Fermi Case Study Flow Variance ≥ 2% Obtaining NRC relief to allow for use of a 2.78% total flow reference band for Fermi RHR Service Water Pumps ISTOG December 2010Fermi Case Study – Flow Variance > 2%Ed Cavey RHRSW PUMP B RHRSW PUMP D TO RHR RESERVOIR RHR HX B FROM D2 EDGSW FROM D2 EESW TO RHR RESERVOIR MECHANICAL DRAFT COOLING TOWERS

10 CFR 50.55a Fermi Relief Request PRR-009 “Relief from Fixed Reference Value Testing” E1151C001ARHR Service Water Pump A E1151C001BRHR Service Water Pump B E1151C001CRHR Service Water Pump C E1151C001DRHR Service Water Pump D P4400C002AEmergency Equip Cooling Water Makeup Div 1 Pump P4400C002BEmergency Equip Cooling Water Makeup Div 2 Pump T4100C040Chilled Water Pump T4100C041Chilled Water Pump ASME Code Component(s) Affected The OM Code requires during subsequent inservice testing, after the establishment of reference values, that the flow rate or differential pressure be set to the exact reference value. The Code does not acknowledge the possibility that there may be limitations in the ability to do this. This issue is discussed in NUREG 1482 Revision 1, Section 5.3, “Allowable Variance from Reference Points and Fixed- Resistance Systems.” ISTOG December 2010Fermi Case Study – Flow Variance > 2%Ed Cavey

Residual Heat Removal Service Water (RHRSW) COMBINING THIS INDICATED FLOW TOLERANCE WITH THE INSTRUMENTATION ACCURACY AND ROUNDING FOR READABILITY YIELDS A TOTAL TOLERANCE AROUND THE “SET PARAMETER” REFERENCE VALUE OF 2.78%. These vertical line shaft pumps are tested at a flow rate of 5400 gpm. Testing is performed by throttling a large 24 inch motor operated globe valve to set flow to the specified reference value. Flow adjustments using these valves are somewhat coarse and motor damage is possible with multiple start/stops. Additionally, these throttle valves (E1150F068A/B) have historic vibration damage. Test equipment has been substituted for the installed plant flow instrumentation to improve the accuracy of the system flow measurement. With the M&TE, total loop flow measurement accuracy is improved to 0.8 percent. However, with the limited throttling capabilities, a tolerance around the reference flow rate of 5400 gpm is needed. In an effort to achieve the reference value and yet not require excessive starts/stops of the throttle valve with the potential for motor damage, an optimum flow tolerance of ±100 gpm or approximately 1.85% has been determined. COMBINING THIS INDICATED FLOW TOLERANCE WITH THE INSTRUMENTATION ACCURACY AND ROUNDING FOR READABILITY YIELDS A TOTAL TOLERANCE AROUND THE “SET PARAMETER” REFERENCE VALUE OF 2.78%. ISTOG December 2010Fermi Case Study – Flow Variance > 2%Ed Cavey

NRC Staff Evaluation As stated above, Section ISTB of the ASME OM Code does not allow for variance from a fixed reference value for pump testing. However, NUREG-1482, Revision 1, Section 5.3, acknowledges that certain pump system designs do not allow for the licensee to set the flow at an exact value because of limitations in the instruments and controls for maintaining steady flow. The licensee has presented a case that each of the pumps listed in Table 1 falls into this category. NUREG-1482, Revision 1, Section 5.3 states that "[the allowed tolerance for setting the fixed parameter must be established for each case individually, including the accuracy of the instrument and the precision of its display." It also states that for a flow measurement, a total tolerance of +2 percent of the reference value is allowed without prior NRC approval. This agrees with Table ISTB , which states that the required instrument accuracy for flow rate is +2 percent. NUREG-1482, Revision 1 also states that "[for a tolerance greater than the allowed percent (which may be necessary depending on the precision of the instrument), the licensee may make a corresponding adjustment to acceptance criteria to compensate for the uncertainty, or may perform and document an evaluation to justify a greater tolerance." ISTOG December 2010Fermi Case Study – Flow Variance > 2%Ed Cavey

The NRC staff concludes that this slight increase in reference value tolerance from ±2 to ±2.78 percent for the RHRSW pumps is acceptable for the following reasons. The intent of this ASME OM Code testing is to assess the hydraulic and mechanical condition of a pump and monitor for degradation. The licensee continues to perform ASME OM Code Group A pump tests using the reference point acceptance criteria given in ASME OM Code, 2004 Edition, Tables ISTB and ISTB The larger flow tolerance will yield increased scatter in the differential pressure data. H OWEVER, THE LICENSEE PERFORMS ANALYSIS OF THE RESULTS USING A DATA NORMALIZATION METHODOLOGY TO MINIMIZE THE EFFECT OF THE SCATTER IN ASSESSING PUMP DEGRADATION H OWEVER, THE LICENSEE PERFORMS ANALYSIS OF THE RESULTS USING A DATA NORMALIZATION METHODOLOGY TO MINIMIZE THE EFFECT OF THE SCATTER IN ASSESSING PUMP DEGRADATION. In addition, any small increases in undetected degradation will not affect pump operability as the RHRSW pumps are tested at a 6.09 to 8.43 percent margin above the IST Required Action Low criteria. Furthermore, the licensee has observed that the RHRSW pumps degrade uniformly over a 12- to 15-year period, which gives adequate time to detect and correct degrading trends in performance. ISTOG December 2010Fermi Case Study – Flow Variance > 2%Ed Cavey

In conclusion: Plants can decide to implement the use of curve-based criteria with less emphasis on achieving a repeatable reference flow rate - or – Control the allowable flow band to the lowest possible variance and attempt to test at a specific flow value (may need relief if >2% total inclusive of accuracy) In either case there is a need to accurately trend pump performance - both short and long term. You must know the healthy pump FLOW to DP relationship and normalize your measured data to account for variance from the reference point. QUESTIONS???????? ISTOG December 2010Fermi Case Study – Flow Variance > 2%Ed Cavey