K epENCLOSURE 3

REGULATORY ANALYSIS

REVISION 2 TO REGULATORY GUIDE 1.99

RADIATION EMBRITTLEMENT OF REACTOR VESSEL MATERIALS

-- Al

/I

~ /

/-" ++

....... 1 ./. 8 .......Z B ........9703060433 871120PDR REQQD01,099 R 0M

r MarK Kim - Keg kiuicie 1.99 Page 1 11II M~~K i'yrK - ~eg (.~u ide 1 kiY Page 1

From: PDRTo: Mark KirkDate: 6/6/2007 1:55:38 PMSubject: Reg Guide 1.99

Hello Mark,

I found few documents on Task ME 305-4 this may interest you.Document below is the exact document which you were looking for. This document is on microfiche andcard number is91984:258-91984:320. You can get this microfiche from PDR.

see youSardar

Accession Number: 9703060433Title: Rev 2 to reg guide 01.099,task ME 305-4, "Radiation Embrittlement of Reactor Vessel Matls."Document Date: 11/20/1987Estimated Page Count: 63Document Type: REGULATORY GUIDE (REG GUIDE), REGULATORY GUIDES & STD. REVIEWPLANS-TEXTDocument/Report Number: REGGD-01.099, REGGD-1.099Author Affiliation: NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)Microform Addresses: 91984:258-91984:320Physical File Location: PDR:REGGD7-01.099-R-871120,PDR:REGGD//01.099 R 871120,CF:SUBJ//I&P-11GUI&MAN 871120Package Number: 9703060433

TABLE OF CONTENTS

1. STATEMENT OF THE PROBLEM ........................................... 1

2. OBJECTIVE ....................................................... 3

3. ALTERNAT IVES ....................................................... 4

4. CONSEQUENCES ....................................................... 4

4.a. Costs and Benefits of Alternatives ........................... 4

4.a.1 Pressurized Water Reactors ............................ 5

4 .a.1.1 Background ................................... 54.a.1.2 Operating Problem with LTOP Systems .......... 54 .a.1 .3 Remedies ............. ! ....................... 8

" Hardware and Operational Remedies ........... 8" Possible Regulatory Relief ................. 9

4.a.1.4 Conclusion .................................. 11

4.a.2 Boiling Water Reactors ................................ 11

4.a.2.1 Regulatory Background ........................ 124.a.2,2 Analysis of Margins .......................... 134.a.2.3 Other Means for Reducing the Impact on BWRs.. 164.a.2. 4 Conclusion ................................... 18

4.a.3 Other Applications of Regulatory Guide 1.99 ........... 184.a,4 Risk Avoided by Using Revision 2 (See Section 5 (3)).. 194.a.5 Costs to NRC (See Section 5 (7)) ...................... 19

4.b. Impact on Other Requirements ................................. 19

4.b.1 Impa,.t on the Pressurized Thermal Shock (PTS) Rule(10 CFR 50,61) ........................................ 19

4.b.2 Impact on Generic Issue 94 ............................ 234.b.3 Impact on Material Selection .......................... 23

4.c. Constraints ................................................. 24

5. BACKFIT ANALYSIS FOR POWER REACTORS ................................ 24

(1) Object ive ........................ ............................. 24(2) Actions by Licensee .......................................... 24(3) Risk Reduction ................................................ 24

" Normal Operation-PWRs ................................... 24" Hydrotests and Leak Tests-BWRs .............................. 27

11/20/87 RG 1.99 REV 2 REG ANAL

TABLE OF CONTENTS (Continued)

* LTOP System *Transients-PWRs ................................. 27* PTS Transients-PWRs......................................... 28* Evaluation of Vessel Integrity After a Transient ............. 29* Evaluation of Vessel Integrity In The Presence of a Flaw .... 30* Summary of Risk ............................................. 30

(4) Radiological Exposure ....................................... 30(5) Cozts - PWRs .................................................. 30(5) Costs - BWRs .................................................. 31(6) Safety Impact of Changes in Plant Complexity .................. 32(7) NRC Resource Burden ........................................... 32(8) Impact of Differences in Facilities.... ........... 33(9) Whether Backfit is Interim or Final ............. ......... 33

6. Decision Rationale ......................................... I ...... 347. Implementation .................................................... 348. Other Impacts .......................................... ......... 35

LIST OF TABLES

P age

Table 1 CHANGES IN PRESSURE - TEMPERATURE LIMITS FOR 5EFPY BEYONDJANUARY 1986 CAUSED BY USE OF REVISION 2 INSTEAD OFREVISION 1 OF R.G. 1.99 ........................................ 36

Table 2 MINIMUM PERMISSIBLE PRESSURE-TEST TEMPERATURES FOR BWRS FROM3-1 IN BWR OWNERS GROUP REPORT "EVALUATION OF REGULATORYGUIDE 1.99, REVISION 2 IMPACT ON BWRS, PART 2, MAY, 1987 ....... 42

Table 3 CHANGES IN RTPTS FOR ALL PWRS IF REVISION 2 WERE USED .......... 43

Table 4 SUMMARY OF TABLE 3 FOR PLANTS THAT MAY EXCEED THE SCREENINGCRITERION ACCORDING TO PRESENT ESTIMATES ....................... 48

LIST OF FIGURES

Page

FIGURE 1 P-T LIMIT SEPARATES THE OPERATING ZONE FROM CONDITIONSHAZARDOUS TO VESSEL INTEGRITY... ............................. 49

11/20/87 ii i RG 1.99 REV 2 REG ANAL

TABLE.OF CONTENTS (Continued)

LIST OF FIGURES (Continued)

P ae

FIGURE 2 A FIXED LTOP SET POINT ACCENTUATES THE IMPACT OF LARGEADJUSTMENTS OF REFERENCE TEMPERATURE ON THE OPERATINGWINDOW IN TEMPERATURE .......................................... 50

FIGURE 3 SEVERAL FACTORS REDUCE THE OPERATING WINDOW IN PRESSUREWHEN THE P-T LIMIT IS PROTECTED BY AN LTOP SYSTEM ............. 51

FIGURE 4 OCONEE 1 -- EFFECT OF TREND CURVE FORMULA USED IN CALCULATINGTHE FAILURE PROBABILITY FOR THREE TRANSIENTS DESCRIBED INNUREG/CR 3770 ................................................ 52

FIGURE 5 CALVERT CLIFFS 1 - EFFECT OF TREND CURVE FORMULA USED INCALCULATING THE FAILURE PROBABILITY FOR THREE TRANSIENTSDESCRIBED IN NUREG/CR 4022 .................................... 53

FIGURE 6 H. B. ROBINSON • - EFFECT OF TREND CURVE FORMULA USED INCALCULATING THE FAILURE PROBABILITY FOR THREE TRANSIENTSDESCRIBED IN NUREG/CR 4183. A HYPOTHETICAL CHEMICALCOMPOSITION WAS USED TO ACKIEVE MORE FAILURES WITHIN THELIMITATIONS OF THE MONTE CARLO METHOD ......................... 54

FIGURE 7 EFFECT OF PRESSURE ON FAILURE PROBABILITY FOR LTOP EVENTS AT1500 F AND 250OF WATER TEMPERATURE AFTER 10 AND 32 EFPY ........ 55

FIGURE 8 FOR THE REACTOR VESSEL REPRESENTED IN THIS ILLUSTRATION,THERE IS A LARGE DIFFERENCE IN THE PREDICTED PROBABILITYOF VESSEL FAILURE FROM PTS EVENTS DEPENDING ON WHETHERRT IS 235OF (CALCULATED USING THE PTS RULE) OR 302OF(CU EULATLD USING REVISION 2) ................................. 56

APPENDICES

Appendix A - Denton/Minogue Memo to Dircks, Aug. 12, 1985 .............. A-1

11/20/87 i v RG 1.99 REV 2 REG ANAL

ENCLOSURE 3

REGULATORY ANALYSIS

REVISION 2 TO REGULATORY GUIDE 1.99

RADIATION E14BRITTLEMENT OF REACTOR VESSEL MATERIALS

1. STATEMENT OF THE PROBLEM

Prevention of fracture of reactor vessels is accomplished, in part, by

warming them before pressurization, following the pressure-temperature (P-T)

limits gi-ven in the Technical Specifications. Neutron radiation embrittlement

of the reactor vessel is compensated for by shifting the P-T limits up the tem-

perature scale every few years by an amount corresponding to the shift in the

Charpy test transition temperature produced by the accumulated neutron fluence.

The NRC regulates this process on the basis of Appendices G and H, 10 CFR

Part 50. Paragraph V.A of Appendix G requires: "The effects of neutron radia-

tion... are to be predicted from the results of pertinent radiation effects

studies...."

Since Revision 1 of Regulatory Guide 1.99 was published ten years ago,

there has been a significant accumulation of power reactor surveillance data,

which constitutes a muco more pertinent basis for the Guide than was available

when Revision 1 was written. Revision 2 is based entirely on the surveillance

data, and its issuance will provide a basis for licensing decisions that con-

stitutes the most pertinent results available, in conformance with the

regulation.

The Guide is needed even though many plants now have surveillance data of

their own. (Of course, for the newer plants such data are not yet available.)

For many older plants, unfortunately, the materials in the surveillance capsules

are not the controlling materials for that reactor-according to our present day

understanding. Thus, instead of. using the plants' own surveillance results

directly, the staff must rely on calculated values based on the chemical com-

position of the vessel materials and the neutron fluence. Regulatory Guide

1.99 Revision 2 upgrades and expands the calculative procedures that are accept-

able to the NRC, and it describes acceptable procedures for using plant-specific

surveillance data when they become available.

11/20/87 1

'I

The Guide is used in any regulatory action that requires knowledge of the

fracture toughness of reactor vessel beltline materials, other than the calcu-

lation of RTPTS for comparison with the screening criteria in the PTS role,

10 CFR 50.61. Three examples of such actions are: (1) setting P-T limits for

heatup ond cooldown, (2) evaluating transients that threaten the integrity of

the reactor vessel, such as low temperature overpressurization and pressurized

thermal shock events, (3) evaluating flaws found during inspection. In any of

these analyses, a key input to the calculation is the fracture toughness of the

material as a function of temperature. The ASME Code gives reference values of

toughness as a function of temperature relative to RTNDTI the "reference temper-

ature nil-ductility transition" of the material. The Code also describes how

to measure the initial RTNDT for the unirradiated material. This Guide gives

calculative procedures for ARTNOT, the adjustment of RTNDT caused by neutron

radiation. The Guide also describes how to combine the initial and the "delta"

with a suitable margin to obtain a value of RTNDT that covers the uncertainties

in both.

From analysis of the new data base, and from experience in applying the

Guide, the need for certain changes became clear.

Nickel, in the presence of copper, has been found to increase the Charpy

shift and should be a factor in the calculations. Thus, some reactor

vessels with high nickel welds, which were made when nickel was added in

the welding process, have more susceptibility to radiation than previously

thought. Conversely, some early reactor vessels that were made with no

deliberate alloying additions of nickel have lower sensitivity to radia-

tion. Implementation of Revision 2 will remedy these situations.

The effects of copper and nickel content on the sensitivity of welds to

radiation embrittlement are different than they are for base metal--so

different as to require separate treatment of welds and base metal in this

Guide.

. The fluence function needs revision to fit the surveillance data.

The calculative procedure needs to be amended to prescribe mean values

instead of upper bound values and to state the margin separately.

11/20/87 2

* Procedures for calculating the attenuatiun of radiation damage through the

vessel wall need to be stated specifically.

Improved knowledge of scatter in the surveillance data base made it necessary

to rewrite the criteria for use of plant-specific surveillance data in

setting P-T limits for that reactor.

I, summary, "better science" reveals that the Guide should be revised

after 10 years of use if margins of safety against fracture are to be

maintained.

Enclosure 9, "Basis for _..," gives the technical justification for the

procedures given in the Guide for calculation of radiation embrittlement.

2. OBJECTIVE

The objective of the guide is best described by reference to the schematic

P-T diagram, Figure 1. The following discussion is mainly applicable to heatup

and cooldown for normal operation of PWRs and to pressure tests for both PWRs

and BWRs. The upper-left boundary of the operating zone, the P-T limit, appears

in the Technical Specifications for all plants together with certain limits on

heatup/cooldown rates. The P-T limits are based on Appendix G, 10 CFR 50,

which incorporates Appendix G and parts of Section III of the ASME Boiler and

Pressure Vessel Code. And, the P-T limits are affected by Regulatory Guide 1.99

as described in the previous Section.

In the upper-left corner of Figure I is a region labelled "hazardous to

vessel integrity" which is bounded by a set of curves to indicate that higher

cooldown rates increase the hazards. This boundary moves upscale in tempera-

ture and downscale in pressure as radiation embrittlement accumulates during

the operaLing life of the vessel. The objective of the margins added in cal-

culating P-T limits is to place the operating zone for heatup/cooldown far

enough from the hazardous region to provide the operator time to diagnose and

correct system transients such as low temperature overpressurizations and rapid

cooldown events that could threaten vessel integrity.

The safety benefit of Revision 2 is also illustrated in Figure 1. For some

plants, depending on their copper and nickel contents an;d fluence levels, we

11/20/87 3

now believe that the boundary of the hazardous region is as shown by the dashed

curves labelled "Rev. 2." That is, the hazardous region is closer to the P-T

limit than expected when Rev. 1 was used in the calculations of P-T limits. To

maintain the safety margins, the P-T limits need to be recalculated using Rev. 2.

3. ALTERNATIVES

The alternatives to issuance of Revision 2 are to leave' Revision ý in place

or to eliminate the Guide altogether. The latter can be disposed of quickly:

the staff reviews several P-T limits per year, plus an occasional transient and

flaw indicatien and clearly needs a published basis for its reviews. There is

at present nothing equivalent to Regulatory Guide 1.99 in the ASME Code. ASTM

Standard Guide E-900-87 contains a calculative procedure relating the Charpy

shift to copper and nickel contents and fluence that is the same as the procedure

given in Revision 2, but the Standard does not contain specific guidance on the

use of plant-specific surveillance data or the calculation of attenuation th-rough

the vessel wall or the amount of margin tj be added. The alternative of con-

tinuing to use Revision 1 has a safety impact on most plants; a detailed

analysis of these consequences is given in the following section.

4. CONSEQUENCES

4.a Costs and Benefits of Alternatives

Technically, the benefit ot using Revision 2 in place of Revision I to

R.G. 1.99 is the fact that Revision 2 has a much better basis. The passage of

nearly 10 years has yielded a better data base and better understanding of

radiation embrittlement mechanisms. From a regulatory standpoint, the benefit

of Revision 2 is that its use restores safety margins to their intended levels.

Revisionil often gives estimates of radiation embrittlement that are low when

compared to the results of surveillance testing of materials irradiated in com-

mercial power reactors. Calculations involving Revision 2 enter into three

kinds of safety evaluations where NRC regulations require use of the best

information available: (1) Evaluation of f"laws found in service, (2) Evalua-

tion of reactor vessel integrity after transients have occurred to see if they

should be inspected for flaws that may have been popped in" (enlarged) by the

11/20/87 4•

transient, and (3) Evaluation of updated P-T limit submittals to see if they

provide operators a realistic picture of the pressure-temperature conditions

that are to be avoided in plant operation, as discussed in Section 2 of this

analysis. The impact and the safety benefits of Revision 2 on P-T limits for

PWRs and BWRs are discussed separately in the following sections.

4.a.1 Pressurized Water, Reactors

4.a.1.1 Background. The immediate impact of using Revision 2 as the basis

for calculating P-T limits for PWRs will be to move the limits upscale typically

about 65'F. Based on the public comments received from PWR owners and vendors,

the impact of the upward adjustment of reference temperature is caused by the

combination of the regulatory requirements in Appendix G, 10 CFR 50 and Regula-

tory Guide 1.99 and the requirement •hat the P-T limits be automatically pro-

tected by a relief-valve system at low temperatures. The LTOP (low temperature

overpressure protection) requirements are contained in the following:

(a) Standard Review Plan, NUREG-O800, Section 5.2.2 "Overpressure

Protection"

(b) Branch Technical Position RSB 5-2

(c) Standard Technical Specifications for LTOP systems

(d) Various generic letters sent to all PWR licensees in 1976.

These requirements were established in the late 1970's when it became evident

that there were an unacceptable number of violations of P-T limits during heat-

ups and cooldowns of PWRs. Plant experience shows there are approximately

3 startups from cold shutdown per reactor year.

