ML20198H317
| ML20198H317 | |
| Person / Time | |
|---|---|
| Site: | Prairie Island  |
| Issue date: | 09/15/1997 |
| From: | Wadley M NORTHERN STATES POWER CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| Shared Package | |
| ML20198H322 | List:
|
| References | |
| GL-96-06, GL-96-6, TAC-M96854, TAC-M96855, NUDOCS 9709180001 | |
| Download: ML20198H317 (14) | |
Text
.m.. _ am wg m
Northern States Power Company Prairie Island Nuclear Generating Plant 1717 Wakonade Dr. East Welch, Minnesota 55089
=
M f
September 15,1997 Generic Letter 96-06 l
i 11 1
U S Nuclear Regulatory Commission 3 Attn: Document Control Desk 3 Washington, DC 20555
}
PRAIRIE ISLAND NUCLEAR GENERATING PLANT Docket Nos.50-282 License Nos. DPR-42
]j 50-306 DPR-60 5 i
Response to Request for Additional Information Related to Generic $
Letter 96-06, " Assurance of Equipment Operability and Containment Integrity 3 During Design-Basis Accident Conditions"(TAC Nos. M96854 and M96855) {
i 4
The purpose of this letter is to respond to the NRC request for additional information, J
dated August 14,1997. Our response is attached to this letter, k f
With this letter we have made no new NRC commitments. Please contact Jack Leveille f (612-388-1121, Ext. 4662) if you have any questions related to this letter. A e
a
-e
$c vb '
\
f k Michael D Wadley i Vice President j.
Nuclear Generation -
j p;} }
'l c: (see next page) h'* 7" '
il n
3 g &O lll,El{I${0!fll[III,l!!I[$II.!I 9709190001 970915 t PDR ADOCK 05000282 l!"
P PDR k k
US NRC
. Page 2 -
c: Regional Administrator - Region ill, NRC Senior Resident Inspector, NRC NRR Project Manager, NRC J E Silberg (without enclosures)
Attachments:
- 1. Affidavit
- 2. Response to RAI for Staff Review of GL 96-06, which contains enclosures as noted OL9606,2. DOC ta h i g _ _ _ _ _ _ _ _
l i
UNITED STATES NUCLEAR REGULATORY COMMISSION !
l NORTHERN STATES POWER COMPANY PRAIRIE ISLAND NUCLEAR GENERATING PLANT DOCKET NO. 50-282 50-306 l
GENERIC LETTER 96-06, Assurance of Equipment Operability and Containment Integrity During Design-Basis Accident Conditions i Northem States Power Company, a Minnesota corporation, with this letter is submittirig information requested by NRC Generic Letter 96-06.
i This letter contains no restricted or other defense information.
4 NORTHERN STATES POVER COMPANY BY o h4 Michael D Wadley Vice President Nuclear Generation On this Nay ofs IIW14 nbMA -
,p [ before me a notary public in and for said County, personally appearef Michael D Wadley, Vice 9 resident, Nuclear Generation; and being first duly sworn scknowledgeo that hi is authorized to execute this document on behalf of Northern States Power Company, that he knows the content hereof, and that to the best of his knowledge, information, and 4
belief the statements ade ' it tru that it is ot interposed for delay.
v , ---
MAACIA K. LaCORE d
@ NOTARYPUBUC4NNNESOTA HENNEPN COUNTY i i;
. 3. ..
_ND_3I.f " 3 E 1
CL9606,2, DOC
Resoonse to RAI for Staff Review of GL 96-06 <
This report is provided in response to NRC Staff Request for Additional Information for Review of the Two Phase Flow and Waterhammer Issues in GL 96-06, dated August 14,1997. The information requested is attached and described below. The discussion below describes the analyses methodologies, inputs and assumptions (Information Request No.1). The specific requested analysis and diagrams are noted in the list of enclosures and are included (Information Requests No. 2 and 3).
- 1. Summary To date, the analyses described below have confirmed the previous judgments and conclusions described in the 120 day response to Generic Letter 96-06.
The Cooling Water System is considered operable and capable of performing its design basis functions.
Actions which have yet to be completed to finish addressing these issues are as follows:
Complete piping and pipe support analyses for the piping associated with the CFCUs for the potential hydrodynamic loadings. To date, these analyses have shown that the loadings on the piping and supports are within the acceptance criteria in the USAR. If any situations are discovered where the loadings exceed the USAR acceptance criteria, a prompt operability determination will be made and modifications implemented to correct the condition.
