table 6.11 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Bri ll in water-gas flow(6” BH-BG)

Flow Regime Holdup



101th section (riser bottom)

110th section (pipe outlet)


Stratified

Segregated

Annular

Annular

Segregated

0

0.01

0.23

0 0.16


Too low (close to zero) Too low (close to zero, fluctuating)

0.3 m/s


2.03 m/s

-0.00448 ,/s (fluctuating)


0.015

0.22 m/s

2.19 m/s


70


6.1.2.2.3 6” BF-BL

th

Figure 6.27 Flow regime, holdup, and fluid velocity at 50 section in 6” BF-BL (water-gas flow)

Figure 6.28 Flow regime, holdup, and fluid velocity at riser bottom in 6” BF-BL (water-gas flow)

71


Figure 6.29 Flow regime, holdup, and fluid velocity at pipe outlet in 6” BF-BL (water-gas flow) th

a. 50 Section (Horizontal Line) The phenomenon is exactly the same with 12” BL-BA. The flow at 50

th

section is predicted as

stratified flow, while the other two sections are annular. Holdup at T = 0 s in horizontal and riser are 0.36 and 0.95. Before water can flow through the pipe outlet, the pressure moves from 14.41 bar to 11.03 bar. It makes pressure drop is low and so does the fluid velocity (only 0.002 m/s average for gas and zero for liquid).

b. 101th and 110th Section (Riser) At the bottom of riser, water velocity is in negative value, means that water still flows back. In this section, holdup is zero, same with pipe outlet. Instead of water, gas has positive value in velocity even the velocity is very low (0.002 m/s). Gas and water has the same velocity at 0.002 m/s at the outlet.

72


Table 6.13 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Bri ll in water-gas flow (6” BF-BL)






Flow Regime Holdup

Stratified

Segregated

Annular

Annular

Segregated

0.36

0.16 m/s

0

0

0.23


Too low, fluctuating,back flow 0.004 m/s (fluctuating, back flow)

0.77 m/s

Too low, fluctuating,back flow


0.54 m/s


Too low, fluctuating,b ack flow

4.4 m/s

4.04 m/s


6.1.2.2.4 6” BJ-BB

th

Figure 6.30 Flow regime, holdup, and fluid velocity at 50 section in 6” BJ-BB (water-gas flow)

73


Figure 6.31 Flow regime, holdup, and fluid velocity at riser bottom in 6” BJ-BB (water-gas flow)

Figure 6.32 Flow regime, holdup, and fluid velocity at pipe outlet in 6” BJ-BB (water-gas flow) In OLGA, 6”BJ-BB simulation was developed at lower value than the real liquid flow rate in field (0.79 stb/d). Assuming that OLGA model is an exact representation of the flow data set, the phenomenon is exactly similar with the others which have been described (see Figure 6.29-6.31). Considering the value of liquid flow rate that is very low, the flow can be assumed only composed by gas phase, so that the behaviour is closed to single flow.

74


As can be seen from Figure 6.29, the regime is stratified in horizontal line. Gas velocity in this situation is very low, only for 0.0014 m/s while the liquid flow rate can be negligible. In the outlet 3

section, the gas volumetric flow rate can be represented by a value of 0.006 m /s, since the velocity of gas is very low (0.0002 m/s). Annular regime is occurred along the riser with liquid holdup is about zero. Water is likely to be coalesced in gas phase and carried into the pipe outlet. Table 6.14 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Bri ll in water-gas flow (6” BJ-BB)







Stratified

Segregated

Annular

Annular

Segregated

0.17

0.02 m/s

0

0

0.03

Too low, fluctuating, back flow

0.01 m/s

Negative, back flow


0 m/s


0.57 m/s



0.58 m/s

From Table 6.12, it can be concluded that Beggs & Brill has the same flow regime with OLGA although the other values are not match. The holdup values obtained from Beggs & B rill are not as low as OLGA, moreover, the actual gas velocity are rather too high. The possible reasons to explain its difference includes: the geometry of pipe, source location, and different principal between OLGA and Beggs & Brill to determine holdup and velocity.

