ML20125B006
| ML20125B006 | |
| Person / Time | |
|---|---|
| Site: | Millstone  |
| Issue date: | 02/01/1966 |
| From: | US ATOMIC ENERGY COMMISSION (AEC) |
| To: | Advisory Committee on Reactor Safeguards |
| Shared Package | |
| ML20125A422 | List:
|
| References | |
| FOIA-92-198 NUDOCS 9212090168 | |
| Download: ML20125B006 (40) | |
Text
. _ _ _ _ _ _ _ _ _ _ _____. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
t ..
- fRCKC USE 6NLT c,/ /
)~
/ {
n * *
- (
11 . S. ATONIC ENERGY C(M11SS10N DIVISION OF REACTOR LICENSING REPORT TO ADVISORY C0k41TTEE ON REACTOR SAYEGilARDS IN Tile MATTER OF THE CONNECTICUT LICHT AND POWER COMPANT THE IMRTFORD ELRCTRIC LIGHT CCMPANY THE VESTERN MASSACHUSETTS ELECTRIC CCMPANY ,
THE MILLSTONE POINT COMPANY AS PARTICIPAlftS IN THE MILLSTONE NUCL_ EAR POWER STATION
/
4 l
Note by the Director, Division of Reactor Licensing Th2 attached report has been prepared by the Division of Reactor Licensing for
, sideration by the Advisory Conunittee on Reactor Safeguards at its March,1966 wa t ing. '
- g20;ggedom citat use omity S LAWRENC92-190 PDR
ffbbbbbb h -
BILL l=VW4 INT RODU CT I ON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 COMPARISON WITil DRESDEN 11 STATION. . . ................ 2 STAFF REVIEW APPROACll . . . . . . . . . ................ 3 S IT E CON SI DERAT I ON S . . . . . . . . . . ................ 4 S i t e De s c ri p t i on . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Heteorology . . . . . . . . .*. . . . . . . . . . . . . . . . . . .- $
ll y d r o l o gy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Geology and Seismology. . . . . . . . . . . . . . . . . . . . . . . 11 CONFORMANCE OF Tile MILLSTONE STATION DESIGN TO STAFF'S CENERAL C RIT E RI A . . . . . - . . . . . . . . . . . . . . . . . . . . .-. . . . 11 DESIGN FEATURES RELATED TO CONCERNS PPIVIOUSLY EXPRESSED BY. Tile ACRS. . 22 ACCIDENT ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . 24 Ceneral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24-Control Rod Drop. . . .......................27 Fuel Loading Accident . . . . . . . . . . . . . . . . . . , . . . . 29 Steam Line Break Accident . . ... ... . . . . . . . . . . . . . . 30-Loss of Coolant Inside the Drywell. . . . . . . . . . . . . . . . 32
. CONCLUSIONS . . . . . . . . ._. . . . . . . . . . . .-. . . . . . . . 34 TABLE 1. - Potential - Of f-Site Doses (Staf f Analysis) . . . . . . . . . - . 25 TABLE 2 - Susumery of Accident Analysis by Applicant -for Millstone. .-_ . 36 JJ)FFHCHAL USE ONLY,
. - . . _ _ - - _ . . _ ~ - _ _ _ - - - _
i 4
-iFRGEUSE-ONLF in(RODUCTION A preliminary site report for the Millstone Nuclear Power Station was submitted on May 14, 1965, by the Connecticut Light and Power Company, the Hartford Electric Light Company, and the Western Massachusetts Electric Company. These companies are otso participating in the Connecticut Yankee project. The Connecticut Yankee site is approximately 20 miles northwest of the Hilletone site. The Millstone site is located on the northern shoreline of Long Island Sound near Waterford, Connecticut.
The ACRS and the staff completed a preliminary evaluation to determine the gen-oral suitability of the location and the site area for construction of a reactor plent. At the time of the review, the capacity of the plant, which was stated to be cither a PWR or a BWR, was set at approximately 2500 MWt. In the letter of July 19, 1965, the ACES concluded, "that the Millstone Point site is acceptable for a reac-t . . . . if adequate containment and associated engineered safeguards are provided."
On November 15, 1965, the Connecticut Light and Power Company, the Hartford Electric Light Company, the Western Massachusetts Electric Company, and the newly formed Millstone Point Company (herein collectively referred to as the applicant) formally submitted their application for a construction permit to build a BWR nuclear powar station at the Millstone Point site. The applicant specifies that the station will have a net electrical power output of 549,200 kw (1730 MWt) and is scheduled for completion in mid 1969. The General Electric Company has the responsibility to furnish the complete nuclear power s tation. The station is to be substantially sinilar to the Dresden 11 facility except for the power level.
Following a review of the preliminary Design and Analysis Report (DAR), Vols.
I and II, a meeting was held on December 22-23, 1965 to discuss the safety aspects cnd contents of the DAR. Supplemental information (Addendum No.1) was received on aruary 2,1966 in answer to ques tions submitted by the staff on January 14, 1966.
Another meeting was held on February 17, 1966 to review the supplemental information.
p~TW1IOII u .w -O II: IIu-a f%7 @. F:UI u- .V u
e .
.-9FRCHAUUSE-ONFA 2
This meeting was followed by an ACRS Subconnittee meeting on February 18, 1966.
As a part of the overall safety review for the Hillstone station, the advice of the following consultants has been requested (1) Geological Survey; U. S. Department of the Interior, (2) Fish and Wildlife Services U. S. Department of the Interior, (3) U. S. Weather Bureau; U. S. Department of Commerce, (4) U. S. Coast & Geodetic Survey; U. S. Department of Commerce, (5) Dr. Nathan H. Newmark (Seismic Design) .
The reports of the consultants will be made available for the Acts's use prior to the March, 1966 toeeting.
COMPARISON WITH DP.ESDEN 11 STATION The applicant has proposed to construct a BWR(CE) reactor plant with a power cting including stretch capacity of 2010 MWt. The operating power as proposed, however, will be approximately 1730 MWt. Our review of the various characteristic parameters indicates that the Hi11 stone station will be basically identical to that proposed for the Dresden 11 e tation with due allowance given to the lower power level, i.e., 2255 MWt for Dienden 11 and 1730 MWt for H111 stone. The applicant reports that any differences between the two stations are the result of continuud ef forts to improve the reliability of the safety systems. These details will be discussed in later sections of this repor t.
Since for all practical purposes, the Dresden 11 and Hi11 stone Plants are of the same design, the unique features of this project are concerned with locating the recctor plant at the Hillstone site. In this regard, several site-related differ-onces in overall station design can be identified. The first is the available oloctrical power sources required for the engineering safeguards, while another is tha ventilation stack. In the latter case, the M111 stone station is to be provided with a 375 foot stack while Dresden 11 proposed to use the existing 300 foot stack, hbiJ UhbMNb.N
e .
0FFHCHAL-USE ONE L 3
In addition, the seismic design parameters are slightly different. The Hillstone station criteria are based on a ground : notion acceleration of 0.073 with assurance of a esfe plant shutdown up to 0.17g without a loss of function of components important to safety. The corresponding values for Dresden 11 were 0.1 and 0.23 ,
respectively. The seismic design features have been reviewed by our consultants and found to be acceptable for the Millstone site.
We will discuss the results of our evaluation of the other principal site-related differences in later sections of this report.
STAFF REVIEW APPROACH The first phase of the staf f's review of the Millstone station was to evaluate the station's design and performance characteristics in conjunction with the General Design criteria and in comparison with Dresden II. The Dresden 11 station has already been evaluated on this basis and found acceptable and a comparable review for the Hillstone station revealed, as one would expect, the same degree of conformity.
To avoid any potential danger or shortcomings of a routine review, the staff effectively conducted a second phase review; i.e., still using the criteria as a guide, but pursuing a more extensive analysis in depth for certain of the criteria wherein additional information was believed to be required. The additional areas of concern reviewed by the staf f included certain aspects of the engineered safeguard systems, instrumentation, containment leakage rate, analytical models used for-safety evaluation, and the quantitative aspects related to station stability.
