ML19338C750

Core Degradation Program,Vol 1:Hydrogen Study.

ML19338C750
Person / Time
Site:	Sequoyah
Issue date:	09/02/1980
From:	TENNESSEE VALLEY AUTHORITY
To:
Shared Package
ML19338C749	List: ML19338C748 ML19338C750 ML19338C751
References
NUDOCS 8009030664
Download: ML19338C750 (100)

Text

_ _ _ _ _ _ . _ _ _ . _ _ _ _ . _ _ _ _ --- . _______ _ ___.___ _ _ _

l

'O O TENNESSEE VALLEY AUTHORITY l

l SEQUOYAH NUCLEAR PLANT l

CORE DEGRADATION PROGRAM VOLUME 1 i

O  !

HYDROGEN STUDY l l

l SEPTEMBER 2,1980 8009030664 1

Principal Contributors:

R. H. Bryan J. C. Carter R. F. Christie F. A. Koontz C. A. Myers D. G. Renfro Outside Contributors:

Burns and Roe, Inc.

Sargent & Lundy Engineers Stone & Webster Engineering Corporation Westinghouse Electric Corporation 9

S

DISCLAIMER

C) The study of degraded coro accidents at the Sequoyah Nuclear Plant, their effects, and their mitigations, described in the accompanying report was performed in the 6-week period concluded April 15, 1980.

j Of necessity, due to the limited duration of the study and the state-of-the-art information available to TVA, several limitations were in order.

First, the calculational techniques to analyze the course of degraded core accidents and their effects on containment were not available since those events are not included in the current plant design bases.

As a result, simplifying assumptions and methods were relied on when 1 necessary to estimate the accident consequences.

Second, the evaluations of the overall risk posed by Sequoyah with

, and without additional accident mitigations are somewhat subjective

due to the limited time and data available for probability and l consequence estimation. Only a limited number of accident sequences were examined and best-estimate judgments were made for the associated

containment failure probabilities.

j Third, the level of detail and degree of feasibility achievable in the conceptual designs of the mitigations were restricted by the available time and personnel. Further effort might result in somewhat different design;. .

Fourth, the costs that were originally estimated for the design, construction, and operation of each of the mitigation concepts have not been included in this version of the report.

i i

i t

l l 'I

\ 1

r TABLE OF CONTENTS S umma ry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v i i i T e c h n i c a l S u=ma ry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x 1.0 Purpose................................................. 1-1 2.0 Scope................................................... 2-1 30 Introduction............................................ 3-1 4.0 Background of the Original Design. . . . . . . . . . . . . . . . . . . . . . . 4-1 4.1 Phys ic al Pla n t Fea tu re s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 -1 4.1.1 Containment........................................... 4-1 2

4.1.2 Passive Containment Heat Removal Systems. . . . . . . . . . . . . . 4-1 4.1 3 Active Containment Heat Removal Systems. . . . . . . . . . . . . . . 4-1 4.1.4 Hydrogen Control Systems.............................. 4-4 4.2 Present Design Bases.................................... 4-4

() 4.2.1 Accidents.............................................".4-4 1

4.2.1.1 Mass and Energy Releases............................ 4-4 4.2.1.2 Design Basis Noncondensible Gas Production. . . . . . . . . . 4-5 4.2.2 Analytical Too1s...................................... 4-7 4.2.2.1 Primary System Thermal Hydraulic Ptograms. . . . . . . . . . . 4-7 4.2.2.2 Containment Thermal Hydraulic Programs.............. 4-7 4.2 3 Predicted Plant Response.............................. 4-7 4.3 Ramifications of TMI on Current Design Bases............ 4-8 4.3 1 TMI Challenge to Containment Design. . . . . . . . . . . . . . . . . . . 4-13 432 Consideration of Inadequate Core Cooling. . . . . . . . . . . . . . 4-16 5.0 Risk Analysis of Present Sequoyah Capability for Key Ac cid ent Sequ enc es. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1 5.1 General Description..................................... 5-1 5.2 Analysis Assumptions.................................... 5-1 A

U

~ .

53 Dominant Sequences...................................... 5-10

'" "*'" '"" ""' """"" " ="""*" "" " " -'

O 6.0 Accidents Beyond the Design Bases....................... 6-1 6.1 Phenomenologial Consequences of Key Accident Scenarios.. 6-1 6.1.1 Hydrogen and Other Noncondensible Gas Production. . . . .. 6-1 6.1.2 Hydrogen Conbustion................................... 6-2 6.1 3 Core Melt Energy Consideration Due to Stea'1. . . . . . . . . . . 6-4 6.2 Analytical Too1s........................................ 6-5 6.2.1 Limitations of Present Computer Codes................. 6-5 6.2.1.1 Prima ry Syst em Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-6 6.2.1.2 Containment Codes................................... 6-6 6.2.2 The MAR CH Co=puter Cod e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-7 l 6.2 3 Physical Analysis of Present Sequoyah Capability for Accident Sequences.................................... 6-8 6.2 3 1 MARCH Noncondensibles............................... 6-8 O

6.2 3 2 Con tain me nt Re sp on se . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-8 6.2 3.2.1 Twenty-Percent Zr-H 0 Reaction - Rapid Burn...... 6-20 2

6.2 3.2.2 Sixty-Percent Zr-H 0 Reaction - Rapid Burn....... 6-20 2

6.2 3 2 3 Burn Hydrogen as Generated........................ 6-20 6.3 Plant Capabilities to Withstand Extended Design Basis Events..................................................6-20 631 Con tainm 2 nt Stru ctu re . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-27 1

632 Ic e Co nd en s e r . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 -27 6.3 3 Con tainme nt Spray Systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-27 7.0 Patisation Concepts..................................... 7-1 7.1 Inerted Containment..................................... 7-1 7.2 Halon Injection......................................... 7-2 73 Ig n i t io n Sou rc es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2 7.4 Augme n ted Coolin g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2

7.5 Filtered Vented Contain=ent............................. 7-2

\

7.6 A d dit io na l Con tainme n t. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3

, 7.7 Coupled containment..................................... 7-3 7.8 0xygen Removal by Reaction.............................. 7-3 7.9 P a s s iv e En c lo su re . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7- 3 7.10 P a s s iv e St ru c tu ra l . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7- 3 7.11 Other Concepts......................................... 7-4 8.0 Risk Assessment of Mitigation Concepts. . . . . . . . . . . . . . . . . . 8-1 9.0 Mitigativ e conceptual De signs. . . . . . . . . . . . . . . . . . . . . . . . . . . 9-1 91 Filtered Vented Containment............................. 9-1 1

9.1.1 General Description................................... 9-1 9.1.2 S ys t em Op e ra ting Mod e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-5 9.1.2.1 High Capacity / Pressure Venting. . . . . . . . . . . . . . . . . . . . . . 9-5 9.1.2.2 Low Capacity / Pressure Venting....................... .

