ML19256B764
| ML19256B764 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 08/09/1979 |
| From: | Moseley N NRC OFFICE OF INSPECTION & ENFORCEMENT (IE) |
| To: | Haynes R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| References | |
| TASK-TF, TASK-TMR NUDOCS 7908290023 | |
| Download: ML19256B764 (52) | |
Text
4 1
nc r eo.
ROUTING AND TRANSR',,TTAL SLIP
/ s w~
1---~
To(u...,,,,,,7,,,,,,,,,,,,,,,,,
3,,,,,,,,
so j
,TCn /f8 y,pr3
, j.'l o
a
c
.....r..
s-
... r i sa rea
,4
,. e mm ! ', T W
oars
.. ro.u a ri e
?*w N
I
/ '9, 2
... r.4a gc,,...
?.' c, a W",
- '.".u"...
I.%
4
...ri...
=as =a f
4,
^
t' d
oars
...=4rv.
/*** en m
'* n
'l *.
REMARKS
/
fye [c /* r &'C'fn
- ^P f'
h !,
. t..*
- cc,,., - : - I c,, O /n e A d -c
- r-
~
o i
.[f'/SS/
(mc/
rY O :"' 'M ?
Y'
~
lj
~/
t s.
A (M
C/
p w~,-=ppy,sr e s'C'k<-y,'c/cc7#n/ /$n m W lb
l E k' '
~
(( [/f%[ 5f'4;j' V^
" Y # '0 '
,w,;efefe<<'WM*Y
-Whru
$F $W
~
(""
,Q,,,
f }pn fs A hEfW A OM'/ O
- -x, y,,,,,c.=,,4 er=re n me - QM' =
,,de p em i S W f X"'L='"* '"W
),(e,/ ',utzc401' A s- /N~n nhe A'^W h(+
, /
._% /_),/r~f'sl /Y 0l'*)
- r* Y a
Do NOT use this form as a RECORD of approvals, concurrences, disapprovals, clearances, and similar actions y
fRCM tNam*. Ott*<
- ar'nhol or I*callo")
h[f~/h /*
-~ f
- a OPTIONAL FORM di ese-m-etnes-t s:2-sea are 5041-100 avousr tes7 G3 4 FPum ( 4 t C FR3 100'91.200
\\
\\ k
($g@ n p 1911 283 e
V 79082900 s
DRAFL' 1.
Problem:
The licensee's emergency classification system does not provide for dgraded response to potential emergencies.
Presently, there are only three classes of emergencies:
Personnel / Local Si.te General Response to Personnel /t.ccal emergencies is, naturally, limited in scope.
Response to a Site Emergency is extensive, involving full iglementation of the EP, i.e., notification to offsite agencies, on and offsite emergency environmental monitoring, accountability, etc.
The licensee's General Emergency category, while declared based upon action levels which indicate more severe emergencies, does not result in a response any different than the response to a Site Emergency.
This "all or none" type response philisophy may result in a psycholog-ical tendency to avoid declaring an emergency because the conditions for declaring the emergency, while met, are not interpreted to Warrant the extensive response action that would occur if the emergency were declared.
If this psychnionical tendency is ccupled with ill-defined action levels and an "it can' t really be happening" evaluation, it is understandable how delays in emergency plan implementation can occur.
During the early stages of the THI emergency, tne call-in of selected management by the operating crew resembled what would be expected to cccur in response to the "Fmergency Alert" category discussed in Regulatory Guide 1.101.
It is assumed that, if the licensee had had such a classification (with an appropriate low threshold action 3
(f Q xlyp'#
\\'
fS'h 1911 284-
= = - - - - -
- =
DRAFT 2
level) that senior licensee management, the NRC, State and local officials would have b.en involved much earlier.
Consequently, when indications escalated to those obviously indicative of an emergency, the aforementioned groups could have been "in place" and in commun-ication with the facility and each other.
It is reasonable to assume that an " alert" would have helped to reduce the appearance of conflict-ing information being provided by the licensee and NRC and, with this expertise avuflabic early, perhaps core damage could have been mitigated.
Recomcendation:
a.
The Three Mile Island Emergency Classification System should be
\\p modified to include the classes of emergencies described in
,[
Regulatory Guide 1.101, Annex A, Section 4.
,;;m\\
3
\\
q-w C g b.
The licensee should develop more specific action levels applicaole 4
to the various classes of emergencies.
\\
^
c.
The R. G. 1.33 emergency operating procedures should have an action statement inserted which leads tof the declaration of the proper class of emergency.
2.
problem:
Site management does not appear to follow ident fied emergency planning i
proolems to timely resolutinn.
Interviews with licensee employees indicate that problems related to training, equipment and procedures were identified but not resolved.
In general, responsibilities for maintaining preparedness are delegated to individuals in different departments.
Within a particular department, the responsibilities are further delegated.
r;(Whh' 1911.285
DRAFT 3
.The Supervisor, Radiation Protection and Chemistry, according to the Station Manager, is the individual who has responsibility for the entire emergency plan.
This assignment, while not formal in nature, i.e., sade via memo to all personnel, seems to be understood by most employees, with the exception of the Supervisor, Radiation Protection and Chemistry himself.
Itishisunderstandingthgtheisonlyresponsi-ble for those areas where the procedures indicates he is responsible.
"ar those areas in which the procedures assign him dresponsibil ty, he has delegated the responsibilities to two individuals - the Radiation Protection Supervisor and an Engineer.
These individuals, the Radiation Protection Supervisor, in' particular, rurther delegate a portion of the responsibilities to four Radiation Protection Foreman.
From there, the responsibilities are either assumed or further delegated to the Radiation / Chemistry Technicianc.
The Supervisor. Radiation Protection and Chemistry does not " bring together" the delegated responsibilities within his department to ensure they are accomplished.
The Supervisor, Radiation Protection and Chemistry appears to avoid say involvement in g
e::ergency planning activities that are not the responsibility of his department.
With the recent addition of a Superintendent or Administration and Technical Support interposed in the reporting chain between the Supervisor.
Radiation Protection and Chemistry and the Station Manager, there is no direct link to senior management for the Supervisor, Radiation Protection and Chemistry to accomplished assigned tasks.
The Emergency Plan at all powor r'eactors encompass all departments and disciplines on the site - operations, chemistry, administration, training, maintenance, etc.
Unless a single individual is appointed as an emergency coordinator, and is charged with the responsibity for A
'\\
pss@*
1911 286
DRAFT 4
managing the overall Emergency Plan effort, parochialism rievelops and certain things are not done.
,b' e. Recomendation:
V' Y
C
[ [ The licensee should be directed to appoint / designate one individual as g
an Emergency Planning Coordinator.
