ML20204G294

Forwards Results of Analysis of Station Electrical Distribution Sys Voltages & Comparison of Test & Analytical Results.Differences Between Test Voltages & Analytically Derived Voltages within 3% Value

ML20204G294
Person / Time
Site:	Rancho Seco
Issue date:	04/28/1983
From:	Reinaldo Rodriguez SACRAMENTO MUNICIPAL UTILITY DISTRICT
To:	Stolz J Office of Nuclear Reactor Regulation
References
TAC-12746, NUDOCS 8305020415
Download: ML20204G294 (7)

Text

s

($SMUD SACRAMENTO MUNICIPAL UTILITY DISTRICT O 6201 S Street, Box 15830, Sacramento, Califomia 95813; (916) 452-3211 l

April 28,1983 I

DIRECTOR OF NUCLEAR REACTOR REGULATION ATTENTION JOHN F STOLZ CHIEF OPERATING REACTORS BRANCH 4 US NUCLEAR REGULATORY COMMISSION WASHINGTON DC 20555 DOCKET 50-312 RANCHO SEC0 NUCLEAR GENERATING STATION UNIT N0 1 ADEQUACY GF STATION ELECTRICAL DISTRIBUTION VOLTAGES In our letter of November 8, 1982, the Sacramento Municipal Utility District committed to verify the assumptions and computer program used in the analysis of station electrical distribution system voltages. Attached are the results of these tests including a comparison of test results and analytical results.

\

R. J. odrigu Executive Dir c or, Nuclear i

Enclosure i

lots' A

P305020415 830428 MR ADOCK 05000 P

Ah ELECTRfC SYSTEM SERVlNG M0RE THAN 600.000 IN THE HEART OF CAllf0RNiA

. l I

i i

1. INTRODUCTION The NRC's August 8, 1979 letter to all reactor licensees (except Humbolt Bay) requested that "The adequacy of the onsite distribution of power from the offsite circuits shall be verified by test to assure that analysis results are valid. Please provide: (1) a description of the method for performing this verification, and (2) the test results. If previous tests verify the results of the analysis, then test results should be submitted and additional tests need not be performed."

By letter dated February 17, 1981 the District responded as follows:

"The District will perform a test during the refueling outage expected to take place in mid-1982 to verify the assumptions and computer program used in the analysis.

The test will be performed with the plant shutdown and with the available grid voltage. The nuclear service buses for Train "A" and "B" will be fed from the offsite sources, loads added to thea~and the bus voltages recorded.

The impedance of the loads that are being served.during the test and the recorded switchyard voltage will be entered as data for a computer run. The calculated voltages will then be compared with the test results".

The NRC by letter dated September 20, 1982 set the following criteria for the test verification. "As a minimum the staff requires the following criteria be met".

a. Loading the station distribution buses, including all Class 1E buses down to the 120/208 volt level, to at least 30%;

b. Recording the existing grid and Class lE bus voltages and bus loading down to the 120/208 volt level at steady state conditions and during starting of both a large Class lE and non-Class 1E motor (not concurrently).

Note: To minimize the number of instrumented locations (recorders) during the motor starting transient tests, the bus voltage and loading need only be recorded on that string of buses which previously showed the lowest analyzed voltages.

c. Using the analytical techniques and assumptions of the previous voltage analyses, and the measured existing grid voltage and bus loading conditions recorded during conduct of test, calculate a new set of voltages for all Class lE buses down to the 120/208 volt level.

d. Compare the analytically derived voltage values against the test results.

e. With good correlation between the analytical and test results the test requirements will be met. In general, the test results should not be more than 3% lower than the analytical results; however, the difference between the two when subtracted from the voltage levels 1

-i determined in the original analysis should never be less than the Class 1E equipment rated voltages.

By letter of November 8, 1982 the District commented on the requirement to record voltages'at the 120/208 level as follows:

In reviewing your minimum acceptable test criteria, we have the following comments. Our Class 1E buses at:the 120/208 volt levels are part of a four channel 120V AC/125 DC system. Each channel has a 125 volt battery, a battery charger fed.from a diesel backed Class 1E 480 volt bus, a 125V DC panel, an inverter, and a 120V AC panel. This system is isolated from voltage fluctuation on the AC system by the battery charger. The only.

Class 1E, 120V AC bus that is transformed down from a 480 volt bus is used' exclusively for heat tracing and freeze _ protection. Since low voltage operation and protection of these piping systems has been adequately addressed in our analysis, the District does not propose to load, monitor,

-or analyze this bus as part of our verification test.

The District completed the voltage verification tests on February 22, 1983.

- The purpose of this report is to present the test results and show the cor-relation between voltages obtained by analysis and test.

BUSES TESTED A one line diagram of the buses of the RANCHO SECO electrical system that were used in the analysis verification tests is shown in Figure 1. The 4160V, Class 1E bus 4B and the non Class 1E bus 4D are energized through startup transformer No. 2. Non Class 1E 4160V buses 4E1 and 4E2 are disconnected. from startup transformer No. 2. The 480V class 1E buses 3B and 2B1 are energized through station service transformer X43B2.

In the previous analysis the lowest voltage levels were essentially the same on the train A and B string of buses. Tests were made on the train B string of buses.