4.a.1.2 Operating Problems with LTOP Systems. Allowable pressure during

heatup and cooldown is constrained on the high side by P-T limits based on

fractust considerations and on' the low side by the minimum pressures required

to prevent seal failure and cavitation in the reactor coolant pumps, as illus- -

trated in Figiure 2. Moreover, our requirement for protection of the P-T limits

at low temperatures by using relief valves with a low-set-point pressure narrows

the operating pressure window at startup and shutdown even fu-ther. The set-

point .pressure is normally about 50 psi or so below the P-T limi* to allow for

11/20/87 5

an overshoot during a transient, and the consequences of lifting the valve are

such that the operating window in pressure is well below the set point.

For some plants, there is an even more serious restriction in the window

in temperature, as illustrated in Figure 2. The regulatory requirements cited

above call for automatic protection "at low temperatures," i.e., the upper

temperature limit at which the LTOP system must be enabled is not specified.

Some utilities have chosen to protect the P-i limit over its entire range; in

which case, the enable temperature is the P-T limit at the relief valve setting

for power operation (approximately 2400 psig). When this procedure is coupled

with a fixed LTOP setpoint, the impact of an upward adjustment of the P-T limits

may be to close the operating temperature window in the manner shown in Figure 2.

There is no regulatory requirement for the LTOP system to have a fixed low-

pressure setpoint device. Some plants, mostly Westinghouse NTOL plants, have

installed variable-setpoint electronics, which considerably enlarge the operat-

ing window.

The impact of Revision 2 was calculated for each plant in terms of both a

temperature increase, relative to P-T limits based on Revision 1 of Reg.

Guide 1.99, and a decrease in allowable pressure at startup. The basis for

these calculations was as follows. First, it was assumed that credible surveil-

lance data were not available and present P-T limits were based on Revision 1.

(For the few plants that have credihle surveillance data from 2 capsules, the

effect of a change for Revision 1 to Revision 2 will be small.) Second, the

fluence values used in the calculation were Ihose expected to be reached 5 EFPY

(effective full power years) after January, 198t the date for which fluence

values were given in the utilities' PTS submittal. Third, the adjusted refer-

ence temperatures were calculated at the 3/4 t location, because the P-T limits

for heatup were controlling. By doing so, the differences in calculative proce-

dures for attenuation of fluence through the vessel wall were factored into the

comparison.

A water temperature of 120'F was chosen for comparison of the allowable

pressures calculate using Rev. I and Rev, 2. This temperature is about a5 low

as one could expect to establish the thermal gradient corresponding to a chosen

heatup rate. The rate chosen was 25 0F/hr, a typical value for these startup

conditions. The corresponding temperature difference between the water and the

3/4 t location in a PWR wall was 15'F and KI thermal was 4.0 ksi %fin.

11/20/87 6

Table 1 gives the results of these calculations. For P-T limits based on

Revision 2 there are 15 plants for which the allowable pressure is 500 psig or

below, the lowest being 463 psig. For another 14 plants,. the allowable pressure

is between 500 and 550 psig. Eight plants are in the range 550-600 and another

8 plants are in the range 600 to 650 psig. These pressures are from 100 to

300 psig lower than calculated values using Revision 1.

The significance of allowable P-T limit pressures in the neighborhood of

500 psig can be seen in Figure 3. With a 200 psi AP required across the.pump

seals, comments from the WOG and CEOG indicate that the lower end of the operat-

ing window is between 250 and 300 psig after allowing for the system pressure

difference between the gage and the pump seal plus the gage error on the posi-

tive side. If the operating window is 100 psi, that leaves 100-150 psi for

overshoot, margin to avoid lifting the relief valve and gage error on the nega-

tive side. Comments from PWR owners and vendors did not specify an actual

limit for operability. A canvas of Tech. Specs. in 1982 by L. B. Marsh showed

that setpoint pressures ranged from 375 to 505 psig, with most of them in the

400-450 psig range. The AEOD analysis* of the Turkey Point 4 LTOP event in

November 1981, reported that their system required a minimum pressure of

375 psig to prevent excessive pump seal wear. The PORV setpoint was 415 psig,

.and the P-T limit pressure was 480 psig. It appears that allowable P-T limit

pressures below 500 psig will cause some delay in startup. P-T limits are less

restrictive for low heatup rates. However, from data given in EPRI NP-1139

Vol. 3, the average startup time from cold shutdown to hot standby conditions

is 20 hours or an. average heatup rate of 250 F/hr, from which it appears that

most plants could not achieve higher allowable pressures by lowering the heatup

rate. In the extreme case where the operating window was shut and the, reactor

coolant pumps (RCPs) could not be started at the prevailing temperature the

result would be as described in the WOG comments.

"The loss in terms of time for heatup corresponds to the heatup rate

achievable on pressurizer heaters and RHR pumps alone (estimated to

be typically on the order of 50 F/hr) versus the typical heatup rate

achievable with RCP's running (estimated conservatively low as 25F

*Wayne D. Lanning, "Low T .L ure Overpressure Events at Turkey Point

Unit 4" August 1983.

11/20/87 7

pe.r hour). For a shift of 50"F which results in a corresponding

segme:nt of plant heatup that must be covered without RCP's, the seg-

ment will take an estimated 10 hours with RHR pumps and pressurizer

heaters whereas it would have taken only 2 hours with RCP'S."

Table I also gives the change in RTNDT atthe 3/4 t location in the vessel

wall when calculated using Revision 2 instead of Revision I for the conditions

listed above. The median change is 65'F, with a range of 10 to 140'F increase

in. RTNDT for all plants. The temperature window of operation will be narrowed

by this amount for each plant. The greatest LTOP impact occurs at plants that

do not have an adjustable set point. Of course, those plants that use the

relief valve in the RHR system (or SDCS, shutdown cooling system) to provide

LTOP protection for the reactor vessel cannot raise the set point beyond the

pressure limit for the RHR system.

4.a.1.3 Remedies. To mitigate the impact on plant startup/shutdowii de-

scribed above, licensees may propose certain hardware or operational changes to

the primary coolant system or they may request regulatory relief from some re-

quirements, or a combination of both. With the exception of two B&W plants

(see below) none have submitted specific requests. The following discussion of

possible approaches is intended to show that there are ways to minimize the

impact of the upward adjustment of P-T limits that can be followed by each

licensee when action is required for a specific plant. It is also intended to

suggest some guidelines that the staff could adopt for the case-by-case review

of plant submittals.

Hardware and Operational Remedies

The most effective remedy is provision of a nitrogen or steam bubble in

the pressurizer to avoid operation in the water solid condition.whenever an

LTOP event could occur. Buying additional time for operator action reduces the

number of challenges to the low-setpoint relief systems and the bubble reduces

the probability of severe overpressurization if the relief system fails to func-

tion, All B&W plants have this feature, which (coupled with a nearly clean

record on LTOP events) has led to staff acceptance, in two cases, of LTOP

11/20/87 8

set-points that are somewhat above the P-T limit at low temperature. The basis

for this was that the probability of an LTOP event in those two cases was deemed

to be low enough to prevent its classification as an anticipated operational

occurrence, which is defined in 10 CFR 50, Appendix A as "...expected to occur

one or more times during the life of the nuclear power unit ...

Westinghouse and Combustion Engineering systems have traditionally been

water solid on startup for various reasons including oxygen control in the

reactor coolant and minimizing thermal fatigue in the pressurier 1urge piping,

A discussion of this whole issue is given by Lanning.*The most expensive modification is probably that required to install PORVs

with low, variable setpoints for some CE plants that have no PORVs should that

option be chosen. Generic Issue 84 addresses the need for PORVs in CE plants

that don't have them. Cost figures for such a system have been reported 7n

NUREG-1044, "Evaluation of the Need for a Rapid Depressurization Capability for

Combustion Engineering Plants." The installation of 2 PORW , 2 block valves,

a quench tank and necessary piping and instrumentation was estimated to cost

$2,300,000 for a new plant and $4,300,000 for an operating plant. This system

solves the problem of narrowing the operating window in temperature, and is

probably all that would be needed for some years, until the operating window in

pressure became too narrow at temperat -s at which the reactor coolant pumps

were to be started. For plants that ha\- existing PORVs, the costs of providing

a low setpoint system with vari-able setpoint should be less than the above and

the exposure to radiation would be much less if no piping changes were required.

Possible Regulatory Relief

The first step toward granting regulatory relief is to define what is

meant by the phrase "while operating at low temperatures" in Standard Review

Plan 5.2.2 and Branch Position RSB 5-2. This is being done by addition a new

Branch Position B.2, which states:

"The low temperature overpressure protection system should be operable

during startup and shutdown conditions below the enable temperature,

*Wayne D. Lanning "Low Temperature Overpressure Events at Turkey Point Unit 4,AEOD Draft Case Study Report, August 1983.

11/20/87 9

defined as the water temperature corresponding to a metal temperature

of at least RTNDT + 90OF at the beltline location (1/4t or 3/4t) that

is controlling in the Appendix G limit calculations."

The corresponding pressure is about 1600 psig in a typical vessel. This crite-

rioh will give ioughly 60OF relief on the enable temperature used by those

plants that protect the entire P-T limits up to the PORV setpoint pressure

(-2400 psig) for normal operation. This proposal does not reduce any previously

drawn egulatory safety margin, because the phrase "at low temperatures" was

not defined in Std. Review Plan 5.2.2. For the plants where operating flexi-

bility is reduced because the window in temperature is reduced, this criterion

would put the enable temperature above 200%. Only one of the 30 LTOP events

tabulated in the original studies of this issue were above 200'F. The median

temperature was 1300F.

A second option for granting regulatory relief involves reduction of margin

in calculating the required set-point pressure for LTOP systems. This is needed

where operating flexibility is reduced at startup by a narrowing of the window

in pressure. It would be necessary to maintain P-T limits as presently calcu-

lated and to enforce them administratively as is presently done. Discussion of

this subject with the Westinghouse and Combustion Engineering Owners Groups has

been underway since July 1986 with attention focused on the possibility of

justifying the reduced margin. The staff has concluded that justification must

consist of demonstration that the probability of an LTOP event for a particular

plant is low enough that it need not be considered an anticipated operational

occurrence for rpsons of design or operational procedures, particularly those

that avoid operdtion in the water solid condition. A "clean" record over many

years Dn LTOP challenges to the safety system in that plant or that category of

plants would also be significant.

It is important to note that the requirements of Appendix G, 10 CFR 50

would not be changed, i.e., the P-T limits would not be changed And, if they

are not protected by an automatic-relief system, but only by administrative

requirements, there will be more violations of P-T limits. Licensees should

therefore not make use of the higher set pcint unless the impact on startup is

severe.

11120/87 10

The acceptability of this second option will be decided on a case-by-case

basis. In addition to justifying the estimate of probability of occurrence,

the proposed set point pres-ure must be justified from the standpoint of

fracture preven.ion.

Recovery of the hardware costs through redL vd time for startup/shutdown

operations will be a factor in the licensee's decision. The time saved by

opening the windows in pressure and temperature is plant specific, depending

on how severely operating procedures are impeded and how much relief is

achieved by the hardware change that is being considered.

The relationship of Revision 2 to parts of the Standard Review Plan (SRP)

was reviewed with the cognizant branches of the Office of Nuclear Reactor Regu-

lation. The changes to SRP 5.2.2, "Overpressure Protection," have been

described. Some proposed changes to SRP 5.3.2, "Pressure-Temperature Limits,,"

given in Enclosure /, include the editorial changes needed to reference Revi-

sion 2 plus a paragraph to be added to assure coordination between the Materials

Engineering Branch and the Reactor Systems Branch, who have primary responsibil-

ity for SRP 5.2.2. This SRP clarification assures that the staff position in

Branch Technical Position RSB 5-2, "Overpressure Protection of Pressurized Water

Reactors While Operating at Low Temperatures," regarding the LTOP setpoints is

consistent with the Pressure-Temperature limits (based on Appendix G, 10 CFR 50)

throughout the plant lifetime. This modification does not represent any change

in the staff requirements.

4.a.1.4 Conclusion

1. None of the comments from the industry called for changes to the basic

calculative procedures of Revision 2, despite the fact that there is some impact

on plant operation.

2. The impact is aggravated by LTOP requirements, and some relief has

been provided in terms of clarification and relaxation of those requirements

combined with possible hardware changes.

3. Resolution of this issue will be on a plant specific basis.

ý4.a.2 Boiling Water Reactors

For BWRs the impact of using Revision 2 as the basis for calculation of

pressure-temperature limits has been reported by the BWR Owners Group (BWROG)

as discussed in the Analysis of Comments, Enclosure 2. Pressure-temperature

11/20/87 11

limits will move upscale 50-100'F for some plants. For normal operation, the

impact of this change is small. There being a vapor space in the reactor vessel,

he.atup/cooldown follow the saturation curve, which is well below the P-T limit.

For pressure tests, however, there is a serious impact, according to the BWROG.

The impact on BWRs derives from two factors. Pump heat is used to raise

the temperature, and the rate of temperature rise possible from this source is

only a few degrees per hour at the higher temperatures. If the leak test is

on the critical path for restart, this prolongs the outage, and the lost power

production is costly. Second, there is a requirement in plant Technical

Specifications that when coolant temperature exceeds 2001F, the dry well head

must be in place, the containment must be closed and the safety systems must be

operable. Therefore, the pressure test must be done at the end of the outage

when the time to run the test and the time to fix leaks adds to the outage.

Moreover, the visual inspection for leaks is virtually impossible inside the

drywell and difficult and time consuming at other locations.

Table 2 lists the 35 BWRs in order, starting with the highest required

leak test temperature calculated for P-T limits that would be good for 5 years,

using Revision 2. The Table is taken from comment letter No. 22, S. Ranganath,

General Elec. ic Company. Six of those plants will have to do the leak test

above 2000F, the highest being 2556F, assuming all of the assumptions in the

calculations are correct. At end of life, the number rises to 19, or more than

half.

4.a.2.1 Regulatory Background. The NRC requirements for leak test '.nd

hydrotest temperatures are contained in 10 CFR 50, Appendix G, which incor-

porates by reference Appendix G of the ASME Boiler and Pressure Vessel Code.

The NRC also endorses Section XI c~f the Code, which describes what pressure

tests are to be conducted in service. Leak tests, done at operating pressure

,nominally 1000 psig), are required following opening and reclosing of a com-

ponent (e.g., replacing the reactor vessel head after refueling). Hydrotests

at nominally 1100 psig are required by the Code following certain repairs and

as part of the 10-year 1SI.

When there is no fuel ii the reactor vessel, both Appendix G of the Code

and Appendix G, 10 CFR 50 require that the temperature for hydrotests be not

lower than RTNDT + 60 0 F. Subsequent to loading fuel, the required test tempera-

ture is given by the fracture mechanics calculation described in Appendix G of

11/20/87 12

the ASME Code. For a leak test pressure of 1000 psig, the required temperature

is about RTNOT + 95'F for a typical BWR reactor vessel.

With regard to the flange regions of the vessel, which are highly stressed

by bolt preload, 10 CFR 50, Appendix G requires a temperature of RTNDT + 90OF

for hydrotests and RTNDT + 120'F for normal operation. In this case, RTNDT is

that of the material stressed by bolt preload, which receives almost no radiation.

For pressures below 20 percent of the preservice system hydrostatic test pres-

sure (about 310 psig) the required temperature is RTNDT.

Another benchmark can be found in the ASME Code, Section III requirements

for piping pumps and valves over 2 1/2 in. thick. Unless a lower value can be

justified by using Appendix G, the lowest service temperature shall be

RTNDT * 100 0 F.

Yet another benchmark from ASME Code requirements can be found in the

requirements for Class 2 vessels given in Paragrdph NC-2331 and Appendix R.

For a 6-inch thick vessel, typical of BWR reactor vessels, the required lowest

service temperature is 62'F above the drop weight NOT. By comparison, it seems

reasonable to require a temperature of RTNDT + 95%F for the hydrotest of a

Class I reactor vessel that contains fuel for which residual heat removal is

required.

The foregoing discussion shows that there are several regulato-y pressure-

temperature requirements, which are fairly consistent with each other. There-

fore, to reduce the requirements for pressure tests would mean also reducing

the requirements for other pressure-temperature limits.

4.a.2.2 Analysis 6f Margins. At a meeting on December 9, 1986 with NRC

staff from the BWR Licensing Division and the author, the BWROG proposed that

the impact be reduced by reducing the margins for pressure tests. Specifically,

it was proposed to substitute the Kic curve from Section XI of the ASME Code

for the KIR curve now used. The effect of this change would be to reduce the

required metal temperature for a 1000 psi leak test following refueling from

about RTNDT + 95%F to about R1NDT + 35'F, a reduction sufficient to permit

almost all of the BWR's to do the leak test at temperatures below 200'F.

