ML093100463
| ML093100463 | |
| Person / Time | |
|---|---|
| Site: | Surry  |
| Issue date: | 11/04/2009 |
| From: | Geoffrey Miller Dominion |
| To: | Cotton K Plant Licensing Branch II |
| Cotton, Karen -301-415-1438 NRR/DORL | |
| References | |
| NRC GL 04-02 | |
| Download: ML093100463 (4) | |
Text
Revised Dominion Draft Response to NRC GL 04-02 RAI Question #4 Surry Power Station Units 1 and 2 (11/04/2009)
NRC Question 4 Please provide an evaluation similar to that provided for North Anna to show that the results of both Rig 33 and Rig 89 tests, and the magnitude of plant-specific conservatisms for Surry, ensure that the strainers will function under design conditions.
Dominion Response The differences in debris-only head loss testing results for the two different test rigs (Rig 33 and Rig 89) were evaluated during the North Anna Chemical Effects Audit performed by NRC staff in 2008 (Reference North Anna Power Station Audit Report dated February 10, 2009, ADAMS ML090410626). The NRC staff ultimately concluded that, although the reasons for differences in head loss for the two test rigs could not be definitively identified, the significant conservatisms incorporated into the sump strainer performance analysis bound the uncertainties associated with the different test results.
A similar case can also be made for Surry sump strainer performance; however, a modification to the Low Head Safety Injection (LHSI) pumps will need to be implemented to retain the requisite margin to conservatively compensate for the different head loss testing results for the Rig 33 and Rig 89 test rigs. A discussion of the conservatisms associated with the Surry sump strainer performance analysis and the proposed LHSI pump modification is provided below.
Sections 1.a and 1.b of the Surry 2009 updated supplemental response dated February 27, 2009 (ADAMS ML090641018) describe the extensive plant conservatisms associated with the design of the containment sump strainer for each unit. Additional margins are discussed at the end of Section 3f in the Surry 2008 supplemental response dated February 29, 2008 (ADAMS ML080650562). A summary of these conservatisms is reiterated in the Safety Case provided in Attachment 2 to this letter.
The overall magnitude of these conservatisms cannot be quantified; however, they are viewed to be significant, particularly for the factors listed below:
- 5% margin was added to the debris quantities generated from the ZOI.
- A sacrificial strainer area of 150 ft² was assigned for both the Recirculation Spray (RS) and LHSI strainer for each unit.
- The actual installed strainer area (effective) is larger than the test modeled strainer area as indicated in the following table:
Page 1 of 4
Rig 33 Test Rig 89 Test Actual Installed Strainer Modeled Modeled Station Area (Effective) a Strainer Area Strainer Area (ft²)
(ft²) (ft²)
Surry 1 RS 5584 5310 5597 Surry 1 LHSI 2040 1814 2044 Surry 2 RS 5584 5310 5640 Surry 2 LHSI 2040 1814 2091 Note: a) The actual installed strainer area (effective) equals the total installed strainer area minus the required sacrificial area of 150 ft².
- 100% debris transport was assumed for coating and latent debris.
- In both Rig 33 and Rig 89 testing, fibrous debris was conservatively prepared as single fine.
Sump strainer reduced-scale thin bed tests were initially conducted in Rig 33 to determine the total strainer surface area for each station unit. Rig 89 test loops were used to investigate the influence of chemical precipitates on the debris bed head loss.
The strainer supplier, Atomic Energy of Canada Limited (AECL), has prepared a detailed analysis report to evaluate the different results observed for the head loss tests performed in Rigs 33 and 89. The evaluation focused on the test rig configurations, flow patterns, debris compositions and quantities, debris preparation, air bubble generation, chemical environment, and debris bed formation. AECL and Dominion believe that the Rig 89 test results provide conservative evidence to verify that the installed strainer for each unit will function under short-term and long-term design conditions. Rig 89 tests incorporate lessons learned from the earlier Rig 33 testing, such as biological growth and testing fluid impurity, and consequently provide more accurate results. The AECL/Dominion testing program has concluded that the Rig 89 head loss test results are bounding and, furthermore, that test results from both rigs are conservative in that they produced higher head losses than predicted by NUREG/CR-6224.