75


6.1.2.2.5 8” BK-BP1

th

Figure 6.33 Flow regime, holdup, and fluid velocity at 50 section in 8” BK-BP1 (water-gas flow)

Figure 6.34 Flow regime, holdup, and fluid velocity at riser bottom in 8” BK-BP1 (water-gas flow)

76


Figure 6.35 Flow regime, holdup, and fluid velocity at pipe outlet in 8” BK-BP1 (water-gas flow) In addition to compare results that obtained from OLGA with Beggs & Brill, there are three sections which are observed further. The previous pipelines show significance values change through time. Most of them are unstable due to the involvement of various factors and effects of pipe geometry. Results of BK-BP1 have totally different characteristics based OLGA simulation. Holdup, liquid velocity and gas velocity have reached steady state within 14 minutes. Then their values are constant along 48 hours. The regimes are observed as annular in the outlet pipe and stratified in riser bottom and horizontal section. The results can be seen in Table 6.15.

77


Table 6.15 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Bri ll in water-gas flow(8” BK-BP1)







Stratified

Segregated

Stratified

Annular

Segregated

0.093

0.06

0.025

0.019

0.1

0.06 m/s

0.06 m/s

0.06 m/s

Back flow

0.04 m/s

0.99 m/s

0.93 m/s

1.23 m/s

0.99 m/s

1.19 m/s

This time the comparison results show promising progress, since the flow regime and liquid velocity for horizontal are similar. Holdup by Beggs & Brill seems too high for vertical flow according to OLGA, but for horizontal the result is still allowable. Besides, gas velocities are quite different between both of them. For pipe outlet, liquid velocity is negative, means the water is move downhill through pipe wall.

6.1.2.2.6

12” BB-BP1

Figure 6.36 Flow regime, holdup, and fluid velocity at 50 th section in 12” BB-BP1 (water-gas flow)

78


Figure 6.37 Flow regime, holdup, and fluid velocity at riser bottom in 12” BB-BP1 (water-gas flow)

Figure 6.38 Flow regime, holdup, and fluid velocity at pipe outlet in 12” BB-BP1 (water-gas flow)

Slug and dispersed regimes dominates this pipeline flow until 48 hours. It is difficult to obtain any information about liquid holdup and fluid velocity since the values are really inconsistent. The pressure varies between 10.4-11.6 bar. Liquid holdup for for horizontal section fluctuates fluctuates at average values of 0.4. The worst situation occurred in the bottom riser which has zero holdup until 1.0. It means water has fulfilled the horizontal pipe, hindered the gas path. As consequences, the

79


pressure will arise drastically to escape gas from water blockage. The maximum value of liquid and gas velocity in this section are 5.8 and 1.7 m/s respectively. The flow regime changes almost every 2 hours.

In general, the outlet section shows the same trend with riser bottom. In this section, the liquid and gas velocity are high enough to create some turbulences (see Figure 6.36 and 6.37). Holdup increases sharply when the flow regime turns into slug or dispersed. Table 6.16 Flow regime, holdup, and fluid velocity comparisons between OLGA and Beggs & Brill in water-gas flow(12” BB-BP1) th 50 section Beggs & 101th section 110th section Beggs & (horizontal line) Brill (riser bottom) (pipe outlet) Brill (horizontal) (vertical) Flow Stratified Segregated Annular Annular Segregated Regime Holdup 0.093 0.27 0.025 0.02 0.34


0.06 m/s

0.99 m/s

0.21 m/s

0.06 m/s

-0.28 m/s (back flow)

1.16 m/s

0.99 m/s

0.93 m/s

0.17 m/s

1.28 m/s

Both of them shows totally different for all variables observed. This may be accepted due to some reasons that have been explained above.

6.1.3 Main Finding 6.1.3.1 Experimental Correlation (Mandhane, Aziz et al . versus Beggs & Brill) Except OLGA, other model like Beggs & Brill and flow regime map depend strongly on the origin fluid and situation which the experiment defined. Many correction have been built to generalize such models, however, they were all developed and tested under limiting operating conditions.

This is a very challenging subject because gas-liquid-solid flows are usually very complex, due to the large number of variables involved in the transport processes, and typically poorly understood interaction between the variables. Many of the earliest investigations of this flow focused only on the settling tendency of solid particles without enough consideration about the sand behavior.