9-5 9.1.2 3 P a r t ia l Va c uum Ve n t in g . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 - 5 9.13 Evaluation............................................ 9-6 s 9.1 3 1 Physic al Ef fectivene ss. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-6 9132 Schedule............................................ 9-7

, 9133 Cost................................................ 9-7 9.2 Additional con tainment Vo1ume . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-7 1

9.2.1 Ge ne ral De sc rip tion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -7 9.2.2 S yst em Op era ting Mod e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-7 9.2 3 Evaluation............................................ 9-9 9.2 3 1 Physical Effectiveness.............................. 9-9 9.2 3 2 S c h e d u le . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 - 9 9.2 3 3 Cost................................................9-9 93 coupled containment..................................... 9-9

9 3.1 General Description................................... 9-9 j'~3 9 3.2 S yst em op era ting Mod e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-9 V

9.33 Evaluation............................................ 9-9 9 3.3.1 Physical Ef fe c tiveness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-9 9.3.3.2 Schedule............................................ 9-11 9 3.3.3 cost................................................ 9-11 9.4 Augmented containment cooling. . . . . . . . . . . . . . . . . . . . . . . . . . . 9-11 9.4.1 General Description................................... 9-11

. 9.4.2 S y st e m Op e ra t ing Mod e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -11 9.4.3 Evaluation............................................ 9-11 9.4.3.1 Phy sic al' Ef fe c t iv ene s s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -11 9.4.3.2 S c h edu le . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -11 9.4.3 3 Cost................................................ 9-13 9.5 Cen tainme nt Ine rting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -13 f~]

m.

9.5.1 General Desvription...................................'.9-13 9.5.2 S yst em Op e ra ting Mod es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 - 15 9.5 3 E v alu a t io n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -15 9.5 3 1 P hy s ic al E f fe c t iv e ne s s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -15 9.5 3.2 S c h ed u le . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -16 9 . 5 3 3 Co s t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -16 9.6 Halon Injection......................................... 9-16 .

9.6.1 General Description................................... 9-16 9.6.2 S y s t em op e ra t ing Mo d e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -16 9.6.3 Evaluation............................................ 9-17 l

9.6 3.1 Phys ic al Ef fe c tiv ene ss . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -17 i

9.6.3.2 S c h ed u l e . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -17 l

9.6 3 3 co s t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -17

+

' (m q,) 97 Ig n i t ion sou rc es . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 -17

9 7.1 General Description................................... 9-17

) 9.7.2 syst em Op era ting Mod e s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-22 9.7 3 Evaluation............................................ 9-22 9.7 3.1 Physical Ef fec tiv eness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-22 9.7 3.2 Schedule............................................ 9-22 9733 Cost................................................ 9-24 10.0 Evalua tion of the Alterna tive Concepts. . . . . . . . . . . . . . . . . 10-1 10.1 0v e rv i e w . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 -1 10.2 Evaluation Basis....................................... 10-1 10 3 Concepts to Prevent or Minimize Hydrogen Combustion.... 10-1 10.3.1 Containment Inerting with Nitrogen. . . . . . . . . . . . . . . . . . . 10-2 10 3 2 Halon................................................ 10-3 10.3 3 Ignition sources..................................... 10-4 10.4 Cor.cepts to Increase Containment Capacity for Overpressure Events................................................. 10-5

(-}

v 10.4.1 Filt ered Ven ted Con tainment. . . . . . . . . . . . . . . . . . . . . . . . . . 10-5 e

10.4.2 A d d i t io nal Con tain me n t . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 -6 10.4.3 Coupled Containment.................................. 10-7 10.4.4 Augmented Cooling.................................... 10-8 10.5 Concepts to Mitigate the Combined Effects of Hydrogen Combustion and Containment Overpressure................ 10-8 11.0 Conclusions and Recommendat ions. . . . . . . . . . . . . . . . . . . . . . . . 11-1 11.1 Conclusions............................................ 11-1 11.2 R e c omme nd a t io n s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11-1 Appendices

(~)

v

n. n- .-,

- - , + - -y-'- . e-

SUMMARY

( A preliminary evaluation of events beyond the design basis leading to reactor fuel element damage and generation of hydrogen at the Sequoyah Nuclear Plant has been performed. The purpose of this evaluation was to develop additional information about such events ,

and to ccmpare the relative merits of concepts that have been suggested to mitigate the consequences of such events. This information serves as a basis for further TVA action.

Our ccnclusions based on this study are:

1. The risk due to core damage events is sufficiently small to allow continued operation of Sequoyah. This is based on the following findings: i

a. The risk from these events is comparable to that at Surry which was used in the Ras =ussen study (WASH-1400); recent information has indicated that the risk is less than originally estimated in that study.

The differences in containment design between Sequoyah and b.

PWR's with large " dry" containments do not lead to increased l; risk.

c. The Sequoyah containment can be expected to' survive a Three Mile Island-type event with burning of hydrogen up to at I

() least that produced by 20-percent metal-water reaction.

4

d. TVA has taken substantial measures since TMI to reduce the i likelihood of core damage due to small loss-of-coolant accidents.

2. If the licensing process allows, TVA should not proceed with implementing a mitigation concept at this time. This is based on the above assessment and the following findings.

a. The one concept (inerting containment with nitrogen during operation) that fully prevents hydrogen combustion has serious safety problems and major costs due to large plant d owntimes.

b. Other concepts provide partial protection, but involve additional risks or. major technical problems.

c. Because no concept was identified which provides substantial additional capability to handle hydrogen without major additional risks or technical problems, further study of options could lead to a better decision.

i i

viii

"- W ' -- -m- , .+7

%:s% a N . =,hh ,

= / s - - - - -

Venting of containment, especially filtered vented containment, The 3

provides very little protection for hydrogen effects. '

() pressure increase rate is too fast for realistic vent sizes.

However, these concepts can provide protection for other slower overpressure transients.

4. If a mitigation concept must be selected now, we recommendEven though igni installation of the ignition sources.

not provide complete protection, they would give the plant operators a way to ensure that the hydrogen burns at low Work needed to implement this concept hydrogen is slow.

includes:

a.

A development program for the ignitere. g i'gg

b. A program for analyzing a range of events involving core damage to identify timing of events, the rate of release of hydrogen to containment, and the time and spatial i

dependence of hydrogen and oxygen concentrations.

4

c. Selection of locations for igniters.

d. A testing program to verify the system.

e. A program for analyzing the effects of combustion on equipment in containment and the containment itself, f.

Procurement of ignters and additional hydrogen analyzers,

g. Installation and testing.

Our preferred approach is:

1.

For Sequoyah unit 1, full power license, demonstrate by analysis that the risk is no greater than for current operating FWR's and verify by more detailed analysis the containment's capacity

~

to tolerate limited hydrogen burning. l~

2.

Find out if some of the technical problems with the concepts can be resolved and the additional risks minimized through a research and development program.

3 Continue our investigation of the basic events to improve ourTh understanding.

and may give additional insight into how the events may be prevented.

4. Actively participate in NRC's proposed rulemaking proceedings A probable by developing and sharing information on the subject.

outcome is a decision on how severe an event (degree of core damage / core melt) should be designed for.

O ix ,

/,_ . 7 ~;7"'~

TECHNICAL

SUMMARY

Purpose The purpose of this report is to explore the phenomena associated with the generation and release to containment of substantial amounts of hydrogen due to events leading to core damage and to evaluate the relative potential of various concepts put forward to control these phenctena. The report addresses these matters for the Sequoyah Nuclear Plant. The overall objective is to provide guidance for further TVA action in regard to these matters.

