Chapter | three
Acceptance Tests with Centrifugal Pumps 3.1 PRELIMINARY REMARKS The The purp purpos ose e of an ac acce cept ptan ance ce test test is to demo demons nstr trat ate e the the fulf fulfil illm lmen entt of the the technica technical, l, hydrau hydraulic lic and mechan mechanical ical guaran guarantees tees agreed agreed between between the purchaser and the pump maker. Every proof of guarantee entails costs and should therefore be confined to the data essential to satisfactory service. For standard pumps, only type tests are usually performed.
3.2 PUMP TESTS These tests are classified as follows: 1. The works or factory acceptance test is performed on the manufacturer’s test stand and is very accurate thanks to the technique of measurement employed and the reproducible conditions. Measure remen ments ts on a pump pump ca carr rrie ied d out out in the the fiel field d (fie (field ld test tests) s) 2. Measu are subject to the influence of the prevailing service conditions. Accuracy Accur acy and reliability reliab ility of the measured measu red results result s depend largely largel y on the instrumentation instrumentation and measuring measuring positions. Where contractually agreed, however, acceptance testing can be carried out in the field. Periodic ic field field tests tests serve serve to detect detect changes changes or wear wear on the pump pump and 3. Period shou should ld be ca carr rrie ied d out out to a fixe fixed d sc sche hedu dule le usin using g inst instru rume ment nts s of unva unvary ry-ing standard. Model tests tests ca call ll for for high high me meas asur urin ing g ac accu cura racy cy.. The The mode modell test test must must be 4. Model defi define ned d so that that it se serv rves es as a sub substit stitu ute for for the the ac acce cept ptan ance ce test test on the the full-scale pump. One hundred percent hydraulic model fidelity is a prerequisite.
69 Centrifugal Pump Handbook. DOI: 10.1016/B978-0-7506-8612-9.00003-6
70
Acceptance Acceptance Tests with Centrifugal Centrifugal Pumps
3.2.1 Acceptanc Acceptance e rules rules The rules are laid down in standards in order to simplify understanding between pump maker and purchaser. They contain in general: 1. Definitions of all variables needed to define the functions of a centrifugal pump and to fix the guarantees for its hydraulic performance, flow capacity and total dynamic head and pump efficiency or power. Definitions of the technical technical guarantees and their implementation implementation.. 2. Definitions Recomm mmen endat datio ions ns for for prep prepar arin ing g and and perf perfor ormi ming ng accept acceptan ance ce tests tests to 3. Reco verify the guaranteed data. 4. Rules for comparing measured results with guaranteed data and for the conclusions to be drawn from the comparison. Requirements for curves and test reports. reports. 5. Requirements Descrip ipti tion on of the the prin princi cipa pall me measu asuri ring ng tec techn hniq ique ues s used used for for guara guarant ntee ee 6. Descr demonstrations. In this connection, the terms ‘‘guarantee’’ and ‘‘acceptance’’ in the stan stand dard ards are are to be under nderst stoo ood d in the the tech techn nical ical and and not not in the the leg legal se sen nse se.. Most standards applied nowadays contain two aspects: 1. The standards are intended for determining and analyzing measured results on a statistical basis and must ensure that the true value of the measured variable is ascertained with a statistical reliability of at least 95%. 2. The quality classes are defined in similar fashion in most standards, e.g. ISO 9906 accuracy grades 1 and 2. 2. Note Note:: Alth Althou ough gh stan standa dard rds s such such as ISO, ISO, API, API, HI and and DIN DIN are are beco becomi ming ng incr increas easin ingl gly y harmo harmoni nize zed d we refe referr to them them indi indivi vidu dual ally ly here here as the the indu indust stry ry continues to commonly use this terminology. Top class (ISO 5198), only the smallest deviations are tolerated: Model pumps The ISO accuracy grade is employed chiefly for research, development and scientific work in laboratories, where very high measuring accuracy accuracy is demanded. demanded. In exceptional cases for very large pumps (e.g. >10 MW*) for energy generation, water transport and feed pumps, where the engineering grade 1 is too inaccurate. Higher class (ISO 9906 accuracy grade 1), medium deviations are allowed: Pumps for conveying liquids, for injection and industrial power generation at medium outputs (0.5–10 MW*). Lower class (ISO 9906 accuracy grade 2), larger deviations are possible: Standard pumps manufactured in series for industrial duties (e.g. with type testing). In ISO 9906 the above accuracy standards define the required accuracy of the instruments used in the acceptance test. .