This individual should be given Mb overall responsibility, and authority commensurate with this responsi-Q bil i ty, for maintaining the emergency planning pi ogram at the site.
C' In this manner, he would insure that any delegated tasks are accomplished in a timely and complete way.
This recommendation is applicable to all power reactor licensees.
3.
Problem:
The licensee's provisions fer personnel decontamination facilities do not adequately consider the need for alternate locations offsite.
While the licensee has designated two "Washdown Areas" that are located offsite, these areas are intended to be used for vehicle decontam-ination.
Twn "Washdown Area" emergency kits exist and are maintained at the site.
However, these kits do not appear to contain equipment and supplies that would even permit effective decontamination of vehicles.
The following is a listing of equipment set aside in the Washdown Area kits:
Small envelopes 50 Smear papers 1 box RadiaLion Lape I rol!
, c('
.ms4
} h
\\
DRAFT 5 Information notebooks 1 Masking tape "2" 5 rolls Pads 4 Pens / pencils 10 Felt tip marker 1 GM Survey Meter I High range dosimeters 5 Dosimeter charger 1 there are no huses, buckets, brushes or readily accessible water supply. Additionally, t'here is no equipment set aside that Will permit these areas to be used for personnel decontamination, i.e., no source of water, clothing, soap or other decontamination agents, towels, etc. During the TMI investigation, several inte, views of licensee personnel indicated that the adequacy of the 'icensee's "Washdown Areas" had been discussed in the past. Of particular concarn was the handling uf potentially contaminated rur>-off. Apparently, the licensee did not consider that the normal in piant pe sonnel decontam-ination facilities would ever be unuseable or tnat evacuees from the site might be contaminated. Consequently, when the in-plant personnel decontamination facilities were inaccessible'and evac.lcos from the site were contaminated, a persnnnel decontamination station was established at the 500dwitchyard. lhis was not a designated Washdown Area and was also not equipped to support persunnel decontamination. xe v.- g,C Recomendation: ?$ 9 J Q g The licensee should be directed to establish an adequately equipped ~\\ 'O y personnel decontamination facility offsite. b o m J p? ' g (. 1911 288
UKAr I 6 4. Problem: The licensee's Emergency Repair Team is composed of shift maintenance werkers. During emergencies, it is reasonable to assume that valves must be operated, pumps started, or other functions performed of which shift maintenance personnel would have no knowledge. During the THI accident, valve alignments, racwaste system operations, etc., were performed. However, since the licensee's emergency organization did not include Auxiliary Operators, these functions were not performed by the Repair Party Team, but by auxiliary operators and others who functioned at the direction of the Unit, 2 Supervisor of Operations. These individuals did not take appropriate precautions or coordinate their " repair" activities as they would have were they a part of the normal emergency repair team. .g Recommendation: wJ x,k, Include, as repair party team members, individuals with varied background y in operational aspects of the facility. \\ .. C' 5. Problem: The licensee's site protection force does not have provisions for implementing contingency security measures under conditions where normal measures must be abandoned for reasons of safety.
- Further, there are no provisions for maintaining a log of incividuals (employees) either entering or leaving the site once an emergency has been declared.
Consequent.ly, accountability and assem.bly proceaures could not be implemented it conditions deteriorated. os e m\\ _,m -ua- @(g$j6%@-)6 tug-1911 289
V DRAFT \\ 7 N}v NB Rocommendation: - b ontingency security measures should be developed to include provisions C for continued accountability of all persons onsite. 6. Problem: Recovery plans and the functions of the 1icensee's corporate HQ support groups (Division Support and GPU) are not detailed to any great extent in the licensee's emergency procedures. Consequently, errective use of their support or a smonth transition from emergency to recovery operations did not occur. There was some confusion regarding who was in charge of various of the recovery operations. This had a partic-ularly negative impact upon the radiation protection aspects of the emergency organization. Non-licenseepersonnelwerebrougEintothe picture and placed in positions of authority over the normal licensee personnel and supervisnes. L 5 Recommendation: b y / Detailed recovery plans are not suggested. However,he' licensee's ,v emergency plan should address the major functional areas of the recovery [ organization, specifying, to the extent possible, the individuals who will be responsible for coordinating each area. Since emergencies do V not have a discrete endpoint, the relationship (i.e. authority, reporting chain, etc.) of the emergency organization elements to the elements of the Division Support / Recovery organization should be described. g 1911.290
DRAFI 8 7. Problem: The NRC has traditiona'lly emphasized the offsite effects of accidents. The consequences of accidents analyzed for siting purposes are used as the planning base for development of Emergency Plans and Implementing Procedures. The analyses of these accidents, however, only address the offsite/ site boundary consequences. As such, emergency plans and implementing procedures have evolved to more comprehensively address offsite related response actions while actions related to in plant consequences are much less clearly defined. This has resulted in a generic weakness. The configuration of emergency organizations, the scope and content of Implementing Procedures the design of the facility and equipping and selection of emergency facilities do not adequately consider the in-plant consequences which may result from tne analyzed accidents. ~ In the case of TMT, the emergency organization was not structured to cope with the extensive in plant radiological conditions which existed. Procedures did not address access controls, in plant radiation surveys, dosimetry, personnel decontamination, and general exposure controls. The licensee had not cunsidered which, i r any, of the nonnal radie en protection procedures were intended to be applicable under emerger,: y conditions and Emergency, Plan Implementing Procedures did not address This lackhre-planning, organization Nn$ procedur @ " " N these areas. necessitated their being developed on-the-spot. This, in turn, caused confusion within the emergency organi:.ation since the methods imple-mented were unfamiliar La those required to implement them. g,is h' 9 1911 291
DRAFT g Recommendation: The NRC should develop a scenario presenting a postulated set of in-plant radiological conditiuns t.u be used as a planning base against which licensees should-develop theigganization, facilities, equipment and procedures to cope. g[ W 8. Problem: Licensee's fomulate their own drill scenarios, initiate and teminate the drills, select the dato and time and sometimes pre-assign the drill participants to emergency functions. Some licensees rehearse their drills several times in advance.f notifying the NRC that an annual drill has been scheduled. The.efore, if the NRC observers this drill it may not accurately reflect the maintained readiness of the site, but rather a yearly peak attained readiness. Drill scenarios are written to reflect the scope of accidents, analy:ed in the FSAR. The scenarios include events that the planners think are realistic and would occur during an act.ual occurrence of the event. Often, multiple failures are included in the scenario so that the consequences are severe enough to warr&nt a response. These " failures" are almost always of a type that result in significant offsite effects. The precipitating event is usually " classic" in nature, i.e., broken pipe, stuck valve, etc. It is of significance to note that the TMI conducted an emergency drill on November 2,1978. The scenario for this crill involved a LOCA wf th severe core damage, fuel melting and little core cooling in Unit 2. The same scenario was run in Unit 1 on November 5,1978. The C i 1911 292
