ML18045A746

Evaluation of Palisades Plant Waterhammer Test & Potential Waterhammer Effects.

ML18045A746
Person / Time
Site:	Palisades
Issue date:	09/30/1980
From:	ABB COMBUSTION ENGINEERING NUCLEAR FUEL (FORMERLY
To:
Shared Package
ML18045A745	List: ML18045A744 ML18045A746
References
NUDOCS 8010070609
Download: ML18045A746 (12)

Text

J G Keppler, Director Palisades Plant September 30, 1980 Attachment l EVALUAT.ION OF PALISADES PLANT WATERHAMMER TEST

• AND POTENTIAL WATERHAMMER EFFECTS Prepared for Consumers power Company By Combustion Engineering, Inc.

Nuclear Power Systems September 30, 1980 80I00706Q9

Attachment 1 ~

J G Keppler, Director Palisades Plant September 30, 1980 EVALUATION OF PALISADES WATERHAMMER TEST DATA 1.0 Introduction This report has been prepared for the Consumers .Power Company by Combustion Engineering, Inc. (C-E). It contains the re-sults of an evaluation of the test data from the waterharnrner test conducted on.May 10, 1980. The test was performed at a steam pressure of 200 psia and at feedwater flowrates of 150 and 300 gpm. This report also includes an evaluation of poten-tial waterhammer effects at hot standby conditi9ns.

2.0 Test Data Prior to the test, special instrumentation was installed on the feedwater pipe and feedwater nozzle of steam generator B.

Six strain gauges were installed on the feedwater pipe next to .the nozzle. Four of the gauges provided bending and tensile strains, while the remaining two gauges recorded hoop strains.

The analysis was performed using the output of one gauge which measured bending and tensile strains in the horizontal plane.

The outputs of the five remaining gauges were lost because of insufficient exposure of the light sensitive paper. The feed-water pipe was also instrumented with an accelerometer, which was located on the first downward sloping elbow of the feegwater pipe. This provided an independent measurement of piping dis-placement and natural frequency.

3. O. Analysis

• An elastic model of the feedwater piping was made using the C-E DAGS76 Dynamic Piping Program. This code is the standard C-E computer program for seismic and LOCA dynamic piping analyses for the PWR primary system. In order to simulate the measured response of the strain gauge, various hydraulic forcing functions were assumed for each elbow.

/

The initial compressive strain was postulated to have occurred as the water slug passed through the first elbow. A tensile stress was then assumed as *the pressure on the second elbow increased due to slug retardation downstream of the first elbow.

Figure l contains a diagram of the location of the reactive

~Qi::c.es,_ their directions.and the locations .of.the instrumenta-tion, It was first assumed that classical waterhamrner had occurred during the 300 gpm test. Applying the resultant forcing function, Figure 2, to the piping system model produces piping strains, Figure 3, which are 15 times greater than the corres-ponding field test data. It was, therefore, concluded that classical waterharnrner did not occur.

J G Keppler, Director Palisades Plant September 30, 1980 2 The structural analysis was repeated, with various pressure profiles. The* results were then compared to the test data.

Figure 4 presents the final forcing function and Figure 5 .

shows the corresponding calculated response along with the

. test measurements. As can be seen, this forcing function produces a piping response which compares quite favorably with the test results. This forcing function corresponds to a water slug of approximately 6 feet with an impact velo-city of about .20 ft/sec. The velocity is significantly less than that which would.be calculated using classical rapid condensation rates. Consequent'iy I it appears th'at slow con-densation occurred throughout the process rather than rapid condensation which is usually associated with waterhammer experienced in other nucl~ar units.

A review of the accelerometer data substantiates the con-clusions based on the strain gauge readings. The indicated displacements and frequency of the accelerometer agree reason-ably well with the strain gauge data. Hence, further credence.

is given to the conclusion that a relatively mild waterhammer transient occurred involving only slow film condensation. The

• calculated hydraulically induced stresses are less than the allowable stresses corresponding to.pressure plus Operating Basis Earthquake (OBE) seismic limits.

4.0 Extrapolation of Test Results to Operating Conditions The analysis of the 200 psia test data indicates that pressure spikes on the order of 600 psi occurred during the test. Ex-trapolation of the peak pressure to hot standby conditions using the square root of the pressure ratio produces an over-pressure of 1260 psi. The stress resulting from the test conditions are within theFSAR (OBE) limits as shown in Table 1. (Other examples of faulted limits are also shown in Table 1) .