.
.
71
Pump Tests
No cons constr truc ucti tion on tole tolera ranc nces es are are laid laid down down in the the ISO ISO ac acce cept ptan ance ce rule rules. s. On the other hand, the overall tolerances are defined, taking into account the admissible measuring uncertainties and the construction tolerance.
3.2. 3.2.2 2 Comp Compar aris ison on of ISO, ISO, API API 610 610 and and Hydr Hydrau auli lic c Inst Instit itut ute e Standards with regard to guarantee points and measuring uncertainties
ISO IS O st stan anda dard rd ac accu cura racy cy gr grad ade e
5198 5198 precision grade top
9906 grade 1 middle
9906 grade 2 low
The construction tolerance is included in the increased overall measuring uncertainty and shown in the adjacent method for determining the service data and efficiency. To verify guarantee fulfillment a straight line is drawn through the points Q N, H N and Q = Q = 0, H = H = 0, giving the intervals DQ and Q and DH as H as horizontal and vertical distances from the measured head curve H ( H ( Q Q ). A perpendicular through the intersection of the straight lines 0 – Q N H N with the head curve determines the efficiency h0 .
The service guarantee is thus fulfilled
HN Á XH DH
XH XQ 0
> x h Á hN X NPSH X NPSH $ 3% or
QN Á XQ þ DQ
2
!
Æ 0.03 Æ 0.045
(0.015) (0.025) (0.978) (0.03 < 0.15 m)
h
2
!
!1 0.05 0.08 0.95 0.06 < 0.3 m
0.97 0.03 < 0.15 m
max. NPSH deviation Max. admissible measured value scatter based on 95% statistical accuracy: Number of observations per measuring point
3
5
7
9
3
5
7
9
3
5
7
9
n in % (speed) Æ n in P in % Æ Q , H , P in
0.25 0.5 0.7 0.9 0.3 0.5 0.7 0.8 0.6 1.0 1.4 1.6 0.8 1.6 2.2 2.8 0.8 1.6 2.2 2.8 1.8 3.5 4.5 5.8
(service data) Admissible total uncertainty for measuring instruments and measurement: Values for measuring instruments (...) Flow rate Head Power input Efficiency Speed 1)
For determining pump efficiency
( Æ 1.0) Æ 1.5% ( Æ 0.5) Æ 1.0% ( Æ 0.6) Æ 1.3% Æ 2.25% ( Æ 0.1) Æ 0.2%
( Æ 1.5) Æ 2.0% ( Æ 1.0) Æ 1.5% ( Æ 1.0) Æ 2.0% Æ 3.2% ( Æ 0.35) Æ 0.5%
( Æ 2.5) Æ 3.5% ( Æ 1.5 ) 1.5 ) Æ 3.5% ( Æ 2.0)1) Æ 4.0% Æ 6.4% ( Æ 1.4) Æ 2.0%
72
Acceptance Acceptance Tests with Centrifugal Centrifugal Pumps
Further to the ISO acceptance conditions, pumps used in the petrochem chemic ical al indu indust stry ry are are norm normal ally ly orde ordere red d to API API 610 610 (ISO (ISO 1370 13709) 9) and and othe otherr pumps may be ordered to the more generalized standard Hydraulic Institute.
API 610 (ISO 13709) at n N and Q N: Nominal differential head
Guarantee point
Shut off
0–152.4 m
+5% À2%
Æ 10%1)
152. 152.4– 4–30 304. 4.8 8m
+3% +3% À2%
Æ 8%
>304.8 m
Æ 2%
Æ 5%
hN
Not a rating value
P N
+4%
NPSHreq
+0
Hydraulic Institute Standards test code 1.6 and 2.6 Pumps must lie within the following tolerances for acceptance to level A: At rated head: +10% of nominal flow rate or at rated flow rate/rpm +5% of head <152.4 m +3% of head >152.4 m Conformity with one of the above tolerances is required. Test tolerances: In tests according to these rules the results must show no minus tolerances with regard to flow rate, head or nominal efficiency at duty point.
