ML20086A650
| ML20086A650 | |
| Person / Time | |
|---|---|
| Site: | Washington Public Power Supply System |
| Issue date: | 09/30/1983 |
| From: | BABCOCK & WILCOX CO. |
| To: | |
| Shared Package | |
| ML20086A642 | List: |
| References | |
| BAW-1742, BAW-1742-R01, BAW-1742-R1, GL-81-21, NUDOCS 8311160058 | |
| Download: ML20086A650 (40) | |
Text
- - - - - - - - - - - - - - - - - - - - - - - - - - -
I RESPONSE TO NRC LETTER 81-21
[
[ NATURAL CIRCULATION COOLDOWN TASK 3A]
FOR
)
TENNESSEE VALLEY AUTHORITY AND WASHINGTON PUBLIC POWER SUPPLY SYSTEM W
====='
3 CONTRACT NO. 600-5278 BAW-1742, REV.1 AUGUST 1983 E,
E Ru 1888R 4 8as220 sabcock awiscox J
A PDR a McDermott company
BAW-1742, Rev.1 September 1983 i
l NATURAL CIRCULATION C00LDOWN TASK 3A Prepared for
[
WASHINGTON PUBLIC POWER SUPPLY SYSTEM L
WNP-1 Project THE TENNESSEE VALLEY AUTHORITY Bellefonte Unit 1 and 2 Projects B&W Contract No. 600-5278
(
L E
r L
BABCOCK & WILC0X Utility Power Generation Division P. O. Box 1260 Lynchburg, Virginia 24505 Babcock &Wilcox W
l l
l l
l l
l CONTENTS Page 1
1.
I NT RO D U CT I ON.........................
1-1 1.1.
Objective 1-1 1.2.
Scope 1-1 1.3.
B ac k gr ou n d....................... 1-2 r
l 1.4.
Sumary and Conclusions 1-3 2.
ANALYSES........................... 2-1
[
2.1.
In trodu ction...................... 2-1 L
2.1.1.
Bubble Avoidance................ 2-1 2.1.2.
Bubble Mitigation 2-2 2.2.
Analytical Method 2-3 2.2.1.
RV Up pe r Head Cool down............. 2-3 2.2.2.
RV Upper Head Void Mitigation 2-6 2.3.
Results of Analyses 2-8 2.3.1.
RV Uppe r Head Cool down............. 2-8 2.3.2.
RV Upper Head Void Mitigation 2-10
[
3.
BASIS FOR OPERATIONAL GUIDELINES FOR NATURAL CIRCULATION l
C0 0 '.DOW N........................... 3-1 3.1.
Objective 3-1
(
3.2.
B ac k gro u nd....................... 3-1 3.3 Control Functions for Natural Circulation Cooldown... 3-3 3.3.1.
General Sequence of Events........... 3 3.3.2.
Reactivity Control............... 3-5 3.3.3.
RCS Inventory Control 3-6 3.3.4.
RCS Pres sure Control.............. 3-8 3.3.5.
Steam Generator Pressure Control........ 3-10 7
L 3.3.6.
Steam Generator Inventory Control 3-11 4.
OPERATOR TRAINING AND PROCEDURE DEVELOPMENT r
L
( By TVA a nd S S )....................... 4-1 5.
R E FE R E N CE S.......................... A-1 r
1 h
Babcock & Wilcox
. =o
List of Figures Figure Page 1.
Reactor Vessel and Internal s................
2-11 2.
Schematic Diagram of Upper Portion of Reactor Ves sel a nd Inte rnal s....................
2-12 3.
Upper Plenum and RV Upper Head Mass Transfer Model.....
2-13 4
Heat Transfer Model 2-14 5.
Nodi ng Di a gram.......................
2-15 6.
Pressurizer Model of RV Upper Head.............
2-16 7.
Required AFW Inventory Vs Time to DHRS Cut-In 2-17 8.
Natural Circulation Cooldown Temperature Ys Time......
2-18 9.
Typical Natural Circulation Cooldown -- Pressure-Temperature Diagram....................
2-19
- 10. Natural Circulation Cooldown Control Functions.......
3-12
]
1 Babcock & Wilcox.
)
1.
INTRODUCTION 1.1.
Objective This report is in response to the Eisenhut Letter of May 5, 1981.1 The first objective of this report is to show by analyses that the Bellefonte 1 & 2 and WNP-1 Plants can be controlled and cooled down by natural circulation, with-out forming a void in the reactor vessel (RV) upper head.
A second objective
{
is to show by analyses that if a void fonns in the RV upper head, it can be detected and mitigated so the cooldown can be continued.
A third objective is to show that the analyses above form a satisfactory basis for development of operator guidelines on natural circulation cooldown and that the operator will be satisfactorily trained to implement these guidelines.
1.2.
Scope Several analytical activities on the natural circulation cooldown process are summarized in this report.
The cooldown starts with a loss of all reactor coolant pumps (RCPs) and goes to the cut-in of the decay heat removal systen.
System dynamic analyses on natural circulation cool downs, with subcooled bleed fram the RV head vent, are provided in Section 2.
The dyna'nics of RV k
head voiding and mitigation are also provided in Section 2.
Section 3 provides the basis for operational guidelines for natural circula-tion cooldown.
This basis is adequate to demonstrate the application of the methods analyzed in Section 2.
Section 4 describes the operator training r'
L program and procedure development for natural circulation cooldown.
1.3.
Background
On June 11, 1980, a total loss of component cooling water flow to the reactor coolant pumps occurred at the St. Lucie Unit 1 plant.2 When the coolant flow L
could not be restored in the time limit specified by the plant's Technical Specifications, the reactor was manually tripped.
Within two minutes af ter
(
the reactor trip, the RCPs were also manually tripped.
Natural circulation cooldown was initiated approximately 27 minutes later.
Babcock & Wilcox 1-1
. mo--.=
- m
4 Based on pressurizer level and primary system pressure response, it has been concluded that void formation occurred in the upper head region during the natural circulation cooldown.
At St. Lucie, the measured hot and cold leg temperatures at the time of voiding were highly subcooled.
It appears that the fluid in the upper head was much hotter, relatively stagnant and in poor communication with the rest of the primary system.
It is postulated that the steam bubble in the upper head area was produced when the system pressure dropped below the saturation pressure corresponding to the temperature of the fluid in the upper head.
On May 5,
- 1981, D.G. Eisenhut (NRC) sent a letterl to all Licensees and applicants making several requests:
1.
A demonstration (e.g. analysis and/or test) that controlled natural circulation cooldown from operating conditions to cold shutdown condi-tions, conducted in accordance with your procedure, should not result in reactor vessel voiding.
2.
Verification that supplies of condensate-grade auxiliary feedwater are suf ficient to support your cooldown method.
3.
A description of your training program and the provisions of your pro-cedures (e.g. limited cooldown rate, response to rapid change in pressur-izer level) that deal with prevention or mitigation of reactor vessel voiding.
This report is in response to the May 5 letter.
1.4.
Summary and Conclusions The objectives of this report have been met.
A method has been developed from analyses to permit natural circulation (NC) cooldown at up to 50F/hr in the core and RC loops without voiding in the RV upper head.
This avoidance of a bubble in the RV upper head does not increase the supplies of AFW other-wise required.
This method limits the rate of depressurization to prevent voids while cooling the RV upper head by bleeding of the subcooled coolant from the upper head out the RV high point vent and by heat transfer to the m
reactor coolant system and containment.
