ML13330A612
| ML13330A612 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 04/30/1986 |
| From: | Rahmeyer W, Tullis J UTAH STATE UNIV., LOGAN, UT |
| To: | |
| Shared Package | |
| ML13324A845 | List: |
| References | |
| BPC-SCE-86-4178, TAC-62010, NUDOCS 8605070218 | |
| Download: ML13330A612 (60) | |
Text
Enclosure May 2, 1986 Log BPC/SCE-86-4178 FINAL REPORT Performance Tests on the Feed Water Check Valves for the San Onofre Unit I Nuclear Power Plant Phase A Submitted to:
Bechtel Power Corporation Norwalk, California 90650 and Southern California Edison By:
J.
Paul Tullis and William J. Rahmeyer Utah State University Foundation Logan, Utah 84322-9300 April 1986 8605070218 860505 PDR ADOCK 05000206 P
ABSTRACT Flow tests were conducted 3t the Utah Water Research Labora tory, Utah State University to evaluate the stability of the 10 and 12" Atwood & Morrill swing check valves to be installed in the feedwater system for Unit I at the San Onofre power plant.
The valve tested was a 10" Atwood & Morrill swing check valve that is scheduled to be installed in the plant.
The valve was tested in three separate piping configurations to simulate:
- 1) the 10" check valve to be installed inside containment; 2) the 10" check valve to be installed outside containment, downstream from the flow control valve and 3) the 12" check valve to be installed downstream from the main faedwater pumps.
Tests were conducted with cold water at low pressure with the data adjusted to plant operating conditions by properly scaling the density of the fluids.
Tests on the valve to be installed inside the containment showed that light tapping started at 2700 gpm (all flows are adjusted to plant operating conditions).
The tapping stopped at 3050 gpm.
Between these two flow rates, the tapping had a light metalic sound which could be heard by amplifying the output of the accelerometer on a speaker, using a stethoscope or placing the ear in direct contact with the valve body. The tapping could not be distinguished by someone merely standing close to the valve.
Based on tests on numerous other valves, it is our conclusion that this valve is as good as any valve that has ever been tested by us.
The tapping of the disc against the valve body is not hard enough to cause any noticeable deformation of 1
the body or disc.
The only wear expected would be the very slow wear on the hinge pin caused by the slight rotation of the disc when it is operating in the tapping range.
At low discharges between 40 and 150 gpm, which simulate auxiliary feedwater pump flows, the valve was totally stable.
No tapping or any movement of the disc could be determined by ear or by the accelerometer.
The pressure transducer installed down stream showed that there was virtually no pressure fluctuations being generated by the valve.
It is our opinion that the valve can operate indefinitely at these low flow rates with no detrimental effects to the valve or the piping system.
For the 10" Atwood & Morrill swing check valve installed five diameters downstream from the flow control valve, tapping started at 2600 gpm and was almost stopped at 3900 gpm.
The tapping was slightly more intense and at a somewhat higher frequency than for the inside containment tests.
The slight increase in the tapping frequency and intensity is caused by the flow disturbances generated by the flow control valve.
The tapping was not severe enough to cause damage to the valve body or the disc, but slightly more rapid wear would be expected on the hinge pin.
We would however, expect years of trouble free operation.
The tests made on the 10" valve operating as a model of the 12" valve installed downstream from the injection pumps started tapping at 3650 gpm and was essentially stopped at 5700 gpm.
The tapping conditions were slightly more intense than the inside the containment tests where the flow was totally uniform, but not as 2
intense as the tests downstream from the control valve.
This is what was anticipated prior to the tests and it was our opinion that the turbulence and non-uniformity of flow generated by the pump would not be as severe as that generated by the flow control valve.
The 12" check valve should operate trouble free for many years.
INTRODUCTION Failure of three 10" and two 12" Pacific swing check valves at the San Onofre, Unit I Nuclaar Power Plant generated a severe water hammer on November 21, 1985.
The five damaged Pacific valves plus three additional redundant valves to be placed inside a containment are being replaced with Atwood & Morrill swing check valves.
Furthermore, three 4" Pacific swing check valves in the auxiliary feedwater system are being replaced with Atwood
& Morrill valves (these 4" valves are not being tested).
Before returning the unit to service, it was felt prudent to conduct laboratory tests on the new check valve to determine its stability in the various flow environments where it will be operating.
Tests were conducted at the Utah Water Research Laboratory on one 10" Atwood & Morrill swing check valve (plant tag no.
345) that is scheduled to be installed in Unit I. Tests were con ducted to simulate operation of the 10" check valve inside containment with special emphasis on tests at low flow rates produced by the auxiliary feedwater pump.
For these tests, the 3
check valve was installed with straight pipe to provide uniform flow upstream and downstream.
A second series of tests were conducted on a 10" check valve to simulate its operation outside containment with the valve located five diameters downstream from the flow control valve.
A third set of tests were performed using the 10" check valve as a model of the 12" check valve installed just downstream from the feedwater pumps.
A piping configuration was assembled which simulated the general flow pattern downstream from the main feedwater pump.
The actual pump was not modeled, but was simulated by a piping configuration upstream that was judged to produce a condition equal to or worse than the pump.
Since it was not possible to see inside the test valve or to O
perform any wear tests for this phase of the tests, acceptance criteria was based on a qualitative evaluation of the performance of the valve.
It generally depended upon an engineering judgement as to whether the tapping or any other noises generated by the valve constitute a situation which would result in rapid wear of the valve. A second phase of testing is planned on a new valve currently under construction, which will allow visualiza tion of the stability of the disc and quantitative wear tests.
MODELING CRITERIA For laboratory tests to properly simulate in-plant opera tion requires geometric and dynamic similitude between the two operating conditions. For the two 10" valve tests inside and outside the containment, one of the actual valves to be installed 4
in the plant was furnished for the testing. Therefore, exact geometric similitude was achieved. The 10" valve is not an exact model of the 12" valve, but representatives from Atwood & Morrill indicated that the sizes of the internal passages and the weight of the disc are nearly proportional and should be very similar.
It was therefore considered that geometric similitude was achieved for these tests.
Dynamic similitude requires that the forces causing movement of the disc be properly modeled.
The forces causing the disc to open are pressure and inertia forces.
The ratio of the pressure forces to the inertia forces produces a dimensionless similarity parameter referred to as the Euler number defined as:
AP Eu = pV(1) in which AP is the pressure drop across the valve, P is the mass density of the fluid and V is the average pipe velocity.
Dynamic similitude is achieved if the test Euler number (Eut) is equal to the Euler number in the plant (Eup).
The ratio of the plant Euler number to the test Euler number must be 1.0 since they must be equal.
This produces:
p Apt p
y2 = 1.0 (2) pt Vt2 This equation (2) provides the means of scaling the test data to plant operating conditions. During a test, one measures the pressure drop corresponding to a given velocity. Knowing the
density of the test fluid and the plant fluid, one then desires to know what velocity or discharge in the plant would produce the same pressure drop and therefore the same Euler number. For this condition, equation (2) simplifies to:
V P p
Pt (3)
Vt t'p This shows that the flow rate in the plant is increased by the square root of the density ratio between the two fluids.
For the scaling of the test data to plant operation for the 10" valve inside a containment, equation (3) is the only criteria to be considered.
This is because both the plant flow and the laboratory flow'used to simulate the plant operation provide essentially ideal flow approaching the check valve.
For the 10" check valve outside containment, located downstream from the flow control valve and the 12" check valve located downstream from the pumps, other considerations are necessary.
For the 10" check valve outside containment, a valve identical to the in-plant control valves was provided for the tests. The check valve was located five diameters downstream from the flow control valve so that its placement and the piping in the near vicinity was essentially identical to the in-plant installation. The only difference between in-plant operation and the tests conditions (except for the temperature difference) is that the tests were run at lower pressures and lower pressure drop across the control valve.
The test flow control valve was
- operated at pressure drops up to approximately 60 psi.