Admissible measuring uncertainty (instr.) fluctuation: The measur measuring ing uncerta uncertaint inties ies are includ included ed in the values quoted above.
Flow rate Head Suct Suctiion hea head Speed Power input
(Æ 1.5) Æ 2% (Æ 1.0) Æ 2% ( Æ 0.5) Æ 2% (Æ 0.3) Æ 0.3% (Æ 1.5) Æ 2%
1)
If a rising Q/H curve is specified, the minus tolerance given here is allowed only if the test curve displays a characteristic rising.
3.2.3 3.2.3 Test Test beds beds The pump must be set up on the test bed with proper suction and discharge discharge piping, so that it corresponds corresponds to the actual installation installation layout. Most importantly, care shall be taken to assure proper inlet conditions. Espe Especi ciall ally y for for ve vert rtica icall mixed mixed and and axial axial-fl -flow ow pump pumps s (pos (possi sibl bly y with with inle inlett bends) operating operating with high flow rates and low heads the inlet configuraconfiguration tion on the the test test bed bed must must be ma main inta tain ined ed to as assu sure re prop proper er inle inlett cond condit itio ions ns and when possible matched to actual service conditions as closely as possible possible ( Fig. Fig. 3.1 ). 3.1 ). Ideal Ideally ly the the flow flow rate rate thro throug ugh h the the press pressur ure e measu measurem remen entt cros crossssection should satisfy the following requirements: regular velocity velocity distributio distribution n in the axial direction; direction; 1. regular equalized in the measuring measuring plane; 2. static pressure equalized 3. the inflow must be free of vortices.
Pump Tests
73
Figure Figure 3.1 Measuring Measuring set-up set-up for determining determining the pressure pressure head head
Ofte Often n thes these e cond condit itio ions ns ca can n hard hardly ly be me met, t, but but the the impo import rtan antt thin thing g is to obtain a good velocity distribution at the pressure measuring point by suitable choice of pipe layout and if necessary by means of flow conditioners and straighteners. Gene General rally ly at leas leastt seven seven diam diamete eters rs of stra straig ight ht sucti suction on pipe pipe are are neede needed d after a 90 bend or other pipe fitting to obtain an acceptable velocity distribution at the pump’s inlet. Greater lengths should be used if a throttling valve is used in the suction pipe to calm the disturbance. To assu as sure re unifo uniform rm ve velo locit city y dist distrib ribut utio ion n at the the pres pressu sure re me measu asuri ring ng poin point, t, flow flow conditioners or straighteners are fitted at sufficient distance from the measuring point. The The test test stan stand d conf config igur urat atio ion n at the the suct suctio ion n end end depe depend nds s on the the natu nature re of the pump inlet (see Fig. 3.2 ). 3.2 ).
Figure Figure 3.2 Test stand stand config configuratio urations ns
74
Acceptance Acceptance Tests with Centrifugal Centrifugal Pumps
Figure 3.3
On pumps with an axial inlet it is necessary to suppress the prerotation set-up by the impeller by providing a second straightener for discharge discharge rates Q / Q00 < 0.5, 0.5, otherw otherwise ise the static static pressu pressure re measure measuremen mentt will be falsified. For example, depending on the connection between pressure gauge and pipe (water or air filled), z1 should be taken as shown in Fig. 3.3 or related to the pipe tap location when air filled. On the other hand, the suction heads may be plotted against the square of the discharge rate. The tangent to the curve obtained in this way, which when discharge rate Q = 0 passes through the geodetic sucti suction on head, head, dete determ rmin ines es the the sucti suction on head head incr increas ease e unde underr part part load load flow flow rates (see Fig. 3.4 ). 3.4 ). The installations for measuring the NPSH of a pump are shown in Table 3.1. 3.1. For For NPSH NPSH test tests, s, depe depend ndin ing g on the the infl inflow ow cond condit itio ions ns,, a boos booste terr pump pump may be required along with a throttling valve. Other test rigs are more
Figure Figure 3.4 Start of pre-ro pre-rotation tation
Pump Tests
75
Table 3.1 System NPSH range
Open systems
>2.5 m with suction throttle valve
Closed systems
2.5–8 m with controlled water level
Practically unlimited for closed loops
Installation principle
Flow rates
All flow rated
Low flow rates
Low, medium flow rates and heads
Control mode
Suction pressure throttling
Wate Waterr leve levell vari variat atio ionn
Syst System em pres pressu sure re vari variat atio ionn
Evaluation method at constant discharge rates. Q ( Q QG ), H 0 , n are n are the measured values
Q G = constant flow rate Q = Q = total flow rate H 0 = head first stage
costly to operate. Furthermore, for higher pressure pumps (stage heads >600 m) detailed NPSH values and also bubble patterns for the suction stage may be determined in model tests.
