ML20214U013
ML20214U013 | |
Person / Time | |
---|---|
Site: | Fort Saint Vrain ![]() |
Issue date: | 09/26/1986 |
From: | Office of Nuclear Reactor Regulation |
To: | |
Shared Package | |
ML20214T997 | List: |
References | |
TAC-61491, NUDOCS 8609300470 | |
Download: ML20214U013 (8) | |
See also: IR 05000267/1986017
Text
_ _ _ - __
_
e
[h#"'*4"I
,
n
UNITED STATES
NUCLEAR REGULATORY COMMISSION
WASH;NGTON, D. C. 20555
.....
SAFETY EVALUATION (PRELIMINARY) BY THE OFFICE OF NUCLEAR REACTOR REGULATION
RELATING TO LER NO. 50-267/86-017
>
PUBLIC SERVICE COMPA'NY OF COLORADO
FORT ST. VRAIN NUCLEAR GENERATING STATION
DOCKET NO. 50-267
1.0 BACKGROUND
On April 3, 1986, with the reactor at 25% power Fort St. Vrain (FSV)
experienced a period of grid fluctuations during a storm which resulted in
several short periods of loss of forced circulation. While an actual
loss of off-site power did not occur during the event, off-site voltage
transients were transmitted into, and through the plant's electrical
distribution system. This initially caused the trip of an undervoltage
protection relay in the helium circulator bearing water pump system,
resulting in circulator trips and loss of the Loop-2 forced circulation.
The electrical transients also caused the isolation of the helium
purification system. It was this latter failure that led the operating
staff to manually scram the reactor. Further bearing water trips,
apparently due to undervoltage relay operation, later resulted in four
brief periods of complete loss of forced circulation. In all cases, the
'
plant operating staff promptly restored the circulators to operation.
The maximum core heatup during the loss of forced circulation was
estimated to be 19.6 F. No fuel damage occurred, however, a small
release occurred due to leakage past a circulator shaft shutdown seal.
These events were reported to the NRC in Licensee Event Report (LER)
50-267/86-017 submitted by letter dated May 3, 1986 (P-86349).
2.0 EVALUATION
i
As part of their project, " Analysis and Evaluation of Operational
'
Experience from Fort St. Vrain," Oak Ridge National Laboratory (ORNL)
provided a brief evaluation (copy attached) of the event described in the
LER identified above and the current status of the licensee's followup
efforts. The ORNL evaluation provided the event scenario, details of the
protection systems which tripped due to electrical transients, and a
review of similar past occurrences. The evaluation indicated some doubt
as to whether the FSAR bearing water protection logic had indeed been
present in the previous events. The evaluation also noted that backup
l bearing water was not available for this incident, since it is not
required by plant procedures at low reactor power levels.
The ORNL evaluation did provide information on some preliminary
' corrective actions by the licensee, which had confirmed that the pump
undervoltage relays were already powered by uninterruptable DC buses,
i
'
W2nuME7W
S
.
s -2-
These DC buses were not affected by the grid transient. However,
the AC undervoltage sensing coil was being powered by the off-site grid,
and the preliminary evaluation by the licensee indicated that the low-
voltage setpoint was set too high. The ORNL evaluation indicates that
these equipment protection trips occurred, eve'n though the off-site power
transients were never severe enough to cause an automatic start of on-
site emergency power. The evaluation also suggests that in a situation
with such severe grid instability, electrical isolation of the plant and
reversion to the plant's emergency AC power system may have been prudent.
ORNL points out the lack of any specific guidance to the plant operating
personnel on when to isolate from the grid during adverse transient
conditions.
ORNL has also indicated that a licensee response to this LER remains
open. Currently the licensee plans that corrective actions will be
closed out in the closeout package assembled for the review of the NRC
Senior Resident Inspector. There appear to be no plans by the licensee
1.0 provide any additional followup to the staff on the event analysis or
corrective actions.
3.0 STAFF ASSESSMENT
The staff's review of the ORNL evaluation and the LER clearly indicates
that no significant- risk to the public existed during this event.
However, this event also indicates the relative sensitivity of Fort St.
