ML20214D465

Response to NRC Questions on Steam Generator Tube Rupture Chapter of B&W Owners Group Emergency Operating Procedures Technical Bases Document

ML20214D465
Person / Time
Site:	Three Mile Island
Issue date:	11/18/1986
From:	BABCOCK & WILCOX CO.
To:
Shared Package
ML20214D448	List: ML20214D446 ML20214D465 ML20214D503
References
47-1164369, 47-1164369-00, NUDOCS 8611240107
Download: ML20214D465 (24)

Text

. -- __ _

47 1164369-00 o 1 t

e LOWNERS GROUP I Operator Support Committee l

.b

a .raamrsPt1r. N TI. r ~gpst- "w ery sw 1

1 l Response to NRC Quest:]ns on the Steam Generator Tube Rupture Chapter l of the B&W Owners Group Emergency Operating Procedures Technical Bases Document Babcock & Wilcox

• "* " ' " P "Y mnx . .

6.

e,pocr 00 u3 -,39 gcP

47 1164369 00 ia I -

g "" OWNERS GROUP I Operator Support Committee 3 5. N D c, 44 . ki-D.ad[M9.7. j 'I;[g' ; * $216'~

i I Response to NRC Questions on the Steam Generator Tube Rupture Chapter of the B&W Owners Group Emergency Operating Procedures Technical Bases Document i

Babcock &Wilcox a McDermott company l kDR D K 89 V PDR l

7_

d

! 47-1164369-00 1-  ;

1 4

1 1

i 1

RESPONSE TO NRC QUESTIONS ON THE i STEAM GENERATOR TUBE RUPTURE CHAPTER 0F THE B&W OWNERS GROUP EMERGENCY OPERATING PROCEDURES TECHNICAL BASES DOCUMENT i

i J

1 l l

1 i ,

\

i l 4

---

,,---,,,,n--,,,.n,__ . _. _ ,,,_-,- , , _ . , _ . , _ _ _ -, , , _ _ -

ResDoDSe to NRC Staff Request for Additional Information on SGTR Guidelines

1. Request Appended to the TBD is a list of references; however, the text of the TBD does not refer to any of these references. When a statement in the text is intended to be supported by one of the references, this reference should be identified' in the text.

Response

Consideration was given during the development of the TBD to including reference numbers in the text. This was ultimately not done for two basic reasons: firs t , the concise technical nature of the TBD would result in a large number of annotations throughout the document. This would not only be cumbersome, but is also not necessary considering the expected infrequent use of the references. Second, many statements in the text are in reality a consolidation of information from several references. To fully justify statements within the context given requires unde r s tanding all of the references listed in the SGTR section.

2. Request Sections 2.1.1 and 3.1 discuss the means for diagnosis of a SGTR and identification of the affected SGs. Since the TBD provides for subsequent recovery strategies not reviewed in the generically approved Oconee ATOG, and since selection and prioritization of the subsequent recovery strategies can rely both on the information available and the timeliness of that information, the TBD should identify timeframes that the alarms and indications should be available from each of the monitoring methods discussed. Bases for these estimates should be identified or referenced in the text.

Response

The subsequent recovery strategies are those based on the TRACC limits. The time to reach the SG level limit is dependent on the leak rate, decay heat load, steaming and/or draining capability, and initial SG level. A data point is given for SG fill rate in section 3.3.3.2 from which the user can estimate the time availble for a specific transient.

The time to reach the BWST limit is dependent on the leak rate, available initial volume. in the BWST, and the plant specific low level limit based on plant specific steam line volumes out to the first set of isolation valves. The user can estimate the time available based on the leak rate or HPI flow 1

rate for a specific transient.

The time to reach the radiation limit is dependent on many factors based on the monitoring method used, which is plant specific. Therefore it is not possible to include meaningful estimates for this limit.

A conservative calculation has been performed for a double-ended rupture of a single tube in one SG using a set of assumptions that may not be applicable to any one plant. This calculation, therefore, only provides one data point out of an infinite spectrum but may be useful in obtaining a rough estimate of the relationship of TRACC to a conservative double-ended rupture of a single tube. The assumptions used in this calculation were as follows:

a. A failed fuel level of .05% was used since this is a realistic bound of B&W operating plant history.

b. A constant iodine spike factor of 29.9 times the equilibrium full power release rate was assumed to exist over the first ten hours of the cooldown. This iodine spike assumes that all of the fuel rod gap iodine activity in the failed fuel is released in the first ten hours.

c. A best estimate atmospheric dispersion factor (x/q) of 50%

cumulative probability was used from the most restrictive B&W plant site.

d. The entire cooldown was performed by steaming to the atmosphere in natural circulation.

e. No credit was taken for iodine partitioning in the SG.

f. The leak rate (400 gpm) was assumed to remain constant throughout the cooldown even at reduced primary to secondary differential pressures.

The results of this calculation shcwed that the cooldown could be performed in this manner for 691.7 minutes (11.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />) before the TRACC limit on radiation release is reached. This calculation was performed to assess the radiological consequences and thus did not address the TRACC limits on SG level or BWST level. However, using the conservative constant leak rate assumed (item f. above), it is probable that these limits would have been reached during the latter stages of the cooldown. The use of a more realistic leak rate assumption would probably result in not reaching either the SG level or the BWST level TRACC limit during the 11.5< hours of steaming.