When the NRC staff questioned the safety implications of pressure testing

at RTNDT + 35%F, the BWROG representatives expressed a willingness to set a

lower cutoff at RTNDT + 600 F. However, they did point out that testing at the

11/20/87 13

6A

lower temperature made the pressure test a more severe proof test and was in

that sense more beneficial to safety in subsequent operation of the vessel.

The staff disagreed with the view that the pressure tests constitute proof tests.

In written comments to the BWROG on the meeting the staff said: "despite the

fact the the Code-required hydrostatic test is intended to reveal gross negligence

in a repair operation, the principal use of the hydrotest (1100 psig) as well

as the leak test is to perform a Visual leak test at a pressure equal to or

slightly higherthan operating pressure. Given the history of pipe cracks in

BWRs, the staff feels that the visual inspection for leaks during pressure tests

is very important. Although the hydrotest may provide some information on

reactor vessel integrity, it is doubtful that hydrotests serve as a proof test,

ard they surely do not serve as or substitute for adequate inspection when

assessing future reactor pressure vessel integrity. Prevention of failure of

the vessel during these pressure tests is an overriding consideration, and

therefore, the required test temperatures are set on that basis."

To given full consideration to the basic question: "Is there an acceptable

basis for reducing present safety margins for pressare tests," a panel of expe-

rienced people was assembled for a day's discussion of the issue. Panel members

were: Dr. Frank Loss, Materials Engineering Associates, John Merkle, Oak Ridge

National Laboratory, Edward Wessel, private consultant, Warren Hazelton, Mate-

rials Engineering Branch, Division of BWR Licensing, NRR, and M. Vagins and

P. N. Randall of the Materials Branch, DES, RES. The basis for selection of

the members was: expertise in fracture mechanics and fracture control pro-

cedures, extensive background in development of ASME Code requirements in this

area, and availability under present contracts for a prompt, "one-meeting"

effort to provide advice on the subject.

The issue raised by the BWROG was discussed from several angles: (a) the

basis for the traditional safety margins that are given in terms of NDT + 30,

60 or 1200; depending on the application, (b) the basis for each of the factors

in the LEFM (linear elastic fracture mechanics) approach followed in Appendix G

(allowable toughness, assumed flaw size, and margin on stress intensity factor),

and (c) the consequences of a reactor vessel failure during hydrotest. The

Panel noted that traditional margins, which are used in certain places in

Appendix G, 10 CFR 50, as well as in the ASME Code requirements for Class I Pip-

ing Pumps and Values and for Class 2 vessels, originated from analyses of World

11/20/87 14

Q

War 11 Liberty ship failures by the Bureau of Standards and ship and pressure

vessel failures by the Naval Research Laboratory, which led to the development

of the Irop weight test and its correlation to service failures and explosion

bulge . 2sults. The "yield criterion," as it was later called, was

NDT + 60', )r 1 in. thick material meaning that nil-ductility temperature from

the drop weight test should be 60OF below the service temperature if yield

before fracture was the desired criterion. Rules. for thick sections (and largC

flaws), derived from service failures of turbine rotors and later from the early

HSST test program done at NRL, lead to a yield criterion of NDT + 120OF for

heavy sections, 6 inches and over. The panel agreed that the basis for these

traditional margins was the set of assumptions: stress levels near yield when

all sources were considered, high strain rates arising from dynamic loading or

from initiation of a running crack emanating from a pop-in type flaw, and con-

siderable uncertainty about the size of flaws and the locP, toughness of the

metal surrounding the flaw.

It is interesting to note that the same variables were di,.cussed in more

detail in the context of an examination of the factors that make up the LEFM

approach called for by Appendix G, 10 CFR 50. The BWROG had raised the issue

of high-strain-rate loading by recommending that the allowable toughness values

be taken from the KIc curve of Appendix G, which is for static loading. (The

KIR curve was drawn to bound all valid data from static, dynamic and crack-

arrest tests that were available for reactor vessel materials at the time

(1971)). Extensive panel discussion about the need to use the KIR curve

centered on the probability of existence of a pop-in flaw. Two types were

considered. A local brittle region could be caused by improper weld stre-s

relief or by allowing segregation of elements that enhance radiation embrittle-

ment, but these were considered unlikely. Multiple flaws in close proximity,

which might link up at. some stress level and form a running crack were con-

sidered more likely. The absence of information from non-destructive examina-

tions of the BWR vessel beltline was a major stumbling block, as discussed below.

The panel also noted that a Iwer-bound KIc is difficult to determine because

of scatter and size effects including the variation of toughness along an

extended crack front, and the use of K IR is often justified as a lower bound of

static toughness measurements that has been shown to include recent test data.

In the face of the uncertainties, the panel was unwilling to accept the use of

the Klc curve for the calculation of P-T limits.

11/20/87 15

In considering the basis for the 1/4 t x 1-1/2 t semielliptical flaw assumed

in Appendix G calculations, the Panel felt there was no basis for considering

relief in this assumption, because:

(a) the beltlines of BWR vessels are not, given inservice inspection by

ultrasonic means (UT), because it is impractical, and

(b) even if they were inspected, the ability to declare a beltline free

of defects above a certain size based on presently required UT methods

is much in doubt. It was recalled that the use of acoustic emission

during hydrotest had been considered some years ago, but nothing came

of it.

Concern was expressed over the possible presence of stress corrosion cracks

in the base metal, perhaps extending from cracks in stainless cladding that had

been sensitized by the stress relief treatment given the vessel. Service expe-

rience with cracking in the steam generator head at Garigliano and in the steam

generator shell closure welds at Indian Point and Surry tended to make Panel

members cautious, although they felt there was not much chance that the reactor

vessels would have been stress relieved at too low a temperature, as were the

steam generator welds.

There was considerable discussion of the possibility that previous operating

pressure-temperature conditions would reduce the severity of cracks by a warm

prestress mechanism. It was concluded that there were too many uncertainties

about path effects of the unloading/cooli-g sequence to make use of this factor.

Moreover, it was recalled that earlier tests of warm prestress effects in large

spin test specimens of turbine rotor material did not give the improvemnent

expected, based on tests of small specimens.

The safety factor of 1.5 on KI membrane for hydrotest was discussed only

briefly. it covers the neglect of weld residual stress, cladding stre~sss and

other uncertainties in the calculated KI values. There seemed to be no basis

to reduce it.

In summary, the Panel concluded: especially in the absence of inservice

inspection of the BWR vessel beltline., and considering the consequences of a

failure with spent fuel in the vessel, no basis for reduction of margins for

hydrotest had been found.

11/20/87

4.a-2.3 Other Means For Reducing the, Impact on BWRS. In written comments

to the BWROG the staff has solicited information on three means to mitigate the

impact of implementation of Revision 2.

1. Comments were solicited from the BWROCG on the safety implications and

operational aspects of a relaxation of lech. Spec. requirements that the drywel,

head be in place, containment secured, and safety systems operable when

temperatures exceed 200*F, for system leak and hydrostatic pressure tests con-

ducted when the vessel is water solid and all control rods are fully inserted.

2. Comments were also solicited from BWROG on the following as an alter-

native to present system leak test procedures: following present pressure-

temperature requirements (10 CFR 50. Appendix G),, pressurize to the highest

pressure attainable at 200OF (but not over Code requirements, of course) and

perform the visual inspection at that pressure.

The BWROG was asked to address both the safety implications and the opera-

tional aspects of leaks that might be overlooked at the reduced pressure but

yet show up when the system is brought to full pressure during startup. Relief

in this form is not contemplated for s-tuations where a nydrostatic testis

required following a repair or by the required 10-year hydrotest.

3. Finally, it was asked if the BWROG would include in their impact anal-

ysis of Revision 2, a detailed, realistic analysis of the use of the auxiliary

heat sources during pressure tests to reduce heatup time, especially when

required to reach higher temperatures.

The reply from the BWRDG, dated May 20, 1987 included NEDC-31363, which

is Part 2 of their evaluation of t• Revision 2 impact onBWRs. Part I was

NEDC-31140. In response to item 1.(the 200OF limitation) the BWROG stated

S... relaxation of plant technical specifications to allow the dry-well head

to be removeo during a BWR pressure test (even at temperatures exceeding

200'F)... will provide relief during pressure testing..." In support of their

statement ". .. we know of no adverse safety consequences resulting from this

relief" they provided a systems evaluation (Enclosure 3 to their letter of

May 20, 1987) that listeo the systems conditions that should accompany the

relaxation. However, after further consideration.. te staff is not willing tv

grant relief on the requirement that containment should be closed when the

temperature exceeds 200'F.

11/20/87 17

In response to item 2, testing at a pressure below that required by the

AW.E Code, but consistent with the 200QF limitation, the BWROG supported the

idea but not-d that it may be necessary to implement ASME Code changes. This

being so,. this remedy is long-range and will be Jiven further study by the

NRC members on the appropriate ASME Boiler and Pressure Vessel Code working

groups and above.

In response to item 3, cost of an auxiliary boiler, the BWROG provided

two case studies for BWRs, one, actual cost figures and one a conceptual study,

plus the results of a utility survey of delays in heatup time if higher test

temperatuves were required. These are analyzed in Section 5.(5).

4.a.2.4. Conclusion.

I. Thei.• has been no suggestion from BWR owners that the impact on BWRs

that will be caused by use of Rerision 2 should be mitigated by making changes

in Revision 2.

2. The safety margins for pressure tests, which appear in Appendix G,

10 CFR 50 and Appendix G of the ASME Code, are justified and provide a safety

benefit that is the overriding consideration.

3. There will be a cost impact from longer neatup time. Utilities can

reduce the time by installing an auxiliary heat source when its cost is justi-

fied by the time saved. (See staff analysis in Section 5.(5).)

4.a.3 Other Applica4 ion, of Regulatory Guide 1.99

Regulatory Guid& 1.9S is used whenever RTNDT must be calculated as part

of an analysis of a transient that has actually occurred. The artalysis provides

the basis for deciding if the possibility that the vessel has been damaged is

sufficiently high to warrant an inspection before returning it to sevice. Such

inspections are time consuming and costly in terms of power replacement.

Another application of Regulatory Guide 1.99 is in the analysis of flaws

found by inservice inipection of the reactor vessel beltline. A recent example

occurred at Indian Point 2. First, t draft Revision 2 was used to calculate

RINDT at the inside surface, based on the f.luence and the weld and plate chemi-

stry. Second, because the flaw was near the outside surface, the formula for

attenuation of ARTNDT through the vessel wall, (a feature of Revision 2), wasused to -alculate ARTNDT at the tip of the flaw. The evaluation is the basis

for deciding if the vessel must be repaired before being put back in service.

11/20/87 18

.J

In evaluations of transients and flaws there may be significant safety

questions; the cost impact may be high, and contention over the decisions may

surface. Clearly, it is important to have a published basis for the calcula-

tion of radiation embrittlement effects.

4.a.A Risk Avoided by Using Revision 2

A detailed evaluation of the risk avoided under normal operation and during

transients is given in Section 5, Backfit Analysis, Subsection (3).

4.a.5 Costs to NRC.

See Section 5, Backfit Analysis, Subsection (7).

4.b. Impact On Other Requirements

4.b.1 Impact on the Pressurized Thermal Shock (PTS) Rule (10 CFR 50-61)

When the subject Regulatory Guide 1.99, Revision 2 was submitted to the

CRGR for review on July 24, 1985, (prior to publication for public comment),

considerable concern was expressed regarding its effect on the recently promul-

gated PTS rule. CRGR's concerns were answered in a joint memorandum to the EDO

from the Director of NRR and the Director of RES, dated August 12, 1985 (see

Appendix A). That memorandum provides a discussion regarding the background

and bases for the PTS rule's RTPTS correlation, as well as the background and

bases for the RTNDT correlations in Reg. Guide 1.99 Rev. 1 and 2, and it pro-

vides a discussion of reasons why the correlations are different. The memo-

randum concludes with the staff's plans to deal with the differences. It

states that the PTS rule will remain as it is (with the RT correlation which

is different from the RTNDT correlation contained in the subject proposed Reg.

Guide Revision) during the public commentperiod for the proposed Guide. It

further states:

"Following resolution of comments, once general agreement is reached

regarding the best way to calculate RTNDT, then it will be appropriate to

re-evaluate the overall conservatism of th- PTS rule, and consider whether

amendment of the PTS rule is desirable. At that time, we will have avail-

able the plant-specific materials and fluence values that the PTS rule

11/20/17 19

requires all PWR Licensees to submit early in 1986. This will allow us to

much more accurately assess the effects on all plants that WLýUld result

from a changeover to the Rev. 2 method."

We are now at the time mentioned in the above memorandum: -we have reached

agreement regarding the best way to calculate RTNDT (as described throughout

this paper). And we have in hand the plant-specific material and fluence values

from the PTS-rule-required submittals. This information has allowed us to

determine the effects on all PWR plants of a changeover to the "Rev. 2" method

shown in Table 3. The effects are the result of differences between the RTNDT

correlation used in Rev. 2 of the Reg. Guide (described throughout this paper)

and the PTS rule's RTPTS correlation, in which: (1) weld and base metal data

were analyzed as one data base yielding one correlation function; (2) the flu-

ence function was of simpler form; and (3) there was a second equation which

gave bounding values. As shown in Table 3, changing the PTS rule to require

use of the "Rev. 2" correlation would cause over half of the plants for which

RTPTS exceeds 2006F by end-of-license to see an apparent decrease in RTNDT,

i.e., an increase in perceived safety margin. However, as summarized in

Tab-le 4, it would also cause up to 5 plants to fail to meet the screening cri-

teria before end-of-license. One of the 5 plants fails to ,,eet the criteria

according to the present PTS rule and one uther plant fails to meet the criteria

under the present rule but meets the criteria if Rev. 2 'ý used. One of the

5 plants will reach this screening limit in the 1988-1992 time period, depend~ig

on the efficiency of the proposed flux reduction program, if Revision 2 is used.

Another will reach the screening limit in 1993. using the Rev. 2 method, and one

in 1997. The others will reach the limit after the year 2000. Note from

Table 3 that the 4 plants that are most adversely affected show an 'increased

RTPTS" of 68, 67, 63, and 52F0 , respectively. Note also that these numbers may

change when the final reviews of chemistry and fluence are complete and as a

result of extensions of license life to O.L. + 40 from C.P. + 40, (i.e., to 40

years after the operating license date from 40 years after the construction

permit date). Fluences at end of life are estimates, and some may well increase

enough to cause the screening criterion to be exceeded. For example, there are

5 plants listed in Table 4, in addition to the two discussed above, for which

RTpTS calculated by the PTS Rule falls within one degree of the screeningcriterion.

11/20/87 20

In the opinion of the NRC staff, the above quoted changes in the non-

conservative direction are greater than can be absorbed by the uncertainties

believed to exist and taken into account by the staff when the RTNDT-based

screening limit was set. Based on this new information (i.e., that RTNDT

values nay actually be over 60OF higher than previously thought for 3 plants)

it is not possible to justify the previous conclusion that risk due to PTS is

acceptable so long as RTPTS (evaluated by the method given in the PTS rule)

stays below 2700F. Moreover, consistency in the calculation of radiation

embrittlement throughout the NRC's regulatory documents is important. It is

necessary to remove this newly discovered non-conservatism by incorporating the

new (and believed "better" or "more pertinent") RTNDT correlation into the PTS

rule. The immediate action required of the plants that would then be caused to

exceed the screening limit before end-of-license would be to evaluate feasi-

bility of further flux reduction and report such flux reduction plans to the

NRC. The long term action, to be taken at least three years before reaching

the screening criterion, involves performing a plant-specific probabilistic

analysis foll(.wing the guidance given in Reg. Guide 1.154 to justify operating

beyond the screening criteria. (Only two plants will need to do this in the

near future (-1990)).

Once Revision 2 is approved by the CRGR, the Division of Engineering, RES,

will request EDO approval to initiate the following chang to the PTS Rule-

replace the RTpTS correlation presently in the PTS rule with a new one that

is identical with the RTNDT correlation in Reg. Guide 1.99, Rev. 2. When

approved, the Division of Engineering will prepare for formal office concur-

rence a resubmission to the CRGR of the above information in the proper format

asking for permission to submit the proposed PTS rule change to the Commission

so that it can be published for Public Comment.