3.2.4 Conversio Conversion n of test results results All measured results determined at a speed n speed n and and with with test test flui fluid d devia deviati ting ng from the nominal speed n speed nN and fluid density r N must must be conv conver erte ted d to the latter. The admiss admissibl ible e speed speed deviati deviations ons ac acco cord rdin ing g to HI and and ISO ISO are are give given n in the table below.
Speed range Test values
HI
HI >225 kW
ISO 9906
Accordi rding to experience nce
Servic vice data efficiency ncy
80 to 120%
60 to 140%
50 to 120% +20%
Æ 20% Ä À50% Æ 20%
NPSH À3%
80 to 120%
60 to 140%
80 to 120%
Æ 20%
76
Acceptance Acceptance Tests with Centrifugal Centrifugal Pumps Converting the measured values (index u denotes converted values):
Q u = Q ( Q ( n nN /nT )
n N = nominal speed
n T = test speed
H u = H ( H ( n nN / n n T )2 P u = P ( P ( r n N / n n T )3 r N / r r T ) ( n hu
r N
= nominal density
r T
= test density
=h x
NPSHu = NPSH ( nN / nT ) ; values of exponent x between 1.3 and 2 have been observed. observed. The correction formulas allow for the following ratios: test and actual speed; test and actual fluid density; viscosities viscosities of the pumped pumped liquids; liquids; model and full-scale diameters. To effect efficiency correction formulas exist which relate to pump efficiency h, hydraulic hydraulic efficiency hh or hhR as defined in section 1.2. All correction methods are applied solely to the efficiency at best point: h00 = hhR  hv  hm h00 for full scale hT00 for test Table of common common formulas: formulas: .
.
.
.
Origin Sulzer Karassik
Formula 1Àh1 1Àh2 1Àh1 1Àh2
Notes 0:224
0:07
0:07
¼ Â Â ¼ Â Â D2 D1
h1 h2
N2 N1
N2 N1
0:17
n 1 n 2
n 1
0:07
Stage efficiency (only for efficiencies greater than 25%) Pump overall efficiency
n 2
3.2.5 Measuring Measuring instrume instrument nt uncertainty uncertainty The abstracted guidelines below for measuring uncertainties of particular values apply to acceptance tests class 1 (ISO 5198 Annex A).
Measuring the flow rate Q
fq
1. with tank and clocked measuring time
t ! 50 s (measuring the level difference Dz in z in m by siphon gauge when the filling jet is swung in and out)
Æ 0D: Z 3 %
Pump Tests 2.
with differential pressure instruments
77
Æ 1.0 Ä Æ 1.5%
(according to ISO 5167) 3.
magnetic, ultrasonic and mass flow meters
Æ 1.5%
(according to ISO 9104) Measuring Measuring the pressure pressure head p /( r r Á g )
f p
( p p / r ) is to be inserted below) r Á g 1.
with liquid columns between 0.1 and 1.5 m
Æ p=0ð:r 1: g % gÞ
and Dh fluctuations h fluctuations in m Æ 10À3 m, with wider Æ102 Á D h p=ðr Á gÞ fluctuations of Dh in h in m 2.
deadweight manometer
Æ 0.1%
3.
with calibrated spring manometer of
final value Æ0:6 read % -out value
accuracy class 0.6 4.