In conjunction with the s taf f's review involving the criteria, the overall safety aspects associated with operation of the reactor at the proposed site have baen analyzed. The results of the analyses of minor accidents which did not lead to e release of fission products were found to be comparabia for both the Dresden 11 facility and the Millstone station, comparability is least in the case of a major accident leading to a large release of fission products due to the dif ference in h kbhh hNb
-OFFHCHAL USE-Olh' 4
overall site characteristics.
In addition to the above, the staff also reviewed the Millstone station in conjunction with the concerns of the Acts as expressed in its letter of November 24, 1965. In this regard, we requested the applicant to discuss the consideration which has been given in the design to recommendations contained in the ACES letter of November 24, 1965 related to reactor pressure vessels. The applicant was also requested to consider the effects of pipe whipping, missiles, and fuel movement as discussed in the ACRS letter dated November 24, 1965 concerning its review of the Dresden 11 facility.
SITE CONSIDERATIONS site Description >
The H111 stone Point site consists of approximately $00 acres, partly owned
/ the Millstone Point Company, and the remainder held under option. Tha site is located in the town uf Waterford, Connecticut, on the east side of Niantic Bay and on the south shore of the state. It is approximately 1.5 miles from Niantic, which together with the other beaches on Black Point on the west side of Niantic Bay, have a total permanent population of approximately 7,200, all residing within three miles of the site. In the summer, this number increases to 7,800 people. The site is also 1.1 mile from Pleasure Beach on the east side of Jordan Cove. The area along this shore within 2 miles of the site has a permanent population approximately 680, which increases to 1,200 in the summer. The nearest large city is New London (1965 estimated population 33,300), and its nearest boundary is 3.2 miles from the site. The total permanent population within 4 mil'es of the site, according to 1965 estimates, is 25,774. This increases to 33,000 during the summer months when beach cottages are occupied.
The nearest residence is located 0.55 mile northeast of the' proposed reac-ter location, and is in a group of approximately 40 houses located less than 1 mile
~b[bbbk )hb'hhk '
- . - - - . _ - . - - - - _ _ . _ -_ =
JFFECML USE ONHi- !
from the reactor. The nearest site boundary will be in this same direction, at a j distance of approximately 2100 feet *. Between these two distances is a narrow strip ,
of beach on which contractual arrangements prohibit the construction of any houses suitable for h a bit ation. Current ac tivities at the site include a small naval iosearch laboratory (4 employees), and a pilot desalination plant (12 employees).
Th3 applicant states that these facilities may be permitted to remain.
It is to be noted that the preliminary site review was based on a minimur site boundary of 0.5 miles whereas the applicant now states that this has been reduced to 0.4 miles. Our analysis is therefore based upon the latter distance.
To the closest boundary of New London, 3.2 miles, the population is uniformally spread through eleven of the sixteen cardinal directions with few or no persons within the other five directions. No 22h' sector has more than 5,000 persons. The avail-
. ole low population zone distance, the refore, is 3.2 miles which is also the distance to the closest boundary of New London. The 3.2 miles is also the population center -
distance.
Meteorology The applicant has furnished a description of the meteorological environment for this site based on climatological records from U. S. Weather Bureau stations in the area, and on data collected at the Connecticut Yankee site and at Brookhaven National Laboratory. In addition, 6 weeks of measurements made at the site have been partially analyzed. Since this is a coastal location, the maximum wind speed l associated with hurricanes is of significance with respect to the safety design considerations of the facility. The maximum wind speed reported in the area is 115 mph, with gusts to 140 mph, reported at Montauk Point, Long Island. Similar winds are reported for Block Island, located near the middle of the entrance to k
'* Note: The applicant's report which presently states 1150 feat to the nearest site boundary has been determined to be in error.
-OFFECPAEUSE ONLY .
-3FFECHAH. RUSE ONLC 6-Long !aland Sound. In addition, there are occasional tornedoes in the state. The i
applicant propo.,,e a design wind speed of 115 mph with gusts to 140 mph for the plant, l which compares f avorably with the maximum wind speeds reported for the area. The maximum wind velocity during a tornado, however, could be expected to exceed this j design value. The applicant has considered the possibility for a tornado striking ths plant, and has s tated that it is one in every 1,250 to 2,000 years. Design footures are proposed for protection of those comporents vital to reactor safety to the extent that those required for core cooling vill be designed to withstand 300 reph winds.
The applicant presented models for the atmospheric dispersion character-istics of the Millstone site, based mostly on genersi assumptions. Inversion frequency information was obtained from Brookhaven and Connecticut Yankee. The ispersion models are based on experimental work performed at Hanford, originally reported in IN 54128 and IN-80204, and subsequently republished in the references given by the applicant. For the horizontal spread of the plume, measured values of sigma theta (the standard deviation of the wind direction observed over the period of the plume passage) for various wind speeds are used. Hawever, the six weeks of data obtained at the site are not very meaningful, and the applicant prefers to rely principally on the distribution of data obtained over a one-year Tariod at the Dresden site. We have discussed ou r r e s e r ya ti on s r e c a r ding the validity of this data with a number of interested groups in the pas t. The vertical apread of the plume is calculated using the original form of Sutton diffusion parameters during neutral and unstable conditions. During inversion conditions, the vertical growth of the plume is described by a special formula developed at Hanford to fit the condi-tions experimentally found to exist at the Hanford site. Its applicability to this site is not discussed, but in maoy cases inversions do not produce the highest dose for this facility because of the high stack.
"hbbbb bbb hW '
-9FFHCHAirUSEOl& '
We have examined these models, arJ when the suggested parameters are used, they give results which agree closely with the results obtained using pasquill's model with a wind speed of one meter per second. For the horizontal dispersion parameters, the applicant has assumed values of sigma theta times wind speed of 0.16 and 1.0 t;cdian-meters /second corresponding to wind speeds of 2 and 10 mph, to represent dispersion for a two-hour period. The relationship of these values to the site data cannot be well established at this time, but they compare reasonably with values derived at Hanford. In this respect, Hanford meteorologists have suggested that at a wind speed of 2 aph, the value of sigma theta times speed could have a cinimum value of 0.05 for an exposure period of two hours.
In the applicant's viewpoint, the duration of an exposure hazard to the public from an accident is determined by the wind direction persistence whenever se accident would occur. In this respect, the applicant has chosen as a basis of enalysis a wind direction persistence of 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> based on a sampling of data from vcrious cities around the country, none of which are in New England. The stsff has obtained similar data for Kennedy Airport, which shows a greater persistence than any of the cities for which the appli:: ant gave data; i.e., persistence in one direction of 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> or more occurs on the average of about 3 tires per month. A persistence of 30 hours3.472222e-4 days <br />0.00833 hours <br />4.960317e-5 weeks <br />1.1415e-5 months <br /> or more occurs on th; average of about 2 times per month, and a maximum persistence of 76 hours8.796296e-4 days <br />0.0211 hours <br />1.256614e-4 weeks <br />2.8918e-5 months <br /> has been reported. In view of this information, t..e staf f believes that the wind persistence assumed for the Hillstone Point site is not sufficiently conservative for the purpose of estimating the consequences of an acci-dont.
The U. S. Weather Bureau has reviewed the meteorological information sub-citted by the applicant, and their coments emphasize the attitude which is 'taken in the discussion above; the dispersion models used are not sufficiently related to the existing site data, and do not take into account in a realistic way the unique physical
- ktT M.U. n .~ N O..l.i U,I@
umu.7 O S E %
l
\
- fFHCHAL-USEONLb differences of this peninsular site, as compared to the continental sites from which most of the information is taken. Specifically, the long path of the air over water ca it approaches the site gives it properties different from that of continental air.
For one thing, the turbulence level decreases, as the initial site data indicates, so that the frequency of occurrdace of the horizontal diffusion parameter proposed
, by the applicant is about 25%, rather than the 1% which they found at continental i sitos. In addition, there are periods during the year when the water cools the air, forcing a s trong inversion condition. Upon reaching the warmer land, thermal turbu-lance is generated, producing fumigation conditions, just as occurs af ter sunrise at intend sites. However, the significant dif ference here is that the fumigation condi- 5 tion can persist for several hours whereas at inland locations the duration is a T
auch shorter time, usually less than one hour. Under such conditions, the air scontration will be approximately an order of magnitude higher than any estimated by the applicant. The Weather Bureau report, which also covers some other meteoro-logical areas, has been furnished to the Consmittee.