DRAFT 10 in plant involvement was very limited, and dose rates, " selected" for nw the areas were all low in comparison with Wh&-wTg encountered during the March 28 accident. Recommendation: The NRC shnuld develop a set of standard drill scenarios that could be administered on a surprise basis. , 9. Problem: Effluent monitors used to assess projected consequences do not have a range adequate to evaluate consequences of releases resulting from accidents. The majority of the instruments installed in reactor facility gaseous 6 effluent pathways have a maximum range of 10 counts per minute. This maximum count rate, modified by such knowrd as flow rates, detector efficiencies and atmosphoric dispersion are translated into an initial dose projection or concentration applicable to a partir.ular location. An evaluation of these monitors indicates that, at their extreme range, the corresponding whole body and thyroid dose projections (or concentration projection if action guides are in factors of MPCg) are below the lowest level at which protective actions or notification would be considered. At THI Unit 2, the HP-R-219 monitor is used to project offsite conse-quences from accidental releases. The maximum range of the gas channel 0 10 cpm, corresponds to a 2 hour whole body projected dose of 180 mrem at the area of max X/Q. This number was derived from a X/Q value (maximum) of 6 x 10" sec/m3 and a vent flow of 90,500 crm. h og@pfM@efu'm g bcu 1911 293
DRAFT ~ U The iodine channel on these monitors presents a different problem in determining projected thyroid doses. The detector looking at the charcoal cartridge will also detect the cc=pton spectrum of noble gases. While charcoal does not retain noble gases in appreciable quantities, the high noble gas to radiciudine ratio [s'till results in a significant amount being retained in charcoal. As the charcoal becomes " loaded" the total activity on it causes the meter output ta be pegged. This necessitates a change out of the filter, to bring the meter back on scale and to permit gamma isotopic analysis of the cha rcoal. In most facilities, however, the location of the charcoal cartridge is such that the high general area dose rates will make accessibility to the cartridge risky or impossible. Recemmendatien: The ranges of the gas and lodine channel be adequate to permit the determination of at least 10 times the protective action guide for which evacuation is considered as the protective action. Silver zeaolite should he used in gaseous effluent streams for sampling radionuclides. 10. Problem: Tne use of either a charcoal or a selective medium for radiciodine may result in inadequate surveys. The use of charcoal can result in the overprojection of iodine concentra-tion due to the retention of noble gases in the charcoal. If the charcoai is counted in the field with a GM instrument, everything gu@, kkW@$WW M \\ g s )) V UO 1911 294
DRAFI 12 detected may incorrectly be assured to be iodine. fhe extremely low efficiency of the GM detector also introduces additional problems relative to MDA, sampling time, etc. Counting of charcoal with a single channel analyzer can also lead to similiar overprojections due to the errect of the compton spectrum from retained noble gases. The sole use of a selective medium has an even more serious shortcoming. Since it has almost no_ affinity for noble gases, because a sample indicatesnoactivity(<MDA),itmaynotbeassyedthatnoactivity / is present in a plume. Usually, the location for sampling is selected based upon direct radiation levels on the ground. The sample is taken at an area indicating the highest radiation level. If the plume is a,d ul ,,36.. M vuYeea2lv have no way of knowing he is ruuding " shine". Consequently, if he were to take air sample using a selective medium and it indicated no detectable activity the surveyor may reach the conclusion that the plume contains no iodine, when, in fact, he did nut even sample in the plume. Recommendation: Emergency environmental air sampling should make use of a dual media (inparallel)samplingtechnique. A charcoal cartridge would be used as i an indicator to verify that the sample was, in fact, taken in the plume. The selective medium, silver rea91ite, would be counted to determine iodine concentration. p w ~ A-Qls m ' \\. \\p a m w 1911 295
DRAFT 13 11. Problem: Protective Action Guides and Protective Action Levels do not adequately consider potential radiological impacts. Under existing guidance from the EPA and NRC, an actual release of radioactive materials must occur before the guidance is considered to be applicable. Licensees calculate projected consequences based on actual source terms or field measure-cents then compare the results with adopted protective action guides, caking appropriate recommendations to the state ocr.cer-ing_pr4testiver" aettens,f' The TMI accident, however, indicates that the aforementioned philosophy cay be short s ighted. In comparing the projected doses calculated during the periods when uncontrolled releases were occurring, one finds that no protective actions were warranted. (While selective evacuation of children and pr9gnant woman was recommended by the Governor, it was intended to be conservative, but not necessary as a result of the comparison of projected doses with previously adopted action guides.) On the other hand, the volume of radioactivity in containment was of such a magnitude that had it been released, the action guides for evacuation may have been exceeded even before the ' tate could have been notified. Uncer such conditions, Drutective Action Guides no longer re resent an avoidable dose, but more correctly a dese ccmmitment. (With this perspective, it may have been more saprcpriate to institute protective actions arouna TMI that were more extensive than selective evacuation of children and pregnant woman.) It would thus appear that, under certain conditions, it may be prudent to icolement various of the range of protective actions, including w.cJo. o w a re.:.a e o w '.m c h.Jcd.r_ evacuation, as a precautionary measure'even though projected dosesdo not indicate that such actions would be appropriate. 1911 296 gn ?%??h ~ aw -e - + -
~ DRAFT 14 Recommendation: The NRC, in coordination with the EPA Office of Radiation program, should formulate guidance for the implementation of protective actions in circumstances where the potential for exceeding action " guides warrants the ireplementation of the corresponding protective actions. In addition, the NRC should provide guidance to licensees regarding the need to have provisions for assessing the potential consequences under such circumstances and the need to provide state and local officials with the results of such an assessment. 12. Problem: A gaseous effluent monitoring system is needed for noble gases and iodines which (a) responds only to effluents being released via the pathway being measured, and (b) has.* -ange sufficient to quantify effluents resulting from the most serious design basis accident. (Ncble gas effluents from TMI could not be adequately munitored because of one or both of these considerations.) Recor.menda tio n: Revise NRC design criteria and guidance to specirically refer La gaseous effluent monitoring system whicn has the following capability (cr equivalent): the system should have redundant cetectors, one of which is positioned in the effluent stream being monitored and the other is positioned outside of the duct or conduit containing the effluent. The second (or background) detector, should: 1911 297 g Q s\\ W ayw=f%d %y y s,v wann W
DRAFT 15 a. be housed in material identified to that used for construction of the duct or conduit walls; b. be located rar enough away from (or shielded from) the effluent duct so direct radiation rrom material in the duct does not contribute significantly to the radiation it measures; be located close enough to the other detector so that the background c. radiation measured by both detectors will be essentially the same; and, d. have its output signal of such a form that it can be directly subtracted from the output of the other detector, i.e., the net signal would correspond to the concentration or radioactive material in the duct.