80


Based on literature study and calculation that have been analyzed, Beggs & Brill is better than Mandhane and Aziz et al . to predict the flow regime in Bekapai pipelines. According to this, all of All Bekapai pipelines consist of stratified flow in horizontal sections. This regime type is common in oil-gas transportation system in pipe. With gas and liquid velocities which are low and separated phase existence between liquid and gas, favorable condition for sand settling is likely created. Hence, in vertical pipe, slug and segregated flow regime are predicted. Stratified is not possibly occurred in vertical flow so that annular is estimated to occur in this position. Gas is the continuous phase in annular regime so that gas behaviour will mostly affect the sand transportation in Bekapai pipelines.

6.1.3.2 OLGA versus Beggs & Brill Comparative study between OLGA and Beggs & Brill has been done to learn the performance of both methods to determine multiphase flow properties (flow regime, liquid holdup, and fluid velocity) if they are applied in real cases in field. The flow regime results are found match in 6”BFBL, 6”BH-BG, 12”BL-BA, and 6”BJ-BB. In 8”BK-BP1 and 12”BB-BP1 pipelines, the similarity between OLGA and Beggs & Brill only exist in horizontal line since more complicated regime are shown by OLGA at riser bottom and pipe outlet. It is also proven that OLGA and Beggs & Brill prediction are found identical for flow regime determination either in oil-gas or water-gas flow.

Liquid holdup and fluid actual velocity that obtained from OLGA simulation are rather complex and dynamic. It makes own difficulties to compare them with Beggs & Brill. Thus, both results are not closed each other for all pipelines in Bekapai. The main reason lies in their different views of model application. Beggs & Brill indeed has different correlation for each inclination angle, but it does not consider the effect of inclination change over flow pattern itself. By the division of pipe into smaller segments including time consideration, OLGA is one step ahead from Beggs & Brill in this case.

Slug is found in 12” BB -BP1 and 8” BK -BP1 (gas-oil flow). In the other hand, 6” BF -BL, 6”BH-BG, 12” BL-BA and 6” BJ-BB do not experience slug, but fluid velocity in their horizontal section are unstable (turn into back flow within seconds). This turbulence or wave will occur and make sand settling easier because flow do not tends go forward smoothly.

81


Many factors or parameters that affect flow regime are included here, but for some reasons they are not described further in this report. Sand behavior in this section is not completely explained because the critical velocity to avoid sand se ttling has not been discussed.

6.2

Analysis of Sand Settling Condition Sand settling is an issue of concern at low velocities of oil-gas mixture along pipeline. Together with sand erosion and sand monitoring, sand settling is important elements of any effective sand production management strategy. Many literatures that reported about sand settling describe the phenomena as two phase flow of solid-liquid like Newitt, Stuhmiller, and Nunziato. Studies about oil-gas-water phase flow with sand existences are very rare; most of them used experimental investigation to determine critical flow velocity to avoid sand settling. D.G. Thomas (1961) defined the critical velocity as the mean stream velocity required to prevent the accumulation of a layer or either stationary or sliding particles on the bottom of a horizontal conduit.

Salama (1989) reported his investigation about sand production management which define sand settling as the transition between scouring and moving dunes (i.e. sand is on the bottom of the pipe but moving along the pipe). The flow velocity at this condition would be lower than the velocity to disperse the sand, but high enough to transport the sand through the pipeline. In this analysis, critical flow velocity definition in horizontal pipe is based on Salama. Comparative study between Salama and Bekapai cases can be seen in Table 6.14. Salama was chosen because his investigation is most closely approximates the condition of Bekapai.

Table 6.17 Salama versus Bekapai case Investigation by Salama (1989)

Bekapai case

100, 280, and 500 m

140, 180, 240, 235, 240, 256, 264, and 1000 m

4 in

6, 8, 12 in

Media

Water, gas (CO2, N2, air), oil, inhibitors, sand

Water, oil, gas, sand

Water cut

1%, 10%, 50%, and 100%

46%, 53%, 86%, 87%, 98%, and 100%

Pressures

4 and 8 bara

11, 12, 14 , 57 bara

Ambient

Wall (60 C)

Sand particle size Pipe diameter

Temperature





o

82


Table 6.17 Salama versus Bekapai case (continued) Liquid flow rate

0.03, 0.1, 0.2, 0.3, 0.4 m/s

0, 0.06, 0.09, 0.14, 0.24 m/s

Varied during tests

0.43, 0.83, 0.1.11, 1.95, 4.91 m/s

Gas flow rate

Same with flow regime experimental correlation, Salama equation cannot be used in vertical flow. Sand behavior observed in horizontal flow is completely different that in vertical pipe. Chien (1994) developed new correlation to predict the settling velocity of irregularly shaped particles in Newtonian and non-Newtonian fluid for all types of slip regimes. This model is recommended for using in vertical case.