Background

Light water cooled nuclear power plants are designed to prevent and mitigate accidents which could lead to significant core damge (i.e.,

reduced fuel pin integrity due to overheating and oxidation of the 4 zirconium leading to cladding failure). Major design measures, i analysis, and equipment have been used to provide assurance that such damage won't occur. In addition to the attention given the subject by the nuclear steam supply system vendors and the Nuclear Regulatory Commission (and its predecessor, the Atomic Energy Commission), TVA has expended considerable attention by inhouse

experts on accident prevention and mitigation. Nonetheless, significant core damage did, in fact, occur at Three Mile Island unit 2 (TMI-2). This event called for a reassessment of the likelihood and consequences of such an event even though the'

() ccnsequences at TMI-2 were much less severe than other events considered in plant design.

i Hydregen Effects l

In order to produce reasonable thermal efficiencies, reactor coolant systems operate at moderately high pressures and temperatures. If

, the reactor coolant is lost due to rupture of the system or other reasons, cooling water must be injected to carry off the energy released by the decay of fission products contained in the irradiated fuel. If cooling water is not injected or is not effectively conveyed to the core, the fuel pins heat up. At elevated temperatures, the zirconium clad oxidizes (if steam is present), resulting in a loss of clad strength, release of energy, and production of hydrogen.

The hydrogen is released to the containment building. -If the amount is sufficient, the hydrogen can burn in the containment, resulting in a large energy release to containment, increasing the containment temperature and pressure. If the burn is-big enough and fast enough, the containment could fail due to excessive pressure, releasing fission products to the environment.

. If core cooling is reestablished in time (as was done at Three Mile

, Island), the damage can be limited; if not, progressive core damage occurs, leading to complete collapse of the core (core melt).

g- If complete core melt occurs, the hydrogen and. other noncondensible

(,j gases could cause containment failure due to overpressure even without X

-. . c .

, . - ~ - , , - - - , , - . -- . - , . , - - . - , - - - - ~ -.n n

f hydrogen combustion.

Events Which Could Lead to Core Damage O)

Core damage cannot occur if core cooling systems function after an accident. -The cooling systems are redundant and independent with redundant sources of power. Therefore, multiple equipment failures or operator errors are required to occur for failure of the cooling systems. We have studied the three most important initiating events:

. the large loss-of-coolant accident; the small loss-of-coolant accident, and failure of secondary side heat removal system. If these inititing events are accompanied by failure of all core cooling systems, core damage will occur. There is some time, before significant core damage exists, for the operator to take remedial action to restore cooling (from a few minutes to about a couple of hours, depending en the event), thereby arresting damage. In this time, the water in the reactor boils away and the core heats up.

If remedial action is not taken, oxidation of the fuel proceeds fairly rapidly (in a few minutes time) until a significant fraction of the clad is oxidized (50-75 percent). If remedial action is taken in this time period, damage can be arrested while the core is a coolable geometry. If remedial action does not take place, melting occurs and the core fragments fall to the bottom of the reactor vessel.

Af ter a period of time, vessel failure can occur due to heating and melting, dropping the fuel-clad-vessel metal mixture (corium) into the reactor vessel cavity where it reacts with water to produce steam and with the concrete to produce noncondensible gases. Hydrogen is

{s ~)s produced by oxidation of the core supports, the vessel itself, and the remaining clad.

It has been postulated that eventually the core could melt its way through the containment concrete leading to releases of fission products to the ground water. Recent information indicates that this will probably not occur.

Seoucyah Containment Design -

The Sequoyah containment, along with several other plants, differs from the containment used at TMI-2. Sequoyah uses a pressure suppression type containment, wherein a large bed of ice is used to condense steam and remove energy. Some other plants use a pool of water instead of ice. The pressure suppression concept leads to smaller containments with lower design pressures (typically 1/2 to 1/3 the size and 1/3 to 1/4 the design pressure of " dry containments")

and increased compartmentation. It is expected that these '

characteristics could lead to a different response to significant hydrogen generation. The large, passive heat sink (ice or pool) is

an excellent raeans of removing heat; on the other hand, the lower volume results in higher hydrogen concentrations (for the same amount of hydrogen produced) while the lower design pressure allows less pressure buildup in containment.

i l Xi i i

I

Prevention Prevention is more useful than mitigation; TVA has responded t o the TMI event (through its Nuclear Program Review) with a number of steps 4

f~s to prevent accidents like that at TMI. These steps included:

1. Increased training of operators, improved procedures, and improved information to the operator to: (a) reduce the likelihood of operator errors negating the safety features provided to prevent and mitigate accidents; (b) help the operator diagnose events; and (c) help the operator respond to unexpected conditions.

2. Features to remove gases from the reactor coolant system which might prevent cooling water from getting to the core.

3 Improved equipment to further prevent the occurrence of a small loss-of-coolant event like TMI-2.

4. A special test program at Sequoyah to demonstrate natural circulation cooling of the core and provide experience to the operators.

Risk TVA has evaluated the risk due to core damage using methodology similar to that used in the Reactor Safety Study, WASH-1400. The results indicated that the risk for Sequoyah is roughly the same as for other pressurized water reactors; that is, the containment has somewhat less capability for accommodating noncondensibles, but this

() small additional risk is balanced by the presence of passive heat removal capability and other advantages. We believe that the preventative measures incorporated since the Three Mile Island event further reduce the risk.

Inherent Capability Our analysis demonstrates that Sequoyah could safely withstand an ~

event like Three Mile Island unit 2 which included combustion of hydrogen equivalent to 20 percent metal-water reaction. That is, if all the hydrogen produced as a result of oxidation of 20 percent of the fuel clad burned, the resulting containment pressure would be approximately equal to the predicted ultimate strength of the containment. When coupled with TVA's preventative measures, this inherent capability provides increased confidence that the risk from core damage events is acceptably low.

Mitigation Potential Notwithstanding the above, we have performed a preliminary evaluation of various concepts suggested for mitigating the effects of core damage (including the effects of hydrogen). These concepts are listed in table l'along with comments on their effectiveness and major d rawbacks. In general, our conclusions were:

- Some concepts promise to be effective for protecting against

() failure due to hydrogen burning or failure due to other xii

'V cverpressure causes, but no single concept is effective for both, r~ - Scme concepts, although effective for either hydrogen or (s,s) overpressure, pose significant safety concerns.

- To provide substantial protection for all core damage events would require a combination of mitigating concepts (i.e., one concept to mitigate hydrogen combustion and one to mitigate other overpressure causes). <

- The most widely publicized concept, filtered vented containment, provides little added protection from the effects of hydrogen but involves a very significant risk of its own since it uses deliberate release of radioactive gases to the environment.

- The potentially most effective mitigator of hydrogen effects (containment inerting) is used by TVA in its Browns Ferry Nuclear Plant. However, we have concluded that its use at Sequoyah could actu- '

ally increase risk due to reduced inspections inside containment and increased hazards to personnel.

- It appears that with additional research and development, an acceptable mitigation scheme could be developed if required.

Mitigation Strategies Based on what we know at this time, we believe that the following levels of added protection might prove feasible if required.

O - Increase the Amount of Hydrogen that can be Accommodated 1.- Use distributed ignition sources to burn hydrogen at low concentration (4-6 percent) to minimize buildup.

2. Use halon injection to suppress combustion.

3 Provide an additional containment structure with a high capacity mixing system to reduce the concentration.

These can be used individually or in combination. Item 2 requires researca to identify the long range incontainment behavior of halon to assure that essential equipment would not be damaged. It appears that these steps singly or in combination could increase the capacity to withstand hydrogen effects by 50 to 300 percent.

Ignition sources are preferred over halon injection because of ccacern over the affect of decomposition of halon in the long term.