Although considered highly unlikely, if rapid condensation were postulated to occur at hot standby, the slug could theo-retically achieve a high velocity (approximately 260 ft/sec) and the calculated resultant peak stresses would exceed the allowable stress limits.of the piping.

--- The Palisades -Plant has never experienced -a waterhammer during -

• actual plant operation even though there were a number of*

occasions when the feedwater sparger was drained arid conditions conducive to waterharraner existed. A comparison of the Palisades

y G Keppler, Director Palisades Plant 3

September 30, 1980 piping with other plants indicates that there are two unique design features which may mitigate or eliminate the effects of water hammer. The feedwater ring is two feet below the feedwater nozzle and the horizontal run of feedwater piping

. has a 90° bend in the horizontal plane approximately 8 feet from the feedwater nozzle. The lower feedwater ring reduces.

the number of times the ririg is exposed to a steam environ-ment. If rapid condensation were to occur, the 90° bend in the horizontal plane may eliminate or mitigate the waterhammer by breaking up the slug as it passes through the bend. Figure 6 contains a curve which shows that if rapid condensation occurs, the slug would be in a cavitating condition as it passed through the horizontal elbow. The cavitation in turn would allow steam to bypass the slug, restoring the downstream pressure. This theory is supported through experimental data on cavitating elbows (reference l) . Based on discussions with operators on duty when the 200 psia waterhammer testing was performed, it is extremely unlikely that waterhammer of simil.ar magnitude could have previously occurred undetected. To further postulate that waterhammer of greater magnitude (based on either extrapolation of the 200 psia data, using the square root of the pressure ratio~- or classical waterhammer theory) has gone undetected, is virtually inconceivable.

During the recent feedwater pipe cracking repairs at Palisades, the area of the thermal sleeve in the feedwater nozzle'. was closely inspected on several occasions. No deformation of the thermal sleeve was noted. Deformation (expansion) of the thermal sleeve has been reported during inspection of a steam generator which had been subjeqted to waterhammer at another plant. Lack of such deformation forther corroborates the conclusion that waterhammer has not occurred at Palisades, other than during the test performed at off-normal conditions.

5.0 Conclusions A. Although conditions which are generally conducive to water-hammer existed on a number of occasions there have been no observed occurrences of waterhammer under normal operating conditions at Palisades. This finding is based on:

l. The operators would have noted a waterhammer even if it were as mild as the one observed during the 200 psia

-test-. -- - - - - - - - - --

2. No deformation of the thermal sleeve in the feedwater nozzle was observed druing examinations perforr.ted in conjunction with the repair of the feedwater piping.

J G Keppler, Director Palisades Plant Septeml:>er 30, 1980 4 B. If a waterhammer were to occur at hot standby, with the same condensation rate as inferred from the 200 psia test, the resulting stresses would be within faulted allowables.

C. If one were to postulate a more rapid condensation mechanism at h~t standby (rates consistent with classic waterhammer theory), the increased slug velocities associated with the waterhammer are predicted to result in cavitation as the slug passes through the horizontal elbow. This .would prevent the generation of very large impact forces at the downturning elbow. * *

• IN LIGHT OF THE ABOVE, IT IS CONSIDERED HIGHLY UNLIKELY THAT FEEDING A SINGLE STEAM GENERATOR AT FLOW RATES OF UP TO 300 GPM AT HOT STANDBY CONDITIONS WOULD RESULT IN AN UNACCEPTABLE WATER-HAMMER AT PALISADES PLANT.

6 .0 References (1) D S Miller, "Internal Flow Systems", Volume 5, Published by BHRA Fluid. Engineering, 1978

TABLE 1 STRESS LIMITS

l. Test Conditions Pressure Stress @ 200 psi = 1150 psi Dead Weight Stress = 95 psi Waterhanuner stress = E x Strain

= 27.0 x 106 psi x (550 - 1150 27.9 x

= 13,977 psi FSAR OBE CRITERIA (18,000 psi) 1150 psi+ 95 psi+ 13,977 psiC::l.2Sh 15 , 2 2 2 < 18 , OOO ps i

2.

• Extrapolated Conditions Pressure Stress @ 900 psi = 5175 psi Deadweight Stress = 95 psi Waterhammer Stress = E x Strain x J9/2.