Even there is a relationship between flow regime and sand transportation, as stated earlier; it does not mean that it will affect sand settling condition directly. Salama (2000) and Newitt (1962) have already reported about four flow patterns defined in slurry flow, however, they are different with flow patterns for multiphase flow as showed in previous section. Ruano (2008) came with his thesis to understand the sand behavior in multiphase flow in horizontal and near horizontal. He analyzed explicitly about flow regime relation with flow critical velocity under various sand production rates. The results indicate that there are always possible for sand to settle in each regime, depends on fluid velocity and rates of sand production.

Therefore, “sand settling” subject in correlation with flow regime can only be investigated by experiment until now. Hence, there are some other factors that also affect sand settling phenomena. They will be analyzed further in this section.

6.2.1 Horizontal Pipe o

This is easier to observe sand settling in horizontal line (inclination angle = 0 ) than vertical one o

(inclination angle = 90 ). There are many papers reported about the phenomena although only few of them fit with multiphase flow. Quantitatively, flow critical velocity in Bekapai pipeline can be seen in table below.

83

84


Table 6.18 Flow critical velocity in several Bekapai pipelines using Salama equation

8" BK-BP1

Particle

Vm range (m/s) vm<0.066

B

6" BJ-BB %w

Vm range

%w

Vm range

6" BH-BG %w

Vm range

12" BH-BG %w

Vm range

%w

d (mm

1.2

vm<0.015

1.2

vm<0.171

1.2

vm<0.053

1.2

vm<0.157

1.2

vm<0.27

1.2

0.1

0.015
0.1

0.171
0.1

0.053
0.1

0.157
0.1

0.27
0.1

0,038
0.07
14

0.015
14

0.181
14

0.056
14

0.167
14

0.286
14

0,063
0.074
26

0.016
26

0.192
26

0.06
26

0.177
26

0.303
26

0,106
0.077
44.5

0.017
44.5

0.2
44.5

0.062
44.5

0.183
44.5

0.315
44.5

0.081
9

0.018
9

0.211
9

0.066
9

0.194
9

0.333
9

0.084
3.9

0.019
3.9

0.22
3.9

0.068
3.9

0.202
3.9

0.347
3.9

vm<0.067 0.067
vm>0.02 6.98

vm<0.015

6.7

0.015
vm>0.233 6.98 6.7

vm<0.175 0.175
vm>0.072 6.98 6.7

vm<0.054 0.054
vm>0.214 6.98

vm<0.16

vm>0.367 6.98

6.7

0.16
6.7

d<0,038

0,15
6.98

d<0,038

0.275
6.7

0,038
vm<0.275

13.36

0.016
13.36

0.185
13.36

0.057
13.36

0.17
13.36

0.291
13.36

0,063
0.075
23.59

0.016
23.59

0.196
23.59

0.061
23.59

0.18
23.59

0.308
23.59

0,106
0.078
22.29

0.017
22.29

0.203
22.29

0.063
22.29

0.187
22.29

0.321
22.29

0,15
0.083
5.15

0.018
5.15

0.215
5.15

0.067
5.15

0.198
5.15

0.339
5.15

0,25
0.086
7.51

0.019
7.51

0.224
7.51

0.069
7.51

0.206
7.51

0.353
7.51

0,355
0.071
vm>0.091 vm<0.066

vm>0.02

vm>0.237

vm<0.015

0.87

0.015
6.16

0.016
6.16

0.182
6.16

0.074
39.2

0.016
39.2

0.193
0.077
14.5

0.017
14.5

0.201
0.082
17.11

0.018
17.11

0.085
22.1

0.019
22.1

0.070
vm>0.090

0.06 0.87

vm>0.02

vm<0.172

vm>0.074

0.06

0.066
D

%w

6" BF-BL

0.066
vm>0.09

C

Vm range

12" BB-BP1

vm>0.374

0,6
0.06

vm<0.158

0.06

vm<0.272

0.06

d<0,038

0.87

0.158
0.87

0.272
0.87

0,038
0.057
6.16

0.167
6.16

0.287
6.16

0,063
39.2

0.06
39.2

0.177
39.2

0.304
39.2

0,106
14.5

0.062
14.5

0.184
14.5

0.316
14.5

0,15
0.212
17.11

0.066
17.11

0.195
17.11

0.335
17.11

0.221
22.1

0.069
22.1

0.203
22.1

0.348
22.1

0.172
0.06 0.87

vm>0.234

vm<0.054

vm>0.218

0.054
vm>0.073

vm>0.215

vm>0.9369

0,25
85


Table 6.18 Flow critical velocity in several Bekapai pipelines using Salama equation (continued) vm<0.066 0.066
vm<0.016