The plant 6
valve operates at pressure drops normally above 300 psi.
The differences between the influence of the resulting flow down stream of the flow control valve on the operation of the check valve is not considered significant. This is primarily because the high velocity jets discharging from the v-ports of the double ported globe valve are dissipated inside the valve body. The turbulence intensity inside the valve body at the higher pressure drops would be greater but the general flow pattern exiting the valve would be expected to be similar.
The small high velocity Jets would create high frequency, intense turbulence fluctuations inside the valve.
These would dissipate rapidly with distance.
It is therefore felt that the test valve operating at the same flow rate but at a lower pressure drop will produce flow condi tions at the check valve very similar to those which would be present in tha plant.
One other factor that must be considered in making this evaluation is the intensity of cavitation in the plant and in the test valve.
Cavitation conditions can be quantified with a similarity parameter referred to as the cavitation index.
It is defined as:
Pd + Pb Pv u
Pd (4) in which Pu is the pressure upstream from the control valve, Pd is the pressure downstream, Pb is the barometric pressure and Pv is the absolute vapor pressure of the fluid at operating 7
temperature. For the 10" valves at normal plant operation, the value of a is:
a = 650 + 14.3 -
247
= 1.34 300 To determine if the valve is cavitating, one must compare this value of a to experimental data that identifies when the valve starts to cavitate.
No data are presently available on the cavitation performance of this particular valve.
However, the authors have cavitation data on other globe valves.
For the test control valve to properly model or cause slightly more disturbance than the plant valve, the cavitation level during the tests was kept within the light to moderate cavitation range. This limited the pressure drop across the test valve to 60 psi and less.
During the tests, the intensity of cavitation at the control valve was always slightly worse than expected in the plant valve.
In summary, scaling the test data on the 10" valve to the two in-plant 10" valves, merely requires that the test discharge or velocity be multiplied by the square root of the density ratio. Since the test data was taken at 42 0 F and the plant operates at 400 0, the flow was increased by 1.079.
For the 12" check valve, the flow is corrected by the area ratio times the square root of the density ratio.
The plant temperature at the pump discharge is 330 0 F.
This means that the test discharge must
- b 8
be multiplied by:
.706
- 62.43.5 1
I
____ I 1.502
.495 56.27 in order to correct it to plant operating conditions.
TEST INSTALLATIONS Piping For 10" Valve Inside Containment Figure 1. shows photographs of the piping installation for the 10" check valve simulating operation inside containment.
Figure
- 2.
shows a line drawing with dimensions.
The long straight approach of 12 and 10" pipe upstream and downstream provided near ideal flow conditions for the check valve. With O
the valve located inside containment, having at least 10 pipe diameters of straight pipe upstream, and approximately 5 pipe diameters downstream, it was felt that this piping configuration would accurately simulate the in-plant operating conditions.
A downstream valve was used to control the test flow rate.
For the auxiliary feedwater flow rates between 40 and 150 gpm, flow was supplied by a low head reservoir.
This provided flow which was essentially free of flow disturbances and allowed an evaluation of any disturbances or pressure fluctuations that were induced by the check valve.
Tests at higher flow rates were
-done using a 100 horsepower vertical turbine pump.
The pump was required to allow the system to be operated at higher pressures so that cavitation at the check valve was suppressed.
Even with the pump operation, light cavitation at the check valve occurred at higher flow rates.
9
0r FIGURE 1. Photographs of Test Piping for
.10" Valve Inside Containment
FIGURE 1. (Cont.)
Photographs of Test Piping for 10" Valve Inside Containment
FIGURE 1. (Cont.)
Photographs of Test Piping for 10" Valve Inside Containment
8x 12" Pu d
Orifice 18 12" 12" O
O 2
2-7'6"1 3'
20'6" 41" 13, 16" dia. Pipe To Weigh and Volumetric Tanks FIGURE 2. Line Drawing for 10" Valve Inside Containment
Piping for 10" Check Valve Outside Containment Figure 3.
is a schematic drawing showing the location of the 10" and 12" check valves in the feedwater system.
Figure 4.
shows the valve layout.
The 10" check valves are labeled number
- 393, 346 and 345.
The 12" check valves are 433 and 439.
The test piping was reproduced from the two elbows upstream of the control valve to the straight piping downstream from the check valve. The block valves were not included because they are full ported gate valves and their influence on the operation of the check valve would be negligable. A control valve with the valve body and internals identical to the valves in the plant was provided for the testing.
It was installed five diameters upstream from the check valve as shown in Figure 3.
Flow was supplied for these tests with a 200 horsepower vertcal turbine pu;np.
This produced pressure drops up to 60 psi across the control valve and adequate pressure at the check valve to limit the cavitation level to light cavitation. A downstream control valve was used in conjunction with a pump bypass valve to control the flow and pressure in the system.
The control valve was always operated fully open. The control valve operated from light to moderate cavitation during.the testing.
Piping for 12" Check Valve at Feed Water Pump The two 12" check valves are in different orientations downstream from the two feed water pumps.
It was considered that the most severe operating condition would be for the valve installed with an elbow between the pump discharge and the check valve (Pump B). This is because testing performed for previous 14
91"
-0" Control Check Valve Valve a
4 2"
I ~
~
3 49 s
8 o
o s
P-P, II2" 51 I
16"dia. Pipe 6
To Weigh and Volumetric Tanks FIGURE 3. Line Drawing for 10" Valve Outside Containment
Main Feedwater System San Onofre Unit 1 E
MFW PUMP l
From To S Condensate From SatetyIneto HV' 852B IA SG MOV FCV FWS FWS H
22 458 398 396 FWS HV 456 HV 438 852A 654A Ist PT HP HTR FE Pl E.
MrW
~457 CV-144 FWS FWS 11,T O
TB Sydsate Injection MOV FCV FWS FWS System20 457 346 342 pi H
F FS HV 4%CV-142 FWS FWSToS HV 439 852B37 37 8!>4B PT U
MOV FCV FWS FWS From Auxiliary 21 456 345 343 Feedwater System FIGURE 4. Schematic of Main Feedwater System
clients has shown that elbows can induce significant swirl which can cause instability of check valves. Photographs in Figure 5.
and the line drawing in Figure 6. show the configuration of the testing.
A pump of the right type and capacity was not available at the laboratory. It was therefore necessary to simulate the pump conditions with piping available at the laboratory.
Our experience with flow conditions at the discharge of a horizontal centrifical pump indicates that one expects a reasonably uniform velocity distribution with little swirl but a rather high intensity of turbulence. With this in.mind, the piping configuration shown in Figures 5.
and 6. was selected to simulate pump operation.
Note that the piping from the discharge O
tee and elbow to the elbow and 10 X 8 reducer upstream from the check valve are geometrically similar to the plant. The test dimensions were scaled by the ratio of the inside diameter of the 10" pipe to the inside diameter of the 12" pipe.
Flow from the pump was simulated by the 10 X 8 reducer and 8" elbow.
To simulate turbulence generated by the impeller of the pump, a multi-hole orifice plate was installed just upstream from the 10 X 3 reducer. The plate (Figure 5.) had thirty-one inch diameter holes. It was felt that the turbulence produced by these sub merged jets would be greater than the turbulence from the pump discharge. The orifice plate also served to reduce any swirl generated by the three elbows upstream.
To determine the sensitivity of the check valve to the flow conditions from our piping simulating the pump, the multi-hole 17
FIGURE 5. Photographs of Test Piping for 12" Valve Tests
FIGURE 5. (Cont.)
Photographs of Test Piping for 12" Valve Tests
LL FIGURE 5. (Cont.)
Photographs -of Test Piping for 12" Valve Tests
1110 psee g@ro 10 Orif ice 1727 111 8
9 SE0.
16 dia. Pipe FIGURE 6. Line Drawing for 12" Valve
orifice plate was removed and the test repeated.
This configura tion would have produced flow that had significant swirl because of the four elbows and the tee from our main pipe distribution system.
