ML19211D124
| ML19211D124 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah  |
| Issue date: | 01/11/1980 |
| From: | Mills L TENNESSEE VALLEY AUTHORITY |
| To: | Rubenstein L Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8001160510 | |
| Download: ML19211D124 (17) | |
Text
.
TENNESSEE VALLEY AUTHOrtlTY CH ATTANOOG A, TENNESSEE 374o1 400 Chestnut Street Tower II January 11, 1980 Director of Nuclear Reactor Regulation Attention:
Mr. L. S. Rubenstein, Acting Chief Light Water Reactors Branch No. 4 Division of Project Management U.S. Nuclear Regulatory Commission Washington, DC 20555
Dear Mr. Rubenstein:
In the Matter of the Application of
)
Docket Nos. 50-327 Tennessee Valley Authority
)
50-328 Enclosed are TVA's revised responses to the questions on water level measurement systems inside containment transmitted by your letter to H. G. Parris dated October 5, 1979.
These revisions have been made as a result of our telephone conversation with members of the Instrument and Control Systems Branch staff on December 17, 1979. Also enclosed is the additional information requested in the December 7, 1979, tele-phone conversation on steam flow instrumentation and the Auxiliary Feedwater System. The Sequoyah Nuclear Plant Final Safety Analysis Report will be revised to be consistent with the enclosed responses.
The enclosed responses are based on modifications to the steam generator reference leg design (i.e., insulation) to be completed before initial criticality.
Very truly yours, TENNESSEE VALLEY AUTHORITY
'/ -!I i,i i.< ' l t
- \\
J' 3,
f L. M. Mills, Manager Nuclear Regulation and Safety Enclosures i75(074 t
Of ;
i 56 Y An Ecud Oc;:crtunity Em;:loyer g
yg l
ENCLOSURE RESPONSE TO L. S. RUBENSTEIN'S LETTER TO H. G. PARRIS DATED OCTOBER 5, 1979 REVISED PER DECEdbER 17, 1979, TELEPHONE CONVERSATION WATER LEVEL MEASUREMENT SYSTEMS INSIDE CONTAINMENT
~
1.
Describe the liquid level measuring systems within containment that are used to initiate safety actions or are used to provide post-accident monitoring information.
Provide a description of the type of reference leg used, i.e., open column or seal refer-ence leg.
Response
Two types of level measurement systems used inside containment are described below along with the particular application:
A.
An open column reference leg is used for steam generator (SG) level measurement.
The instrument is connected to the SG liquid by a condensate chamber at the upper tap.
The liquid in the reference leg will be at essentially ambient temperature.
Steam Generator Narrow Range Water Level Safety Functions lutoine trip and feedwater isolation on high-high steam generator water level Reactor trip on low steam generator water level in coinci-dence with steam flow - feed flow mismatch Reactor trip on low-low steam generator water level Auxiliary feedwater pump initiation on low-low steam genera-tor water level Post-accident monitoring function Steam Generator Wide Range Water Level Safety Function Post-accident monitoring function B.
A sealed reference leg is used for pressurizer level measurement.
The instrument uses a seal liquid which is not part of the pressurizer liquid and has a physical barrier (a diaphragm) that transmits the hydraulic pressure from the liquid (steam) to the seal liquid.
The diaphragm is located a sufficient distance from the condensate chamber to be at ambient temperature.
175 075
as.
The pressurizer uses a sealed reference leg.
The vertical portion of the sealed sense line is outside the cavity that shields the pressurizer and therefore is not subject to short-term heatup.
Since Sequoyah is an ice condenser plant, it does not experience peak temperatures in the upper compartment as high as non-ice condenser plants. The long-term heatup is the only concern.
The bellows sensor is 12 inches below the condensate pot and even with the upper tap.
This means if the sense line " boils" dry, there is a 4.3 percent error introduced because of reference leg height. Refer to figures 1 and 2.
Pressurizer Water Level Safety Function Reactor trip on high water level Post-accident monitoring function 175 076
2.