0.48 2

vm<0.170 0.170
0.48 2

vm<0.053 0.053
0.48

vm<0.156

0.48

vm<0.268

0.48

d<0,038

2

0.156
2

0.268
2

0,038
3.91

0.165
3.91

0.284
3.91

0, 063
3.91

3.91

0.180
0.074
4.61

0.016
4.61

0.191
4.61

0.06
4.61

0.175
4.61

0.301
4.61

0,106
0.077
6.52

0.017
6.52

0.198
6.52

0.062
6.52

0.182
6.52

0.312
6.52

0,15
0.082
3.73

0.018
3.73

0.21
3.73

0.066
3.73

0.193
3.73

0.331
3.73

0,25
0.085
35.22

0.019
35.22

0.218
35.22

0.068
35.22

0.2
35.22

0.344
35.22

0,355
vm>0.090

43.54

vm>0.02

43.54

vm>0.231

43.54

vm>0.072

43.54

vm>0.212

43.54

vm>0.364

43.54

0,6
vm<0.065

4.36

vm<0.015

4.36

vm<0.17

4.36

vm<0.053

4.36

vm<0.156

4.36

vm<0.268

4.36

d<0,038

0.065
7.02

vm<0.016

7.02

0.17
7.02

0.053
7.02

0.156
7.02

0.268
7.02

0, 038
19.13

0.165
19.13

0.284
19.13

0,063
0.069
F

2

vm<0.015

0.015
0.070
E

0.48

19.13

0.015
19.13

0.18
3.91

19.13

0.056
0.056
0.073
15.81

0.016
15.81

0.191
15.81

0.059
15.81

0.175
15.81

0.301< vm<0.312

15.81

0,106
0.076
17.34

0.017
17.34

0.198
17.34

0.062
17.34

0.182
17.34

0.312
17.34

0,15
0.081
13.08

0.018
13.08

0.21
13.08

0.065
13.08

0.193
13.08

0.331
13.08

0,25
0.084
21.55

0.019
21.55

0.218
21.55

0.068
21.55

0.2
21.55

0.344
21.55

0,355
vm>0.089

1.71

vm>0.02

1.71

vm>0.231

1.71

vm>0.072

1.71

vm>0.212

1.71

vm>0.364

1.71

0,6
vm<0.069

3.51

vm<0.015

3.51

vm<0.17

3.51

vm<0.053

3.51

vm<0.156

3.51

vm<0.268

3.51

d<0,038

10.76

vm<0.016

10.76

10.76

0.156
10.76

0.268
10.76

0,038
17.24

0.015
17.24

0.165
17.24

0.284
17.24

0,063
0.066
17.24

0.17
10.76 17.24

0.053
85


Table 6.18 Flow critical velocity in several Bekapai pipelines using Salama equation (continued) vm<0.066 0.066
2