Bus Loading Test During the bus loading test, voltages, currents and phase angle were measured and the bus loads calculated from the measured values. The string of buses included in the test was analyzed using the Voltanal (TE503) plant voltage calculation program for the loading condition measured in the bus loading test above. The voltanal (TE503) program is the computer program that was used to review by analysis the adequacy of the station electric distribution system submitted to the NRC by letter of April 2,1982.

I Table 1 lists the loads in percent of normal bus load and voltages obtained from measurements made-during the analysis verification test and the voltages taken from the results of the analysis by computer. The percentage differences between the test results and the analytical results are listed in Table 1.

i 2

i

220KVSwlTCMYARD y STARTUP TRANSFORutR to.2

.i.

.,, 321 - 4J5J.35 KV

,v-^^^~virv^--~vi TAP 5230 KV gi ytNDING - 42.8 NVA Y 2 Y wihDING - 34 NVA I wihDING - 22.6 NVA TO 416C V BUS $ts all & 4t2

=

4180 V BUS 4D 1 r E18 SYSTir 416C V BUS 48 l

STAT 60k SE RVICE am, , , TRANS.34382 Z Z Z 418C - 400 V av' 1 i yap.gggg y L12BVA 40C V BUS 38 1 r l

W V BUS 251 1 P Figure 1 l

l 3

c:

a,.

TABLE 1 PERCENT OF' TEST ANALYSIS LOAD IN VOLTAGE VOLTAGE DIFFERENCE BUS FSAR TABLE 8.2-2 kV 'kV  %

220 kV

'Swyd -

232.04 232.03 -

4B 37 -4.36 4.35 0.2 3B 64 0.478 0.479 0.2

2B1- 53 0.477 - -

i 4D 42* 4.34 4.35 0.2

• A nominal load of 5446 kVA was assumed for bus 4D Since power factor was not measured on bus 2B1-during the first load test,

.a followup test was carried out to determine the correlation between-

' test voltages and analytical results for buses 3B and 2Bl. The results listed in table 2 below also show good correlation.

[ TABLE 2 i

PERCENT OF TEST ANALYSIS LOAD IN VOLTAGE VOLTAGE DIFFERENCE BUS FSAR TABLE 8.2-2 kV kV  %

3B 52 0.483 0.483 -

I~

2B1 59 0.482 0.482 0.0 In all cases the difference between test and analytical results is less than one half of one percent and well below the required 3%. The differ-

, ence between the voltage'by analysis and test when subtracted from the original analysis minimum voltage level resulted in only~a slightly

! -lower voltage level than the voltage determined in the original analysis.

In the worst case the minimum equipment terminal voltage with the grid at

~

j 214 kV was determined in the original steady state analysis to be 85 percent of rated motor voltage. Allowing for the 0.2 percent difference between test and analysis r'esults, the minimum voltage is' reduced to 84.8 percent.

At this voltage level, the demonstration in the original analysis that safety related equipment can perform its required function is still valid.

t 4

i

'9 -

Class 1E Motor Starting Test In the Class 1E motor starting test the 1000 hp auxiliary feedwater pump motor was started on Class 1E bus 4B and bus loads and voltages measured.

Voltage levels on the buses were then analytically found'using the Voltanal TE-503 computer program for _the starting condition of the auxiliary feedwater pump motor and bus loads as measured during the steady state test. The test and analytical results which show a difference of less than the required 3%

are listed in table 3 below. In the original analysis a minimum transient-class 1E terminal voltage of 75 percent occurs with a minimum expected grid-voltage of 214 kV during the block loading of the class 1E buses with the nuclear service air coolers and the nuclear service cooling water pumps.

The application of the difference between test and analysis results to the minimum transient voltage of 75 percent-has the effect of reducing the' starting torque of the nuclear service air coolers and the nuclear service water pumps by 1.21 percent.- This difference does not prevent the motors from starting. _

TABLE 3 TEST VOLTAGE ANALYSIS VOLTAGE DIFFERENCE BUS kV - kV  %

220 kV

=Swyd 232.06 232.03 -

4B 4.22 4.21 0.2 3B 0.460 0.464 0.9 4D 4.2 4.22 0.5 Non Class 1E Motor Starting Test In this test the 3500 hp condensate pump motor was started on non-class 1E bus 4D. A motor starting _ voltage drop analysis using the Voltanal TE-503 computer program was also made for the condensate pump. The test and analysis results showing good correlation are listed below in table 4.

In the original analysis the worst case minimum transient class 1E equip-ment terminal voltage is 82 percent of 4160V on bus 4B during the starting of the condensate pump motor. When the difference of 0.06 kV is subtracted from the original analysis minimum voltage of 3.432 kV the corrected voltage value is 81.1 percent. This voltage transient as explained in the original analysis lasts for approximately 2 seconds and does not prevent the class 1E equipment from performing its function.

5 i

m-TABLE 4 TEST VOLTAGE ANALYSIS VOLTAGE. DIFFERENCE BUS kV kV -  %

220 kV Swyd 232.04 232.03 -

4B 3.92 3.98 1.5 3B' O.433 0.437 0.9 4D 3.87 3.98 2.8 Conclusion The test results listed in tables 1 through 4 show that there is a good correlation between the results by analysis and the results by test. The difference between the test voltages and the analytically derived voltages is within the 3 percent value given in the test criteria outlined in the NRC letter of September 20, 1982. Furthermore, when the program' error is applied to the results -of the TE-503 computer program results the safety related equipment can still perform its-required function.

6 i