ML19329A394
| ML19329A394 | |
| Person / Time | |
|---|---|
| Site: | Oconee  |
| Issue date: | 05/27/1977 |
| From: | Eisenhut D Office of Nuclear Reactor Regulation |
| To: | Goller K Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML19329A395 | List: |
| References | |
| NUDOCS 8001030965 | |
| Download: ML19329A394 (10) | |
Text
_
y e.
NAY f 7 377
..l 1
i j
M MDRANDtM POR: Karl 1. Coller, Assistant Director for Operating Reactors,
'l DDR l
PROM:
Darrell G. Eisenhut, Assistant Director for Operational t
Technology, DDR c;
SURJECT:
LIQUID EFFLUENT CONTROL MONIIORS Our review of a proposed amendment for the Oconee Nuclear Power Station has brought to our attention a possible generic problem involving liquid affluent control nonitors. The proposed Oconse = mand = ant would have allowed discharges at concentrations up to 35 times those specified in 10 CFR Part 20, Appendix B, Table II, for unrestricted areas. The basis i
l for the proposed==anAmant was that the hfth background at the liquid effluent control monitor and the inability of station personnel to cor-l relate the monitor's readings with affluent concentrations prevented the l_
monitor from responding to concentrations near the valves specified in l
10 CFR Part 20, Appendix B, Table II.
The licensee had stated that the monitor readings could not be correlated with the affluent concentrations because of the widely varying types of l
l effluent from the three units on the site. This problem has been discussed at length with A. Cibson and A. Kowalczuk of IfnE, Region II, and the li-consee. Because there was a difference of opinion between Region II and the licensee as to whether this problem could be so M d EEB/ DOR requested P.- Stoddart, ETSB/DSE (who has special e::pertise in t. La area) to visit the site. On March 16, 1977 P. Stoddart visited Oconee Station Units 1, 2 and 3 to discuss with station personnel the problems that they have had in correlating ths liquid effluent concentrations with the liquid effluent.
control monitor readings (trip report enclosed).
During the visit, P. Stoddart was informed that station personnel have recently di scovered defects in the NaI crystals in the liquid affluent l
l control monitor including crystal fracture, separation of the crystal-glass interface and discoloration of crystals. This has resulted in a decrease in gross count against standard radiation sources and a shift
{
in peak positions in the energy spectrum. These defects are believed to l
i contact:
J. Donohaw, EEB/ DOR 28066 g/y 8-
- orrscE>
772210244 sunNa mE >
y l
~
~ ~
m~
NRC PORM 318 (9 76) NRCN 0240
' W u. s. ooVERNMENT PRINTING oFFICEa 1,78 - 636434
-. -80010309Q. --.-
t
i.
j l
Karl R. Galler IIU f 71977 I
be caused by very rapid heating and cooling of the crystal during the dischsrae of liquid affluents. The crystal is in good thermal contact with the discharge pipe.
It is not insulated. The discharge of heated liquid radweste has raised the temperature of the crystal as much as 30*F in a few minutes which is an order of magnitude greater than the manufacturer's specification. Flushing the pipe immediately after dis-charge then lowers the N erature rapidly.
{'
Raising the temperature combined with the observed defects of the NaI l
crystal during discharge will reduce the sensitivity of the affluent moni-tot during discharge. This will affect the correlation between the effluent acuitor readings and the effluent concentrations, i.e., the monitor cali-bration. This reduced sensitivity of the monitor may allow discharges of concentrations in liquid effluents in excess of the concentrations listed in 10 CFR Part 20, Appendix B, Table II.
To be affective, the NaI crystal should be maintained at a near-constant temperatura during discharge and at a temperature which is compatible with the temperature at which the t
monitor was calibrated.
We recommend that this temperature problem for NaI crystals in affluent control monitors be investigated at other facilities. The method lican-sees;could use to determino if a NaI crystal is cracked is to corpare l
the responses of the monitor to a single solid source placed at five
[
different locations on the front face of the crystal. The source would be placed along the circumference at top center, left center, right center and bottom center, and at the center of the crystal (see attachments of the enclosed trip report). If the responses for the different locations around the circumference are significantly different or if the response for the center location is lower than those for the circumferential loca-tions, the NaI crystal is suspect.
We suggest issuance of an IE information circular with followup by IE inspectors as an appropriate way to proceed in this matter. We are pro-coeding with the review of the Oconee amendment as a separate matter.