vm<0.170 0.170
0.48 2

vm<0.053 0.053
0.48

vm<0.156

0.48

vm<0.268

0.48

d<0,038

2

0.156
2

0.268
2

0,038
3.91

0.165
3.91

0.284
3.91

0, 063
3.91

0.180
0.074
4.61

0.016
4.61

0.191
4.61

0.06
4.61

0.175
4.61

0.301
4.61

0,106
0.077
6.52

0.017
6.52

0.198
6.52

0.062
6.52

0.182
6.52

0.312
6.52

0,15
0.082
3.73

0.018
3.73

0.21
3.73

0.066
3.73

0.193
3.73

0.331
3.73

0,25
0.085
35.22

0.019
35.22

0.218
35.22

0.068
35.22

0.2
35.22

0.344
35.22

0,355
vm>0.090

43.54

vm>0.02

43.54

vm>0.231

43.54

vm>0.072

43.54

vm>0.212

43.54

vm>0.364

43.54

0,6
vm<0.065

4.36

vm<0.015

4.36

vm<0.17

4.36

vm<0.053

4.36

vm<0.156

4.36

vm<0.268

4.36

d<0,038

0.065
7.02

vm<0.016

7.02

0.17
7.02

0.053
7.02

0.156
7.02

0.268
7.02

0, 038
19.13

0.165
19.13

0.284
19.13

0,063
19.13

0.015
19.13

0.18
3.91

19.13

0.056
0.056
0.073
15.81

0.016
15.81

0.191
15.81

0.059
15.81

0.175
15.81

0.301< vm<0.312

15.81

0,106
0.076
17.34

0.017
17.34

0.198
17.34

0.062
17.34

0.182
17.34

0.312
17.34

0,15
0.081
13.08

0.018
13.08

0.21
13.08

0.065
13.08

0.193
13.08

0.331
13.08

0,25
0.084
21.55

0.019
21.55

0.218
21.55

0.068
21.55

0.2
21.55

0.344
21.55

0,355
vm>0.089

1.71

vm>0.02

1.71

vm>0.231

1.71

vm>0.072

1.71

vm>0.212

1.71

vm>0.364

1.71

0,6
vm<0.069

3.51

vm<0.015

3.51

vm<0.17

3.51

vm<0.053

3.51

vm<0.156

3.51

vm<0.268

3.51

d<0,038

10.76

vm<0.016

10.76

10.76

0.156
10.76

0.268
10.76

0,038
17.24

0.015
17.24

0.18
17.24

0.165
17.24

0.284
17.24

0,063
0.066
G

vm<0.016

0.48

3.91

0.069
F

2

vm<0.015

0.015
0.070
E

0.48

0.17
10.76 17.24

0.053
0.074
13.13

0.016
13.13

0.191
13.13

0.059
13.13

0.175
13.13

0.301< vm<0.312

13.13

0,106
0.077
18.32

0.017
18.32

0.198
18.32

0.062
18.32

0.182
18.32

0.312
18.32

0,15
0.082
13

0.018
13

0.21
13

0.065
13

0.193
13

0.331
13

0,25
0.085
18.78

0.019
18.78

0.218
18.78

0.068
18.78

0.2
18.78

0.344
18.78

0,355
vm>0.212

5.26

vm>0.364

vm>0.090

5.26

vm>0.02

5.26

vm>0.231

5.26

vm>0.072

5.26

5.26

0,6
86


Table 6.18 Flow critical velocity in several Bekapai pipelines using Salama equation (continued)

H

vm<0.065

0.05

vm<0.015

0.05

vm<0.17

0.05

vm<0.053

0.05

vm<0.156

0.05

vm<0.268

0.05

d<0,038

0.065
2.08

vm<0.016

2.08

0.17
2.08

0.053
2.08

0.156
2.08

0.268
2.08

0,038
0.069
10.07

0.015
10.07

0.18
10.07

0.056
10.07

0.165
10.07

0.284
10.07

0,063
17.32

0.016
17.32

0.191
17.32

0.059
17.32

0.175
0.076
0.301
17.32

0,106
0.076
26.8

0.017
26.8

0.198
26.8

0.062
26.8

26.8

0.312
26.8

0,15
0.081
19.18

0.018
19.18

0.21
19.18

0.065
19.18

0.193
19.18

0.331
19.18

0,25
0.084
23.81

0.019
23.81

0.218
23.81

0.068
23.81

0.2
23.81

0.344
23.81

0,355
vm>0.089

0.69

vm>0.02

0.69

vm>0.231

0.69

vm>0.072

*Particle A is not included in this calculation because it has no sieve analysis.

