ML20042F977
| ML20042F977 | |
| Person / Time | |
|---|---|
| Site: | Fort Saint Vrain  |
| Issue date: | 05/04/1990 |
| From: | Crawford A PUBLIC SERVICE CO. OF COLORADO |
| To: | Weiss S NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM), Office of Nuclear Reactor Regulation |
| References | |
| P-90128, NUDOCS 9005100250 | |
| Download: ML20042F977 (8) | |
Text
t m,y.
O Public Service' "O'QL
. P.O. Box 840 -
Denver CO 80201 0840 t
May 4 1990-Fort St.'Vrain A. Clegg Crawford Unit No. 1-Vice President -
-,-90128 Nucler oproHons -
P s
U.S.~ Nuclear Regulatory Commission
- ATTN: _ Document Control Desk Washi_ngton, D.C. :20555 ATTN:
Mr. Seymour H. Weiss, Director Non-Power Reactor, Decommissioning and Environmental Project Directorate Docket.No. 50-267
SUBJECT:
REACTIVITY EFFECTS OF WATER INGRESS
Dear Mr. Weiss:
The purpose:of this letter is to address questions raised by Mr. Pete_
Erickson of-the NRC staff, in several phone conversations, regarding the' reactivity effects of water in the FSV core. Attached, for your-
/information, is an assessment of possible sources of water in-the core, means of detecting water in the PCRV, and the effects of water l ingress.on. reactivity.
'If you have any questions related to this information, please contact -
Mr..M.H. Holmes-at-(303)480-6960.
.Very truly yours, A. C. Crawford
'Vice President
- Nuclear Operations ACC/JRJ:faw Attachment.
cc: Regional Administrator, Region IV
~ Attention: Mr.
J.. B Baird, Technical Assistant-Division of Reactor Projects Mr. R. E. Farrell
. Senior Resident Inspector Fort St. Vrain n e a u niir fg e.
n
~
-Attachacnt to P-90128
~
- ,Page 1 of 6.
c.
Assessment of the Effects of Water Ingress on the FSV Core Introduction Ingress of water into the FSV core causes an increase in reactivity.
Under defueling conditions, a water vapor / helium gas mixture in - the core will have an insignificant effect on reactivity.
- However, introductionLof liquid water into the core has the potential for causing a significant reactivity increase.
PSC has examined possible means of liquid water entering the core. The likelihood of each of these. means of -water entry is evaluated and water detection capabilities are 4
'1 discussed.
Reactivity Effects of Water Ingress The presence of water in the FSV core causes an increase in reactivity due to reduction of neutron leakage, reduction of the Th-232 neutron capture rate and. reduction of the effectiveness of control poison. The net coefficient of reactivity for water is approximately 2.2 E-5' delta k per lbm of water in the core (FSAR tection 14.2.1.4).
This value is accurate whether water is in the liquid or vapor phase, and is not-significantly influenced by core temperature. The presence of water-t vapor in the core would have a relatively small effect on reactivity, less than 0.01' delta k,
when the reactor is operating at power (the worst case steam ingress is evaluated in FSAR Section 14.2.1.4) and less than-0.001 delta k under shutdown /defueling conditions (Section 5.2.2 of the FSV Defueling SAR, submitted in PSC letter dated August' 16, 1989,
- .Crawford to Weiss,P-89287). The presence of liquid water in the core has the potential for larger reactivity contributions than water vapor, since a ' greater mass of liquid water could accumulate in the core. As discussed below, the introduction of quantities of liquid water in excess of 1000 lbm into the core is considered to be incredible..
7 Water Ingress Detection Capabilities o
PSC. has several diverse and redundant means to determine that water is' entering the PCRV. These include the analytical moisture monitors, the receiver tanks in the new PCRV pressure control system, and malfunction of an optrating helium circulator should its suction become flooded with water.-
The analytical moisture monitors are described in FSAR Section 7-3.2.1.
These moisture monitors are reliable, and their indication is monitored regularly by the reactor operators (control room panel I-9305 and' data-logger)
PSC is in the process of installing equipment for a modified PCRV pressure control system, scheduled for completion in June, 1990.
This system will maintain the PCRV at atmospheric pressure by removing helium from the PCRV at essentially the same rate it is entering from circulator seals and control rod drive purges (estimated to be approximately 10 scfm).
Helium will leave the PCRV through the existing helium purification lines, flow through the High Temperature Filter Adsorber (HTFA) and the helium purification cooler of the in-service w
j
.v AttachirGnt' to P-90128
,*-.Page 2.of 6-c-
train, then through one of two helium compressors and into the high pressure helium storage bottles.
