ML20035H168
| ML20035H168 | |
| Person / Time | |
|---|---|
| Site: | Zion File:ZionSolutions icon.png |
| Issue date: | 04/23/1993 |
| From: | Farrar D COMMONWEALTH EDISON CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9305030231 | |
| Download: ML20035H168 (9) | |
Text
.
r Ccmmenwsalth Edison 1400 Opus Place Downers Grove, Illinois 60515 a
April 23,1993 f
U.S. Nuclear Regulatory Commission Washington, DC 20555 Attn: Document Control Desk
Subject:
Zion Nuclear Power Station Units 1 and 2 Response to Unresolved issues inspection Report Nos. 50-295/93002; 50-304/93002 NRC Docket Nos. 50-295 and 50-304 i
References:
(1)
D.L. Farrar letter to US NRC dated March 18,1993, transmitting Commonwealth Edison Compar.y response to NRC inspection Report Nos. 50-295/93002; 50-304/93002 (2)
G.C. Wright letter to L.O. DelGeorge dated February 19,1993, transmitting NRC Inspection Report Nos. 50-295/93002; 50-304/93002 i
Enclosed is Commonwealth Edison Company's (CECO) response to the unresolved issues identified in the referenced inspection Report. The Inspection Report contained the Notice of Violation (NOV) that cited one Severity Level IV violation and two i
unresolved issues requiring a written response. CECO's response to the NOV was fonvarded i
in Reference 1. As requested in Referenced letter 2, the two unresolved issues concerning l
the use of the pump curves for Inservice Testing (IST) and the basis for testing of several safety injection check valves is attached.
If there are any questions or comments concerning this letter, please refer them to Marcia Jackson, Regulatory Performance Administrator at (708) 663-7287.
Respectfully, j
h'!
,is s_n-D.L. Farrar Regulatory Services Manager l
t MJ: sci mjtypg Attachment l
cc:
A.B. Davis, Regional Administrator-Rill l
C. Shiraki, Project-NRR J. Smith, Senior Resident inspector-Zion Station j
9305030231 930423 g
I PDR ADOCK 05000295 O
PDR a
4 RESPONSE TO UNRESOLVED ITEM 50-295/93002-01 (DRS); 50-304/93002-01 (DRS)
PUMP REFERENCE CURVES DESCRIPTION OF CONDITION The inservice test procedures used for the auxiliary feedwater (AFH) pumps used reference curves to define acceptance criteria for flow and differential pressure (d/p), which was not in accordance with ASME Section XI Code.
The Safety Evaluation Report (SER), dated June 4, 1991, on the IST program stated that reference curves may not be equivalent to the ASME Section XI Code requirement for performance of IST for pumps and relief requests should be individually requested for each pump.
The licensee developed Pump Position-08 submitted to NRC on January 7, 1993, which stated reference values would be established with a tolerance band. A review of this position in test procedure PT-7A-ST, " Auxiliary Feedwater Pump A System Checks and Tests,"
indicated that the flow tolerance for the AFW pump was +30, -20 gpm from the reference value.
The associated d/p acceptance criteria was then based on the pump curve for the reference flow and associated tolerances. As such, the licensee was still using a portion of a pump curve and relief from the Code requirement that had not been requested as stated in the SER.
Test procedure PT-2A-ST, " Safety Injection System Test," also included a tolerance on the flow reference value for the safety injection (SI) pump; i
however, no attempt was made to adjust the flow rate during the test to obtain the flow reference value.
In this case, the d/p acceptance criteria was based only on the flow reference value without taking into account the flow tolerances.
This was based on testing the pump on the flat part of the pump curve and as such, the d/p did not vary significantly over the flow range.
