ML19207B369
| ML19207B369 | |
| Person / Time | |
|---|---|
| Site: | Dresden, Quad Cities  |
| Issue date: | 07/30/1979 |
| From: | Reed C COMMONWEALTH EDISON CO. |
| To: | James Keppler NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| References | |
| IEB-79-12, NUDOCS 7908240513 | |
| Download: ML19207B369 (9) | |
Text
"
Commonwealth Edison One First National Plaza. Chicago, liknois
,/
b Address Reply to: Post Office box 767 Chicago. Illinois 60690 July 30, 1979 Mr. James G. Keppler, Director Directorate of Inspection and Enforcement - Region III U.S. Nuclear Regulatory Commission 799 Roosevelt Road Glen Ellyn, Illinois 60137
Subject:
Dresden Station Units 1, 2 and 3 Quad-Cities Station Units 1 and 2 Response to IE Bulletin 79-12 "Short Period Scrams at BWR Facilities" NRC Docket Nos. 50-10/237/249/254/265 Reference ( a) :
J. G.
Keppler letter to B.
Lee, Jr.
dated May 31, 1979
Dear Mr. Keppler:
Reference (a) transmitted IE Bulletin 79-12 which requested a response within 60 days to items of concern addressed in the Bulletin.
Attachments 1 and 2 to this letter provide our response for Dresden Station Units 1, 2 and 3 and Quad-Cities Station Units 1 and 2, respectively.
Please address any additional questions you may have concerning this matter to this office.
Very truly yours, Cordell Reed Assistant Vice-President
,a ygo
.)
cc:
Director, Division of Reactor Operations Inspection "790;82,40s3
Commonwealth Edison NRC Docket Nos. 50-10/237/249 ATTACHMENT 1 DRESDEN STATION As noted in IE Circular 77-07, Dresden Unit 2 experienced a short period scram in December, 1976.
In order to prevent recurrence, all procedures involving reactor criticality, control rod withdrawal sequences, and the operator training program were modified during 1977 to:
1.
Provide precautionary notes to operators; 2.
Increase operator awareness of potential short period c0u3itions; 3.
Specify operator action in the event of a short period; 4.
Restrict use of notch override during approach to critical (particularly during high xenon conditions) ;
and 5.
Minimize control rod notch worth while adhering to Rod Drop Accident sequencing constraints and prudent operational practices.
Since implementation of these changes, Dresden has experienced criticali;y over a wide range of reactivity con-ditions, and we have had no difficulties with shorm periods or excessive notch worths.
We feel that our current operating practice meets the intent of IE Bulletin 79-12 and ninimizes the probability of short reactor periods.
However, we will implement more restrictive sequences (response to Ite,n 3a) to further lessen the likelihood of short periods.
Responses to the individual items listed in the Bulletin are provided below.
Item 1 - Units 1, 2 and 3 The point at which criticality occurs in a BWR is a function of many variables, including moderator tamperature, core xenon concentration and distribution, core exposure and exposure distribution, and control rod density.
Most of these variables are strong functions of time, indicating that gross core 7*9
_no
Commonwealth Edison NRC Docket Nos. 50-10/237/245 reactivity during an approach to criticality is itself a time-dependent variable.
Experience has shown that even critical rod patterns generated by rophisticated state-of-the-art d imensional BWR simulation codes cannot be relied on, particularly during changing xenon conditions where the exact time of criticality must also be predicted.
A simple equipment-induced delay in startup would require complete re-analysis of the core's xenon reactivity.
Prediction of the point of criticality, accurate or inaccurate, will not in itself prevent short periods.
In fact, reliance on an inaccurate prediction may reduce an operator's attentiveness prior to the point of predicted criticality, thereby enhancing the possibility of short-period scrams.
Although an accurate prediction may alert an operator to the time at which he should be more careful, we believe the more prudent approach is to restrict reactivity inserations and increase operator awareness during the entire rod withdrawal sequence.
Dresden's implementation of this concept in lieu of criticality predictions is further described under Items 2 through 5 below.
Item 2 A.
Units 2 and 3 Our operating procedures currently provide detailed cautions regarding the potential for high notch worths during high xenon conditions.
Use of the notch override switch is prohibited from 50% control rod density through heatup (thus assuring notch-step withdrawal) during high xenon criticals and cautioned against during all approaches to criticality (see response to 3a for further notch withdrawal restrictions which will be implemented at Dresden).
Our startup procedures also require that prior to control rod withdrawal, all operable SRMs and IRMs must be fully inserted with the IRMs on the most sensitive range.
In addition, the two highest SRMs and an IRM from each scram channel must be displayed on chart recorders operating on high speed.