The allowable rate of depressuriza-tion is detennined by analyzing the hottest fluid in the RV upper head versus time.
The bleeding results in a faster cooldown rate in the upper head, and consequently, a faster allowable rate of depressurization.
1-2
The RV upper head is cooled without flashing by bleeding at least 3000 gallons of coolant from the RV head vent for each 50F of RCS loop cooldown.
The cooltwn of the RV upper head as a result of heat transfer from a stag-nant upper head was also stu died, but was not shown to be viable.
The calculated cooldown rate of the upper head without a RV head vent was less than 2F/hr.
These results are therefore, not discussed fu rther in this f
report.
If a 50 gpm average bleed flowrate is maintained during cooldown, the RV upper head cooldown does not limit the RCS depressurization corresponding to
(
the loop design cooldown rate of 50F/hr.
(The WNP-1 unit has a head vent design flowrate of 50 gpm and the Bellefonte 1 and 2 units have a head vent design maximum flow rate of 200 gpm).
A RV head bubble alone presents no safety problem.
In the event upper head voiding does occur, methods are included herein for its recognition and mitigation.
The void may be mitigated by venting through the RV high point vent.
The RV thermal stresses associated with this venting have been qualitatively assessed to be acceptable based on adherence to the basis for guidelines provided, (specifically the RV upper head and hot leg temperature differences anticipated with use of the guidelines).
The need to perfonn an NC cooldown is remote.
The operational guideline infonnation (Section 3) recognizes this, as well as the fact that the Bellefonte 1 and 2 and WNP-1 RCPs are capable of running intermittently with
(
a loss of all cooling water.
Information is provided on operator control functions that must be addressed during NC plant cooldowns.
The control functions include reactivity, RCS pressure, RCS inventory, SG inventory and SG pressure. Procedure development and operator training on these procedures for NC cooldown are described.
[
[
1-3 Babcock & Wilcox
. me n
9
2.
ANALYSES 2.1.
Introduction These analyses were directed toward (1) avoiding a steam bubble in the RV
/
upper head during a NC cool down/deprassurization and (2) mitigating the bubble if it occur:.
(The upper head region design is shown in Figures 1 and 2.)
2.1.1.
Bubble Avoidance Several mechanisms can have a significant effect on the formation of a void (bubble) in the B&W designed reactor vessel upper head region during NC cool-
~
down/depressurization transients:
(
Cooldown/Depressurization Rate - Voids will not form in the RV upper head l
until the coolant in the upper head reaches saturation conditions.
(
Natural Circulation Flow - NC flow is from the upper plenum to the top of
{
the plenum cover and back down to the outer annulus.
The flow from the upper plenum is at the temperature of the hot leg (core outlet).
During a NC plant cooldown, the coolant temperature in the upper head region of
(
the reactor vessel will not follow the loop coolant because the flow rate through this region is small (or stagnant in the uppermost part) without
[
forced flow in the RCS.
Heat Transfer - The heat transfer is from the upper head fluid and netal
(
to the cooler fluid at the top of the plenum cover, the control rod drive (CRD) nozzles, the column weldnents and other cooler surroundings.
/
L Forced RCP Flow Prior to Cooldown - The RV upper head region temperature is about equal to the hot leg temperature during forced flow conditions in the RCS. Operating RCPs af ter a reactor trip will reduce the RV upper head temperature from the full-power hot leg temperature down toward the lower post-trip hot leg temperature.
The amount of upper head tempera-ture reduction will be dependent upon the length of RCP operation assumed af ter the reactor trip and the post-trip decrease in hot leg temperature.
2-1 Babcock & Wilcox a esc 0erniest company
RV Head Bleed - Bleeding through the RV head vent replaces the stagnant upper head fluid with cooler, subcooled RC hot leg fluid.
This minimizes the temperature lag, allows continued cooldown without flashing and cools the RV upper head metal, which is a heat source for the RV upper head fluid.
Hence, these flows directly affect the temperatures in the upper head region.
Analyses were perfomed to determine the maximum allowable cooldown/depressurization rate of the coolant in the RV upper head during an NC cooldown following 100*. power operation without forming a steam bubble.
All the mechanisms above were considered in this analysis.
RV head bleed was found to be the most effective upper head cooling mechanism.
Tne upper head cooldown rate from these analyses was used to
~
detemine the time at which the decay heat removal system (DHRS) could te cut in without flashing in the RV upper head.
Two cases of RV head bleed (rate) were analyzed; bleed at average rates of 50 and 200 gpm.
2.1.2.
Bubble Mitigation Several mechanisms can mitigate a steam bubble in the RV upper head, should it occur:
Heat Transfer - The bubble will eventually cool and condense due to heat transfer to the cooler surroundings.
This is a very slow process because of the insulation around the RV upper head and the low heat transfe r coefficients of steam.
Mixing - The bubble can be condensed by mixing with subcooled coolant.
This mixing is very small under NC conditions.
RV Head Vent - The bubble may be displaced from the RV head by venting through the head vent.
Collapsing - It is very difficult to completely collapse a steam bubble in the RV upper head by simply raising the RCS pressure.
In fact, in an ideal adiabatic system no condensation, as a result of this compression, will occur.
The RV upper head, because of the low heat losses without mixing, approximates an adiabatic system.
]
.s h
Babcock &Wilcox 2'-2
. mo. -.n A:'
.m.
^
Analyses were perfonned to evaluate the mitigation of a steam bubble in the i
RV upper head in the event that one is formed during an NC cooldown.
The primary mechanism considered in the analyses was venting the bubble through the RV ba vent.
Because of the relatively short times required for unting, these other mechanisms were of minor importance.
The venting was analyzed at two different pressures (1020 and 2500 psig) for comparison.
2.2.
Analytical Method 2.2.1.
RV Upper Head Cooldown The cooldown of the RV upper head under NC conditions is a result of both heat and mass transfer (flow) from the upper' head.
The flow paths are shown in Figure 3 and the heat transfer model for the analyses is shown in Figure 4.
tiass transfer from the upper head is through the plenum cover drain holes and the RV head vent.
Heat transfer from the RV upper head is by a coa-bination of conduction, convection and and a very small amount of radiation.
The model for heat losses through the metallic insulation yiel ded heat transfer values very similar to those measured in the field, however this mode of heat loss was very minor compared to the heat transfer due to RV head bleed.
The coolant rising from the fuel assent)1y upper end fitting into the upper plenum and column weltents is assumed to be the same temperature as the RCS hot leg since the bulk of this coolant goes directly to the hot leg.
'nitially, the upper head water temperature was assumed to be 590F, based on an assumed RCP trip concurrent with reactor trip and subsequent RCP coast-down. The RV head metal temperature was assumed to be 626F. This is approxi-mately the hot leg temperature for 100% power.
The RCS loops, including the hot leg, were assumed to cooldown at 50F/hr.
Some of the other assumptions used in this analysis are listed below:
1.
Convective heat losses from the CRDs were modeled as conductive
[
heat l os ses.
This is a standard analytical method for modeling fi ns.5 r
L E
L F
2-3 Babcock & Wilcox e ak0ermott compeay l
2.
The flow rate for natural circulation was 3% of full flow.
Based on operating experience and the NATURAL code,
this flowrate is considered to be conservatively low.
3.
Where conductive heat trans fer exists between adj acent nodes of different materials with different thermal conductivities, the lesser value or limiting value was used.
4.