- 13.
Problem: \\/ At icw wind speeds (< 2 mph), frequent changes of wind direction can be experienced, resulting in an ill-defined or meandering ef fluent plume. Consequently, radiation levels for any given area may tend to undergo significant changes with time, making the assessment of radio-logical impact a somewhat difficult task. (TMI experienced this prcblem on numerous occasions following the March 28 event. ) Recarmendation: Licensees should be required to install exposure rate monitoring devices (which read out in the control room) at the centerline of each of 16 sectors corresponding to the principal compass points. These r, r T , ~ & h\\ 0 f},\\ ' ni \\ j) c 1911 298
DRAFT 15 monitors should be located on the site boundary (or at a greater distance if the maximum dose rate is expected to occur beyond the site boundary). The range of these monitors should be from 0.1 - 104 mR/hr, as suggested in a proposed ANSI Standard (ANSI N13/42 WG6, March 1978) entitled " Performance Specifications for Reactor Emergency Radiological Monitoring Instrumentation". 14. Problem: Although a loss of electrical power to the meteorological tower was not a problem associated with the TMI March 28 event, the impurtant:e of onsite wind direction and speed data to the individual directing the environmental survey teams can not be overemphasized. The meteorolog-ical data are also an integral part of the offsite dose calculation. The loss of such data (because the tower is not connected to the plant's vital power system) would likely have severely hindered the assessment of offsite impact. Recomenda t ion: The NRC should re examine its position and determine whether meteorolog-ical data is not sufficiently important to justify having the complete meteorological data system (feca tower to the control room recorder) connected to the vital power system. Ib. Preblem: Many nuclear power plants like TMI do not have a computing system which permits plant operators to perform real-time environmental radiation dose estimates based on real-time meteorological data. With -n $~ 1 ggga y a eu i911 29.9
~ DRAFT 17 regard to the noble gases released following the March 28 event at TMI, such dose computation capability would have been a valuable adjunct to the radiation exposure rate data being gathered by the survey team (or vice-versa) and to the environmental impact assessment performed retrospectively by readout of TLDs. Recc.m endation: Licensees should be required to develop and maintain real-time dose computation capability onsite which is based on a real-time data input from the meteorological tower and from the plant effluent monitoring system (including any effluent samples collected and analyzed). It should also accept inputs of field survey data (e.g., air samples, expression rate measurements) so that a comparison of predicted and ceasured valves can be made. Such computing capability could also be used to perform the routine estimates of radiological impact on man discussed in Regulatory Guide 1.21. 16. Arcblem: The TMI Environmental Technical Specifications require TLDs to be located at twenty sites, fifteen'of which are indicator stations and five are background stations. This number of sites may not permit a thorough estimate of doses following an accident such as the March 28 event at TMI a point Which was made by the Ad Hoc Population Dose Asses 5 ment Group:* "...there are only 20 dusimeters, so that three sectors (NE, ESE, W) have no measurements at all and seven (NNW, $$E, $$V, SW, WSW, WNW, NW) have only one..." 1911 300 g e Q1$$ t v,
DRAFT 18 Recommendation: The NRC should re-examine its guidance relative to the number of TLD locations required in routine radiological environmental monitoring programs, keeping in mind that these dosimeters w.11 also be used for accident assessment. 17. Problem: Following the March 28 incident, the THI nuclear plant gamma spectros-copy system was inoperable because of a high background radiation level due to noble gases. Such in-house counting capability would have very quickly informed the licensee that radioXenon - not radio ~ iodine - was the primary radioactivity on the charcoal cartridges collected of fsite. (This was confirmed at 1400 on March 28 after the State of Pennsylvania laboratory completed a Ge(Li) scan on the initial charcoal cartridge sent to them that morning by the licensee). Recommendation: Regulatory guidance should be amended to recommend that licensees maintain emergency counting capability (an Nal detector plus multi-channel analyzer might suffice) at a location outside or the racility (at THI, the Observation Center : night be suitable for this purpose). Once set up, it shnuld be required that the emergency counting capability be maintained in a state of readiness (nperable and calibrated) and that it be used perindically to ensure that individuals who would use the equipment during an emergency are familiar with its operation. 1911 301 VyfLgg pWT%- wax
DRAFT 19 l'. Problem: Ouring the period foilowing the March 28 event at THI when the initial offsite dose projection was made, the in plant monitor (HP-R-214) used as a basis for the dose projection was misread, resulting in a factor of 1000 overestimate in t.'e dose rate estimate for the nearest downwind community, The monitor used at THI for this dose projection has an eight-decade range from 10'I trom 10 mR/hr and can be read out in two 7 primary modes: (a) with the range selector in "all", the meter readout covers the entire eight decades or (b) with the range selector switch 2 3 7 en 10, 10 ... or 10, any of six 3-decade expanded scale readouts are available (the switen position indicates the top decade of the 3-decade scale). Reccmmenda tion: Regulatory guidance should be amended to discourage the use of such conitors. An eight-decade monitor having a single logarithmic scale appears adequate (can be read in a manner consistent with the accuracy reuuired in the calculation) and would not be as subject to misinter-pretation during an emergency. 19. Problem: The lack of pre printed forms on which to record radiation survey data (both hy the individual recording results of surveys received via radio and by the individual perfncming the surveys) likely resulted in a record of offsite radiation surveys which is incomplete. In general, the survey team name (or preferably, the individuals performing the survey) and the specific survey instrument and t.he manner in which 1911 302 %39gvG9% e-
~ DRAFT 20 it was used (e.g., open or closed window) were not recorded. The lack of a complete survey record hindered the concurrent as well as after-the-fact assessment of radiological impact. Recommendation: A survey form should be prepared and placed in the licensee's emergency kits. The forms should contain the following as a minimum: Date and time of the survey. a. b. l.ccation of survey (make refererne to pre-selected survey points, street intersections, mileage and direc. tion from known landmarks, etc.). Individual (s) performing the survey. c. d. Instrument used to perform the survey (include serial no. or other identifier) and the mode in which the instrument was used (e.g., window open or window closed), The duration of the meter reading (i.e. 20 mR/hr for 5 minutes). e. 20. Problem: Lack of quality control in vendor supplied radiological health and 1 ih' safety support. services. I The NRC has no way of assuring that the health physics support service provided to licensees can meet minimum standards of accuracy and p)li 303 <Nem cWS$ d \\ gww
DRAFT 21 precision. Se' vices such as instrument calibration, whole body counting, bicassay, and external dosimetry are provided to licensees without NRC ability to ensure their adequacy. Recommendation: Prepare regulatory guides in the areas of whole body counting bioassay, external radiation dosimetry, instrument Calibration, radiochemical analysis, environmental monitoring, etc. which set forth currently acceptable methods in these. areas and through tech. specs. require that licensees meet or exceed these methods. This will require licensees to select vendors who meet the requirements and will allow NRC to conduct inspections to assure that the minimum requirements are met. 21. Preblem: Technical Specification 6.2, " Organization" does not require the licensee to provide a radiation protection organization capable of implementing an adequate radiation safety prugram. It would be nak([tto assume that one " college trained" individual with two supervisors and twelve technicians could perform all the actions and evaluations required at a dual unit operating reactor facility in the areas of not only radiation prutection but chemistry as well. The Radiation Protection and Chemistry staff at TMI Included 9 college trained individuals, 25 technicians and assorted consultants. This entourage did not intact implement an dequate radiation protection program. 1911 504 l\\ [mnssu] V "i, jt 3 a n@ A q; u a - _ - ~ - ~.