- Provide Complete Protection from Hydrogen Combustion

1. Only containment inerting will provide complete protection.

Containment inerting has additional inherent risks (reduced safety system reliability because of reduced containment access for inspection, maintenance, and repair)

() and high cost.

xiii

- Provide Protection from all Core Damage / Core Melt Phenomena that can Lead to Overpressure Failure, Including Hydrogen Combustion

() 1. Sig'nificant protection by providing ignition sources and/or halon injection and one of the vented containment concepts.

2.- Complete protection by providing containment inerting and additional containment The costs for these options would be on the order of to per plant plus an operating c6st of up to

, . dollars per year. Note that there are other containment failure modes (isolation failure, . steam explosions, missiles) that would not be mitigated by these measures, although these failure modes are considered less likely than overpressure.

Summary .

Our investigation shows:

- While ice condensers and other suppression type containments react differently than " dry" conta,inments, the overall risk is not significantly different.

- Sequoyah has inherent ability to withstand the effect of burning the hydrogen produced by oxidation of about 20 percent of the fuel clad.

(

- Several concepts investigated appear to have some potential for mitigating the effects of hydrogen combustion or containment overpressure. However, all of the concepts capable of mitigating hydrogen effects pose additional safety concerns or have major technical problems.

- Some concepts investigated appear to be effective in preventing uncentrolled combustion; some appear to-be effective in preventing overpressure failure due to,noncondensible gases. No concept identified appears to be effective for both.

- A combination of concepts might provide additional protection, but substantial research and development will be required to obtain an acceptable design.

I e

s Xiv

[m A -

TECHNICAL

SUMMARY

TABLE 1

( MITIGATION CONCEPTS I 1

i Concept Comments-Inerted Containment Effective for preventing hydrogen combustion; significant safety j concerns due to reduced access to containment equipment and hazard to

personnel. Ineffective for containment overpressure protection.

Halen Injection Moderately effective for preventing hydrogen combustion; questions exist in regard to Halon's long term behavior and effect on equipment; ineffective (may aggravate the situation) for

containment overpressure protection.

Ignition Sources Moderately effective for all but rapid releases; potential for misoperation; minor impact on containment overpressure.

i

() Filtered Vented Centainment Limited benefit for protection from hydrogen combustion effects; moderately effective for containment overpressure protection; the deliberate release of noble gases is a major safety concern.

Additional Containment Somewhat effective for protection against the effects of (and preventing)

, hydrogen combustion. Effective for containment overpressure protection; expensive.

t Coupled Containment Similar to additional containment but somewhat less effective because second containment may not be available at all times the first unit is opertting; safety concern in regard to impact on nonaccident unit.

E50105.01 n

v XV l

., . n-z , =  ; ._ 3g , ,,. , _ 33

• ' du 1s :a:7 + - , ni; ^ - .m :S:: m ,;,,, ., .; g

. , - . _ ._. . _ _ , m.. - . _ . ,. , _, . - _ _ . , _ _ ,, . .

1.0 Purpose n

( ,) The purpose of this study was to assess current containment capabilities and to identify and evaluate concepts for the mitigation of the consequences of core damage, including hydregen generation, in order to prevent failure of the containment. The results are intended to provide additional insight into the events that could lead to significant hydrogen generation, the effects of such quantities of hydrogen, and the relative benefit, cost, and drawbacl:s of various mitigation concepts. While the subject is applicable to all TVA nuclear plants (particularly those incorporating pressure suppression type centainments), Sequcyah plant characteristics were used in this study.

A N..]

e

~> 1-1

2.0 -Scope F1 4

() ~

The scope of this study included six major areas of effort:

2 .1 Identificaticn of key accident sequences - - A discrete i number of accident sequences were selected for use as representative I

cf a wide range of types (small loss-of-coolant accident (LOCA), large LOCA, etc.) and severities (moderate core damage, severe core damage,

.4 etc.) of potential accidents at Sequoyah.

2.2 Analysis of present Sequoyah containment for key accident sequences - - The physical consequences of the key accidents and their effects on the present SQN containment were analyzed. The current

. risks from these key accidents without additional mitigation were

! also estimated.

i 23 Identification of mitigation concepts for key accident sequences 2

- - Concepts for mitigation of various consequences of the key accidents were identified based on prior industry ,

experience, feasibility, or current regulatory attention.

2.4 Development of mitigative conceptual designs - Conceptual designs were produced for the identified mitigation scheces. This included

! general arrangement of components, bill of materials, and identification of principal engineering factors.

2.5 Evaluation of mitigative concepts- - The conceptual designs were i cach evaluated from several standpoints - physical effectiveness, risk reduction, cost, design and construction schedules, nuclear j safety, operational impact, feasibility, etc.

l 2.6 Recommendations for future study - - The experience gained in these efforts has made it possible to recommend specific areas uhere further research and development would be useful.

I l

i i

j-e 4

2-1 l

. . _ _ _ . . . ~ , _ _ _ .._ _ _ _ ._. _ _ . _ , . . . . ~ . _ , _ . _ . . . _ . _ ,

f

/"T k/

s 3.0 Introduction One of the design basis accidents for Sequoyah is a large LOCA. The consequences of this accident would include large mass and energy releases from the-primary system rupture followed by the generation of s=all amounts of hydrogen in the reactor core and in the containment. Sequoyah safety systems currently include both active and passive measures to mitigate these accident consequences. To mitigate the primary fluid loss and the accompanying pressure and temperature effects, high and low pressure injection systems actively reflood and cool the primary system, containment spray systems actively cool the containment atmosphere, the ice condenser passively cools the containment atmosphere, and the containment structure itself is designed to cetain its integrity at pressures above those expected. To handle the hydrogen produced, recombiners remove it from the containment atmosphere. These systems have the capability to alleviate all of the consequences of the current design basis 3

accidents with adequate margin.

The accident at Three Mile Island involved moderate mass and energy releases from the primary system through a stur,k-open relief valve.

This was effectively a small LOCA, and the appr opriate safety systems were automatically initiated as designed. Manual override of the f high pressure injection system by the operators at TMI led to uncovering the core for longer periods than in any of the design basis

() accidents. This resulted in extensive core damage accompanied by hydrogen generation significantly above that considered in design. For this reason, TVA is reassessing the effect of hydrogen on plant safety, i The best approach to prevent hydrogen from jeopardizing plant safety l 1s to prevent its formation in significant quantities by maintaining

! adequate core cooling. The emergency core cooling systems are redundant, qualified systems whose perforsange has been evaluated conservatively in great detail to ensure their effectiveness. They have received the most attention of any plant system. Nothing in the TMI experience indicated that they were not adequate. TVA therefore believes that the Emergency Core Cooling Systems (ECCS) at Sequoyah are sufficiently reliable to perform their intended functions, and that complete failure of the ECCS is of sufficiently low probability. However, TVA is responding to the TMI accident through commitments made in its TVA Nuclear Program Review (" Blue Book" and " Green Book") to make improvements in several areas. These include instrumentation changes to make more information available to the cperator, hardware changes to make improved system operating modes available for the operator's use, and intensified training of the operators to respond more effectively to inadequate core cooling incidents. As a result of these post-TMI efforts to incorporate imprcvements in both hardware and procedures, we feel that the likelihood of Sequoyah experiencing inadequate core cooling has been significantly reduced. The Nuclear Safety Analysis Center of the l

() Electric Power Research Institute has estimated that risk has already 3-1 ,

i l

bccn reduced by a factor of three to five for U.S. light tater

- reacters followir.g TMI, and further improven.ents will reduce the risk i by another facter of two to four below h' ASH-1400 (3-1) estimates.