An anticipated public comment regarding that proposed correlation change

is that the RTPTS-based screening limit specified in the PTS rule was signi-

ficantly influenced by the RTPTS correlation then in use, and changing that

correlation to a different one wo Id necessitate a change in the PTS screening

limit. We co not plan to make such a change, for the following reasons (this

same discussion will be reiterated in the CRGR package that will propose chang-

ing the RTPTS correlation).

11/20/87 21

The NRC staff selected the PTS screening limit after a careful consideration

of:

(1) The operational PTS events that had occurred in U.S. PWRs, parti-

cularly including the eight most severe such events, considering

their demonstrated frequency and their severity; and

(2) Results of some preliminary PRA analyses, to account for risk due

to PTS events that could be more severe than those that have occurred

(but less frequent and, therefore, not represented in the operating

experience).

The staff considered both types of information, considered the degree of

uncertainty involved, and considered the amount of margin desired (due partly

to the perceived uncertainty) to achieve acceptable risk. The result was

selection of the present screening criterion which was not selected in a

cotally quantitative way but rather involved considerable educated judgment.

The RTPTS correlation change would not affect the operational experience

data base, and it would not affect the judgment regarding uncertainties. The

change would somewhat affect the results of the PRA analyses used in the screen-

ing limit selection process, and would affect the results of later confirmatory

PRA analyses. The VISA Code Sensitivity Studyl, 2 contains a study of through-

wall-crack frequencies calculated for three transients for each of three differ-

ent plants, using the PTS rule correlation (the "std" correlation) and using a"weld" correlation which is substantially identical to the Reg. Guide 1.99,

Rev. 2 correlation. Section 4 of that reference shows some of the available

•esults in figures 7 through 9. However, those figures are incomplete and

contain a significant error. Complete, corrected results are included here on

Figures 4 through 6. These Figures show that the.effect of a correlation

change is a shift in the "equal risk" RTNOT that ranges from 10 to 20%F for

Oconee 1, from 15 to 40'F for Calvert Cliffs 1 and from 20 to 35%F for H.B.

Robinson 2. Because the exact amount of the change varies greatly from

1E. P. Simonen, R. I. Johnson, and F. A. Simonen" Vessel Integrity Simulation

Analysis (V'SP) Code Sensitivity Study," NUREG/CR-4267, December 1985.2 Communicatio'is from E. P. Simonen

11/20/87 22

should be represented in the surveillance capsules. However, there should not

be any extra cost involved.

4.c. Constraints

We have not identified any constraints such as scheduling or enforceabil-

ity that affect the implementation of Revision 2.

5. BACKFIT ANALYSIS FOR POWER REACTORS T _ l Cr. ýu •r

Revision 2 to Reg. Guide 1.99 fits the definition of backfitting in that

it is a regulatory staff position interpreting the Commission rules that is new

and different from a previously applicable staff position. Revision 2 provides

substantial increase in the overall protection of the public health and safety

at costs that are justified in view of the increased protection. These factors

have been discussed in Section 4.a.1 and 4.b. for PWRs and in Section 4.a.2 for

BWRs. A summary of that discussion will be repeated here as part of the dis-

cussion of risks averted by using Revision 2. The format is that defined by

10 CFR 56. 109, paragraph (c)(1) through (c)(9).

(1) The specific objective of Revision 2 is to provide an updated basis

for estimating the extent of radiation damage when calculating pressure-

temperature limits for reac.tor vessels, and when evaluating vessel integrity

fallowing transients such as LTOP and PTS events or following detection of a

flaw in the beltline.,

(2) When Revision 2 is implemented, mosf nsees will have to recal-

culate their pressure-temperature limits and thereafter operate with whatever

constraints that imposes. Owners of PWRs may have to change the setpoint of

the low-temperature overpressure protection system. Most BWR owners will have

to perform leak tests and hydrotests at higher temperatures.

(3) The reduction in risk of fracture of the reactor vessel produced by

implementation of Revision 2 derives from several sources as discussed under

the following headings.

Normal Operation - PWRs

There is a risk of vessel failure during a heatup!coo'down operation if

the P-T limits followed in that operation are lower then they shouId he. This

1/20/87 24

increment of risk averted by using Revision 2 instead of Revision 1 was quan-

tified in a study by Pacific Northwest. Laboratories (PNL) included in the

Regulatory Analysis done for the For-Comment version of Revision 2. Based on

the increased risk, PNL calculated the public exposure to radiation as a

consequence of vessel failure, and calculated the costs resulting from a change

to Revision 2 as the basis for P-T limits. The PNL report is Enclosure 5.

The PNL study considered one of the plants for which the P-T limits based

on Revision 1 are 100'F below those based on Revision 2. If a plant in that

situation continued to operate with P-T limits based on.Revision 1, and if the

operator followed the limits closely in a heatup-cooldown sequence, there would

be greater probability of fracture of the vessel.

The change in the probability of fracture(Revision 1 - Revision 2) was

. calculated using a Monte Carlo technique and VISA code, which originated at the

NRC and was further developed at PNL. In this analysis, initial RT NDT tough-

ness, copper content, fluence, and flaw size were treated probabilistically.

Actually, in most runs, the flaw size had to be treated as fixed at a large

value (a 2 in. deep continuous flaw, probability of one) to get the Monte Carlo

procedure to produce failures frequently enough to keep the required total runs

to a reasonable number. Then the probability was reduced by a factor of 3500

to account for the more realistic ilaw size distribution normally used. This

factor was the ratio of the results qiven by two VISA runs: the first with the

adjusted flaw size and the second with the original flaw size distribution.

The conclusion reached by PNL, based on the probabilistic fracture mechanics

analysis, was that the best estimate of the increase in probability of vessel

failure resulting from following P-T limits .that were I000 F too low was 2.5 E-7

per heatup/cooldown cycle.

The PNL estimate of 6 startup/shutdowns per plant, year was based on reactor

scram data available to PNL. However, after a scram, the plant does not neces-

sarily go to cold shutdown. In the plant operating data given in NUREG-0020,

shutdowns greater than 72 hours are tabulated for each month. After correcting

for the refueling shutdown, which extends over several months, the average over

a three-year period, 1982-4 was 4.3 shutdowns greater than 72 hours per reactor

year. From an ORNL analysis of plant operation for 1982, there were approxi-

mately 2.0 shutdowns per i.eactor year greater than 120 hours. Based on these

data, and assuming that some of the 72-hour shutdowns were not cold shutdowns,

11/20/87 25

it seems reasonable to assume that there are approximately three startups from

cold shutdown per reactor year.

The risk averted per reactor year would thus be 7.5 x 10-7 for those

ratcheted 100OF (i.e., for those plants having P-T limits based on Revision 1

of RG 1.99 where, for reasons of chemical composition and fluence level, the

limits based on Revision 2 would be 1000 F higher). No probabilistic analyses

were calculated for plants ratcheted by lower amounts, so it has been assumed

that for the 44 PWRs ratcheted by 506F or more the risk averted was 5 x 10-7

per reactor year. Actually, Table 1 reveals there are 3 PWRs ratcheted 100F

or more, the maximum being 139*F. There are 41 plants ratcheted between 50 and

99°F and 17 plants ratcheted 8-49%F. Neglecting the latter group, the total

risk averted is about 2 x 10-5 per year for the reactors now in operation or a

total risk reduction of about 6 x 10-4 for these plants during their lifetime.

Using the techniques described in Enclosure 5, PNL calculated the resulting

man-rem avoided to be 11,000 man rem for the lifetime of these plants.

Cost estimates by PNL for implementation of Revision 2 were based on the

costs of replacement power, assuming that plants ratcheted 50-100*F would take

2 hours longer per heatup/cooldown cycle. Public comments did not provide any

support for a general plant-average estimate of the delay, and the PNL estimate

is regarded as a very arbitrary one. Based on the costs and the 11,000 man rem

avoided, PNL concluded that the cost per man rem avoided (approx. $5000) was

too high to warrant a recommendation that Revision 2 be implemented. However,

because the only source of risk considered in the PNL study was that incurred

in normal operation (heatup/cooldown) and because a good basis for the cost

estimates was lacking, the decision was made in July 1985 to publish Revision 2

for public comment.

Rather than pursue additional funding and further delay implementation of

Revision 2 to update and expand the PNL study, sufficient rationale exists to

warrant a recommendation to go forward with Revision 2. Risks not considered

in the PNL study are those caused by accident conditions (LTOP transients and

PTS transients) where improper knowledge of the state of embrittlement of the

vessel may delay corrective operator action, plus the risks associated with

evaluations of vessel integrity after a transient has occurred or a flaw has

been detected. These sources of risk are treated in the following paragraphs.

In particular, discussion regarding the need for the Guide revision to reduce

PTS risk at several plants is presented in sufficient detail to justify

issuance of Revision 2.

11/20/8.7 26

Hydrotests and Leak Tests--BWRs

From the discussion given in Section 4.a.2 there is risk averted by using

Revision 2 as the basis for calculating the required temperature for BWRs during

pressure testing. Based on estimates reported in the comment from the BWR Owners

Group (NEDC 31140), 8 BWRs will be ratcheted 50-100OF and 21 BWRs 30-50'F. The

number of pressure tests per year is approximately 1. Assuming the risk averted

per B1WR presure test is similar to that calculated by PNL for a PWR heatup/

cooldown, the total risk averted for BWRs is somewhat over 8 x 2.5 x 10-=

2 x 10-6 Per year, a small addition to the total for PWRs of 2 x 10-s for normal

operation.

LTOP System Transients--PWRs

For PWRs, there is a considerable history of low temperature overpressuri-

zation (LTOP) transients as discussed in Section 4.a.1.2. From data collected

by Ed Throm, Task Manager of Generic Issue 94, the probability of a challenge

to the low setpoint relief valve system is dbout 0.1 per reactor year. The

probability that the relief system is unavailable is about 0.1 per reactor

year. Consequently, the probability that an LTOP event will exceed the P-T

limits is about 0.01 per reactor year. For the 60-70 reactors, this amounts to

about 20 LTOP events over their remaining 30-year lifetime. To obtain the

probability of vessel failure resulting from an LTOP transient, one must esti-

mate the pressure attained before operator action terminates the transient,

vessel temperature reative to RTNDT and the size of defects present, all in

probabilistic terms such as. those used in the VISA Code.

To calculate risk averted in LTOP transients if the P-T limits were based

on Revision 2 instead of Revision 1, one must estimate the probable reduction

in peak pressure provided by more prompt operator action. From Tab•' I it

appears that the P-T limit in the temperature range of interest w-k be 100-

350 psi lower, using Revision 2.

We assume that a drop in *the P-T limit of 100-350 psig (Revision 2 relative

to Revision 1) causes the same beneficial reduction in peak pressure, based on

the assumption that operator response begins as the P-T limit is approached in

11/20/87 27

either case. From an early probabilistic analysis of LTOP events* Figures 5-1

and 5-2 have been cross-plotted (See Figure 7) to show the decreased probabil-

ity of failure per event as a function of peak pressure.

Reading from the upper part of Figure 7, an order of magnitude decrea5e in

probability of vessel failure from 10-1 to 10-1 per event is produced by reduc-

tions in peak pressure ranging from 120 psi for the most severe conditions (a

low temperature event occurring in a highly embvittled vessel) to 360 psi for

the least severe case examined. In rough terms, therefore, implementation of

Revision 2 will reduce the probability of vessel failure by an order of magni-

tude for perhaps 20 LTOP events in the lifetime of the present population of

PWRs. To give absolute numbers would require analysis of the probable tempera-

tures and susceptibilities to fracture of these reactors.

PTS Transients -- PWRs

Pressurized thermal shock events represent the other significant category

of transients (in addition to LTOP events) that threaten reactor vessel integr-

ity and for which the analysis requires knowledge of the reference temperature,

adjusted for radiation. As indicated in Section 4.b.1, there is increased risk

in operation if the sensitivity of the reactor vessel to thermal shock has been

underestimated. Some plant operating procedures contain information about a

"PTS region" on the P-T diagram that is to be avoided by reducing pressure if

the cooldown rate exceeds the Technical Specification limit.

An astimate of the risk avoided by using Revision 2 as the ba-is for amend-

ing the PTS rule can be made from Figure 1 of Regulatory Guide 1.154 (the "PTS

Guide"), which is a semilogarithmic plot of frequency per reactor yearversus

mean surface RTNDT. In the.frequency range 10-6 to 10-4, the slope of the curve

labeled "PRA total" is about 45 0 F per factor of 10 in frequency. From

Table 111, there arE 3 plants for which the increase in RTPTS if calculated by

Revision 2 instead of by the present PTS rule exceeds 60'F and 6 more plants

for which the difference is between 20 and 42%F. For these 9 plants, therefore,

the risk avoided would range from a factor of 3 to a factor of 30. Figure 8

*R. M. Gamble and J. Strosnider, Jr., "An Assessment of the Failure Rate forthe Beltline Region of PWR Pressure Vessels During Normal Operation and CertainTransient Conditions," NUREG-0778, June 1981.

11/20/87 28

illustrates this point for a plant "ratcheted" 67°F. It is a replot of Fig•ire 1

from RG 1.154 to convert the abscissa to EFPY and plot the frequency on a linear

scale. Figure 8 shows the increasing probability of failure per EFPY as the

plant ages. For this particular plant, at end of licensed life., the p,.obability

is actually shown to be about 2.4 x 10-1 per year (using Revision 2) whereas it

is thought to be about 1 x 10-6 per year according to the present PTS rule.

To actually reduce the probability of vessel failure from PTS for this

plant, which is something approaching 2 x 10-4 during its lifetime, (a) the

hazard must be known to owner and operator, and (b) some action must be taken

to reduce vessel brittleness, to improve plant systems or modify operator action

to reduce the frequency/severity of PTS events. It has not been possible to

quantify the risk averted by these means, but it does seem necessary to make

sure the extent of the hazard is known.

Evaluation Of Vessel Integrity After A Transient

A significant source of risk averted after an actual LTOP or PTS event is

to be found in the evaluation to determine if the probability of vessel damage

is high enough to warrant a vessel inspection for cracks that could have "popped

in" during the transient and arrested at such depth in the vessel that they

would pose a threat-when full operating pressure was applied. Appendix E to

Section XI of the ASME Code provides criteria for judging when such an inspection

is needed. For example, suppose an LTOP transient occurred with a maximum

pressure of 11500 psig at a temperature of RTNDT +100 F, calculated using Revi-

sion 1 of RG 1.99. From Table E-1, the acceptable pressure is 2250 psig, hence

no inspection would be required. If, however, using Revision 2, the temperature

of the event was found to be RTNDT -50'F, the allowable pressure would be

1250 psig and inspection or further investigation would be clearly indicated.

The difference in RTNDT, Revision 2 - Revision 1, cited in this example would

be equalled or exceeded in roughly half of the PWRs listed in Table 1.

Appendix E to Section XI also provides a criterion for evaluating the

severity of a PTS event. It shows: if T-final falls below RTNDT + 55 0 F,

further analysis and inspection is required. Parametric studies made in thermal

shock studies have shown that for an event in which the pressure is 1500 psi or

greater, the probability of crack pop-in at that temperature is too low to

obtain by the Monte Carlo technique. However, if Tfinal-RTNDT were -,OF, the

11/20/87 29

studies show the probability of failure (given the transient) to be 10-4

(extrapolated). Thus it is clear that the differences in RTNDT calculated by

the ethods of Revision 1 and Revision 2 can make a significant contribution

to risk averted.

Evaluation of Vessel Integrity in the Presence of a Flaw

Inservice inspection of the reactor vessel beltline has, on occasion,

revealed a flaw of significant size. In its evaluation, it is obviously neces-

sary to know RT NDT for the material in which the flaw is embedded. Without

repeating the previous discussions of risk it seems clear that here again

there is a significant risk averted by using Revision 2 instead of Revisior 1

to obtain RT for that material.

Summary of Risk

The risk averted by implementing Revision 2 has been shown to derive

from: (a) better operator awareness of where the fracture limits are so that

he has more concern and more time to take action to controI transients such as

LIOP and PTS events, and (b) better knowledge of the degree of vessel embrittle-

ment when evaluating a transient that has occurred or a flaw that has been

detected. Even in normal operation and pressure testing there is a significant

risk averted by not having pressure-temperature limits 50-100*F lower than they

should be, which will be the case it Revision 2 is not implemented.

(4) No significant impact on radiological exposure of facility employees.brought on by the implementation of Revision 2 has been identified, with the

possible exception of exposure to personnel inspecting for leaks during pressure

tests. The pressure boundary will be hotter as a result of using-Revision 2 as

the basis for P-T limits, which increases the difficulty of inspection. Based

on the absence of comments from utilities, this effect is judged to be small.