DRAFT 22 Reccancndatiun: Technical Specification "6.2" should be reviewed and amended as necessary to insure: The radiation pre tection staff reports directly to the senior a. site management representative, b. A technicially qualified supervisor is maintained for each major discipline (i.e., dosimetry, environmental, in plant). An adequate number of qualified technicians are assigned to meet c. the demands of each discipline. 22. Problem: Technical Specification 6.3, " Unit Staff Qualif f eation" requires that each member of the Unit staff meet or exceed the minimum qualifica-tions of ANSI N-18.1-1971 for their particular jobs. ANSI N-18.1-lS71 requirements appear to have been met at TMI before the accident. The performance of the radiation protection staff clearly demonstrates the inadequacy of this standard to assure qualified personnel are available during all pnases of operation including emergencies. Reccmmendation: A Regulatory Guide that clearly defines minimum qualifications for each discreet job function should be pre. pared and implemented as a license condition. 1911 305 rerg y b, ha sj[jj d h ObSs +y7
DRAFT 23 These minimum qualifications must not be expressed in such nebulous terms as " college trained or college graduate". Qualification require-cents should be based on a combination of formal training, experience and actual demonstration of proficiency. The demonstration of proficiency should be performed by the licensee according to a program reviewed and approved by NRR. On a random basis I&! would then ask an individual to demonstrate his proficiency according to the approved program. - 23. DrobTem: Technical Specification 5.4 " Training" requires that a retraining and replacement training program for the unit staff be maintained that caets or exceeds the requirements and reenmmendations of Section 5.5. cf ANSI H-18.1-1971. Section 5.5 lacks specificity and is inadequate guidance to insure the qualification of the operating staff is maintained. Actiens of the nontechnician members of TMI radiation protection and chemistry staff taken prior to and during the accident clearly indicata an inadequate retraining program. Reco mendation: A Reculatory Guide that Clearly defines a acceptable training program for each jeb description at a power reactor facility should be prepared and implemented as a license conditinn. y)\\\\ 506 hb Ibkk
~ DPsAFT 24 24 Problem: Technical Specification 6.5 " Review and Audit" does not adr2quately define a specific audii. technique for nonsafety related areas such as radiation proLuction and training. The internal audit if properly regulated would dictate improvement in areas like radiation protecticn. Recumenda tion: Add a statement to this of the Technical specifications section requiring that audits be performed according to the QA plan. 25. Preblem: Technical Specification G.ll " Radiation Protection Program" is not sufficient to require a adequate program is developed. Especially for those licensees who, for instance, perform their own onsite dosimetry program. Recomendation: Prepare a Regulatory Guide which describes an adequate radiation protection program. This should include such items as instrument inventory and calibration, etc. 26. Problem: The licensee performed his own onsite TLD program. }C)11
- 07 D
a
~ DRAFT 25 Recommendation: The NRC or some other agency should without further delay require implemeatstion of regulations covering personnel dosimetry syst. ems. In either case, the NRC should implement a Technical Specification requirements describing minimum standard for those licensees who chose to perform their own dosimetry. 27. Problem: Individuals were exposed to very high concentrations of xenon-133, and xenon-135. In some instances the whole body exposure rates were reported in hundreds of R/hr. Wedonotknowhowwellgi'gvarious survey instruments respond to the low energy gamma radiation frem xenon-123. The teletector may have under-responded by more than 25%. Self reading pccket dosimeters may also have under-responded. o-TLD dat/ is not consistent with survey instrument readings. T1.Ds J-usually read less that would have been expected. Beta response on the TLDs is not consistent. Recommendation: The NRC should commission an experiment, t.o establish the response characteristics of survey instruments and dosimetry equipment to highly radioactive noble gas clouds. This experiment must be designed to resolve possible dosimetry errors due to geometwie considerations. Completion of this experiment is critical to correct assessment of the in plant personnel exposures. 1911 308 yhpa @g@um{@a g o
DRAFT 26 28. Preblem: The licensen had no acceptable method to insure planned dose accumula-tion in rapidly changing exposure fields was monitored. Recommendatinn: Every licensee should be required to provide and use suitable alarming dosimetrs under these types of conditions. 29. Problem: Aporopriate respiratory protective devices were not readily available during this emergency. Reccemendation: a. Emergency s:<ppifes of f adine remaining cant isters shou'ld be available at all facilities, b. The inventory of 30 minute self contained areathing apparatus supplles :ihould be ini;reased at, each faci!ity. c. Self contained breathing apparatus with a useful operation thte of more than two hours should be available. to' d. Respiratory protective devices #use in the control room should be w specifically modi fied to enhance communic ation. bb bO$d/ ], A 1911'309
DRAFT 27 Respiratorypro{ectivedevicesusedforsearchandrescueshouldbe specifically adppted for automatic radio communication. 30. Problem: flow do you monitor access to entrances to vital or camgerous areasin a general radiation amergency. Recomnendation: The exist.ing t.elevision munitoring syst,em could be expanded to observe critical portals or operations areas. A remote location orfsite could then be used to observe movement through these areas and possible deterioration of plant conditions such as fire, steam or water leaks. 31. Pr_oblem: Ouring a declared " General Emergency" should the limits of 10 CFR 10 and 10 CFR 20 apply? Reco=endation: The requirements of 10 CFR 19 should always apply. 10 CFR 20 should be revised to include an Emergency Action statement. 32. Problem: 10 CFR 20.l(c) "As Low As Is Reasonably Achievable" can not be enforced. inus eaumuuel enemm pMMM a 1 1911 310
DRAFT 28 Recomendation: It is critically important that tnis section of the regulation be made a useful enforcement tool. 33. Problem: 10 CFR 20.101 does not address annualYxposure limits to skin and -== extremitics. Recomendation: Revise 10 CFR 20.101 to include annual skin and extremity dose limits. 34. Problom: 10 CFR 20.401 does r.ot address retention of routinely performed whole body counting data. Recomenda tion: Revise and clarify 10 CFR 20.401 to include this data. 35. Problem: There is no clear regulatory requirement for methods, procedures and equipment for sampling airborne effluents containing radioactive caterials. The results of these samples are the bases for significant analyses and conclusions, but there are nonspecific regulatcry require-ments on tne sampling system, which as the foundation of the conclu - siens. ( un (M[js'Y!S @l f !IlblI D @W@((I) Jni., 1911,311
