IR 05000272/1984010
| ML18092A479 | |
| Person / Time | |
|---|---|
| Site: | Salem  |
| Issue date: | 01/30/1985 |
| From: | Martin T NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| To: | Uderitz R Public Service Enterprise Group |
| References | |
| NUDOCS 8502080410 | |
| Download: ML18092A479 (12) | |
Text
Docket No Public Service Electric and Gas Company ATTN:
Mr. Richard A. Uderitz Vice President - Nuclear P. 0. Box 236 Hancock 1 s Bridge, New Jersey 08038 Gentlemen:
JAN301985
\\E f\\LE COP'f Subject:
Combined Inspection Nos. 50-272/~nd 50-31~
This refers to your letter dated November 30, 1984, in response to our letter dated May 5, 198 Thank you for informing us of the functional description of the emergency dose assessment model documented in your lette This information will be examined during a future inspection of your licensed progra *(our cooperation with us is appreciate
Sincerely, Oric;ir~d Sigr..ed By:
fen 4 le( K.&-//;;!Ji if 1t*?. Thomas T. Martin, Director Division of Engineering and Technical Programs cc:
R. L. Mittl, General Manager - Nuclear Assurance and Regulation J. M. Zupko, Jr., General Manager - Salem Operations E. A. Liden, Manager - Nuclear Licensing and Regulation C. P. Johnson, Manager - Quality Assurance Nuclear Operations P. M. Krishna, Manager - Nuclear Review Board M. J. Wetterhahn, Esquire R. Fryling, Jr.,.Esquire Public Document Room (PDR)
Local Public Document Room (LPDR)
Nuclear Safety Information Center (NSIC)
NRC Resident Inspector State of New Jersey 8502080410 850130 PDR ADOCK 05000272 G
PDR OFFICIAL RECORD COPY RL50-272/84-10 - 0001. /15/85
Public Service Electric and Gas Company bee:
Region I Docket Room (with concurrences)
Senior Operations Officer DPRP Section Chief J. J. Hawxhurst RI:DETP Haw'\\~/ ms 1/'lt"V/85 RI~TP tin 1/ /85
~
RI:DETP Bellamy 1;1J\\j35
OFFICIAL RECORD COPY J/J.N u***,.,
v
- C* ~)
Harpster 1/;:>6'85 RL50-272/84-10 - 0002. /15/85
OPS~G Nuclear Department November 30, 1984 U.S. Nuclear Regulatory Commission Region I 631 Park Avenue King of Prussia, PA 19406 Attention:
Mr. Thomas T. Martin, Director Division of Engineering and Technical Programs Gentlemen:
RESPONSE TO NRC COMBINED INSPECTION 50-272/84-10 AND 50-311/84-10 SALEM GENERATING STATION FEBRUARY 27 TO MARCH 2, 1984 DOCKET NOS. 50-272 AND 50-311 Attached is a functional description of "MIDAS" which is a refined dispersion model for emergency dose assessmen This response meets our commitment of a request regarding item 4a of your inspection letter dated Ju~e 8, 198 This description reflects the current status of the system as well as its future development and incorporation into our emergency dose assessment progra CC Mr. Donald C. Fischer Licensing Project Manager Mr. James Linville Senior Resident Inspector The Energy People
Sincerely, E. A. Liden Manager -
Nuclear Licensing and Regulation
Tf.CHNICAL Df.SCRIPTIOt..: Of THE COr.\\P\\1TFRI ZF.n M[TFOROL()GICAL n~FORMP..TION N~fl nns ASSF.SS~if~:T SYSTFr'.
( t1In~s)
AT SALnl l,f-~~~ERl..Tit:c::
STATim~
f'ose assessments in the event of an er-*er':.)ency at the the Salem Generating Station are acc01"'1f)lishec h]' usinc; a co011 1uterizerl system on a tir:e share basi Meteorolot;Jical ar.d Rndiological
~ata are entered manuall Site meteorological and ~lant radiation data will be accessed directly by the computerized dose assessment model when installed on the station computers (discussed be low) *
The software package is referred to as
~lI DAS (for ~eteorological Information and Dose ~ssessment ~ystem) and is provided by Pickard, Lowe and Garrick, Inc. (PLG).
