ML20206B158
| ML20206B158 | |
| Person / Time | |
|---|---|
| Site: | Peach Bottom  |
| Issue date: | 10/31/1988 |
| From: | INTERNATIONAL TECHNICAL SERVICES, INC. |
| To: | NRC |
| Shared Package | |
| ML20206B161 | List: |
| References | |
| ITS-NRC-88-9, NUDOCS 8811150355 | |
| Download: ML20206B158 (12) | |
Text
. _ _ _
ITS/NRC/88 9 October 1988 Technical Evaluations i.
PECO's Topical Report PEC0 MS-0004 "Cethods for Performing BWR Systems Transient Analysif Peach Bottom Atomic Power Station Units 2 and 3 r
h 1
l 1
i L
4 l
l i
l l
1 i
i j
r I
International Technical Services. Inc.
l i
420 Lexington Avenue
(
I r
f i
N l
92///5Y 353 NN
o ITS/NRC/88 9 November 1988 i
Technical Evaluation:
Egg,,Ipaical Reoort PEco-FMS 0004 "tigthods for Performina BWR Systems Transient Analysis" Peach Bottom Atomic Power Station Units 2 and 3 I
I 1.0 INTRODUCTIOil In submission of its +,opical report, PEco FMS 0004 entitled "Methods for f
Performing BWR Systems leansient Analysis," (1) Philadelphia Electric Company i
(PEco) stated ;ts objectives to present descriptions a J qualification of its i
i RETRAN02 based plant model for Peach Bottom Atomic Power Station (PBAPS) i Units 2 and 3 as a best estimate, general purpose, and system analysis tool.
l l
TowarJ tFase objectives, PECo presented in the topical report start up test
{
an al.n.2 s to benchmark control systms and component performance and responses, two operations.1 transient : "ts and an NRC specified problem f
(which is a Turbine Trip i:ithout Bypass (TTWOBP) transient).
The latter was provided as comparison with analyses performed by General Electric (GE) using
)
ODYN and Brookhaven National Laboratory (BNL) using RELAP-3B with BNL TWIGL.
4 Although PEco intends to use the base model for reload analysis and licensing f
calculations, neither a statistical analysis to determine the operating limit i
l minimum CPR (MCPR) nor any discussion of conservatism were contained in this l
subject topical.
These are intended to be presented in topical report PECo-j FMS 0006 (3).
i This review, *,herefore, ovaluates the licensee's analysos performed with the l
PLCo version of the RETRAN02 MOD 004 computer code for eventual use in l
licensing analysis, but does r.ot address issues related to conservatism or to
(
l the required statistical uncertainty analyses.
The information reviewed 1
1 i
consisted of the subject topical report and supplements (2,3) thereto pr'ovided by the licensee.
2.0 EVALUATION 2.1 Elca Version of RETRAN 02 M00004 i
four modifications were made to the officially distributed version of RETRAN02 M00004.
Two of these were made to improve an input / output i
capability and the other two to improve initialization and steady state l
computations.
These latter changes were made more as user convenience 1
i features (to prevent the code from failure) due to limitations of the code I
rather than as error corrections ar.J do not impact analyses.
The licensee, as required by the license agreement, notified EPRI and changed the designation of the code to the "PEco version of RETRAN 02" to distinguish it l
from the officially distributed version.
Upon receipt of the change notice, j
the code developers examined the nature of changes and felt that since these j
were not errors in the code they do not warrant general implementation of j
these changes. We find these modifications acceptable.
l l
j The SIM1RAN E and SIMULATE E codes were also m:dified by PEco and these j
modifications are documented in PEco FMS-0005 (3) and reviewed separately.
]
Two additional changed made to the code are designed to result in the consistunt prediction of the cor9 reactivity components and initial power distribution between the 3D simulator model and the 10 RETRAN mode.
Details of these modifications will be presented ir PEco FMS 0006 (3).
L I
i With respect to the subject review, the use of the PECo version of RETRAll 02 i
is acceptable, f
i t
l 2.2 Modal Descriotion i
f Peach Bottom Units 2 and 3 start-up tests were used to b
, the nodalization, code model options and control systems and to demonstrate thi adequacy of calculational procedures and the PBAPS ba;e model.
l 3
i
[
2 1
l i
i i
J
2;2,1 Plant Nodalization 3
i l
The PBAPS base model was developed to represent the PMPS at the rated operational conditions in the steady state mode. The initia conditions were based on measured or, if not measured, assumed values.