This calculation provides, for one set of assumptions, a time to reach the TRACC radiation limit. Even with the same set of assumptions, the timing may be different depending on the 2

individual plant's monitoring methods. Therefore, for this reason and the reasons discussed in the response to item 10, the results of specific analyzed data points are not included in the TBD unless they have generic applicability.

This calculation is very conservative due to assumptions c, e, and f. The atmospheric dispersion factor used was obtained from LOCA analyses and is for ground releases. The releases from a tube rupture are elevated in a steam plume which results in a " chimney effect." Recent analyses by GPU Nuclear show that a x/q based on this chimney ef fect is significantly less than the ground release value depending on the steam exit velocity. The chimney effect x/q is ~1000 times less for main steam safety valves and ~20-225 times less for modulating atmospheric dump valves.

In addition, the analysis takes no credit for iodine partitioning in the SG. Preliminary results of current EPRI analysis supports a conservative iodine decontamination factor (DF) of 4 in the once-through SG over the first two hours of the event and a DF as great as 10 to 15 over the event duration. Therefore, the iodine released to the atmosphere would be 4-10 times less than that assumed in this calculation.

Thus, assuming more realistic values for x/q, DF, and leak i flow rates, the actual thyroid doses at the EAB would be ~100-a 500 times less than shown in this calculation. As a result, the TRACC limit on radiation release should not be reached during a natural circulation cooldown on both SG's with a

double-ended rupture of a single tube. Please refer to the response for item 11 for an example of the reduction in

calculated doses that results from use of more realistic values.

3. Request Section 2.2.4 recommends that if heat transfer doesn't exis t to either SG then restore heat transfer to at least one SG as soon as possible, even if it is the affected, or most affected SG. Further justification should be provided for this strategy (e.g., are there situations where this response may not be mandated?).

Response

The alternative to restoring heat transf er to at least one SG is to establish HPI cooling. HPI cooling is not a preferred heat removal method and at this point in the guideline (2.2.4) the initial cooldown to below the lowest MSSV setpoint has not been performed. This initial cooldown is always performed on both SGs when they are available (also true in ATOG).

J d

3 i

1

Therefore, restoring heat transfer to the af fected or most affected SG is consistent with the plant status that would exist had heat transfer never been lost. In addition, once heat transfer is restored, the TRACC limits are still applicable to ensure that steaming of the affected or most affected SG is bounded.

4. Request

a. Section 2.3.1 and 3.3 discuss cooldown and depressurization to below the MSSV setpoint. To support this strategy, address typical effects (in terms of time to reach the MSSV setpoint, approach to Tube Rupture Alternate Control Criteria (TRACC) limits, SG fill, etc.) that can be expected from this maneuver in the case of a single tube rupture with loss of off-site power.

Response

Again, this strategy is the same as in ATOG . If the affected or most affected SG was not steamed during this initial cooldown the MSSVs on that SG would lift. The MSSVc and the ADVs (which may be in use due to the loss of off-site power) both relieve to the atmosphere but the ADV can be isolated if it sticks open; a stuck open MSSV cannot be isolated. Thus, the use of the ADVs, if necessary, does not result in an increase in radiation release and may prevent an uncontrolled release.

A compilation of data from previous analyses and dose calculations provides the following assumptions and conclusions:

a. The assumptions used were the same as those listed in the response to item 2 except that the leak rate was calculated as a function of primary and secondary pressures.

b. The tube rupture was initiated at 100% power and a runback was performed to 15% power before the reactor was tripped.

A loss of of f-site power occurred at the time of reactor trip. The inclusion of a runback increased the overall transient time before completion of the cooldown which resulted in higher iodine concentrations and greater BWST depletion. The initial cooldown to below the MSSV setpoint was performed in natural circulation which resulted in a longer cooldown due to the higher initial hot leg temperature and a greater radiation release due to steaming to the atmosphere.

c. At the completion of the initial cooldown to below the MSSV setpoint, the status of the three TRACC parameters was as follows:

4

i. SG level was being controlled at the natural circulation setpoint of ~50% on the operate range.

The steaming of the affected SG was sufficient to maintain SG 1evel.

ii. Total BWST depletion at this point was 11,680 gallons which is less than 10% of the available BWST inventory for the most restrictive plant specific TRACC limit.

iii. The integrated thyroid dose at the EAB was 314 mrem which is 1.05% of the TRACC limit. The integrated whole body dose was less than 1 mrem which is less than 0.004% of the TRACC limit.

Thus, for a rupture of a single tube using some conservative assumptions, none of the TRACC limits are approached during the initial cooldown to below the MSSV setpoint.

Request

b. Expand the guidance in section 2.3.1 to address the depressurization alternatives across the B&W product line.

Include an evaluation of pressurizer sprays, PORVs, vents, and any other viable options, with identification of priorities. Relate the approved ATOG guidance of an additional 50F SCM (if PORV is used) to the absence of such guidance in the TBD.

Response

The requested information is provided in the remainder of the TBD which was not available for NRC review at the time of submittal of the SGTR guidelines. The TBD was subsequently submitted in its entirety on September 11, 1985. In particular, Chapter III.G of the TBD, "Cooldown Methods,"

provides information on alternate depressurization methods. In addition, the TBD provides additional guidance on RCS voids not covered by ATOG which is also contained in Chapter III.G.