To provide independent verification of these requirements, a " Nuclear Engineer's Pre-Startup Surveillance" is procedurally performed to check the status of startup instrumentation prior to commencing control rod withdrawal.
rq L
.s
_u i
e
Commonwealth Edison NRC Docket Nos. 50-10/237/249 B.
Unit 1 On the initial critical of each operating cycle, each control rod shall be notched out between positions 0 and 4.
On subsequent startups, the unit operator will b3 required to begin notch withdrawal (between positions O and 4), three rods preceding the most recent xenon-free critical number of rods.
Since Unit 1 fuel does not use any burnable poison rods, each critical requires additional control rod withdrawal as exposure is added.
The remaining control rods will then be notched (positions 0 - 4) from that designated step until the units is at heating power.
This method will assure that every rod is notched through its high worth region when high worths are possible.
Operating procedures will require that the nuclear engineer determine the step to begin the control rod notching before the first rod is pulled and instruct the nuclear engineer how to determine this based on previous criticals.
Procedures also require that the SRMs be inserted and are reading greater than 100 CPM.
Item 3a A.
Units 2 and 3 The most significant deterrent of short periods during high xenon startups is use of the " infinite lattice" sequencing technique.
In this technique, peripheral rods are banked out along with their internal counter-parts while complying with Rod Drop Accident sequencing constraints rather than Drmediately after attaining 50%
rod density.
This serves to reduce the notch worths of internal arrays which previosuly were the only control rods controlling 3.arge sections of the core near the periphery.
Since implementation of this technique, no short periods during high xenon startups have been observed.
-, n s O
Commonwealth Edi:;on NRC Docket Nos. 50-10/237/249 4-Prior to 50% rod density, incremental reactivity insertions are normally small, provided the use of notch override is restricted.
At Dresden, Rod Drop Accident constraints require banked withdrawal of groups 3 and 4 to positions 04, 08, and 12 Further restrictions to minimize short periods will be added to Dresden sequences by requiring that group 3 and 4 reds be notch-step withdrawn between positions 04 and 12.
This essentially precludes continuous withdrawal until the rod is beyond position 12 where notch worths are substantially reduced.
B.
Unit 1 Unit 1 sequences have been developed using the infinite lattice array concept.
Each operating cycle, the core management organization for Dresden 1 will review the sequence to verify the notch worths are as low as practicable and also verify RDA compliance.
Therefore, the use of these sequences, as well as the rod notching, will minimize the possibility of any short periods.
Item 3b - Units 1, 2 and 3 At Dresden, each operator withdrawing control rods for criticality must read and initial a full page of precautions and related operator actions addressing potential high notch worth conditions.
In addition, a "Contro.1 Rod maneuver Request Form" is completed by a qualified nuclear engineer and approved by a licensed shift supervisor prior to withdrawing control rods for criticality.
This form describes the maneuver to be performed (in this case reactor criticality and heatup) and provides specific cautions on potential or anticipated problems (i. 4e,
potential Sigh notch worth and short periods) associated with the described maneuver.
We feel that this has been successful in increasing operator awareness and in assuring prudent operator actions during and subsequent to criticality.
Item 4 - Units 2 and 3 Dresden has had excellent performance to date of the
" emergency rod in" switch.
Since the switch is more often used to aid in settling a control rod in its desired notch position (rather than to mitigate a short period), it is highly unlikely that the initial malfuntion of this switch would occur no V
J
Commonwealth Edison NRC Docket Nos. 50-10/237/249 simultaneously with the unlikely occurrence of a short period.
(This item is not applicable to Unit 1, as it does not have an
" emergency rod in" switch).
Item 5 - Units 1, 2 and 3 Following the earlier short period experienced on Dresden Unit 2, all licensed operators received a special training course on short periods,why they have occurred, and what actions to follow if they do occur.
All first-time license trainees receive the same training,and this subject has also been incorporated into our normal retraining program.
This program and our procedural cautions have served to increase operator understanding and awareness of potential short period conditions.
We believe that this enhanced operator awareness, coupled with the restrictions on reactivity insertions described above, provide the most effective means of preventing the occurrence of short reactor periods.
I
.- g qqQ
_ - ~
" w 6-
=%,e em -
-.e-.
~
Commonwealth Edison NRC Docket Nos. 50-254/265 ATTACHMENT 2 QUAD-CITIES STATION
Background
IE Circular No. 77-07, concerning short periods during reactor startup, was issued April 14, 1977.
At the time the circular was received at the station, Quad-Cities had not experienced a high notch worth induced period of less than 30 seconds during startup since January, 1973.