INSTRUMENTATION Pressures upstream and downstream from the check valve were measured with a precision Heise dial gauye.
The dial gauge was calibrated at the laboratory with a dead weight tester just prior to the tests.
The calibration data are contained in Appendix 1.
Flow rate was measured either volunetrically, by weight, or with a calibrated orifice.
The weight and volumetric tanks had been previously calibrated with weights traceable to the National Bureau of Standards. The calibration information is contained in
.*Appendix 1.
An 8" X 12" orifice was installed upstream for the tests on the 10" check valve simulating the inside containment operation. This orifice was calibrated in place during the tests with the weight tanks. The orifice was used to measure flow rates at some higher discharges.
All flow rates for the second and third piping installations were measured volumetrically.
The temperature of the water was measured with a mercury thermometer.
A Valedyne pressure transducer was used to measure pressure fluctuations downstream from the check valve.
The transducer had a 20 psi diaphragm which was calibrated with a dead weight tester just prior to the testing program.
The output from -the transducer could be read on a digital indicator and was recorded 2Z
on a strip chart recorder.
This-allowed scaling of the pressure fluctuations from the strip chart recorder or noting them from the digital indicator.
The calibration of the pressure transducer is also contained in Appendix 1.
An accelerometer was attached to the body of the valve in the area where the hammer on the disc arm would contact the valve body. The output from the accelerometer was monitored by a volt meter.
It was also amplified with a speaker so that the tapping noise and the flow noise could be easily detected.
In addition, a stethoscope was used to listen for flow noise and tapping
-sounds.
Audio tape and video tape was made of most of the test runs.
TEST RESULTS 10" Valve Inside Containment Tests Runs 1-31 in Table 1. contain the tests on the 10" check valve installed with straight piping upstream and downstream to simulate the flow inside the containment.
Runs 1-15 were preliminary runs and were not witnessed.
- Table
- 2.
lists the equations used in the spread sheet program to process the data for Table 1.
The data were checked manually to verify the accuracy of the equations and the computations.
At low flows simulating auxiliary pump flows, no noise of any kind could be heard coming from the valve.
Never once did the disc tap against the seat. Based on an audible evaluation, we conclude that the valve can operate in this flow rate 23
To
- 1.
FLOW CALIBRATION OF 10* ATWOOD-MORRILL SWING CHECK VALVE FOR CONTAINMENT A
8 C
D E
F 6
11 J
K L
M N
0 P
q 1
1 I
I TEST PLA NT I
TS
-i T EST R
UN WEIGHT TIME ORIF CONY FLOW ORIF Pu MANO.
COI P
C CHART PK PRESSI OBSERATIONS 3NO.
lb m e reafirgj to psi 1jpn CO F~~
i ISAEFUTP1 5 ~ ~ ~
~
~
~
~~~.
N. b se redn topi pmICEF qpm 1 IDIN !fp pai ireadingt alrpi1 a
28 1986 I
1 L 1..!
j _
I 5
1 47640 157.5 15.85 0.179 2185 I 1297 2357 19.5621 9.76 813 1 650 0.179 1.16 6.97 2025 l.130 0.14 NO NOISE S2 3540 171.6 110.70 0.01066 1406 1294 1516 I9.562i 6.28 10.10070.5010.01066 0.75 1 9.35 I 1621 i.230j 0.24 NO NOISE 7
3 20000 116.7 185.00 0.01066 1233 12951 1330 9.56
.1 9.72 i 1553
.20 0 0.21 NO NOISE I
8 4
47920 145.8 118.70 0.179 2363 1292 2549 9.562 10.56 8.10 1 7.45 1_0.179 ' 1.33 1 6.77 2047
.130 0.14 1
NO NOISE I
9 27 2759 i 1 2 976 9.5624 12.33 6.10 1 9.80 i 0.179 1 1.75 1
4.35 0
.250 0.26 i LIGHT TAPPING 10 6
29.30 0.179 2963 1 1294 3197 19.562 113.24 5.25 111.351 0.179 I 2.03. 3.22 12079
.150 0.16 1 LIGHT TAPPING I 11 1 7 23.00 0.179 26267, 1294 2832 19.562 11.73 029.80 8.801 0179 1 1.58 i2822 20921.50 10.37 I LIGHTTAPPING 12 8
18.601 0.179 2361 1294 2547
.1 10.55 33.60 1
7.35 0.179 1 1.32 f 32.28 2058
.450 0.47 NO TAPPING 3
9 20.80 0.179 2497 1294 2693 9.562 11.16 31.851 8.00 1 0.179 43 130.42! 2087
.350 0.37 START TAPPING 14 10 26.501 0.179 2818 1294 3040 9.562 12.59 i265011.25 0.179 1.83 12 4.67 2081
.400 0.42 STOPTAPPING 15 11 42.501 0.179 3569 129g4 3850 9.562' 15.95 2.5155079 2.96 19.69127
.380 04 OAPN 16
- 12.
52.20 0.179 3955 1294 4267 19.5621 17.67 19.20 20.00 0.179 1 3.58 115.62 2091 1.500 0.53 I
NO TAPPINGI 17 MAR 31 1986 1
18 13 9700 272.3 3.70 0.01066 256 1289 276 9.562 1.14
- 1.
24.10 012 0.24 505
.000 0.00 NO TAPPING 19 14 11000 219.3 7.05 0.01066 361 11315 389 9.562 1.61 10.4027.400.01066 0.29667
.000 000I NOTAPPING 20 15 14000 208.1 112.90 0.01066 484 13041 522 19.5621 2.16 10.20 31.30 0.01066j 0.33 9.87 837
.000 i 0.00 21 22 16 2000 357.7 0.01066 40 43 9.562 0.18 110.95i14.70 0.010661 0.16 110.79 102 iMETER 0.03 NO NOISE
!INCREASE Q 23 17 1900 223.91 0.01066 61 1 66 9.5S62F0.27110.85192510.0O66 0.2t 10.64135IMETERI 0.031 NONOISE INREASEQ 24 18 2000 131.3 0.01066 109 118 9.5 62 0.49 110.88120.25 10.01066 0.22 10.661 236 !METER!
0.00 i
NO NOISE INCREASE Q 25 19 5500 232.4 1.50 0.01066 170 1345 184 I9.562 I 0.76 11080i21.60 0.01066 0.23 i10.571 355 IMETERi 0.03 NO NOISE INCREASE Q 26 20 9500 172.1 8.85 0.01066 397 1292 428 09.562 177i6626I0.01661 0.30 10.361 7191.METER 00 NO NOISE INCREASE Q 27 21 9000 203.6 5.60 0.01066 318 1300 343 9.562 1.42 10.71125.2510.01066 0.27 10.44 612 1METERI 0.03 1
NO NOISE DECREASEQ 28 22 1900 254.1 10.01066 54 58 9.562 0.24 i110.82114.850.01066i 0.16 10.66 135 IMETERi 0.03 NO NOISE DECREASE Q 29 23 17000 182.1 25.00 0.01066 671 1300 724 9.562 1
0 1.6 1058 1200 021 NONDISE INCREASE0 30 24 17000 94.75 92.20 0.01066 1290 1301 1391 9.562 1 5.76 i 9.80 162.001 0.01066 1 0.66 9.14 1587 1.300 0.32 1
NO NOISE ilNCREASE Q 31 25120000 99.22 117.10 0.01066 1449 1297 1563 9.562 6.47 19.48 172.4010.0106 0.77 8.71 1649 1.380 0.40 1
NONOISE 1INCREASEQ 32 25 20000 99.22 6.90 0.179 1449 11304 1563 19.562 6.47 9.48 172.4010.01066 0.77 8.71 11649
.380 0.40 1
NO NOISE NINCREASE Q 33 26 13.00'0.179 1300 2139 9.562 8.86 8.26 198.60 10.010661 1.05 7.21 1934 I.250 0.26 1 LIGHT CAY.