Two plume dispersion models are availabl One that utilizes the straightline Gaussian technique and a second that estimates plume trajectory utilizing a segmented approac Results are displayed in tabular or graphic form as do~e rates or dose projections such that comparison with the EPA Protective Action Guides (PAGs) can be mad HARDWARE The software package is presently located on a VAX 11/750 located at Digital Graphics Inc., in Rockville Marylan Communication between the host and user terminals is conducted by dedicated phone lines, with dial-up, back-up capabilit By December 31, 1986, the software package will be operational on a Systems Engineering Labs, (SEL) computer, located in the Technical Support Center, Salem Sit Disk storage will be on a fixed disk and magnetic tape drive will be provided for data archive functions and systen utilitie The con~uter runs in a multi-tasked environment under a virtual memory operating system which can accommodate many users ~imultaneously via local or
rer.iote terminal A Tektronix 4115 19" color graphics terminal with hard copy devices (color and grey scale) is provided in the P.O In addition, a Tektronix 4107 13" color terminal and a yrey scale copier are provided in the Control Roo~ and TS All terrr:inal co:TW!Unic2tions o~'erate at a speed of 1200 haurl or greate G FY F"R l~ L SOFTV.. i\\R F CH 1\\ RACTt:R I STI C ~
Syster operation~ is rrovided by a us0r friendly menu concept and is conf iuureci for rapid calculation of off-site dose site specific para~eters are storerl in disk files under system manager contro Dispersion and dose calculations can be initiated by a single operator action on the CRT keyboar The Gaussian straight line model is used for initial dose estimates and projections while the plume segment model can be used for more refined estimates of plume location, dose rates and integrated dos Results are displayed on site maps extending to a 50-rnile distance from the plan Maps can be stored in the terminal to enable more rapid display of result Tabular information is provided to supplement the graphics outpu CRT displays may be copied on printer The software provides for manual entry of both meteorological and radiological effluent data if real time monitor data are unavailable through direct transmission to the host compute Files are available for operator entry (edits) of simulated data for use during drill This also enables trainees to practice using predetermined scenarios in a mode that does not interfere with the on-line emergency mode that would be used during a real emergenc All software is written in FORTRA"' Listings are maintained on-site for use in interpretation or problem solving.
FILF. STRl.lCTURF Meteorological and rarliological data bases are stored on disc for use in making the re~uired rlose calculation Files cont a i n i r. g " constant " i n f o rn a t ion spec i f i c to the s i t e Ip 1 a n t situation are changed by privileged edit routines provided in the packa9 A series of routines performs these calculations using both the "fixerl" anc "time-dependent" cJata and stores the results of the calculations as necessary in files used by other routine The user can inst~uct the software to schedule runs th8t automatically read and display results fro~ these files sequentially without significant operator interventio METEOROLOGICAL AND EFFLUENT DATA MAINTENANCE TASKS A series of tasks is provided to inspect, maintain and archive the data bases created by the syste Examples follow: A task is provided to print the hourly or 15-minute meteorological parameter averages over any specified time period (within the bounds of the file). The "bad data" task can display the areas of bad data recovery for quick inspectio.
The "joint frequency" task categorizes and prints the meteorological data (in joint frequency form) by direction, speed group and stability class for use in Regulatory Guide 1.21 report.
summaries of total release by isotope can.be printe.