Vrain's helium circulation system to off-site grid upsets. While no
probabilistic risk assessment is available for this plant, studies
completed for LWR's indicate that loss of off-site power transients
contribute significantly to plant risk. The response of a high
temperature gas cooled reactor to loss of off-site power would, of
course, be somewhat different than light water reactors. However, even
with a fundamentally more forgiving design, the potential seriousness of
the complete loss of forced circulation cooling cannot be overlooked.
These concerns must be coupled with the fact that the licensee had
apparently not considered grid transients such as occurred on April 3,
and consequcntly, had no operator guidance in place to minimize its
impact on plant systems. ~
4.0 STAFF CONCLUSIONS AND RECOMMENDATION
Review of the ORNL evaluation leads to the following preliminary
conclusions:
1. The event did not of itself pose a significant threat to the
public.
2. The event did, however, suggest the sensitivity of various
plant systems to off-site AC power grid instabilities, most
notably the potential for complete loss of forced circulation
cooling for grid transients which did not cause electrical
- isolation of the plant, and initiation of emergency AC power. ;
. _ _ _ _ . _ _ _ _ _ _ ._. _ . _
_
'
.
'
-3-
3. The availability of backup bearing water at lower reactor power
levels should be considered.
4. The plant operational staff responded well to the event, but
apparently did not have adequate guidance for considering the
possibility of manual isolation from the grid and initiation of
emergency AC power sources.
5. The licensee had not yet completed their response to the event,
and has no plans for any formal submittal to the staff on
planned corrective actions.
Based upon the above factors, we conclude that the April 3 ever.t raises
sufficient concerns regarding Fort St. Vrain, and that a more formal
review of the licensee closeout should be conducted. Therefore, we
recommend that the licensee provide details of their evaluation and
response to this event, and that this material be reviewed by the Plant,
Electrical, Instrumentation, and Control Systems Branch, in the Division
of PWR Licensing-B.
Principal Contributor: M. P. Rubin, DPWRL-B/F0B
Date: September 26, 1986
Attachment: Analysis and Evaluation of
Operational Experience from Fort St. Vrain
- - . . . -_ -- -- - __- -_- _ _ _ ~ - -- _-
.
.
~
, ,
'
.
' Attachment
j
.
PROJECT: Analysis and Evaluation of Operational
Experience from Fort St. Vrain
CONTRACTOR: Oak Ridge National Laboratory BUDGET PERIOD: May 1986
'
CONTRACT MONITOR: Frederick J. Hebdon BUDGET AMOUNT: 44 K (FY-86) i
i!
PERCENT EXPENDED: 62
-TECHNICAL MONITOR: Peter S. Lam
J. R. Buchanan PHONE: FTS 624-0393 or
CONTRACTOR PROGRAM MANAGER:
(615) 574-0393
David L. Moses PHONE: FTS 624-6103 or
! PRINCIPAL INVESTIGATOR:
(615) 574-6103
PROGRAM OBJECTIVES:
l/
!1 The NRC Office for Analysis and Evaluation of Operational Data (AE00)
has established an extensive program for screening, analyzing, and
ll:i 4}( evaluating the operational experience data, particularly the data
reported in Licensee Event Reports (LERs), and other information from
' ' .) all connercial nuclear power plants in the United States. Because of
the unique design of the Fort St. Vrain.HTGR, the technical expertise at
- i
the Oak Ridge National Laboratory (ORNL) is being used to fulfill the
required AE0D staff engineer functions for following and assessing the
operational experience developed at Fort St. Vrain.
,
.
'
ACTIVITIES DURING REPORTING PERIOD:
- 1. LER Screening
!
l
During the current reporting period, two LERs covering 2 events have
i been received and screened. One LER (R0 86-016) has been classified The
as reporting a Category 4 event requiring no further action.
l
other LER (R0 86-017) has been classified as reporting a Category 3
I event requiring more information to make a final classification and
i
recommendation for disposition. R0 86-016 reported' a missed com-
'
.
pletion date for the hydraulic power system functional test due to'
personnel error in scheduling delayed portions of the quarterly
,
surveillance. R0 86-017 is discussed in more detail as follows.