Vendor supplied transference values were used at insulation / air inter-
~
faces.
5.
The ant >ient (reactor building) temperature was assumed constant at 120F.
For the analyses, the upper portion of the reactor vessel and internals was divided into a multinode representation as shown in Figure 5.
The mass trans-fer model was superimposed on this multinode model as shown by the solid and dotted flow paths.
A solid line from one node to another signifies mixing.
A dotted line signifies no mixing, such as the case for coolant rising inside the column weldments from the upper plenum to the RV upper head.
The distri-j bution of the coolant flow in the RV upper head was based on realistic assump-(
tions. Water enters the upper head region from the coolant circulating above the plenum cover.
Fluid rises upward from the row of nodes until it reaches the dome.
At this point, the flow path connects with the next radially inward path.
This pattern continues until all flow has been directed to the center flow path, at which time it exits out the top water node through the RV head vent.
Figure 5 shows that the upper head coolant which exits back through the plenum cover drain holes comes from only the first layer of nodes in the upper head.
This assumption is believed to be somewhat conservative and was used since a more realistic flow pattern was not determined in the
]
analyses.
The dimensions of the nodes were chosen to accommodate the dimensions of the actual hardware as much as possible.
1 5
Babcock & Wilcox r
2-4
As explained on Figure 5, each node represents a three dimensional ring in the analysis.
Finite dif ference equations were then written for each node volume.
This set of finite difference equations was then solved simultane-ously for each discrete time step.
A 25 second time step was chosen based upon conventional stability cri teri a.5 Future node temperatures were cal-culated based upon the current temperature plus the heat and mass transfer over the time step. The general fonn of the finite difference equations is:
Future = T resent + at[Qk + Oh + Om + Or3 T
P oxCpxV Where:
at = time step 1
i L
Qk = conduction heat transfer = -kA AT/aX l
Qh = convection heat transfer = hA AT
(
Qm = heat transferred with mass = o x V x Cp x (Tnew-T resent)/at p
4 Qr = radiation heat transfer = B x A x c x [(T resent)4 - (Tadj.I 3.
p o = density of node material r
L Cp = specific heat of node material c = emmissivity of metalic insulation V = volume of node k = thermal conductivity of node material A = horizontal cross sectional area AT = temperature dif ference across interface
(
aT/aX = temperature gradient across node h = convection coefficient B = Stefman-Boltzman Constant Tadj = temperature of adjacent node (both T resent and Tadj p
in Qr equation are absolute temperatures)
Tnew = new temperature of node due to mass transfer, present - T n) exP [(-6/o x V)at] + T n.
= (T i
i
[
6 = nass flow rate into node, T n = weighted mass average incoming temperature, i
{
T resent = present node temperature.
p I
[
2-5 Babcock s.Wilcox
.=<o n....
f Conductive heat trans fer was considered at boundaries of similar media (air-air, steel-steel, water-water, i nsulation-i nsulation).
Convective heat losses were considered at other boundari es,
( ai r-steel,
air-insulation, steel-wa ter).
Radiative heat transfer was also considered at the surface of the insulation to the nearest gray body, but this heat transfer was very minor.
The CRD convective heat losses (to the service structure region) were modeled as conductive heat losses. Ambient conditions of 120F were assumed.
Qk = -kA(dT/dx)
Qk = Heat transferred by conduction k = thermal conductivity of carton steel A = horizontal-cross sectional area dT/dx = linear temperature gradient along CRD length.
This additional Q tem was added to the finite difference equations for nodes in the RV head (dome) that contain CRD nozzles.
The leadscrews and column weldnents were modeled similarly.
The masses and volumes of these components were distributed among the applicable node rings to take into account the cooling by these components.
All sources of heat -- both into and out of each node -- were summed and then divided by the mass and Cp of the node.
This tenn was added to the present temperature to obtain the future node temperature.
The process was carried out for all nodes before continuing on to the next time step.
2.2.2.
RV Upper Head Void Mitigation Analyses were performed to determine under what conditions a void, if it occurred, could be vented from the RV head vent.
KPRZ, a non-equilibrium digital pressurizer model, was utilized to analyze the reactor vessel upper head region as a pressurizer.6 The reactor vessel high point vent valve was modeled as a pressurizer relief valve, and flow through the RV upper plenum cover was approximated by flow through a pressur-izer surge line.
J N
Babcock & WHcox 2-6
..o
The pressurizer model of the RV upper head is shown in Figure 6.
KPRZ eval-uates three thermodynamic regions in the pressurizer -- one for steam, one for subcooled liquid, and one for two-phase mixture.
Heat transfer to and from the vessel walls and heat and mass transfer at the steam-liquid inter-face were evaluated.
Steam separation from the liquid regions, condensate rainout from the steam region, and mixing betwcon the two liquid regions were modeled.
The deletion and subsequent recreation of the steam, mixture, and buf fer regions were analyzed.
At initiation, only two regions were consid-ered, corresponding to the steam and mixture regions.
However, during surge transients, a third region (called the buffer liquid region) was employed to represent the accumulation of the cooler primary system fluid as it flows into the bottom of the pressurizer.
Regions were deleted and created based on mass inventory criteria.
The numerical solution consisted of converting the partial dif ferential equations describing the system into sets of coupled ordinary dif ferential equations. These equations were then integrated contin-L uously in time. Additional details of the KPRZ model are documented in refer-ence 6.
Some of the more important assumptions used in the RV upper head void mitiga-tion analysis are as follows:
1.
The geometry rearrangement necessary to model the reactor vessel upper head as a pressurizer does not affect the physical processes occurring c
1
(
during the venting operation.
The pressurizer volume used in the KPRZ 3
nodel was the actual volume of the upper head region (690.7 f t ),
2.
The plant operator has throttling control of HPI during the venting oper-ation so as to maintain constant pressure in the RV.
The RCS pressure during venting was assumed to be 1020 psig for case 1 and 2500 pisg for case 2.
The case 1 pressure is arbitrary, but it is near the expected pressure at which voiding might occur were a cooldown to have occurred and drawn an RV upper head steam bubble.
The case 2 pressure is the RCS
[
system design pressure.
The steam is assumed to be saturated before venti ng.
[
3.
The venting process is assumed to be constant-volume.
For each cubic foot of steam that is displaced through the high point vent nozzle, one r
cubic foot of subcooled liquid flows from the RV into the upper head region (through the upper plenum cover plate).
L Babcock & Wilcox 2-7
. u.o...n......,
2.3.
Results of Analyses The results of these analyses show that the RV upper head coolant can be cooled down sufficiently during a 50F/hr cooldown of the core and RC loops with natural circulation to allow DHRS cut-in without flashing in the upper head.
The cooldown of the RV upper head without flashing is achieved by bleeding at least 3000 gallons of coolant from the RV head vent for each 50F
(
of RCS loop cooldown.
This upper head cooldown is accomplished within the time required for an RCS loop cooldown at 50F/hr by bleeding 50 gpm from the RV head vent during the RCS loop cooldown.
These results also show that the supplies of auxiliary feedwater requi red will not be increased by the requirement to avoid a steam bubble in the upper head if a 50 gpm bleed rate is maintained from the RV head vent.
The auxil-iary feedwater requirements are shown in Figure 7.
3 In the event the reactor vessel upper head is completely voided (690 ft ),
venting calculations show that it would take less than 60 minutes for WNP-1 to completely vent the upper head and only about 12 minutes for Bellefonte 1 and 2.