UNAH 29 It is clear that in the case of TMI that the licensee would have ins *iled stack sampling system that met ANSI N13.1-1969 if tnere had been a clear regulatory requirement. The applicable Regulatory Guide 1.21 states that stack sampling guidance in ANSI N13.1-1969 is generally acceptable, but this Regulatory Guide recomendat. ion is not a regula-tory requirement. Recommendation: KM ftSuiq? ion >
- = --
E2.be amended to establish strict specifications on stact sampling methods, equipcent and eqisipment calibration. There are well established and proven methods in use in nenradiological airborne effluent sampling that could be adopted for use in sampling radio-logical effluents, such as G0 CFR, Appendix A. In these regulations, the EPA clearly establisnes the acceptable system, but does leave provision for exception, if necessary, but only on approval of the EPA Administrator. Thus, the requirements are clear and realistic. e 1911. 312 @k@9m @@s@an$ = fi fR au
DRAFI 30 38. Problem: There is no specific regulatory requirement or guide for bioassays at Nuclear Power Plants. 10 CFR 20.108 only refers to bicassays required by licenses. Technical Specification requirements for bicassay are nonspecific. The current Regulatory Guide 8.9 for bicassays only refers to a series of ICRP report.s in a very general-way. Many of these ICRP reports are in conflict with one another. Recornmendation: A Regulatory Guide for bicassays at Nuclear Power Plants be issued. D] } Dq@
DRAFT 31 This guide should establish a program for bicassays based on air sample results, as Regulatory Guide 8.11 for uranium establishes. 39. Problem: There is no regulatory guidance concerning acceptable programs for air sampling at nuclear power plant.s. At present, 10 CFR 20.103 and 20.201(b) have only broad survey requirements, leading to a great deal of disagreement between licensees and the NRC as to what is acceptable. Recomendation: Develop a Regulatory Guide that defines more specifically an acceptable air sampling program. This guide snculd clearly state that air samples must be taken near the brea.'1g zones of workers and should require that lapel air samples at some frequency to confirm the adequacy of other air sampling systems. OSHA. in 29 CFR 1910, sets by regulation sampling methods, frequencies, analyses, etc. 40. Problem: Area radiation monitors in the reactor building provided little reliable indication of actual dnse rates. The containment dome monitor (ionization chamber) was too heavily shielded to measure the dose rate from xenon-133 and the GM tube monitors were offscale high or saturated. Recomendation: Recommended high range, non-saturating type instruments be placed in reactor building. Locations should be selected to provide as much as [h3$ $ $ hE=
DRAFT s e 32 process information as possible in addition to area dose rates. It kould be desireable to supplement ARM's with pan and zoom TV cameras capable of 1 coking at certain key components (press valves, RCDT RB sump, etc.). 41. Preblem: Effluent monitoring systems cannot differentiate between todines, particulates and noble gases in higt activity samples or with high background. This problem was icent'ried by the NRC's Lessons Learned Task Force. We endorse their recon endations and in addition request the following items be considered. Recemmenda tion: Encourage the production of systems with multichannel analyzer features and preferably automatic background correction for iodines and important particulates. This is certainly within the state or the art and should be economically feasible. Locate monitors (detectors) where they are least likely to be affected by hign gaseous background and direct radiation from systems containing highly' radioactive liquids or gases. (Example, auxiliary building roof might be a good place for gaseous effluent monitors.) 42. .Dechle :: Cbtaining in plant effluent samples was hazardous be.cause of high levels of radiation and airborne activity in the auxiliary building. !911 315 m- ,mmnc B
DRAFT 33 Recermendation: Location of the samplers and associated detecturs in a low background area (See 2) would solve this problem. It would also be feasible to install separate grab sample points in such pathways as the vent stack and certain ventilation systems. Such systems could consist of only a sa.'::ple probe (s), a sample holder and a pump. Provisions to take a gaseous or tritium sample should be included. To provide for represen-tative sampling piping r 'ns should be minimized. (Example: A particulate, iodine and noble gas syt,em could be installed directly on the side of the vent stack for use with a portable or permanent pump.) 43. Problem: Onsite counting room instruments could not be used because of high rartiation background (primarily noble gases) and very radioactive samples. Recoe.mendation: Identify alternate sites where counting facilitys may be setup. Have necessary regulated power, onvironment controls, etc., to operate r:ulLichannel analyzers and associated computer equipment. Have redundant counting equipment or easily portable equipment. Recommend that a very low range (pR/hr) insutrument, in the counting room to provide a direct assessment of background radiation levels. r u;e a w ueuu m l 1911 316
J 6 34 Accident sample geometries should be developed along with special sampling techniques to aid in the analysis of " hot" samples. Shie10ing should be designed to protect against direct radiation frem t other samples etc.. but would not be offeetive in a high noble gas background. 45. Problem: The plant chemistry procedure for boron analysis did not require encugh information be taken to allow a determination of whether or not sodium hydroxide hat! been injected into the RCS. i911 317 q q m <! M @ uur m rN s lr wa tam
a en OsAt-T as Recomendation: Have inspectors determine if normal procadures for it:portant analyses (boron, chloride, etc.) will be reliable if used under accident accidents. 46. Problem: In accidents similar to the THI-2 accident, it is pass. :a that high dose rates in the auxiliary and fuel handling buildings could make these areas too dange.rous to enter (Example: If ventilation system fafled). Beccc endatien: L]- ~ ( E trvaluate the consequences of not being able tc enter the auxiliary building. 19i1 318 g m cT 9 fM! O ~ p@M touwwubn m r
~ O.Nk ?_t. 48. Problem There is the potential for significant radiation protection problems in the turbine building for THI-2 type accidents. High dose rates fecm unshielded equipment and less efficient ventilation could impede placing the plant in a safe condition. Rece cendation: &tJMsYb Evtluate the secondary side radiological hazards for an accident which has both a large primary to secondary leak and extensive fuel dataage. Considerations are: a. Will the lack of shielding make any key equipment inaccessible? b.- Are equipment drains piped up such that leakage on the floor would be minimized? Do plants have the capability to pump the turbine building sump c. to the normal radwaste system? d. Are all gaseous and liquid discharge points monitored? 15 liquid radwaste capacity adequate to handle secondary side e. leakage? f. Is the condensate storage tank likely to become a radiation hazard at a source of radioactive effluents? ke W$$h[M w-