(5) COSTS - PWRS

The principal cost impact at the time of -)IVerntation of Revision 2 will

be that required to install PORVs with variabl sef:points for those Combustion

11/20/87 30

Engineering plants that have no PORVs but find the operating window in tempera-ture is nearly closed. No utility has presented such a plan, but some studies

have been made as reported in Section 4.a.1,3.

The source of continuing costs is the cost of replacement power for delays

in heatup/cooldown caused by narrower operating windows. These costs range

from about $10,000 to $20,000 per hour for most plants. In the PNL analysis

done for the "For Comment" version of Revision 2, the additional time was esti-

mated to be 2 hours for the plants ratcheted 50-100'F and half that for the

others. Comments from vendors and utilities did not support this estimate nor

offer any other. The approach described in Section 4.a.1.3 is to outline all

the remedies to be considered when aiutility's submittal describing a hardship

is received.

(5) COSTS - BWRS

As described in Section 4.a.2, the impact on BWRs is in the time to warm

the vessel for pressure testing. In response to an NRC request for cost

figures on an auxiliary boiler system, the BWROG provided cost breakdowns from

a BWR/3 and a BWR/2 study. For the latter, "Plant B" had two electric boilers

and shutdown cooling heat exchangers already in place. The actual cost of

modifications to accomplish the heatup function was $850,000. For Plant A,

the BWR/3, a conceptual study produced estimated total costs of $2,900,000 -

4,000,000 for a permanent boiler of 100,000 lb. capacity housed in its own

building, plus heat exchangers, piping etc., and vessel drain piping to

equalize upper and lower head temperatures, Alternatively, if a rental boiler

was used, the costs came to $1,600,000 - 1,900,000 plus $70,000 for a two-month

rental each time a boiler was required. In addition to these two specific

cases, Section 3.3 of NEDC 31363 gives a summary of cost statements obtained

from a survey of BWR owners. Half of the utilities responding already have an

auxiliary boiler in place. "Cost estimates to modify the boiler systems and

to provide condensate cleanup and return, averaged about $500,000."

Balanced against the installation costs is the. saving in power replacement

costs plus some saving in pump maintenance costs achieved by supplementing the

pump heat with heat from the auxiliary boiler. Power costs $10,000 - 20,000

per hour (NUREG/CR 4012 Vol. 2). In Appendix B to NEDC 31363 it states: "A

11/20/R7 31

one-hour reduction in heatup time for each outage for the remaining life has a

present worth of $100,000." Using this figure, the cost of installation would

be balanced by the savings in replacement power if the average heatup time

saved is 5 hours per heatup for plants that already have parts of the

auxiliary system in place to 20 hours for a rental boiler system and up to

40 hours for a complete system.

To estimate the number of hours heatup time saved by an auxiliary boiler,

its capacity is assumed to be 20%F/hour, based on the information provided by

the BWROG. Heatup times using pumps were obtained from Tables 6-2 and 6-3 in

NEDC-31140, which summarize utility replies on this issue. Times to get to

200'F ranged from 12-48 hours. Assuming the starting temperature was 80'F,

the time to get to 200'F using an auxiliary boiler (and no pump heat) would

be 6 hours or a saving of 6 to 42 hours. Times to get to 220'F ranged from

20 to 96 hours which works out to a saving of 13 to 89 hours, (using the same

assumptions',

In addition to the replacement power savings, there is a significant, but

unquantified cost saving in the reduced seal wear maintenance costs. In Part I

of its comi.ents on Revision 2 (NEDC-31140) the BWROG said:

"The recirculation pumps are designed to operate best at 1000 psig..

Operation with the vessel at low pressures is an off-design condition so seals

are not loaded as designed. While this off-design conuition is not considered

severe to the extent that it should be avoided, operation at low pressures

should be minimized to improve seal life."

In conclusion, the BWROG submittals commenting on Revision 2 provided some

cost figures but no comprehensive analysis of the cost effectiveness of an

auxiliary boiler as a function of the heatup time with pumps only. The foregoing

analysis makes use of bits ani' pieces taken from the information submitted. It

may not be applicaLle in every case, but it appears from this analysis that

many of the BWRs that face the greatest delays in heatup when P-T limits are

based on Revision 2 would find it cost effective to install an auxiliary boiler.

(6) The potential safety impact of changes in plant or operational com-

plexity, including the relationship to proposed and existing regulatory require-

ments such as Appendix G to 10 CFR 50 has been discussed in Section 4.

(7) An itemized list of the NRC resource burden is as follows.

11/20/87 32

NRC RESOURCE BURDEN FORTHE THREE-YEARPERIOD AFTER THE GUIDE BELJMES EFFECTIVE

Page Responsible ResourceReference Organization Burdenin Encl. 3

Staff ContractHours Costs

1. Prepare SERs on licensee 5-8 and NRR/EMTB 4400revision of P-T limits, 11-16PWR and BWR

2. Prepare SERs on PWR 5-8 NRR/SRXB 300 $150,000iicensee revision ofLTOP set points andenable temperatures

3. Prepare SERs on licensee 8-11 NRR/SRXB 600requests for relief fromrequirement to protectthe App. G curve based onsystems changes to reducefrequency of LTOP events.

4. 10 CFR 50.61 PTS Rule - 19-22 RES/DE 1500Amend Formula for RTDTto be consistent with E"

Revision 2.

(8) The potential impact of differences in facility age and between PWRs

and BWRs has been given in detail in Section 4.

(9) The proposed implementation of Revision 2 is final in the sense that

there is no intent to return to Revision 1. There is, however, a strong prob-

ability that revisions will continue to be made over the lifetime of the plants

as more data are added around the "fringes" of the data base: high nickel mate-

rials, low copper "modern" steels, and high fluence conditions encountered in

plant life'.extension. In addition, the effects of the low fluence rate in BWRs

may require special'treatment. Other chemical elements, notably phosphorous,

may someday become a factor. Where there is now a substantial data base, little

change is anticipated.

11/20/87 33

6. DECISION RATIONALE

Based on the foregoing analyses of the safety issues, system impacts and

costs, it is recommended that Revision 2 be issued in final form. The analysis

has shown that the Guide is needed, because it provides part of the basis for

ensuring safe operation of reactors during startup and shutdown and for the

evaluation of transients and flaws found in service. Periodic updating of the

Guide is consistent with the requirements of Appendix G, 10 CFR Part 50. From

a regulatory stand point, the benefit of Revision 2 is that its use restores

safety margins to their intended levels.

Adoption of Revision 2 will raise the P-T limits for most operating reactors

that now use Revision 1 as the basis for these limits. In preparing the value/

impact analysis, the most significant source of risk was not quantifiable. As

discussed in Section 5(3) the principal safety impact of operating with lower

P-T limits (i.e., continuing to use limits based on Revision 1) will occur dur-

ing a transient, because the operator (a) will not have accurate information of

the potential hazard to vessel integrity, and (b) will have less time to take

corrective action.

Finally, we believe that industry is receptive to the proposed "trend

curves," as they are commonly called. Based on industry response to presenta-

tions to ASTM Committee E-1O, the Metal Properties Council, and ASME Boiler

Code Section XI working grouPs concerned with this subject, there appears to

be general agreement on the need for new curves. Based on comments received,

we know of no serious objection to the calculative procedures given in

Revision 2. The calculative procedures for ARTNRT in Revision 2 have been

adopted in ASTM Standard Guide E900 and will be put in Section XI of the ASME

Boiler Code, but this will take at least a year. There will continue to be a

need for Regulatory Guide 1.99 to provide acceptable treatment of the question

of margin, the treatment of plant specific surveillance data, and the calcula-

tion of attenuation of damage through the vessel wall.

7, .MPLEMENTATION

No immediately effective regulatory actions are required to ensure that

facilities now in operation pose no undue risk to public health and safety.

However, Revision 2 should be issued as soon as possible. Being widely

11/20/87 34

accepted in the industry, it is the basis at least in part for the updates

of presgure-temperature limits that are being submitted. The staff needs to

be able to review the submittals to Revision 2 without equivocation.

Parag.raph D.1. of the Implementation Section of Revision 2 reads much as

it did for Revision 1, the key sentence being: "...the methods described in

regulatory position C.1 and C.2 will be used in evaluating all predictions of

radiation embrittlement needed to implement General Design Criterion 31 or as

called for in Appendices G ano H to 10 CFR Part 50 submitted after the effective

date of publication of the Guide." This means that plants will be allowed to

continue to follow the present schedule for updating their P-T limits, but only

within a 3-year period. Generic Letter 88-XX requires all operating plants to

review the limits within a 3-year period after the Guide becomes effective and

to revise them if necessary. The decision to forego prompt implementation

across the board is based on the existence of significant margins. In the

staff's judgment, the risk of allowing some plants to operate three more years

with present P-T limits was justified by the reduceo impact on the industry and

the NRC that will oe achieved by not requiring a complete review of all P-T

limits in a 6-12 month period. No staff actions will be required to implement

the Guide other than making sure that each facility gets a copy.

8. OTHER IMPACTS

There are no other actions, systems or prior analyses known to need

reassessment as a result of publication of Revision 2.

This Guide does not add to the reporting or information collection require-

ments of licensees, nor does it affect small entities as defined in the Regula-

tory Flexibility Act.

11/20/87 35

Table 1 CHANGES IN PRESSURE -5 EFPY BEYOND JANUARYREVISION 2 INSTEAD OF

TEMPERATURE LIMITS FOR1986 CAUSED BY USE OFREVISION 1 OF R.G. 1.99

RTNDT Pressure, psig,

Plant Name at 3/4 t, 'F at Twater = 120OF

Arkansas Nuclear Rev. 1 72 754One, Unit I Rev. 2 146 520

Diff: +75 -234

Arkansas Nuclear Rev. 1 45 918One, Unit 2 Rev. 2 109 606

Diff. +63 -312

Beaver Valley Power Rev. 1 104 620Station Unit 1 Rev. 2 158 501

Diff. +53 -119

Byron Station Rev. 1 54 859Unit 1 Rev. 2 86 685

Diff. +33 -174

Callaway Plant Rev. 1 67 778Rev. 2 109 605Diff. +42 -173

Calvert Cliffs Rev. 1 57 840Nuclear Power Plant Rev. 2 160 497Unit No. 1 Diff. +104 -343

Calvert rliffs Rev. 1 68 771Nuclear Power Plant Rev. 2 122 570Unit No. 2 Diff. +54 -201

Catawba Nuclear Rev. 1 -1 1415Station, Unit 1 Rev. 2 27 1076

Diff. +28 -339

Catawba Nuclear Rev. 1 42 944Station, Unit 2 Rev. 2 82 702

Diff. +40 -242

Donald C. Cook Rev. 1 55 848Nuclear Plant Rev. 2 129 553Unit No. 1 Diff. +74 -296

Donald C. Cook Rev. 1 97 646Nuclear Plant Rev. 2 144 523Unit No. 2 Diff. +48 -123

11/20/87 36

W

Table I (Continued)

RTNDT Pressure, psig,

Plant Name at 3/4 t, OF at Twater = 120OF

Crystal River Rev. 1 73 745Unit 3 Rev. 2 142 528

Diff. +69 -218

Davis Besse Rev. 1 58 828Unit No. 1 Rev. 2 135 541

Diff. +76 -287

Diablo Canyon Rev. 1 - 6 1496Unit 1 Rev. 2 105 618

Diff. +111 -877

Diablo Canyon Rev. 1 92 664Unit 2 Rev. 2 137 536

Diff. +46 -128

Joseph M. Farley Rev. 1 79 715Nuclear Plant Rev. 2 121 572Unit 1 Diff. +42 -143

Joseph M. Farley Rev. 1 82 702Nuclear Plant Rev. 2 128 557Unit 2 Diff. +45 -145

Fort Calhoun Rev. 1 32 1030Rev. 2 167 489Diff. +135 -541

R. E. Ginna Rev. 1 142 526Nuclear Power Plant Rev. 2 210 447

Diff. +67 - 79

Haddam Neck Rev. 1 78 721Plant Rev. 2 108 608

Diff. +30 -113

Indian Point Rev. 1 82 702Unit 2 Rev. 2 139 533

Diff. +56 -169

Inriian Point Rev. 1 128 556Unit 3 Rev. 2 177 476

DifV. +49 -80

11/20/87 37

r

•4V

Table 1 (Continued)

RTNDT Pressure, psig,

Plant Name at 3/4 t, OF at T = 120OFwater

Kewaunee Rev. 1 130 552Nuclear Power Rev. 2 200 454Plant Diff. + 70 - 97

Maine Yankee Rev. 1 31 1039Atomic Power Rev. 2 125 564Plant Diff. +94 -476

Millstone 2 Rev. 1 69 765Rev. 2 95 653Diff. +25 -113

McGuire Rev. 1 20 1148Unit 1 Rev. 2 107 611

Diff. 87 -537

Mcruire Rev. 1 37 985Unit 2 Rev. 2 93 661

Diff. +56 -324

North Anna Rev. 1 88 679Power Station Rev. 2 151 512Unit 1 Diff. +63 -168

North Anna Rev. 1 104 620Power Station Rev. 2 154 506Unit 2 Diff. +50 -114

Oconee Nuclear Rev. 1 68 772Station Rev. 2 141 529Unit I Diff. +73 -243

Oconee Nuclear Rev. 1 84 696Station Rev. 2 161 496Unit 2 Diff. +77 -199

Oconee Nuclear Rev. 1 73 747Station Rev. 2 149 515Unit 3 Diff. +76 -232

Palisades Rev. 1 101 632Plant Rev. 2 153 509

Diff. +52 -123

11/20/87 38

Table 1 (Continued)

RTNDT Pressure, psig,

Plant Name at 3/4 t, OF *at Twater = 120OF

Pale Verde Rev. 1 43 935Unit 1 Rev. 2 89 676

Diff. +46 -260

Point Beach Rev. 1 113 593Nuclear Plant Rev. 2 189 465Unit 1 Diff. +75 -128

Point Beach Rev. 1 177 477Nuclear Plant Rev. 2 219 441Unit 2 Diff. +42 -35

Prairie Island Rev. 1 85 692Unit 1 Rev, 2 136 538

Diff. +52 -154

Prairie Isiand Rev. 1 118 579Unit 2 Rev. 2 152 510

Diff. +34 - 69

Rancho Seco Rev. 1 76 730Rev. 2 145 522Diff. +69 -208

H. B. Robinson Rev. 1 92 664Steam Electric Rev. 2 1L2 510Plant, Unit No. 2 Diff. +60 -154

Salem Generating Rev. 1 107 610Station Rev. 2 158 500Unit 1 Diff. +51 -110

Salem Generating Rev. 1 -13 1600Station Rev. 2 80 714Unit 2 Diff. +92 -886

San Onofre Rev. 1 185 468Nuclear Generating Rev. 2 185 468Station, Unit 1 Diff. 0 0

San Onofre Rev. 1 39 968Nuclear Generating Rev. 2 53 701Station, Unit 2 Diff. +44 -267

San Onofre Rev. 1 56 840Nuclear Generating Rev. 2 92 663Station, Unit 3 Diff. +35 -177

11/20/87 39

Table I (Continued)

RTNDT Pressure, psig,

Plant Name at 3/4 t, OF at Twater = 1200f

Sequoyah Rev. 1 78 720Nuclear Plant Rev. 2 134 543Unit 1 Diff. +55 -177

Sequoyah Rev. 1 42 946Nuclear Plant Rev. 2 91 666Unit 2 Diff. +50 -280

St. Lucie Rev. 1 15 1201Unit 1 Rev. 2 101 632

Diff. +85 -569

St. Lucie Rev. 1 47 905Unit 2 Rev. 2 97 645

Diff. +50 -260

Virgil Summer Rev. 1 54 854Rev. 2 104 622Diff. +49 -233

Surry Rev. 1 69 767Power Station Rev. 2 144 524Unit 1 Diff. +75 -243

Surry Rev. 1 59 826Power Station Rev. 2 134 543Unit 2 Diff, +75 -283

Three Mile Island Rev. 1 64 794Nuclear Station Rev. 2 142 528Unit 1 Diff. +77 -266

Trojan Rev. 1 54 855Nuclear Plant Rev. 2 108 608

Diff. +54 -248

Turkey Point Rev. 1 144 523Unit 3 Rev. 2 199 455

Diff. +55 -68

Turkey Point Rev. 1 144 523Unit 4 Rev, 2 199 455

Diff. +55 -68

Waterford 3 Rev. 1 32 1028Rev. 2 65 790Diff. +33 -238

11/20/87 40

Table 1 (Continued)

RTNDT Pressure, psig,

Plant Name at 3/4 t, OF at T water= 120OF

Wolf Creek Rev. 1 50 884Generating Station Rev. 2 93 658Unit No. 1 Diff. +43 -226

Yankee Rowe Rev. 1 128 556Rev. 2 179 474Diff. +51 -82

Zion Station Rev. 1 104 622Unit 1 Rev. 2 171 483

Diff. +68 -139

Zion Station Rev. 1 .64 796Unit 2 Rev. 2 132 548

Diff. +68 -248

11/20/87 41

Table 2 MINIMUM PERMISSIBLE PRESSURE-TEST TEMPERATURES FOR BWRsFrom Table 3-1 in BWR Owners Group Report "Evaluationof Regulatory Guide 1.99, Revision 2 Impact on BWRs,Part 2, May, 1987.

