PY-CEI-NRR-2221, Forwards Addl Info Re tornado-generated Missiles to Suppl Info Contained in Submittal Dtd 970814.Addl Info Is Provided to Aid NRC in Review of Issue
| ML20211F415 | |
| Person / Time | |
|---|---|
| Site: | Perry  |
| Issue date: | 09/26/1997 |
| From: | Myers L CENTERIOR ENERGY |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| PY-CEI-NRR-2221, NUDOCS 9710010055 | |
| Download: ML20211F415 (5) | |
Text
. _ _ _
G R
Power Generation Group Perry Nuclear Power Plant Mall Address:
216 280-691$
Low W. Myws 10 Center Road P.o. Boa 97 FAX: 216.m Vice Prescent Perry. ONo 44081 Perry, oH 44081 September 26,1997 PY-CEl/NRR 2221L l
United States Nuclear Regulatory Commission Document Control Desk Washington, D.C. 20555 Perry Nuclear Power Plant Docket No. 50-440 AdWtional Information On Tomado Missile Submittal r
Ladies and Gentlemen:
The attachment to this letter contains additional information regarding tomado-generated missiles, to supplement the information contained in a submittal dated August 14,1997 (PY-CEl/NRR 2200L).
This additional information is provided to aid the Nuclear Regulatory Commission staff in their review of thisissue.
If you have questions or require additional information, please contact Mr. Henry L. Hegrat, Manager - Regulatory Affairs, at (440) 280-5606.
Very truly yours, f&
Y?
Attachment cc: NRC Project Manager
/
NRC Resident inspector NRC Region ill State of Ohio Aw/
9710010055 970926 PDR ADOCK 05000440 iih i
i i,
i
Attachnant PY-CEl/NRR 2221L Paga 1 of 4 SUPPLEMENTAL INFORMATION ON TORNADO MISSILE PROBABILITY ANALYSES Additional information is being provided to the Nuclear Regulatory Commission (NRC) on the subject of tornado generated missiles, to aid in the review of a Perry Nuclear Power Plant (PNPP) submittal dated August 14,1997 (PY-CEl/NRR-2200L). The following supplemental information is provided:
1.
One clarification is on the acceptance criteria against which the probability calculation results will be compared. Two of the sentences in the proposed Updated Safety Analysis Report (USAR) words that were docketed in the August 14,1997 letter (see page 5 of Attachment 1), statec' (emphasis added) "the PNPP-specific acceptance criteria is that the totalprobability of tornado missiles simply striking an important system or component must be shown by analysis to be less than 10 per year", and, "Should PNPP evaluations using the TORMIS methodology provide results indicating that the plant configuration equals or exceeds PNPP's 104 acceptance criteria, then unique protective barriers will be utilized to reduce the total probability value to below the acceptance criteria ". These two statements commit to maintaining the total (cumulative) probability from unprotected systems and 4
components to less than 10 per year. The total probability value is calculated by determining specific system / component probabilities using the TORMIS methodology, then simply adding these small probability values together, and 4
comparing the resulting sum against the 10 acceptance criteria.
2.
A second clarification is how the PNPP analyses address redundant components.
The TORMIS probability analyses quantify the total (cumulative) probability that the unprotected components (whether redundant or not) would be impacted, and maintains this probability to a very low level (below 10 ). As noted above, this is 4
done by determining the Individual probabilities of systems or components being struck, then adding the various probabilities together to determine the total probability. Therefore, the probability of specific redundant components being impacted is even lower than the total probability. As noted in the NRC Safety Evaluation Report (SER) on the Electric Power Research Institute (EPRI) Topical Reports, positive (physical) tornado missile protection need not be installed "if it can be demonstrated that the probability of damage to unprotected essential safety-related features is sufficiently small."
With respect to redundant systems / components, the TORMIS analyses were performed by grouping redundant systems / components that are in the same general area, and then determining the probability of the group being struck. Therefore, the specific redundant, unprotected components were addressed in the TORMIS analyses, and, as discussed above, the probability of specific redundant components being impacted is even lower than the total probability. This represents an extremely low level of probability, and is considered to be acceptable as a means of showing that redundant, unprotected components are not susceptible to being struck from tornado missiles.
Attachment PY-CEl/NRR 2221L Page 2 of 4 it should be noted, as was done in the in:tial submittal, that the probability calculations are focused on the simple probability of " strike", versus " damage" or failure. The probability of damage to redundant components sufficient to result in common cause failure resulting from the strike would be even lower.
3.
There are several items being evaluated by the TORMIS analyses during the current refueling outage. After NRC approval of use of the TORMIS methodology and of the acceptance criteria, and upon subsequent completion and approval of the TORMIS analyses during the ongoing refueling outage, the USAR will be revised to identify the plant systems or components that were analyzed. These USAR lists will be maintained under the normal provisions of 10 CFR 50.59.
The item that was identified during the NRC Architect / Engineer inspection was the Condensate Storage Tank (CST) piping and instrumentation. These items are being evaluated to determine the associated probability; however, due to administrative controls that have been incorporated into Off-Normal Instruction ONI ZZZ-1 "Tornedo or High Winds", the CST piping and Instrumentation are not being added into the total (cumulative) probability. The other systems / components currently being evaluated and included in the total probability are physically protected from tornado missiles by Seismic Category I buildings except for limited portions that have a short run outside, or that are located near a doorway or a pipe penetration.
l Limited portions of the following systems / components are being evaluated:
Emergency Service Water system, the Division 1 motor control center (MCC) switchgear room, the diesel generators, the Control Room, Division 1 and 2 cable chases, and the spent fuel pool. As noted in the August 14,1997 letter, there are no missiles that can directly impact on irradiated fuel, even in the spent fuel pool.
