ML20052D534
| ML20052D534 | |
| Person / Time | |
|---|---|
| Site: | Arkansas Nuclear  |
| Issue date: | 04/21/1982 |
| From: | Stolz J Office of Nuclear Reactor Regulation |
| To: | Cavanaugh W ARKANSAS POWER & LIGHT CO. |
| References | |
| NUDOCS 8205060561 | |
| Download: ML20052D534 (4) | |
Text
_ _ _ _
M APR 21 1982 W
s DISTRIBUTION:
ORB #4 Rdg 0EL,D
+
/
Docket File DEisenhuf AE0D 2,
Docket flo. 50-313 Gray File JTsao 2
APR271982m EBlac od lir. William Cavanaugh, III Senior Vice President, N
Energy Supply Arkansas Power & Light Company P. O. Box 551 Little Rock, Arkansas 72203
Dear fir. Cavanaugh:
lle and our consultants have reviewed your report "Energency Feedwater Systen Upgrade Reliability Analysis for Arkansas fluclear One, Unit flo.1" which was transmitted to us by letter dated October 7,1981. We find that in order to continue our review we need additional information identified in the enclosed Request for Additional Information. We request the infornation within 30 days fron receipt of this letter.
Sincerely.
- 0RIGINAL SIGNED BY John f."5312[dief Operating Reactors Branch #4 Division of Licensing
Enclosure:
Request for Additional Information cc w/ enclosure:
See next page 8205060561 820421 PDR ADOCH 05000313 P
pop ORB 4:DL C-0 B#4:DL omcs >
l son ~-o...g.g.g.94cb..j.g? p...
..x..
l W ronu se oo-soi uncu oua OFFICIAL RECORD COPY usem e-awe
Arkansas Power & Light Company cc w/ enclosure (s):
fir. John R. Marshall Manager, Licensing Arkansas Power &. Light Company
~
P. O. Box 551 Director, Bureau of En'vironment'a1 Little Rock, Arkansas 72203 Health Services 4815 West Markham Street :
Mr. James P. O'Hanlon Little. Rock, Arkansas -72201 General Manager Arkansas Nuclear One Pr.0. Box 608 m z.n.
=. -:-
-e Russellville, Arkansas.72801..
- =
. ; : _r : n-e : -,.-
. n Mr. William Johnson
- e?
U.S. Nuclear Regulatory. Commission P. O. Box 2090 7.2801 n
2-Russellville, Arkansas Mr. Robert B. Borsum Babcock & Wilcox
^
Nuclear Power Generation Division Suite 220, 7910 Woodmont Avenue
~~-
~
Bethesda,' Maryland 20814
~
~~
Nr. Nicholas S. Reynolds i_.~
.Debevoise & Liberinan 120017th Street, NU Washington, DC 20036
~i
~
Arkansas Tech University
~
Russellville, Arkaris.as - 72801 Honorable Ermil Grant Acting Co~unty Judge of; Pope County;
~"
~
~
~
Pope County. Courthouse Russellville, Arkansas 72801 Regional Radiation Representati.ve EPA Region VI 1201 Elm Street Dallas, Texas 75270 tir. John T. Collins, Regional Adminiscrator U. S. Nuclear Regulatory Commission, Region IV 51 rling on xa s
I
-~ ~
~ ~ ;**
, - * ~ ^ -
Enclosure REQUEST FOR ADDITIONAL INFORMATION i
CONCERNING EMERGENCY FEEDWATEP. UPGRADE RELIABILITY ANALYSIS FOR ARKANSAS NUCLEAR ONE, UNIT NO. 1 DOCKET NO. 50-313 1.-
The report's results are not presented in a manner that lends itself to a "NUREG-0611-type" comparison.
For example, unavailability given LOOP is not tabulated. LOOP appears as a basic event on the fault tree; was this entered as the probability of LOOP given LMFW? How was this handled?
2.
A narrative description of cut sets was provided, but no quantitative details were given. What were the contributions from the dominant cut sets?
3.
Failure data were not given. What are they?
4; The following questions pertain to the scenario discussed in point 4 of Section 3.2.1.
At the bottom of p. 3-1, it is stated that "...AC powered valves CV-X2 and CV-X3 will not open."
(Fig. D-7 seems to indicate that CV-X3 is DC-powered.
Is this correct?)
In paragraph D.2.3 (p. D-3), it is stated that "The flow of EFW to each SG is controlled by redundant normally-open modulating solenoid motor operated control valves in parallel paths."
The fact that loss of AC fails the control valves closed suggests that they are normally closed ~, but the subsequent discussion says that they are normally open. What are the normal positions of all valves in the '
four discharge paths? In this scenario, why were they presumed closed?
Section 3.2.1.4 also talks about the o'pening of recirculation valves CV-2815 and CV-2816 during the loss-of-all AC power event which causes portion of turbine driven pump flow bypassing to the condensate storage tank. The ANO report recommends that an analysis be done to determine if adequate flow will be available to the SGs with this bypass flow.
Has the analysis been performed? How serious is the bypass in degrading the total flow to the SGs?
This raises further questions, whose direction is suggested by the following:
5.
What physical measurement (s) actually regulate the recirculation valves'.
(CV-2815 and CV-2816)? According to the text (p. D-3, paragraph D.2.5),
flow elements upstream of the control valves are used to decide whether flow is being demanded from each pump; however, the diagram suggests that the posi' ions of the control valves (C,V,X2 and CV-X3) are sensed.
6.
If actual flow is the parameter to determine the recirculation (i.e., open-ing of valves CV-2815/2816) then a simple failure of EMD pump will also cause the recirculatio.n in TD pump flow during the loss-of-main feedpump event.
This failure mode should b.e included in the fault tree.
e e
e 4.gewbe m
. i
)
7.
Pump Trip:
Under what conditions do the pumps trip? Are there trips which are supposed to operate during a test but not during an emergency?
The fault tree takes credit for operator recovery" of suction source in the event that it had been inappropriately left valved off after l
maintenance? How much time was assumed to be available for this before pump damage occurred?
8.
Maintenance: What valves are closed for pump maintenance? Are any of the discharge valves disabled for this purpose? What action is taken to isolate steam from the turbine driven pump during maintenance?
9.
Providing steam to the turbine pump: One " steam unavailable" scenario involves the following: One of the valves in the pressure reducing station fails open, and one or the other of the relief valves fails to reseat.
According to the fault tree, the relief capacity is such that there is insufficient pressure to drive the turbine if a relief valve is stuck open.
It seems likely that given a wide-open control valve, the relief valve will cycle open and closed, so that even a small cyclic failure probability will lead to a substantial overall failure probability at the valve is repetitively cycled.
Doesn 't this scenario therefore have essentially the probability of either of the control valves failing high?
10.
Isolation of the discharge paths: Are there single failures in the vector logic that can isolate both dischrage paths from a given pump?
Example:
Channel D logic can isolate both discharga paths from the turbine driven pump. Are there failur'es in Channel D (e.g., power failures) that isolate both paths?
Are. these valves commanded open by vector logic given a simple LMFW, or does the logic assume that they are already open as they are supposed to be?
~
e 9
m