4 i

l 1

i t

t 4

1 i

! l i

l h

! l j

i I

i I;S AliYD.P4 i

h 3-2 i

i

O. 4.0 Eackground of the Original Design Sequcych inccrporates numerous features to provide for safe operation and to mitigate a wide range of accident and transient sequences.

These features have been studied previously to determine their effcetiveness in dealing with the plant design basis events. This section briefly describes the present design bases, analytical tools for evaluating events, and the implications of TMI on the present design bases.

4.1 Physical. Plant Features There are three principal barriers to the release of radioactivity during normal cperation and accidents. These barriers, the fuel rod cladding, the reactor coolant system, and the containment perform indepcndently and have been designed to be highly reliable and effective for current design basis events. A detailed discussion of the fuel and reactor coolant system may be found in Appendix A.l.

4.1.1 Containment

. The containment is a large structural volume designed to contain the

! radiation, mass, and energy released from a breach of the reactor j coolant system and the nuclear fuel cladding. The Sequoyah j

() containment is a free-standing steel structure with a free volume of 1.2 million cubic feet. The design pressure of this structure is 26.4 psia which exceeds the calculated pressures resulting from any of the present design basis events.

The containment is divided into three compartments: The lover compartment, the upper compartment, and the ice condenser compartment (figures 4-1 and 4-2) . The lower compartment completely encloses the reactor coolant system equipment. The upper compartment contains the refueling canal, refueling equipment, and the polar crane used curing refueling and maintenance operations. The ice condenser connects the lower compartment to the upper.

4.1.2 Passive Containment Heat Removal Systems The ice condenser contains a, proximately two and one-half million pounds of ice. .This quantity of ice is capable of removing all of the heat released by a large LOCA for the first hour. The steel centainment shell and the walls and equipment located inside centainment are heat sinks that remove energy from the containment atmosphere, thcreby reducing the containment pressure and temperature during an accident.

4.1 3 Active Containment Heat Removal Systems

,s In additi,on to the ice condenser, several systems remove heat from

?

4-1

t

~

/

( (

%.v) i I

. , i I f f.Y

( J 5-a- ,,,,

ISP 0fCE A t1 .i cuq wi%10" ( ['

- o/..'- - - -

l a = :-

f

- - - 1 1-y q, _ . .. m . . , mi - - .

=4__ _ . -

,, m s c .~

w,m....,..,

s .3'5 . .,. x. ~

I, - . _J -

i cCoJas l , +- .___.

==7 t sounsaar L,._ ,

4

- -ict c e,este tavle cc 1anw4m-

. D',

l t i,

,ss;rj'g3 li , WMWi. -

K i D _g Icl tal.11 L ,g,.

ext ,wis l

,u_niv,t

), i

' i

"'{. 3-,SLAcfCR

. . .ktA0 ,

,. , l y j seau , , , . , me, . ,

; j i  !

-l ,,,,,,,,,,,,,,,,

.umai cmss t.

s,

,f._.- . . _ ,

Aj w' l i cceanat soo ca,sv .

(

l. --- - - \ "'55'J 5"'fL8 }

i  ! r* i

/ .

% = <== 1 I I f.I  !

l .

.l g -

.~

.'- - -  : e.

p 5".rM/t ict sto .

1 } w.tr

%, , if24fLai (a! vtm I ,

l Dxas I l , .- i.:,;- i

\ %y a..

__~ ,,,,.

.yg ,-y,

_ .r , _l '""

p s '.

I, ll . ~ .

,. i c 3 J

/ y N,

. N,!

m

t. .

s\  ; ...

~--~

,m , _e _,,m,..,..,,.,

.C 4 su .

p . _

((_ ~._ .g. ^

y ,,' 8 IWW'4C

'Ax '

W ..w T

. ,w..: .

.m.

q y

k g'.

g;s.

c 69 ..

yV s

j m.

_.y'*'

.N .. .

..9 1 1 1. .

e ca stAtC%1. .3 rc a ma tt g 4.it 0 g ggg gg , _ j cLAe t w3 vb=(

f: .

e ..

.~

FIGUEE 4-1 g) t w

CDIJTADGENT CROSS SECTICIT 4-2

_- - m . u _ -.

a n .o r k n -- ; m - m e m ---

,.  ; .;c. - . n L::,&h:4 a .. -

I l'D, q, .

l. l 8

t

- I ItQ e

'1

)' -

i:, uca - ,,

• PCF.A8 CRAT , 64 - F LT ,

a:_

] g.-[ , p. ,

cau o ut -. ,

,u _ _ -

s t -t .e.. - ~ att COs0thilt ,

=

= s m:

} I l g'tLssLatete 3 -

, l

" i l~ /

/

S]q

,e .

a us I L gf- ]

[- ._

s C.u; .sa m.t q ,'! $ cerantwc

-4 i

j yd52f , ,

gg n== g r-L r .._.. ~ 4,.t w... . . -

. r us. & .. . mu-.

/'n(

e _

vt.v q 3 i d u.al ocons ,

wll _w-  ; , e. . ._ ,' ( - - -- p 'y l

IDi j --

-r l:

.m - -

[/// ' .6 p

l

,f 3y i i 3i l.

im'*L5 t)Q._' .

j i. .;!lj,i i

i

_ '~*5

.=

u--- s -A.v 7

/ w

.' 4 m 1l .o . ._. v a f :fx.

- - - + .

, ,/ - ; g:.

-f .,a f;

% . -- y 1.,g

. f,

/,/

e

' W b.- I'. ._il .,

!' t*,

. cn n tw r

-~ ,., ..

- .y %, , , ,. ,

ir

- . . ~ . . +

.:a ._ ;--____

, .l,

,,, .  ; -o.n ext - ,_

,. / *

} l .,( ,'

~I .." 'o; =. '.

\, f

• 3,

,1 ~

'1

• 5 . '- 4- N *

, [- /. :.. k -

g4 : . . . . .

-. utssvacia  ;

y

! 4 ,

• s

,,', stuutau , .

- 8,  ;

/ _.*. a ., /,'.*. ,

' ,.; f.

/ ewtos - .

cias aic<e ,

sea evic a ts staa ns ecu, a .f., .

. . .s.

r ,

FIGURE 4-2 CQETAIIGENT CROSS SECTION .

D, '

,s ,

l 4-3

--- - -._~. . . _ _ . _ _ _ _ , . _ _ , , , _ "" '

w- , ,

_B- _

-- =$ 5.sh a ,

Y

the containment during a LOCA. The containment spray system removes heat frcm the containment atmosphere initially by the addition of O' unter from the refueling water storage tank (RWST). Eten the water supply in the RUST is exhausted, the supply for the pumps is changed from the tank to the emergency sump located in the containment.

Kater from.the sucp is run through heat exchangers in the residual i heat removal (RHR) system and the containment spray system to remove

! heat frca the containment sump water and thereby from the containment.

4.1.4 Hydrocen Control Systems Several safety systems ar e incorporated into the design of the Sequcyah containment to mitigate the effects of hydrogen production following the original design basis accident. The hydrogen mitigation systems function following the accident to detect, mix, and consume the hydrogen produced. Each subsystem presently included in the

, Sequoyah design is briefly described in Appendix A.5 as to function 1

and operation principle. The systems employed will adequately process all of the hydrogen postulated in the original design basis accident scenario with 100-percent margin. They include hydrogen analyzers for postaccident monitoring, electric hydrogen reccabiners for removal of hydrogen, a containment mixing system to prevent pockets of hydrogen at higher concentrations than the general containment airspcce, and a small hydrogen purge system as a backup to the recombiners.