- Run the two phase hydraulic model with initial pressures and flow corresponding to the Cooling Water Pump operating at 93% of the IST Pump Curve. The current analysis has been performed to confirm operability at the present pump performance capability. This additional analysis at reduced Cooling W >. r Pump performance provides the limiting case for evaluating th< ? CU heat removal capabilities.
The following actions are being considered to enhance system performance.
- Complete the modification to the control logic for the Cooling Water isolation valves for non-essential Turbine Building loads. This modification will increase the margin currently available in the Cooling Water System and subsequently reduce the two phase flow conditions at the CFCUs.
J
Att:chment 31pt;mber 15.1997 Page 2 Evaluation of methods to move the two phase flow condition outside of containment are being performed. Further efforts are contingent on the results of these evaluations.
- 2. Cooling Water System Response to a Safety injection Signal
- a. Cooling Water System Automatic Response The attached figures (B35-01 and D35-23) provide a simplified system schematic suitable for the purposes of this discussion. Additional details are provided on the attached system Piping and Instrument Diagrams. The purpose of this discussion is to provide background information regarding system realignment in response to a Safety injection (SI) sigital; which would be generated within the initial seconds of a large break LOCA. For the purposes of discussion, a coincident Loss of Off Site Power (LOOP)is assumed.
The Cooling Water System is designed to provide cooling water supplies to auxiliary feedwater pumps, diesel generators, control room chillers, component cooling heat exchangers and containment fan coil units. The des!gn basis for the Cooling Water System is that one pump supplying water to a single header will provide sufficient water for one unit in post accident mitigation and the other unit in hot shutdown.
As shown on Figure B35-01 the Cooling Water System is normally operated with both main headers cross connected through the four isolation MOVs in the Screenhouse (MV-32034,32035,32036 and 32037) and the two isolation MOVs in the Auxiliary Building (MV-32144 and 32159). On a SI signal, two of the four MOVs in the Screenhouse reposition to align 121 Motor Driven Cooling Water Pump (MDCLP) to the accident unit and split the main headers. For example, if the accident is in Unit 1, MV-32036 and MV-32037 close. Also, due to the SI
, signal, the two MOVs in the Auxiliary Building close to split the headers. This i alignment provides two independent trains of Cooling Water; each of which is
- capable of providing sufficient cooling.
l As shown on Figure B35-23, during normal operations, cooling to the
! Containment Fan Coil Units can be provided by either the Cooling Water System (safeguards supply) or the Chilled Water System (non-safeguards supply).
Specifically, the attached figure shows the alignment for the Chilled Water l System to the CFCUs. On the SI signal, the Control Valves are repositioned as follows (if Chilled Water is initially lined up to the CFCUs):
- Chilled Water Supply and Return Valves (CV-39401 through 39404 and CV-39409 through 39412) reposition to realign Cooling Water to the CFCUs.
GL9C06_2. DOC J
Attachment S:pt:mber 15,1997 Page 3
+
Valves CV-39405 through 39408 close to isolate the Shroud Cooling Coil
- Assembly Valves CV-39201 and 39203 open to allow full flow through the CFCUs This alignment on the SI signal ensures that the water supply to the CFCUs is from a safeguards water supply.
The Cooling Water System has five pumps, two non safeguards and three safeguards. The two non-safeguards pumps are normally operated, with the three safeguards pumps in standby. During certain times of the year, as intake water
, temperature increases, it is necessary to operate a third pump to support unit operation (121 MDCLP). The non-safeguards pumps are powered frora non safety related buses and would be de-energized on a loss of off site power (LOOP).
Two of the three safeguards pumps have their own independent diesel engine drivers (referred to as DDCLPs). The third safeguards pump is motor driven (121 MDCLP) with the safeguards Unit 2 Emergency Diesel Generator back-up power supply (Emergency Diesel Generator D5 or D6). All three safeguards pumps automatically start on the SI signal generated by the accident condition. When
, both DDCLPs start the 121 MDCLP is automatically tripped if running or blocked from starting if idle. As described above, the design bases for the Cooling Water System is that one pump aligned to one header is sufficient for the heat loads of one unit in post accident mitigation and the other unit in hot shutdown.
