ML20244C648
| ML20244C648 | |
| Person / Time | |
|---|---|
| Issue date: | 04/05/1989 |
| From: | Novak T NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | Rossi C Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML19306D150 | List: |
| References | |
| NUDOCS 8904200321 | |
| Download: ML20244C648 (5) | |
Text
_
,gn;.*
- "J'* -
/
.,0, UNITED STATES
, i N-i.
NUCLEAR REGULATORY COMMISSION
,a I
WASHINGTON D. C.205b5 t
i g
j
.A 05 389 l
MEMORANLLM FOR: Charles E. Rossi, Director Division of Operational Events Assessment Office of Nut. lear Reactor Fegulation FROM:
Thomas M. Novak, Director Division of Safety Frograus Office for Analysis anc Evaluatien of Operaticral Data SLEJECT:
COMMENTS ON THE CEAFT BACKFIT ANALYSIS FOR THE PROPOSED GENERIC LETTER ON SERVICE WATER FROBLEMS
, We are pleased to learn that the proposed subfect Generic Letter plans to irecrporate cil the recorrnendations ceveloped in the AEOD case study report on service water cystem probiens (NUREG-1275, kol. 3). These reccomendations aim to significantly reouce reactor accident risks caused by failures end degrada-tiens of service water systems. These recommendations are consistent with onguing research activities on service weter system fouling by the Office of Nuciear Regulatory Research and a recc uendation by Region II for a generic communication. We believe these recommendations constitute prudent anc good tractices ter the safe operation oT the plants.
We have reviewea the craft backfit analysis for the actions:
(1) which erbrace the recommencetion that hett exchancers, which are cooled by the servie water system ed perform a safety function, be tested periodically to look i thermal cegrao
'on, ano (2) which recuire inspection of service veter system piping and ccmpont E.
The draft benefit-cost estimate in the proposed Generic Letter packroe for the octicn in olving thermal testing cf heat exchangers is 0.003 persen-rea/$1000.
Cur t: understanding is that this is baseo cn an assumed baseline core melt frecuency oT 4E-6/ry, a conditional offsite dose of TE+5 persor.-rem / evert, a core iceir reduction ef about 1", and a unit cost of S11,000/ry.
We Niieve that the baseline core melt frecuency, the ccnditienal offsite dose, and the ccre melt recuction fraction are significantly uncerestimeted, and tre cost of the action is significantly overestimated, when compared to salues in craft NUREG-1150.
First, results in existing comprehensive risk studies end in craft HUREG-1150 give a rerge ut core melt frecuercies from 2E-3/ry to aE-6/ry (see Enc 1r,sures 1 anc 2). We believe a baseline core melt f recuency
/
31
}
y,a
~
i e'
,3 Charles E. Rossi
-2 ef IE 4/ry shoulo be used, which is approximately an average velue of the plants studiec in oraft NUREG-1150, as dccumenteo in SECY-89-058, Figures 1 and 2 (Enclosure 1).
Second, the off site ecse (7E+5 person-rem / event) is underestimated when cenpareo to values from dratt NUREG-1150 which reflect uncertainties in how the containment fails. A more realistic value, from draft HUREG-1150, fcr average cor.ditional coses for a containment failure tolicwing a core treit is of the crder of 3E+6 person-rem / event.
Third, the impact of the action was calcu-lated to be 6 bout 1% and could easily be higher.
We are of the opinicn that a value of 5% can be.iustifiec. The justification is derived from the observation that f ailures and degocations of the service water system have adverse impact en a large numoer of safety systems required to prevent and mitigate serious reactor accirents.
For example, ecuipment cependent on service water ccoling are ECCS pur.rs, ECCS room coolers, errergency diesel generator coolers, RHR heat exchangers, containment spray heat exchangers, electrical equipment room coolers, and control room chillers.
Some of these ecuipment, when failed, are rot rettily recoverable; ano their failures play a very importart role in the contribution tc reactor accident risks.
Finally, the cost of concutting heet exchancer thermal testings in the drait backfit analysis should be lower.
For exemple, if the flow tests are reduced to the same frecuency as the heat exchanger thermai tests, only cre train is tested, and the calibration is eliminateo, the total cost would be redtced by approx 1mately e factor of five.
Adding all these adjustments would result ir nn increase by a factor of 2,500 in the draft benefit-cost ratio from 0.003 to 7.5 person-rem /51000. A benefit-cost ratio of 7.5 person-rem /$1000 would indicate that the proposed corrective action trerits expcolticus consideration.