New dryer units at the outlet of each compressor will have the capability of reducing the moisture concentration in this gas stream exiting the PCRV to a dew point of -75 degrees F.
The in-service helium purification cooler cools the helium to between 40 and 50 degrees F.
Should the moisture concentration in the primary coolant reach a dew point between 40 and 50 degrees F (approximately 10,000 to 15,000 ppm by volume),. moisture will condense, which can then be withdrawn to a receiver tank. The receiver tank has a window to provide for monitoring of condensate from the helium purification coolers on a regular basis.
Build-up of water in the receiver tank serves to warn the operators, in addition to the analytical moisture monitors, that large quantities of moisture are present in.the primary coolant.
In the event water in excess of 5,000 gallons enters the PCRV, an operating helium circulator would be impacted due to water entering the
]
circulator suction.
Operators, who monitor circulator parameters on a t
routine basis, would investigate the cause of the malfunction and j
isolate the source of water to mitigate the water ingress into the PCRV.
j Sources of Water and the Likelihood of Introducing Water into the Core There are three possible sources of water into the PCRV, as follows:
bearing water flowing past the buffer seal or the shutdown seal of a helium circulator, up the shaft and into the PCRV; steam generator tube l
leaks; and water from the PCRV liner cooling ~ system or the Fuel Handling i
aiachine (FHM) liner cooling system flowing through a cracked tube and j
- through a postulated breach in the adjoining liner into the PCRV.
FSV
-has experienced water ingress via the helium circulator shafts on a q
e number of occasions and has implemented various measures to ' reduce the 4
l probability of these events.
Pin hole size-leaks occurred in the superheater bundle of two different steam generator modules during power o
operation over the life of FSV, one in 1977'and tr.e other in 1982. The subheaders feeding the leaking tubes were cut and capped.
. Water has:
entered the PCRV through a crack in the core support floor (CSF) liner i
due to several leaking PCRV liner cooling tubes in the CSF.
All known 1
'f leaking cooling tubes in the CSF have been isoiated. There have been no cooling tube leaks in.the FHM.
i Three means by which _ liquid water from these sources could enter the
{
E reactor core were cosidered. These are:
13 flooding the PCRV with j
water and subme:ging part or all of the core under water; 2) rupture of 1
cooling tub % in either the PCRV top head liner cooling system or the f
FHM cooling system, with postulated breaches in the liners, such that 1
water falls down on top of the core; and 3) circulation of primary j
coolant containing moisture in the vapcr phase, with condensation i
postulated to occur on the PCRV top head, and water droplets falling l
onto the top of the core.
If core temperatures are assumed to be lower than primary coolant temperatures, which is not anticipated, condensation could occur in the core coolant channels.
'( J Attachment to'P-90128
_.-Page 3 of 6 L..
3 The likelihood of introducing liquid water into the core from each of x
these three means is assessed in the following paragraphs.
- 1) Submergence l
4 Calculations have determined that 150,000 gallons of water would have to enter the PCRV beforo water would contact the. bottom layer of fuel blocks.
FSV Administrative Procedure D-2 requires that at least one operable analytical moisture monitor be in operation, analyzing primary coolant moisture content, and, if this is not possible, a helium circulator must be operating.
Normal helium circulator operation provides indication that gross flooding of the PCRV has not occurred.
Operators monitor makeup water to the bearing water surge tanks,. PCRV liner cooling surge tanks and the condensate storage tanks and would investigate abnormal makeup water Jnventories.
Operators.would be warned of substantial amounts of water in the PCRV by the analytical moisture monitoring system, buildup of condensation in the in-service helium purification cooler, malfunction of the operating helium circulator and abnormal makeup requirements.
The operators would be able to investigate and isolate the source of water into the PCRV.
Based on the numerous warnings and the length of time operators would have to isolate the sources of water ingress (on the order of days for liner cooling tube or steam generator tube ruptures), PSC considers ingress of 150,000 gallons of water into the PCRV to be incredible.
- 2) Spray l
1 In addition to flooding the core, PSC has also considered : the j
possibility of a water spray or mist entering the core, since such I
l partial water densities within the core coolant channels could potentially result in significant reactivity increases. Hypothetically, spraying water into. the core could come from the PCRV top. head liner cooling system or the Fuel. Handling Machine (FHM) liner cooling system, i
Several barriers exist between each of these sources of water and the l
core. The walls of the square steel liner cooling tubes are 0.120 inch
- j thick in both the PCRV top head and in the FHM. The PCRV top head l
carbon steel liner is 0.75 inch thick and the refueling penetration
-s carbon steel liners are 0.59 inch thick. Since the cooling tubes are 4
welded to the outside of these liners, leaking cooling tubes would not i
result in a water spray on the core unless the top head liner was also l
breached. Surveillances (SR 5.2.14) of PCRV liner thickness have not e
detected any reduction in liner thickness.