5 year futureleak test (1000 psig) EOL leak test

Plant temperature, OF temperature, OF

I Monticello 255 2832 Nine Mile Point 1 223 2513 Cooper 210 2434 Browns Ferry 1 209 2525 Duane Arnold 205 2276 Hatch 1 202 2397 Millstone 1 198 2348 Shoreham 187 2819 Hatch 2 177 240

10 Fitzpatrick 176 19911 Oyster Creek 173 18812 Dresden 3 172 21713 Dresden 2 168 19214 Brunswick 2 167 19415 La SaTle 2 165 19116 Hanford 2 163 21117 Browns Ferry 2 162 21018 Peach Bottom 3 162 18919 Browns Ferry 3 159 21020 Pilgrim 158 18221 Clinton 154 22822 Vermont Yankee 151 17523 Peach Bottom 2 150 18224 Quad Cities 1 148 17625 Grand Gulf 1 144 16226 Nine Mile Point 2 143 16727 La Salle 1 141 22728 Quad Cities 2 141 16629 Hope Creek 1 138 16830 Susquehanna 1 138 17431 Brunswick 1 137 17932 Fermi 2 134 23533 Susquehanna 2 130 16634 River Bend 1 129 20135 Grand Gulf 2 125 162

11/20/87 42

ASC

Plant name d

Arkansas NuclearOne, Unit I

Arkansas NuclearOne, Unit 2

Beaver Valley PowerStation, Unit 1

Byron StationUnit I

Callaway Plant

Calvert CliffsNuclear Power PlantUnit No. 1

Table 3 CHANGES IN RTPTS FOR ALL PWRs

IF REVISION 2 WERE USED

End ofLicensed

*pplicable RTPT at end Lifecreening Of censed life, (CP + 40)riterion, deg. F unlesseg. F PTS Rule Rev. 2 noted)

300 264 268 2008

270 180 172 2012

270 .258 257 2010

270 123 113 2024(OL-+ 40)

270 161 .152 2024.(OL + 40)

270 238 301 2014(OL + 40)

EstimatedYear ScreeningCriterion willbe reachedPTS Rule Rev. 2

2049 2049

>2050 >2050

2032 2048

>2050 >2050

>2050 >2050

2039 1997

Calvert Cliffs 270 199 197 2016 >2050 >2050Nuclear Power Plant (OL + 40)Unit No. 2

Catawba Nuclear 270 104 87 2024 >2050 >2050Station, Unit 1 (OL + 40)

Catawba Nuclear 270 127 120 2024 >2050 >2050Station, Unit 2 (OL + 40)

Donald C. Cook 300 251 260 2009 2033 2037Nuclear PlantUnit No. 1

Donald C. Cook 270 205 210 2009- >2050 >2050Nuclear PlantUnit No. 2

Crystal River 300 .267 257 2008 2029 2039Unit 3

Davis Besse 300 217 249 2011 >2050 >2050Unit No. 1

11/20/87 43

Table 3 (Continued)

Plant name

Diablo CanyonUnit No. 1

Diablo CanyonUnit No. 2

Joseph M. FarleyNuclear PlantUnit No. 1

Joseph M. FarleyNuclear PlantUnit No. 2

Fort Calhoun

R. E. GinnaNuclear Power P1

Haddam NeckPlant

Indian PointUnit 2

Indian PointUnit 3

KewauneeNuclear PowerPlant

Maine YankeeAtomic PowerPlant

Millstone 2

McGuireUnit 1

Appl icableScreeningCriterion,deg. F

270

270

270

RTpTC at endof 1%censed life,deg. FPTS Rule Rev. 2

202 270

209 211

191 186

End ofLicensedLife(CP + 40)unlessnoted)

2008

2010

2012

EstimatedYear ScreeningCriterion willbe reachedPTS Rule Rev. 2

>2050 2008

>2050 >2050

>2050 >2050

270 233 233 2012 >2050 >2050

270 235 302 2008 2030 1993

300 266 283 2006 2032 2026ant

.270

270

270

300

165

214

269

329

159

226

269

313

2004

2006

2009

2013(OL + 40)

> 2050

> 2050

2010

2004

>2050

>2040

2010

2006

300 243 252 2008 >2050 >2050

300 197 187 2010 >2050 >2050

270 24/ 256 2013 2038 2038

11/20/87 44

Table 3 (Continued)

End of.Licensed Estimated

Applicable RT at end Life Year ScreeningScreening ofmT•censed life, (CP + 40) Criterion willCriterion, deg. F unliess be reached

Plant name deg. F PTS Rule Rev. 2 noted) PTS Rule Rev. 2

McGuire 270 194 188 2013 >2050 >2050Unit No. 2

North Anna 270 225 22.1 2011 >2050 >2050Power StationUnit 1

North Anna 270 228 217 2011 >2050 >2050Power StatounUnit 2

Oconee Nuclear 270 239 249 2007 2046 2034StationUnit 1

Oconee Nuclear 300 299 292 2013 2014 2019Station (OL + 40)Unit 2

Oconee Nuclear 300 233 261 2007 >2050 >2050StationUnit,3

PalisadesPlant

270 270 322* 2007 2007* 1988-*1992

Palo Verde 270 142 128 2024 >2050 >2050UJ,• 1 (OL + 40)

Point Beach 270 249 269 2010 2029 2011NL-iear Plant (OL + 40)Unit 1

Point Beach 300 293 305 2013 2018 2008Nuclear Plant (OL + 40)Unit 2

Prairie Island 300 178 181 2008 >2050 >2050Unit I

*Depending on the efficacy of the proposed

of life will be reduced and the Screeningearly 1992.

flux reduction program, RTPTS

Criterion will not be reached

at end

until

11/20/87 45

Table 3 (Continued)

End ofLicensed Estimated

Applicable RTi at end Life Year ScreeningScreening of ]•censed life, (CP + 40) Criterion willCriterion, jeg, F unless be reached

Plant name deg. F PTS Rule Rev. 2 noted) PTS Rule Rev. 2

Prairie Island 300 222 208 2008 >2050 >2050Unit 2

Rancho Seco 270 264 255 2008 2012 201R

H. B. Robinson 300 269 262 2007 2032 >205-0Steam ElectricPlant, Unit No. 2

Salem Generating 2?0 255 256 2008 2021 2020StationUnit I

Slem Generating 270 160 202 2008 >2050 >2050Stati 0oUnit 2

San Onofre 270 265 228 2004 2010 >2050Nuclear GeneratingStation, Unit 1

San Onofre 270 145 137 2013 >2050 >2050Nuclear GeneratingStation, Unit 2

Ian Onofre 270 131 124 2013 >2050 >2050Nuclear GeneratingStation, Unit 3

Sequoyah 270 240 225 2010 >2050 >2050Nuclear PlantUnit 1

Sequoyah 270 172 166 2010 >2050 >2050Nuclear PlantUnit 2

St. Lucie 270 231 239 2010 2050 >2050Unit 1

St. Liucie 270 179 172 2023 >2050 >2050Unit 2 (OL + 40)

11/20/87 46

Table 3 (Cont;inue,)

Plan'

Vi rgi I

SurryPowerUnit I

SurryPowerUnit 2

ThreeNucleaUnit I

TrojanNuclea

TurkeyUnit 3

TurkeyUnit 4

Wateri

WoIf CGeneraUnit N

Yankee

Zion IUnit I

ZionUnit

tname,

Summer

Stati on

ApplicableS.reeni ngCriterion,dg. F

270

270

of kv-e'nsed' I i fe,

269 260

Licensed

(UP * 40)unessnoted)

2013

2012

Estimatedvearr setreeii no

be- r'eiahe..TS Nfle Re..2'

2013 2019'

270 225 21-3 2013 >2050 >2010Station

Mlile Island 270 274 262 2OOS 2007 2011r Station

270 1.91 1% 2011 32050 >2050

.r Plant

Point 300 .. 263 263 2007 2035 2020

Point

ord 3

•reekiting Station0o. 1

Rowe

tationi

Star i o

300

270

270

263

84

140

283

83

131

2007

2024(OL + 40)

2025(OL + 40)

1997.

2008

2008

203.5

>2050

>2050

2020

>2050

>2050

270

300

270

239

311

259

249

299

249

2029

2005

2017

2025

2011

2023

11/20/87 47

TABLE 4 SUMARY OF TABLE 3 FOR PLANTS THATMAY EXCEED THE SCREENING CRITERIONACCORDING TO PRESENT ESTIMATES

P'1ant

CalvePlant

Di ab I

Fcrt

India

Ocone

Palis

Point

Point

Surry

ThreeUnit

Zi on

a - ab-b

c- C

Willbefor

When

calct.Name presE

rt Cliffs Nuclear PowerUnit No. I

0 Canyon Unit No. I

violate the screening criterion'e the end of licensed life

RTPTS is If RTPTS were

Jlated by the calculated by•nt PTS Rule Revision 2, RG 1-99

b

•b

Calhoun

n Point Ui it 3 c c

nee Nuclear Power Plant a a

)e Nuclear Station Unit 2 c

ades Plant b a

Beach Nuclear Plant, Unit 1 c

Beach Nuclear Plant, Unit 2 a

Power Station, Unit 1 c

Mile Island Nuclear Station,I b

Station, Unit I a c

ictually violates screening criteriaiarely (RTPTS = 270 or 300)

lose (RTPTS = 269 or 299)

11/20/87 48

e1

-rF

Temperature, dleg, F

"Note: The three curve• fepre-Seni d~ftered coo~ing rales,Curve I being fihe stowesl."

Figure 1 P-T Limit Separates the Operating Zone fromConditions Hazardous to Vessel Integrity

11/20/87 49

0)

9L

2400

?000a-!C..0

4-,

(f):

0

0

C:1200 P-T Umit

250 F/hrHeatup

400

0100 200 300 400

Temperature, deg. F

Figure 2 A Fixed LTOP Set Point Accentuates the Impact of Large Adjustmentsof Reference Temperature on the Operating Window in Temperature

11/20/87 50

T >

"0 OVERSHOOT0 LTOP

SETPOINT. MARGIN TO AVOIDLUFTING VALVE /

40 -- GAUGE ERROR /

OPERATING /S1WINDOW A'

300 Cý

GAUGE ERROR

PRESSURE DIFFERENCE(2001 SEAL TO GAUGE /

100 REQUIREDAP SATURATICACROSS PUMPLRAI

•EA L

0 100 200 300 400

Temperature, deg. F

Figure 3. Several. Factors Reduce the Operating Window in PressureWhen the P-T Limit is Protected by an LTOP System

11/20/87 •I

)N PRESSURE

600

,,,J J.

10-2

5

10-3

t

LL

0

C.

0

*0

10)-4

10-5

200

RTPTS or RTNDT,Rev. 2, OF

FIGURE 4 OCONEE I - EFFECT OF TREND CURVE FORMULA USED IN CALCULATING THE FAILURE PROBABILITYFOR THREE TRANSIENTS DESCRIBED IN NUREG/CR 3770

14SLB - 1 Main Steam Line Break with OTSG Overfesd for 20 Min.TBVG4 Four Turbine Bypass Valves Stuck Open with OTSG Overfeed.6A Similar to TBVG4.

I0-3 _10-2

5

)xxx

~~ Rev.2

0

10.4 -L..I

'D PTS Pulp PTS /RU.C

/Rev. 210

.5i

2.1 2.4 43

200 250 300 350 200 250 300 350 200 250 300 350

RTPTS or RTNDT,Rev. 2. OF

FIGURE 5 CALVERT CLIFFS 1 - EFFECT OF TREND CURVE FORMULA USED IN CALCULATING THE FAILURE PROBABILITY

FOR THREE TRANSIENTS DESCRIBED TN NUREG/CR 40222.1 Small Steam Line Break at Hot Zero Power.2.4 Steam Line Break at Hot Zero Power with Repressurization Not Controlled, with SG overfeed.8.3 Small Break LOCA with Loss of Natural Circulation, Repressurization Not Controlled.

I

10-2--

5-

// IIf

/• / .I5 PTS Rule PTS Rule / PTS Rule /-o

n-

Rev 2 Rev 2 ev. 2.

//I/O.

/

6.6 6.9 8.6

150 200 250 300 350 150 200 250 300 350 150 200 250 300 350

RTpTs or RTNOT,Rev. 2, 0F

FIGURE 6 H. B. ROBINSON 2 - EFFECT OF TREND CURVE FORMULA USED IN CALCULATING THE FAILURE PROBABILITY

FOR THREE TRANSIENTS DESCRIBED IN NUREG/OR 4183. A HYPOTHETICAL CHEMICAL COMPOSITION WAS USED

TO ACHIEVE MORE FAILURES WITHIN THE LIMITATIONS OF THE MONTE CARLO METHOD.

6.6 Large Steam Line Break at Full Power with SG Overfeed.

6.9 Similar to 6.6.8.6 Main Ste~m Line Break at Hot Zero P'ower with SG Overfeed.

E-"

4 .. I I I I " ' I!. ..

1500 10 EFPY1l 250W 32 EFPYI 2500 10 EFPYI1500 3.2 E FPY •

10 -

//// // /

•/ // // /

p // /I I/ /

I I1 /

Io• /

10-8 I 1 . I 1

200 400 600 800 ic 1400 1600.

Pressure, Psig

Figure 7 EFFECT OF PRESSURE ON FAILURE PROBABILITY FORLTOP EVENTS at 150OF AND 250OF WATER TEMPERATUREAFTER 10 AND 32 EFPY.From NUREG-0778, (amble and Strosnider

11/20/87 55

I

8x10-

7x 101

isI i I

flevision

I

6x 10-4-

Lnd of Licernsed Life1

RTp3T aS end o1licensed life is235OF per tile PTS rule,302'F per Re-vision 2.Screening Cr ;--ion - 2701T

/,1/

//cc~

LL

U:)0

CO)

10)CL

5x10"5

4x00- 5-

3xl10 -

///

//

1do

2_x10-5 ///

Probability Associatedwilh tile ScreeningCriterion

Jan. 1986G

Ix10•5 - 710000,PTS Rule-

1)(10-6ý10 20 30 40

EI IY

Figure 8 For the reactor vessel represented in this illustration,there is a large difference in the predicted probabilityof vessel failure from PTS events depending on whetherRTpTS is 235'F (calculated using the PTS rule) or 302*F

(calculated using Revision 2).

11/20/87 56

.4 . . ,

AUG 1 2 1985

MEMORANDUM FOR: William J. Dircks, ExecutiveDirector for Operations

FROM: Harold R. Denton, DirectorOffice of Nuclear Reactor Regulation

Robert B. Minogue, DirectorOffice of Nuclear Regulatory Research

SUBJECT: RELATIONSHIP OF RTND, CORRELATION IN THE PTSRULE 10 OTHER RTI "CORRELATIONS (YOUR MEMORANDUMOF JULY 30. 19857"D

This mmorandum explains the relationship bet en the recently promulgatedPTS. rule, Reg. Guide 1.99, Rev. 1, and the recently proposed Reg. Guide1.99. Rev. 2. This memorandum also contains RES and NER's recommendationsregarding how to proceed with the proposed Reg. Guide 1.99, Rev. 2.

The recently promulgated PTS rule (published in the Federal Register on"July 23, 1985) contains its own RT correlation (to distinguish it fromother procedures for calculating RPTS) which the rule requires licenseesto use when evaluating their, pý7 ,tls RT PT for comparison to therule-specified screening criterion. The RT P orrelation in the PTS rulewas taken from a regression analysis performe'Jd on the PWR surveillance dataavailable at the time of the initial formulation of the PTS rule in late1982 (SECY 82-465, November 1982). The RT correlation in Reg. Guide1-99, Rev. 1, dated April 1977, (whichN•h used primarily to setpressure-temperature limits for normal operation), was token from abounding curve for the data available at that time, consisting of about 2/3experimental data from research reactors and only 1/3 surveillance datafrom power reactors.