A natural circulation cooldown can also be accomplished without violating other pressure-temperature limits.
More detailed results are reported in the following sections.
2.3.1.
RV Upper Head Cooldown The results are shown in Figure 8 for both the 50 and 200 gpm bleed cases.
This shows both the temperature of the hottest upper head coolant node and the temperature of the circulating coolant nodes (hot leg temperature) as a function of time.
Water temperatures adjacent to the RV dome wall initially rose to 614F from 590F for the 50 gpm bleed case (and to 600F for the 200 gpm bleed case).
The circulating reactor coolant that flowed above the plenum cover weldment and down into the outer annulus cooled the node represented by the ring at the head base.
As the ring cooled, it, drew heat from the dome shell, thus cooling it and the water adjacent to it.
After several hours, the main heat transfer paths became (1) heat transferred to water near the top of the RV upper head from the RV dome, (2) heat transferred down the dome to the RV metal above the hot legs, (3) heat transferred from metal to water n
in the RV uppe r head at the periphery of the plenum cover, and most H
importantly, (4) heat removed by the bleeding coolant.
Babcock a,Wilcox 2-8
This bleed is replaced by coolant at the hot leg temperature.
In addition, this bleeding promotes mixing up through the entire upper head as shown in Figure 5.
The cooldown rate of the upper head coolant is 38F/hr for 50 gpn continuous bleed and 45F/hr for 200 gpm continuous bleed.
These results are both with an RCS loop cooldown of 50F/hr. Note that an increase in the total bleed of 4007, increases the RV upper head cool down rate less than 2 0*,.
)
Higher bleed rates are not that much more effective since the RV upper head cooldown would never quite match the RCS 1 cop cooldown rate.
Either of these cases cools the upper head coolant sufficiently within the time required for RCS loop cooldown at 50F/hr so that the DHRS can be cut in when the RCS hot leg is at 305F.
The 50 gpm bleed rate may be effectively obtained by venting a rate of 200 gpm for 15 minutes for each hour of cooldown or 100 gpm for 30 minutes, etc.
The bleed rates of 50 gpm and 200 gpm were chosen as being representative of the maximum design rates of WNP-1 and Bellefonte 1 and 2 respectively at nonnal system operating pressure.
The results indicate that a bleed rate less than 50 gpm may also provide acceptable RV upper head cooldown rates.
The temperature of the RCS loop at DHRS cut-in was assumed to be 305F.
The maximum RCS pressure at DHRS cut-in was assumed to be 420 psig to prevent lif ting the DH system relief valves.
To achieve DHRS cut-in without flashing in the RV upper head would then require that the RV upper head temperature be no higher than 449F.
(
The analysis assumed the only flow above the first layer of nodes over the plenum cover was the bleed flow which continued upward and out the RV head vent.
This first layer of nodes over the plenum cover cooled at about the same rate (50F/hr) as the RC loops as a result of this mixing.
Hence, if NC
{
flow (in addition to the bleed) were to ev. tend farther above the plenum cover, the upper head cooldown rate would be nearer to that of the RC loops.
However, flow velocities into the upper head region from below the plenum cover are expected to be less than two feet per second and could not cause appreciable penetration up into the dome region.
It should be pointed out
[
that the dome region is a large volume, approximately 690 ft.3, with a plenum cover to dome top distance of over 6 feet.
Even if twice the penetration had
(
been assumed, the results of the cooldown would not change dramatically.
Typical pressure-temperature limits are shown in Figure 9.
This case is
[
based on RCP trip concurrent with reactor trip.
Babcock & Wilcox 2-9
. me u.,
2.3.2.
RV Upper Head Void flitigation The tabulation below defines the time required to completely vent steam voids from the RV upper head and refill the region with subcooled liquid for vari-ous initial bubble sizes.
These analyses, using the KPRZ code, assumed that makeup is throttled to keep the pressure in the reactor vessel upper head nearly constant during the venting process.
Case 1 - 1020 PSIG: Time Required to Clear Reactor Yessel Upper Head Fraction of Upper Head Voided Time Required to Vent, Minutes At Start of Venting process Bellefonte 1 & 2 WNP-1 One-Fourth 2.9 14.5 One-Hal f 5.7 28.2 Three-Fou rths 8.5 42.3 Completely Voided 11.0 54.8 Another case was also evaluated for comparison.
Case 2 (at 2500 psig) might be encountered if the plant operators attempted to raise the pressure before opening the high point vent valve.
A completely voided RV upper head would not actually be encountered under these conditions due to compression of the steam bubble.
The tabulation below summarizes the time required to vent the RV upper head region for these conditions.
Case 2 - 2500 PSIG: Time Required to Clear Reactor Vessel Upper Head Fraction of Upper Head Voided After Time Required to Vent, Minutes Pressurizing Bubble to 2500 psig Bellefonte 1 & 2 WNP-1 One-Hal f 6.7 33.2 Completely Voided 12.0 59.8 These results are relatively insensitive to the RCS pressure.
Lower pres-sures will take slightly less time since less mass is required to be vented.
s Also, it will take about 50% of the time to vent the upper head if it is 50%
voided, etc.
1 2-10 Babcock & Wilcox
. m.o.,-n..,
Figure 1.
Reictor Vessel and Internals
(
w=wwww=
, Mi u p %y y i I
p,r
..x
+
==r 3
i 7n M==-
r m nannw x-e y.=
y wl1 9 0 0 M 0 0 0 e
(
.u -
i pH W
- ~""X..y !
n/
N ^
- ;c, ~L ff/s, i%
U\\
/U
= r w
C
- l e,
wI o-lg,l i
a
- n t <
mrnoeaan
- f
(
~3
= - 4,t.
..=,
=
5,b s
i.
i i
- f. :;;
a au.
. q.',I g'll
'5
'N gg
'mv -
i:
u en-W i
,i i!
y b! **,s i,
I d
%wame em g
l
'4 r'
9 i 4 -
h]
%l,,).-
f
.4 buer assarrarr waverw a vec _ d
+
i w
-a o u ruas
.l Ocraren m cuerr)
>1 l
l SY j
-n m.
- !I
.h [ " ""
3,wsrmama 3
y h
l i I A
s~ u u,o m Nl I
'3"*
- h Ad W I
'C.
py L.
,m 4 T: y !!'jLji!I j w.
3_L,,is i 3_i j p
i
_U !J_l I
i!_"
ti1 A ly a
~ %' :: d 4 g p K, 7
g I
i l
']
2-11 Babcock s.Wilcox
Figure 2.
Schematic Diagram of Upper Portion of Reactor Vessel and Internals RV HEAD VENT
. INSULATIDM CRDM NcZZLE T
~~
RV UPPER HEAD l
REGION
~
~
PLENUM C0VER
!--DRAIN HOLE (TYP OF 36)
RV HEAD %
(DONE) m
-00TER ANNULUS "H
(
F 8)
):0 00 Fd3e i
UPPER PLENUM r
g.
HOT d,,,, - -
I GL COLUMN
\\'!-
1 3
/~
wELDMENT k
]
/
(81 WITM R00 GUIDE
/
STRUCTURES, 8 EMPTY)
[/.