IINCREASE Q 34 27 19.80 0.179 2447 1300 2640 9.562 10.94 30.61 7.8 5 1 0. 179 1.41 29.20 2065
.350 NDNOISE NCREASEQ 35 28 21.45 0.179 2547 1300 2748 9.562 11.38 29.101 8.40 i 0.179 1 1.50 27.60 2077 1.400 0.42 STARTTAPPING 4INCREASEQ 36 29 26.70 0.179 2842 1300 3066 9.562 12.70 24.31 10.401 0.179 11.86 22.45 2083
.350 0.00 iSTOPPED TAPPINGiNCREASEQ 37330 0179 3359 1300 3
9562 15.01 20.6914.35 0.179
_ 2.57 118.12 2096 30 NO N 38 131 47.601 0.179 3795 1300 4093 9.562 16.95 116.05. 18.30
- 0. 179 I3.28 j1i2.77 2097 1.600 0.00 INO NOIS ICAIATN 591 T _I m _42deqF I
II I
TABLE 2.
FLOW CALIBRATION OF 10" ATWOOD-MORRILL SWING CHECK VALVE FOR CONTAINMENT A
C 0
E F
H IK TEST
- !PLANT I TEST 2
RUN [WEIGHT TIME ORIF CONW FLOW OI LWPP VLP 3 NO.
lb isec -rednit psi qpm iCOEF Ppmn IDN i_
4 MAR 28 1986 5
1 47840 157.5 15.85 179
=B5/CS/62.4*448.8 F0.5 lF5*(SQRT(62.43/53.65) 962
=F5/(448.8*0.7854*(I5/1 2)'2)
F8.1 3 6
5 L 25.4 0.179
- I294S6E6)^.5
=f6/((E6*D6
.)
I=F6*(SQRT(62.43/S3.65))
9.562 F6/(448.8*0.7 54*(I6/12)'2) 6.1 7 MARM3 1986 I
9700 272.29 3.7 0.01066 *B8/C8/62.4*448.8 E8*D8
=05
=F6*(SQR62.43/53.65 9.562
=F8/ 446.8*0.7854*(18/12)2) 10.5 9126 13 0.179
-1300*(D9*E9)-0.5 F9/( E9'D9 '0.5)
F9*(SQRT(62.43/53.65))
9.562 I-F9/ 448.8*O.7854*(I9/t2)'2 8.26 FLOW CALIBRATION OF 10- ATWOOD-MORRILL SWING CHECK VALVE FOR CONTAINMENT L
H 0
P g
R S
T I
ITEST 2
MANO.
CN.
DP(gross Pd Cv CHART PK PRESS OBSERYATIONS 5 readno to pst psi p
SCALE FLUCTPSI 1 4
5 6.5 0.179
=L5*M5 wK5-N5
- F5/SQRT(N5) 0.13
=1.05*5 INO NOISE 6
9.0 0.179
=L6#M6
-K6-N6
[F6/SQRT(N6) 0.25
=-1.05*Q6 ILIGHT TAPPING 8
24.1 0.01066 *L8*M6
- K6-N8 F85TN o
s*ae NOTAPPING 9
98.6 0.01066 *L9*M9
-K9-N9
- F9/SQRT(N9)
!0.25 I-.Io5*Q9 ILIGHT CAY.
IINCREASE Q
indefinitely with essentially no wear and no detrimental effects to the system or the valve.
The pressure fluctuations generated by the check valve were below the measuring capability of the pressure transducer.
Runs 16-22 show that the maximum peak pressure fluctuations were on the order of 0.03 pis.
The pressure transducer was a 20 psi diaphragm and could accurately measure to approximately
.1 psi.
This documents the fact that there is no undesirable flow distur bances generated by the check valve at low flow rates.
As the flow was increased, light tapping started at a plant flow of 2740 gpm. The tapping stopped at a plant flow rate of 3066 gpm. The tapping noise in this range had a metalic sound and would be termed "very light".
It is our opinion that even for extended periods of time, no damage would occur to the valve body or the hammer on the valve disc do to this impact.
The only concern would be a slight amount of wear damage on the pin and the bushings after many years of operation.
Based on our experience of testing numerous other check valves for various manufacturers, this valve was as quiet as any valve we have ever tested. Based on the test results to date, it is our opinion that the new 10" check valves to be installed inside. containment will provide trouble free operation for many years.
10" Valve Outside Containment Tests Table 3.
shows the data for five test runs for the 10" check valve installed outside containment downstream from the control valve. Data were taken only at higher flow rates.
Below a plant 26
FLOW CALIBRATION OF ATWOOD-MORRILL SWING CHECK VALVE OUTSIDE CONTAINMENT A
8
-C D
F F
H 1
J K
L M
N 0
P_
q R
s CTRL.
CTRL.
TEST CTRL PLANT TESTI DP I
CAVITATION TESL
.UN VOL TIME ER.
VALVE VALVE FLOW VALVE FLOW PIPE VEL Pu VMANO.
C o
Pd SIGMA LEVEL AT Cv OBSERVATIONS 3
NO. cu-ft sec VOLTS.Pu p1 Pd I _g COEF VgmID1N fps p
irq topsI psi p
YALVE 4 APR 1
1986 1
T5 32 840 161.13 4.73 50.60 24.8 2423 477 2614 9.562 10.83 24. 80
.50 0.179 23.28 1.43 NOCAY.
1965 START TAPPING 6
33 e40 129.43 5.57 70.80 30.05 3D35 475 327419.562 13.56 430.05 3
0.1t79 2.47 2G.531 CA 1931 LIGHT TAPPING AT ABOUT 2 HZ 7
34 840- 117.79 6.16 68.00 18.5 3349 476 3613 9.562 14.97 1 1850i6 0.1796. 2.
I5530.
LIGHTCAY.
1943 1 TAPPING 8 35840 109.28 6.58 66.10 8.4 3621 477 3906 9.62 16.18
.1.79 31 5.251 0.36 MOD.CAY.
2040 INFREQUENT TAPPING ALM5T SEATED 9
__I t
I I
SIGMA 1 1.34 IN PLANT FLOW CALIBRATION OF ATWOOD-MORRILL SWING CHECK VALVE OUTSIDE CONTAINMENT A
8 C
D I
F G
I I
CRL._CTRL.
TEST ICTRL IPLANT 2
RUN IVOL TIME YER.
YALYE VALVE FLOW EYALY IFLOW 3
ND.
cu-ft sec YOLTS Pu.psi Pd~psi Igpm COEF
- qpm 4 APR I
1986 i
5 32 840 161.13 43450.6 4*D)/C
)0.5 J-G5*(SQRT( 62.43/53.65))
6 33 840 129.43 5.57 170.8 130.05
=448.8*(86+6.34*D6)/C6 I=G6/(E6-F6)'0.5 IG6*(SQRT 62.43/53.65))
7 34 840 117.79 68 18.5 488B6 7_)
GSQRT62.43/53.65))
8 35 840 109.28 6.58 66.1 8.4 1=448.8*(B8+6.34*D8)/C8 I=G8/(E8-F8)'0.5
=G8*(sqRT 62.43/53.65))
FLOW CALIBRATION OF ATWOOD-MORRILL SWING CHECK VALVE OUTSIDE CONTAINMENT J
K L
m N
0 P
q 1
ITEST DP!TEST 2
PIPE VEL Pu MANO.
CONY.
qross Pd
_CYOBSERVATIONS 3 IDN fps pl Treading to pa psipal 5
9.562
-G5/448.8*0.7854*J5/12)'2)
-S 8.5 10.179 jMS*NS
-L5-05
=G56/SRT(OS)
START TAPPING 6
9.562
=06/ 440.8.7854 J612:L
- 2)
F6 13.8 0.179 jM6*N6
-L6-06
=G6/SQRT(06)
TAPPING 7 9.562
=G7/(448.8*0.7854.J7/12)zy
- F7 16.6 10.179
=M7N
=L7-07
=G7/SQRT(07)
ITAPPING a
9.562
=08/ 448.8*0.7854' J8/12 2)
-Fe 17.6 0.179
-M8*N8 I-L8-08
=G8/50RT(08)
ILESS TAPPING
flow rate of 2300 gpm, no noise was heard from the valve.