37 g. Is the turbine building ventilation system adequate? 49. Problem: 4 c. e, N % :es:iJc, &Pt S b ' Reach rods on some important valves were inoperable. The operation of at least one valve in the letdown system was interferred with by nearby angle iron. Recommendation: in Require that all potentially critical. valves and. reach rods Aany system .s e ca -ne. :.a. . _. e:. a..u: that might contain reactor coolant be M f 'j-maintained and designed for some abuse (over torqued etc.). Bypas.s valves around filters, demineralizers etc. in the makeup and purification system are examples of such valves. 50. Preblem: Exposure rate gradients were observed to be extreme; from a few R/hr to 1000 R/hr within a few feet. rer such gradients, the direction or the field relative to the surveyer be+5cg extiremely important. 8 Recommendation: Inspectors snould review plant training programs to ensure that informa-tion relative to potential accident radiation levels is included. Plant personnel (anyone whu might enter, not just HP's) should be trained to be aware of the expusure rate to their body as well as at the detector on the end of an extended probe. Extendable probes are required for the type dose rates experienced at TMI-2. ryM s w g bjdW4 1911,320
1 51. Problem: Some instruments and supplies stored in the Unit 2 auxiliary building became hignly contaminated due to high airborne activity levels. Recomendation: L...... sL,-Id 6 adme d ef 1 li: P
- al-
"' "Cl Inspectors should ensure that supplies and equipment which would be needed during an accident are stored in an "always clean" area. 52. Probig: Few % extremity dosimeters were used although plant personnel were handling some samples of > 1000 R/hr on contact. Extremity dosimeters comenly used at TMI-2 (taped to the wrist), W%> w~ not reliable indicators of the dose to the fingers from handling s::all highly radioactive samples. Recommendation: ,-e-a w e.s Inspection ecdule: should ensure that an extremity monitoring program is part of normal operatinns as well as emergency operations. Guidance on the recommended methods of using extremity dosimeters should be provided to the licensees. 53. Prnblem: ,..:,,, :.. ) ,-& %eadeoncEe P e c e ' $. ' '" S = 4 :5 * * ; Enosures due to long term skin contamination.werc-not -adeq"iM1 3 addrc; sed. r 2-1911 52) _, 6 x p... c eA
- n.,.
S mn M 'i m %g a ' ' ' ',) \\ 1 i. '.\\ 'W {V(li j p ' D #pg. c , g '[0 \\
DRAFT Recc=endation: t,u o,., s.w. A 4., 4-e 4, f 4.,. 9,,, y _, proc b. Inspection moedes should ensure that HP programs adequately address skin exposures. Instruments used to measure contamination should be calibrated for beta dose and gamma dose contribution to the contaminated a rea. Licensee should be required to have a formal procedure for evaluating and documenting all contamination exposures. 1911 322 f g e n m. .k' - gy - s m. \\g); g;,i(u t' ' ~
40 56. Problem: For the first two or three days of the accident, supplies and equipment were inadequate for sustaining ongoing environmental sampling and radiation protection programs. Recommendation: Establish emergency planning guidance that defines a minimum acceptable levelofmonitoringandprovidefarecommendedlistofsuppliesnecessary to accomplish the sampling (could be done as part of emergency plan approval precedures). Items like power inverters, samples pumps, charcoal and particulate filters need to be present in adequate a=uunts considering damage, failures, etc. It is also desireable to have portable systems for compressing air into a pressure bulb of established counting geometry for improving sensitivities. It is also desireable that iodine species samplers be available for environmenta'l samples. 1911 323 q E\\ T e m_,c M f' f h yj\\1 h n. y #e bJU \\. ' fs
DRAFT ~ ici ga,&e c, is cy p'2m should also ensure initial requirementG can be met for health protecting equipment, t.e., dosimeters, respiratory protection equipment (filterr,), anti *c clothing, high range instrumentation, etc. 57. Problem: The 1icensees abiiity to. refill the gelf contained breathing units was inadequate. k..t1 1, s ' M-b N " d M'N
- d I '* * * # " 3 -
,9 e A 9 p m - d u s h Emergency plant need to address the means by which SCBA's will be fflied. Cascade systems generaliy are too slow. Licensees snould use their own compressors or have arrangements with a Mcal Support organiza-tion (fire dept., other industry, etc.) to obtain immediate use of a compressor. The location of such equipment needs to be considered to avoid filling SCBA's with contaminated air. 58. Preblem: dd-g;4 9 g, Control room activity levels increased whenever auxiliary building and fuel handling building ventilation systems were secured. Reccmmendation: Evaluate the consequencas if the control ronm becomes uninhaoitable in a situation (similar to TMI-2) where continued operator actions were necessary to keep the plant, in a stable condition. 1911 324 m Q_,\\g e" A =\\ gy\\\\
D ~o',%A)CT n.rs [ 42 Consider making the auxiliary building and fuel handling building ventilation systems safety related for protection of the control room. 59. Preblem: $1 Specific plant data relative to sump capacities was not. in a readily useable form. 'l Capacities of lower elevations in the auxiliary building and reactor building were not accurately known nor the highest tolerable e (' level before critical instruments and pumps would become inoperable. B Recem.mendati o n: hl'g /f 7'"8*'c P " 6 /'" Lifuik s & mg% QCt @ O lf 3 in %C)wth$ Q V k o v1 IC S v..w w ou.;ies s,d 'c r chvat dur'.r.gg m .e. 5t h,v:a kce n s. ion-capact,ies-shotnebe tietermirmd~ ^ . i lo g tMg. All critical instruments and equipment should in ot r h d e r Q vt ". Q U A. Ita c 1 bL>:l V S ceTdentified and the water level of which they would become inoperable. E0. Problem: The time interval for testing charcoal adsorbers is tou long (18 months). Charcoal adsorbers may be rendered ineffective by conditions not readily recogni::ed by the licensee resulting in potentially long periods of time when adsorbers on standby or in service are actually inadequate. Recomendation: Decrease the periods between tests and require acceptance criteria to be cet priot to exceeding 1% power. ..g } } ' N b
- v %cQM,\\
x_;p ( y (W,' ~, 9%* Li v c.; "
3.,,.,.. ee -. a %,; i C 2.L 4.J. s L. 43 Develop a simple indicator test for charcoal adsorders which can be performed rapidly by onsite personnel. If sample results "Indteate" a problem then full scale testing should be performed. G1. Problem: Difficulties in maintaining a conder.ser vacuum because of an inadequate supply of auxiliary steam caused the plant.to have f.o use the main steam atmospheric reliefs more than a,2v frn:.<e d w e es t, l e. This could have been a serious problem if the secondary side had become contaminated by the primary to secondary leak. Recommendation: Evaluate the need for a more reliable source of auxiliary steam. 62. Preblem: There is the likelihood that, wat.er entered the waste gas system vent header in relatively large amount.s from the reactor coolant drain tank and the reactor coolant bleed holdup tanks. In addition to causing operational problems with the waste gas compressors, this is a potenti.11 liquid leakage pathway via drains on'the vent header, compressors and waste gae. decay tanks. Recommendation: Evaluate the feasibility of automatically isolating the vent header from certain components when licuid (or steam) flow is iminent. An example might be high pressure of high level in the RCOT. 1911 326 pppWmmM w;% ylj ~,' ~..