In sunmaary, the diffusion climatology of this site is not as well established as it could be. The particular parameters selected may be_ appropriate, but some additional site data would be helpful in establishing this more definitely. It-seems that there may be a somewhat higher than average frequency of poor dispersion conditions. We believe that for the final review, considerable additional analysis, and-perhaps more data, may be necessary to establish more firmly the effect of fumigation conditions induced by cold winds from .ong Island Sound. However, even-
\
using appropriately conservative meteorological assumptions. as diocussed in t hi s r e p or t , p o t e ntial exposures under accident conditions are within ecceptable limits.
ONECEAL USE ONLY
. - _ - ~ . - _ ~ . - - - - - . - . - -.-. _~ . - _ . _ - _ . - ---
d 9
Hydro 1ory '
An important consideration for this site is high water caused by storm tides and wave runup on the coastline. In recent times severe hurricanes have passed over the area with their centers a few miles both to the east and to the west of the site, so measured ef fects f rom both situations have been obtained, it was found that the highest water levels result f rom hurricanes whose centers are a few miles vest of the sito, so that the site is subjected to the winds in the severest, northeastern sector I
of the storm, which also tends to fill up Long Island Sound due to east winds backing to southeast. Such storms have produced water levels as much as 10 feet abore sea lovs1 at Niantic. In addition, it is stated that astronomical tides can rise as much as 3 feet above the mecn tide range about once a year. Since the mean tide range is 2.7 feet , we interpret this to mean that the water can rise 4.3 feet above nean sea 4 eval once a year without storm action. Data is presented which indicates that the larFest umasured storm surge is 7.3 feet. This, combined with the largest astronomical tide gives a height of 11.6 feet above nean sea level. The applicant has estimated a maximum wave runup of 5.1 feet for an 80 mph wind with a certain shore slope. We have no means of checking the accuracy of this estimate at the present tire, but anticipate comments f rom our con Jultants on this matter. The sum of these ef fects is 16.7 feet above mean sea level, which is within the range of ef fects observed at other coastal locations. The applicant states that the natural grade at the plant location is 15 feet, but Tigures 26 and 30 in the DAR indicate the finished grade will be only 14 feet. It is also stated that the water level due to storm runup in not expected to exceed 14 feet at the particular location of the plant, but no explanation for the discrepancy is given. Presumably, there -is some dif ference in slope of the shore at this point Sich r+ duces the wave runup.
9FFECHAL USE ONLA We do note that the simultaneous occurrence of the maxima of all ef fects 1
invched is regttiied to produce a water IcVel as high as the proposed grade, lloweve r ,
.it ic act clear that in the event of such a simultaneous occurrence, the site vocid j rmain cbove water. We ftel that this should be resolved by further infor.astion from the appiscatta and our consultants, and that in any avant the dif ferences are small enough that shey can be easily resolved.
The overburden at the site is so impervious that motion of water in it will i
be extremely slow in any direction. The expected ground water gradient is in such a direction th.a it is not expected that water can move from the site toward the main-lend. Ilsnee, we do not feel that any public water supplies will be affected by the operation of this plant. The nearest well is 3/4 mile northeast of the plant, and is nrotected by an intervening ridge. This well supplies only 40 homes, so it is an easily replaceable supply if such action were necessary.
Bottom contours in Twotree !aland Channel were measured along two lines f rom the site. In addition, continuous records were obtained of the water velocity at three different depths at a number of poirits along each line. Similar measurerents were obtained in Niantic Bay. In this way, reasonable estimates have been made of the total water flow in the vicinity of the site, which are about 100 times the dis-charge flow from the plant. It is anticipated by the applicant that the concentration added to the chant.el will be less than present background level so that reconcentration in the Niantic Bay Scallops will not be a problem.
In conclusion, we believe that 1) the discrepancy in maximum water levels is small enough to be resolved easily, 2) there appears to be adequate dilution in Twotree Island Channni to obviate any problems with the concentrations which are anticipated for iis plant .md a monitortug program is proposed which will provide backup protection in this area.
-OFFECEAL USE ONLY-
P 4FMGHAL-USE-ONEiA Geolorv and Seismolocv The bedrock in this arnc. consists of Westerly granite and associated Monson gnsias which extend to the surf ace in '.he area of the quarry. The impervious nature of, the rock is demonstrated by the f act that even though the quarry has been closed since 1960, the water level in it is still some 17 feet below Long Island Sound. It orgcars that the granite is a sheet, interlayered within the gneiss, which, at the quarry, has become tnickened to a bulbous forn. It is anticipated that the plant will be placed on bedrock, and that the formations could support as much as 100 tons per
- square f oot , much more than required by the plant.
A seismic ref raction survey of the site has been performed, and this revealed no unusual or extreme subsurf ace conditions. Correlation with boring data was good.
The only earthquake of significance in this area was the Moodus quake of 1991, which was assigned an intensity of as much as VIII on the Modified Nrcalli Scale by some writers. This intensity would be redu:ed to VI at Millstone Point, or approximately 0.05g. The applicant has proposed that the design of structures and equipnent important to safety will be based on ground motion due to an acceleration of 0.07g and an acceleration of O 17g without a loss of components important to safety. These values are in agreement with those recommended by the U. S. Coast &
Geodetic Survey. The seismic design criteria have been reviewed by our consultant, Dr. Newmark, who has concluded that they "will provide an adequate margin of safety for seismic resist aace,"
CONFORMANCE OF THE MILLSTONE STATION DESICN TO STAFFf S CENERAL CRITERI A We have evaluated the Millstone Station to determine if the " General Criteria for Nucicar Power Plants" as published by the Commission on November 22,1965, have seen satisfied, A comparative study was made of the data submitted in support of the
~
bhI._b bb .G
. 4 wFFHCHAL USE-ONLb.
criteria for the Dresden 11 f acility, and that submitted in support of the criteria for the Millstone Point facility. k'hile these f acilities are very similar in design, i in several instances the data on Millstone Point differed and/or was inore specific with respect to certain of the criteria. These dif ferences will be identified and discussed in this section. In all other . espects, the discussions of how the criteria are fulfilled given in our report to the ACRS on Dresden 11 dated September 20, 1965, era applicable to the Millstone Point f acility. !
Erlt e rion 5 The reactor must be deniened so that power or process variable oscillations or transients that could cause fucI f ailure or t>rimary systen damare are not possible or can 50.readily suppressed. ,
- Additional informatipn has been received to further support the objectives that the overall reactor plant as designed would be stabic. The applicant has ,stovided desi6n s'dalines in terms of parameter values for acceptable dynamic response. The guideline valuca are based on the criteria that in all cases the minimum gain margin would be 13 db and the minimum phase margin would be 55' (the same as Dresden II) .
The dynamics of the reactor plant have been investigated using both digital and analon cotq-u e rs. Input perturbations simulating of f-normal conditions have been used to detstrine the response of reactor variables. In addition, studies have been made in tne ircquency dornain for stability evaluation.
The following table provides a comparison of design data with the desirn guidelines for the M: '1 stone station.
Parameter duideline Desinn
- 1. Core average exit quality 412% 10.8%
- 2. Core inlet subcooling <30 Btu /lb 21 Btu /lb
- 3. Void reactivity coefficient <10d/% void 7.5d/% void 4 Core two phase transit time 40.65 sec 0.6/. 'se c
- 5. Ratio of core sing 1s phase pressure drop to two phase pressure drop > 0. 4 0.51
<OFFECHAbUSEMLY-
._JFFHCHAL USE ONUc Parameter Cuideline Design 6 Rate of change of recirculation > 0.9 1.0 loop characteristic curve (head /% f1w)
- 7. Tuel pel*et diameter > 0.375" 0.488"
- 8. Active fuel length <150" 144" 9
The applicant reports that the results of frequency analysis of the Hillstone station based on the above design parameters indicates that the gain margin to 17.5 db and the phase margin is 70' . These values are well above the design margins of 13 db and $5', respectively. The applicant has also investigated the offects of off-normal conditions due to changes in power level or f ailure of equip-nen t. In the case of a pwer increase on the order of 20% the margins were 15.9 db and 56'. Both of these values are still above the design limits. Failure in citler to flow control system, pressure regulator, or loss of feedwater heating yielded a
. uction in gain :nargin of 3 db. This results in a margin above the design limits and thereby indicates the system to be stable within the limits of the analysis. The cpplicant states that appropriate testing is to be conducted during the operational program for the M111stons station to determine the overall system response to various parturbations.