pressure transducer
Æ 0.1%
Measuring the velocity head c 2 /2gf /2gf e with circular cross-sections (exactly measurable)
1.5f q% Æ 1.5f
Measuring the speed n
fn
1.
with hand revolution counters
Æ 0.5%
2.
with electronic counters
Æ 0.1%
Measuring the power input P 1.
fp
power input to 3-phase motor: P N 25 kW
Æ 1.5%
at nominal power: 25 kW < P N 250 kW
Æ 1.0%
P N > 250 kW
Æ 0.8%
2. from torque and speed: a .
swivel motor
b.
torque meter a
! 0.75amax
0.5amax < a < 0.75amax
p ffiffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi
Æ 1:02 þ f n2
p ffiffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi q ffiffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi
Æ 1:22 þ f n2 Æ 1:52 þ f n n2
Determining the density r
f r r
of water at temperatures of up to 100 C
Æ 0.1%
The measuring uncertainties for uncertainties for the service data and efficiency are obtained from the individually measured values: 1.
for the flow rate Q :
f Q = f q
78 2.
Acceptance Acceptance Tests with Centrifugal Centrifugal Pumps for the head H : H :
f H
r ffiÀffiffi ffiffi ffiffi ffiffi Áffiffi ffiffiffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffiffi ffi ffi ffi ffi ffi ffi ¼ Á þ Á þ Á þ Á Z d À Z s 2 H
f 2z
Pd gÁH r Á g
2
2 f pd
Ps gÁH r Á g
2
2 f ps
ðCd ÀCs Þ4 ð gÁHÞ
f 2e
For pumps with high heads and low ps/pd values it may be assumed that:
f H ¼% f p 3.
for the efficiency: f h
q ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi q ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ffi ¼Æ þ þ ffiÆ þ þ f 2q
f 2H
f 2p
f 2q
f 2H
f 2p
If, for example, the head curve H ( Q ) of measured service data is converted converted to nominal nominal speed, the measuring measuring uncertainty is .
for the flow rate Qu: f Qu
.
for the head Hu:
q ffiffi ffi ffi ffi ffi ffi ffi ¼ þ q ffiffi ffi ffi ffi ffi ffi ffi ffi
f Hu ¼
f 2Q
f n2
f 2H þ 4f n2
3.2.6 Fulfillmen Fulfillmentt of guarantee guarantee 3.2.6.1 ALLOWING FOR THE MEASURING UNCERTAINTIES IN THE TEST EVALUATION In a diag diagra ram m ea each ch me meas asur urin ing g poin pointt appe appear ars s as an elli ellips pse e if the the unce uncert rtai ainnties ties for for both both coor coordi dina nate te va vari riab able les s are are take taken n into into ac acco coun unt. t. The The ma majo jorr axes axes of these ellipses emerge from the product of uncertainty and measured value for the variable in question, i.e.:
.
for the flow rate
f Qx Qx Á Q x
.
for the head
f Hx Hx Á H x
.
for the power input
f Px Px Á P x
.
for the efficiency
f hx Á hx
However, it is sufficiently accurate to replace the area of each measuring uncertainty ellipse by the four corners of its two axes. The characteristics H ( Q ) and h ( Q ) thus emerge as bands bounded by continuous upper and lower curves, which must be plotted, so that each of the two limiting limiting curves intersects intersects or touches at least one of the two axes at each measuring point (see Figs 3.5a and 3.5b ). 3.5b ).
3.2.6.2 FULFILLING FULFILLING GUARANTE GUARANTEES ES WITH RADIAL, RADIAL, MIXED AND AXIAL-FLOW PUMPS ACCORDING TO ISO The performance guarantee is fulfilled if the head curve passes through the tolerance cross representing the guaranteed requirements in accordance with section 3.2.2 (see Figs 3.6a and 3.6b ). 3.6b ). The head, efficiency and power power curves curves are to be develop developed ed using using a polyn polynomi omial al curve curve fitted fitted to the third or fourth order or a spline curve fitted for comparison to the guarantee limits.