The "trend plot" task can be used to plot meteorological or radiological effluent cata which enables checking for pro~lem areas in the data *
fi *
Other tasks can be useci to summarize the delta-T and winci rose dat P,.:PllT DATA RF.OU I Rf.MF.NTS Presently, meteorological anci radiological ciata is input manually into the compute Upon completion of the hardware installation, the computer will store "real time" meteorological and radiological effluent monitor dat \\Jind speed and wind direction at the 33, 150 and 300 ft levels along with vertical tenperature difference between the 33 ana 150 ft, and 33 anu 300 ft levels will h~ derived fron insttunents on the meteorological towe Digital signals characterizing these data are then sent to the computer via data link Radiological data will be received from the effluent process monitor Releases can be assigned to any of four release points from which a release rate (microcuries/sec) is compute Since these effluent monitors do not provide an isotopic breakdown, the fraction of the total release for each isotope is determined fron ~ulti-channel analyzer (~tCA) spectra or as a relative isotopic concentration based on accident typ ACCIDENT DISPERSIO!~ AND DOSF. CALCULATION Results of real-time atmospheric dispersion and dose calculations for accidents are available in tabular and graphical for ~HDAS software is available for the models referred to in NUREG-0654, Appendix 2 as Class A and or enhanced Class The following two sections describe these model The Class A Model The Class A model used for real time assessment of dispersion is the standard Gaussian mode The graphical isopleth output, representing a straightline Gaussian shaped plume, was designed
- to n'rlace the plastic overlays (for maps of the same scale)
currently found in the efficrgency kits in response facilitie The o v e r 1 a y s a re re t a i n e cl a s ha ck - up.
/\\.
b a c kg round rn a p o f t r <<:*
s i t e i s p 10 t t cc a 10 n c;
..,,. i th th t:
i.so ri 1 et h s s. n th a t hot h a p ;-: e <"' r on tlw saP.1e plo *
P..11 accirlent calculations are llnrlcr T"lenu control for ease of us~'
by the operato The map scale, release point (alonG ~ith vent flows), level for data on the rnoteorological tower and terrain height (suhtracte~ froTTI plu;ne height for elevatec1 releases) can all he pres~lecterl, using system erlit defaults or selecte~
during the input portion of the computer analysi Averages for any previous hour (or the last 15-minute period) can be selected for the calculatio Certain self-checks are provided to warn the user of problem For example, if meteorological data are "bad" the user is notified and asked if data from some other source are availabl If so they are entered by the operato Likewise, if dose results are selected and there are no effluent release values present (from real time effluent monitors) or the data are bad, the user will be prompted for inpu Beta, gamma and/or thyroid-inhalation doses are computed after all input data have been entere A calculation will not be completed without contemporaneous meteorological and effluent dat Results are in printed or plotted for several choices are available to the user for the source ter If a Design Ba~is Accident is assumed, but the rele9se rate is unknown, preset release scenarios can be used for up to ten accident scenario Otherwise, real time data fro~ effluent
~or.itors can be useo.
- Enhanced Clas~ A Model NUREG-0654, Appendix 2 also refers to a more complex model for estimating diffusion and exposures out to greater distanc~ The rnoccl currently plannc-c tor use at
~aleTT-is a plunc:-
se~;r-0:-:t model hased on a program developeci by PLG callerl CRACIT (For Calculation of Reactor Accident Consequences Including Trajectory) which is similar in conc~pt to that of the CRAC program which was written for the Reactor ~afety Study (\\*:.l.\\SH-1400).
The "front end" source term and "run" rnenu options provideri for the Class A ~odel are also used to drive the enhanced A model versio Thus, the operator interface is essentially unchanged in switching from one model to the othe THE PLUM SEG~1ENT MODEL The basic functions of the plume segment model are the calculation of meteorological dispersion of the released radioactive material as it travels downwind and the estimation of the resulting doses from this materia The meteorological dispersion is modeled assuming Gaussian diffusion and variable trajectory transpor At the present time the model uses generic correction factors for dispersio After review of site data, PLG will modify the software program, if necessary, to include site-specific correction factors when the hardware system becomes operational on-sit The transport portion of the dispersion model allows the plume travel direction to vary as the wind direction varie The model divides Ble plume segments called spatial intervals according to the travel distance for each hour (or quarter-hour).