Tapping started at a flow rate of 2614 gpm and was essentially stopped at 3906 gpm. The valve never totally seated at higher flow rates but the tapping reduced in frequency so that there were several seconds between the light taps.
Between 2600 and 3900 gpm, the tapping was slightly more severe and more rapid than for the tests with uniform flow upstream.
As expected, the turbulence generated by the control valve increased the severity of tapping slightly.
The 5 pipe diameters between the valves seems adequate.
Again, based on our experience with tests on other check valves previously, it is our opinion that the tapping noises generated by the valve in this disturbed flow condition are not excessive and the valve should have relatively little wear and operate for years without serious maintenance problems.
Table 3. also lists the equations included in the spread sheet program for processing the data.
12" valve Downstream of Pump B Table 4.
lists two sets of tests for the 12" check valve installed downstream from piping simulating the flow configura tion for feedwater pump B.
Tables 5. and 6. list the equations used in the spread sheet program for processing the data in Table
- 4.
Runs 36-40 were made without an orifice installed upstream.
Runs 41-44 were for the multi-holed orifice plate installed to simulate pump turbulence.
The noises generated by the valve were almost identical for the two installations.
If
- anything, there 28
TABLE 4.
FLOW CALIBRATION OF ATWOOD-MORRILL SWING CHECK VALVE OUTSIDE CONTAINMENT A
B Ca E
F 6
I d
K L
m N
ORIF.
TEST ORIF.
TEST PLANT ! TEST I PLANT 2
RUN4 VOL TIME VER. YorN? FLOW YEL.
PIPE PIPE IVEL I
FLOW Pi Pd_
OBSERVATIONS 3
NO. Icu-ftI sec IYOLTS qpm fps IDIn IDi fp sq psi III OBRATIPONS 4 APR 3
1986 1
1 5
316 1 840 1 160.47 4.66 no 12432 1 10.9 19.562 11.374 10.9 3653 26.8 23.401 START OF LIGHT INFREQUENT TAPPING 6
371 040 1 110.85 5.49 no 3542 15.6 9.562 11.3741 15.8 5321 I 9.91 I 2.80 LIGHT INFREQUENT TAPPING 7
38 840 I 103.33 6.4 no 3825 17.1 9.5621 11374 1711 5.60 ALMOST NO TAPPING, ABOUT I IN 10 SEC.
8 39 840 1 119.37 5.13 o
2801 14.6 19.56211.37414 j
122.001 15.80 LIGHT TAPPING ABOUT 2 HZ.
9 40 I40 I 144.4 1 4.6 -
no 2701 I 12.1 19.562 11.374j 12.1 i 4058 128.83124.501 LIGHT TAPPING ABOUT 3-4 HZ. MAX. TAPPING Fzt-lI T-43.7 DEG F FLOW CALIBRATION OF ATWOOD-MIORRILL SWING CHECK VALVE OUTSIDE CONTAINMENT A
B C
D E
F G I H
I J
K JL M
N I
ijORIF. I TEST ORIF. 1 TEST PLANT I TEST
- PLANT I I
2 RUN OLi TIME YER. YorNI FLOWj VEL. 1 PIPE PIPE 1 YEL+/- FLOW I Pu IPd OBSERVATIONS 3 NO.pc-sec YOLTS qpm fps i I D ni II D.1 nop_____
______m ps_
_ps 4
APR 3 1 1986 J11 1
t(
l 5
411 840 1 55.85 4.6 6 jyes2504 31.0 49.562 1.374j 11.19 3762 j46.05143.401 START Of LIGHT INFREQUENT TAPPING 6
42 I 840 114.23 6.31 s
3457 42.8 9.562 11.3741 15.451 5194 j61.70!54.95 LIGKT TAPPING ABOUT 1 1N 3-5 SEC 7
43 840 100.15 8.2 13997 149.5 19.562 111.374 1 17.86 6005 153.35i44,431 ALMOST NO TAPPING.,ABOUT I IN IOSEC.
8 44 840 144.841 4.6 yes 2693 33.3 19.5621 11.3741 12.03 ! 4046 143.83F39.701 LIGHT TAPPING ABOUT 3-4 HZ.
9i I
I I
T-43.7 DEG F
TABLE 5.
FLOW CALIBRATION OF ATWOOD-MORRILL SWING CHECK VALVE OUTSIDE CONTAINMENT A J C D E
F G
H I
J 1
IORIF. TEST IORIF.
ITEST IPLANT iTEST 2
RUN VOL ITIME IYER.
YorNTFLOW iVEL.
IPIPE PIPE ZYEL 3
NO.
cu-ft sc IYOLTSI iqpm Ifps IIDin ID in ifps 4
APR 3 119861 1
1 1
1 1
5 6 840 160.47 4-448.8*(B5+6.34*D5)/C5
=(FS/448.8)/0.499 9.562 111.374 1-F5/(448.8*0.7854*(H5/12)"2) 7 K
L m
M IPLANT I
2 FLOW Pu IPd 1OBSERVATIONS 3
GPm Psi3Ipsi
.4 5
- (0.706/0.495*(62.43/56.27)'0.5*F5 26.8 23.4 ISTART OF LIGHT INFREQUENT TAPPING 6
1 E
7 IT=43.7 DEG F
TABLE 6.
FLOW CALIBRATION OF 12" ATWOOD-MORRILL SWING A
B F
G H
I J G 1 ORIF. iTEST ORIf.
TEST IPLANT ITEST 2 RUN POL TIM YERNEI YorN?!FLOW IVEL.
PIP I P E 1YEL T NO.
cu-fl3e j!
IJLTZZ jq if~
.I~ 1L Inr _______
4 APR 3
.198611-i_
41 40 155.85 4.66
- 448.8*
85+6.34*5/
=0f5/448.)/0.18.562 1.374.=F5/(448.8*0.7854#(H5/I2)'2) 642 le40 114.23 16.31 Ve3 448.e* B6.6.3D)C 6(488.8 D6)/C6 *(F6/448.8)/0.18 i9.562 111.374 f/(448.8*0.7854*(H6/12)2) 140 100.158.2 1
=48.8(B7+6.34D77/C7
=(7/448./.18 9.562
.11
(.8.7854(H7/12)2) 44 6.3T44 8 8 *.7aD7_,
2
-U64 0 141J44.84r14.6 lues u448.8'(R
.34'D)/CS
=(FB/448.8)/O.18 !9.562 11.374 =2F8/(448.8*0.7854(H8/12)-2)
FLOW CALIBRATIONOF 12"ATWOOD-MORRILL SWING K.
L N
1 PLANT 2
FLOW lp Pd JOBSERYATIONS 3qpm psi P31 4
4 95.
5 = 0.706/0.495 *(62.43/56.27).*F 46.05 43+/-.4 STfAGRT O F LGHT IWNEQUENT TAPPING 6
- (0.706/0.495) 62.43/56.2710.5*F6 61.7 54.95LIGHT TAPPING ABOUT IN3-5SEC 7
=(0.706/0.495)* 62.43/56.27)'O.5*F7 53.35 44.43 1ALMOST NO TAPPING r ABOUT I IN 10 SEC.
8 = 0.706/0.495)*(62.43/56.27$"0.5*F8 43.83 39.7 iLIGHT TAPPING ABOUT 3-4 HZ.
9 143.7 DEG F
was slightly more tapping:
A greater intensity and higher frequency with the orifice installed.
In general, the tapping started at approximately 2400-2500 gpm, increasing in frequency and amplitude to a maximum of about 3-4 hertz and then almost disappeared at a flow rate between 5700-6000 gpm.