On the basis of the foregoing, we believe that the design of the Millstone station is based on conservative stability limits and in view of the applicant's intention to perform dynamic testing the actual plant characteristics would be deter-mined to provide further assurance of. stable operation and therefort beM eve t.,4 Lt hjectiva of Criterion 5 is antiaf ted.
Criterion 7 The maximum reactivity worth of control rods or elements and the rates with which reactivity can be inserted must be held to values such that no sdnale credible rechan-ical or electrical control system ma' function could cause a reactivity transient ecpcble of damaging the primary system or causing significant fuel failure.
A support structure is provided just below the control rod drive flanges.
The system is designed to prevent excessive control rod withdrawal in the event of a 0W 0 h Y hb
JFFHCHAL-USE ONE thimble f ailure wherein the unbalanced reactor pressure is applied across the control rod. The support systeta is designed to limit the total downward movement of
- drive and thimble to a maximum of 3 inches. The design performance of the support is pre-dicated on the assumption that the collet fingers are in engaged with the index tube of the control rod at the time of thimble failure.
We believe that the capability of the support structure to withstand the totc1 forces in the event that the rod has been accelerated over a distance of 12 inches (2 notches) should be considered in the design. Further the capability of the collet fingers to engage and hold the rod under the above conditions should also be considered. If the collet fingers had been initially disengaged in anticipation of rod motion and then a thimble rupture occurred, it is possible that the rod could move 12 inches before the collet fingers engaged the rod. Double notching is not an mf requent occurrence based on control rod operating experience with C. E. systems.
The applicant states that although the fingers may be damaged, their f ailure mods is to bend inward and thereby join and lock the index tube. The structure is designed to absorb the energy of the complete drive line, thimble, and gui e tube, accelerated through one inch by reactor pressure. .The total energy is predicted to ,
be 105,000 in.lbs. We believe that the total energy of the accelerated muchanism undsr the postulated mode of f ailure may be greater than current design due to the additional energy of the accelerated control rod. At this stage of review, we believe that the design concept is acceptable but that further evaluation of design details will be required at a later date.
In addition, the applicant was asked to discuss safety system capability to protect the reactor (no fuel damage) against excursions induced by the simultaneous
'thd aval of any two rods under all conditions of allowed bypars of protection channels and individual chambers. (This question was prompted by the f act that a single failure, OFFHCTAirUSEOMMT
0FFHCHAL USE ONU._
i.e. , shor t circuit, within the rod selector network would allow the simultaneous withdrawal of two rods). The applicant indicates that even under the worst condi-tions of allowable ins trument bypass, a rod block automatically initiated by the ApRM system would prevent fuel damage. We believe that the analysis should be made avail-cble at a later date but is not required at this time since this accident would be loss severe than the already evaluated rod dropout incident. We believe that Criterion 7 can be satisfied on the basis of the design objectives.
Criterion 17 The containment structure, including access openings and penetrations, must be d3 signed and fabricated to accowoodate or dissipate without failure the pressures and torperatures associated with the largest credible energy release including the effects of credible metal-water or other chemical reactions uninhibited by active quenching systems. If part of the primary coolant system is outside the primary reactor contain-ment, appropriate safeguards must be provided for that part if necessary, to protect th3 health and safety of the public, in case of an accidental rupture in that part of th2 system. The appropriateness of safeguards such as isolation valves, aCJitional antainment, etc. , will depend on environmental and population conditions surrounding tha site.
The applicant states that in the event of a rupture of one of the main recirculation lines, and without the benefit of any core cooling, a metal-water repc-tion on the order of 27.5% could be expected. The model used for the analysis assumes that the extent of netal-water reaction is based on a terminating mechanism initiated when the zirconium clad reaches the net ting ooint of the metal (i.e."'3300 F).
Other analytical mod :le assume a terminating point at the melting point of airconium oxide (formed by the zirconium reacting with steam) which is approximately 48000 r.
In view of the possibility of a larger than predicted amount of reaction, we have esked the applicant to provide jus tification for the model used. The applicant s tated and we agree, that there exists no experimental data to support a claim that the cledding would remain intact above the metal melting temperature and therefore have sdopted the boundary condition that the cladding would fail at the metal melting torperature. It is further stated that even under the conditions of the formation of Zr07, a " crucible" ef fec t could not be justified be that gross cracking
@FFHCHAL USE M?LY
- .- - - _ . - - . - . - - - - - - . - - ---_ -.-..=~___--
4RCHAL~USE ONLF would cause segments to f all from the core region. The basis for- the applicant's model is derived ' rom both the work at various tvitional laboratories and at General ,
Ele ct ri c. The effect of assuming the oxide molting rather than that of the metal, as th's terminating point, would be to increase the predicted amount of reetal-water re-cetion from approximately 25% to 45%. This increase would result in a pressure in the containment in excess of design: 1.e. 85 psig compared to a design pressure of 62 poig.
We have reviewed the model and basis used for the Millstone evaluation (s en:2 for Dresden 11 and Humboldt) and believe that in the absence of any experimental data t'o show a higher terminating temperature, the model as presented is acceptable.
We believe, however, that in view of the rapidly changing scope of knowledge
. this field, co n a e rva t i s m ahou1d be inherent in the containment capability end in the calculation of the extent of reaction. Based on the following, we believe "
^
that suf ficient conservatism both in the calculational model and in containment capability margin has been provided. This provides aesurance that the design would not be found inadequate to withstand metal-water reaction consequences at some future date.
(1) It is our opinion that the short tirse calculated for the completion of the metal-water reaction is very conservative. To transfer steam to the upper parts of the core in stoichiometric quantities would seem to require some cooline of the lower regions, thus retar _ing the reaction in the lower portion of the core.
(2) The reaction rate would be limited by the rate of steam evolution from the bottom plenum, thus extending the length of time- over which energy would be released t'o the containment. A possible mechanism for terminating the reaction could be local clad swelling or melting which would block steam flow channels.
TTMGAEe UE OFRY
,FFEML-USE ONLL (3) The capability of the drywell-suppression chamber coatainment to withstand the censequences of a retal-water reaction has been analyzed by the applicant and a plot of the amount of metal-water reaction which the containment can tolerate as a function of time has been provided (Figure 118, Millstone Design and Analysis Ite po rt ) . The hydrogen and all energy associated with a given percent reaction (as well as decay heat energy) was assumed to be released to the containment uniformly over the duration of the event. The analysis illustrates the very high f raction of retal-water reaction which the containnent can tolerate with an inert at mos phe re. For exawple, the containment could withstand a 70% metal-water reaction over a three hour period assuming all energ) was released to the contain-ment atmosphe re. The only safeguard assumed operable for the analysis is one containment spray system.
The applicant has not been requested to analyze the final disposition of the molten core since the 100% core oc1t plus metal-water reaction was considered a " design basis" accident. Ilowever, the following sequence can be postulated for this reactors (1) the core melts, (2) large amounts of heat are transferred to the vessel walls by radiction, (3) the internal structure within the vessel (lower c. rid plate, control rods, rod thimbles) melt and collapse, (3) any remainine water in the lower plenum is flashed to steam, (5) ti.e bottom of the insulated primary vessel loses structural integrity, (6) the molten mass is held in and between the 145 control rod thinbles and portions finally melt through the thimble support s t ructure , (7) masses of metal drop into the concrete-bottomed pool of water in the drywell sump to which heat has been continuously t ransferred by radiation, (8) water is flashed to steam and the energy thus t ransferred to the suppression pool . If the molten core is assumed to drop into 4
vater over a period of time, the drywell will not be endangered by rapid steam evolution. The core would be quenched completely within a few hours because of *.he ai;;n cont ainment spray syster f1cu rate (i.e., each loop has a spray capacity of 6000 c.pm).