,
The LER for R0 86-017 describes in reasonable detail the series of
incidents which cccurred during the period 0900 hours0.0104 days <br />0.25 hours <br />0.00149 weeks <br />3.4245e-4 months <br />.-(MST).. During this
April 3, 1986, and 2030 hours0.0235 days <br />0.564 hours <br />0.00336 weeks <br />7.72415e-4 months <br /> (MST) April 4,1986.
period, " heavy snowfa'll and severe wind conditions caused numerous
ground faults, voltage transients, and breaker trip actuations to ,
occur on three of the five 230 kv transmission lines supplying
outside electrical power to the Fort St. Vrain Station." At the
beginning of the period of transient conditions, the reactor was
operating at only 25% power, and so the turbine-generator set was
During
not operating and not loaded onto the reactor steam supply.
the transient conditions, total loss of offsite electrical power
never occurred, and low voltage swings were never low enough to
the automatic start of the standby / emergency diesel
actuate
generator sets. However, the offsite voltage transients were
,
,----=-,e - - - .-%.e-wey,.--- -- .---------------------w.v.---..-v..*.-e.-. ----..,,,...--,.,---,--1--,,,,w-wr--
e ,
-
.
- 2
.
- transmitted into and through the plant's electrical distribution
System. As a result, " electrical noise" caused the automatic
actuation and closure of the respective helium purification trains'
isolation. valves and the actuation of the helium circulators'
bearing water pump trip via undervoltage protection. The under-
voltage protection aytomatically closed the isolation valve on the
bearing water supply line, thereby leading to a low differential
pressure (loss of flow) trip of the bearing water pump (see FSAR
Section 4.2.2.3.7, pages 4.2-26). -Trips of the bearing water pumps
in turn caused circulator trips and loop shutdowns since backup
bearing water was not available a'nd is not normally made available
below about 25-30% power per plant procedures. Following a shutdown
on Loop 2, which occurred early during the period of transient con-
ditions, an apparent failure of the 'D' helium circulator shaft
shutdown seal to seat properly led to a small release of primary
coolant. The radioactive inventory of the small release was found
to be well below maximum permissible concentrations for such
releases. .
As a result of the electrical transients, operators performed a
manual scram of the reactor soon after the transients began. During
-
the remainder of the period of experienced transient conditions, the
plant experienced four brief periods of total loss of forced cir-
culation with the operators responding in each case within a few
minutes to restore circulator and steam generator operability.
Prior to this investigator's receipt of this LER, a telephone
discussion was held with NRC-NRR's Mr. K. L. Heitner and NRC-AE00's
Dr. R. R. Tripathi with regard to NRR's concern about this event.
In particular, NRR expressed concern over the apparent weak follow-
through on corrective actions with regard to (1) the electrical
supply circuit reliability for the relays for the bearing water
pumps (i .e. , improved power supplies off uninterruptable dc bus),
(2) means to avoid spurious control circuit transients on the inlet
valve to the helium purification trains (i .e. , possible improved
a
power supplies), and (3) means to effect adequate monitoring of such
transients. Since the text of the LER had stated that "the event
' initiating actions were directly storm related," NRR was also con-
cerned that the LER had not adequately addressed what will be done
by the licensee to prevent such occurrences in the future.
( ,
Also. prior to discussing th'e LER with the licensee, this inves-
l
tigator researched the history of the effects of electrical
l transients on loop shutdowns and circulator trips. This research
was performed using the LER file in the computerized Nuclear Safety
>
Information F)le Data Base. Other than technician-induced transi-
ents (i.e., opening the wrong breaker) such as reported in R0 85-
027, which was reviewed in our monthly report for January 1986, this
investigator found only two other apparently relevant LERs, both of
'
which were also of recent date. These LERs are RO 84-007 and R0 84-
e 008 (the partial f ailure scram event of June 1984). Both of these
!
previous events were initiated by the faulty actuation of a rapid The
(pressure) rise relay on the ~4160/480 volt transformer.
i
resulting transformer trip led to the temporary loss of 480 V
I
.
- - -. . - --. .- . --- - .-. .. - - -. - .-
-
. ..
-
..
3
,
'
!