The standard Gaussian model is used to estimate plume dispersion based on the wind speed, wino direction anc delta temperature measured on the meteorological towe The
plume, therefore, is represented by a series of segments, each of which has different characteristics hased on the meteorology at the time the segments are in their respective location TrE"* mociel simulates plume rise, building wake effects, dry riernsition anri wet riepnsition a~ a function of rain rat If necessary, tbe software program may he mociifierl to consider seasonal anci riiurnal/nocturnal changes as well as mesoscale circulation effects on the mixing heigh The nodel can be run using hourly or quarter-hourly averages (updates) ot meteorological condition nn~ E' p ROJ F,CT I (lt.: s The results of the plume trajectory and dispersion modeling and the calculated doses can be plotted on the site ma Plume spread (horizontal and vertical) varies as a function of wind speed and atmospheric stabilit Plume direction is controlled by the wind directions based on meteorological dat Characteristics of each spatial interval can also be printed in tahular for The dose calculations in both models provide information used in making protective action decision Projected integrated organ doses for the whole body and thyroid organs are computed as a function of distance downwind for a given short-term (usually 15-minute) releas Three pathways are used in the plume segment model including plume shine, inhalation and ground shin The plume model includes both finite and semi-infinite cloud assessmen The whole body dose consists of the sum of plume shine dose due to plume passage overhead, inhalation dose due to inhaling airborne radioactive material and ground shine dose from particulates deposited on the groun The thyroid inhalation dose is reported separately for use in comparison with the PAG's, although the plume shine and ground shine components are available in table The average dose rate (mre~/hour) is also estimated from the three pathway component The Class A model does not treat
,.
R particulates (other than iodine) or ground shin The model can be ru~ in a predictive mode using the most recent meteorological dat Thus, persistence of meteorological conditions is used for all dose projection Long-term Releases -
Yistorical Data Calculation routines are also provided for making longer-terG dose estimates (for more than one 15-minute release).
Versions are availahle for either thP straightline or th~ rlu~e seQment rorlel Releases are simulated as several short-tern releases, each having a separate plu~c movinu away fror the site according to the meteorological conditions for that tim In the enhanced A model (variable trajectory), each successive release is controlled by a different weather sequenc noses over the area are integrated for grid locations (a fine mesh grid is used for the enhanced model).
The dose over all releases for each of up to four requested projection time periods can be displayed graphically as isopleths on the CR REMOTE USER EMERGENCY REPORTS (the Broadcast Function)
Software is provided which can automatically transmit output ("broadcasts") to predefined remote terminal The "Broadcast" reports include meteorological data in the format specified originally in NUREG-065 The remote terninal operator does not have to constantly monitor the terminal and schedule tasks to receive 15-minute updates *
'
Attachment I U-Bolt Anchor Test Program and Development of Stiffness Values The program at the Franklin Research Center performed in respon~e to NRC inspection findings consisted of (2) two interacting phases:
(1) The analytic phase consisted of performing finite element structural analysis on models representing the U-Bolt assemblies, pipe lengths, and contact/friction mechanisms, for three representative types of 6 inch diameter pipe anchor Deformation and load-pitch patterns were determined for piping system loads applied independently in all six directions plus a U-bolt pretension loa From this data, the (6 by 6) stiffness matrix for the center of each assembly was determine (2) The testing phase consisted of measuring the load-deformation characteristics, under controlled pretension, of nine different U-Bolt anchor assemblies for 2, 4 and 6 inch diameter pip Axial and torsional load applications determined the frictional constraint capability of each assembl The incremental deformation under these loads, as well as under lateral load, yielded flexibility data for correlation with the analytic phase of this progra Both the analytic and testing data were utilized to compute flexibility characteristics for sizes and directions not directly determined from analysis or testin In addition, the constraint capability for combined axial load and torsion was estimated for each assembly tested and a general procedure wa-s formulated to treat any size assembl Selected piping analyses, modeling U-Bolt assemblies as both full anchors and with the stiffness matrices developed under this program, yielded comparable, and for the most part conservative, results for static, thermal and dynamic loads.