-OFFEAh-USRONLY-
- ~ . - - - - - - - _. - _ . - . - - -. _ - - - - - - . .--
d e MFHCHAL USE ONLY-18 -
W2 believe that the proposed containment design satisfies this criterion.
Criterion 18 !
provisions must be made for the removal of heat from within the containment structure as necessary to maintain the integrity of the structure under the conditions d: scribed in Criterion 17 above. If engineered safeguards are needed to prevent con-toinment vessel failure due to heat released under such conditions, at least two 4 ind3 pendent systems must be provided, preferably of different principlos, gackup equipment (e.g. , water and power systene) to such engineered safeguards must also be redundant.
Initially, the design of the Containment Spray and Core Spray systems-included certain valves that would have been manually locked open without remote position indication. In several cases these valves were located inside the containment.
Our review indicated that although the systems as designed included redundancy in the purps, pipe lines, and valves, ultimate operation of these systems would be predicated on cdministrative control to assure that these manual valves were open prior to reactor arction. Recognition of potential problems by the applicant resulted in certain changes which we believe will better assure the operational status of the systems.
Th2 changes that have been made are discussed below.
(1) The manually operated locked open valvts on the pump suction of the Contain-ment Spray System and also on the pump discharge lines just outside the containment hsve been replaced with motor operated valves with position indication.
(2) The manually operated locked open valves on the pump suctions in the Core Spray System have been replaced with locked open motor operated valves with position
-indication. Motor operated valves will also be used in the test lines and interlocks will be provided so that the test valves will receive a closing signal when an opening !
signs 1 is received by the system block v.alve between the pump and the reactor- vessel. t (3) Components in the Standby Liquid Control System heva the capability af being tested during reactor operation. Since this system does not have rapid cetuction requirements, certain test valves have been lef t manually operated with
. procedural control requirements that they be locked in the proper position for
-OFMCML USE ONLY -
-~"FRCHAL USE ONUc-operation when not under tes t. All valves , except one check valve and one shut-of f velve, are outside the containment and accessible to plant personnel. There would be sufficient means to correct in time those conditions that could interfere with proper system operation. The shut-of f valve inside the contaitument would be closed only for maintecance of its associated check valve. Proper positioning of this valve will be assumed by procedural control and by position indication lights that will be provided.
In addition to the valve changes discussed above, tne overall design of the Containment Spray System has been modified for Ge Millstone station. The applicant s tates that the Containment Spray System is composed of two independent systems and that within each system, there are two subsystems containing redundancy in components.
Eoch subsystem has its own pump and heat exchanger (total of 4) whereas a single heat xchanger in each loop (total of 2) and redundancy in pump capacity for each system was proposed for Dresden II. The significance of the Millstone design is that any two of four provided subsys tems are required to assure containment integrity in the event o f a n )CA . If one assumes no failure of the spray headers in the containment, the system provided is reduadant. Hewever, in the event of a rupture or damage by r. missile to one set of spray headers in the containment, the assurance of redundancy is lost.
All componentt. in the remaining system must function to achieve design performance for containment integrity in the event of the ICA. We believe that the applicant should review their proposed design in view of the foregoing potential failure mode but that the conceptual design is adequate to permit approval at this stage of review and that the criterion is satisfied for the Millstone piant.
Criterion 21 Sufficient normal and emergency sources of electrical power must be provided to assure a capability for prompt shutdown and continued maintenance of the reactor fccility in a safe condition under all credible circumstances.
[ E @u FI a w.I..A..,_Us W
{I _I N @u - @v u a-
4FFHCHAEUSE'ONL" '
In addition to the turbine generator, the applicant indicates that auxiliary
- power for the station will be provided from three separate and independent sources.
These are: 1) two- 345 kv transmission lines , 2) one 27.6 kv line, and 3) one standby diesel.
The 345 kv transmission system will be supplied from either the Connecticut Ycnkee station or else from the Manchester substation. Either substation can supply sufficient power for the Millstone emergency loads.
Tne 27.6 kv subtransmission line will be installed to provide construction power and will be retained as an alternate auxiliary power source. Installation of the lines will be made over a different route than that utilized by the 345 kv lines.
^
The supply is to be from another substation provided with circuit breakers to protect against failures on the 345 kv system.
The capacity of the standby diesel is 2500 kva with a power factor of 0.8.
This capacity is adequate to handle the loads of the safety systems in the event of a loss of all power.
We have reviewed the auxiliary power systems proposed by the applicant and believe that redundant and independent sources would be available for all safety systems. In the event of a complete loss of power from all the auxiliary sources, .
including the diesel generator, core after-heat removal can be accomplished for.one hour by the isolation condensers without makeup water. Af ter one hour, . makeup water can be supplied by an auxiliary pumping system.
To further -evaluate the safety. capability of the station design we have con-sidered the condition of a power " blackout." - In the event of loss of auxiliary power from the off-site 345 kv and 27.6 kv supply lines, power would be available from either the turbine generator or from the diesel generator.
~
The 100% load rejection capability (formerly 50%) has been proposed by the applicant in Amendment 1.- The capability will be made available by an increase in the bypass system, Although a 0FFECEAL USE~ONLY^
AFFHCHAL USE ONL "
complete analysis of the effects of this change on the turbine trip, isolation valve closure, and load rejection incidents have not as yet been subraitted by the applicant, we believe that the change would not adversely affect the safety characteristics of the Hillstone Point reactor with respect to those particular occurrences.
In the event of a scram and a loss of off-site power, power would be supplied from the diessi generator with the additional safety feature of heat removal by the isolation condensers. The emergency diesel generator will not be tripped off *he line in the event of (current) overload. It is important that any automatic and
. manual load-rejecting systems be reliable to ensure that unnecessary loads which could damage and disable the generator are dropped during emergency operatior.. We believe that we can resolve the question by revicwing at the final review stage the procedures for loading the diesel generator to prevent overload.
In view of the foregoing, we believe that the diversity of auxiliary power systems proposed by the applicant would assure an adequate source of power for the s tation and that the criterion is satisfied.-
OFFECHAL USE ONLY.
.+ ;
=._ ..
.-9FMCHAkUSE6H
- 22.-
l DESIC FEATURES RElATED TO CONCERNS PREVIOUS 1,Y EXPRESSED BY THE ACMS The letter dated November 24, 1965 f rom the ACRS . to the Chairman ' state ^ that _ the AEC nVuld rive further attention to means for evaluating the factors that may af fect the nil ductility transition temperature and the propagation of the flaws durint vessel life. Aaother letter dated November 24, 1965 to the Chairman in regard to
- Dresden Il from the ACRS stated that it recommends continued study. of the problers associated with pipe-whipping, the generation of missiles which might violate the containment during a postulated accident involving failure of the primary system piping, and the possible ef fect of fuel movement upon reactivity transients.
The applicant has stated that the initial NDI temperature of the reactor head and shell flange material and the plate material connecting to these flanges is 10*F Sile the NDT of all the other material is 40'F. These values have been obtained from 1/4 thickness specimens. The applicant has also stated that the total integrated -
- neut ron exposure that is considered practical to maintain will be less than 1 x 1019 nyt with an expected exposbre of 4.85 1017 nyt. It.was further stated that the 1 x 1019 nvt will result in a NDT shif t of 260*F while the 4.85 x 1017 nyt. will result in a shif t .
of 55*F. The applicant.has stated that under operational conditions the 55'F or 260'T shift would not cause brittle f ailure of the vessel. The staf f has reviewed the information given above and agrees with- the applicant's- estimates of NDT shif ts.- -
The applicant has stated that all large pipes which penetrate the containment are designed so that they have anchors or limit stops located outside of the contain-
, ment to limit the movement of-the pipe. These stops are. stated to be designed to with-' ~l stand the jet forces associated with the clean break of the pipe and- thus maintain- the
' integrity of the . containment. The recirculation line within.the primary containment ill be provided with a system of pipe supports sufficient to withstand reaction forces from a circumferential break and longtudinal split. The applicant 'has stated that-CFFECHAL USE ONLY '
. ~ . . - _ . . - - -. -. ~ .- - -.- ~ .. .. - . . - - -
q
.. :a JFMGAL-USFrONHi ,
l potential missiles such as valve bonnets, valve stems, recirculation pumps, lthermowells,-
occ. will be appropriately. designed against if found necessary to do so in;the future.