- - Essential Bus 1 and to the trip of the bearing water pumps, in the
case of R0 84-007, the reactor was operating at less than 2% power
'
without backup bearing water. available, and a Loop 1 shutdown
occurred. In the case of R0 84-008, the reactor was operating at
about 50% power with backup bearing water available, and an 'A'
!
'
circulator trip occurred on buffer-mid-buffer differential pressure
as the result of a surge of backup bearing water following trip of
the bearing water . pump. The surge of backup bearing water caused a
water ingress event via the circulator sh' aft.
FSAR (Revision 3) Section 4.2.2.3.4, pages 4.2-17 and 4.2-18, states
'
that "if one of the operating bearing water pumps should fail, its
standby pump is automatically started by a low differential pressure
switch." This same section also describes the backup bearing water
system function. However, FSAR Section 4.2.2.3.7, pages 4.2-25 and
- 4.2-26, indicates that pump failure due to loss of power on Bus 1
L
(Loop 1) or Bus 3 (Loop 2) will trip all three bearing water pumps
, in the loop affected and initiate undervoltage protection. The
- undervoltage logic is implied to cause a throwover to Bus 2 accom-
panied by ths isolation of the normal bearing water supply line so
i that pumps can be restarted automatically off of Bus 2 but will not .
t cause an overs"pply of bearing water since backup bearing water
would have beer nutomatically initiated if in service. In R0 84-007
and R0 84-008, there was a loss of bus-supplied power. In R0 86-
017, there was initiation of the undervoltage protection without
total loss of bus power. The undervoltage protection closed the
valve in the supply line and thereby caused the low differential
pressure (loss of flow) trip on the bearing water pumps.
However, in the two earlier LERs, it is not clear whether the FSAR-
described bearing water protection logic was in force. Subsequent
, discussions with the licensee (M. Joseph) tends to indicate that the
system described in the FSAR was in fact the result of at least R0
l
84-008, wherein a reverse water ingress was experienced. Notably,
- R0 84-008 corrective actions addressed only control rod-related
problem resolution and not that of the initiating event'.
Following preliminary analysis, the LER for R0 86-017 was thbn
!
discussed at some length with the licensee (D. Goss and M. Joseph).
l
'
The licensee (M. Joseph) explained again, as this investigator has
heard often before, that the corrective actions for the LER will be
'
closed out in the closeout package assembled for the review of the
NRC Senior Resident. Inspector (SRI). The licensee has ao plans to
submit a supplemental LER or necessarily any other type report to
j NRC on this event and its follow-up. This position is consistent
with that developed and expressed previously following the licen-
'
+
see's participation in an industry course for utilities. The
- industry course recommended the avoidance of consnitments to LER
supplements as an unnecessary burden on licensee commitment track-
I ing. The " weak follow-through" noted by NRR in R0 86-017 apparently
j results from the complicated nature of the events and the fact that
investigations are still ongoing into long-term corrective
j actions. The licensee had simply not completed corrective actions
'
within the 30-day period regulated by 10 CFR 50.73 for submitting
.
i
- -.- -_-..._-- - - -_ - . - -_ _ - . - _ . - _ - - _ ___ - -__.
. .
,
. 4
n'
'
. LERs and now defers the assurance of completion of corrective
actions to the SRI on an unspecified time schedule.
As discussed in the LER, the licensee (M. Joseph) again indicated
that the protection setpoint for the bearing water pump undervoltage
relay was being addressed as part of the corrective action. The
licensee emphasized that this undervoltage relay was already on an
uninterruptable de bus, which .was unaffected by the electrical
transients. Instead, the sensing coil for ac undervoltage (where
the ac bus was affected by offsite power transients) was surmised to
have possibly been too high a low voltage setpoint (30 V ac). Since
none of the ac buses is immune to offsite electrical power transi-
ents unless it is already isolated onto an onsite power source, the
possible reduction of this low voltage setpoint appears to be the
primary practicable option under consideration.
Also, the licensee did not choose to go onto onsite power during the
snowstorm because the severity of the storm's effects was claimed
not to be predictable from the past 30 years of experience with the
grid. The storm-induced loss of offsite power event (R0 83-018) in
May 1983 was apparently not factored into the choice to remain tied
to the grid except that the result of the previous event had been
for fiRC to officially discourage the carrying of common loads on
both onsite and offsite power simultaneously during grid distur-
bances. Also, as mentioned previously, the offsite power transients
were never severe enough to cause an automatic start of onsite
emergency power. Adecisiontoisolatecertainessentialbusesfromf
the grid would appear to be a judgement call unless specific rulesI
, can be developed for such cases.