PSC has evaluated the potential for PCRV top head liner cooling tubes leaking into the core, and has concluded that such leaks are incredible.
t In assessing the potential for PCRV liner cooling tube leaks affecting i
the core, PSC has considered past leaks from the Core Support Floor (CSF) liner cooling tubes into the PCRV.
The PCRV top head liner geometry and stresses are very different from those of the' CSF liner.
While any liner cooling tube has the potential to leak, the top head liner is designed for the stresses which could occur if a leak developed.
Unlike the sidewall portion of the CSF liner, the PCRV top head uses anchor bolts to attach the top head cavity liner and penetration liners to the concrete.
These anchor bolts are
)
I 3
Attachment-to P-90128
,, Page 4_of 6 conservatively designed to withstand. forces which could develop if a liner cooling tube leak occurs, so _ that pressurized water behind the liner could not push the liner away from the concrete and stress liner cooling tubes or their welds. The sidewall portion of the CSF liner, l
where previous liner cooling tube leaks have occurred, has no anchor bolts to anchor it to the concrete.
Forces due to any cooling tube leaks,would push the CSF sidewall liner away from the concrete, resulting in high stresses in the sidewall, CSF liner cooling tubes, and the cooling tube welds to the CSF liner.
In other respects, the PCRV top head liner and cooling tubes are not likely to degrade or leak compared to the CSF leaks which have occurred.
As ' discussed in FSAR Section 5.9.2.8, the PCRV liner hot spot situation in the PCRV top head (maximum concrete temperatures of 201 degrees F) is much less severe than the CSF hot spot situation under the core barrel (maximum concrete. temperatures up to 326 degrees F)._ The PCRV top head only-experienced cold side gas temperatures during operation, whereas the CSF experienced larger temperature differentials, being subject to both hot and cold side gas tempertures on different faces during l
operations. The~ future potential for leaks or degradation developing i
during defueling is minimal compared to that experienced during plant operation since there will be no thermal cycling of the PCRV top head liner during defueling. Overall, the potential for PCRV top head liner cooling-tubes developing leaks and affecting the core during defueling i
is so minimal as to be incredible.
The possibility of FHM liner cooling tube ruptures was also considered.
The FHM storage racks which house the fuel elements are cooled by water 1
which flows-through square cooling tubes attached to the outside of the
. fuel storage compartment dividers. The cooling tubes are thus separated from the FHM interspace, which may be open to the core, by.0.25 inch thick steel. There are alarms and drains on the FHM storage racks so that a leak will be detected and water can be drained out of the FHM.
l Based on the above information, PSC considers the development of a liner cooling tube leak and concurrent liner breach, in a location'which would j
result in water falling on top of the core over the course of defueling,
?
to be incredible.
- 3) Condensation PSC has considered the possibility of liquid water entering the core as the result of condensation of water vapor.
In the event of high l'
moisture concentrations in the primary coolant helium, water would 9
condense on surfaces whose temperatures are below the temperature of the primary coolant and below the dew point temperature corresponding to the water vapor concentration.
It is not anticipated that temperatures of the reactor core would decrease below primary coolant temperatures during the course of defueling, since the core is a heat source in the PCRV. However, should this occur when the primary coolant is saturated, or nearly saturated, with water vapor, water could condense on core surfaces, such as in the coolant holes and on the sides of the fuel blocks. Another mechenism by which condensate could enter the core l
Attachment to P-90128
,, i Page 5 of 6 involves condensation of water vapor on the PCRV top head, and water droplets falling down upon the top of the core.
Water falling vertically from the PCRV top head would not enter the orifice-valves and would therefore not be expected to-enter coolant
[
holes in the fuel blocks.
Water falling on the upper metal plenum elements could accumulate and run down the sides of these elements and down the sides of a fuel column.
FSV _has.' conducted numerous startups with high moisture levels in the PCRV, sometimes with large quantities of water having been in the PCRV for months.
In all but one startup the reactivity discrepancies for these conditions were well within the 0.01 delta k allowed by Technical
. Specifications.
PSC considers introduction of water into the core by the mechanism of condensation to be credible, though anlikely, over the course of defueling.
The following paragraphs discuss the single occurrence of condensation in the core that significantly influenced reactivity.