Pump Tests
Figure Figure 3.5a 3.5a Allowin Allowing g for the meameasurement uncertainties in the H ( Q ) curve
79
Figu Figure re 3.5b 3.5b Allow Allowin ing g for for the the meameasurement surement uncertain uncertainties ties in the h ( Q ) curve
Figure Figure 3.6a Fulfillment Fulfillment of the guarante guarantees es for head, efficienc efficiency y and power
80
Acceptance Acceptance Tests with Centrifugal Centrifugal Pumps
Figure Figure 3.6b Fulfillment Fulfillment of the guarante guarantees es for head and power power only
3.2.7 Test beds beds and measuring measuring instrument instruments s In all Sulzer manufacturing plants there are pump test beds for acceptance tests and in-house development tests ( Fig. Fig. 3.7 ). 3.7 ). Their size and equipment are governed by the product ranges of the particular facility. Basic research and the development of new pumps including model development development are chiefly chiefly carried carried out in Winterthur. Winterthur.
3 . 3 A X I A L TH T H R U S T M E A S U R EM EM E N T S Here Here axial axial thru thrust st deno denotes tes the resid residua uall thru thrust st that that still still act acts s on the the thrust bearing after allowing for the partial compensation by the balancing ancing device device (usual (usually ly a piston piston)) (for (for theoreti theoretical cal consid considerat eration ions s see section 1.7).
Axial Thrust Measurements
81
Figure Figure 3.7 One of four test test beds at Sulzer Sulzer Pumps’ Pumps’ UK facility facility
Hydraulic axial thrust is the resultant of all forces acting on the rotor. These are mainly: the forces induced by the pressures on the suction and delivery-side shrouds of the impeller; the impulse force (deflection of the flow); the forces resulting from the pressures in front of and behind the shaft seal; the forces acting on the balancing device. .
.
.
.
3.3.1 Reasons Reasons for measuring measuring the axial axial thrust thrust .
.
.
For larger pumps (special versions) a guarantee is often given for the direction direction and maximum maximum amount amount of the axial thrust. The measurement must demonstrate fulfillment of this guarantee. It is usually performed on the test stand at the same time as the acceptance test for the Q / H charact characteri eristic stic.. If, however however,, the test test stand stand condit condition ions s deviate deviate too too much much from from the the act actual ual inst instal alla latio tion n with with rega regard rd to speed speed and and tem temper peratu ature re,, fiel field d measurement is possible. Usin Using g the the me meas asur ured ed va valu lues es,, the the ca calc lcul ulat ated ed va valu lues es ca can n be chec checke ked d and and the existing existing computer computer programs programs confirmed and expanded. expanded. If the the pump pump is equi equipp pped ed with with new new hydr hydrau aulilic c comp compon onen ents ts not not previ previou ousl sly y measured in full scale, or with new impeller configurations, the axial thrust is measured on the first-off pump.
82
Acceptance Acceptance Tests with Centrifugal Centrifugal Pumps
Figure Figure 3.8 Diaphragm Diaphragm as axial axial load load transduce transducerr
.
On all all mode modell pump pumps, s, the the axia axiall thru thrust st is as a firm firm pri princip nciple le me meas asur ured ed on the final variant variant intended intended for standardizati standardization on or full scale-up.
3.3.2 Axial Axial force transduc transducers ers employed employed When When selecti selecting ng axial axial force force transdu transducer cers, s, a fundam fundamenta entall distin distincti ction on is made between models and full-scale pumps. With a model pump the design can be adapted largely to the demands of the measuring technique used. On the other hand, on a full-scale pump more concessions must must be ma made de to the the desi design gn,, i.e. i.e. it must must be poss possib ible le to fit fit the the tran transd sduc ucer er in the bearing bearing housing housing with minimal alterations. alterations. In eith either er ca case se the the elem elemen entt fitt fitted ed must must be suff suffic icie ient ntly ly elas elasti tic c to prov provid ide e an adequately strong electrical signal via strain gauges. Figure Figu re 3.8 shows shows a diaphragm with integrate integrated d bearing bearing housing housing for measuring the axial thrust on model pumps. With this device the thrust can ca n be me meas asur ured ed in both both load load dire direct ctio ions ns.. By me mean ans s of suit suitab able le strai strain n gaug gauge e circuitry the influence of temperature can be eliminated almost entirely, giving a nearly constant sensitivity throughout the measuring range. Full Full-s -sca cale le pumps pumps are are usua usualllly y fitte fitted d with with annu annular lar load load trans transdu duce cers rs as in Fig. Fi g. 3. 3.9 9. This This type type me meas asur ures es only only comp compre ress ssiv ive e forc forces es.. The The ring rings s are are fitt fitted ed behind the supporting rings of the axial thrust bearing in the normal bearing housing (see Fig. 3.10 ). 3.10 ). If the thrust acts in one direction only, it will suffice to fit one annular load transducer.