I If one of the DDCLPs fails to automatically start, the 121 MDCLP starts and is automatically aligned to the header associated with the accident unit. In addition, per Technical Specifications, it is permissible to pre-align the MDCLP in place of a DDCLP without entering a LCO. In this instance, valves are positioned such that the MDCLP and the operable DDCLP each supply separate headers.
- b. Sequence of Events On a LOCA+ LOOP scenario the CFCU fans and the operating non safeguards cooling water pumps would lose power. The following are the expected key time parameters during the initial time period following the accident:
CL9406). DOC
Att:chment S:pttmber 15,1997 Page4 Ilme Descriotion 0 sec LOCA + LOOP 2.4 see SI Signal 4 - 5 see DDCLP start (typical)
~ 10 see Containment temperature at 250F SI + 10 see EDG up to speed and voltage
~ % i:,ec Containment temperature at peak v6 if 270F S4 + 20 see 121 MDCLP start (if one DDCLP falls b. aart)
S' + 25 see FCU fans start During plant startup testing, the time delay associated with the low pressure start of the DDCLPs was set at 15 seconds to preclude water hammer due to column rejoining. This is longer than the time period for typical DDCLP start time in response to a SI signal shown above. Therefore, if both DDCLPs start, hydrodynamic loadings due to water column rejoining would not be expected. As discussed later, the hydrodynamic loads are calculated assuming column rejoining does occur.
O The diesel driven cooling water pumps have historically been reliable machines, and can be expected to be available to perform their post accident mitigation function. Reliability information for the DDCLPs is as follows:
Failure to Start 4.46E-3 per demand Failure to Run 4.85E-3 per hour These values indicate that these are very reliable machines, and that failure is not anticipated, in addition, it is not a common practice (except for Preventative Maintenance) to remove a DDCLP from service; i.e., the normal alignment for the system is to maintain both DDCLPs operable. For these pumps, the reliability centered maintenance program has a target out of service time of seven days (168 hours0.00194 days <br />0.0467 hours <br />2.777778e-4 weeks <br />6.3924e-5 months <br />) per year.
- 3. Two Phase Flow and Water Hammer Analyses The following discussion provides a summary of the methodology used and the conclusions from the analysis of two phase flow and hydrodynamic loadings in the Cooling Water System.
Initially a detailed assessment was performed of different water hammer mechanisms which could exist in the Cooling Water System under both subcooled and two phase flow conditions. This assessment was performed to GL9606.2. DOC j
Attichment srptImbe' 15,1997 Page5 investigate the potential severity of postulated water hammer events during mitigation of design basis events. This assessment is provided in the attached report FAl/96-89. Based on the findings in this assessment, it was concluded that the limiting hydrodynamic loadings result from the column rejoining phenomena in the piping associated with the CFCUs (see FAl/96-89 for further details regarding the conclusions). This assessment also looks at the potential for non-condensible gases coming out of solution either during the voiding or steaming phase and concludes that the calculated hydrodynamic loadings (FAl/96-77) are very conservative.
During non LOCA (e.g., LOOP only) and smaller break LOCA events the energy transferred to the CFCUs from the containment atmosphere is not sufficient to cause two phase flow downstream of the CFCUs. Voiding would occur due to system draining (column separation) until flow is restored. During the refill of the CFCUs, the flow would be in a single phase subcooled state. Hydrodynamic loadings in the system could occur due to the subsequent column rejoining. For these cases, the resultant pressure pulse is determined in FAI Calculation FAI/96-77," Assessment of Prairie Island Fan Cooler Piping Loads." This evaluation reviewed the piping configurations to evaluate the possibility of void formation, and assessed the nature of the piping and fan cooler refill to determine the potential for producing a stratified condition of steam and subcooled water. The water velocity in the pipes was based on the design conditions for the CFCUs. The refill velocities are sufficient (based on the Froude Number) to ensure that the horizontal lines run full during refill. The calculation presumed that the vertical downward flowing sections of piping would flow in an annular pattern and fill from the bottom up; which results in void collapse near the top of the pipe section. The calculated pressure pulse in the eight inch pipe is 145 psi using the Joukowski equation. The methodology used in this part of the evaluation provides reasonable assurance of a conservative estimate of the hydrodynamic loads. These hydrodynamic loads are then used in the analysis of the loadings on the pipe supports. Further experimental work indicated that the vertical downward flowing sections of piping would also run full; ,
which would preclude the column rejoining in these sections. However, for conservatism, the piping and support analysis assumes these column rejoining events occur at these locations.