Provide an evaluation of the ef fect of post-accident ambient temperatures on the indicated water level to determine the change in indicated level relative to actual water level. This evaluation must include other sources of error including the effects of varying fluid pressure and flashing of reference leg to steam on the water level measurements.
Recponse A.
Re ference Leg Heatup High energy line breaks inside containment can result in heatup o f level measurement re ference legs.
Increased re ference leg water column temperature will result in a decrease o f the water column density with a consequent apparent increase in the indi-cated steam generator water level (i.e., apparent level exceed-ing actual level).
The following formula can be used to calculate the magnituc'e of this bias:
L L, cal - ol H
P E
77- (p 7, cal - o, cal) g where:
level error due to reference leg heatup, as a E
=
fraction o f level span, level span = vertical distance between narrow H
=
range taps on steam generator, height o f reference leg, HL
=
maximum vertical distance from lower tap to
=
water level in condensing pot on upper tap.
This must be determined for the limiting instru-ment connec.tions, water density at containment temperature and pt, cal
=
steam generator or pressurizer pressure for which the level indication system was calib-r a ted.
If this in formation is not available, an upper-bound density (lower-bound temperature) must be assumed.
water density in reference leg at the time of
=
ot interest (p r, cal - og, cal) = di fference between saturated water density and dry saturated steam density at the steam generator or pressurizer pressure for hich the level indication system was calib-w rated.
An upper-bound pressure must be assumed.
175/077
This procedure is based on the assumption that the tubing from the upper and lower taps, below the elevation o f the lower tap, have the same temperature at all times.
B.
Reference leg Boiling In addition to the above reference leg density change under subcooled conditions, boiling could conceivably occur in the reference leg following depressurization of any steam generator with high containment temperature.
This combination of condi-tions could only occur following a steamline or feedline rupture inside containment.
If such boiling were to occur, it could cause a major bias in the indicated level for a short time period, in the extreme case indicating 100 percent level when the vessel is actually empty.
C.
Coolan't Density Changes A bias in indicated water level may also be introduced by changes in pressurizer or steam generator pressure, due to changes in the density of the saturated water and steam within those vessels. While prediction o f the e f fects o f rapid depres-surization requires complex calculations for eacn speci fic case, the bias which would exist at low power under quiescent condi-tions can be calculated directly, using the following formula:
, cal - p[ - o, cal + a H
p og-p q)
- II ng, cal - o, cal g
I
-1}
" II- (
g, cal - o, cal p
g g
where:
level error due to density changes in both the ves-E
=
sel and the reference leg, as a fraction of level
- span, true water level in the vessel, above the lower L
=
level tap, saturated water density at the pressure of interest, or
=
og dry saturated steam density at the pressure of
=
interest, and other symbols have the same meanings as in Section A above.
175f078
D.
Seal Bellows Pressure For the pressurizer water level measurement sys t em, the spring force of the sealing bellows, located between the condensing pot and the top of the reference leg, of fers a further effect of reference leg fluid expansion on indicated water level. As the fluid in the reference leg, seal bellows, and sensor bellows expands due to a rise in containment temperature, the seal bel-lows must expand to accomodate the increased fluid volume.
This expansion is resisted by the force of the bellows acting as a spring under tension. The resultant increase in pressure in the bellows above that outside the bellows, i.e.
in the top of the pressurizer, is felt in the sensor bellows and results in a reduction in the indicated water level relative the true water level in the pressurizer. This effect is ir. the opposite direction from the simultaneous increase in indicated level (relative to true level) caused by the ef fect of reduced density and thus gravitational head in the vertical reference leg tubing on heatup - the ef fect which has been discussed at such length already.
Bounding calculations assuming a wide range of different temperatures in the upper and lower containment compartments indicate that the combination of these two effects due to hcatup above the normal operating temperature i s always in the direction of indicated level being grea-ter than actual level; that is to say the gra-vity effect always is greater than the bellows spring effect.