\\
PLENUM CYLINDER s g
/l 4
)
(
4
(
2-12 Babcock siWilcox
l Figure 3 Upper Plenum and RV Upper Head Mass Transfer Model s
10
)
o
((
l..
t
- y
\\
gfES
)
i O 'G,, ) @/
f jo e
is
,ff a
Aj w
v e
i w p
-,k 09 M
/
^
(/4 Fe=Je %ee 70 %ee 8 th Oescr4Dtion 0
2 4
3r9 Fuel Assesely upoor ene fitting (UEF) into column =eimment (G) 3 Bru Fuel Assemely UEF fato plemas open arte 4
5 From inside G to plenum ogen area thru louer emit ports 4
5
'Js full column welement 5
7 Antal flow in plenum open aree 5
15 3rv 3-inen wies in olenue cyl. to outlet ws21e 7
9 Astal flow in plenum coen area 7
11
- hru 18-and 20-inen %Ies in stone cyl. to plenum cyl. outer annulus 9
11 thru 20-inch %Ies in plenum cyl to plenus cyl.
outer annulus 9
15 3rv 20.tnen holes in 3 tnum cyl. to outlet nD121e 1
6 10 Bru G too caos to upper need 3
10 Bru emety G 's and Sh.ence holes in piene cover to outlet nozzle 10 11 Bru plenus cover ersta 41es to plenus cyl. outer annulus 10 15 Bru plenue cover drain moles to outlet nozzle 11 15 stenue cyl. outer annulus to outlet nozzle 10 1s r. uocer n... to w se nt eo 2-13 Babcock & Wilcox I
i Figure 4.
Heat Transfer Mode' 6
Ga K
d==g==&~~~T
=
f -f S:
1 i
t 4
~l y
U10 (D
}
w
- y
- %2 O ! y-I o 0
.js 7 ll
@! \\.
l
/
,/
- e..
x3
.iy h
j (Ci
/
t 1
1
/,
/
/
s
/
/r M4
/
i n'
4 We WP OescM otion I
(4) apper Head Coolant *o Lead $ crew 2
(a)
Plenum Cover to upoor plenum 3
(b)
Colem
- element *.o upoor Plenum 4
(b) 00 020 %221e 5
(a) tv %ed to Air Unoer Insulation 5
(c)
Insulatton to contatraent 7
(b) tv dead to Plenum Cover 4
(b)
From Motter to cooler Coolant %oes 9
(a) av gee to coolant in upoor aeee 10 (a) vooer %ed coolant to 71eaum Cover 11 (a)
Air Under Insulation to Insulation 12 (b)
From mtter to cooier woes of Column Welenent 13
- (b)
C20 %zzle to Contatrument 14 (a)
Insulation to contaffuent 15 (a) tV all to upper Pienus Coolant (a) Convection (b) Canouction (c) tediation
- C2O nozzl< was assumed to be 12CF at 3 *eet soove tv mead.
2-14 Babcock & Wilcox
Figure 5 Noding Diagram BLEED h
Jr 7
AXIS OF
- --f AMBIENT
~ ~ g~ ~ ~~'--l i
ROTATION l
l L-b I
i e
i Tn t
_U at _.-d-Li__I
[ INSULATION
[ T l
1I STAGNANT AIR
{-
d-Y-b- /
SPACE L__ _!_ _ _i at..;_ _n.__
J
_J L___ {_ __ 4_ __ _L at_ p ;< _ 9 1
L___ _l. _ _a _ _L_. __; __
J s,,,
V,(, DOME)
I I
l 1
l i
a thin l
J 1
q j
lRVUPPE t HEA0 l 1
f- -- Y'9%- r -
,N d
I PLENUM COVER l
lI I'
'l
\\
l Il
,\\!\\
_ j 1 I!
I bL
-l d-- h 4
RV ABOVE RV UPPlR PLENUN
')
I -
y HOT LEG g
f'f"T-t-T,--iY-h, h-h_
-l PLENUM L_l.
q'L M p."_4_Ly{d y Z CYLINDER I
'_. g!._ _m_. _ _.9 Il OUTER l_ _;_ ! ; _a _q Ji.;_lgt:
i 1
ANNutuS I
I I
II Il l
l__L_ql__MlI
_Lp__q iji i
I COOLANT FLOW
=u HOT LEG
)
Note: This noding diagram is only a one-half cross section of the upper reactor vessel and internals.
In the analysis, this noding dia-s gram is rotated about the left side vertical axis so that the RV and internals are modeled in three dimensions.
Each node thus becomes a ring.
2-15 Babcock s.Wilcox f
Figure 6 Pressurizer Model of RV Upper Head STEAM XNN MIXTURE LE) -@p@
g.
BUFFER REGION
/
- o O O H #9
'/
8
\\
w\\
l g(/g 7
<:lP,r
[,
v
/
l
\\
/
Al-
/r M
/
l
/
K4
{
l.
RAIN 00T 2.
BUBBLE RISE 3.
MASS & ENERGY TRANSFER AT INTERFACE 4.
CONDENSATION FROM DONE 5.
PRESSURE RELIEF 6.
BUBBLE RISE 7.
MIXING 8.
SURGE FLOW 9.
HEAT TRANSFER BETWEEN RV HEAD (DOME) AND BUFFER MIXTURE AND STEAM, (3 REGIONS) c 2-16 Babcock & Wilcox
Figure 7 Required AFW Inventory Vs Time to DHRS Cut-In
- 1. 2 1.0 e
S t
Oco 0.8 N 0.6 u.
O l
g 0.4 w
NOTE:
TO THIS CWVE, 40000 GAL MUST BE 0.2 ADDED TO ACCOUNT FOR SENSIBLE HEAT C00LDOWN TO 305'F 0.0 i
i a
i e
i e
O 1
2 3
4 5
6 7
TIME AFTER REACTOR TRIP-DAYS
(
(
2-17 Babcock & Wilcox
o Figure 8 Natural Circulation Cooldown Temperature Vs Time 600 590 t
558 MAX RV UPPER HEAD COOLANT 50 GPU BLEE0 500 RV UPPER HEAD
[ CUT IN TERPERATURE 3
FOR OMRS
=
450 h
BAI RV UPPER 4
EAD CGOLANT HOT LEE 200 GPu BLEED 400 TEMPERATURE 358 s
/
3g, O
I 2
3(
4 5
6 7
Ties (Hours)
(
2-18 Babcock & Wilcox
Figure 9 Typical Natural Circulation Cooldown -
Pressure-Temperature Diagram NOTES: I) THIS OIAGMAR 13 BASE 0 DN A 3000 GAL 8LEEJ FRod THE RV HEA0 VENT Fall EACH 50F NOT LEG MOLDOWN. C00LOOWN 13 CONTINUE 0 AND ELEE0 13 lNTERRUPTE0 OURING OEPitt$$UltilATION.
NOT LIMITS
- 2) NOT LEE C00LOOWN 13 LlulTE0 TO THl3#
l SiOE OF THE NOT LEE P T DIAGRAM.
l j
2000
/
I r -- - ' r-- -
p l
l l
/
/
I
/
/'
MINIMUM 1500 p-1--
su7C00 LING 3
7 p - --
MAnciN COLD LEc -
I
=
E MOT l
/
/
LEc f
i C00LDOWN j
& BLEED
/
g j
F-r-
r - --
~
C00LDOWN f I
l s DEPREs-l l
g SATURATION I
sualIATION 7
av f
f UPPER MEAD Y
I J
r-J i
r-500 1
I l
i
==-
DNR$
/
OPERATION minimum RV Upper Head Subc00 ling Margin (if temperature indication available)
I I
100 200 300 400 500 E,0 700 Temperature (*F)
(
(
2-19 Babcock & Wilcox
3.