The information in the circular was considered significant, however, and all the indicated recommended actions were Dmplemented at Quad-Cities.
Subsequent to the circular, the station implemented the Banked Position Withdrawal Sequence (BPW3) method of control rod sequencing.
The BPNS sequencing constraints, however, were designed to assure compliance with Rod Drop Accident assumptions and did not specifically address minimizing notch worths for the purpose of preventing fast period scrams.
Subsequently, an IRM hi-hi scram was experienced on Unit 2 on October 26, 1977 during the withdrawal of two notches in a low control density peripheral region.
The specific condit. ions present at the time of the scram indicated the need to modify the sequencing techniques employed for high xenon, hot reactor moderator temperature, and fast scram recovery criticals.
After a cceprehensive study of BWR experience on short periods, the appropriate further corrective action was deten ined to include a revision in the treatment of withdrawal of peripheral control rods.
Quad-Cities implemented the technique of banking peripheral control rods with their corresponding interior rod groups in what we refer to as an infinite lattice approach to peripheral rod withdrawal.
This approach minimizes notch worths of the type which resulted in short periods under fast scram recovery conditions.
The notch worth minimization approach was deemed the most effective way of dealing with scrams caused by fast periods.
In addition to changing sequencing techniques, an additional restriction on the continuous withdrawal of contro] rods was bnplemented.
This additional restriction disallows use of the notch override control switch on startups from the time half the control rods are fully withdrawn (50% density) until heatup is completed.
Since November, 1977, with the implementation of the above corrective actions, Quad-Cities has not experienced a scram from a reactor short period resulting frca a high notch worth.
The approach taken to prevent short periods (i.e., combining notch worth minimization with restrictionn on continuous rod withdrawal) has been effective.
7o
'q
, o au T
Commonwealth Edison NRC Docket Nos. 50-254/265 i.
Current Action to be Taken After receiving and reviewing IE Bulletin No. 79-12, Quad-Cities has concluded that additional protection from excessive continuous rod withdrawals should be implemented at the station.
The area of concern is withdrawal of rods from position 04 to 12 in rod groups 3 and 4.
Continuous rod withdrawal protection is already provided for by previous corrective action bmplementation once 50% rod density is reached.
This additional action will provide added protection by requiring notch withdrawal of rods from position 04 to 12 in rod groups 3 and 4, the last two rod groups withdrawn prior to reaching 50% control rod density.
This additional action, coupled with notch worth minimization techniques used in control rod sequencing at Quad-cities, should prevent scrams from fast period during reactor startups.
Response to Specific Items Item 1 Scrams during startups are an operational problem which we are extremely interested in preventing.
The station has carefully considered the most effective method of preventing scrams resulting from fast periods.
The station's conclusion is that an estimate of the critical rod pattern is not an effective i
method foi preventing these scrams due to the inaccuracy involved in any existing method for critical rod pattern prediction.
In fact, reliance on an inaccurate prediction may reduce an operator's attentiveness prior to the point of predicted criticality, thereby enhancing the possibility of short-period scrams.
The inability to prevent short-period scrams by obtaining a critical prediction is well documented Ly utility experience.
Quad-Cities, therefore, feels any commitment to provide a critical rod pattern prediction with the intent of preventing short periods would be an ineffective and, therefore, inappropriate response to the problem.
Item 2 (For response, see Background and Current Action to be Taken Sections.)
SRM, as well as IRM, monitoring is already an existing part of the station's startup procedures.
r e
j
~ n '. t
~
Commonwealth Edison NRC Docket Nos. 50-254/265,
Item 3A The sequences prepared for use at Quad-Cities are reviewed,in part, specifically for xenon-override criticality conditions.
Notch worths are minimized in this region of the sequence by implementation of BPWS, an infinite lattice technique for peripheral rod withdrawal and requiring notch withdrawal of rods from 50% rod density until heating is completed.
Item 3B Operator training and retraining includes general discussion of rod and notch worths.
A specific caution on notch worths for excess xenon conditions is already provided in the station's startup procedures.
Item 4 The emergency rod in switches at Quad-Cities have historically performed very well.
The station has concluded that should a problem arise with the switches, it can be adequately detected from the normal use t'e switches receive.
Should maintenance be required on the switches, it can be accomplished via normal work request procedures.
Item 5 Operator training, including discussion of general notch worths, differential and integral rod worths, excess xenon effects on rod worth, and the operator's role in core monitoring during approach to critical, was performed as part of previous corrective action to short periods experienced at Quad-Cities.
Corrective training and retraining continues to emphasize discussions on rod worth and operator awareness during approach to critical.. This bulletin will also be included as part of operator retraining.
., e e,
q q 'i
_.