Experiments were run using scale models of a representative configuration for the Prairie Island CFCU piping (FAI Report Numbers 96-89,96-107 and 97-88). The scale models were constructed to evaluate the following conditions:
- 1) column separation could occur once the imposed flow rate was removed,
- 2) steam generation would be added to the piping configuration,
- 3) a significant loop seal exists in the simulated fan cooler,
- 4) a significant size loop seal exists at the bottom of the piping, CL9606_2. DOC l
1 l
Att chm:nt Siptimb:t 16,1997 Page 6
- 5) the supply piping can experience column separation as well as some drainage, and
- 6) water refill rates for Froude numbers typical of those of interest in the Prairie Island Cooling Water System.
Using the scale models, several input parameters (steam flow rate, water refill rate, voiding time, pipe size) were varied to determine the relative sensitivity to the resultant water hammer. Water hammer activity was exhibited by the experimental results; however, these results demonstrated that the calculated hydrodynamic loads used in the piping support analyses (FAI No. 96-77) are conservative by a factor of 3 ta 20 compared to the experimental results. This provides objective evidence regarding the conservative nature of the loadings used in the piping and support analysis. The justification for these differences are described in FAl/96-89 and are attributed primarily to the evolution of non-condensible gases out of solution during the voiding / steaming phase.
During a design basis large break LOCA, the heat transferred to the CFCUs is sufficient to boil the water in the CFCUs and cause two phase flow in the return piping from the CFCUs to the main Cooling Water System return header. In order to evaluate the effects of two phase flow conditions, a transient thermal hydraulic model (TREMELO) was developed for Prairie Island which is capable of modeling the increased resistance due to the two phase flow condition. The TREMELO model incorporates the momentum transfer process which cause waterhammer, the transient heat transfer, and steam generation in the fan cooler coils. The model evaluates displacement of water due to steam generation in the CFCUs and associated piping, the energy transfer to the supply and return piping and the influence of two-phase frictional and momentum pressure drop including those which could lead to two-phase critical flow. The modelis benchmarked with relevant data in these areas, including water hammer events. The following inputs were used to evaluate a bounding case with respect to the potential for two phase flow at the CFCUs for the current operating condition of the Cooling Water Pumps:
LOCA at peak conditions (maximum heat input to the FCUs)
Loss of Otf-Site Power CFCU fouling factor of 0.0000 (maximum heat transfer capability)
One Cooling Water Pump per header (headers separated)
Loss of Instrument Air System (results in maximum flow demand on the system)
At present there are no flow limiting devices (orifice, throttle valve, etc.) used during post accident mitigation to provide back pressure for the water flow through the CFCUs. Orifices are installed in the return lines from the CFCUs. However, during an accident, an air operated bypass valve provides a flow path around the GL9606 ,2, DOC
Att:chment S:pt:mb:r 15,1997 Page 7 orifice. This bypass valve opens on the SI signal to maximhe flow through the CFCUs.
Two trains of CFCUs, consisting of two CFCUs each, are located within each containment. To simplify the two phase flow analysis, a train of FCUs are modeled as a single unit located at the highest elevation of the two units. The current analysis models both FCUs 22 and 24 as a single unit and FCUs 21 and 23 as a single unit. Using the elevation of the highest cooler for the analysis of the cooler circuit is conservative since it is more conducive to column separation and two phase flow.
The two phase flow model (TREMELO) was initially benchmarked against the results from the single phase Cooling Water System hydraulic model for Prairie Island (PROTO-FLO) with no heat input to the CFCUs to verify accurate development of the model. These results are indicated within the attached analysis. The methodology and inputs used in the model are noted therein. The results in the analysis substantiate the preliminary results discussed in the response to Generic Letter 96-06.
The TREMELO model is used to evaluate the system hydraulic response during the transient from flow cessation (due to the LOOP), to flow reinitiation, then to steady state conditiens. Transient analysis is performed with time period of 5 and 30 seconds from pump stop on the loss of power to flow reinitiation. The time period for flow reinitiation is varied to evaluate the different possible pump combinations; i.e., typical starting time for the DDCLPs is on the order of 5 seconds, and 121 MDCLP will be up to speed within 30 seconds, in addition to the transient analyses, steady state analyses are performed varying several of the inputs (train of CFCUs, orifice valve position, and CFCU fouling factor).
These results are attached.