The SGTR guidelines reference Chapter III.G in sections 2.4.6, 2.5.4, 2.5.5, 2.5.6, 2.5.7, 2.5.8, 2.5.9, 2.5.10, 2.6.1, 2.6.2, 3.3.1.2, 3.6.1, 3.6.2, and 3.7. l

5. Request Section 2.5.2 states the RCP NPSH may require a higher RC pressure than the subcooling margin. Evaluate the effect of this difference in pressure on overall tube leakage and .

discuss the viability of less conservative NPSH requirements (i.e., emergency limits for the RCP's on B&W plants).

5

Response

B&W plants have RC pumps made by three manufacturers: ,

Westinghouse (W_), Bingham, and Byron-Jackson (B-J). The NPSH curves for these pumps (in the pump combinations requiring the highest NPSH) are shown on Figure 1. Please note that the curves for Westinghouse and Bingham pumps are essentially

, identical. Figure 1 also shows a constant 25F subcooling margin curve with a minimum RC pressure of 200 psia required for DHRS operation. The volumetric leak rates for a double-ended rupture of a single tube at these conditions are shown below and on Figure 2.

RCS Temo (F) Leak Flow @ NPSH (com) Leak Flow-@ SCM (com)

W B-J 500 171.2 141.0 147.4 450 169.7 140.7 126.4 400 169.5 139.7 108.5 350 168.7 138.9 100.8 300 168.6 138.9 125.0 250 160.1 127.1 124.2 The difference in tube leak flow between the SCM curve and the NPSH curves is not significant enough to warrant violation of

NPSH requirements thus the TBD states that the NPSH curves should be observed. However, some plants, especially those

with Westinghouse or Bingham pumps, may wish to pursue emergency NPSH curves as GPU Nuclear has done. Emergency NPSH 1

curves can be accomodated within the TBD criteria.

6. Request Section 2.5.10 discusses void mitigation. The Owners Group recommends that during natural circulation, if a void develops, the operator should hold the cooldown and eliminate the void. Justification should be provided for this " hold-the-cooldown" strategy considering the variety of situations that could exist (i.e., why hold up cooldown if the void is not severe?).

Response

l Normally, a natural circulation cooldown is not recommended.

If possible, natural circulation is used to just remove decay 4 heat and maintain plant conditions until a forced circulation cooldown can be performed. If a natural circulation cooldown i is performed, then actions are recommended to prevent void formation. If a void does form then the normal course of action would be to hold the 'cooldown while the void is eliminated. This information is provided in Chapter III.G of the TBD.

6 i

I s._ - .. _- .-. - - - - - , - _ _ _ _ - . _ _ _ . , _ - , _ _ _ - ______-. __ _ __ - _ _ - _ _ _ _ _

4 However, this normally recommended method of dealing with voids is not the only option available in Chapter III.G and may not be the preferred method under other circumstances.

Therefore, the statement in 2.5.10 of the SGTR guidelines to '

hold the cooldown and eliminate the void will be removed. This section will reference Chapter III.G for the various options available to manage the void.

7. Request

a. Section 3.3.1.1 states that the cooldown rate may be increased to a maximum of 240F/hr (4F/ min.) if neither of the following is experienced before reaching 520F: a) the affected SG could have carryover or b) radiation release could exceed the TRACC limit; and that "a fas ter rate may increase the tube-to-shell delta T...The tube-to-shell delta T is still limited to a maximum value..." Later, this maximum value is identified as 100F for normal cooldowns and 150F for emergency cooldowns.

a) Provide an evaluation of the typical tube-to-shell delta T that could be expected for a 240F/hr cooldown under the above conditions, b) If the resultant tube-to-shell delta T is greater than 100F, justify that a suitable balance is being achieved between the radiation criterion for invoking the higher criterion and the concern that the higher cooldown rate might result in increased tube damage and subsequent greater releases (compared to a scenario in which a lower cooldown rate is being observed to maintain the tube-to-shell delta T at 1

less than 100F).

Response

The emergency cooldown is performed only until hot leg temperature reaches 500F. If the SG shell does not cool at all during this cooldown the maximum tube-to-shell delta T would be ~50-55F from post-trip conditions or ~100F if the cooldown is performed immediately following a trip from 100% power (177FA plants; 205FA plants could realize a maximum delta T of

~120F but the normal delta T limit on 205FA plants is 160F).

Some SG shell cooling will occur during this cooldown so the actual delta T will be somewhat lower than these values. Thus the statement in the TBD is misleading in that it implies a larger tube-to-shell delta T than will exist for the emergency cooldown.

expected delta T.

This statement will be revised to reflect the t

The summary of significant changes in tube rupture guidelines 7

provided with these responses notes that the emergency cooldown is recommended for fewer cases than originally used in ATOG (item 5 in the summary) . If the emergency cooldown is not used for the two cases identified in the TBD, the SG will overfill and/or the dose limits will'be exceeded before the SG can be isolated using the normal cooldown rate. This' consideration far outweighs the consequences of a slightly higher tube-to-shell delta T. ,

Rcquest

b. The term " rapidly" on page 28 should be better defined to allow a clearer action level at which the operator should initiate the 240F/hr e: srgency cooldown.

Response

The term " rapidly" will be deleted and the intent of this statement will be clarified. The emergency cooldown should be performed if the SG fill rate will reach the TRACC SG level limit defined in section 3.6.1.1.b before the SG can be isolated using the normal cooldown rate, including the use of SG drains if available.