In the extreme extension condi-tion, with all components in both compartments assumed to be at 340 F, the extension of the bellows had not yet reached the restraints with-in the housing, the burst strength of the bel-lows had not been approached, and the level error due to the spring force of the bellows compensated for less than one-fourth of that due the gravity effect of the heatup. The compen-sating effect of the bellows spring force has been conservatively neglected in calculating the total level errors.
E. ErFFE8"c5 L8G H ggwD ULO C ( At r.o t Arte '(.
fbs r 4cc ior ni Amaie vr Tempranroar 2 60" E= N
( f a.c at - 9.
\\
L H
(_e
. g e,,,
cat g
H. 14 4
H = 144.4
t Recu rot.ll 52 H
D 3
Tem P c 120 F 9 - 57.924 6 /g(5 t
(ffcac 9yd = (46.'79r */n - 2.oS4 '%g5 )
s 44.70)'blH'
=
D it o M C AL 92C pssy RT 5*37'F E=
144.4 (al. 7/32 - 77. 92 + I 144 44*101
)
Er
- 0. 0 8 T*
% ERROR = 8. T 2 E x 10 0
=
):~,
[co t. A N T D Cwstry Cnneses Cacouvoc
- E= S (Lcat-9, acht&Va L
Qg - qq H _
Pe cu.-gtm. )
+
-l H
et ent - gc.% t.
H: )+4 4 t
8: 14 4 etent - ci.mz S/W (R; tat-9gtAt)=94.701 ib/ce3 - ( %. 7.?5 - 2. 09 4 )
L = ( ASS o m i 28. S"]
() : o.095~ %lge s Rg= 58, 30(, Ib/ft1 I" *i [ ( l-713 Z. - S 7. 9 2.4 0 9 4 + 0. O S T E:
+
144
(
94 701
- 28. [_ / f 8. 5o 4 - c.o 9 5' _ g \\
m
(
49.Toi I
175$080 e-e. 0 4 0 + o.o c o E = 0.1 O O 7e E AIO P-IO7e Ex \\O0
=
f Atsam G W psi) i.e s. G.
3.
Provide an analysis of the impact. that the level measurement errors in control and protection systems (question 2) have on the as,sumptions used in the plant transient and accident analysis. This should include a review of all safety and contrn1 getpoints derived from level signals to verify that the setpoi'nts will initiate the action required by the plant safety analyses throughout the range of ambient temperatures encountered by the instrumentation, including accident temperatures.
If this analysis demonstrates that level measurement errors ara. greater than assumed in the safety analysis, address the corrective action to be taken. The corrective actions considered should include design changes that could be made to ensure that containment temperature effects are automatically accounted for. These measures may include setpoint changes as an acceptable corrective action for the short term. However, some form of temperature compensation or modification to eliminate or reduce temperature errors should be investigated as a long term solution.
Rcsponse A.
Steam Generator Narrow Range Water Level Trip Setpoints The only high-energy line rupture within the containment for which 'the steam generator water level provides the crimary trip function is a reedline rupture.
For such a case the laa or low-low water level trip must be actuated when the pressure di fference between the narrow range level taps corresponds to a zero-level value.
Thus the trip setpoints nust be at or above the value that would be indicated at zero true level.
Because large steam generator pressure changes are not expected before trip, only the re ference leg heatup e ffects need be considered, and not the effects of system pressure changes.
The determination of the steam generator low-low level trip for Sequoyah is as follows:
Bottom of span (percent) 0 Normal channel accuracy (percent)
+5 Post accident transmitter error (percent)
-+To Insula ted re ference leg e f fects (post accident heatup) (percent)
+3, -0 TOTAL ERROR, percent o f span
+18, -15 Trip setpoint
~18 percent o f span Allowable setpoint 17 percent o f span
]JC Q@]
The value o f +3 percent, -0 percent used for re ference leg effects was obtained from the formula in the response to question 2 part A, assuming that the reference leg temperature does not exceed 340 F before reaching the High containment pressure setpoint.
i
As the steam generator narrow range reference legs will be-insulated and bounding temperatures are available, the formula in the response to question 2 part A has been used for each section of vertical length to which a discreet temperature can be assigned.