For a given cycle, i
the base model was generally used to analyze operational transients requiring l
I only changes to the transient dependent input to the trip data or certalii
]
plant parameters when analyzing events at off rated conditions.
i j
Base Plant Nodalization
[
j PEco developed a base model for the PBAPS containing, among other volumes, a 24 hydraulic node and 24 neutronic node core, one six-volume steam line representing four steam lines, a steam separator and a non equilibrium volume f
representing the upper downcor.er.
The algebraic ", lip option and subcooled void model with one-dimensional kinetics option were selected in the core volumes.
The RETRAN separator option utilizing a non-equilibrium method was selected for a separator component representing the 211 individual steam l
separators.
A non-equilibrium pressurizer model simulates the upper downcomer region where a steam water interfact. exists.
The base model also contains two recircul tion loops driven 'y one control system.
The centrifugal pump model was used in the recirculation pump volumes and the jet pump model was selected based upon plant metsured data for jet pumps, Safety l
valves were modeled using a fill table based on a standard ASME coefficient i
of discharge calculational method, The turbine was modeled not by a fluid j
volume but through use of control logic to compute the turbine speed and
[
steam flow via a fill, 7
Vessel Nodalization t
{
fatt i
i PEco's nodalization of the reactor core is equal or more detailed than that
)
used for comparable BWR cores by other licensees.
On that basis we find the 3
r
- - -., - _ - _, - - ~ _ _ - _. - -.. _ _ _,. - - _.
j core nodalization acceptable.
1
- Kinetics Ontion I
Kinetics input data were generated to reflect the end of cycle conditions.
PEco stated that generation and qualification of the 1-0 kinetics input will be described in detail in PECO FMS 0006 (3).
The limiting transient was I
analyzed using the one-dimensional (1 0) kinetics option in the RETRAN02 M00004 version.
The 1-0 kinetics option in the RETRAN02 M00004 was a
previously reviewed and found to be acceptable for use in this context of this review (4).
Seoarator Model PEco's use of the separator model was consistent with the vendor's data for carryunder and carryover and was within the restrictions placed upon use of this. model in the RETRAN02 SER (5) relating to (a) fluid transient time through the separator and (b) on not using the model in a reverse flow mode.
We therefora find PECo's use of the separator model acceptable.
Use of Non-eavilibrium Pressurizer in Downcomer Reoion PEco's nodali:ation of the downcomer as a three voluma representation consists of a single volume for each of upper, middle and lower downcomer regions and is consistent with the need for non equilibrium modeling in the two phase region in the upper downcomer.and equilibrium modeling in thw lower downcomer regions as in both cases done by PECo.
Because of the sc, called "pancake" problem that would occur in RETRAN calculations for staded two-phase nodes, PEco.: hose not to subdivide the upper downcomer region further.
This approach is consistant with the approach used by the industry, and we find this nodalization acceptable.
PECo's us2 of the non equilibrium model in the upper downtomer region is acceptable since the use of non equilibriun modeling is necessary in that region to represent the steam flowing in from the separator and the cooler 4
. Itquid present in the lower portion of tnat volume.
However, as restricted in' the RETRAN02 SER (5), the licensee should provide additional justification for conditions with single phase fluid in the upper downcomer if such conditions occur.
Balance of Plant PEco's nodalization of the balance of the reactor system (recirculation loops, etc.) is such that each significant component or connecting pipe is represented by a separate volume. We find this approach acceptable.
1:1 flalL1idil Use of M jet pump model is restricted by the RETRAN02 SER (5) to the normal quadrant of forward flows.
Although ia PEco's analysis of the single motor-generator trip test, a reverse flow was computed to occur, due to the timing (flow reversal occurs after the time of MCPR), the small magnitude of the flow reversal, and its insensitivity to the predicted transient MCPR, the
{
licensee concluded that its use of the model have no significant impact on the determination of conservative plant operating limits.
We find this
]
argument acceptable, however, we recommend that this be checked during the PEco FMS 0006 review.
2.2.2 Control Systems i
Control systems developed by PEco are: the reactor water level calculation, feedwater control system, rectreulation control system, and the turbine l
electro hydraulic control (EHC) system.
l i
Qualification of threo of these control systems (feedwater control system, I
recirculation control system, and the turbine electro hydraulic control (EHC) f system) were presented in the topical report using start-up test data, and is discussed in detail in Section 2.3 of this riport.
The reactor water level calculator was also exercised in these start up tests.
5 1
1
)
We find that the results of the comparison indicate the PEco's models are reasonable models of the actual control systems.
We therefore find these
- models acceptable.