8. Request Section 3.3.1.4, " Summary of Limits During Cooldown," states that the fuel pin compression limit should be observed. An earlier section implies that this limit is not a safety concern for the cooldown in progress and identifies situations in which it recommends allowing violation of the limit.

Assuming that observing the limit could slow down SGTR mitigation, and considering the previous paragraph which indicated that the limit is not considered a safety limit, the conflicting recommendation in Section 3.3.1.4 should be deleted or modified.

Response

Section 3.3.1.3 discusses the fuel pin in compression limit and notes cases where the limit should be observed. Thus the statement in 3. 3.1. 4 is consistent but will be clarified to note the exceptions identified in section 3.3.1.3.

9. Request Section 3.3.2.1 discusses the disadvantages of SG isolation.

Further quantification is needed to evaluate the timing of SG isolation and for procedure writers to properly assess the relative merits of isolation versus alternative mitigation strategies . In particular:

8

a) The discussion states that an isolated SG becomes a heat source, causing the cooldown to take longer.

Quantitatively compare the cooldown times with and without SG isolation. .

b) The discussion states that the loop with an isolated SG becomes stagnant during natural circulation and the loop may void, hindering pressure control and delaying cooldown. Justify this qualitative conclusion against the consideration that since the isolated steam generator becomes a heat source, reverse flow could occur in the loop. Consider in this justification the potential for reverse flow to sweep voids from the loop.

If it is determined that a void would develop and not be swept away, place a perspective on this consideration by identifying how much voiding would be required to stop natural circulation flow and by quantifying how long it would take to remove the void, identifying various void removal techniques.

Response

There are numerous variables involved that differ plant to plant such as TBV and ADV capacities, venting capability and capacity, condenser availability, DHRS cut-in conditions, etc.

For this reason the TBD has avoided specific numerical quantification in the interests of maintaining the generic aspects of the guidelines. The generic considerations are provided to enable each plant to quantify the options for their particular design.

For example, natural circulat'.on cooldowns to DHRS operation can range in duration from ~6 nours with large TBV and RV vent capacities and use of two loops to ~110 hours with small ADV capacity and using only one loop. There are a large number of possibilities within these bounds. Forced circulation cooldowns can range from ~4.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> to ~13 hours. The significant differences in the forced circulation range are that stagnant RCS regions do not exist that can delay the cooldown as in natural circulation and that the condenser is assumed to be .trailable which provides greater steaming capacity.

There is no indication from existing analyses that reverse natural circulation will occur in a stagnant loop nor is it considered likely. Therefore the formation of a void or the void size does not impact flow in the stagnant loop. The void will, however, impact the ability to depressurize by acting as a second pressurizer. As RC pressure is reduced additional flashing will occur in the voided hot leg which will slow the pressure decrease. The void may be removed in a matter of 9

\

l seconds if RC pumps are available and if the void is within j the siz range of available inventory f rom the pressurizer

(~500 f tg) . The timing required to remove the void by venting l depends on the vent size, void size, and whether a continuous or intermittent vent procedure is used. Guidelines for void management are provided in Chapter III.G of the TBD.

10. Request Section 3.3.2.2, " Continued Steaming-Advantages," Item 1 states that two loop cooldown allows the fastest possible cooldown. In order to assess the relative merits of two loop cooldown to allow a procedure writer to select his option with a more complete understanding of the advantages and disadvantages, additional quantification should be provided comparing cooldown times (to DHRS cut-in conditions) for two loop cooldown, one loop cooldown, HPI cooling, etc. (including all options permitted by the TBD). For these analyses, identify all cooldown considerations (e.g., maximum cooldown rate limitation, ADV capacity, etc.) for various phases of the cooldown.

Response

As stated in the response to item 9, the intent of the TBD is to provide the generic considerations in the guidelines with each plant providing the quantification based on their particular design. The TBD provides quantified information when the quantification can be presented generically. Total quantification within the TBD, which is the apparent thrust of the request for additional information, will destroy the generic nature of the document and result in large sections not being applicable to a given plant. However, the guidelines did consider quantified information in their development and discrete action points, such as TRACC, are quantified in the guidelines.

11. Request Section 3.3.2.2, " Continued Steaming-Disadvantages," Item 1 states that higher integrated releases result from steaming both SGs than if the affected SG were isolated and allowed to fill. Quantify the considerations that warrant toleration of these higher releases. Compare the expected typical release for a single tube rupture event with continued steaming of the ruptured SG versus the value of a comparable parameter calculated with isolation of the ruptured SG.

Response

Please refer to the document entitled " Comparison of Transient Mitigation Philosophies for Steam Generator Tube Rupture 10 9 .- - . - . - _ ,. -

Between the Technical Bases Document and Abnormal Transient Operating Guidelines" provided with these responses. Any action that delays the RCS cooldown to DHRS operation must be weighed against the consequences of alternative actions that facilitate a faster cooldown. The TBD has attempted to establish an acceptable balance between the conflicting goals of minimization of releases and performance of an orderly, expeditious cooldown.

Several dose calculations were performed in establishing the TRACC limits. The initial goal of the TRACC radiation limit was to meet 10% of the 10CFR100 limits over the duration of the cooldown regardless of the number of ruptured tubes. The dose calculations demonstrated the 10% limit is appropriate.

