ML20043H897
| ML20043H897 | |
| Person / Time | |
|---|---|
| Site: | South Texas  |
| Issue date: | 06/18/1990 |
| From: | Dick G Office of Nuclear Reactor Regulation |
| To: | Hall D HOUSTON LIGHTING & POWER CO. |
| References | |
| TAC-73009, TAC-73010, NUDOCS 9006270038 | |
| Download: ML20043H897 (7) | |
Text
_
o e
t.:
June 18, 1990 1
'Docia No. 50 498 l
and 50 499 i
Mr. Donald P. Hall Group Vice. President, Nuclear Houston Lighting & Power Company P. O. Box 1700 Houston, Texas 77251 i
e
Dear'Mr. Hall:
l
SUBJECT:
INTERIM REPORT, SOUTH TEXAS PSA REVIEW:
FIRE REVIEW STATUS 1
REPORT, SOUTH TEXAS PROJECY, UNITS 1 AND 2 (TAC NOS. 73009 AND 73010) l Enclosed is a copy of ~ the preliminary evaluation of the fire occurrence portion of the South Texas Project Probabilistic Safety Analysis.
The review is being conducted by our contractor ".andia National Laboratories (SNL).
- The evaluation re) ort has a number of questions and open items which we believe can be resolved t1 rough a meeting with members of your staff, the NRC staff, and SNL'. Af ter your staff has had an opportunity to review the enclosed report and estimate the time. required to provide the information indicated by the open items, please have them contact me regarding an appropriate time for a meeting..
Sincerely, i
Original signed by a George F. Dick, Jr., Project Manager Project Directorate IV.2 Division of Reactor Projects - III,.
IV, V and Special.P_rojects Office of Nuclear Reactor Regulation t
Enclosure:
DISTRIBUTION As stated EtetW Md!
E. Peyton NRC PDR G. Dick cc w/ enclosure:
Local PDR OGC See next page D. Crutchfield E. Jordan G. Holehan ACRS(10)
J. Weihe PDIV.E Plant File r
E, Chd liah G. Kelly f
~~
IFC
- PDIV-2/LA
- PDIV-2/PM 15TV.2/D c....:.....,...g.y.....:..........
Li'.L.......
- .............:.............. n......e.....
AME. t.EPeyt W
- GDick:ye..
s:,6 rimes <
.....:................:..............t...............:..............:..............:....e..........
GTE
- 6/18 /90
- 6//P/90
- 6/4/90
_0FFICIAL RECORD COPY kf@
l Document Name: SOUTH TEXAS PSA REVIEW
[)
9006270038 900610 l
F
~
m r
'Mr. Donald P. Hall 2
cc:
Senior Resident Inspector Jack R. Newman Esq U.S. Nuclear Regulatory Commission Newman& Holt 2Inger., P.C.
P. O. Box 910 1615 L Street, N.W.
Bay City, Texas 77414 Washington, D.C.
20036 Mr. J. C. Lanier Licensing Representative Director of Generation Houston Lighting and Power Company City of Austin Electric Utility Suite 610 721 Barton Springs Road
. Three Metro Center Austin, Texas 78704 Bethesda, Maryland 20814 Mr. R. J. Coste11e Bureau of Radiation Control Mr. M. T. Hardt State of Texas City Public Service Board 1101 West 49th Street
'P. O. Box 1771 Austin, Texas 78756 San Antonio, Texas 78296 Rufus S. Scott Mr. R. P. Verret Associate General Counsel Mr. D. E. Ward Houston Lighting & Power Company Central Power and Light Company P. O. Box 61867 P. O. Box 2121 Houston, Texas 77208 Corpus Christi, Texas 78403 INPO Records Center I
1100 Circle 75 Parkway Atlanta, Georgia-30339-3064
-Regional Administrator, Region IV U.S. Nuclear Regulatory Commission 611 Ryan Plaza Drive, Suite 1000 Arlington, Texas 76011 Mr. Joseph M. Hendrie 50 Be11 port Lane Bellport, New York 11713 1
Judge, Matagorda County Matagorda County Courthouse 1700 Seventh Street Bay City, Texas 77414 Mr. M. A. McBurnett Manager, Operations Support Licensing Houston Lighting & Power Company i
P. O. Box 289 i
Wadsworth, Texas 77483
Introduction In December 1989, a plant visit and walkdown was conducted.