A qualitative comparison between the tapping noises generated during these tests, with the other two previous sets, would rank this installation between the tests on the valve inside the containment and the tests on the valve downstream of the control valve.
It is our opinion tnat the check valve functioned satisfactorily in this location and should provide years of trouble free service.
Leak Tests A reverse flow leak test at a differential pressure of 65 psi produced a leakage rate of 0.016 gpm.
Future Wear Tests The phase B tests to be conducted in the future will allow visualization of the stability of the disc as well as provide information on the wear rates of the valve under various flow conditions. That information will provide more quantitative data for evaluating the performance of the check valves and give relative wear data.
This relative wear data, when used in con junction with the first inspection of the valves at the next refueling, will allow a reasonable maintenance schedule to be developed.
32
Witness of Tests The tests on the first two configurations were witnessed by the following individuals:
John Hosmer, Jeff Statton and Ramesh
- Shah, Bechtel; Paul Tullis and William Rahmeyer, Utah State University; Herb Rockhold, INEL/NRC; Brian Watts, Shawn Bailey and Bruce Duncil, Southern California Edison: Edgar J. Bolton, AtwooJ
& Morrill; E.H.
- Trohier, USNRC and Rex A.
- Elder, Consultant.
The tests simulating operation of the 12" check valve were witnessed by Herb Rockhold, Jeff Statton and Brian Watts.
33
APPENDIX I Calibration Data
CAL:-B-IA'TION OF FLOW METERS AT UWRL Calibration Facilities The hydraulic testing laboratory at the Utah Water Research Laboratory consists of 50,000 square feet of snace that contains an 8. foot wide, and 6 feet deep flume, several 3 foot wide flumes that run in both the north-south and the east-west directions throughout the test floor, a network of steel pipes located under the floor that allow a water supply to be obtained from a large number of locations, and a variety of other items needed for testing and calibration of varies devices.
The water supply for such testing is supplied by gravi ty from an upstream reservoir through a 48 inch diame ter pipe.
On the floor of the hydraulics laboratory the head available from the reservoir is 30 feet.
For applications requiring larger heads the flow from the reservoir can be directed through several available pumps before being put into the piping network under the laboratory floor.
The primary flow measurement equipment consist of two weighing tanks, each with a capacity of 30,000 pounds, and two volumetric tanks for larger flowrates.
Each of the volumetric tanks consists of a large tank that spills into a small vernier tank. The smaller of the two main tanks spills over into the larger of these tanks. The combined capacity of these tanks is about 3,600 cubic feet or 225,000 pounds of water.
Electronic controls automatically divert the flow from the volumetric tanks to the waste channel when the flow from the tank first begins to spill into its vernier tank over a sharp edge weir between the two. When the spillage from the large tank into its vernier tank has ceased, the volume of water in the vernier tank is ava ilable as a function of a voltage that is displayed on the operating console. This volume is added to the vo lume of the larger tank to get the volume collected in the recorded time. The reading of one division on the voltage readout is equivalent to 0.2 cubic feet of water in the vernier tank.
The operation of the two 30,000 pound weighing tanks is also optionally controlled manually, or auto-
Page 2 matically, from the console.
The accuracy of the weicht
-easurements is within plus or minus 5 pounds.
Tc increase the cll ection time the two weighing tmn:s can be operated serially to accumulate the weight of any number of tanksful. What limits the magnitude of flowrates that can be adequately accomodated by the weiching tanks is the time required to drain the first tank while the second tank is filling, etc.
This maxi mum flowrate is from 5 to 6 cfs.
These primary flow measurement facilities are per iodically recalibrated, or their calibrations verified, against National Bureau of Standards Weights that UWRL has purchased for this purpose. The last such verifi cation of the calibrations against these standard weights was accomplished during June, 1984. Before commencing the calibrations of the flow meters for Henry Pratt Company we plan to verify the calibrations of both the weighing and volumetric tanks.
Meter Calibration Procedures Installation of Meters for Testing:
The flow meter to be calibrated is installed in an appropriate location that depends upon its size, and other tests that may be occupying facilities, so that the water can be adequately directed to the meter, and the discharge from the meter can be directed to either the weighing and/or volumetric tanks through a distance as short as practical.
A valve is installed downstream from the meter to control the flowrate through the assembly, and manometers are installed at the pressure taps specified for the tests being performed.
These manometer meas urements always include the differential head of the meter, but may also include other differential heads to determine headlosses due to the meter and/or control valves, etc.
attached to the meter.
Depending upon the tests being performed, the size of the meter, and other special requirements, the water supply may come directly from the reservoir supply, that has approxi mately 30 feet cf head, or it may be directed through pumps in order to increase the working pressures at the meter. The working pressure at the meter must be large enough to prevent anything but minor negative pressures from occuring at the meter's throat where the high vel ocities cause a reduction in pressure.
Page 3 Performing the Calibration:
With the meter prop erly installed the piping system is filled and all air is e:rplled through bleed taps in the system, and water is allowed to flow through the meter to get an idea of what valve settings should be for differential manome ter readings, when negative pressures may occur at the meters throat, and if and when the pressure may need to be increased by pumps. Included in this checkout is a verification that no leaks exist at the connections made to install the meter, and to determine the approx imate flowrates at which it will be necessary to switch from air over water manometers to water over mercury manometers, when series of such water over mercury ma-nometers may be needed. The control valve is closed completely and all manometers are checked to verify that they produce zero readings.
After the system is checked out and any problems have been corrected, the calibration can proceed. The calibration procedure consists of setting the downstre am control valve to achieve a desired flowrate, and wa iting until the flowrate is completely stablized by noting when the manometer across the meter reaches a constant reading. When the flow is stable the flow measurement begins, and the manometers are read period ically during the flow measurement to:
- 1. ensure that the flowrate is steady, and 2. to recognize any errors in reading the manometers. When flowrates are not too large to permit it, the collection time for each flow measurement is 300 seconds, or longer. Generally three consecutive flowrate measurements, and manometer read ings are obtained with a given setting of the downstre am control valve. This procedure provides information related to the repeatibility of the measurements, and helps eliminate any operator errors. The control valve is then adjusted and the procedure repeated for from 10 to 20 separate flowrates to cover the range of Reynolds numbers agreed upon with the customer.
As the data is recorded during the above calibra tion procedure, computations of the meter's discharge ccefficient and differential head are made, and plotted on a laboratory graph. Should any such data show in consistencies, or fall outside an expected error band, the measurements are repeated.
Page 4 After the tests are comPleted the data collected in the laboratory are entered into a computer, and ana lyzed. Tables of data are printed, as well as a graph constructed that gives a plot of the data showing the relationship of the meter discharge coefficient to Rey nolds Number.
The customer is provided two copies of a report that describes the calibration procedures, a brief summary of fluid mechanics principles associated with flowmeters, the data obtained during the testing, the analysis of this data in determining the flow meter discharge coefficient, and any special problems ob served.
The sheets on which the data were recorded in the laboratory are provided as an Appendix to this re port.
Upgrading of Laboratory Calibration Accuracy Most of the past testing and calibration at the UWRL has been with accuracies within plus or minus 1 percent. To ensure accuracies within plus or minus.25 percent we plan to upgrade our test facilities.
This upgrading will include the following:
- 1. The timing system that records the collection times will be replaced by one capable of recording times to the nearest one-hundredth of one second.
The current timing system records times to the nearest one-tenth of a second.
- 2. An electronic monitoring system will be in stalled to provide a record of the movement of the valves that control whether the flow is directed to the waste channel or flow measuring tanks. The controls of the valves with be adjusted so that complete symmetry of their operation is achieved to eliminate errors as sociated with switching the flow from the waste channel to the measuring tanks and back again.
- 3. Addition large bore manometers will be ac quired for the air over water measurements. At present UWRL has only one U-tube air over water manometer. Our other manometers are U-tube manometers designed for ei ther water over Merian Blue fluid (specific gravity
=1.75), or water over mercury.
Page 5
- 4.