This is a significant improvement over the earlier SGTR guidelines in ATOG in that no limit on releases existed in the ATOG version and the ATOG guidelines only covered rupture of a single tube. A summary of the dose calculations is as follows:

Cooldown Mode Thyroid Dose (rem) Whole Body Dose (rem)

Two loop PC 1.0x10-4 2.1x10-3 Affected SG isolated at 540F Two loop FC 8.29x10-4 2.5x10-3 Both SGs Steamed to DHRS Two loop NC 1 1.27 4.55x10-3 Affected SG isolated at 540F Two loop NC 19.1 2.34x10-2 Both SGs Steamed to DHRS 1Cooldown mode is single loop NC following SG isolation.

The assumptions used for these calculations were as follows:

1. Dispersion factor (x/q) of 50% for most limiting B&W plant site.

2. Failed fuel of .05% (realistic upper bound based on B&W plant fuel history).

3. Continuous iodine spike factor of 29.9 for a ten hour period.

11

4. No credit taken for iodine partitioning in the SG.

5. Double-ended rupture of a single tube. The difference between these calculations and the one described in the response to item 2 is that pr edic ted leak flows we re used for these calculations while the calculation in item 2 assumed constant leak flow.

The lowest releases are obtained by isolation of the af fected f SG as soon as possible. What these calculations do not show are the possibilities for higher releases due to other complications that can arise during the cooldown such as loss of heat transfer to the remaining SG. The philosophy of the guidelines is to minimize the " window" for these other complications by shortening the duration of the cooldown and by maintaining the plant in as normal a, configuration as possible as long as the TRACC limits are observed.

As stated in the response to item 2, these dose calculations are conservative. If a SG iodine DF of 5 is applied and the x/q is reduced by a factor of 25 to account for the chimney effect, the offsite doses at the EAB for the duration of the event become:

Cooldown Mode Thyroid Dose (rem) Whole Body Dose (rem)

Two loop FC <0.1x10-3 <2.10x10-3 Affected SG isolated at 540F Two loop FC <0.1x10-3 <2.50x10-3 Both SGs Steamed to DHRS Two loop NC 1.01x10-2 <0.2x10-3 Affected SG isolated at 540F Two loop NC to DHRS 1.53x10-1 <1.0x10-3

12. Request Sections 3.4.1 and 1.3.2 state that minimizing the amount of radiation released to the environment is one of the objectives for the SGTR TBD. Since intentional steaming of an affected steam generator to the environment is one of the recommended strategies (i.e., Section 3.3.2.2, Disadvantages, Item 1 states "...this method results in higher integrated l releases..."), this objective is misleading and should be reworded.

l 12 i

l l

Response

Section 1.3.2 (section 1.1.2 in the chapter provided with the TBD) states "The intent is to steam the affected SG(s) as much as possible to aid the RCS cooldown, but limit the releases to less than pre-determined limits." This is not misleading.

Section 3.4.1.3 describes the philosophy of the SGTR guideline in achieving a balance between the conflicting goals of a controlled cooldown and what can be considered acceptable levels of radiation release. The TBD contains no statement to the effect that releases will be kept at the minimum possible levels regardless of what this does to the plant configuration. Isolating the af fected SG as soon as possible, thereby placing the plant in an abnormal configuration and lengthening the duration of the cooldown, entails an element of risk that no complications will subsequently arise that could lead to higher releases than if both SGs were continuously steamed. The TBD endorses the philosophy that there is an acceptable level of releases before this risk must be taken. Please note that the TRACC limit is an integrated limit over the duration of the cooldown. In the natural circulation example given in the response to item 11 (steaming both SGs to atmosphere all the way down to DHRS) , a person would have to stand at the EAB for 19 hours2.199074e-4 days <br />0.00528 hours <br />3.141534e-5 weeks <br />7.2295e-6 months <br /> and 55 minutes to receive 19.1 rem to the thyroid. Current licensing bases for this size tube leak would " allow" 30R over the first 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />.

13. Request Section 3.4.1.1, Item a , states that " studies show that the thyroid doses, which are due to iodine, are limiting for tube ruptures." Discuss the relevance of this statement to licensing bases of older B&W plants. Some of the studies should be identified in the text. Also, the calculations mentioned in Section 3.4.1.1 should be referenced.

Response

The statement in 3.4.1.1 refers to realistic tube ruptu r e scenarios and was not intended to address licensing bases.

Realistic tube rupture cases, with some conservatism, such as those described in the response to item 11 show that the whole body doses are a higher percentage of the respective limit if all steaming is done to the condenser. However, these doses are extremely low relative to the limit (~0.1% of the TRACC limit) and therefore are not very limiting. If, however, all steaming is done to the atmosphere, the thyroid doses are a higher percentage of the respective limit and would reach the limit sooner. Thus the conclusion that the thyroid doses are limiting for real, " worst case" tube ruptures.

13

The use of the TRACC radiation limit is more conservative than the licensing bases for the older B&W plants and the TBD covers multiple tube ruptures which is beyond the design bases for any plant. Therefore, the owners Group position is that the TBD SGTR guidelines are more conservative than the licensing bases.

References for the SGTR guidelines are provided in the TBD.

Please see the response to item 1 for the reasoning for not including references within the TBD text.

14. Request Section 3.4.1.2 cautions against overfilling steam generators into steamlines in condition (a), and identifies an exception in condition (b). Neither item mentions other related considerations, such as waterhammer in steam lines and steamline qualification for water loads. Provide a discussion of these conditions. Also the loadings should be included in the SGTR TBD to provide the procedure writer with a more complete perspective of the issues.