This visit was performed without benefit of review of the PRA (which
)
had not yet arrived).
Eased on the plant walkdown and information provided by the utility at that time, a letter requesting additional information was written.
This letter was written in December 1989.
A reply to this information request was not received until May 1, 1990.
A copy of these questions and the sTP reply is provided as part of Attachment 1.
A review of the PRA was conducted in March and April 1990.
In l
early May, once the STP response '..o the initial questions was.
l received, a review of thess responses was also conducted.
Summary and Conclusions to Date 1.
The fire report as it is currently written is unreviewable.
a.
There is conflicting information between what was prcvided by the utility in initial presentations (including the plant visit), their response to the initial questions of December, and the PRA itself.
b.
There is insufficient information to provide justification for many questionable factors that appear in the PRA.
I 2.
The screaning criteria utilized to eliminate fire areas under consideration potentially has screened dominant contributers l
to fire-induced core damage frequency.
(It must be noted that a similar screening criteria was applied to most other external events, and therefore it is not known whether other potentially dominant (non-fire) contributors.to external.
event core damage frequency have also been screened).
specific observations t
1.
The descriotion of the area and severity factor assessment
~
crocedures described in the recort raises many nuestions.
These procedures are not consistent with typical fire PRA methodology.
Based on our interpretation of this methodology in some instances these factors are potentially non-conservative.
c,.
DRAFT i
I It is stated on page 9.3-2 of the PRA that "We do not know j
exactly how the cables are routed through the fire tone".
If it was not known where equipment and cables were located in j
any fire areas, assignment of geometry and severity factors is tenuous at best.
The combined reduction factors given in Table 9.3-8 of the PRA are non-conservative in comparison with values developed in previous PRAs utilizing fire propagation code simulations.
Also, the possibility exists such as in a.switchgear fire, that these factors could both be 1.0.
Therefore, it is recommended that cable location identification within some fire-areas in conjunction with fire code runs be performed to confirm the actual area and severity ratics and determine whether the generic assumptions made in regards to these factors are at least conservative.
2.
The centrol room analysis has emeleved ruestionable and potentially non-conservative assumetions in reaards to cabinet fire scenaries, This fire area (control room) has been shown in past fire risk assessments to be one of the most important contributors to fire risk.
The STP-control room fire analysis has ignored a potentially dominant scenario which has been quantified in past PRAs.
This scenario involves abandonment of the control room with subsequent control of the plant from the remote shutdown
~
- panel,
'The PRA has also not taken credit for suppression of the fire in:the control room.
Fire testing experience indicates that unsuppressed cabinet fires damage an entire cabinet within a relatively short period of time (6-8 minutes).
In the analysis of the geometry and severity factors, it is stated on page 9.4-2 of the PRA that if a large area must be affected the severity factor must be very small.
It is known from testing that virtually any unsuppressed fire can damage all electrical components within a cabinet.
In the example on page 9.4-7 of the PRA severity factors from 7.2E-2 to 1.5E-3 are used in the. qualification of control room scenario 6.
A typical severity factor for an unsuppressed fire should be 1.0.
It also seems immaterial where within the cabinet the fire originated as long as it goes unsuppressed.
Therefore, utiliaing this geometry / severity factor for cabinet fires seems inappropriate and also non-conservative.
In derivation of control room fire frequency on page 8.5-15 of the PRA, it is stated that 83% of the total frequency
f.
DRAFT (4.08E-3/yr) is due to transient fuel.
However, on page i
9.4-1 of the PRA it is also stated that only control room fires that start inside control panels are considered.
In fact all control room fires that have occurred to date have been in cabinets.- There seems to be an inconsistency here.
If 80% of the derived fire frequency is somehow attributable to transient fuels why is it being counted in the quantification of control room cabinet fire frequency.