= 27.9 x 106 psi (550 1150 > x io-6 x J9/2 21.9 x io6*

= 29,800 psi FSAR SSE CRITERIA (36,000 psi) 5175 psi + 95 psi + 29, 800 psi< 2. 4Sh 3 5 , O7 O psi < 36 , OO0 psi

. \

Note: Other faulted condition allowables could be 3.0sh = 45,000 psi and .7 s~lt = 42,000 psi

PIPING DIAGRAM FORCE AND INSTRUMENTATION LOCATIONS STEAM GENERA TOR B

x y

~

FB ---"--I-~.

\ p

\

FA X - LOCATION OF STRAIN GAUGE Y - LOCATION OF ACCELEROMETER FIGURE 1

CLASSICAL WA TEA HAMMER FORCING FUNCTION FOR 300 GPM TEST 0

• ~ 3000L----+-~-+--~+--+--.....+-~+-~l---+~-+-~+---t~--;

<(c.< 2000 LI..

1000L---+-~....+:-~+--+--l-ll-+-~+-~l---+~-+-~+---t~--;

.1 .2 .3 .. .4 .5 TIME, SECONDS

- .l .2 .3 . - -- -*'." .5 TIME, SECONDS FIGURE 2

9LCULATED STRAIN GAUGE .PONSE ASSUMING RAPID CONDENSATION STRAIN AT STEAM GENERATOR NOZZLE ANALYSIS RESULT (3000 psi) STRAIN GAUGE DAT A

• 11000 10,000 9000 8000 7000

• aooo 5000 4000

\0 0 I

~ 3000 '

I

~ 2000

~

E-1 U) 1000 J r\

0 'l r"\. -

u 11 / v

-1000

-2000 u

-3000

-4000 M

-5000

.1 .2 .3 .4 .5 .6 .7 .8 .1 .2 .3 .4 .5 .6 .7 TIME, SECONDS TIME; SECONDS - - - -- -

FREQUENCY ~ 15 hz FREQUENCY = 11.4 MAX. DISPLACEMENT . MAX. DISPLACEMENT AT Y = 2.67 IN AT Y = .18 IN FIGURE 3

SLOW CONDENSATION FORCING FUNCTION FOR*

300 GPM TEST

• 200 '

I --- r----...,._

.02 .04 .06 .08 .10 .12 .14 .16 .18 .20 SECONDS TIME

• -a>*

~ soo1----+-_;_~-4-----1---+1--+----1----+--+--+--+--+----t

~1~4001----+--~---4-----1---M-+----1----+--+--+--+---+----t 3001---+--+--+--+--+-t-+---+--+---t--+--+---t--

200 1001---+---+--+---l---+-+-+---l---+--+--+-'----+--+---t

.02 .04 .06 .08 . 10 . 12 . 14 . 16- . 18 .20 SECONDS TIME FIGURE 4

~ALCULATED STRAIN GAUG-ESPONSE ASSUMING SLOW CONDENSATION STRAIN AT STEAM GENERATOR NOZZLE ANALYSIS RESULTS STRAIN GAUGE DAT A (600Psi) 800 20 700 600 15 500

• 400 10 300 n co 200 , 5 0

.,.. r x 100

• ~ ~

~

c(

0 0

~

....a:

"' \\ / l \- "'"'a:w

-100 v u

\J V' ....

-200 -5 "'

~

-300

-400 -10

.1 .2 .3 .4 .5 .6 .1 .2 .3 .4 .5 TIME-SECONDS TIME-SECONDS FREQUENCY FROM ACCELEROMETER FRE:OUENCY ..

11.4 hz ~19 hz MAX~ DISPLACEMENT MAX. DISPLACEMENT AT Y=.18 IN. AT Y=.20 IN.

FIGURE 5

CAVITATION CRITERIA FOR ELBOWS r

I

~p ISADES BEND GE METRY

.Q I

b 3r----~-t-~-+----if--~~-+--+-~--l~~~~~~_;.J I

a:

w CD N~ CAVIT noN

~ I

> 2i--~~-r-~~--i~~~-+--t--~--l~---~-+----~~--I z

z 0

j:::

<C

.... 1 OPERAT NG

<C

(.)

.5 1.0 1.5 2.0 rid FIGURE 6