BASIS FOR OPERATIONAL GUIDELINES FOR NATURAL CIRCULATION C00LD0WN 3.1.
Objective This section provides the basis for operational guidelines for a natural circulation cooldown.
It is intended to demonstrate the application of the methods developed in Section 2.
This demonstration will provide the pro-cedure writer with the basis for writing operational guidelines and, ulti-mately, operational procedures dealing with a natural circulation cooldown in the Bellefonte 1 and 2 and WNP-1 plants.
The emphasis is on operational in-fonnation for avoiding a steam bubble in the RV upper head during NC cooldown and mitigating a bubble if one should occur.
3.2.
Background
Following a reac tor trip, RCP forced circu *' ion and heat transfer to the steam generators is the preferred method of residual and sensible heat re-moval whenever plant temperatures and pressure are above DHRS cut-in condi-tions.
However, there are conditions in which the RCPs will not be avail-able.
The two most likely reasons are loss of power to the RCP motors and loss of services (cooling water and seal injection) to the RCPs.
Experience has shown that loss of power to the RCPs for an extended time is a very unlikely occurrence. Loss of services for an extended time is also unlikely, but if needed, the RC pumps can be operated for short intermittent intervals if all services are lost and can be operated continuously if only seal injec-tion is lost.
Component elevations on B&W plants are such that satisfactory natural circulation decay heat and system-stored sensible heat removal can be obtained by density differences between the thermal centers in the core and steam generators.
This NC flow is governed by decay heat, component elevations, primary-to-sec-ondary heat transfer, loop flow resistance, and the presence of voids.
(
Babcock & Wilcox 3-1 I
The loss of RC puraps coupled with a need to cool down is an even more remote event than the scenarios above, but one which must be taken into account.
A basic question must be answered before initiating a NC cooldown:
Is the stable control of the plant in greater jeopardy at hot shutdown (as compared to cold shutdown) now or is it projected to be in the future?
If the answer is no, then no NC cooldown is recommended since it is much safer to remain at hot shutdown until RCPs can be started.
However, if a NC cooldown and depressurization is chosen, the operator must be aware that the RV upper head will not inherently cool down at the same rate as the RCS loop.
The hot stagnant fluid in the RV upper head can flash as depressurization occurs, even though the loop is subcooled.
Thu s, RCS loop subcooling margin alone is not a good indication of RV upper head condi-tions during NC cooldown. An RV upper head temperature indicator or the ini-tial RV upper head temperature and the amount of bleed from the RV head vent during RCS cooldown is also required to determine RV upper head conditions.
A steam bubble in the RV upper head is a control problem, but it is not in itsel f a safety problem.
It can become a safety problem if it restricts a mandatory RCS cooldown and depressurization.
Therefore, actions should be taken to avoid a steam bubble in the RV upper head during a NC cooldown.
The initial RV upper head temperature is about equal to the hot leg tempera-ture duri ng forced-flow condi tions in the RCS.
Operating RCPs after a reactor trip will reduce the RV upper head temperature from the full-power
/
hot leg temperature towards the lower post-trip hot leg temperature.
The amount of upper head temperature reduction will depend on the length of RCP operation af ter the reactor trip and the post-trip decrease in hot leg temper-a ture.
RCP operation for several minutes beyond the reactor trip is effec-tive in reducing the RV upper head temperature and thus reduces the RCS pres-sure for voiding in the upper head during NC cooldown.
The operator must be careful not to exceed the maximum RCP operating time if LOCA conditions occur which require RCP trip.
This RCP operation after reactor trip does not relieve the requirement to cool down and depressurize in accordance with the NC cooldown P-T curve.
1 3-2 hM&hx e m em 2
The basis for operational guidelines is provided within the context of ATOG plant control functions.
This approach is consistent with the ATOG guide-lines being developed.
3.3.
Control Functions for Natural Circulation Cooldown Five main AT0G control functions must be addressed throughout a plant cool-down using natural circulation flow in the RCS. The control functions are:
1.
Reactivi ty 2.
Reactor coolant system inventory 3.
Reactor coolant system pressure 4.
Steam generator inventory 5.
Steam generator pressure.
i Incorrect operator response in any of these five control functions can lead to interruption of NC core cooling. This report provides guidance throughout the plant boration, NC cooldown and depressurization to DHRS cut-in condi-tions.
The general sequence of events is fi rst discussed, followed by a separate discussion on each of the AT0G control functions.
Figure 10 sumar-izes the use of the control functions for NC cooldown.
3.3.1.
General Sequence of Events This section discusses the general sequence of events upon loss of forced RCS flow including the initiation of NC cooldown.
Natural circulation cooldown can be complicated by other equipment failures; however, the discussion in this report assumes that 1oss of reactor coolant pumped flow is the only failure unless otherwise specified.
[
The general sequence of events upon loss of forced RCS flow should be:
1.
Immediate actions required at or upon reactor trip.
2.
Verify the status of all vital systems.
3.
Monitor the four AT0G symptoms:
RCS loop subcooling, overheating, overcooling and the special symptom of SGTR.
Respond to these symptoms by use of the five control functions.
{
(
3-3 Babcock & Wilcox
.m.o.,
n
4.
Verify that natural circulation flow is established and maintained in the RCS by observing the indications below (Plant responses to a paraneter change will not be accomplished for several minutes due to increased loop cycle times),
a.
Loop AT, (Th - Tc), less than nonnal full-power AT (with both steam generators operating).
b.
Cold leg temperatures constant or decreasing.
c.
Cold leg temperatures approximately equal to steam generator secondary temperature.
d.
Hot leg temperature constant or decreasing, e.
Th and Tc decrease as steam generator Tsat is decreased.
5.
If natural circulation degradation is suspected, enhance natural circulation flow by a.
Increasing turbine bypass or atmospheric steam dump flow to reduce RCS temperatures.
b.
Increasing RCS pressure by operating makeup pumps or with pressurizer heaters if available.
c.
Raising secondary water level.
d.
Establishing adequate primary water inventory with no voids.
e.
Establishing adequate subcooled margin.
6.
Bring the plant under control with the five control functions.
Stabilize the plant if possible.
7.
Try to restart at least one RCP in each loop, if possible, whenever a.
At least one steam generator is removing heat from the RCS and, b.
Proper RCS pressure-tenperature conditions are established and, c.
RCP support system services have been restored.
8.
Determine whether the plant should be cooled down or kept in hot standby.
This is a critical decision and the answer will depend
~
upon the exact set of circumstances.
However, this decision is usually not irrevocable and should be re-evaluated from time to time.
l Some of the reasons for a cooldown 'are a LOCA (including a steam
(
generator tube rupture), a loss of power which is expected to be extended or a loss of additional safety equipment is anticipated or 7
iminent.
3-4 Babcock & Wilcox
. me a
9.
Detemine how the plant can be cooled down.
Check all systems required to determine if a NC cooldown can be accomplished.
Engage backup systems as required. Prepare a status list of systems.
- 10. Determine whether a steam bubble exists in the RV upper head.
If a bubble exists, vent it in accordance with procedures (Section 3.3.3).
- 11. Commence NC cooldown.
Use the five AT0G control functions (discussed separately in the remainder of this section) to control the plant.
- 12. Perform follow-up actions as required.
Continue follow-up actions required by the specific conditions in parallel with plant cooldown procedures.
Applicable emergency procedure requirements may supersede parts of the guidance suggested in the following sections.
3.3.2.