3.
The results have been aivan without 4ustification for the probability anslanment of various facters used in their derivation.
while many examples were found in the PRA to illustrate this point, fiv,e examples will be given as to how the report (PRA) i is unreviewable.
a.
In the screening analysis take scenario 2004-TS-01 in Table.D.6 of the PRA.
This scenario is a. fire in l
the EST-A switchgear room.
A scenario frequency of l
1.33E-3/yr is assigned.
A computer printout of equipment located within the area is given.
The impact category LLA is assigned.
It-is stated that this area is a class 1, 3 with the potential to
)
affect one or more trains of more than one system.
screening criteria 6 is used.
This screening criteria is that if the scenario frequency is. a small fraction (say 14) of the applicable plant damage state frequency then it can be eliminated from further consideration.
If one refers to page l
8.6-1 of the PRA it is stated that if a Class 1, 3
scenario is greater than 1.0E-4/yr the corresponding scenario is considered for further analysis.
An additional statement made is that based on judgement the additional failures leading to core damage contribute 10*8 If further analysis was performed it is not given in the PR7 So in this instance the information we are left with is that a fire scenario of potentially 1.33E-5/yr has either been screened without explanation on what additional random failures are required to lead to core damage or it has undergone a further screening analysis that is not given and, hence, not reviewable.
For the two factors in the scenario 2004-FS-01, fire frequency and the probability of additional random failures required to lead to core damage, no justification of the methodology employed in quantification is given.
In the case of fire
?
4 l
I,,,
i
~
DRAFT frequency, Table 8.5-2 of the PRA lists frequency and other factors that were probably employed in derivation.
However, no example in the PRA is given to allow review of the fire frequency derivation procedure.
For the probability assignment of additional random failures required to lead to core damage no justification is given at all.
Hence, no
]
review is possible of even these relatively sinple frequency derivations.
- b. _ Referring to page 6 of Attachment 1, it is stated by STP that fire areas are not screened by application of a frequency screening criteria of 2.0E-7/yr.
However, on page 9 3-4 of the PRA the second level of fire screening is stated to be at a frequency of 2.0E-7/yr.
In Table 9.3-9 of the PRA, 11 out of the 24 andstates are apparently screened using this criteria.
While this criteria by itself has not conpletely eliminated the fire area given in the example, it did significantly contribute to its elimination in that approximately 50% of the endstates were screened.
c.
Referring to page 6 of Attachment 1 it is stated that the total core damage frequency for fire initiated events is 5.06E-7/yr.
However, on page 9.4-23 of the PRA a simple summation of control room core damage frequency yields s.02E-7/yr.
- Also, during the plant visit yet another value for total fire-induced core damage frequency was given.
d.
It is stated on page 9 of the Attachment 1 that detailed exanples of each dominant fire scenario are given.
Since the screening criteria eliminated many potential scenarios at orders of magnitude greater than their ultimate total fire-induced core damage frequency, it cannot be known whether this statement is accurate.
a.
On page 9.3-4 of the PRA endstates are= quoted as being screened at 10% of the equivalent internal event frequency.
However, on page 7 of Attachment 1, STP states that fire areas are not screened by application of this criteria.
Referring to the example given in Chapter 9 of the PRA, approximately one-third of the fire area andstates are screened by this (10%) criteria.
It must be noted that typically much more chance for recovery exists for internal event failures as compared to fire-related 2
L.
i l..
DRAFT l
failures.
Therefore, the potential exists that if further development of the fire scenarios occurred L
there relative contribution with respect to sin.ilar l
internal events endstates might significantly be altered.
Rece nandatiems 1.
A plant visit is highly recommended to walkdown a number of fire areas which were screened from further analysis.. This would allow for e more robust review of some of the potentially questionable and non-conservative plant-layout dependent factors employed in this analysis.
2.
The fire PRA as currently written is unrevievable.
More i
adequate documentation needs to be provided in the fire PRA to make any meaningful review possible.
1 t
O l
4 i
4
,.--y.
.,.-_,--,...~,,.--.,_,-4-..._-,-
- - - -