1
The transients experienced with the automatic isolation of the
j helium purification train are also still under review and analysis
by the licensee. There are apparently many inputs to and interlocks
on the valve closure control for the inlet to the helium purifica-
, tion train. The licensee could have overriden all , prohibits to
l opening the valve in order to effect a DBA-1 depressurization and
exhaust cleanup; however, the electrical transients would not permit
normal operation for coolant cleanup and recirculation of purified
helium. This situation led to the licensee's management decision to
l scram the reactor manually and also kept the purification train from
l performing normal coolant cleanup and recirculation for some two
hours. The valve closure signals may not have been due to any
instability on the dc instrument bus but rather may have been due to
feedback signals front various components operating on unstable ac
power. As illustrated in the drawing PI-23-1, Issue AT, the
interlocks on the control of the inlet valve to each purification
train are extremely complex because of the complexity in the equip-
ll
i ment configuration for the downstream processes. The licensee
't (M. Joseph) asserts that there is a continuing review and assessment
of this particular problem but that this problem is an operability
'
issue and not a safety issue since the capability to perform DBA-1
depressurization and coolant exhaust cleanup were not affected. *
- - = - _ --. - _ - . . _ _ _ _ _ - - . _ . - _ _ - - _ _ . _ _ _ - _ _ - _ - _ _ _ - _ _ _ _ _ . __ _ _
_ _ . _
,
' '
-
. .
,
.,
.. 5
.
.
Finally, the licensee (M. Joseph) notei that initial failure of the
'D' circulator shutdown seal to set properly cannot now be dupli- I
cated in' testing. Such failure to seat shutdown seals occurs
randomly but infrequently on various circulator shafts. Presumably,
some'small particle of foreign material is the source of this type
problem which typically disappears when t.he seal is reseated.
This investigator will continue to monitor the progress of solutions
to the corrective actions on R0 86-017; however, this investigator
finds generally that the cause of the problems are probably unique
to the Fort St. Vrain design complexity, particularly for the water-
lubricated circulator bearings and the coolant purification train.
Other than perhaps pump shaft seals on LWRs, there is little equiva-
lent complexity in similar functional systems on other reactors.
Preliminary information from the licensee also indicates that appar-
ently the uninterruptable dc buses and circuits were not affected.
Until the licensee's investigations are completed, this investigator
classified this event as Category 3.
Copies of NRC Form 423 are attached .for filing.
2. Review of Other Documents
During the current reporting period, other potential sources of
operating data have been screened. Some of these are listed as
follows:
a. Inspection and Enforcent (IE) Report 50-267/86-07 with
regard to the March 1985 inspection of the Radiological
'
.
Environmental Monitoring Program (the IE Report noted that
'
the release documented in R0 85-004 and which occurred on
March 17,1985, had not been summarized in the subsequent
Semiannual Radioactive Effluent Report for the affected ,
period).
l ,
I
b. Early Reporting of Events. A 10 CFR 50.72 report, event
number 04516, dated May 6,1986, described a brief power
transient due to the drift open of a tagged out Loop 1
'
drag block valve on the. main steam system. Reactor power
i rose from 34.5% to about 39% in response to automatic
'
reactor regulation against the steam demand. The transi-
i ent lasted 15 minutes until operators returned the valve
position and syst,em to normal. At the time, NRC-mandated
limits required that reactor power not exceed 35% of
l rated. A Region IV daily report of May 7,1986, described
- the same event.
c. NRC-NRR letter, April 10, 1986, requesting of the licensee
additional information on the proposed fuel surveillance
j program.
l d. NRC-NRR ' letter and accompanying safety evaluation,
! April 15, 1986, on the Fort St. Vrain 1985-86 licensed
j operator requalification program.
.
1
I
- - . . - - . _ . . - , . . . . . . , . _ __ . - . - . - . - . - . - - . _ . - _ - _ - - - -