Aceitiotally, the thermal testinc cf safety-related heat exchangers is consistent with General Desion Criterion 46, " Testing of Coolir? Water Systems." This particular test. like other surveillance tests on the service water system, should not be considereo a backfit, especially when the proposed test provides an irtegrateo assessment, cf the function of safety-related heat excnangt rs to verify the FSAR commitments, consistent with NRC regulations.
An examinat;cn of the operating events referenced in Appendix A of NUREG-7275.
Vol. 3, indicates tnat a2 of the 276 sig.ificant cperatino events are related to inacecuate flow end/cr heat trant.fer capabiiity of various heat excnangers supported by service water.
These events cccurred at over 20 differert plants. Of particular note is the Oconee event in 1987 when they oiscovered that there was sigt.1ficant degradation of the heat transfer capability of the I
certainment coolers and the LPI heat exchangers.
This degradation, which occurred over a reriod of many yrars, resulted in the heat exchancers not beir.g able to assure plant operatinr within the safety limits at full power. Similer heat exchanger fouling was ncted at the Catawba and McGuire plants in the same time frames.
At a rEetinc with the HRC in 1985 to discuss oegr?ced performance of the tervice water system, tre licensee nade a commitment to implement periedic j
thermal testing 07 the above referenced cooiers at Oconee.
Similar testings
7
.)
Ly..,p.. -
Charles E. Rossi I have been instituted at ANO-2 and Turkey Point because of experience with
' degraded heat exchangers.
An industry group has also-suggested that thermal performance testing be conducted to detect heat exchanger degradations.
Based on these experiences, implementation of-a thermai performance testing,
L program is essential'for assuring that plant performa.ce complies with stated safety limits.
The experiences at Oconee, McGuire, ANO-2 and Turkey Point clearly illustrate that flow measurements alone do not provide an adequate indicatio'n of whether the safety margins are maintained.
As to the proposed Generic Letter action that requires inspection of service water system piping and components, a benefit-cost estimate higher than the
_ draft backfit analysis would also be obtained.
The related AEOD recommendation in our case study report (NUREG-1275, Vol. 3) emphasizes inspection of the service water system piping where a failure would disable or significantly oegrade the service water system.
These selective inspections would dramatic-ally lower the costs involved.
Furthermore, as stated earlier, the baseline core melt frequency, the conditional offsite dose, and the core.. melt reduction-fraction should be higher than those which were assumed in the draft oackfit analysis.
These considerations-would' lead to a benefit-cost estimi..e in the same range as that involving thermal testing of service water system heat exchangers, representing an increase by approximately three orders of magnitude.
If you have questions or comments, please contact me or Peter Lam -(x24436) of my-staff.
Q%
W" Thomas M. Novak, Director Division of Safety Programs Office for Analysis and Evaluation of Operational Data
Enclosures:
As stat,ed.
cc w/ enclosures:
L. C. Shao A. C. Thadani C. H. Berlinger
' R. L. Baer A. A. El-Passioni C. V. Hodge T.. Y., Chang R. H. Bernharo i
i
2 0
" E
' 0 1
F L
' 3 U
0 G
" E D
S
' 0 N
A 1
EN R
m GO G
AI
\\'
' 4Y
^
0C T
i M U N
" E A B
- 0. E U
M
\\
D s
\\
1 Q
O I
- R i
\\
E T
E T R
T R S
/
' 50 F O
I B
O
' E H
E D
C k
- 0. G C N A
A Y
1 O
C M
E L
I A
P Z A
N N
i i,
' 6D E
s 0 E I
^
R U
E
/
" E R
., l
- 0. O Q
T C
E 1
N R
l
- F
' 7 H
0 A
Y 4
" E Y O
"' 0 R U R Q 1
U E S S
' 8 0
l
-E 5
3 5
5 1
5 0 0 3
0 2
0 0
1 1
0 0
0 d
0LATIE eH0iYkT l
l
SE G
N
- N A
_I O
~
m
- Z I
R Y
C H
N Y
A E
O l-O Ue
~
- U n
- Q a
l E
Qit i
En
- S d
e e
Rc Fe M
r P
D N
i F
E L
h A
Gt o
i.
U A
l R
5 G
9 G
Md n
a An e
a D
M M
h H O t
C ES
~
T O
~
O A
R
~
E TO O
P B
C L
A
- Y N
- R
~
J R
~
ri
- R U
E S
T N
55 -
i: !
3 4
5 6
7 I
0 0
0 0
0 E
E E
E E
0 0
0 0
0 1
1 1
1 1
co0 n AO F,,E{Enh l
!