Also the applicant has conc *Jded that large, massive rotating components- such as the ,
reactor recirculating pump motors would not have sufficient -energy to move this mass to the containment wall.
We believe that the applicant's design criteria are acceptable at this stage of design but that analysis of detailed designs will be necessary in the future by the staff to conclude that the pipe. whipping and missile problems have been satisfactorily resolved.
In regard to the concern of the ACRS as to the consequences of a major pressure vessel rupture, the applicant states in Amendment No.1 that ir "is asidered to be incredible; therefore, the design approaches mentioned in paragraph 2 of the ACRS '
tetter are not included as design requirements of this plant." It is to be noted _ that t
the-design of the containment is based solely on the basis of a rupture (and sub-sequent consequences) of one 28" recirculation line. We believe thern is insufficient margin in the design to withstand the effects of a large vessel break. This is_true for all G. E. reactors employing the pressure suppression containment concept.
In reply to the ef fects of fuel movement, the-applicant states that the results
.of Lits evaluation indicate that the consequences either--lead .to a negative- reactivity.
condition or else do not significantly affect the results .of the postulated accident; .
e.g. , rod drop transient. G. E. , however, is continuing its studies- of this- ef fect :
by both analyses = and in conjunction with' experimental information being obtained at TREAT and SPERT.
-.0FFHCHAL USE ONLY-
JFFECH/s" u USFrONLb -
ACCIDENT ANAI,YSIS renorni The four major accidents postulated by the applicant (those involving the potantial release. of significant amounts of fission products) explore four possible routes by which fission products might escape from the containment. The rod drop -
accident releases fission products from the fuel within the confines of the. primary coolant system. The refueling (fuci drop) accident releases fission products from the fuel directly to the refueling building, the secondary containment. The steam line break accident releases fission products entrained in the primary coolant directly to the atmosphere. The coolant loss accident releases fission products om the fuel to the pressure-suppression (primary) containment. ,
it is the worst foreseeable case of this last accident (rupture of a large line and a 100% meltdown of the core with associated metal-water reaction) that forms the design basir, for the pressure-suppression containment. _
OFFHQAL USE?ONLY-
. . . - - - - . - . . - . - - . - . _ - - . --_-.-- . ~ . . _ - .-
AFRCHAL~USRONh5
~ 25 ~
The staff analysis of the major accidents is given in the following sections ce well as a description of the applicant's assumptions. Credit was given by the
- staff for release from the 115 meter stack in all accidents except the steam line break. For the refueling and loss of coolant accidents, a Standby Cas Treatment System halogen filter efficiency of 90% was assumed. We believe that a higher filtration effic'iency may well be justified on the basis of final design and test information for this system and by considering current work at ORNL in fission pro-duct release, transport, and filtration evaluation. Table I below sunenarizes the potential doses calculated by the staff.
PCTfENTIAL OFF-SITE DOSES (RDQ.
ACCIDENT First Two Hours Course of Accident at 0.4 mile at 2.4 miles Thyroid Whole Body Thyroid Whole Body Steam Line Break 30 41 5 41 Control Rod Drop 90 7 35 3 Re fue ling
- 6 41 25 1 Loss of Coolant * - <1 <1 240 10-
- Standby Cas Treatment halogen filter efficiency of 901 assumed l
_ The excursion model used in the calculation _ of: the -fuel response to the rod drop and refueling accidents appears to be very senstive to the input values of '
-axial peaking and doppler weighting factors.
For example the fuel: drop accident with !
a total reactivity insertion of 2.3%'and a peak gravity fall velocity _of 13 to 15 f t/
3 see .is calculated to have a peak fuel enthalpy (average across the peak pellet) of .
about 200 cal /gm.- If theselsame' values are superimposed on Figure 109 of the Design: H i
Snd Analysis Report (the parametric s tudy of the rod drop-accident) a value _ of 300- .
l I
to 350 cal /gm is obtained. These-latter energy values imply _ complete melting of i OFMCHEUSEWLY-
_LA*ECHAlrUSErONLY some fuel and the potential for rapid d!spersal of molten fuel in the water of the refueling pool.
We believe it unlikely that a pressure pulse great enough to damage the primary system or accelerate a water column in the refueling pool could be generated by the above peak energies. We also believe that any steam generated would be quenched in the 50 foot deep refueling pool. It is our opinion, however, that the apparent sensitivity of the excursion model discussed above warrants further inves-tigation in the future.
The leakage rate assumed by the s taff for transport of fission products from the cont ainment to the reactor refueling building was 0.71/ day at a pressure of 47 psig. The applicant has stated that the containment will be tested at the design pressure of 62 psig before penetrations are installed and subsequently ok rate tested at the calculated peak accident pressure (currently 47 psig) and again at lower pressures to establish the leakage characteristics of the system.
The acceptance criterion stated is that the leakage rate at the calculated peak acci-dent pressure would not exceed 0.5%/ day during the test. Since the test is performed under different (dry) conditions than the postulated accident conditions (steam-nitrogen-hydrogen) we believe it appropriate to apply a correction factor to relate accident to test conditions in the accident dose calculations and have therefore used a leakage rate of 0.7%/ day. Since a turbulent flow regime has been assumed by the applicant and the a taff, the leakage rate would not vary appreciably over the pressure range from peak accident to approximately 10 psig. In addition, most of the release occurs at pressures above 10 psig. On this basis, we have assumed a constant leakage rate of 0.7% per day from the containment structure.
The meteorology used for the staff calculations was that described in ORO-545 which provides a parametric approach to establish the " worst" atmospheric dif fusion condtions as a func tion of dis tance from the reactor. The " worst case"
-OFFECWrUSE ONLY-
\
Tp]p p(r HT A R TiTO T J o Bn LA v1 M f7 %*
%sL au uvunu.7 dif fusion condition was sisumed to exis t throughout the particular time duration chosen. The two hour doses were calculated for a distance of 0.4 mile from the recetor.
The staff believes that this coastal site may be subject to fumigation conditions of several hours duration as discussed previously. We estimate that the fumigation turbulence could widen the plume from an elevated release in the vertical direction rapidly enough to cause it to reach ground level as close as 1/2 mile.
Va estiamte that this could increase the concentration of fission produces inhaled by a recipient at 1/2 mile by an order of magnitude. In the case of the control rod drop accident this could raise the thyroid dose significantly above 10 CPR Port 100 guidelines. We do not believe that the fumigation conditions would greatly increase the doses calculated at the low population outer boundary of 2.4 miles.
While a closer analysis might be able to " trim" the conservatism from the rod drop accident fission product release to the point where 10 CPR Part 100 guide-linas would not be exceeded, we believe that the existing " sneak path" for fission product release should be eliminated. Clarification of the " sneak path" wherein fission products are released to the atmosphere without filtration is given in our discussion of the Control Rod Drop accident. Elimination of this route and using exis ting systems would reduce the of f-site doses by providing a 30 minute holdup and at least a f ac tor of 10 from halogen filtration in the Standby Gas Treatment Sys tem.
Control Rod Drop In the control rod drop accident it is assumed that the bottom entry rod has been fully inserted and has stuck in this position unknown to the reactor opera-tor. The collet fingers are assumed to unlatch and leave the rod in the core when the drive mechanism is withdrawn. Subsequently, the rod falls out of the core inserting on amount of reactivity corresponding to the worth of the rod.
P T J/fHi A U TT TAD /fMT1T V
% u uvacua.a v eJR %t/ A N u u--
28 -
The applicant has stated that hot standby is the worst operating condition at which the accident could happen both because a higher energy release is calculated for this condition and because a path for the unfiltered release of fission products could oxist (through the mechanical vacuum pump on the condenser). A rod worth of 2.5%
in reactivity was chosen as the highest worth rod permitted by the rod worth isinimiter.
The cpplicant's analysis indicates that the peak fuel energy density would be about 200 cal /gm (average across the peak fuel pellet) resulting from a minimum reactor period of 8.5 milliseconds and a total energy generation of 4000 Mw-sec.