4.2 Present Design Eases j The centainment is designed to assure that an acceptable upper limit j of leakage of radioactive material is not exceeded under design basis accident conditions. The design basis for the containment is generally taken to be the complete severance of a large reactor coolant system pipe at a locstion which results in the worst t consequences for the containment. See Appendices A.3 through A.3 2 for further discussion of the containment design bases including mass and energy release and fornation of hydrogen.

4.2.1 Accidents The limiting design basis accidents for the containment are loss-of-coolant accidents (LOCA) due to their potential release of radioactivity. A spectrum of LOCA's are studied from small pipe breaks (pipes less than 2" in diameter) to a break in the largest main ecolant pipe (approximately 30" in diameter). Most studies perfcened to date concentrated on the large pipe break because it was believed to present the greatest challenge to the fuel cladding, the containment, and the plant safety systems. The analysis of such accidents is extremely complex and requires the consideration of heat transfer, fluid flow, neutronics, and the interaction of safety systems.

4.2.1.1 I; ass and Energy Releases Followingia postulated rupture of the Reactor Coolant System (RCS),

, A 4-h

- _ . , ~ - _ _ _ _ , _.

steam and water are released into the containment. Initially, the water in the RCS is subcooled at a high pressure. When the break cecurs, the water passes through the break where a portion flashes O. to steam at the lower pressure of the containment. These releases continue until the RCS depressurizes to the pressure in the containment (end of blowdown). At that time, the vessel is refilled i by water from the accuculators and Safety Injection (SI) pumps. Water i flowing from the accuculators and SI pumps start to fill the downco=er causing a driving head across the vessel which forces water into the

. hot core. When' a level two feet below the top of the core is reached,

~

the core is assumed to be reflooded and totally quenched, leaving only decay heat to generate steam.

The LOCA analysis calculational model is typically divided into three phases: (1) blowdown, which includes the period from accident occurrence (when the reactor is at steady state full power operation) to the time when zero break flow is first calculated; (2) refill, which is from the end of blowdown to the time the ECCS fills the vessel lower plenum; and-(3) reflood, which begins when water starts moving into the core and continues until the end of the transient.

4.2.1.2 Design Basis Noncondensible Gas Production Another source of mass that must be considered in containment design is the' production of noncondensible gases during the accident. It is possible for nitrogen gas to enter the containment from the passive emergency core cooling system components. This source, while not significant for containment pressurization, is considered in the

(} containment analysis. More importantly, the production of flammable gases is of concern. The more severe loss-of-coolant accidents postulated for plant design may result in the production of excess l hydrogen which ultimately ends up in the containment air space.

Depending on the final volumetric concentration of the hydrogen and the quantity of oxygen present, the gas can be a fire hazard, an explosion hazard, or too diffuse to chemically react.

T l

Several sources of hydrogen have been identLfied for the LOCA. These include dissolved hydrogen in the reactor coolant system, which is released during' blowdown; radiolysis by fission product decomposition of the coolant; fuel cladding oxidation from high fuel cladding temperatures concurrent with steam flow over the clad; and corrosion

of zinc-coated surfaces in the containment interior by spray water concurrent with high ambient containment temperatures. Approximate quantities of gas produced by each mechanism are shown in table 4-1.

It should be noted for- the design basis event that the zirconium cladding reaction is important in the short term and the other production mechanisms are important in the long term.

See ' Appendices A.3 through A.3 2 for further discussion of the containment design bases, including mass and energy release and hydrogen formation.

i

O 4-5 s -

n - .- --vv v,-+- = e.nn, , - - e vm- p.<ew,~o-r y e - - - v

TABLE 4-1

("8

)

DESIGN BASIS HYDROGEN SOURCES Source Potential Hg (1bm)

Zircenium (43,204 lbm Zr) 3 percent RXN 57 Primary System Dissolved Gas 6 Aluminum Corrosion (long term) 142 Zine (long term - paint, galvanized steel corrosion)* 1800

'Most of the eine is in the ice condenser components which remain at a low temperature and low corrosion rates until late in the accident.

s.

N,. ,/

E50107.01 4-6 e

(,

N ./

4.2.2 Analytical Tools Some of the most difficult problems to analyze are the thermal 73

() hydraulic problems encountered in reactor safety. Even with advanced conputers, it is difficult to completely analyze all but one or two dicensional thermal-hydraulic problems. A brief introduction to the ar.alytical tools available to study the loss of coolant accident and their limitations is necessary to a complete overview of the predicted plant response to the events described in this report.

4.2.2.1 Primary System Thermal Hydraulic Programs The primary system pressure, temperature, and flow response to various postulated accidents is calculated using state-of-the-art computer codes. The software tools numerically approximate the mass, momentum, and er.crgy conservation equations for a series of interconnected volumes and flow paths simulating the reactor system. The Westinghouse WFLASH (4-1) and SATAN (4-2) codes typify the design tools used for Sequoyah primary system analysis. These codes, as well as more recently developed programs that encompass greater detail, are discussed in Appendix A.4.1.

The currently available codes treat the behavior of steam and water mixtures but do not follow the presence of a noncondensible gas (e.g. ,

nitrogen, hydrogen, and other noncondensibles) in the primary system.

This licitation is not a problem for the design basis accidents since the production of large quantities of noncondensibles is not predicted.

('  ;

4.2.2.2 Containment Thermal Hydraulic Programs The current design basis analyses performed to determine containment design parameters consider the addition of water or steam from a spectrum of pipe breaks in either the RCS or the secondary side of the steam generators. Cc=puter codes have been written which again use the conservation of cass, energy, an< momentum to evaluate the pressure and temperature inside the cent intent at any point in time.

Heat transfer coefficients, ice melt rates, and other critical factors required to evaluate the performance of the ice condenser are determined empirically for various mass and energy release rates.

The LOTIC (long term ice condenser code) computer code (4-3) was developed by Westinghouse specifically to evaluate ice condenser containtents and was used to evaluate the Sequoyah containment. As in the case of primary system computer codes, the LOTIC code has not been written to treat hydrogen separately. A detailed discussion of the capabilities of LOTIC may be found in Appendix A.4.

4.2 3 Predicted Plant Response Studies of the design basis LOCA at Sequoyah indicate approximately 1-3 percent of the core zircalny fuel cladding is calculated to

(

U' 4-7 e

exidize based on conservative 10CFR Appendix K computer models.

Althcush the core is calculated to uncover and to heat up to the temperature where zircaloy oxidation rates are high, the core is quickly reflooded and cooled by the emergency core cooling systems,

, thus limiting oxidation. Production of hydrogen from the oxidation 0' . of core metal is therefore not expected to exceed 60 pounds of gas.

No hydrogen generation is predicted from the vessel steel or from stainless steel cladding of the vessel since these materials do not heat up to reaction temperatures during the accident. The containment spray system is designed to operate throughout the accident, and the ccmbination- of spray water and containment temperature will corrode the centainment painted surfaces and any aluminum present during the long term. Additional hydrogen will be produced by fission product radiolysis of water in the core and sump. This source also generates oxygen during the process which must be considered for inerted containment designs.