Based on this inspection and in consultation with NRR, the use of reference curves for pump testing was not in accordance with the Code. This is considered an unresolved item (50-295/93002-01 (DRS); 50-304/93002-01 (DRS))
pending licensee actions to perform pump testing in accordance with the Code requirement or to submit a request of relief from the Code requirement. This item should be evaluated for all pumps included in the IST program.
e j
ZNLD/2586/2
' RESPONSE-Following re ceipt and review of the NRC SER dated June 4,1991, Zion's intent was to set-it reference value at a value that was repeatable and representative
-of good pump operation as specifically stated in Section XI.
The acceptable pump head limits were also reduced from previous Technical Specification pump head limits to ensure Zion would be in accordance with the Code. As explained in Technical Approach and Position PP-08, Zion's ECCS flow analysis limits are based on the pump head curve.
Zion used the pump head curve in conjuction with the ECCS flow analysis to determine the limits of the acceptable / alert / action ranges.
PP-08 results in Zion obtaining corresponding differential pressures which conservatively meet or exceed the requirements of table INP-3100-2.
Due to instrument inaccuracies (calibration tolerances, readability, repeatability, etc.) and system flow adjustment capabilities, any measurement is unlikely to be exactly equal to a previously obtained reading.
Zion recognized that Section XI, Article IHP-3100, does not specify nor imply any tolerance on reference values such as flow. A literal interpretation of the word " equals" as used in INP-3100 would result in an inability to complete the surveillance and ultimately a declaration of inoperability for the pumps if the plant operators were unable to exactly match the flow reference value in subsequent testing.
Another outcome of extremely tight or non-existent tolerance is the unnecessary wear and shortened expected service life of equipment.
For instance, Zion's AFH System uses motor operated valves for flow control. Up until the modification of the valves in the last refueling outage of each unit, these valves had developed poor throttle control over the years.
It was considered that extended start /stop operation of-the valve operator motors to achieve a tight flow tolerance during testing would result in early (and~
unnecessary) failure of safety related equipment.
To ensure that either the pumps were not declared inoperable unnecessarily or the equipment was not unnecessarily degraded, a flow tolerance was considered necessary although not delineated in Section XI.
ZNLD/2586/3-
L Zion imposed tolerances on the flow reference values for all IST Pumps, as
~
' explained in PP-08. These flow tolerances were determined individually by system and were based not only on known instrument inaccuracy but also on the ability of the equipment used to throttle flow.
This was important to preclude unnecessarily declaring the. pumps inoperable due to inability to establish an exact flow value.
Zion believes that the flow tolerances chosen were based on sound technical' justification.
As discussed in PP-08, Zion developed two approaches that were believed to be within Code requirements.
(1) The first (traditional) method (See Attachment 1) describes those pumps where. fixed flow limiting orifices (miniflow) allow for nearly. repeatable flow values each test without valve manipulation (ie. as found flows).
Any attempt to throttle the flow below the value obtained during the test (eg. Safety Injection pumps) endangers the pump from lack of cooling (beyond manufacturer's guidelines) and is the subject of NRC Bulletin 88-04.
This method uses a_ fixed range of differential pressure as stated in the Code with owner defined tolerances on flow.
(2) Engineering evaivations of the flow reference tolerances, for those pumps which are tested under conditions of substantial flow, revealed that the traditional analysis method (fixed d/p range) was inadequate (See ).
Most pump head curves are relatively flat at low flow rates and then become increasingly vertical with increasing flow. On some' pumps tested at substantial flow conditions. Zion found that the H
imposition of dual test limits (Code limits & ECCS flow analysis limits) as discussed in PP-08 and using the traditional-analysis method resulted in a very small band of acceptable operation.
This small band of acceptable operation was not representative of the pump's actual condition, but rather an accumulation of convergent limits.
l Therefore, the'second method is_ utilized for pumps (Component _ Cooling and
.l Auxiliary Feedwater) that are tested near design flow and have limited' i
throttle control capability. This method uses a narrow variable' flow 1
range to ensure the pump is not declared inoperable unnecessarily and that
.I the limits are conservative with respect to assessment of operational
'l readiness. What was not specifically identified in_this position was the exact limiting tolerances given for each of the pumps to' establish the minor impact this has on acceptance criteria for the pumps.