0.69

0.182
17.32

vm>0.212

0.69

vm>0.364

0.69

0,6
86


Table 6.18 Flow critical velocity in several Bekapai pipelines using Salama equation (continued)

H

vm<0.065

0.05

vm<0.015

0.05

vm<0.17

0.05

vm<0.053

0.05

vm<0.156

0.05

vm<0.268

0.05

d<0,038

0.065
2.08

vm<0.016

2.08

0.17
2.08

0.053
2.08

0.156
2.08

0.268
2.08

0,038
0.069
10.07

0.015
10.07

0.18
10.07

0.056
10.07

0.165
10.07

0.284
10.07

0,063
17.32

0.016
17.32

0.191
17.32

0.059
17.32

0.175
0.076
0.301
17.32

0,106
0.076
26.8

0.017
26.8

0.198
26.8

0.062
26.8

26.8

0.312
26.8

0,15
0.081
19.18

0.018
19.18

0.21
19.18

0.065
19.18

0.193
19.18

0.331
19.18

0,25
0.084
23.81

0.019
23.81

0.218
23.81

0.068
23.81

0.2
23.81

0.344
23.81

0,355
vm>0.089

0.69

vm>0.02

0.69

vm>0.231

0.69

vm>0.072

0.69

0.182
17.32

vm>0.212

0.69

vm>0.364

0.69

0,6
*Particle A is not included in this calculation because it has no sieve analysis.


In general, sand will not settle in all Bekapai pipelines investigated here because actual mixture velocity is larger than its critical value for each pipe. Fluid properties factor is negligible since all flows are assumed to have same properties. The remaining factors are pipe diameter, particle diameter, and liquid superficial velocity. These three parameters are proportional with critical value. For example, the relationship between liquid superficial velocity and critical flow velocity is illustrated in figure 6.17.

0.07 0.06 ) s / 0.05 m ( y t 0.04

Particle A Particle B

87


In general, sand will not settle in all Bekapai pipelines investigated here because actual mixture velocity is larger than its critical value for each pipe. Fluid properties factor is negligible since all flows are assumed to have same properties. The remaining factors are pipe diameter, particle diameter, and liquid superficial velocity. These three parameters are proportional with critical value. For example, the relationship between liquid superficial velocity and critical flow velocity is illustrated in figure 6.17.

0.07 0.06 ) s / 0.05 m ( y t i c 0.04 o l e 0.03 v l a c i t 0.02 i r c 0.01

Particle A Particle B Particle C Particle D Particle E Particle F Particle G

0 8.22E-05

9.38E-02

1.41E-01

Particle H

liquid superficial velocity (m/s)

Figure 6.39 Critical velocity profiles in 6” BJ -BB, 6”BF-BL, and 6”BH-BG

Figure 6.17 shows that larger liquid superficial velocity will produce larger critical velocity (except for particle F because the values are too small). This value is specific for each liquid superficial velocity. Besides, the larger particle diameter, the larg er critical velocity occurs.

According to their low values, sand particles are predicted to still move through pipe. Erosion risk will be greater in this condition refer to sand concentration and the mixture velocity. Erosion is not main focus in t his report, but it i s included in sand transportability phenomena along pipe. Erosion is closely related to corrosion, which is defined as the phenomenon of a protective film of corrosion product being eroded away by the erosive action of the process stream, exposing fresh metal which then corrodes (API 14 E).

87


6.2.2 Vertical Pipe In his paper (1998), Salama suggested Chie n’s model to predict sand settling velocity. This model does not considered pipe diameter factor so that the critical velocity value is same for each particle in all pipelines. Considering the diameter range of each particle, information about flow critical velocity can be seen as also in ranges (Figure 6.40).

Table 6.19 Actual mixture velocity i n vertical Bekapai pipeline for each particle

Pipeline 8 inch BK-BP1 6 inch BJ-BB 12 inch BB-BP1 6 inch BF-BL 6 inch BH-BG 12 inch BL-BA

50

106 µm
45 40 35 t 30 h g i e 25 w % 20

15 10

Liquid velocity (m/s) 0.04 0.0008 0.67 1.64 1.14 2.21

mixture velocity (m/s) 1.28 0.48 1.59 5.14 3.26 7.72

d>600 µm 355µm< d<600 µm

150 µm
63 µm
250 µm
38 µm
particle B particle C particle D particle E

d<38 µm

particle F

5

particle G

0

particle H

critical velocity (m/s)

Figure 6.40 Range of critical velocity in several Bekapai B ekapai pipelines based on particle diameter

88


Sand dune formation will be created if critical flow velocity is smaller than mixture velocity. When particle diameter becomes larger, it has bigger possibility to settle down. According to each particle size, one particle tends to have different critical velocity. Table 6.15 shows that only particle with diameter > 600 µm (particle E, F, G, and H) will settle in 6 ”BJ-BB pipelines. O

Because the pipe position is vertical, most of the sand particles will settle at 90 elbow before riser. Particle H seems to be found in high concentrate than the others (43.5%).