The initial release of hydrogen from core cladding oxidation after a design basis LOCA is accommodated by dilution in the containment air space. Long term hydrogen sources are accommodated by the use of the hydrogen recombiners which heat the hydrogen / air mixture initiating chemical recombination to produce water vapor. Typical hydrogen production rates from the above sources are shown in figure 4-3 Assuming at least one recombiner is started in the first day following the accident, the containment hydrogen does not exceed the 4 percent flammability limit as indicated in figure 4-4. If no reccmbiners are available, the hydrogen purge system can be started several days into the accident to maintain hydrogen concentrations within acceptable limits.

() In the event of the design basis LOCA, the maximum containment pressure is 11.8 psig. Figures 4-5 and 4-6 provide the containment pressure and temperature response curves for this event. As can be seen in the pressure response curve, the containment rapidly pressurizes due to large mass and energy releases that occur during the blowdown period. The containment pressure starts dropping as the steam condensation rates in the ice condenser exceed the steam releases from the primary system due to decay heat and steam generator sensible heat. This trend continues until ide melt at approximately 4000 seconds. Over the next several hundred seconds, the containment pressure rises to its maximum value. The containment spray systems then maintain the pressure within the containment design limits once the ice has molted and in the long term gradually reduce the pressure to near atmospheric.

4.3 Ramifications of TMI on Current Design Bases i

The accident that occurred at the Three Mile Island facility on March 28, 1979, has led to the reassessment of the design basis hydrogen

production at Sequoyah. The event was primarily a small break LOCA with inadequate core cooling. The LOCA resulted from a stuck- .

1 i

~

O 48 s e- *u '*M s.sa%2 eMPy.* *9 6n aFr'M 4 % *w e & s*e.K '4 -

, - [ 6=+

.L Mwr ' e " :' a =

,..h

O- n v u)

H DROGE1 T PROD TCTION RATES O. -

+ . . . .

. J;

. # +-

o  :  :

..:......... .... f.

g. ....._.........:................:...............i................:................-  :

M i TiESTINGIIOUSS CALCUIATIONS i i.

i. i.

g

.i i. .i i. NRC BASIS i.

. t

. .................................... ....... .... ............ ................... p.2

~

' O  !. i i i ~+

.i D- . . . . . . .

.p

. O _ .............................................. ...............................

. AJ.l (f)

C v .  :  :  :  :  :  : .

M .,  :

+#

p. . .

i . ., . .

NO i . .

O i . * . .

4

.r............ .... ................................................................. . . . . . . . . . . . . . . ......................................................................

H O

H H i. .

O oA.., . . . .

4..... ....... . . . . . . . . . . . . . . - ....................................................................

(A.

O . .

a N.4 0_ 1

s. . .

\s i 5

m 3 ._ ... ..s.........: N .

5

......... w....r .e o r e .r. .a. d..i. .o. .l.ys.

.i.s. . . ..: . . . . . . . ... . . . . . . .: . . . ... . . . . . . . . . . .: . . . . . . . . . . . . . . . . .: . . . . . . .. . . . . . . . .

5

' i i*

i i

i

' t-  !  :

i_ _ _ _i _______e+_c___rosiott or _ _ _ . ___N .

_+:_ ___ . - - - _ - -. -_- __. _ + ._ _ _ _* ~_ _i _ _ _ _.i. _ _ _ _ _ _ _--_______

W d- - - - - -

%.-----------------------:-------------*-~~:-------'-~~-~~-:----"W*---'--'-*----------~-

N - e sump:radiolysic.

~. . .

O o . .

0.0 10.b 20.0 30.b db.b Gb.b Ob.0 7b.b 00.0 CD.b 100.0 TIlIE days l

FIGURE 4-3

()

\ J b

u ')

e C01rTAINME1T T H 'DROGEir VOLUME PERCE1T T

.e

d. .

.  :  :  :.  :  :  :  :.  :. e +-

. r

-4 r . . . . . .

e ................:................:................:................:................

L,,

• i i i i WESTINGIIOUSE CALCUIATIONS i

q i i i NRC RiSIS i

tQ O

p

...............j...............j................j................j................ 1 H, RECOMBINER @ 24 HRS -j-------- f.1 i i..

i i . .

! M

. . . .A.

6-

. , a.

~ .

c .

tQ g

s f4  :  :  :  :

m m $- ...............i.................!.................!...............!...............!.................!.................!.................!.................i.................

4 -

i i 2 i i :N i  !

i D C- ................:................i.................:..............i.................i.................i.................i...........................

O .

p. .- ................. ................ .......................................................................................................................... .................

a, O .

O .

Q.- ...................................................*..................................................

O c .

N. . ................e................................................................................................................................................................

O  :.  :. . .

O . . .

O. . . . . .

o_ .

50.0 60.0 70.0 00.0 90.0 100.0 0.0 10.0 20.0 30.0 10.0 TIME (days)

FIGURE 4-4

hqungal7 Egbragen 91ubg

% C

, 1 v

. ................. ......................................O -H

.................{..................4...................4....

...................)...................g.................. 6. ....-..--...--.- .----*.-..--..- -

. . . . -l.-  :

r,,  ;

H.4 . . . . .

.)........................................l.........................................

-4  :  :.  :.  : -

O .

o Qt ...................i..(................j...................j...................i...................i................... H

..................j. ................j...................j..................j...................j................... -

-c .

r ...................g...

....... ..........................g...........................

............g...................

..................{... ..... .

....................{................... L

. e  ; 3 .

v . . l .

. I .

m c/)

...................................... ..................... .... ............. 4....................

i.n O

Os-y n v  :  :  ;  : i

o CQ m g

_. 4 os,s

..................i....................q...................:....................

.................;................... . , ia.-t

......... .. ! ...,...... ......g...................;....

...g...................[...................

_. he 4 m

rA g

s . . . .

t-4 W A .

...s...................

, Z  :.  :

C) * . . .

H~

D .

U2 .

F-4 m Ed o  :  :  :  :

,,s w d 02  :  :  :  : .N O

p at3 c{I f"' .

.........g................

c% ...>.......................................>...................

.,,,4

. ..... ;e

6. A to 43 g .

.........t................ ,...:...................;................... ...................

__4 ........

s o  :  :

O ... ....

A y e

.~l

.. ......;...................p...................

MZ9

(_) ........ ~

p A  :  :  :.  :. i l 4

m.c4 ............................. ... ................i.......................................

_ l oy . - 1 oZd  :  :  :  :  :

........ A M u)

I s  :-  :  :-  :  :. -w I

,v) i s a s s .

o p 1 l

04T 0*dT 0*0T O'0 0*9 0't 02 '

(msd) aanssana l 4-11 l l

I

9equngaly Eighragen 9tubg O no

............ ...... ........ . . . . . . . . . . . . . . . . . . . . -4 3r.... .....................

...................:...................q.......

...........;...........l.......t....................;................

. f .

.................. 6...................;..... ............ 6.........f..... ......................)................... _

/ .

t

[Il i

! e'  !

.......................s........................................................

. ,i .

5.-  :  : r  : .

.....................p.....................................:..................

-4 .................................. .

O  !.  !, s ei . i.

~

$ $ 5 $ $

07  :

o

...................f.......l..........j............[......{...................j...................j..~............... v-4 r%f p

4........... 1.........................

.................... ....... ........................g......g.............................. . 4........................ .

......... ........................t..... l

..................g...... ............;.............+ ....p....................;....... ............ ...................

g . . s

.............. . , 4......

g ..................$...................., . .