Table 4-1 below compares the strainer debris head loss test results from Rigs 33 and 89 for debris only test configurations (no chemical effects). The short-term debris head loss acceptance criteria used for the strainer design bounds the non-chemical debris head loss test result from both Rig 33 and Rig 89. For evaluation of long-term pump NPSH margin, Surry confirmed that the Rig 89 test results with chemical effects met the long-term acceptance criteria for strainer head loss (long-term was defined as 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> to 30 days and includes the maximum effect of aluminium precipitation in the debris bed). In addition, it was noted that the differences in the non-chemical test results from Rig 33 and Rig 89 are much less than the long-term pump NPSH margins that were provided in Table 3.g-1 of the Surry February 27, 2009 supplemental response. The Page 2 of 4
key design information from Table 3.g-1 is repeated in Table 4-2 below. Again, the minimum NPSH margin in Table 4-2 accounts for a total strainer design head loss that is bounding for Rig 89 test results including chemical effects. Furthermore, the outside RS and inside RS pumps have 10.3 ft and 12.5 ft of additional NPSH margin to cover the differences between Rig 33 and Rig 89 test results. The LHSI pumps have at least 8.5 ft of long-term NPSH margin. However, the NPSH available hydraulic calculation assumes that dissolved air released in the pump can is removed and that an adequate water level is maintained inside the pump can. The Surry RS pumps have air ejectors that remove dissolved air to ensure an adequate pump can water level that does not impose a constraint on the pump NPSH margin. In contrast, the LHSI strainer long-term head loss is limited to 2.2 ft (see Table 4-2) because of a LHSI pump design constraint that is discussed below.
The Surry LHSI pumps are vertical, two-stage pumps that are located inside pump cans outside of the containment that are connected to the sump strainer by partially buried piping. For the available NPSH determination to remain valid, the water level inside each pump can must be maintained above the piping nozzle inlet to the can from the containment sump. Evaluations determined that the total LHSI strainer allowable head loss of 2.2 ft (Table 4-2) cannot be exceeded to ensure that the water level in the pump can remains above the suction pipe nozzle with adequate design margin. This limitation does not apply to the North Anna LHSI strainer as its associated pumps have air ejectors installed inside the pump cans to maintain a minimum water level in the event air comes out of solution. Currently, the limiting head loss constraint imposed on the Surry LHSI strainer is from the pump can water level analysis. If this constraint is removed, the minimum NPSH margin of 8.5 ft shown in Table 4-2 can be realized for higher strainer head losses. This NPSH margin is more than adequate to accommodate the differences between Rig 33 and Rig 89 test results and test uncertainties.
To recover the significant LHSI pump NPSH margin shown in Table 4-2, Surry Units 1 and 2 will install air ejectors on the LHSI pump cans. This modification will ensure that the water level inside the can will remain well above the suction pipe nozzle, which will preserve the calculated NPSH margin shown in Table 4-2. The air removal design scheme and installed configuration will essentially be the same as that currently used for the Surry RS pumps. A refueling outage on each unit is required to install the modification, as each LHSI pump resides in an approximately 50 ft long pump can such that significant planning and coordination is required to facilitate removal of the pumps and installation of the air ejectors on the pump cans. Consequently, a unit refueling outage is necessary to provide the time required to properly perform these activities.
The modification to install the air ejectors on the LHSI pump cans will be implemented during the fall 2010 refueling outage for Surry Unit 1 and the spring 2011 refueling outage for Surry Unit 2.
In summary, the Surry RS and LHSI strainers meet the short-term head loss acceptance criteria considering both Rig 33 and Rig 89 test results. The RS pumps have significant long-term NPSH margin to accommodate the difference in Rig 33 and Page 3 of 4
Rig 89 test results and to cover test uncertainties. The Surry LHSI pumps have adequate long-term NPSH margin to bound the conservative Rig 89 test results.
However, to increase the long-term LHSI pump NPSH margin to accommodate Rig 33 head loss test results with significant margin, air ejectors will be added to the Surry Units 1 and 2 LHSI pump cans. The other strainer design conservatisms discussed above, and in the Safety Case provided in Attachment 2, provide the basis to conclude that the LHSI and RS strainers for Surry Units 1 and 2 will satisfy their design functions and accommodate the head loss differences observed between Rig 33 and Rig 89 test results with considerable margin.
Table 4-1: Rig 33 and Rig 89 Strainer Test Results - Debris Only Strainer Rig 33 Debris Head Loss (psid) Rig 89 Debris Head Loss (psid)
LHSI 0.53 (Test S2-33) 0.11 (Test SPS-LHSI-C1) 0.24 (Test S2-35)
RS 1.0 (Test S2-28) 0.26 (Test SPS-RS-C2) 1.3 (Test S2-30)
Table 4-2: Summary of Long-term Pump NPSH Margins Pump Minimum NPSH Total Strainer Allowable NPSH Minimum Available Head Loss Required NPSH Margin (ft H2O) (ft H2O) (ft H2O) (ft H2O)
Outside RS 24.48 5.0 9.19 10.29 Inside RS 28.0 5.0 10.5 12.5 LHSI 25.37 2.2 14.6 8.57 Page 4 of 4