The calibration of the weighing tanks and the volumetric tanks will again be verified against Nation al Bureau of Standard weights during February or March, 1985.
Analysis of Accuracy of Calibrations Accuracies of calibration of meters are influenced by a combination of potential inaccuracies in determin ing the flowrates through and the pressure differen tials across the meters. These inaccuracies are much smaller in a laboratory, such as UWRL, than those asso ciated with the field installation of the meter and the field readings and/or recording of the meter differen tial pressures. The accuracies depend upon the magni tude of the flowrate and the performance of the meter.
At smaller flowrates the determination of the flowrate is more precise because collection times can be larger, and the flow itself is more stable; however, the dif ferential head across the meter is small and therefore reading inaccuracies of manometers are a larger frac tion of the total deflection. At larger flowrates, on the other hand, the collection times must be shorter so any inaccuracies associated with time of switching flow from the waste channel to the measuring tanks and back again are larger. At larger flowrates the pressure differential across the meter is larger, but also for most meters this pressure difference begins to fluctu ate with time. If such fluctuations are in the order of several millimeters( or more) as occurs with some meters, then the manometer reading becomes a time aver aged, best judgement of the individual making the read ing.
Table 1 provides an analysis of the accuracies that can be achieved using the large volumetric tank at UWRL for a flowrate of 20 cfs through a meter of 24 inches diameter. In this analysis we assume the above mentioned timer capable of recording times to the near est one-hundredth of one second has been installed.
For smaller flowrates the accuracies are greater both with the volumetric tanks and the weighing tanks.
In Table 1 errors associated with the five sources of inaccuracies are listed, with the errors associated with volume measurements subdivided into 3 separate ca-
Page 6 tagories.
The percent error associated with each of these inaccuracies is provided in the last column of this table. These separate percent errors are squared and summed, and finally the square root of the sum is taken.
This root sum squared equals 0.099 percent, which is well below the required.25 percent accuracy.
TABLE 1.
ANALYSIS OF ACCURAC: IN US:ING 3600 CU3IC FOOT VOLIETRIC TANK lSource of Description Percent
!Error Error I
!1. Volume Measurment A. Depth in Tank Equilibrium depth in tank, 0.03"/66" depth X 100 0.0455 B. Calibration 9 filling with error of 5 lbs each, (5 lb/25000 lb X 100)**2 X 9 0.0600 C. Depth in Venier Accuracy of vernier tank measurement Tank in relation to total volume collected, 0.1"/24"depth X (124/3600)
X 100.f 0.0144
- 2. Timing error Assume flowrate of 20 cfs, Collection time = 3600/20 = 180 sec.
.01/180 X 100
.}
0.0056
- 3.
Switching error Nonsymmetry of directing water from waste channel to volumetric tanks and 0.02 vice versa 3/
- 4. Mercury Manometer Scales are calibrated in millimeters readings and readings to.5 mm, deflection = 1 m.
0.0005 X 100
_4/
0.05 J5.
Fluid Properties Variability of fluid density with temperature & other effects 0.03 Root sum squared = Irej
= V.0097 =0.098 percent 1/ The Calibration of the vernier tank readout is 0.2 cubic feet per division of readout.
On this basis the error equals 0.2/3600 X 100 = 0.0056 %.
A larger estimate is used to account for other possible errors such as hysterisis on the system, surface tension, etc.
2/ The present timer records to the nearest tenth of a second.
This timer will be replaced by one that records to the nearest one-hundredth of a second.
3/ Closure and opening times of waste channel and volumetric tank valves has been adjusted to accuracies obtainable with stop watches.
We plan to install an electronic system that will permit very accurate adjustment to ensure symmetry of the operation between diverting into volumetric tank and back to waste channel.
The percent error shown is thought to be conservative and will be refined once the electronic system is installed.
4/ For smaller deflections, when manometers are very stable, an optical read out will be utilized that gives readings to 0.1 mm or better, 0.01/200 X 100 =.05 % for a 20 cm manometer reading.
A FA4)vk 74>
2s 96>h/
19, 0-944-4 1,a2 0
10 6s e.\\
e
/2 9
o -/9
/8;-19, 853.14 8 5.4 an0 3-
.o 980-5; L
- p 7-4 /of6
/
or
- ?
o o
/0
?94,.2 774.1 -
8 4'l
/2 2
.4 p
?
5 0
.9
-0 8 343,
-,o
-. o 8
/
S
/s o
4 5.2'/
/o298 9
93 ~
-,Oq
-. 18
.4.4 4
2 7 4..
A I/
0
..;1
-l0 5854
- 15j f$4.
40/§
/oligl 1.o
+?
24 P).1.0
.o.1 ff5 4.
ZO/ 2.
- L4.
/}949.8
/
9 41 q)?l
-,06
./
99/*9
/ 4 60
- 22 94 )0 24 0f -go -o?
948 /,4.
89 064.4
/L 214..4
.o2f 40324-N'
.i 2
4
/2-o93 isoS
-9a.
.;/
/
co.
y d
ac>o s4.
./
o
/2tt s.;
/
0 9 2-8 8)*
y8 A' /80-o -/
CALIBRATION OF VALIDYNE PRESSURE TRANSDUCER AND HEISE PRESSURE GAGE A
B C
D E
I -ESSURLFROM DEAD WEEI.G.HTT.TEST.ERpsi PRESSURE G psi
%DIFFERENCE 2F 0.00 0.40i 3
- 6.
6-t pnn
- 5. 2
.4 0...................
1000 -. 1 1601.6...
5 20.00150.
6
.....0.0......~.....................
0....
4
.00
?.......---...................
64.00 10.0 0 --.600 500.000 59.95 69.9 5
.08.
7 9
70.00 4 6998
-0.0 6
....--. 0.......
0.00
- 5. 00..
0.4.0......01....0.
. p
. TRANS..UCER.Y CHART................. l........ IC HAR div/p.........
- 0..
................ a.....N.
.0 1
30.00
.59090 6900 2.0 175.00 10.20
- 5.
1100-......... at. ~v..........
.................. 102r..
20.404
- 1. 00...............j 1.10.20 30.20...6.00.........1.07.201 16 5-Z
.00 39.60.
2.50 03......1....98.....
14 2500.499 0
1...
0 1.0
...00....
15 10.00 j5.90b 16.00 1.03 200 19 40.00 5.0 194.075.20 1 38.50 0.96 18
APPENDIX II Valve Details
APPENDIX II.
DESCRIPTION OF SWING CHECK VALVE By Ed Bolton Atwood & Morrill Salem, Massachusetts The Atwood & Morrill valve is a 10" class 900 "Boiler Feed Pump Check Valve" with "Double Bearing Covers," and pressure-seal bonnet.
The seat bore (minimum flow passage) is 8 1/4 inches in diameter.
There are no external accessories.
The DISC/DISC Arm is cast as one piece for water service.
Seat angle is 200 off
- vertical, the full open angle is 650 +/- 50.
Shaft diameter is 1 3/8 inches and the distance from the centerline of the shaft to the centerline the bore is 5 13/16".
The shaft turns in bushings contained in the bearing covers.
These bearing covers are attached to the side of the body with studs and nuts and the body penetration for the bearing covers are sealed with wound SST/Asbestos gaskets.
The drawing for the valve is Atwood & Morrill 15487-01, similar to 15487-05.
Shaft Diameter -
1.375 t.003" Bushing I.D. -
1.387 t.003"
-.002" Diametrical Clearance -
.007"-.018" Lateral novement of Disc Along Shaft Permissable -
.03"-.09"
0
.04k
.4 0
N
a APPROX S)EE WELD END DETAIL 6.00
).94 oNth REV ML C(
2v 9.T 5
-20.50 i.. t..
2113REQ'D FORS SHAFT REMOVAL E\\THER S\\DE.