Axial Thrust Measurements
83
Figure Figure 3.9 Annular Annular load transduc transducer er for fitting to a full-scale full-scale pump
The strain gauges are attached to the rings so that the following requirements requirements are met optimally: optimally: total load is indicated even if it acts extremely eccentrically; low temperature sensitivity; sensitivity variation within Æ 1% throughout the measuring range. .
.
.
Figure Figure 3.10 Annular Annular load transduce transducerr fitted in the axial thrust thrust bearing bearing
84
Acceptance Acceptance Tests with Centrifugal Centrifugal Pumps
Figure Figure 3.11 Oscillosco Oscilloscope pe display display of radial thrust thrust measurement measurement
3 . 4 R A D I A L T H R U S T M E A S U R EM EM E N T S (For theoretical considerations considerations see section section 1.8.) The radial thrust is measured by means of stay rings fitted between the bearing and its housing. Stay rings are fitted on the bearing near to the impeller for single-entry pumps and on both bearings for doubleentry pumps. The stays are sized to facilitate facilitate determination determination of deformadeformation tion by me mean ans s of of stra strain in gaug gauges. es. The The strain strain gaug gauges es on two two opp oppos osit ite e stays stays are connected to form a Wheatstone bridge, giving a high output signal and good temperature stability. The signals from the two stay pairs at righ rightt angl angles es to ea each ch othe otherr are are disp displa laye yed d on an X/Y X/Y osci oscill llos osco cope pe.. Figure 3.11 shows an example of the force sustained by a stay ring. The stay ring may be calibrated statically or dynamically. For static calibration the ring is loaded via the shaft in different directions. This method demands great expenditure of resources. The dynamic method is less expensive. In this case a known unbalance is used to generate a force that revolves with the shaft. By varying the unbalance or speed the force can be varied throughout the measuring range. The measuring rig is satisfactory if a circle appears in the oscilloscope.
3 . 5 E F F I C I E N CY C Y D E T E RM RM I N A T I O N B Y T H E THERMOMETRIC MEASUREMENTS Losses and a slight compression of the fluid cause a small increase in the temperature of the fluid pumped. If the pump head exceeds 100 to 200 m this this temperat temperature ure increas increase e becomes becomes suffici sufficient ently ly large large to allow allow
Efficiency Determination by the Thermometric Measurements
85
Figure Figure 3.12 3.12 Enthalpy Enthalpy diagram diagram
determination of the inner efficiency hi by very accurately measuring the temperature difference between inlet and outlet of the pump (DIN 1944). Figure 3.12 shows the pumping process in an enthalpy/entropy diagram. gram. Poin Pointt 1 repr represe esent nts s sucti suction on,, poin pointt 2 disch dischar arge ge cond condit ition ion.. The The enthalpy rise D his = gH from 1 to 3 is due to the compression of the fluid (ise (isen ntrop tropiic proc proces ess, s, no losse osses) s) whil while e the the incr increa ease se from from 3 to 2 is due due to the the inne innerr loss losses es of the the pump pump (the (these se incl includ ude e all all exce except pt the the me mech chan anic ical al loss losses es,, see section 1.2). The inner efficiency is given by: hi
¼
D his D heff
¼
D his D his
þ D hloss
ð1Þ
If p If p1, p2, T 1 and T 2 are measured, enthalpy difference in equation equation (1) can be determined from appropriate tables (water/steam table in the case of water). For For cold cold wate waterr the the phys physic ical al prop proper erti ties es ca can n be take taken n from from Fig Fig.. 3.1 3.13 3 and inserted into equation (2) (2):: hi
¼
1 2 ÀT 1 B þ C T P2 ÀP1
ð2Þ
The relation between pump efficiency and internal efficiency is given by the following equation (see section 1.2 for definitions of efficiencies): h
¼ hi Á hm
This This presu presupp ppos oses es that that the the bala balanc ncin ing g water water from from the the bala balanc ncin ing g device, disc or piston (if provided) is fed back to the suction branch.