Response

Section 3.4.1.2 is concerned with preventing uncontrolled releases through the MSSVs and provides guidance for two specific conditions that may challenge the MSSVs. Condition (b) is not an exception but rather a case where overfill can not be prevented. In this case, other actions (reference to section 3.7) are recommended to prevent lifting of the MSSVs due to the overfill.

Steam line configurations are plant specific and detailed analyses of the steam lines are outside the scope of the TBD.

Consideration for steam line loadings due to overfill is addressed in section 3.3.3.2. However, owners do recognize the existence of phenomena such as water hammer and have considered such effects in plant design and procedures to the extent possible.

15. Request Section 3.5, " Considerations For Use of SG Drains," states that for the case in which radiation release rate is the limiting TRACC, drains should not be used if the site boundary dose rates would increase due to the storage location of the contaminated fluid. Section 3.4.4 (2nd paragraph) indicates the likelihood that if one TRACC limit is reached, all limits may be reached. The guidance of the above drain recommendation should be reconsidered in light of this statement in Section 3.4.4, the better control of a storage location (versus the uncertain alternative disposition of the dose), and the 14

overall philosophy in steaming recommendations. .

Response

It is possible that all of the TRACC limits may be reached in a given transient since all of the limits are rate dependent and all of the limits would normally require the rupture of several tubes. However, the third paragraph in section 3.4.4 notes that any one of the three may be more limiting, i.e.,

one TRACC limit may be reached sooner in the transient.

Section 3.5 discusses the merits of the drains relative to each of the TRACC limits. The statement in question refers to the possibility that the storage location may be such that releases from the stored drainage contributes to the site boundary doses. In such a case, the releases from the storage location must be considered in determining the effectiveness of using the SG drains, i.e., use of the drains may not prevent reaching the TRACC limit on releases. The TRACC limit on releases is a fixed limit regardless of whether the SG is steamed, drained, or both. Ultimately, the SG will be isolated if necessary to prevent exceeding the TRACC limit. Therefore, the statement in the request regarding uncertain alternative disposition of the dose does not apply.

16. Request Section 3.6.2, item 2 states that the use of feedwater spray may be beneficial in reducing steam pressure. Discuss how much this maneuver affects tube-to-shell delta T, and consider situations where this action might be detrimental.

Response

The isolated SG must have an established level and heat transfer from the RCS in order to sustain pressure near the MSSV lift pressure. The addition of feedwater spray through the auxiliary feedwater header in this case will not result in large tube-to-shell delta T's. The use of feedwater spray may not be very effective in reducing SG pressure with a high SG level, high SG fill rate (due to the tube leak), and high RC temperatures but it is not expected to be detrimental.

17. Request Section 3.6.2, item 3 states that it may be desirable to augment SG filling with feedwater. Evaluate whether this maneuver could lead to a) backflow into the RCS and unacceptable dilution of the RCS boron concentration, or b) overpressurization of the SG due to malfunction or inadequate pressure / inventory control.

15

l

Response

It is possible that this maneuver could lead to backflow into the RCS or pressurization of the SG to the point of lifting MSSVs. Precautions will be added to item 3 in section 3.6.2 to address these considerations.

18. Request Section 3.6.2, item 4, discusses converting the SG to a water / water heat exchanger. Insufficient guidance is provided to determine which SG is being discussed at any given point in this section, what the status of the other SG should be, and analyses which demonstrate the viability of this cooling mode.

Also, shouldn't limitation a) include the possibility that the drained water can be treated (e.g. , borated , etc. ) ?

Response

Item 4 of section 3.6.2 will be revised to clarify the status of both SGs and which SG is being used as a water / water heat exchanger. In addition, item 3 of section 3.6.1, item 4 of section 3.6.2, and item 3 of section 3.6.3 will be revised to be consistent (two dif fe rent flow paths, one using SG drains and one using TBVs, are discussed). Limitation a) will also be revised to note that the drained water can be treated for reinjection into the RCS as discussed in section 3.5.

No analyses have been performed for the TBD to demonstrate the viability of this cooling mode. As stated in sections 3.6.1 (item 3) and 3.6.2 (item 4), this method should only be used if no other cooling methods are available or are less

%* desirable. Section 3.6.1 also notes that analyses (plant specific) should exist before considering this method. Plant specific analyses were performed in support of using the SGs as water / water heat exchangers at TMI-2. While the TMI-2 method involved the use of specially installed equipment, the calculations performed for TMI-2 will be reviewed for generic applicability and use in the TBD. At the very least, some of the considerations and precautions for TMI-2 may have a generic application.

19. Request Section 3.6.2, Item 4, cautions against backflow through the tube rupture into the RCS. Are there situations where, with supplemental RCS boration, tube rupture backflow may be desirable? If so, further consideration of this maneuver to control the ruptured SG should be explored.

16

Response

There is no particular benefit to backflow into the RCS as there may be in a U-tube SG. Guidelines for U-tube SGs may use high SG levels to reouce or terminate tube leak flow which will cause backflow during RCS depressurization. Backflow may occur in an OTSG if the steam lines are filled sufficiently as discussed in section 3.6.2 but is of no particular benefit in controlling the tube leak rate. However, since backflow can occur, a precaution is in order and was added due to a previous verbal NRC comment.

20. Request Section 3.7 states that the operator may attempt to limit RCS pressure increase by opening high point vents. Additional guidance should be provided by further discussing the relative effectiveness of this action and its prioritization versus other pressure relieving techniques.