Maximum Credible Reactivity Increase Due to Credible Water Ingress-Condensation PSC has experienced a
significant reactivity increase due to condensation once over the life of FSV. This took place during zero power testing, with the reactor at approximately 1 E-5 percent reactor power and adding negligible heat to the primary circuit. Steam from the auxiliary boiler had been supplied to a reheater to heat the primary coolant to approximately 180 degrees F.
While performing a startup on January-23,: 1975, reactivity increased by 0.015 delta k as a result of water in the core. This is discussed in Abnormal Occurrence Report No.75-07A (final), dated 4/4/75.
It was discovered that large quantities of. water had entered the PCRV via the shutdown helium circulators.
It was estimated that 4,250 gallons entered the PCRV. During the ensuing
. reactor startup, it was projected that approximately 750 lbs.
of water were in or on the. graphite of the core.
l As a result of this -event, tests were conducted in which a. graphite sample was suspended in helium with moisture present.
These tests demonstrated that absorption of water into the graphite and condensation of water on the graphite occurred in quantities such that, when compared to the FSV. core, as much.as 1,000 lbm. of water could accumulate in the core. Accumulation of 1,000 lbm. of water would result in a reactivity increase of 0.02 delta
'k, assuming the core to be fully loaded with fuel. This is considered by PSC to be the worst case reactivity effect of moisture in the FSV core.
t s
7 ro Attachment to P290128 Page 6 of 6 l
[
I A number of differences exist between the conditions of zero power testing, when this reactivity excursion occurred, and defueling l
conditions, which tend to make recurrence of a similar event unlikely.
In the 1975 event, operators were unaware of the presence of large quantities of water in the PCRV prior to startup.
Presently, a moisture l
L trending program is in place whose objective is to identify increasing j
moisture levels' 50 that actions will be taken to determine and isolate y-any source (s) of moisture. Whereas the steam generators were being used
[
as a heat source during the 1975 event, during defueling PSC uses the steam generators only to cool the primary coolant.
High water vapor concentrations would lead to condensation on the steam generator tubes, 1
L and on cooler portions of the primary circuit in the bottom of the PCRV i
rather than in the core, since it is expected that core temperatures
- would exceed temperatures of components in this region.
Condensation on L
p the defueling elements, whose temperatures will generally be lower than J
the temperature of fuel blocks, would have little, if any, effect on reactivity. Thus, occurrence of an event similar to the 1975 event over i
the course of defueling is considered to be very unlikely, While PSC cannot foresee circumstances during defueling in which reactivity could increase by 0.02 delta k as discussed above, a review of the shutdown margins during the remainder of defueling (Table 1) shows that this increase would still leave a substantial shutdown
- margin, even during control rod withdrawals for shutdown margin assessments.
Conclusion PSC has considered different water ingress scenarios which would impact core reactivity. Tests concluded that it would be possible, through the
- combined effects of condensation and absorption, to introduce a maximum of 1,000 lbs. of water into the core 'with. a resultant reactivity increase of approximately 0.02 delta k, conservatively assuming a fully fueled core.
PSC hat studied shutdown.argins through the remainder of defueling and determined that reactivit;/ increases of 0.02 delta k would still-leave a substantial shutdown margin (see Table 1). Water ' ingress i
events that would result in partial or complete submersion of the core are considered to be incredible.
I r
r w
9 k
(
\\>
p, Table 1 Shutdown Margins During Remainder of Defueling Sequence.
From 25 Fueled Regions Remaining NO. OF 2 ROD SON
$DM FUELED 2 R00
- PA!R$ OUT ASESS.
ASESS.
REGIONS PAIR $ OUT k(ef f)
ROD k(eff) 25 30+37
.8798 4
.9268 24 37+24
.8717 4
.9267 23 24+31
.8730 4
.9268 22 31+23
.8677 4
.9240 21 23+32
.8621 4
.9234
?O 32+13
.8929 4
.9497 19 13+19
.8936 4
.9498 18 19+12
.8564 4
.8979 17 12+18
.8432 4
.8957 16 18+11
.8530 4
.8926 15 11+17
.8531 4
.8927 14 17+8
.8240 4
.8460 13 8+14
.8135 4
.8423 i
12 14+9
.8175 4
.8452 11 9+15
.8239 4
.8468 10 15+10
.8170 4
.8368 9
10+16
.7703 8
16+2
.7848 7
2+5
.8075 6
5+7
.7869 5
7+4
.7563 4
4+1
.7852 3
1+6
.6472 2
6+3
.6526 1
3
,6526 0
Next 2 regions to be defueled in sequence.
.