2.3 Model Oualification The base model described above was compared to test data for three series of start up tests, safety / relief tests and turbine trip tests.
In addition, PEco submitted results from the NRC Test Problem for a turbine trip without bypass transient as a licensing basis transient.
Since this transient had been previously analyzed by General Electric (GE) and Brookhaven National Laboratory (BNL), using their respective computer codes (GE used ODYN and BNL used RELAP-38 together with the core dynamic code BNL TWIGL) (6), PECo performed analysis using RETRAN02 H00004 to provide basis for further qualification of their plant model.
2.3.1 Peach Bottom Start-uo Tests DECO selectted three series of start up tests to qualify the control system in the Peach Bottom RETRAN plant model: feedwater system transient testi, turbihe electro hydraulic control tests, and reactor recirculation test.
The first two were performed with stand alone models and the latter test vas analyzed with the whole plant model.
Feedwater System Transients The feedwater controller calculates the steam flow to the feedwater turbines which drive the feedwater pumps.
Feedwater pump speed and reactor pressure are used to determine feedwater flow.
The feedwater enthalpy is determined as a function of main turbine steam flow with appropriate delays and time constraints to simulate feedwater heater heat transfer rates and piping delays.
However, this system model was not designed to simulate rapid transients.
The feedwater system transients were used to qualify PEco's stand alone model 6
. of the turbine-pump dynamic login.
The full plant model was used for reanalysis of one of the tests.
The recirculation M G test data were also osed to further qualify the feedwater model.
The resultant analyses indicate that PEco's feedwater system model is acceptable for mild transient.
Turbine Electro Hdqulic Control Transients j
Turbine electro w.aulic contrv transients were analyzed using a stand-l alone model of tim we, M ce9 trol logic.
The control valve servo gain used in the RETRAll t
.c. 1 devrf ned from the recorded control valve error signal.
The servo
. h, w s. nose, that best matched the recorded control valve position data.
a;wse %a eervo gain was found to vary depending on the power level and s v o the e rol valve was stroking open or closed, an average value was cho. ' f ar tic..ETRAN model.
An analysis was reportedly performed with the actuai lve.ervo gain which resulted in better agreement j
with control valve error signal data and only slight improvement in the 1
prediction of the control valve position.
The acceptatility of this model
]
for licensing analyses depends upon the transient being analyzed and PEco i
should be required to justify its use on a transient by transient basis in licensing applications.
Reactor Recirculation Transients i
l These tests were initiated by trips of one or both motor generator (M G) set drive motors; the former at Unit 2 at near full power (95% NBR) and the l
latter at Unit 3 ac near full power (94". NBR).
Due to the limited data I
available from the tests conducted at these plants, certain kinetics l
parameters were shared in the analysis.
These tests were used to exercise the reactor water level control, feedwater system and turbine EHC system.
i In all three series of tests the RETRAN computed plant parameters exhibited 1
the same trend as those exhibited during the test.
The PECo analysis and results adequately demonstrate PECo's knowledge of I
RETRAN and PEco's ability to analyze the test data and computational results.
I i
j i
I i
On that basis, we find that PEco's plant control systems (feedwater, turbine EHC and recirculation) ' t the PBAPS RETRAN base model acceptable in a best-
~ estimate sense.
However, since the current feedwater system model is not designed to simulate rapid feedwater flow excursion, the feedwater system is only acceptable for used in analysis where such excursion is not predicted.
2.3.2 Safetv/ Relief Valve Ten The tests were conducted at 42.4% rated core thermal power and 57.2% rated core flow.
The transient was initiated by the simultaneous opening of two SRVs causing a depressurization and a decrease in core power.
The plant eventually reached a new equilibrium state, then tiie SRVs were closed one at a time.
The PEco analytical results predicted trends well but the magnitude of the neutron flux was overpredicted.
PEco indicated an understanding of the source of the overprediction.
Thus PEco demonstrated the ability to model safety / relief valves and some understanding of the impact of valve behavior on the transient results.
2.3.3 Turbine Trio Tests Three tests were initiated by manually tripping the main turbine during the end of cycle 2 at PBAPS Unit 2.
This is a rapid pressurization event and is similar to a typically limiting event for reload licensing perspective.
The initial conditions for three tests were not the same but were such that the pressurization resulted in a significant positive neutron flux transient.
Since the current PBAPS RETRAN feedwater control logic is not designed to simulate the rapid feedwater flow excursions, the measured feedwater flow was input to RETRAN as a function of time for each test.
In each of the test calculations, RETRAN predicted results were corrected for sensor and sensing line delays.