The presence of the steam in the CFCUs and associated piping could generate hydrodynamic loads during refill and steam bubble collapse. During the refill and subsequent two phase flow condition from the CFCU, the steam flow from the CFCU is sufficient to clear the water in the downstream piping. This precludes the buildup of a stratified condition of steam flowing over subcooled water, and only small water hammer events would occur during the voiding. This was reviewed in detail during the experiments to examine the possibility that certain conditions could lead to significant two-phase instabilities that would cause substantial hydrodynamic loads. From these studies it was concluded that the loadings generated during this phase would be less than loadings generated during column rejoining. This is discussed in detailin FAI Report 96-89.
Interaction and mixing of a two phase flow mixture with the subcooled water in the return header was also considered. Comparing the transient plant conditions GL9606,2. DOC 1
Attachment SIptrmbit 15.1997 Page 8 to such mixing data shows that quenching would occur within one to two diameters of the mixing location and that the mixing zone would behave as a quasi-steady state condition. Thus, while some small pressure pulses could be expected from steam condensation, these are very small compared to that calculated for the column rejoining event. See attached discussion dated January 27,1997," Mixing of Two Fluid Streams With Different Enthalples in the Return Header," for further details.
During normal operation, cooling water is provided to non-safety related shroud coolers inside of containment. The water supply to the shroud coolers comes from a four inch tap at the eight inch supply line to one of the CFCUs inside of containment. The return from the shroud coolers is at the eight inch CFCU return line inside of containment. Isolation valves automatically close on the Safety Injection signal to isolate the non-safety related four inch piping from the safety related eight inch piping. The four inch piping from the eight inch header to and including the isolation valve is safety related. During the loss of flow, the CFCU voids and steam is pushed into the supply and return piping. This steam could migrate into the eight inch supply pipe past the four inch tap. Thus, steam may intrude into and accumulate in the four inch line to the isolation valve. With flow reinitiation the possibility exists for the steam to be trapped by the cooler water. This could result in a water hammer event at the closed four inch valve.
These loads would then be transmitted back into the eight inch header; however, due to the pipe sizes and associated transmission factors, the water hammer loads transmitted back into the eight inch piping would be less than those calculated for the water column rejoining in the eight inch header. The pressure pulse due to the void collapse at the four inch isolation valve could be larger than .
the pulse transmitted into the four inch line from the waterhammer in the eight inch line,- The hydrodynamic loadings from the larger pressure pulse are used in the piping and support analysis for tese specific sections of piping. Based on the experimental evidence, these calculated pressure pulses are very conservative, This is further discussed in the attached discussion, dated
- Januar/ 24,1997,"The Role of Piping Tees on the Supply and Return Headers."
- 4. Conclusion in conclusion, the Cooling Water System is considered operable and capable of performing its design basis functions. As noted in the Summary section of this report, the analysis of the piping and supports is not yet complete. During the completion of these evaluations, if any situations are discovered where the
- loadings exceed the criteria in the USAR, a prompt operability determination will be made and modifications implemented to correct the condition.
GL9606,2. DOC
Att:chment S:pt:mber 15,1997 Page 9
- 5. Enclosures Prairie Island Training Center Schematic B36-01, Cooling Water System.
Prairie Island Training Center Schematic B35 23, Fan Coil Unit Chilled Water Supply.
Calculation No. ENG-ME-301, FAl/96-77," Assessment of Prairie Island Fan Cooler Piping Loads," Revision 0, dated January 31,1997.
FAl/96-107, " Experimental Data to Simulate Possible Water-Hammer Loads in the Prairie Island Service Water System for DBA Conditions."
FAl/97-88, " Verification Experiments for Waterhammer Events in Power Plant Service Water Systems."
FAl/96-89. " Evaluation of Possible Water-Hammer Loads in the Prairie Island Service Water System for DBA Conditions."
FAl/97-68, " Prairie Island Fan Coil Unit Analysis using TREMOLO 1.01,"
Revision 0, dated July 28,1997.
FAI Mem :randum, Bob Henry to File, dated January 24,1997, "The Role of Piping Tees on the Supply and Return Headers."
FAI Memorandum, Bob Henry to File, dated January 27,1997," Mixing of Two Fluid Streams With Different Enthalpies in the Return Header."
Cooling Water System Piping and Instrument Diagrams.
Cooling Water System Isometric Diagrams.
CL9604,2. DOC 1
- _ _ _ _ _ _ _ _ i