- :. : :. i: 3 o i J 'i n 44 Provide overflow points on all liquid radwaste tanks that will prevent liquid from reaching the vent lines.
Accident conditions (abnormally high pressures and liquid input rates) should be considered. 63. Preblem: A number of pathways for liquids and gases to leak into the auxiliary bu'Iding and fuel handling buildings existed which probably cb.itributed to high gaseous activity, particularly iodine activity, during the accident. Reccmmendation: Have licensees evaluate all pctential pathways for liquids including those from liquid drains on waste gas systems, steam traps, tank vents, etc. and ensure that. releases are discharged to a suitable collection point. Determine if the pathway exists before and after R.B. isolation. 64 Problem: Problems in the waste gas system which or developed during the accident such as leaks. loss of compressor seal water, etc. probably contributed substantially to Gaseous releases. Recc:.manda t i on: Develop technical spect rications applicable to leakage in th' waste gas system (or any gaseous leakage). 19il 327 m m; nwg yy -- ~ 'm 9WuG (. t 3
A 45 Include in preep testing a requirement to pressure test the system to its design limits to see if high pressure adversely affects any components such as compressors and to determine overall system leakage. 65. Problem: Ver.tilation riow paths existed between Unit 2 and Unit 1 causing higher than expected activity in Unit 1. It was also noted that Technical Specifications for the operability of the U-2 fuel handling building ventilation system only applied to times when spent fuel was in the SFP even though the Unit 1 SFP contained spent fuel and the two pools were not physically separated. Recemendation: Establish criteria for separation or units and in cases where systems are connected establish tech specs for both units for the most limiting condition in either unit. 66. Droblem: Not being able to describe the exact location of certain critical items of equipment (valves, instruments, etc.) caused unnecessary exposure. Cer tain ind viduals who were aksed to open or close or check certain i ydlves with whlCh they were not familiar, received unnecessary exposure because they had to " search" for the right valve. \\0\\ , M,, a N \\ g 3 pQ f D $5 '
46 Reccmmendation: Photograph all areas of the plant and in particular those critical components which might have to be operated by personnel who are not familiar with the equipment. Label all equipment and valves with large easily readable labels or nearby signs etc. 67. Problem: Extremely high dose rates in valvel alleys and around certain equipment made it impossible to observe the condition or position of the equipment. Recc.t.monda tion: Develop means of remotely viewing certain equipment. Permanently mounted CCT! or remotely controlled portable CCTV are suggested, 'ead glass viewing ports might also be useful in certain areas. 69. Preblem: Post LOCA equipment such as hydrogen reccobiners were too radioactive to work on without large exposures t.u individuals. Reccmmendation: Evaluate the need for more shielding on post LOCA equipment. Move instrumentation for the equipment away from the highly radioactive equipment to allow instrument adjustment without large exposures, 1911 329 off g R $1 M tuu e m A mea e
DRAFT 69. Problem: Certain individuals received doses while installing backup instru-mentstion for monitoring critical RCS parameters such as pressurizer pressure. Rec::mmendation: Identify key parameters (such as pressurizer pressure) and install simple reliable instrumentation in accessible areas. Design should include the provisions to flush the instrument and make a quick change if necessary. 70. Problem: Liquid radwaste capacity was not adequate to keep 1iquic frca overflowing into the lower auxiliary building elevation. Reccmmendation: Spect ry that liquid rad waste capacity be in. stalled and reserved for ea.lergency use. If the licensee has to use the reserve capacity, technic 11 specifications should require the restoration of the full capacity within some reasonably short time frame. Radwaste syste.ms shculd be re-evaluatsa for: --- C ~ .c_.- Equipment availability. a. 1911 330 A g g,y 4 f m@D \\; s ";,. w gv
~ 40 b. Peak demand versus average demand. Required tankage to meet peak demands with some emergency reserve. c. 71. Problem: Certain atmospheric radiction monitors indicated a leak in the waste gas system shortly after the March 28 trip. Proper evaluation of the monitor traces after previous trips mignt have alerted the licensee to look for leaks in the waste gas system. Reccmmendation: Require that licensee trip reports contain a brief evaluation of RMS responses for each trip. 72. Problem: A lack of reliable records of operation (not recorded on strip charts or by the plant computer) made evaluation of the causes and effects of the accident very difficult. Reccamendation: Have a recording system installed In the control rocm and on phone lines. The emergency organization snould provide for a recorder talker to put information on the tape. 73. Problem: Many significant alarms were " lost in the crowd" or not available at 1911 N1 f,,gacyw@3fM .s & ;;j u b
,e ',. 4? 3 all in a timely fashion (RB sump level). Recommendation: Update computers and printers to provide faster printout and make more use of CRT's for critical data. Recording strip chart information in digital form for ease of reproduc-tion is highly recommended. 74. Problem: There were not enough complete plant reference sets available to NRC personnel involved in evaluating radwaste problems and effluent prcblems. Licensee copies were in constant use by licensee. Reccanenda tion: Provide a complete FSAR and set of as built P& ids to each regio for inclusion in an emergency response package. The licensee should also provide regional offices with changes and modifications as they occur instead of annually. Changes and modifications would be marked up on prints by regional personnel. 1911 332 gp (p%N'hb 37 g, ..}}