The total error allowance of 18% of span can be regarded as including 1% for rack module drift (i.e.
"setpoint drift") and 17% allowance for other errors in and uncertainties. Thus the setpoint is to be set at 18% and periodically checked.
If the set-point drifts to 17%, then the drift is as assumed for the worst case, and 17% still remains to accomo-date all other errors and uncertainties.
At this point the Technical Specifications require resetting back to 18%. Thus the technical specifications assure that the drif t error assumed is not exceeded, and that the margin reserved for other errors and uncertainties is preserved, The setpoint revisions recommended to correct for the reference leg heatup errors comply with Regula-tory Guide 1.105, Revision 1, November 1976 as follows:
1.
The neu setpoint is es tablished with suf ficient margin between that value assumed in the safety analyses and the nominal trip setpoints to allow for (a) the inaccuracy of the instrument; (b) uncertainties in the calibration; and (c) the instrument drift that could occur during the interval between calibrations.
These components are all included in the error allowances enumerated.
2.
The new setpoint is established in that portion of the instrument span (i.e. O to 100%) which ensures that the accuracy required is main-tained.
Instruments are calibrated so as to ensure the required accuracy at the se tpoint.
3.
The range selected for the ins trumentati on en-compasses the expected operating range of the process variable being monitored (i.e. O to 100%) to the extent that saturation does not negate the required action of the instrument.
4.
The accuracy of the setpoint is equal to or better than the accuracy assumed in the safety analysis considering the ambient tempe ratur e changes, vibration, and other environmental conditions. The ins truments have been designed so as not to anneal, stress relieves, or work 17s 082 harden under design conditions to the extent that they will not maintain the required accu-racy. Design verification of these ins truments has been demonstrated as part of the instrument qualification program discussed in IEEE 323-1971.
These setpoint revisions are conservative in view of the extremely low probability of the simultaneous occurance of all error conponents enumerated at their extreme valves in an adverse direction.
The above setpoint revisions will be made on those trip set-points that provide primary protection for accidents that results in an adverse environment inside containment.
The recommended setpoints derived above result in operating res trictions. Westinghouse is there fore evaluating two alter-nate long term solutions which will permit the icwering of the steam generator water trip setpoints. The two systems under consideration are as follows:
Mechanical compensation of sealed reference legs Temperature compensation of transmitter output B.
Pressurizer Water Level Trip Setpoint No credit is taken for this reactor trip function following a high energy line rupture inside containment.
Thus the trip setpoint need not be revised to include environmental errors.
C.
SEC0llDARY TRIP FUi!CTIONS A review of the FSAR shows that water level protection ciannels have not been used as
" backup" or " secondary" protection for any analyzed event which would cause an adverse containment environment.
175ll083
4.
Review and indicate the required re ristons, as necessary, of emergency procedures to include specific information obtained from the review and evaluation of items 1, 2, and 3 to ensure that the operators are instructed on the potential for and magnitude of erroreous level signals. Provide a copy of tables, curves, or correction factors that would be applied to non-accident monitoring systems that will be used by plant operators.
Response
As listed in the response to question 1, the steam generator narrow range water level and pressurizer water level instruments are used for post-accident monitoring.
Because of reference leg heatup and process variable changes, the indicated para-meters may provide erroneous infor>ation to the operator fol-lowing a high energy line rrpture.
The limits of allowable indicated water level are provided based on conservative upper bound error calcuations frcm the response to question 2.
Indicated steam generator water level can be maintained within a range that will assure that adequate heat removal capacity exists.
For Sequoyah this range of indicated water level is determined as follows :
Maximum Minimum water level water level Transmitte error, total, adverse environment
+25 percent
-25 percent Level re ference leg, 90* F to 340*F
+15 percent 0
Process pressure error, 800-1100 psia 0 bottom o f span
+1 percent
@ top of span
-4 percent Total bottom o f span
+40 percent Total top o f span
-29 percent The-e fore, to assure the s team generator tubes are covered, t.he indicaud water level must be greater than 40 percent.