The large difference in calculated dose to the public between the staff and applicant's analyses are due to the differences in release fractions and plate-out fcetors. The applicant assumed that only 1% of the noble gases and 0.5% of the halogens were released from rods calculated to have fuel clad damage. The staff analysis 4 2med that 100% of the noble gases and 50% of the halogens would be released from rods with peak fuel enthalpies greater than 170 cal /gm (this corresponds to 1% of the total noble gases and 0.5% of the halogens in the core). We believe that some decontamination would be realized and acknowledge that the amount of fission products corried to the condenser would be limited by the closing of the steam line isolation velve. The 50% halogen plate-out in the primary, the 10% carryover of halogens to the condenser and the 50% plate-out in the condenser assumed in the s taf f analysis lead to a reduction of 40 in the source term. The total reduction in halogens assumed by the applicant is 2.8 x 10 4.
The consequences of the rod drop accident could be greatly reduced by providing automatic isolation of the mechanical vacuum pump system. This would insure that the effluent from the condenser would be routed through the Standby Gas Trootment System if high radiation levels were detected in the line. This would provide a 30 minute holdup of gases as well as at least a factor of 10 reduction in the released halogens. We recomend that provisions be made to automatically isolate evv" ~ n n ~,.y)p mm
_ QQ q~ ;
Q$' E' h s !Q y
e mag e n A if tr e m ee Am uvurm UUb UN 29 -
the mechanical vacuum pump system on the main condenser on a high radiation signal and that subsequent of f-gas from the condenser at hot standby be routed through the Standby Gas Treatment system.
Fuel Loading Accident The fuel loading accident is the postulated drop of a fuel element into a near-critical fuel array obtained by withdrawing two adjacent control rods during loading. The accident would require violation of a number of procedural require-ments and the failure of interlocks which prevent fuel handling over the reactor while rods are withdrawn. The fuel assembly reactivity worth in the postulated
" worst" configuration would be 2.3% and the calculated maximum reactiW ty insertion rate 10% per second.
As in the control rod drop accident, the applicant assumed release of
.ssion products from fuel rods perforated as a consequence of the accident (energy dansities greater than 170 cal /gm) and only 1% of the noble gas and 0.5% of the halogen inventory in each perforated rod.
The applicant has calculated the rate of. steam generation during the excuraion end has s tated that the maximum energy release rate from the peak fuel to the water would be about four times that at steady state operation. This relatively slow release of energy is a consequence of the low thermal conductivity of the 1102 fuel, q The formation of steam would be only about twice the steady state value since the water in the vessel is 100 F below the boiling point and the amount of heat required
.for vaporization is larger at the lower temperature. We therefore expect that no expulsion _of water would be experienced from the 50 foot-deep pool, that steam would be condensed before reaching the pool surface, and that some of the halogen gases would be absorbed in the water. We also believe that any pressure pulses generated in the core would be the result of some mechanism for rapid dispersal of the fuel into the pool water (such as the rupture of _a water-logged fuel element),
bb .be bh N 1
a
= .
RMCML4JSE-ONLY -
The staff analysis assumed 100% of tie noble gases and 50% of the halogens were released from the damaged fuel elements and that a plate-out of 50% of the hologens occurred in the refueling pool. The fission product inventory was released to the environment through the facility stack at the rate of 100% of the building volume per day. With this assumption, 20% of the fission products are released in the first two hours and 75% in the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. A halogen filter efficiency of '9%
w:s also assumed in contrast to the 99% efficiency assumed by the applicant.
Steam Line Break Accident The reactor steam system consists of two 24-inch main steam lines (MSL).
Each MSL is connected through a header to two outlet lines from the reactor. Each outlet line is equipped with a flow restrictor while each MSL is provided with two isolation valves (one on each side of the drywell). The isolation valves are specified i close in from 3 to 10 seconds with a 1 second instrument d? lay.
The applicant has assumed that one of the two steam lines is completely severed in the pipe tunnel. The steam flow would be restricted by the flos limiters within the drywell and backflow through the turbine would be stopped as the turbine stop valves rapidly close. It was es timated that a total of 50,000 lb. (11,00 lb.
steam, 39,000 lb, water) would be released by two phase flow through the steam lines in 11 seconds. The removal of this amount of primary coolant would not uncover the core even neglecting the 20,000 pounds of feedwater which would be added in this time period.
The maximum calculated impulse pressure in the steam lines from a discrete plug of water (unlikely) is 1350 psig for a sudden stoppage. This is greater than tha design pressure of 1250 psig but within the test pressure. A continuous slug of water could exert a maximum pressure of 1070 psig.
The applicant's dose calculation was based on Dresden Unit I iodine
! inventer.ies and an elevated release from a rising steam cloud. We believe that the
-92FEG%EUSE ONLL
Ud"5U A U U YAUT Ul& 11 '
uvurau wa;.4
- 31 -
steam cloud would probably rise a few hundred feet but not in units of miles as the cpplicant claims.
The staff's calculation assumed that (1) the water and steam blowdown totaled 70,000 lb. during the 40 seconds before the pressure has decreased to 185 psig at s which time the engineered safeguards would function, (2) the fission product inventory was assumed to be equivalent to that in 70,000 lb. of reactor water when the stack release rate is at the estimated licensed limit and (3) the isolation valves failed to close. The inventory is a factor of 10 higher than that expected during normal operation of the facility. A total of about 100 curies of I-131 and 1-133 w:s assumed released at ground level. TID meteorology was assumed. The thyroid dese at 0.4 mile was calculated to be '.0 rem. The contribution from noble gases is small in this case since the noble gases are not accumulated in the primary system
.ter but are continuously removed by the off-gas system.
Since the steam line break accident was considered the limiting accident in the staf f's site evaluation report, some further discussion seems appropriate. In the previous site evaluation report, the activity in the primary coolant system vas assumed to be 50 ue/cc with the isotopes in their equilibrium concentration. The 50 uc/cc value was chosen as an estimate of what the primary coolant activity might be at the maximum permitted stack release rate. The dose obtained from this accident for a ground release of steam was about 300 rem at 1/2 mile.
Operating data at Dresden 1 (Ref. Dresden II, Plant Design and Analysis Report, Amendment No. 4, p.17-4) indicate that the total activity when the stack release rate approaches 0.7 curie /second (the allowable limit at the Dresden site) would be about 20 uc/cc. Further, due to the continuous operation of the reactor cleanup system, the longer lived isotopes would be in less relative abundance than in an equilibrium mixture. Using the Dresden I operational data as a basis for an 1
f
-FFECHALJUSErONLL estimate of the maximum halogen concentrations, a dose of 30 rem was calculated at 0.4 mile.
Loss of Coolant Inside the DrYvell The break of a pipe in the primary system within the dryvell is considered th) Maximum Credible Accident because of the large potential for fission product rolease. A break in the "second line of defense" against fission product release, the primary system, could lead to violation of the "first line of defense;" the fual clad. This places complete reliance on the last major barrier to fission product release; the drywell-suppression chamber containment. This accident has also been chosen as the basis on which the containment is designed.
The applicant has stated that a range of coolant loss accidents has been onelyzed and ; hat the largest break, a circumferential rupture of the recirculation 4ne, results in the maximum doses since a small break would result in longer heatup times and allow some fission- product decay. Two engineered safeguards must func tion in this accident to prevent major core melting: (1) one of two core spray systems, and (2) one of two cooling loops in which water is taken from the suppression pool, p::ssed through a heat exchanger, sprayed into the drywell and returned to the suppression pool via the vent lines.
Af ter the blowdown, which has been calculated by the applicant to take more then 20 seconds, the heatup of the fuel is determined by a computer code which calculates fuel and cladding temperatures and other parameters. The code considers decay heat, stored energy, chemical reaction energy and thermal radiation between rods end channels. Peak pressures during the blowdown are 38 psig in the drywell and 15 psig in the suppression chamber as opposed to the design pressure of 62 psig for each chamber.
In the absence of a metal-water reaction the containment pressures are reduced to about
! 3 psig in the first few minutes due to the steam condensation effected by the drywell Q T MM
__-uu w a;f PJ 3Sa $ RJII uX -
a .-
JMGAL4JSE ONLL_
spray. The core spray, initiated by low primary system pressure, was stated to come on within one minute following the break. The core spray would prevent the UO2 from recching 3000 r, the recrystalization temperature, if initiated as late as 5 minutes citer the break. For the case in which the core spray functions properly, the amount of zirconium that reccts would not give rise to a flanusble mixture from the evolved hydrogen even if the containment atmosphere were not inerted. However, the applicant proposes to inert the containment atmosphere to prevent recombination of hydrogen.