Response

The effectiveness of the high point vents depends on the plant-specific vent design. Prioritization, with one exception, is also plant-specific and therefore the TBD just identifies the possible paths for pressure relief. For example, the use of SG drains depends on their availability and design limitations relative to the actual conditions during the transient. The one exception is the use of TBVs and/or ADVs and item d specifically states that this option should be used last.

21. Request Several sections (including but not necessarily limited to Section 2.2.2, Section 2.3.1, Section 3.3.2.1, Section 3.3.2.2, Section 3.3.3.1, Section 3.3.3.2, Section 3.4.4, Section 3.6, Section 3.6.1, Section 3.6.2, Section 3.6.3, Section 3.7) contain discussions related to isolation of af fected steam generator (s) . The f ollowing questions apply to this topic:

a) Throughout the document a clear distinction between filling the SG, overfilling the SG, flooding or filling the steamlines, and involuntary releases is not apparent. These conditions should be clearly defined, identified throughout the . document , and, in each case identified, a discussion of the status, consequences, and influences of various operator actions should be provided. All discussion should be supported by analyses.

17

~

b) In any situation discussing isolation of an affected SG, the discussion should also quantify whether the SG is expected to fi?.1, overfill, etc. (if isolated), and how long the fill or overfill would be expected to take.

This supporting information is important in assessing overall dose considerations and evaluating the feasibility of operator actions.

c) Qualification of steamlines should be considered for each case in which water may carryover from SG into steam lines. Though the qualification of steamlines for water loads may be plant-specific, the discussion should accommodate this consideration, providing the flexibility of alternative actions for plants with qualified lines, and cautions and restrictions for those whose lines are not qualified.

d) In cases of isolation of an affected SG, analyses should be provided to quantify the relative effectiveness of various relieving techniques in limiting or preventing the rise in SG level, the overfilling of the SG, and the rise in SG/steamline pressure.

Response

a) The terms filling, overfilling, and flooding are essentially interchangeable in that all refer to increasing the SG level to the point where water enters the steam lines and the SG may no longer be available for steaming.

Overfilling is normally used to denote an uncontrolled increase in SG level to this point. However, some uses of the term filling refer to intentional or unintentional increases in SG level. The guidelines will be reviewed to ensure consistent use of these terms and definitions will be added. The guidelines will also be reviewed and revised, as necessary, to ensure each use of these terms is clear regarding plant status and the consequences and influences of operator actions. The guidelines are supported by analyses in the sense that all statements are either a) directly based on an analytical result, b) based on an engineering compilation or extrapolation of analytical results, or c) based on engineering judgement derived from analytical and operational experience.

b) As noted in the response to item 2, the time to fill the SG is dependent on several transient-specific parameters and therefore it is not feasible to quantify the timing to overfill of the SG throughout the document. A data point is provided in section 3.3.3.2 from which the user can project fill times for specific transients. The most important consideration, which is stressed throughout the document, 18

~.

la - to depressurize the RCS to below the MSSV setpoint

before the steam lines fill completely. Larger leak rates will cause higher fill rates but also aid RCS depressurization.

i c) Section 3.3.3.2 states that the steam lines should not be allowed to fill until RC pressure and temperature are within the steamline boundaries established by analyses.

Section 3.3.3.2 addresses the optional isolation and filling of the SG where the user has a choice. The limitation should apply whenever the steam lines are filling but in other situations the user may have no choice; the steam lines may fill regardless of other options available. However, if the loading analyses allow filling of the steam lines with the RC pressure and temperature that will exist at 3.3.3.2, then filling at lower pressures and temperatures that will exist at TRACC should be bounded.

d) Various options are discussed for steaming, draining, and the use of alternate pressure relief paths. However, as previously stated, all of these options are plant specific and therefore cannot be quantified generically. The generic considerations and limitations are provided and each user can quantify the options based on the plant specific design.

22. Request In various Sections (including Section 3.3.3.2 and Section 3.4.3) the TBD states that a SG, once filled, should be assumed not to be available for the remainder of the cooldown.

This statement should be clarified and justified. What is meant by " fill the SG?" Why is the steam generator not any i longer available and what phenomena would prevent or discourage its use? Section 3.6.2, Item 4, suggests use of the SG as water / water heat exchanger. Doesn't this maneuver involve " Filling" the steam generator?

Response

" Fill the SG" means filling to the point of water entering the l steam lines. This will be clarified as part of the response to item 21.a. It is not prudent to steam the SG when water can be entrained in the steam flow and damage steam valves. Section 3.3.3.2 notes that use of SG drains may not drain the steam lines, implying that, if the drain and steam line configuration is such that draining of the steam lines can be accomplished, the SG may be available for steaming considering other limitations (TRACC). Section 3.3.3.2 will be revised to make this statement explicity. Section 3.4.3 makes the same implication and will also be clarified. Section 3.6 discusses 19 i

-____ __ _ __ ____ _ _ . - _ - . _ - . . _ _ . , , . . . - . - . _ . . . - , , - - _ - _ . _ . ~ , - , - . . . - .-_ ,-

i special considerations for unisolating a SG and provides the reasoning and limitations for these considerations.

23. Request In discussions of subcooling margin, the SGTR TBD does not adequately identify the desirability of minimizing subcooling margin for SGTR events. This should be added, or justified otherwise.