9.50 9.50
+
- j 0*
@1<
34
APPENDIX III Test Plan
SAN DNOFRE UNIT 1 FEEDWATER SYSTEM CHECK VALVE TEST PRDGRAM
- 1. OBJECTIVES:
This procedure outlines the testing requirements for performance testing of a 100 feedwater check valve.
The objectives of the test are to establish the stability and suitability of the check valve installations as designed in DCP 3400.OBP as well as to develop a wear rate data from which a maintenance schedule can be based.
Phase A of this test program will address the stability testing.
Phase B of this test program develops check valve wear rate data.
II.
TEST PROGRAM OUALITY CONTROL: The person in charge of the test shall be responsible for controlling the work in accordance with this procedure.
Deviations from this procedure shall be fully explained in the final report accompanying the test data.
All instruments used to measure flowrate and pressure in the performance of these tests shall be talibrated prior to the test.
In addition, documentation of the calibration shall also be provided. Instruments such as stethoscopes.
accelerometers-4M do not require calibration beenas a <
153 80 /. 1f O
1538D/l
PHASE A TESTING o..nd This test is to determine the acceptability of the 10' check valves.
The valves will be tested for stability in operation in the low flow range. In addition, other performance data will be taken.
Two specific test configurations will be used.
The first test (Test 1) will model valve response to low flow operation in a straight piping system (Figs. I sagt.
After completion of this test, performance data will be taken at a the flow conditions listed in the run log.(
- %&- I Pa-3D IA Acceptance Criteria - Test 1 Pressure oscillations induced by the check valve when corrected for plant conditions are less than 1.6 at flows between 25 and 150 GPM.
(Refer to calculation NC-791-01 for basis.)
In addition, a qualitative review and acceptance of valve performance will also be provided.
trerequisites - Test 1 The following requirements shall be met prior to commencement of testing:
- Test piping installation completed and in accordance with Figure 1.
- Test instrumentation has been calibrated and installed.
Procedure - Test 1 (Refer to Figure I and lA)
The following test procedure shall be used for the flow conditions listed in the Run Log.
- 1. Establish steady state flow through the check valve and test piping.
- 2. Obtain water temperature reading and record reading.
- 3. Initiate the flow testing by means of calibrated weigh tanks, volumetric tanks, or a calibrated orifice.
- 4. With the system maintained at steady state, obtain average P1 and P2 readings and record the readings.
- 5. With the system maintained at steady state, obtain a reading of the downstream pressure fluctuations (P2) and record this reading.
This can be obtained through review of strip recorder output from PT, or from test oscilloscope.
- 6. Obtain the following additional process data on the check valve by adjustment of the test flow through the check valve when these characteristics are found, flow rate must be determined by metering steady state operation.
A. Identify "tappingO range of the valve.
This can be done through qualitative measurements such as stethoscopes and accelerumeters.
B. Identify the seating velocity of the check valve.
1530D/2
PHASE A TESTINS (continued)
AcceptAnce Criteria - Test 2 (Refer to Figure 2)
A qualitative review and acceptance of valve performance will be provided.
frerequisites - Test 2 The following requirements shall be met prior to commencement of testing.
- Test piping installation completed and in accordance with Figure 2
- Test instrumentation has been calibrated and installed.
Procedure - Test 2 (Refer to Figure 2)
The following test procedure shall be used for the flow conditions listed in the Run Log.
- 1. Establish steady state flow through the check valve and test piping.
- 2. Obtain water temperature reading and record reading.
- 3. Initiate the flow testing by means of calibrated weigh tanks, volumetric tanks, or a calibrated orifice.
- 4. With the system maintained at steady state, obtain average P1 and P2 readings and record the readings.
- 5. With the system maintained at steady state, obtain a reading of the downstream pressure fluctuations (P26) and record this reading.
This can be obtained through review of strip recorder output from PT, or from test oscilloscope.
- 6. Obtain the following additional process data on the check valve by adjustment of the test flow through the check valve when these characteristics are found, flow rate must be determined by metering steady state operation.
A. Identify 'tapping' range of the valve.
This can be done through qualitative measurements such as stethoscopes and accelerumeters.
B. Identify the seating velocity of the check valve.
A O
15380/3
PHASE B TESTING (LATER)
FIGURE 1 10" VALVE MODEL INSIDE CONTAINMENT MINIMUM 300 MINIMUM 200 MINIMUM 100 12NSTRAIGHT PIPE 10STRAIGHT PIPE 10"STRAIGHT PIPE FE PT TI TEST FLOW CONTAINER CONTROL OF KNOWN PI Pi VALVE VOLUME TEST PUMPP1P P1 P2 RUN LOG RUN DESCRIPTION
- 1.
LOW FLOW (25 +/-15GPM)
- 2.
LOW FLOW (55 +/-15GPM)
- 3.
LOW FLOW (150 +/-15GPM)
- 4.
FULL FLOW
- 5.
PART FLOW (90% +/-10%)
- 6.
PART FLOW (80% +/-10%)
- 7.
PART FLOW (70% +/-10%)
- 8.
TAPPING RANGE OPERATION
- 9.
SEATING RANGE I.DGN
FIGURE 1A 12" VALVE MODEL MINIMUM 300 MINIMUM 200 MINIMUM 100 12"STRAIGHT PIPE 10"STRAIGHT PIPE 10"STRAIGHT PIPE FE PT TI WATER RESERVOIR TEST FLOW CONTAINER CONTROL OF KNOWN Pi P I VALVE VOLUME TEST 46 PUMP P 1 P2 RUN LOG RUN
- DESCRIPTION
- 1.
FULL FLOW
- 2.
PART FLOW (90% +/-10%)
- 3.
PART FLOW (80% +/-10%)
- 4.
PART FLOW (70% +/-10%)
- 5.
TAPPING RANGE OPERATION
- 6.
SEATING RANGE SONGS 2 & 3 QS2:1320,302]RUNLOG. DGN
IGURE 2 10" VALVE MODEL OUTSIDE CONTAINMENT 300 50 12"STRAIGHT PIPE STRAIGHT PIPE PT
.TI WATER RESERVOIR N
E FCV TEST FLOW CONTAINER CONTROL OF KNOWN P1 P2 VALVE VOLUME TEST PUMP RUN LOG RUN
- DESCRIPTION
- 1.
FULL FLOW
- 2.
PART FLOW (90% +/-10%)
- 3.
PART FLOW (80% +/-10%)
- 4.
PART FLOW (70% +/-10%)
- 5.
TAPPING RANGE OPERATION
- 6.
SEATING RANGE SONGS 2 & 3
SAN ONOFRE UNIT 1 FEEDWATER SYSTEM CHECK VALVE TEST PROGRAM ADDENDUM 1 PURPOSE: The purpose of the Addendum is to detail the requirements for test
- 3. Test 3 will simulate the 12" feedwater check valve stability under a model of the plant piping layout (see Figure 3).
ACCEPTANCE CRITERIA: A qualitative review and acceptance or valve performance will)be provided.
PREREQUISITES:
Test 3 The following requirements shall be met prior to commencement of testing:
Test piping installation completed and in accordance with Figure 3.
Test instrumentation has been calibrated and installed.
PROCEDURE:
Test 3 (Refer to Figure 3)
The following test procedure shall be used for the flow conditions listed in the Run Log.
- 1. Establish steady state flow through the check valve and test piping.
- 2. Obtain water temperature reading and record reading.
- 3. Initiate the flow testing by means of calibrated weigh tanks, volumetric tanks, or a calibrated orifice.
- 4.
Obtain the following additional process data on the check valve by adjustment of the test flow through the check valve when these characteristics are found, flow rate must be determined by metering steady state operation.
A. Identify "tapping" range of the valve. This can be done through qualitative measurements such as stethoscopes and accelerometers.
B. Identify the seating velocity of the check valve.
1538D/2 Prepared by:
S'O 7'
Dr. J. Paul Tullis Reviewed by:
S E-C f
Jeff Statton, BPC Eng.
Reviewed by:
OR N
L t
tJ A7O Brian Watts, SCE Eng.