The applicant states that proper functioning of the engineered safeguards af ter o large coolant line break would result in clad damage and resultant fission product release from the plenums of about 45% of the fuel rods. The resulting doses would be very small.
It is the s taf f's opinion that while all possible steps should be taken to prevent core melting during the coolant loss accident, the design basis of the contain-ment should be the containment of the fission products from a 100% core meltdown with attendant metal-water reaction. The applicant has es timated the amount of metal-water recction associated with the 100% core melt and has discussed the effects of the energy and hydrogen generated on the containment. It has calculated that about 27.5%
of the zirconium could react.
The staff has analyzed the loss of coolant accident assuming 50% release of the halogens and 100% release of the noble gases. A 50% plateout of halogens was also assumed. The leakage rates were assumed constant over the duration of the accident (0.7% per day from the containment to the refueling building and 100% of the refuel-ing building volume per day through the 115 meter facility stack). Results of the staff calculation are given in Table I.
i *. 54 -*M 4
- hN v u a % . .. _: . w.m mua
{ THC4AL4JSE ONLY-The primary dif ference between the staff's and applicant's calculations is the meteorological assumptions. The applicant has assumed that a 15 hour1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> wind persistence is the maximum to be expected and the staf f has assumed a constant wind direction for th2 duration of the accident. Using Type B (ORO-545) meteorology, the worst 15 hour1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> doso at 1/2 mile would be 30 rem compared to the applicant's calculation of 12 rem.
Thus the " course of the accident" dose of 240 rem to the thyroid at 2.4 miles is an upp2r limit of potential consequences. The staff calculates that the peak rate of dose delivery to the low population boundary (2.4 miles) is 20 rem / day. A dose estimate for shorter wind direction durations may be obtained by multiplying this number by the length of wind persistence assumed.
CONCLUSIONS On the basis of the foregoing safety evaluation of the Millstone Nuclear Power Station, we have concluded that there is reasonable assurance that the facility can be built and operated at the proposed location without undue risk to the health and safety of the public. In the course of our review, we have identified several items which will require particular evaluation at the final review stage prior to reactor operation. These are the following:
- 1. The adequacy of the Control Rod Thimble Support structure.
- 2. The ef fects of a failure in the control rod selector system causing the withdrawal of 2 control rods.
- 3. The effects of a loss of redundancy to the Containment Spray Systems in the event of a rupture of a spray header.
- 4. The transient effects of a sudden loss of load since the bypass system now has a 100% capacity.
- 5. The prevention of diesel generator overload.
GFHCHAL USE ONLY- -
s,.
%7FHCHAL USEONLY-
- 6. Results of meteorological site tests.
- 7. - Plant safety as affected by excessive' tidal water.
- 8. Containment integrity in the event of facility stack failure.
None of the items- listed above are significant-enough to be considered as
-unsurmountable safety problems, and we believe that they may be resolved with addi-tional design details and appropriate analyses during construction and prior to opsration of the facility.
J
-OFFECLAhUSEONLY-
. i n .
TABLE 2 Suinnary of Accident Analysis by Applicant for Millstone Accident Cause Reactivity Initial Power Minimum Peak Power Total- UOy Insertion Level and Temp. Period Generated in' Energy. Melted
, (millisec) Excurs io'n - During- (lb)
, (MW) Excur-
.l sion (MW-sec)
$ Drop-out Rod becomes dis- 2.5%, 3.8%/ 10- rated, 8.4 10 5 4000 ---
connected fros' sec 547*F , ,
. . drive and '. - Os- !
t .,t
', - i out of, c,re. o'.fal.ls. 7
- :, ;-
- 1 4 .. , . 1, -
. , . s ,. p - .
- . le. . ,
. . . .!- . + ? .
, Fuel Losding- Drop fuelassen . '2.3%, 10%/aec' 10~8'inted.' .
- 3.9 1.5.10 5: - 26001 ---
- i ~
, : 'bly into near .
~<200*F
- x I crikical.2 z 4,,,
~
- ). ' ,
1 'd' t
array ,
, , .i ,.
.?
i r
r
- _ i ,e- , , .
.. Steam line break' l Rupture'of Steam Ratgd power,l .,
, Line h eside Dry . .547 F' ,
well- '. ; i. .
. .e e, ,.-
. .olant Los s , No Double-ended ---
Rated power, ',
Core Spray, 100% break of recirc.' 547*F .
Core Melt with line -
. _. i metal-water'reac-' .
t i tion
^
,'~.
9 i
^
2
- 9 '
4 .,
. c.
y- s t
v --
r +
F b
Page 2 (Continued) ,
Summary of Accident Analysis by Applicant for Millstone Accident Rods with Peak Fuel Previous Irradiation Fractions of F. P. Quantity F. P. to Primary ^
- Burned- Temp.-(*F) 'of fuel . Released from fuel ,
out clad' Rod Drop-out 300. .. <5000 500 days.@ 2400 MWt 11 coble gases 0.5% .03 Megacuries Noble' '
- ~
to 30 min. before Halogens in ruptured .02 Megacuries Halogen f .'
, accident rods '
. .r. .,- ~
7, y- . ~
Fuel loading . ^ 224_ <5000 500 days e 2400'MWei- 1% noble gases 0.5% D1 Megacuries Nob'l$
,. ' ; _ -to 24 hrs. before drop Halogens in ruptured
.006 Megacuries Halogen
. l . .. *
, Ms 'e
, kg : '., . , , . , ,.
. - ,4
- i.~ 'I t .,.
- : A -
asteau Line
--- .s -- n. . -.- ---
0.3 Curies 1-131 *.1 Curies '~
, g,, *
,s ..
Break . t {f ..' ,- -
1-133, 17 Curies other I. '44
{-
,. ,, curies noble, ',
.+L '
s ,
i e
._ Coolant Loss, -- . , [- *--- ., 500 days'@ 2400 MWt 100% noble, 50% halo- at 30 min: 350 Megacuries-
~
, No' Core' Spray, *4'
_until accident . gen, 50% volatile noble, 190 Megacuries halogen,.
1001' Core Malt .
( solids, 1% other 180 Megacuries volatile solids' ;
with Metal- ' -
_',, r ,
, solids,10% of halo- .3 Megacuries other solids, ;
- -te'r Reaction ; - '- , gens are organic
{
i 1; . . .
~+ .
% g 4 i / .
O- i I .
4 Page 3 (Continued) Sununary of Accident Analysis by Applicant for Millstone Accident Metal-Water Reaction Noble cas Stack Maximum 2 hr. Maximum Total Assumed Discharge Rate Exposure in Rem ( mile) Exposure in Rem (% mile)
E ,
(Curies /sec)
W ole Body Thyroid W ole Body Thyroid i-
-3 -3 3 Drop-out .27 lb moles B2 evolved 19 1.3 x 10-6 1.3 x 10 1.4 x 10-6 5.3 x 10 from 12.3 lb Zr reac-(at 1 min.) i* :
. 'ted . 30 W sec. -
1 ,
s 4
_3
.. . . _ 3 _t L Fui?'t.oading 24_ W sec -
- i, .11 '
6 x 10. 4 i 5.8 x 10 6 x 10 o 1.6 x 10 (at 1 min.)
1
, { ,
& .,,' ,a 4 4
Steam Line Break '
. i
,--- 2.3 x l'07,6 .1.4 x 10^1 ,
.;- ,, . . . . 4
. . i f-. Coolsat Loss,'1001 625 lb soles H2 evolved 4.'6 1.9 x 10',3 1.3 x 10-2. 3,t ,
12
- Core Melt with from '27,Q Zr to reac- (at i day)
- i Metal-water Reac- ted 7 x 10 W see -
- jtion (27.5%,of Zr reacted) ~
- - _t t.
_ , .. s * - . %
=
s.
i s i j . t.
~.
=
9 g A i- (
4 i j-c
,s , . -
e
- 7 hw