To assure the steam generator is not filled, the indicated water level must be less than 71 percent.
175fU84
Indicated pressurizer water level can be maintained within a range that will ensure that a water level exists in the pressurizer.
For Sequoyah this range is determined as follows (the reference leg error given below is a boundary analysis and may be reduced by further analysis):
Minimum Maximum water level water-level Transmitter error total
+25 percent
-25 percent Level reference leg, 90*F to 340*F 22 percent 0
Process pressure error 200 - 2350 psia 0 bottom of span
+3 percent Stop of span
-5 percent Total bottom of span
+50 percent Total top of span
-30 percent Therefore to assure a water level exists (i.e. not full or empty) the indicated water level must be greater than 50 percent and less than 70 percent.
Furthermore, a remote possibility exists that the fluid in the 'open reference legs may flash to steam in the depressurized steam genera-tors following a secondlary high energy line rupture.
There fore to alert the operator to the possibility of erroneous indications, the following caution will be inserted in all plant emergency instructions for indicated steam generator water level.
CAUTION:
The operator should use several plant indicated variables to veri fy the existence of water level in one or more steam generators due to the possibility o f erroneous level indication due to measurement system errors. The backup variables that should be used include auxiliary feedwater flow, steamline pressure and primary wide range T at and Tcold-h In particular, the operator should not rely upon steam generator water level indications in any depressurized steam generators following a high energy line rupture inside containme.nt.
This is due to the possibility of re ference leg boiling.
O 17EHl[085 e
STEAM FLOW CONDENSING POTS NRC CONCERN What are the temperature ef fects on the steam flow instrumentation?
RESPONSE
The condensing chambers on the steam flow instrumentation are located 43 inches apart vertically as shown on figure 3.
The effect of sense line heatup on steam flow measurement should be in the high flow direction since the low pressure side of the transmitter has a sense point 43 inches above the high pressure side.
The sense lines would be expected to heat up at an approximately equal rate because of their close proximity. An equal heatup rate will cause an apparent high flow shift in the measured signal since the density of the 43-inch water column drops and puts a lower bias pressure on the low pressure side of the transmitter.
AUXILIARY FEEDWATER (AFW) DESCRIPTION NRC CONCERN Is the AFW system controlled by the Steam Generator level measurement (Reference Leg) instrumentation?
If so, how does heatup of the reference leg following a pipe break affect the AFW system?
RESPONSE
The Sequoyah AFW systems use transmitters separate from the NSSS system.
These transmitters share sense lines, so the heatup problem will affect both. TVA believes the AFW system is satisfactory in that the insula-tion on the sense line delays the heatup so operator action will not be required for 10 minutes. After this time, the operator will utilize a correction chart to correct the temperature effects on the reference leg. To further ensure that water is reaching the steam generator, AFW flow to each steam generator is indicated in the control room. The AFW flow transmitters are located outside containment and therefore do not experience a. harsh environment.
/
17EqlP 087
Ctavr Wai.L
..b i(.
-a.:)
.so.
.b \\
.v.
a
- a..
OEllcWS stysoit 1.b., t Perssou t en
.n.
. Wf
'3'. ' ir p y.
- 1
.b
.E PR. E 5 5 0 R.11 E R.
C AV )T'(
WALL Tv Pitat P t.W
\\l iew Op 9h E s su R.12. E L UPPER TB9 LFVet M F s% S u l6 P\\G N T-17 088 fog u r e.,
\\
% ^~
'a 9 s CovoExtarr pot YGdT feeA 4
5 E A LG i.
L 7
a a
g
&\\
V:FC i
bM' "se*s'.e j
.t
,s x
\\
y P.
To TEAra minE A
-)
fg.6 5 5 0 2.i z GR
.a..
. s.
sa.. r t L
1757089 TNPtt.RL
$ ECTio r Vifw Os Paessociteg UPPFL T~R P Levet MF A s o R.E MFWT Fig ure 7-
Fr..w R ESTut.T0 fs 4
43" Y
b
/
O STEa m Ft.ow lust e umerrarior - Fiquar 3 175%090