Response

Section 3.3.1 discusses maintaining minimum subcooling margin and notes that this will minimize the RCS-SG differential

~

4 pressure. This section also discusses observation of RCP NPSH and fuel pin in compression limits. Section 3.3.1.3 discusses the impact on the tube rupture of observing the fuel pin in compression limit versus minimum subcooling margin. The

discussion in section 3.3.1 will be expanded to emphasize the desirability of maintaining minimum RCS-SG differential pressure but, when applicable, the RCP NPSH limit should still ~

be observed.

l

24. Request Section 3.4.1.3, the second to the last paragraph describes dose limits expected by implementation of the TBD recommended mitigation strategies. The paragraph should be reworded to identify the scenarios for the dose limits discussed and the licensing bases for the plants which might reference the TBD.

Response

The paragraph in question in section 3.4.1.3 states that this dose limit applies for all tube leak rates and cooldown times, i.e., all scenarios. Section 3.4.1.3 also discusses the relationship of this dose limit to licensing bases. Licensing bases have evolved over the years to become increasingly more restrictive. Current licensing bases , which do not apply to I the operating plants referencing the TBD, contain an explicit limit of 10% of 10CFR100 doses. This is the same limit used in the TBD thus is conservative for the. operating plants. Current licensing bases are structured around the double-ended rupture of a single tube. The TBD covers multiple tube ruptures in one or both SGs and therefore is more conservative than the licensing bases since the same dose limit applies. Current licensing bases limit the doses to 10% of 10CFR100 over the first two hours of the event at the EAB. The TBD uses the same integrated dose limits but applies them over the duration of the cooldown, resulting in much lower dose rates for scenarios

, up to and including the double-ended rupture of a single tube and thus is more conservative than the design bases. Finally, 20 1

the licensing bases assume certain conditions and predict the radiological consequences such that an event beyond the design bases (e.g., multiple ruptures) involves unknown radiological consequences. The TBD, however, uses a real-time dose limit so that the radiological consequences are defined and limited regardless of the scenario. Realistic calculations, that still employ conservatism, show that the actual doses will be considerably less than 10% of 10CFR100 limits over the entire cooldown to DHRS operation for leak sizes up to and including the double-ended rupture of a single tube (please see response to item 11).

I Current licensing bases refers to Section 15.6.3,

" Radiological Consequences of Steam Generator Tube Failure (PWR)," of the Standard Review Plan, NUREG-0800, Rev. 2, July, 1981

25. Request Section 3.4.4 (" application of TRACC") recommends that if TRACC is expected to be violated due to continued steaming of the unisolated SG, then it should also be isolated and HPI cooling initiated. The implication is that this will prevent violation of TRACC. Clarify whether isolation and HPI cooling does prevent violation of TRACC limit (s).

Response

The three TRACC limit,s are designed to a) prevent ca r ryove r from a steaming SG, b) limit the unrecoverable losses from the BWST, and c) limit the integrated doses at the EAB. The isolation of both SGs prevents carryover from a steaming SG and, assuming no further complications, terminates the radiation release to the environment. The initiation of HPI cooling not only cools the core but allows the further losses from the BWST to be recoverable for reinjection if necessary.

The further complication that could result in additional releases is the subsequent lifting of an MSSV on one of the l

isolated SGs. Prevention of this complication while in HPI cooling is stressed in the guidelines. In addition, it is not I expected to be a problem by the time both SGs are isolated, i.e., RC pressure and temperature and decay heat should be low enough at this point to prevent challenges to the MSSVs. This is a strong consideration for not isolating the SGs before it is absolutely necessary.

t 26. Request l

l There are some obvious differences in Chapter III-E compared to the approved SGTR guidelines in Reference 2, and additional l clarification is needed. Provide a comparison summary of the i significant changes in SGTR strategies and include (or 21

p ,

reference) the bases for these changes. An example is the increase in subcooling margin prior to use of the PORV (for '

HPI cooling) on Reference 2 compared to the absence of this guidance in the proposed TBD Chapter III-E.

Response

Please refer to the attached summa ry of significant changes.

Regarding the specific example, the PORV was used in Reference 2 for RC pressure control during natural circulation with a tube rupture. This use of the PORV and alternate depressurization methods are provided in Chapter III.G, "Cooldown Methods," in the TBD.

27. Request Section 3.8, " Impact of Unisolable Steam Leaks," Item "b" recommends that if the steam leak is inside the RB, the SG should be steamed to the condenser, if available, only while steaming the SG is necessary for other considerations. The discussion should be clarified to address the additional considerations that should be given if the condenser is not available. For example, conditions may dictate steaming for reasons other than TRACC limits, such as an unacceptable rise in containment pressure. Also clarify what feedwater controls are recommended.

Response

Once the SG is isolated, the only release to the RB is the tube leakage which is essentially a SBLOCA. The RB can handle this release without an unacceptable rise in containment pressure. Section 3.8 states that the same basic philosophy for managing steam leaks in Chapter III.D (" Excessive Prima ry to Secondary Heat Transfer") is used for tube ruptures except as noted. Section 3.8 also notes that this philosophy is to steam the SG(s) with the steam leak as long as the cooldown is controllable and states to use the condenser, if available.

This means to steam to the atmosphere if the condenser is not available as long as continued steaming is required or allowed by other considerations in the SGTR guidelines. Control of feedwater while steaming is dictated by the size of the tube leak and the guidelines for steam leaks in Chapter III.D.

Feedwater is isolated when the SG is is olated as noted in section 3.8.

l l

l l 22