3.5.1 Methods Methods of measuring measuring temperatur temperature e When measuring the water temperature in the inlet and delivery pipes, it is nece necess ssar ary y to ma make ke sure sure that that the the me meas asur ured ed va valu lues es repr repres esen entt the the ac actu tual al mean temperature in the measuring measuring cross-section. cross-section.
86
Acceptance Acceptance Tests with Centrifugal Centrifugal Pumps
Figure Figure 3.13 Physical Physical charac characteris teristics tics B and C for water
3.5.1.1 MEASURING MEASURING WITH SAMPLING SAMPLING PROBES AND QUARTZ QUARTZ THERMOMETER THERMOMETER Prob Probes es are are inser inserted ted at both both me measu asuri ring ng cros cross-s s-sect ectio ions ns and and a sm smal alll amount of liquid is bled off through them. This flow through the probe ensures that the sensor of the quartz thermometer is exposed to the actual temperature at the sampling point. By throttling the sampling rate to nil the total pressure can be determined at the same time. The best imme immerrsion sion depth epth for for the the pro probe is abou aboutt 30% 30% of the the pipe pipe radiu adius. s. Figure 3.14 shows a typical probe. The temperature difference can be read off directly from the quartz thermometer thermometer with a resolution resolution of 0.0001 0.0001 C. On hot water pumps the two sensors of the quartz thermometer are inserted inserted in welded-in welded-in temperature temperature sockets.
Efficiency Determination by the Thermometric Measurements
87
Figure Figure 3.14 Probe for for measuring measuring with quartz quartz thermometer thermometer
The probe method is also suitable for fast measurements in the plant provided the pipe cross-section is not too large and the temperature distribution in the measuring cross-sections is sufficiently uniform.
3.5.1.2 MEASURING MEASURING THE OUTSIDE OUTSIDE PIPE TEMPERA TEMPERATURE TURE A simpler but less accurate method is to measure the temperature on the outer wall of the suction suction and discharg discharge e pipe. The great great advantage advantage of this method is that no parts have to be welded into the pipe and that the measurement can be performed in a very short time. The procedure is also well suited to monitor monitor a decrease in pump efficiency efficiency during service (e.g. due to wear). By means of thermocouples wired in opposition, the thermoelectric voltage between delivery and suction is measured directly. Adequate insulation must be provided to ensure that the outside of the pipe wall assumes the temperature of the water. Figure 3.15 shows thermocouples in a differential circuit (thermopile) with three elements. elements. The approximate temperature rise (and thermoelectric voltages) for different efficiencies are plotted as a function of head in Fig. 3.16. 3.16. Power data are based on empirical values. It ma may, y, howe howeve ver, r, be agre agreed ed in the the supp supply ly cont contra ract ct that that the the se serv rvic ice e data data Q and H may be converted according to the above formulas, and the efficiency efficiency according according to a known correction correction formula formula for hydrodynamic hydrodynamic .
.
88
Acceptance Acceptance Tests with Centrifugal Centrifugal Pumps
Figure Figure 3.15 Thermopile Thermopile with with three three elements elements
machines, machines, even if the speed deviations deviations are greater than the standards allo allow. w. A few few comm common on upli uplift ft form formul ulas as are are show shown n in the the tabl table e on page page 76. 76. However, they hold good only if the following conditions are satisfied by both the model and the full-scale machine: – ReD > 3 Â 106 (related to impeller impeller outside outside diameter) diameter) – Impelle Impellerr diameter diameter of model model D2 ! 350 mm.
Figure Figure 3.16 Approximat Approximate e temperature temperature rise in a pump (water (water temperature temperature below 40 C)
