ML20058K907
| ML20058K907 | |
| Person / Time | |
|---|---|
| Site: | Monticello  |
| Issue date: | 12/28/1972 |
| From: | Mayer L NORTHERN STATES POWER CO. |
| To: | Anthony Giambusso US ATOMIC ENERGY COMMISSION (AEC) |
| References | |
| NUDOCS 9104260268 | |
| Download: ML20058K907 (6) | |
Text
[ Regulatory Fila Cy.7
~
- y i
NSF NORTHERN 5TATES POWER COMPANY l
winuseou s, unus e son es4ei i
i S
q
{
g December 26, 1972 g% 19 4,,
m Mr. A Giambusso reputy Director for Reactor Projects 4
Directorate of Licensing United States Atomic Energy Commission dY Washirgton, D C 20545 u. ta
Dear Mr. Giambusso:
UAc MONTICELLO NUCLEAR GENERATING PIAITI'
,,.2 1973 7 Il Dacket No. 50-263 License No. DPR-22 g
f
- EnD*t Report of Performance of the Reactor Codlant D.
LdM E
Icakage Detection System Ygw g
st a
'Ihis report is submitted in accordance with Section 6.6.E.3.h of Appendix A, Technical Specifications, of the Provisional Operating License for the Monticello N2 clear Generating Plant.
A review of the operation of the drywell leak detection and location sys-tem has been co=pleted based on experience gained over the past 18 months of plant operation. As will be discussed in detail, operation of the sys-tem has been completely satisfactory to date.
In an effort to attain the maxi:: ram speed and sensitivity of leak detection, however, an additioral instrumentation system has been developed and is undergoing technical evaluation.
Brief Description of the Present D17well Coolant Leak Detection System Drywell leakage is broadly classified as being from identified or unidenti-fied sources.
Identified leakage is piped from the recirculation pump seals, valve stem leak-offs, reactor vessel flange leak-off, bulkhead and bellows drains, and vent cooler drains to the drywell closed radwaste sump. All other leakage is collected in the drywell open radwaste su=p and classified as unidentified leakage.
Stes= from a reactor vessel crack would condense in the drywell and collect in the open radwaste sump.
Both the open and closed radwaste su=ps are connected by piping to two smaller cumps contain-ing a float switch operated pump and level alarms.
Both sump systems are r
of equal volume and nearly identical in construction.
I i
I 31 9104260268 721228 ~
NORT.
.RN CTATED POWER COPv ANY 4
Mr.IA Giambueso Ibcember 28, 1972 Each of the systems uses two timers to measure the interval between sump pump operation.
One timer actuates an alam if this interval decreases below an adjustable minimum.
The other timer actuates an alam if this interval cor-responds to slightly less than the technical specification limit. Actual leak rate determinations are made daily using flow integrators installed in the pump discharge line of each system.
In addition to the su=p leak measuring system the following additional mea-surements are available to give indication of gross coolant leakage:
a.
Drywell pressure b.
Drywell temperature Drywell iodine and particulate airborne activity (CIF/ CAM campler) c.
Evaluation of the Manticello Leak Ietection and location System To date, identified and unidentified leakage based on daily flow integrator readings has varied from 0.6 to 5 5 gp= and 0.2 to 3 2 gpm respectively.
These rates are a fraction of the Monticello L. hnical Specification limits of 5 gpm unidentified and 25 gpm total leakage. The ability of the system to measure long tem leakage trends with a high degree of accuracy (+ 3%)
has been well established. Performance of the system in detecting short ter= leakage trends using the adjustable timer alam has been less success-ful because of float switch repeatability errors.
Setting the adjustable timer too close to the current leak rate may cause spurious alams.
In gen-eral, however, short term increases of 0.8 gpm can be consistently detected.
Tae response time of the system is easily determined.
It can be shown that 186 gallons of water are pumped during a single cycle of the open or closed i
radvaste sump pumps.
Tbe maximum delay time to alam for leakage in excess of the technical specification limit is therefore:
Unidentified leakage 186 cal 37 2 min
=
5 Epm Identified leakage 186 gal 9 3 min
=
20 gpm Wnen the technical specifications on leakage are exceeded, 2h hours are per-mitted to place the reactor in a cold shutdown condition.
Tbe maximum delay times calculated are satisfactory based on this time scale.
i
4 NORT
!RN QTAT IO POWER Coh ANY Mr! A Giambusso De cembe r 28, 1972 Although we believe that experience has proven the drywell leakage detection system to be satisfactory, a number of alternate methods of detecting coolant leakage nave been undergoing evaluation.
Techniques Considered for Improving Speed and Accuracy of LYywell Coolant Leak Detection i
Acoustic bbnitoring System An acoustic monitoring system has recently been installed on Prairie Island Unit yl. Using specially designed transducers mounted on the reactor vessel, coolant pumps, and primry piping the system can potentially accomplish the following:
Continuously monitor reactor vessel flaws and their growth a.
b.