The staff deliberately chose not to incorporate the Reg. Guide 1.99 RTcorrelation into the PTS rule because: 1) it was desired to have the latuyImethods, acceptably conservative for PTS use, included in the PTS rule, andReg. Guide 1.99, Rev. I was somewhat outdated at the time the PTS proposedrule was drafted (late 1982); and 2) it was recognized that Reg. Guide 1.99might be subject to repeated revision- because the technology is improvingand the surveillance data base is continually expanding, and it was desirednot to subject the PTS screening criteria to constant changes, with theresultant variability in impact, and 3) it was believed. necessary toinclude a prescribed method of calculating RT so that the overallconservatism of the rule would not be upset by lA nsee or NRC changes tothe VT correlation -without giving serious consideration to the wholerule. lo prevent confusion with RTI,. correlations used for otherpurposes, the correlation in the final.lIS rule was defined by the termRTpTS which Is used for PTS screening.

- 11

~q2 AUG 2 1985

At the time the PIS rule was put into final form, it, was believed that theR7pTx correlation it contained was adequately conservative for thr purposeof'ltablishing a screening criterion. During the relatively long timeperiod consumed by the collegial process for rule promulgation, includingtwo separate six month periods by the Commission for their review, theRTpTt correlation remained frozen. During that time, the NRC's evaluationof lwer reactor data and improved correlation techniques continued, withthe most recent product of that effort being the proposed Reg. Guide 1.99.Rev. 2 (recently presented to CRGR) containing RI correlations for usein evaluating pressure-temperature limits during ng1•Ial operations. If theRlev. 2 correlations were used in place of the RT 10T correlation, thenhigher values of RT1 1 would be predicted for a few p 'ants that have highnickel content welds. However. several features in the proposed Rev. 2correlation will likely be the subject of considerable discussion andpotential disagreement in the technical community when Rev. 2 is issued forpublic comment. These features involve assumption of a differentrelationship for data fitting, division of the previous single data setinto one data group for base materials and a separate data group for weldmaterials, and elimination of previously used data from research reactors.They warrant comment by the technical, community, and until the anticipatedcomments are resolved, we do not believe the conservatism of the presentRT IS correlation ine the PTS rule is called into question.

Accurate values of materials and fluence parameters are essential for themeaningful application of either correlation. For example, based oncurrent flux levels and the staff's limited information on assumed materialcomposition, it was reported to the CRGR at its review of Reg. Guide 1.99,Rev. 2 on July 24, 1985 that the Reg. Guide, Rev. 2 correlation predictsthe Ft. Calhoun plant would exceed the screening criterion in 1q87 (thecurrent RTpTt correlation, by contrast, predicts a 1996 date). However, weSunderstand-PT'at for cycle 10, beginning this fall., Ft. Calhoun will loadselected fuel assemblies containing several half length poison rods in themost effective outer row positions, which will cause a significantreduction in flux at the welds that are closest to the RT Oc screeninglimits. The effect of this change on the date of exceeding • screening.-limit has not been quantified, but we believe it will be sionificant. Inaddition, Oraha Public Power DisLrict has at, effort underway to quantifythe material composition of the critical welds. Preliminary resultsindicate that it is likely that significantly lower values of RTp 7 andR1 T will be justifi able following this effort. 'Accurate f'lenceregUltion and materials data will be submitted in early 1986 as required bythe PTS rule. Using that data, we expect that the screening criterion willbe predicted to be exceeded in the mid 1990's or later even if the Reg.Guide, kev. 2 correlation is used.

We reconmiend that the proposed Reg. Guide 1.99, Rev. 2 be issued for publiccorment. During the comrment period, the PTS rule will remain in place.Licensees and the technical community will consider the technical merits ofthe proposal, including Its, effect on their plants for non-PIS purposes,chiefly as the basis for pressure-temperature limits, which must meetAppendix' G, 10 CFR Part 50.

,3.-.. LAUL 1J 2 Bs

In addition. it will be made clear inI the package issued for public cownentthat the licensees may also consider and comment on the proposed change'seffect on the calculated PIS risk at their plant. assuming the- Rev. 2correlation, if Justified, would at .:-,ne future' time replace the VR..correlation in the PTS rule. Following resolution of coaments. ohtgeneral agreement Is reached regarding the best way to calculate RND.r..then It will be appropriate to re-evaluate the overall conservatism of thePTS rule. and consider whether amendnt of the -P1 rule Is desirable. Atthat tire. we will have available the plant-specific materials and fluencevalues that the PT$ rule requires all) PWliR.0e6nsees to submit early 6 Lc1986. This will all&ow us to such mre accurately assess the effects on aolplants that would result from a changeover to the Rev. 2 method.

Wle believe this is the best way to proceed, given the necessity ofregulating In a conservative but realistic and consistent bray while at thesame time encouraging technical progress and Improvement in basicunderstanding of couxplex phenomena such as material property changes(RTD 1)) during neutron Irradiation.

.We believe this plan is consistent with the agreements reached at themeeting with you on August 2. 2985. Ve therefore plan to revise the keg.Guide 1.99. itev. 2 package to make clear the relationship of the Guide tothe PTS Rule as described above. and then resubmit the package to the CRGR..-

Office of Nwuclear Reactor Regulation

Robert B. Miinogue, DirectorCffIice of huclear Regulatory Research

I.. .-~.

RevIalon 2May lo8wU.S. NUCLEAR REGULATORY COMMISSION

REGULATORY GUIDEOFFICE OF NUCLEAR REGULATORY RESEARCH

REGULATORY GUIDE 1.99(Task ME 3054)

RADIATION EMBRITTLEMENT OF REACTOR VESSEL MATERIALS

A. INTRODUCTION

General Design Criterion 31, "Fractur Prevention of ReactorCoolant Pressure Boundary,"! of Appendix A, "General DesignCriteria for Nuclear Power Plants" to 10 CFR Part 50, "DomesticLicensing of Production and Utilization Facilities," requires, inpart, that the reactor coolant pressure boundary be designed withsufficient marin to ensure that, when stressed under operating,maintenance, testing, and postulated accident conditions, (1) theboundary behaves in a notbrittle manner and (2) the probabilityof rapidly propagating fractur is iminimized. General DesignCriterion 31 also requires that the design reflect the mucetaintiesin determining the effects; of irradiation on material properties.Appendix G, "Fracture Toughness Rpquirments," and AppendixH, "Reactm Vessel Material Surveillance Program Requireniests,"which implement, in part, Criterion 31, ncessitat the calculationof changes in fralct toughness of reactr vessel materials causedby neuron~ radiation throghout the service life. This guide describesgCenral dproodue acceptable to the NRC staff for calculating theeffects of neutron radiation embrittlement of the low-alloy steelscurently used for light-water-cooled reactor vessels.

The calculative procedures given in Regulatory Position 1. 1 ofthis guide amrn t the ame as those given in the Pressurirzd ThermalShock rule (J 50.61, "Fracture Toughness Requiements for Pro-tection Against Pressurized hermal Shock Events," of 10 CFRPart 50) for calculating RTp'M, the refrene tmperatr that isto be compared to tih screening criterio given in the rule. Theinformation on which this Revision 2 is based may also affec thebasis for the PTS rule. The staff is presently considering whetherto propose a change to 50.61.

The Advisory Committe on Reactor Safeguards has been con-suited concerning this guide and has concurred in the regulatoryposition.

Any informration collection actvitie mentiboned in this regulatoryguide are contained as requirements in 10 CFR Part 50, which pro-vides the regulatory basis for this guide. The information collec-

don requirements in 10 CFR Part 50 have been cleard under OMBClearance No. 3150-0011.

B. DISCUSSION

Some NRC requrement~s that necessitate calculation of radia-tion embrittlement am:

1. Paragraph V.A of Appendix G requires fth effects of neutzonradiation to be predicted from the rPuls ofperient radiatiotieffectstudies. This guide provides such Resuts in the form of calculativprocedures that am acceptable, to the NRC.

2. Paragraph V.B of Appendix G describes the basis for zettingthe upper limit for pressure as a functio= of temperature duringheatup and cooldown for a given service period in terms of dtepredicted vaue of the adjusted reference temperature; at t endof the service period.

3. The; definition of reactor vessel betldin give in Paragraph1.F of Appendix G reuires identificatio of reigions of the reatorvessel that are predicted to experiece sufficient aeutran radiationcmbrinlement to be considered in the selection of the most limhiingmaterial. Paragraphs M.A and IV.A.1 speify the additional testrequrements for beltlin materials tha mrpplenat die requlrements11for reactor vessel materials generally.

4. Paragraph n.1 of Appendix H incorporat ASTM E 185by mrce. Paragraph 5.1 of ASTM E 1852-, "Standard Prac-tice for Conducting Surveillance Tests for Lig-Water CooedNuclear Power Reactor Vessds" (Re. 1), requires that the materialsto be placed in surveillance be dtose that may limit operatim ofthe reactortduring i l , i.e., those expected to have the highestadjusted referen= temperitre or th lowest ChzW upper-shelfenergy at end oflife. Both mmasures of radiation embrittlement mustbe considered. In Paragraph 7.6 of ASTM E 185-.S, the require.

n for the number of capsules and the withdrawal schedule arebased on the calculated amount of radmton embrnttleient at endof life. ,i

'I

USNRC REGULATORY GUIDES The guides are Issued In the following ten broad divisions:

Regulatory Guides awe Issued to describe and make availabe to thepublic methods acceptable to the NRC staff of ImplempenUng I Power Reactors 6. Productss=ec=ifc parts of the Commission's regulations, to delineate tech- 2. Research and Test Reactors 7. Transportationniques used by the staff In evaluating specific problems or postu- 3. Fuels and Materials Facilities a. Occupational Healthlatsd accidents or to provide guidance to applicants. Regulatory 4. Environmental and Siting 2. Antitrust and Financial ReviewGuides are not substitutes for regulations, and compliance with 5. MaterialsandPlantProtoctlon 10. Generalthem Is not required, Methods and solutions different from those setout In the guides will be acceptable if they provide a basis for thefindings requisite to the issuance or continuance of a permit or Copies of bsued guides may be Purchased from the GovernmentlIcense by the Commission. Printing Office at the curat GPO prim. Information an current

GPO prices may be obtained by contactlng the Superintendent-ofThis guide was Issued after consideration of comments received from Documents, U.S. Government Printing Office, Post Office Boxthe public. Comments and suggestions for improvements In these 37082 Washington, CC 20013-7082, telephone (202)275-2060 orgui~s are encouraged at all times. and guides will be (rsed, aS (202)!75-2171.appropriate, to accommodate comments and to reflect new informn-tion or experience.

issued guides may also be Purchased from the National TechnicalWritten comments may be submitted to the Rules and Procedures Information Service on a standing order basis. Details on this

ranch ORR ADM, U.S. Nuclear Regulatory Commission, Service may be obtained by writing NTIS, 5215 Port Royal Road,=ashington, D 20555. Springfield, VA 22161.

The two measures of radiation embrittlement used in this guideare obtained from the results of the Charpy V-notch impact test.Appcnix to 10 CFR Part 50rqi that afull curve of absorbedenergy versus temperature be obtained through the ductle-o-brittleftrasition temperature region. The adjustment of the referencetempetur, TNDT, is defined in Appendix 0 as the tempera-ture shift in the Charpy curve for the irradiated material relativeto that for the unirradiated material measured at the 30-foot-poundenergy level, and the dat that formed the basis for this guide were

0-foot-pound shift values. The second measure of radiationembrilme is the decrease in the ChaWr uppcr-shelf energy levl,which is defined in ASTM B 185-82. bis Revision 2 updates thecalculative procedure for the adjustment of reference temperature;however, calculative procedrm for the decrease in upper-shelfenergy are unchanged because the preparatory work bad not beencompleted in time to include them in this revision.

ihe basis for Equation 2 for ARTNDT (in Regulatory Position1.1 ofthis guide) is contained in pucations by . L Outhrie (e.2) and G. R. Odefe at al. (Ref. 3). Both of these papers usedsurveillance data from commercial power reactors. The bases fortheir regression correlations were different in that Odee madegreater use of physical models of radiation embrittlement. Yet, thetwo papers contain simila& - ridation: (1) separate correla-don functions should be used for weld and base metal, (2) the func-to should be the product of a chemistry factor and a fluence factor,(3) the parameters in the chemistry hacto should be the deementscopper and nickel, and (4) the fluence facto should provide a trendcurve slope of about 0.25 to 0.30 on log-log paper at 10" n/cm2

(E> 1 MeV), steeper at low fluences and flatt=r at high fluences.Regulatory Position 1.1 is a blend of the correlation functionspresented by these authors. Some test reactor data were used asa guide in establishing a cutoff for the chemistry factor for low-copper materials. The data base for Regulatory Position 1.2 is thatgiven by Spencer H. Bush (Ref. 4).

The mnasureof fluence used in this guide Is the number ofmotrueprsquare cetimetefr aving enrge greatr than 1 million

electron volt ((B> 1 M ). Te diffMerences in energy spectra atthe surveillance capsule and the vessel inner surface locations donot appear to be great enough to warrant the use of a damage func-tion such as displacements per atom (dWp) (Re.t 5) in the analysisof the surveillan data base (Ref. 6).

However, th neutron energy spectrum does change significandylocation in the vessel wall; hence for calculating the attenm-

tion of radiation embrittementý through the vessel wall, it isnecessary to use a damage function to determine ARTNDT versusra distae ino the wal. The mnt widely accepted damage funo-don at this time is dpa, and the attenuation formula (Equation 3)given in Regulatory Position 1.1 is based on the attenuation of dpa

ghrough the vessel wal.

Sensitivity to neutron radiation embrittlement may be affectedby elements other than copper and nick. The original version andRevision 1 of this guide had a phosphorus term in the chemistryfacto, but the studis on which this revision was based found otherelements such as phosphorus to be of secondary importance, I.e.,including them in the analysis did not produce a significantly bet-ter fit of the data.

Scatr ~in'the dat base used for this guide Is relatively signifi-camt, as evidenced by the fact that the standard deviations for

Guthrie's derived formulas (Ref. 2) are 28OF for welds and 170Ffor base metal despite extensive efforts to find a model that reducedthe fitting error. Thus the use of surveillance data from a givenreactor (in place of the calculative pocedures iv in this guide)requires considembe engin•ering judgmet to evaluate the crodil-ity of the data and assign suitable margins. When surveillance datafrom the reactor in question become available, the weight givento them relative to the information in this guide will depend oan thecredibilty of the surveillance data as judged by the followingcrcrieria

1. Materials in the capsules should be those judged most likelyto be controlling with regard to radiation embrittlement accordingto the rcommendaton of this guide.

2. Scatter in the plots of Charpy energy versu tcrrperature forthe irradiated and uniradiated conditions should be small enoughto permit the determination of the 30-fo1o-pound temperatm and.the upper-shelf energy unambiguously.

3. When there an two or more sets of surveillance date fromone reactor, teo scatter of ARTNDT values about a best-fit linedrawn as dcscrnbed in Regulatory Position 2.1 normally should beless than 2S0 F for welds and 17O for base metal. Evenifthe flnencerange is large (two or more ordm of magnitude), te scam shouldnot exceed twice those values. Even if the dam fail this criterionfor use in shift calculations, they may. be credible for determiningdecrease in upper-shelf mergy if the upper shelf car be dearly deter-mined, following the definition given in ASTM B 185-82 (Rcf. 1).

4. The Irradiation tempertr of the Char specimens in thecapsule should match vessel wall temperature at the claddinig/basemetal inteTface within ±25'F.

5. The surveillance dam for the corelation monitor material inthe capsule should fall within the scatter band of the data base forthat material.

To use the surveillance data from a specific plant instead ofRegulatory Position 1, one murst develop a relationship of ARTNDTto fluece for that plant. Because such data are limited in numberand subject to scatter, Rcgulator Position 2 describes a prod-in which the form of Equation 2 is to be used and the fluence fac-tor therein is retained, but the cheniatr factor is determined bythe plant surveillance data. Of several possible ways to fit such datthe method that minimizes the sums of the square of the errwas chosen somewhat arbitrarily. Its use is justified in put by thefact that "least, squares" is a common method for curve fitting.Also, when there am only two data points. the least squares methodSives gCrCTr weight to the poit with the higher ARTNDT; thisseem reasonable for fitting surveillance data, because generallythe.higher data point will be the more recnt and therore will repre.sent more mode= procedure.