The resulting computations compare well with the test data with respect to global trends, but do not predict the fine structure well due to the fact that the measured data were not filtered for 8
W noise.
Good agreement was achieved between the data and prtf.iction for all three tests for the time of peak neutron flux, but the peaks in flux were significantly over predicted.
Similar results were obtained for peak reactivity and the time of such peaks.
Resolution of these differences was deferred to PEco FMS 0006 (3).
2.3,4 HRC Test Problem i
l This transient was compared with results previously obtained by GE using 00YN and BNL using RELAP-3B together with the core dynamic code BNL-TWIGL (6).
For the sake of better comparison, where possible, the PBAPS base model was modified to conformed to the nodalization used by GE and BNL, e.g.,
the turbine bypass system was inactivated, the steam line nodalization was changeo to the GE steam 1tne nodalization consisting of three volumes.
In addition, for the same raason as above, certain key plant parameters were j
modified.
These include: core power, scram setpoints, RPS logic delays, control rod insertion timss, relief valve capacity and setpoints and delays, j
and turbine stop valve stroke time.
1 J
j This transient is defined as a turbine trip without bypass flow event with no
]
recirculation pump trip for PBAPS Unit 2 at the End of cycle 2.
I d
The RETRAN predicticas of core power and heat flux agf ee well with the GE results.
I
3.0 CONCLUSION
S PECo's topical report "Methods for Performing BWR Systems Transient 4
Analysis " and supplemental information provided by PECo in support of its submittal were reviewed.
We find that PECo has adequately demonstrated its knowledge of RETRAN02 built in models, their limitations and range of applicability.
We further 9
. find that the licensee has demonstrated its technical competence to use REfRAN02 M00004 and analyze results obtained by the use of this code for applications of the Peach Bottom plant transients.
i l
We have reviewed the PEco bas 6 model and the special purpose model use I
therein, and we conclude that PEco's nodalization and use of special purpose non equilibrium and separator models in its base model is acceptable.
The b u t model including the algebraic slip model as a best estimate analysis l
tool is acceptable.
However, this does not imply that the algebraic slip model is qualified for use by PEco for PBAPS in licensing analysis. Altnugh
)
PECo submitted results of FRIGG test in relation to qualifying the algebraic slip model as required by the RETRAN02 SER (5), we reconinend that its review be done in context of licensing analysis, i.e.,
the licensee should be required to demonstrate that use of the algebraic slip model provides adequate assurances of conservative results.
This need is further emphasized I
by the fact that in the turbine trip tests the computed peak flux, a computation which depends strongly on the two phase flow modeling as well as i
the physics, differed significantly from the data.
The licensec stated that l
resolution of these differences will be provided in PEco RMS 0006.
We recommend, further, that before any licensing analysis is accepted, the licensee submit (i) statistical analysis; (11) justification of its licensing conditions and safety margins; and (iii) the feedwater system model with j
ability to simulate rapid feedwater flow excursion such as encountered in j
turbine trip tests.
l 1
j Finally, in addition to the foregoing, detailed discussion and justification of the use of (1) the jet pump model with a flow reversal, and (2) the 10 kinetics option and therefore input generation and code modifications, were also defr.cred to PEco FMS 0006 and therefore their acceptability in licensing i
appli.ations have not been assessed, i
]
l 10 i
I 1
a
4.0 REFERENCES
1.
"Methods for Performing BWR Systems Transient Analysis " PEco FMS-0005, Philadelphia Electric Company, September 9,1987.
2.
(.etter from J.W. Gallagher (PF.Co) to W.R. Butler (NRC), "Respense to Request for Additional Information," dated June 6,1988.
3.
Letter from E.P. Fogarty (PEco) to W.R. Butler (NRC), "Response to Request for Additional Information," dated October 11, 1988.
4.
"Technical Evaluation Report on RETRAN02 MOD 003 and M00004,"
!TS/88 7, July 28, 1988.
5.
L.etter from C.O. Thomas (USNRC) to T.W. Schnatz (UGRA), "Acceptance for Referencing of Licensing Topical Report EPRI CCN 5, 'RETRAN A Program for One Dimensional Transient Thermal Hydraulic Analysis of Complex Fluid Flow System,' and EPRI NP 1850 CCM, 'RETRAN 02 A-Pro ram for Transient Thermal Hydraulic Analysis of Complex Fluid Flow Systems,'" September 4, 1984.
6.
"Analysis of Licensing Basis Transients for A BWR/4," M.S. Lu, et al., BNL NUREG 26684, September 1979.
11
,