Darrell C. Eisenhut, Assistant Director for Operational Technology Division of Operating Reactors DISTRIBUTION:
Enclosure:
Central Files J. Guibert As stated EEB Reading R. Cudlin h
D. Eisenhut ces See next page EEB/0 DOR EEB/0T/ DOR EhT/DO[
A%gR orric,,
JIIobhew:br LB Mp h g BVr%e7 DMt i
no N,us,
SNE/77 5/h/77 5/10/77 5/
87 on w NRC FORM HS (9-76) NRCM 0240 N u. s. sovanmusNT ra NT1No orrects sore-eauu v
Karl R.* Go11er '
i, i
I cet V. Stallo t
E. Denton i
R. Vollmer J. Colline l
A. Schwencer l
D. Neighbors l
L. Higginbotham P. Stoddart A. Gibson (Ragion II)
A. Kowalemuk (Region II)
- 3. Grimes L. Barrett E. Menean S. Block J. Donohaw I
l i'
s oPPIC E P eURNAMEM 4
DATE F
. NRC PORM 318 (9 76) NRCM 0240 W in e.novEnnuEnv eninvine orreces novo ese4a4
VISIT TO OCONEE NUCLEAR PLANT MARCH 16, 1977, BY P. G. ST0DDART
' EFFLUENT TREATMENT SYSTEMS BRANCH, DSE (TAR-6406)
On March 16, 1977, I met with members of the staff of Oconee Nuclear Station, Unit Hos.1, 2, and 3, and with representatives of Duke Power Company, as reauested by EEB under TAR-6406. The purpose of this meeting was to discuss effluent radiological monitoring problens associated with the radio-active liouid effluent discharge line.
I also sat-in on a meetino of a staff task force on radiation monitoring held on March 16, 1977.
Members of the Duke Power and Oconee plant staff in attendance at the i
morning meeting were:
Bob Koehler Mary Birch Mike Tuckman Bryan Burton Charles Putnam Ted McMeekin Attending the afternoon meetina were:
Bob Koehler Ted McMeekin Charles Putnam Bill McLean Mary Birch Jim Lona Bryan Burton Don Rogers Steve DeGanae The followina is a summary of the points discussed:
1.
Backaround of Linuid Waste Monitorino Problem The liouid radwaste system effluent from all three units are discharged to the tailrace of Keowee Hydro Station. The hydro station is used approximately 5% of the time; during those periods, the flow is on the order of several hundred thousand gallons per minute, which allows adequate dilution to dilute effluent concentrations on the order of
~4
-10
-II 10 uti/ml down to 10 to 10 uCi/ml. Approximately 20*:
b i
r
of releases can be timed to coincide with Keowee Hydro Stat'on operating periods. When the hydre station is not operating, leakage flow from valves, gates, etc., is on the order of 40 cfs, or approximately 20,000 ppm. During such periods, the only dilution flow is the 20,000 gpn leakage. 'With radwaste system discharges at 10 to 100 gpm, this represents a dilution flow of as much as 2,000:1 or as little uas 200:1.
If the radwaste systen discharne is 100 opn at 10'# uCi/ml, the effective dilution is 200:1, and the effective con-centration is 5 x 10~ uCi/ml, which, for a typical mix of nuclides, approximates the limits set forth in 10 CFR Part 20.
The Oconee staff found that radioactive contamination was building-up in the discharge pipe and in the detector well, causing an increase in radiation background count at the detector and making it difficult to set the monitor to alarm at 10 CFR Part 2011mits. On November 18, 1976, Oconee renuested that NRC approve a Technical Specification change which would allow the liouid effluent monitor to be set at a value which would be equivalent to 35 times 10 CFR Part 20 limits, based on a 40 cfs dilution flow.
As part of Oconee's justification for reouesting the Tech Spec change, they noted that a new lic0id monitor had been procured and installed.
It was noted that the new monitor had a replaceable inner chamber or liner which could be removed for decontamination.
4 C)
L
^1 3-2.
Incidental Radiation Monitorino Problems In Oconee's investication of problems associated with the liouid effluent monitor, a number of factors were observed which may well be of a generic nature rather than being specific to Oconee.
a.
'Plateout' At the present time, Oconee's licuid radwaste treat-ment consists of one stage of evaporation.
The resultant
~
~
condensate is typically about 10' to 10 uti/ml. The activity that is r, resent, hawever, apDears to be of an ionic nature and
.tends to deposit or plateout on the walls of the discharge pipe and on the walls of the licuid waste effluent radioactivity monitor chamber. The contamination builds up pradually over a period of time and is firmly fixed to whatever surface is present.
Station personnel have tried polishing the inner surfaces of the monitor chamber and also have experimented with lining materials such as Teflon; in each case, the contamination buildup rate is about the sane and the contaminent is firmly fixed to the surface. The new licuid effluent monitor noted above has a polished stainless steel chanber liner which can be removed; Oconee has two re-placement liners which permit exchange and decontamination with minimum downtime.