Predict catastrophic failure c.
Ietect sml1 non-crack related leaks d.
Use triangulation techniques to determine flaw location in system The system was placed into operation for the initial primary cold hydrostatic test and is currently undergoing evaluation. Acoustic monitoring offers the unique advantage of directly monitoring reactor vessel integrity and pinpoint-ing flaws and cracks.
All syste=s using acoustic emissicn techniques are still in the development and testing stage.
It would be premature to consider such a system for the tbnticello plant until the Prairie Island system has proven its value.
Detection of Radionuclides in Drywell Atmosphere Use of the radioactive fission products in the coolant as tracers to indicate leak rate and source has often been proposed.
Kr-88, Xe-135, and I-135 have seemed the most likely radionuclides to use for this purpose.
Experience at other plants indice.tes that the low levels involved raise the following dif-ficulties:
Several hours are required to take sa=ples, count them, and a.
perform calculations b.
Iarge errors in calculations result due to uncertainties in accumtely knowing isotope concentrations in various coolant systems
x y
NORT RN STA*T'IS POWER CON ANY Mr. A Giambusco t-Iecember 28, 1972 Techniques usirg multiple sample points have not 1 een succe. sful because of the drywell atmosphere mixing resulting f rom cooler fan cperaticn.
Atte= pts at Monticello to correlate drywell airborne lodine and particulate activity with leak rate hue also been made without success.
Use of bidtare Sensing Tape and Similar Materials on tLe Beactor Vessel Use of moisture sensirg =aterials has also been proposed as a method of detecting and locatira c=all pressure vessel leaks.
No complete 1; satis-factory materia) has yet been found and further evaluation is in progress c
at other plants. We vill continue to follow these tests and consider the technique if it proves to be practical.
Use of Improved Dryvell Sump Instrumentation There are several possibilities for using improved instrumentation to im creabe the speed and sensitivity of leak dstection.
They include:
a.
Installation of flow meters at points in the drainage ;ath d
leadirg to the surps b.
Installation of su=p level transmitters and analog differen-tiators Installation of su=p level transmitters with input to the plant c.
process computer for leax mte calculation Due to the low leakage flows normally experienced, the use if flow meters is not practical. A typical unidentified leak mte is 0 5 gpm.
Construction of special leak colloction vessels with small metering lines and overflow capability to the sump would be required.
A 0 5 gp leak rate into the drywell floo: surp would result in a level charge rate of:
dL/dt 0.0175in/ min 0 5 gpm
=
=
20.6 gal /in A typical level transmitter has an output of 10 to 50 ma over the rance of the instru 1;.t.
'lhe signal charga 'or a 0 5 cpm leak rate into the cump is therefore:
di/dt 0.0175 in/ min
.c 40 ma
=
60 sec/ min 6.5 in 0.00180
=a
=
see I
NORT..ERN CTATEO POWER Coh ANY Mr. A Giambusso fecember 28, 1972 This small rate of change can be dealt with fairly easily using digital pro-l eessing techniques, but introducec some difficulty for analog equip:nent.
Evaluation of Alternate Dryvell Leakage Detection Techniques Leak Iocalization There appears to be no adequate method of in-service drywell leakage locali-zation and direct pressure vessel monitoring availe.ble at thic time. Acoustic monitoring systems such as the Prairie Icland installation appear to be cuper-ior to others under development.
The Praicie Island acoustic monitorine cystem vill be thoroughly evaluated for applicability to the Enticello piant if it meetc its design objectives.
2 ak Detection Technical review is now in progrecs of a proposal to inr t,all level trancmit-ters in the open and ' osed dryvell rsdvaste sun.ps wit
- outputs to a control room level recorder and the plant process computer.
This system vill te independent of and in addition to all other installed equipment.
Dryvell identified and unidentified leak ratec would be calculated periodically by the compr er and prin+. d out in the control room.
Icak rates in excess of technical rpecificaticn limits vould print out as alarms. The major com-ponents c" the proposed system vill be:
Displacement type level transmitter mounted in a stilli:g well a.
from the top of each dryvell sump b.
Pecordere (with integral trancmitter power supply if nececcary) in the control room to continuoucly monitor cump levels Software written for plant process computer to sample level c.
transmitter outputs, calculate leak rate, and generate alarms The proposed. cystem vill provide the following functions:
i on unidentified leakage in excess of the techni 1 cpeci-a.
Alan:1 fication after a delay of only 1 minute b.
Alarm on total leakage in excess of the technical specification after a delay of only 1 minute Provide graphic control room indication of drywell leakage trends c.
9 NORT.
.R N CTATED POWER CON ANY Mr.' A Giambucco Ibcerr.ber 28,197a If the results of our evaluation indicate that the additional inctruments will lead to significant improvement in the speed and sensitivity of cool-ant leakage measuremente, they will be installed.
Yours very truly, Mr.0+
r L 0 Fhyer, P, E.
Director of Ruclear Support Services LOM/D?o'/br ec:
B H Grier