Zion judged this to be more conservative than using pump reference curves.
LZNLD/2586/4'
These-two approaches of test methodology are different from that stated in the l
SER dated June 4, 1991 for two reasons.
(1) The SER cited the pump curve as an infinite set of reference' values _for pump flow rate and head that may not be easily duplicated.
Zion's pump test methodology is easily duplicated because it is over the most limiting j
flow band (tolerance) which fits the operating characteristics of the l
system.
j (2) In addition, the SER stated, "It is important that an insitu curve is developed, or the manufacturer's curve validated, when the pump, is known to be operating acceptably.and is based on, or validr.ted by, an adequate number of measurement points." Zion's methodology does not require an insitu curve but rather the examination of pump operating characteristics t
around a very limited flow band and was considered to meet the intent of--
i the Section XI requirements.
t As the SER explained, since this testing method...(use of manufacturer or pre-operational pump head curves)... MAY not be equivalent to the Code requirements, relief should be individually requested for each pump.
Had Zion.
kept the same test methodology that was previously established, relief l
requests would have been written and curves validated as recommended in the SER.
The approach described in PP-08 allows accurate trending and maintains the 3
intent of the Code to assess operational readiness.
PP-08 clarified Zion's intended methodology of pump testing as a response to the pump testing anomaly
- 1 and # 2 discussed in the NRC SER.
In a conference call ath NRR on March 8. 1993, PP-08 and various pump test acceptance criteria sere discussed.
In general, the NRC Staff agreed with i
Zion's technical approach.
However, based upon experience with similar issues i
and reviews of other licensee's pump test methodologies, and in view of the lack of clear ASME Code guidance on this issue, NRR has interpreted from Section XI requirements that PP-08 should be amended and relief request (s) i issued for staff review. 'NRR_has indicated _that this interpretation will be t
included in the staff's SER related to PP-08 which is under development.
Upon i
receipt of the SER, Zion will amend PP-08 and develop and submit relief
.j request (s), where appropriate,; consistent with the SER recommendations.
Furthermore, in an attempt to preclude future misunderstanding in this area, f
CECO will-appeal to the appropriate ASME Code Committee (s) for developmentiand q
= promulgation of generic guidance or clarification in-this area.
j
.'In addition, Zion's replacement of the motor operated ~AFH throttle valves i
during the recently completed Unit I and 2 refueling outages has enabled Zion:
'l to. reduce the flow tolerance range to +/--10 gpm at a_460 gpm reference-(from a +30, -20 range). The AFH pump system will be discussed in greater'detall_in j
the proposed relief request.
.l l
1 ZNLD/2586/5 j
lN
~
L RESPONSE TO UNRESOLVED ITEM
~
295/93002-03; 304/93002-03 l
CHECK VALVE INSERVICE TESTING DESCRIPTION OF CONDITION
" Check valve testing, in most cases, was in accordance with GL 89-04.
Exceptions were identified and noted below. Testing to verify a valve would go to its full open position was accomplished by establishing the full flow l
l required by the respective system or by D/I" (disassembly and inspection).
l
" Verifying a check valve would close was accomplished by seat leak testing, pressurizing downstream piping and verifying no pressure increase on the l
upstream piping, D/I, or for pump discharge check valves, verifying no reverse l
pump rotation.
j l
j Check valves SI8957A-B and SI9002A-D were not being tested on a quarterly basis as required by the Code.
These valves were addressed in IST program l
I cold shutdown justifications VC-16 and VC-18, which stated that testing could i
not be accomplished on the Code frequency since the discharge pressure of the residual heat removal pumps was lower than the reactor coolant system pressure and there was no recirculation path available during power operations. A l
review of P& ids identified the SI accumulator test return line as a possible recirculation path available during plant operations for partial stroking the valves on a quarterly basis.
This will be considered an unresolved item (50-295/93002-03 (DRS); 50-304/93002-03 (DRS)) pending licensee review of this potential flow path".