Reactivity Control Control rods accomplish the prompt shutdown; however, it is desirable to start the chemical addition system as early as possible.
The initial objec-tive is to borate the RCS, if necessary, to achieve hot shutdown regtirements before cooling below hot shutdown.
Then commence boration for cold shutdown in accordance with the Technical Speci fications.
Reactivity changes that occur upon shutdown and cooldown, (particularly xenon and temperature) should be anticipated in advance through early boration.
The minimum required reactor shutdown margin for the present operating mode must be attained and maintained in accordance with the Technical Specifica-tions prior to NC cooldown.
If the letdown system is not operable, boraticn may be concurrent with the cooldown if the contraction volume is compensated by the boron addition to ensure the reactor is at least 1% aK/K shutdown throughout the cooldown.
The reduced flowrates experi enced during natural ci rculation increase the time required to achieve boron equilibrium throughout the system.
The nat-ural circulation flow rate in B&W Rairad Loop plants is nomally more than 3%
of full forced flow.
This flow rate t roduces adequate boron mixing through-out the RCS.
However, the pressurizer and RV upper head may be at a boron concentration significantly less than the RCS loop boron concentration.
To campensate, increase the RCS boron concentration higher than the required RCS cold shutdown concentration to account for a potential dilution from the pres-surizer or upper head water. Thus, if a outsurge occurs, mixing of the water 3-5 Babcock s.Wilcox
. ma n..,
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _______ _. __________ w
fran the upper head or pressurizer with the RCS loop water will not dilute the boron concentration below the cold shutdown concentration requirement.
Until final mixing of all regions has taken place, sampling from all flowing regions of the RCS should give a boron content that is higher than required.
Sample locations and frequency should be increased over those used during forced coolant circulation.
RCS makeup should also be horated at least as I
high as the RCS itself.
3.3.3.
RCS Inventory Control The main obj ective of reactor coolant system inventory control during NC cooldown and depressurization is to maintain a level in the pressurizer and the remainder of the RCS subcooled during the cooldown.
If possible, the pressurizer level should be maintained between 100 and 195 inches with makeup and letdown.
RCS inventory problems should be corrected before depressurization or while depressurization is interrupted unless indications show that this delay will worsen conditions later. The following sections provide infonnation on bleeding and venting from the RV head vent.
3.3.3.1.
Bleeding
(
The RV high point vent should be used intennittently to bleed coolant from the upper head during the cooldown.
This bleeding enhances the cooldown in
(
the RV upper head and greatly reduces the potential for forming voids in the upper head during depressurization; however, bleeding in the prescribed man-ner will approximately double the RCS makeup requirements during the cool-f down.
This extra makeup must be taken into account in the overall operation of the plant as well as RCS inventory control.
f The general procedure for bleeding is:
1.
Put an HPI pump in standby and secure letdown.
2.
Energize pressurizer heaters as required to maintain RCS pressure.
3.
Open the RV hdad vent as required to keep the RV upper head subcooled if temperature indication is available, otherwise bleed to achieve approximately 3000 gallons of bleed during each 50F of cool down.
This rate of bleed will overpressurize the quench tank of Bellefonte except for slow cooldowns (N 10F/hr).
The quench tank shoul d be j
monitored and the bleeding further limited if conditions pennit.
Babcock & Wilcox 3-6
..,,,, u.,
4.
Balance makeup and bleed to maintain pressurizer level control.
j 5.
Secure bleed during designated RCS depressurization phases (after each RCS cooldown of $50F) or in case of any unexplained increase in pressurizer level, and monitor for a RV head bubble by observing the RV head level indication and unexplained pressurizer level i ncreases.
(NOTE:
Pressurizer level may be expected to increase some when bleed is secured).
6.
Continue the RCS cooldown during bleeding by controlling the turbine bypass valve (TBV) setpoints (if available) or MADY setpoints.
3.3.3.2.
Symptoms of RV Void Formation RCS pressure conditions may result in formation of a steam void in the upper head. Some symptoms of void formation are:
1.
Pressurizer level increases significantly more than expected while operating auxiliary spray.
2.
Pressurizer level decreases while operating makeup pumps.
3.
Letdown flow is unexpectedly greater than makeup flow if the pressurizer level control system is in automatic.
4.
Void level formation is indicated by the reactor vessel level monitor when the RV head vent is not in use.
l l
5.
RV upper head temperature indicates a loss of subcooling if a temperature indication is available.
3.3.3.3.
RV Upper Head Venting The operator should take the following general actions if upper head voiding is indicated:
l 1.
Stop the depressurization.
2.
Start and put second HPI on standby (if not already on).
j 3.
Take manual control of RCS makeup and secure letdown.
4.
Increase the RCS loop subcooling margin with HPI/ makeup (or l
pressurizer heaters if available) to accomodate the subsequent drop in RCS pressure.
Babcock & Wilcox 3-7
5.
Open the reactor upper head vent to eliminate the upper head void.
The entire upper head may be vented in approximately 12 minutes on the Bellefonte plant and approximately 55 minutes on WNP-1.
Stop venting and monitor RV head level periodically.
Pressurizer level should be closely monitored at this. time.
If pressurizer level goes below 100 inches or above 400 inches while venting, the venting must be stopped until pressurizer level is corrected.
6.
Maintain RCS pressure during venting by makeup (or HPI) and.
pressurizer heaters (if available).
Some pressure fluctuations are to be expected under these conditions, particularly when fi rst opening the vent.
7.
Continue the RCS cooldown during venting (to help maintain subcooling margin and NC) by controlling the TBV (if avail able) or MADY setpoi nts.
3.3.4.
RCS Pressure Control The main obj ective of RCS pressure control during a natural circulation cooldown is to maintain enough pressure so the RCS remains subcooled and at as low a pressure as practical to minimize RCS leaks or other conditions which may have made NC cooldown mandatory.
This is accomplished by careful coordination of RV head bleeding, RCS letdown, pressurizer heaters (if avail-able) and auxiliary spray until criteria-for depressurization. have been satisfied.
No void will fonn until the RCS pressure is decreased below the RV head saturation pressure, regardless of the RCS cool down rate.
RCS pressure control for bleeding and steam. venting. has been covered in the previous section.
.This section discusses RCS depressurization, loss of pressurizer heaters.and loss of auxiliary spray.
3.3.4.1.
RCS Depressurization Before starting the RCS cooldown, the RCS may be depressurized to the minimum subcooling margin.
This minimum subcooling margin is based on the tempera-tures of both the hot leg and the RV upper head.
If - the RCPs tripped concur-rently with the reactor trip (from 100% ' power) and then coasted. down, the 1
minimum initial RCS pressure is approximately 1800 psig.
This Jpressure is saturation pressure for 621F.
This temperature is based.on an initial RV '
upper head temperature (af ter RCP coastdown) of. 590F' with margin added.for ~
c subsequent heatup by the RV dome and RV upper; head'subcooling. -If the RCPs Babcock &Wilcox 3-8
. m.o
continue to operate for several ($ 10) minutes, then the RV dome heatup can be neglected and the initial RV upper head temperature is the hot leg tempera-tu re.
If RV upper head temperature indication is available, then the RCS pressure can be reduced to provide a minimum subcooling margin for the RV head.
After each RCS cooldown of approximately 50F or less, the RCS should be depressurized to the minimum subcooling margin as shown in Figure 9.
The general procedure for RCS depressurization is:
1.
Put an HPI pump on standby.