4.
Point 1 in the NRC SER on the EPRI TORMIS analysis concerning tornado missiles states that tornado missile characteristics should be employed for broad regions and small areas around the site, and that the most conservative values should be used in the probability analyses. As noted in the August 14,1997 letter, the most recent USAR value for the probability of torandoes in the small areas around the PNPP site is 3.11x10" per year. The response also noted that the broad region value would be determined using the regional data in TORMIS for NRC Region I, and would be 4
compared to the local area value of 3.11x10. The most consarvative value of these two was proposed to be used in the analysis. The results of the broad region 4
calculation determined a value of 4.25x10. The TORMIS method of determining regional probability was developed to provide a more accurate probability value than had previously been available, based on a more recent database than was utilized in earlier studies such as WASH-1300, " Technical Basis for Interim Regional Tornado Criteria", published by the U.S. Atomic Energy Commission in May of 1974. WASH-1300 had published a value of 5.8x10" for a 5 degree latitude-longitude region around the PNPP site. Although it is technically justified to utilize the more current TORMIS regional probability of 4.25x10", and this would be conservative as compared to the 3.11x10" local value, the PNPP calculations are being performed using the WASH-1300 value (5.Sx10") to ensure conservatism in the calculations.
1
Attachment.
' I PY-CEl/NRR 2221L Page 3 of 4 4
compared to the 3.11x10 local value, the PNPP calculations are being performed 4
using the WASH-1300 value (5.8x10 ) to ensure conservatism in the calculations.
5.
Another area to be clarifed is the PNPP response to Point 3 in the NRC SER on the EPRI TORMIS analysis. The PNPP response noted that the analysis addresses the TORMIS reduction in tornado wind speed near the ground by injecting the potential tornado missiles into the tornado wind field at elevations above the surface of the ground. The following provides an explanation of why this is a conservative
- approach, and how it compares to the concept discussed in the NRC SER of using a higher ratio of Vo/V33 (speed at ground level / speed at 33 feet elevation).
As discussed in Section ll.E of Volume II of the EPRI-2005 Topical Report, the standard " synthesized" TORMIS model of the wind speeds was determined to be nonconservative at near-ground elevations when it was compared against several
(
other windfield models. Therefore, the report performed sensitivity studies by l
varying several parameters to increase and decrease the near-ground velocities.
These were documented in Figure 11-12 of EPRI 2005 _The recommendation of the report was to utilize parametera that increase the near-ground windfields of the synthesized windfield model so that the rotational velocity at the ground is increased to 225 mph. This results in a Vo/V value of 225/300 = 0.75. The TORMIS 33 computer code utilized for the PNPP analyses uses these parameters. It was noted in reviews of several other nuclear plant submittals that a value for Vo/V33 of 0.82 was utilized in their analyses and subsequently approved by the NRC. The parameters that would need to be revised in the TORMIS computer code used for PNPP in order to use the 0.82 value of Vo/V are not normal input parameters (i.e.,
33 the code itself would need to be revised and verified). The Vo/V value of 0.82 33 results in a velocity at ground level of 246 mph (246/300 = 0.82). TORMIS requires that the user specify the height at which potential missiles are located, and requires that a range of injection heights for the potential missiles be specified. At PNPP the potential missiles will be injected into the tornado windfield using a height range where the lower end of the height range corresponds to wind speeds greater than or equal to 246 mph. This is considered to meet the intent of the NRC SER, which desired evaluation of higher velocity profiles for the near-ground missiles.
6.
With respect to Point 4 in the NRC SER for the EPRI TORMIS analysis, further informe"on on the assumed missile density for PNPP is provided to justify the staten..it in the August 14,1997 letter that the number of missiles used is a conservative value. For the PNPP analyses, walkdowns of the actual plant site were performed to_ identify the type and number of missiles in each region of the plant vicinity, and then an additional margin of 10 percent was added on to account for
- increased missile population occurring during plant outage periods. This walkdown included the contents and the construction of warehouses, office buildings, sheds,-
trailers, parking lots, switchyards, and other miscellaneous structures, it also assumed that the primary parking lots at the site were full. The resulting missile population was determined to be approximately 74,000 missiles. This value was compared to the numbers used for analyses at several other nuclear plant sites.
j
Attachment PY-CEi/NRR 2221L Page 4 of 4 Those numbers were in the range of 25,000 to 65,000 missiles. As a result of the above, the PNPP values for missile density are conservative.
Commitments The following table identifies those actions which are considered to be regulatory commitments, Any other actions discussed in this document represent intended or planned actions, are described for the NRC's information, and are not regulatory commitments, Please notify the Manager-Regulatory Affairs at the Perry Nuclear Power Plant of any questions regarding this document or any associated regulatory commitments.
i Commitments After NRC approval of use of the TORMIS methodology and of the acceptance P
criteria, and upon subsequent completion and approval of the TORMIS analyses during the ongoing refueling outage, the USAR will be revised to identify the plant systems or components that were analyzed, These USAR lists will be maintained under the normal provisions of 10 CFR 50.59.
o
.