RESPONSE
The NRC identified a potential flowpath whereby partial stroke testing of Residual Heat Removal (RHR) system check valves could potentially be performed on a quarterly basis.
This flowpath was not accounted for in the IST Program Cold Shutdown Justifications, VC-16 (SI8957A-B) and VC-18 (SI9002A-D).
Upon further review, Zion maintains that these and any similar Reactor Coolant System (RCS) Pressure Isolation Valves (PIVs), can only be flow tested during cold shutdown to maintain an acceptable level of safety.
The basis for this position is as specified in Technical Specification 3.3.3.F and the associated Technical Specification Bases.
4 i
I INLD/2586/6
r
~
' Each of the PIVs are required to be tested for backleakage if the RCS pressure approaches within 100 psig of system design pressure due to the potential for unseating the check valves.
This testing must be performed within 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> if flow is known to have passed through the valve and prior to entering Mode 2 if RCS pressure is reduced to within 100 psig of system design.
The partial stroke testing proposed by this unresolved item does not conform wi'h the intent of the Technical Specifications regarding the RCS PIVs in that the ability of the valves to prevent backleakage is challenged by passing flow through the valve.
In addition, the backleakage testing during unit operation would require Zion to position various spurious valve actuation group (SVAG) valves in other than their safe position for the duration of the testing.
Although this is allowed for periodic stroke testing per Technical Specification 4.8.3.A, the periodic stroke testing requires significantly less j
time than would be required to complete the backleakage testing.
Therefore, i
this backleakage testing would significantly extend the amount of time that RHR is partially disabled (and consequently unable to fully respond to a postulated accident).
Valves SI8957A-B are not defined as PIVs in Technical Specification 3.3.3.F; l
however, Zion conservatively tests these valves in accordance with this Specification since they are a redundant isolation to the PIV's SI9001A-D and SI9002A-D. The PIVs isolating the RHR system from the RCS are considered "high risk valves" as described in Technical Specification Bases 3.3.3 in that they respond to prevent an Event V accident (inter-system Loss of Coolant i
Accident).
Since SIB 957A-B provide a backup function to these PIV's, it is not considered prudent to challenge these valves by unseating them for partial stroke testing quarterly. No SVAG valves are required to be manipulated in testing SI8957A-B for backleakage, however, the suction valve for each of the RHR pumps and the RHR cross-tie valves, MOV-RH8716A-C are required to be j
closed in turn for the duration of the testing, thus potentially reducing the effectiveness of the RHR system in responding to a postulated accident.
j i
i N
9 9
b J
4 2
ZNLD/2586/7
~
ATTACHMENT 1 METHOD 1: TRADITIONAL 2B St Detest Data ACCEPi ABLE ALERT Dwg for Info ONLY - Point Flet hot Required 3532 0 I I I I I I I I
+
3498 0
/
g'g.
e
/i.
10M (f t l
/
3I90 0 I I I I i I I I i
- 9 *9 *9 *. 9 9 9
- O%%%%%%%EE%
f FL0m (spel
ATTACHMENT 2 METHOD 1: TRADITIONAL METHOD 2: SUBSTANTIAL FLOW 14 ar Detail Data sa at Detall Date ACCEPTAe(E ALEAT ACCEPTAOLE ALERT
,J:*0Ny '"Ec ' 'T T^ m a -"
aii i i i i gi%, y_h' yyX ysyy; _v,yyL yQ gssn y
1549.0 X k x
1550 8540 1530 -
p],
1520 C
e 15t0
/, /
3500
/
1890
/-/
see0 1470 3450 TOH(pst.1 TDHtost I
/
/
sa40 l
/
3430 3420 I
Keeceavsn E". g g g g i
---<hxxxxyg g
,.,0 1400
}
3390 3380 2370 33sg,0 g g ( g j I l l l l l l l 1 o o o o o o o o o o o Y
Y Y Y..... Y..
?.
w w
y;;yy,,,,,
Flow (gpol FLOW (gpel e
,-