2.
Close the RV head vent and monitor RV upper head level indication for voids.
3.
Maintain letdown and makeup control in automatic control.
4.
Operate the auxiliary spray to depressurize the RCS down to the minimun subcooling nargin.
5.
Stop the auxiliary spray and go to the procedure for RV head venting if voids are detected in the upper head.
6.
The RCS cool down may continue during depressurization to hel p maintain subcooling and avoid interruption of natural circulation.
The use of pressurizer auxiliary spray should be minimized when the tempera-ture dif ferential between the spray water and the pressurizer is greater than 250F to minimize the thermal stress accumulation factors in the spray nozzle and the Tee-connection between the auxiliary spray and main spray lines.
When auxiliary spray is used with this temperature difference, ensure the 1 gpm bypass spray flow is maintained continuously.
3.3.4.2.
Loss of Pressurizer Heaters If pressurizer heaters are inoperative or insufficient capacity is available to maintain pressurizer pressure as required, follow the guidance provided below for pressure control throughout the plant cooldown:
Maintain the pressurizer bubble as long as possible.
The following actions will help slow the loss of RCS subcooled margin:
1.
Stop any isolable leakage from the pressurizer.
2.
Minimize ese of pressurizer auxiliary spray.
3-9 Babcock & Wilcox
. mo.,m c.,
3.
Stop all pressurizer sampling if not needed.
4.
Minimize pressurizer level fluctuations by careful control of letdown and makeup.
5.
Until RCS temperatures are cooled down to the shutdown cooling system naximum allowed entrance temperature, maintain an RCS cooldown rate (if possible) greater than or equal to the pressurizer cooldown rate caused by the pressurizer ambient heat loss.
If the RCS hot leg minimum subcooled margin is lost with pressurizer level decreasing, initiate the HPI system.
Terminate HPI pump operation only when RCS subcooled margin is greater than the minimum and pressurizer level is indicated and can be maintained.
3.3.4.3.
Loss of Auxiliary Spray If the auxiliary spray is inoperative:
1.
Ensure pressurizer heaters are deenenjized when not needed.
2.
Determine when a plant cooldown must be initiated based on existing plant conditions and requi rements.
If plant conditions pe mi t,
stay at hot standby conditions until main or auxiliary spray flow can be restored.
If a cooldown is necessary without auxiliary spray:
1.
Allow the pressurizer to cool due to ambient heat loss or operate the pressurizer vent to reduce pressurizer pressure as needed.
2.
As a last choice, operate the power operated relief valve (PORV) to reduce pressurizer pressure as needed.
3.3.5.
Steam Generator Fressure Control Maintain RCS temperature at hot shutdown by operating the TBVs (or atmos-pheric dump valves) until ready to begin the RCS cooldown.
After the prepara tory actic's are completed, the RCS cooldown can be initiated and controlled by regulating the steam flow through the TBVs or atmospheric dump valves.
The operator should maintain the maximum possible RCS cooldown rate without exceeding the Technical Specification limits.
A large temperature 1
dif ference between the RV head and the reactor coolant will provide a large thermal gradient and a greater heat transfer rate.
The RCS should be cooled down as far as possible with the steam generators to provide the largest possible thermal gradient.
s 3-10 8abcock a Wilcox
. m e.n
4 L
The operator should maintain the maximum possible RCS cooldown rate without exceeding the Technical Specification limits.
The Technical Specification pressure / temperature NDT limit curve is based on the cold leg rather than the hot leg temperature during NC flow conditions.
3.3.6.
Steam Generator Inventory Control The event that led to the nat"ral circulation condition may also render the nomal heat sink and the itvantorv replacement system unavailable (e.'., loss of off-site power).
Therefore, the ataospheric steam dump system may becone the only method of heat removal and will deplete available inventory.
Condensate availabili1:y may limit the amount of time a plant can maintain hot shutdown conditions before starting a NC ccoidown.
The operator must iditially evaluate the availability of the condensate and condensate supply systems and detennine the total time available before the shutdown cooling system must be initiated.
If the amount of condensate available appears to be inadequate or marginal, the plant cooldown should be started insnediately to avoid running out of condensate before the shutdown cooling system can be placed into operation.
Figure 7 may be used to determine the amount of condensate which will be' required.
The operator must ensure that the secondary water inventory is being nain-tained by feedwater flow to the steam generator (s).
The SG level should be maintained at the level required for NC cooldown by the main or auxiliary feedwater systems.
The AFW controls will automatically maintain SG levels with a level rate contro;
?f adequate RCS coolingLii not being: achieved due to insufficient secondary water, 'an increase in cold leg temperature (T )
c will be the first indication, causing an abnomal and misleading decrease in Loop aT.
This will persist for some time due to the slow loop transport times.
Eventually the hot leg temperature (T ) will' also increase.
During h
natu ral circulation with an adequate secondary water inventory, t T and the e
secondary temperature are approximately equal.
Tc increasing' (followed later by Th increasing) without a corresponding change in isteam generator secondary pressure / temperature signals 'a lack of sufficient secoildary water.
s
\\
~.t
,j c
l Babcock & Wilcox '
3-11 s -
. me. a -,
- I __ _
_______.__1____________
l l
4 figure 10 NATJRAL CIRCULAtl04 C00LDCom C0% TROL FUNCilDh3 REACTOR TRIP LOS$ OF RCP'$
I l
REACTOR TRIP luetDIATE ACTIONS VITAL ST5TE8 STATU1 vi#lFICAfl0%
4ATURAL CIRCutail0N FL0s WERIFIED E0NITCR SL8Cc3 TING UvER*EATitG DVERCOOttNG 840 SG ftBE 'UPfuRE 451 StitlLIIEC Can a RCP
'II POST TRIP SE CPERAtl0h5
$ TARTED, 40 TES 0
to PERFORu g8T uAlqTAIN MOT STANDev CIRC C00LC0em NCT $wuTD08h Comof flon!
C mflhut l
APPLICatLE EvEti RCI 14vf4TCRY RCS PRES $4RE SG m E4 TORY
$G PREt$yRE MINI LLOW WP CONTROL CONTROL CONTROL CONTROL -
EVENT p
ACTIONS I
l r
3 Babcock & Wilcox
/
{
. m.w a
. 1
4.
OPERATOR TRAINIt0 AND PROCEDURE DEVELOPMEllT (By Owner) u
[
r L
(
[
r O
1 L
d 4-1 Babcock & Wilcox
. m e....a s
REFERENCES 1.
NRC letter to all licensees of operating pressurized water nuclear power reactors and applicants for operating licenses (except for St. Lucie, Unit No.1),
Subject:
Natural Circulation Cooldown (Generic Letter No.
81-21), May 5, 1981.
I 2.
NSAC16, 19 02, " Analysis and Evaluation of St. Lucie Unit 1 Natu ral Circulation Cooldown", December 1980.
3.
NPGD-TM-574, " NATURAL - Hybrid Natural Circulation Code".
4.
BAW-10072A, " SAVER, Digital Computer Code to Detemine Pressure Drops".
Septembe r,1975.
5.
J. P. Holman, " Heat Transfer", 4th Edition,1976, New York, McGraw-Hill, Section 4-6, Transient Numerical Method.
f 6.
NP GD-TM-568, "KPRZ, Digital Code for the Simulation of Transient Pressurizer Perfomance", March,1982.
[
[
[
A-1 Babcock & Wilcox
. me n -
P l
__.... - -. _ _ -