'The use of the replacement liners in the new licuid effluent monitor represents an interin solution to the problen of plateout with respect to the level of builduo of background radioactivity i
r 4_
8' l
and the correspondino <9 duction in sensitivity of the monitor.
l l0conee is in the process of re-piping their licuid radwaste treatment system to permit the use of a mixed-bed polishino It is deimineralizer downstream of the radwaste evaporator.
considered that this auament would remove the ionic conticni-nants with a high degree of efficiency and should substantially reduce the plateout problem.
b.
Scintillation Crystal Deoradation. Attempts by licensee staff to correlate analysis results with liouid effluent monitor readings showed variations and deviations which could not be, accounted for by expected statistical errors. Investigation into causes of the observed discrepancies resulted in the discovery of several cases of defects in the scintillation detector crystal assemblies, including crystal fracture, separation of crystal-glass interf aces, and discoloration of crystals due to internal hydration. The outward symptoms of the conditions described were a decrease in gross count against standard radiation sources and a shif t in energy peak positions in energy spectrum analysis. Although not confirmed by experi-mental data, it is believed that the defects are caused by thennal shock in -the liquid waste monitoring application. Under typical conditions, the monitor is at an ambient temperature of about Liquid effluent is discharged from the concensate tanks 70 F.
at 90 -100 F, following which a flush using lake water at i
k t
t f
f
. about 50 F is used to purge the discharge line. The Harshaw Chemical Company, manufacturer of the scintillation crystals used, recommends that the rate of change of temperature of the crystal not exceeded 1 C oer minute.
In reviewing the catalog literature of two liquid effluent monitor manufacturers using similar crystals, we note that the literature does not specify a limitation on temperature change, specifying only an operatino o
range of 0 C to 50 C (32 F to 120 F).
It is my opinion that the licensee's position that the crystal degradations are due to thermal shock is correct.
I recommend that a bulletin be issued to all operating plants describing the problem and asking each licensee to immediately inspect all detector crystals which are subject to such temperature changes and t report any occurrence of damaged crystals. A orocedure for rustine testing to identify damaged crystals without dis-assembly of the detector probe is shown in Attachment A to this memorandum. A routine testing procedure using a pulse height analysis device is shown in Attachment B.
Regular inspection o' detector ' assemblies should be required on f
a schedule to be determined on the basis of the number of defects encountered. On the basis of Oconee experience, cuarterly in-spection would seem warranted.
- l
=
~
On the basis of Oconee experience, it would aopear that the use of multi-channel analyzers in licuid waste monitors to A
initiate the closure of discharge valves is not appropriate.
small degradation in the detector crystal could result in shift of the spectral peak to a point outside the pre-deternined window, makinp such a monitor ineffective as a safety device.
Operation in a gross count mode would make detection of abnormal releases nore reliable.
Oconee has proposed a modification in detector probe design which would encase the crystal in a,iacket of plastic with good thermal Such a jacket should reduce the rate of insulating aualities.
chance of temperature to the crystal. This is only a potential modification and may not be practicable. Another possible solution is pre-heating of the sample stream before entering the detec-tor chamber; at low flow rates, as in the case of offline monitors this could be done electrically.
At this point in time, the problem has been identified and means have been developed to identify defective crystals; however, the probable cause of the problem remains and nothina has been done to mitigate the probien. The problem has widespread aeneric implications and should be resolved at the earliest possible date.
The Oconee Tenperature Sensitiviy of Liauid Effluent Monitors.
c.
staff also reoorted observation of a tercerature-dependent 4
M readout on.the new licuid radwaste monitor.
In one test in-volving measurements at only two temperatures, observed count rates at 87 F and 95 F for a single test source were as follows:
o 87 F 57,814 cpm 95 F 43,500 cpn The results indicate a decrease in count rate of about 1,800 com/ F, or about 5% per
- F.
This observation was reported to be cuite recent and no confirmatory work had been done as of March 16, 1977.
This observation has not been confirmed and may or may not be valid.
It is a point which should be resolved since it has generic significance in liauid radwaste monitoring.
r d.
Correlation of Varyino Inputs to Liouid Radwaste Discharae Monitor.
One difficulty that Oconee has had in calibrating of the liauid effluent monitor is variation of the average energy of wastes from the three Oconee plants. The applicant reports that the average eneroy in wastes varies from 0.2 Mev to about 2 Mev, depending on the source.
It is m,' opinion that while this may present difficulties in calibrating a monitor to read directly in terms of effluent i
i concentration, there should be no difficulty in preparing a
~
calibration curve relating energy to instrument response.
While setting the monitor alarm to respond at different meter settings for each batch of liauid discharged may be a nuisance,
.it should not be an insurmountable problem.
f
-