Since particle size has not been known exactly except in range form, these particles still have chances to settle among other pipelines. In pigging report during year 2010 in Bekapai, sand has

been found in 12” BL-BA, 12” BB-BP1, and 6” BF-BL. This has been proven by fluid velocity in 12” BB-BP1 (1.28 m/s) which is not too high. In BF-BL and 12” BL-BA cases, settling phenomena may be came from particle size and high sand concentration from wellbores.

Annular regime in vertical Bekapai pipelines indicates high gas velocity. This is may be a good news to find that sand settling has not the main concern in oil-gas transportation along sea line and riser yet because sand particles will easily be swept away. But sand usually has higher concentration in liquid phase (liquid velocity in each pipeline can be seen in Table 6.15). Sand particles (E, F, G, and H) are predicted to be found in 8” BK-BP1 and 6” BJ -BB.

6.2.3 Main Finding The important factors that affect sand settling phenomenon are particle diameter, fluid density, liquid velocity, kinematic viscosity, and pipe diameter. Chien was not considered the last two parameters in his correlation so that for same particle, it has the same critical velocity although the pipeline properties are different. In Bekapai case, sand settling is likely occurred in vertical flow than horizontal one. Only BJ-BB has sand settling problem with particle E, F, G, and H while the others cannot be definitely decided. Sand possibility to settle settle in the other other pipelines is still need to be considered and anticipated seriously.

89


Since specific information of sand particles in each pipeline had not been received until this report was finished (all particles used in this analysis are not the real particle found in those pipelines), the above conclusions cannot be fully accepted. It means sand behavior in Bekapai pipelines is still very complex and need to be studied further with the real model of Bekapai pipelines and adequate data about specific sand particles.

90


CHAPTER VII CONCLUSIONS AND RECOMMENDATIONS

7.1

Conclusions Oil-gas-water flow including sand transportation in pipeline is affected by many factors, such as flow regime, liquid holdup, fluid velocity, fluid properties, properties, pipe properties, etc. Sand behavior and flow regime are interrelated but until now there is no exact correlation made to wholly describe the sand settling phenomena in each regime.

In Bekapai case, parameters like pipe diameter and fluid properties should be put into sand transport consideration. They give a big impact of flow regime and flow critical velocity estimation. They may become a good reason why OLGA is chosen between the other models to determine multiphase flow properties. Beggs & Brill and the other correlations depend strongly on the fluid and pipe diameter in their origin experimental investigation.

Another important parameter in sand transportation is the effect of pipe geometry (i.e. pipe diameter). This is the key to solve problem about flow critical velocity determination. Salama and o

Chien provide correlation without sufficient attention about this (i.e. 90 elbow between sea line and riser). As consequences, their results regarding critical velocity to avoid sand bed formation must be ensured with another model that capable to illustrate multiphase flow phenomena, especially in the transition section between sea line and riser. In this analysis, only OLGA has powerful basic and applicable to be used in several Bekapai pipelines which already oversized due to production decline. Prediction like oil blockage and slug formation (8 ”BK-BP1 and 12”BB-BP1) can be used to support fur ther analysis of sand transportability in Bekapai pipelines.

7.2

Recommendations 

It is recommended to take a precaution over sand accumulation, especially at the ri ser bottom or another transition section of pipelines due to analysis results. Fluid mixture velocity should be enhanced until exceed the critical flow velocity to prevent initial sand bed formation.

91




Routine pigging should be done in pipelines that have been detected to experience sand settling. Some pipelines which have low fluid mixture velocity (6 ” BJ-BB, 8” BK-BP1, and 12” BB-BP1) should be placed at top priority.



Because sand settling phenomena strongly depends on the present data of fluid volumetric rate in pipelines, this analysis is recommended to be routinely updated.



It is recommended to use OLGA instead of Beggs & Brill and experimental correlation in application to determine multiphase flow properties, especially flow regime and dynamic behavior of each parameter included.



It is recommended to do further study and analysis about this topic, especially about the other parameters correlation that affecting sand behavior (e.g. pipe geometry and fluid properties).



It is better to use real model of Bekapai pipelines and fluid in order to be applied in the future.

92


93

Analysis of Sand Transportability in Pipelines

Recommend Documents