C. REGULATORY POSmON

1. SURVEILLANCE DATA NOT AVAILABLE

When credible surveillan dam from the reactor in questionare not available, calculation of neutron radiation embrittlement ofthe betline of reactur vessel of light-water reactors should be basedon the procedre in Regulatory Positions 1.1 and 1.2 within thelimitations in Regulatory Position 1.3.

1.99-2

1.1 A4Jsted Reference Tmperatu

The adjusted lef==r= temperature (ART) for each material inthe beltline is gim by the following exresion:

ART - ntial RTNir + ARTNDT + M in (0)

Initial RTND Is the referenee temperature for the unirradiatedmaterial as defined in Paragraph NB-2331 of Section MI of theASME Boiler and ressur Vessel Code Oef. 7). If measured valuesof initial RTNDT for the gaterial in question are not available,generic mgan values for that class* of material may be used if theeare sufficient test results to establish a mean and standard devia-tion for the class.

ARTNDT is the mean value of the adjustment in referecetemperature caused by irradiation and should be calculated asfollows:

RT'NDT . (CF) ft.2s - 0.10 log f) (2)

CF ('F) is the chemistry factor, a function of copper and vickelcontent. CF is given In Table I for welds and in Table 2 for basemetal (plates and forgings). Iinea Intpolion is pmitted. InTables 1 and 2 "weight-Percct copper" and "weightV-rcent

are4- s the best-estimat vilues for dhe material, which willtiral ehe mean of the moeasured values for a plate or forging

= `!orCel samaples made with the weld wire heat number thatmat•h th critical vessel weld. If such valus arm not available,the Wper limitng values given in tbe Material specifications to whichthe vessel was built may be used. if not available, conservativeestimates (mean plus one standard deviation) based on generic datamay be used ifjustification is provided. If there is no informationavailable, 0.35% copper and 1.0% mickl should be assumd.

The neutron flunow at any depth in the vessel wall, f (10' a/ln,E > 1 Mav), is ddrmined as follows:

f - fso'f (a -0.24x) (3)

where ff (10t' n/cmý, E > I Mev) is the calculated value ofthe neutron flinnce at the inner wetted surface of the vessel at thelocation of the postulated defect, and x (In inches) is dhe dceth intodte vessel wall measured from the vessel inner (wetted) surface.Alternatively, If dpa calculations an: made as part of the flnanceanalysis, the ratio of dpa at the depth in question to dpa at the Itnnesurface may be substituted for the exponential attenuation factorin Equation 3.

The fianc factor. fO.28 - 0.1010 loifsj detemined by calanla-tion or from Figure 1.

"Margin" is the quantity, OF, that is tobe added to obtain con-servative, upper-bound values of adjusted reference temperaturefor the calculations required by AppMdix 0 to 10 CFR Part 50.

Margin -2 Vlaor c, (4)/

c dms .- fntl• UW a 3.rT Iis geuimy detmd, d for de wCulswth which "de concuncd, by w0rie of wedn fla (lid 50 o odor);o r e mend, by = he ASTM Suadu Spefa±ian

Hemr, ol is the standard deviation for the initial RTITDT. If ameasured value of iniia RTNIDT for the Material In qUestion isavailable, al is to be estimated from the- precision of thetest methd.If not, a generic -ean values for that class of matra amr used,al Is the standard deviation obtained from the set of data used toestablish the mean.

The standard deviation for ARTNDT, oA, is 29 F for welds and17 OF for base metal, excep that c,& need nct exceed 0.50 timethe mran value of ARTNDT.

1.2 Charpy Upper-Shelf Energy

Charpy uppr-she energy should be assumed to decrease asa funcfion of fluee and copper content as indicated in Figure 2.Lince interpolation is permitd.

1.3 LimitatIons

Application of the foregoing procedures should be subject todIe following limittions:

I1. The Procedures apply to those grades of SA-302, 336, 533,and 508 steels having minitmum specified yield strengths of 50,000Psi and under and to their welds and hea•-•ffected ones.

2. Tfe procedures ar valid for a nomina irradiation tcruPmebof 550"F. Ihrrdiation below 525. F should be considered to pro-duce greater embrittlement, and irradiation above 5W0"F may be

nsidered to produce less cobrittlement- The orction usedshould be justified by refcW to actual data.

3. Application of these procedures to flunce levels or to cop-per or nieml contet beyond ft ranges given In Figur and TablesI and 2 or to materials having chemical compositions beyond therang found in the dama bases used for tIns guide should be Justifiedby submittal of data.

2. SURI"CE DATA AVAILABLE

When two or more credible surveillance data sets (as definedin the Discussion) become available fiom the reactor in question,they may be used to determine the adjusted eference tmperatureand the Ca=Wy upper-shelf necrgy of die betline materials asdescribed in Rgulatory Positions 2.1 and 2.2, rspectively.

2.1 Adjusted Reference Temperature

I

.1

The adjusted reference temperatumr should be obuaned asfollows. First, if there is clear evidence that the copper or nickelcontent of the surveillance weld differs horn that of the vessel weld,i.e., differs from the average for the weld wire beat numberassociated with the vessel weld and the surveillance weld, themeasund values of ARTNDT should be adjusted by multiplyingthem by the ratio of the chmistry factor for the vessel weld to thatfor the surveillance weld. Second, the surveillance data should befitted using Eqution 2 to obtain thde relationship of ARTNDT tofluence. To do so, calculate the chemistry factor, CF, for the begfit by ultiplying each adjusted ARTNDr by its correspndingfinance factor, summing the products, and dividing by the sum ofthe squares of thde fluence fartors. The resulting valu of CF whencntered in Equation 2 will give the relationship of hRTNDT to

1.99-3

TABLE I

CHEMISTRY FACTOR FOR WELDS, P

Cot.er, Nickel, Wt-%Wt- 0 0.20 0.40 0.60 0.80 1.00 1.20

0 20 20 20 20 20 20 200.01 20 20 20 20 20 20 200.02 21 26 27 27 27 27 270.03 22 35 41 41 41 41 410.04 24 43 54" 54 54 54 54

0.05 26 49 67 68 68 68 680.06 29 52 77 82 82 82 820.07 32 55 85 95 95 95 950.08 36 58 90 106 108 108 1080.09 40 61 94 115 122 122 122

0.10 44 65 97 122 133 135 1350.11 49 68 101 130 144 148 1480.12 52 72 103 135 153 161 1610.13 58 76 106 139 162 172 1760.14 61 79 109 142 168 182 188

0.15 66 84 112 146 175 191 2000.16 70 88 115 149 178 199 2110.17 75 92 119 151 184 207 2210.13 79 95 122 154 187 214 2300.19 83 100 126 157 191 220 238

0.20 88 104 129 160 194- 223 2450.21 92 108 133 164 197 229 2520.22 97 112 137 167 200 232 2570.23 101 117 140 169 203 236 2630.24 105 121 144 173 206 A.9MO 268

0.25 110 'b wrl 148 176 209 243 2720.26 113 130 151 180 212 246 2760.27 119 134 155 184 216 249 2800.28 122 138 160 187 218 251 2840.29 128 142 164 191 222 254 287

0.30 131 146 167 194 225 : 257 2900.31 136 151 172 198 228 260 2930.32 140 155 175 202 231 263 2960.33 144 160 180 205 234 266 2990.34 149 164 184 209 238 269 302

0.35 153 168. 187 212 241 272 3050.36 158 172 191 216 245 275 3080.37 162 177 196 220 248 278 3110.38 166 182 200 223 250 281 3140.39 171 185 203 227 254 285 3170.40 175 189 207 231 257 288 320

fluence that fitx the plant surveillance data in such a way as tominimie the sum of the squares of the errors.

To calculate the margin in this case, use Equaton 4; the valuesgive there for OA may be cut in half.

If this procedure gives a higher value of adjusted referencetemperatur than that given by usig the predures of RegulatoryPosition 1.1, t surveillanoe data should be used. If this proceduregives a lower value, either may be used.

For plants having survelance data that e credible in all respectsexcept that the material does not represen the critical material inthe vessel, the calculative prcedures in this guide should be used

to obtain mean values of shift, hRTNj)T. In calculating the margin,the value of aA may be reduced from the values given in the lasparagraph of Regulatory Position 1.1 by an amount to be decidedon a case-by-case basis, depending on where the measured valuesfall relative to the mean calculated for the surveillance matmeial

2.2 Charpy Upper-Shelf Energy

The decrease in upper-shelf energ may be cbtand by plot.ting the reduced plant surveillance data on Figure 2 of this guideand fitting the data with a line drawn parallel to the existing lines \as the upper bound of all the data. This line should be used inpreferece to the existing graph.

1.99-4

TABLE2 -

CHI RY FACTOR FOR BASE METAL, V,

0op40 Ni•l wt-% .

Wo- 20 20 0.40 .6 oso0 0 2

0 20 20 20 20 20 20 200.01 20 20 20 20 20 20 200.02 20 20 20 20 20 20 200.03 20 20 2D 20 2D 20 20

0.04 22 26 26 26 26 26 26

0.05 25 31 31 31 31 31 310.06 28 37 37 37 37 37 370.07 31 43 44 44 44 44 440.08 34 48 51 51 51 51 510.09 37 53 58 58 58 58. 58

0.10 41 58 65 65 67 67 670.11 45 62 72 74 77 77 770.12 49 67 79 83 86 86 860.13 53 71 85 91 96 96 960.14 57 75 91 100 105 106 106

0.15 61 80 99 110 115 117 1170.16 65 84 104 118 123 125 1250.17 69 88 110 127 132 135 1350.18 73 92 115 134 141 144 1440.19 78 97 120 142 150 154 154

0.20 82 102 125 149 159 164 1650.21 86 107 129 155 167 172 1740.22 91 112 134 161 176 181 1840.23 95 117 138 167 184 190 1940.24 100 121 143 172 191 4MA 204

0.25 104 126 148 176 199 208 2140.26 109 130 151 180 205 216 '2210.27 114 134 155 184 211 225 2300.28 119 138 160 187 216 233 2390.29 124 142 164 191 221 241 248

0.30 129 146 167 194 225 249 2570.31 134 151 172 198 228 255 260.32 139 155 175 202 231 260 2740.33 144 160 180 205 234 264 2820.34 149 164 184 209 238 268 290

0.35 153 168 187 212 241 272 2980.36 158 173 191 216 245 275 3030.37 162 .177 196 220 248 278 3080.3" 166 182 200 223 250 281 3130.39 171 185 203 227 254 285 3170.40 175 189 207 231 257. 288 320

3. REQUIMEENT FOR NEW PLAM

For belline materials in the reactor vessel for a new plant, thecontent of rcsidual eleme such as copper, phosphorus, sulfur,and vanadium should be conatroled to low levels.* The copper con-tent shou bech ad ahe calculated adjusted reference temperatat ie 1/4T position in the vessel wall at end of life is less than2 F. In selecting the optimum amont of nic• to be used. itsdeleterious effect on radiation embrittlement should be balancedagainst its 1benefria met:-lurgical efifcts nd its tendency to lowerthe initial RTNDT)T

oFwr norem iiak.o , we &e Appedt ASTM Swodud Spedficawhml A 5330ef. I-

D. IMPLEMENTATION

The purpose of this section is to provide "nfonati to aplicansand licensees regarding the NRC staws plans for using thisregulatory guide. Exceptkin those cases in which an applicant pro-poses ax ac•eptable altative method for compbling with spcif"eportions of the Commission's regulations, the methods descdin this guide wMl be used as follows:

I. The methods dcsc•rbed in Re atoy Positions I and 2 ofthis guide will be used by the NRC gaff in evaluating all predie-tions of radiation embritlenent needed to implement Appendices0 and H to 10 CFR Pan 50.

ii

"' 1.99-5

2. Holder of licnse and permits ShMd use the methodsdesacibed in this ,•ufe to predict the fcd of uu radiation onracor ves ma as as requtred by Paragrph V-A of Appen-dix G to 10 CFR Part 50, unless they can justi the us of dif-f. . -mettodg The us of the R6isicu 2 methodolowy may remitin a modification of the prcssuretzmpturaD limits contained In

Technical Specificaions in order to cantnue, to sas* therequirements of Section V of Appendix 03, 10 CFR Parn 50.

3. The Inl of Regulatory Position 3 am essen-tinAy mnchanged from those used to evaluate constuctim permitapplications docketed on or after June 1,- 1977.

1;99-6

REFERENCES

1. American Society for Testing and Materials,".Standard Prac-fice for Conducting Surveillance Tests for Ught-Water CooedNuclear Power Reactor Vesse" ASIM E 185-82, July 1982.*

2. 0. L Oua, ri., 'Tbapy Trend Curves Based on 177 PWR DatPoints," in -LWR PressM Vessen Surveillance Doslmetry Im-provement Program," NUREGICR-3391, Vol. 2, prepared byHanford Engineering Development Laboratory, HEDL-TME83-22, April 19S4.**

3. G. R. Odette et al., "Physialy Based RegrcssonCorrbelatonof Embrittlement Data from Reactor Pressure VesselSurveillance Programs," Electric Power Research Institute,NP-3319, January 1984.t

4. S. H. Bush, "Sructurl Materials for Nuclear Power Plans,"in Junad of Tting Nwd auuadomn, American Society forTesting and Materials, November 1974.*

S. American Society for Testing and Materials. "Standard Pac-ic= for.Caracterizing Neutron E•posurs in Ferritic Steels inTerms of Displacements per Atom (DPA)," ASrM E 693-79,August 1979.*

6. W. N. McElroy, "LWR Pressure Vessel Survelace Dosimcuyhoprovement Program: LWR Power Reactor SurveillancePhyuics-Dosimetry Darn Base Couipeaum," NUREG/CR-3319, prepared by Hanford Enoginieering DevelopmentLaboratory, HEDL-TME 85-3, August 1985.0*

7. American Society of Mechanical Engineers, Section III,"Nuclea Power Plant Components," of ASME Boikr andPressure Vesel Code, New York (updated freWiutiy).t

8. A=erlcan.Society for Testing and Materials. "Standard.Specification for Pmssor Vessel Plates, Alloy Stee, Quenched

nid Tempered, Manganese-Molybdenum and Magns-Molybdenum-Nickel," ASTM A 533/A 533M.2, September1982.4

*Copies may be "mW~ from dr. Amminrtm godety &w Tesdg and Materials, 1916 Rac Baue4t PA 19103."Copies may be obtaine from tz t..A~ Vof Da . U.S. Govau= Pd of e, Pt 3 7ox 2. Wuuaton, DC 20013-7Mo.

iCepie may be oEmahad fM IM e EcWk Power R•m-Ch In-aStt 3412 IKvfew Avow, Pao MO. CA 9M304.M=ape mAY be aftianed bn Idc Am.erima Sadety of Mwdhankg Eagtneeu 3451L 471h Street, New Task, NY 10017.

1.99-7

U.

0*

C~

-U

Z.U ....... ........ .. .......

....... ......... .... .. .......

-dol l I I I

.1 T-I T .

o i l 7-7... ..... ......... . ....... ............. INS ... ........

zr,1jw ..........7 HUM =V

.......... . ....... 7- 7 .. . ........

7

RH 0,

0

Om

• 1

.-10"3 4 6 5 7 891

1014a 3 4 6 7 8 9 1:

10

Fkaunc.. nlcm2 (E > 1 MoVI

I'.2 3 645 low

FIGURE I Flueace Factor for Use im Equation 2, the Exprsion for ARTNDT

60

50

40

30

'0

4J

~ 20a.

0)C

LU

~10

S(U0)

C

BASE i ILJMETAL WELDS

Lll 0 .35 - 0.30 •

0.30 0.25-0.25 0.20-0.20 0,15--

-- ---- 0.15 - 0.10--E O ...

.0.100.05. I JILIH¶~ till titi I i I I I

. . ........

+ +jHj ...... :17

p ifif

V:

2 X 1017 4 6 8 -1018 2 4 6 8 1019 2 4 6

FLUENCE, nrcm2 (E > 1MeV)

FIoURE 2 Predictd Decrem in Shelf Energy as a Function of Copper Cment and Flec

REGULATORY ANALYSIS

A copy of the regulator analysis prared for th RelaoryGuide 1.99, Revision 2, is availabl for inspection and copying for

a fee m the Commission's Plubli Docum)it Room at 1717 H StretNW., Washington, DC, anda Reulaory Guide 1.99, Revision 2.

1.99-10

p s cgL V%-,14ý ý- -I