9. ..................j...................

. ,1 .

...................... .................................... ........... 1 .......................... .

,s.

c4  :

m.  : ^

(A L. A ...........D....... . .

..................4.....................e..............l,....a.

-y a 4...................

- GA

p. . . . p I .

H r=/ a .

c1 H k  :  :

. et O

(m w.

Q.

C. l .

n O \D n O .:7 i e . . . . 4.. 1 .

7- Q ,, 'M

^

p n

..........g...... ............ ...... ................... ................... .......... 3 ..i... g

......... ......... 4 v y ...................;......................................;...............................g..;...y................... .

.......... k......;......... ........g......................................(...........Q.4...;...................,.

tr'

...........o...... .

................g m o o.. .

. o o o o* ...n o..n oo m o d o..;o.oomoo.no M, U r.-4' p o ..lo . WH r

, o  ; .

4 8 .

~

,4 N

......... a... ...................... ......... .........;...........o..s...:................... .

,_4 d b kI

...!...................!......... 1d $ h-4

. c".

A

.........!..........t5'.i....!....................

t w :

. - 1....i....................

I .

: O .

i

.

---4  : . N .

.........;..............Y.......................

,.1 .

i

'. ~-4

. . b .

: rn -

em F---

l ea  !  :

4 co j .

cJ H

H b $4 q I

.........,...................i................... ;Oi

........ ...... .. ............................ ......... A C .........;............ 4..... ..................

/W ........ .............. ................ . . .......... . . . . . . . . . . .fr.3 o. . . . . . . ....................s......................... -

N e ....

. l *

....;...............g... ................... .

...........t.................

A g gr., ..........................l...y................... ,-

g s

---4 . .  : i .

.... .............;... ...............g......... ,Z o7" O  :  : mp A D -

.,4 .........................3..;....................

:: ~

Q ................'..................l........

.........}................l..'................... .

. $oy  :  : -

.................. i...................:.........

o%d q m in 8 .

. . g . ,,

. . . g

. . . =

. g .

r) o . .

I v i e i e i - ,O ,_4 0'013 0'0T3 0'00T 0'09T O'0ET 0'06 0'09 (4) samvaamal 4-12

cpen pressurizer power operated relief valve which was not recognized by the cperators. - The emergency core ' cooling was severely degraded by improper operator action which was based on conflicting

~O 1==tr==e=t tie = ==a tre1#1=s tuet a1a eet cever the eve #t- Tae degraded cooling led to hydregen generation which exceeded that normally considered in plant design.

4.3.1 TMI Challenge to Containment Desi:;n Following the accident at Three Mile Island, it was known that a significant amount of hydrogen was produced and burned, generating a 28 psig pressure spike in the containment as shown in figure 4-7

(4-4). This pressure spike, had it occurred at Sequoyah, would have

challenged the containment integrity since it exceeded the Sequoyah design pressure. It is not expected that the containment would have failed due to the design margin present. There is considerable uncertainty concerning the actual quantity of hydrogen produced and burned at TMI. Since the total hydrogu influences the maximum containment pressure following a burn, the various estimates for TMI are described below.

.Several separate organizations have examined the plant instrumentation recordings using various approaches to determine the amount of hydrogen produced during the course of the accident. The original analysis was performed by the NRC and indicates that 452 pounds of hydrogen were burned at IMI (4-5). Other organizations more recently have calculated a wide range of hydrogen burning from 400 pounds to 1160 pounds. Most agree that the data is sparse and not very c onclusive. For example, the containment temperaturg instrumentation indicated a temperature increase to a maximum of 190 F during the hydrogen burn, whereas a theoretical temperature of over 1000 F was possible. The results of the various investigations are presented in Table 4-2.

Several indications show that it is possible the hydrogen burn was localized. This could result in significantly higher temperatures in localized areas of the containment without affecting the general atmospheric temperature throughout the volume thus explaining the temperature dilem=a. The presence of steel and concrete heat sinks

. may also have absorbed significant amounts of energy.

! It is important to note that not all of the hydrogen generated was burned. Considerable hydrogen remained in the reactor coolant eystem both in the hydrogen bubble above the core and dissolved in the primary system water as can be seen in Table 4-2. Also, not all of the hydrogen released to the containment atcosphere burned since approximately 1.7 percent hydrogen by volume remained following the i burn. If the same conditions resulted at Sequoyah, more than 452 pounds of hydrogen could probably be accommodated without containment 1 -failure.

1 i+-13

- r, . ----.-.,,,y..- -- , ,--. - r v

9equagall E U brugen @tubg x, C.

t Q H

re H  :  :

:  : q

.......... .......... ........... ........... ........... ........... ........... ........... ........... ...... ... - co H F- .

H  :  :  :.  :.  :.  :.  :.  :  :

O

:  :  :  :  :  :  :  : O 4 .

. . C

............ ........... -d g

&4 r  :

V, . . . . .

.......................... . . .. . . . . . . . . ........ ... ........... o.

-Q F,-  :  :  :  :  :  :.  :  :.  :. ^

. y- . . . . U2 N

e/  :  :  :  :  :  :  :  :

. H .

O - 5 oy e i Q ..........j...........j....... .

...........j...........j...........j...........j...........j. ........ _.

g v :t cc

• H  :*

W =

-a  !.  :  :  :  : 4 3  !..

H o [.-4

..... ........... ........... ........... ....................... ............ .......... g i H  :

A .

c p.

0 .....!...........!............!............!............!............!..........!........... y 9 b

rn 1 H m  :.  :.  :.  :  :.  :.  :

i p- W .

t b--(

~

w  :.  :.  :.  :.  :.

3 (-t A  :

2

.... o . . . . . .

o.

v . . . . c9 A

),

r, A .. . '.

. 4 . . . . . .

y e3  :  :  :  :  :  :

3 H M* O C .

Y H j m  :

l l

l l

l C

s ........

. -. A s.

4 i, . .

s.

. o

\ l

:  :  :  :  : l 6

• e

• a e i 4 a a e 0*62 0*9% 0'C2 0'02 0*a 0*H O'IT 0'8 0'9 0d O'I-(21sd) HHnSSSHd ,

4-3 li t

</- '

l

n

("%

%J TAD 2 1,-2 THREE MILE ISLATO HYCROGEN PRODUCTIC:1 ESTIt',ATES f*e thod M wen Burned Hydrogen Prodcccd 0,.ygen ipletion 400-1160 lb. 670-1285 lb.

dP 1:ethod 554-5S3 lb. 1002 lb.

Core Heatup  !!ot Cetermined 440-1320 lb.

Details of / cove:

Determination Type H2 Burned H in Air RCS RCS H2 Generated Total by Analysis ib. (Urburned) Bubble Dissolved NSAC 1. Preburn 0 Depletion 0 680 300 63 1048 2 "

172 72 1203 2.1tarch 31 820 144

, 3. April 1-2 400 178 20 72 670

4. June 1 "

1030 52 0 2 1134

5. Auqust 2 "

1100 52 0 2 1214 4:-

Ae 158 !a ic4 144 72 1002 k

m Kemony 0 Depletion ses 872 158 184 72 1286 2

Sandia 1. Upper Core Heatup 1320

2. iower " 440

3. Best

- Guess

" 770 Battelle, Colux. bus dP 554 0 Depletion 1034 2

NRC Unknown 452

References:

1. Analysis Center " Analysis of Three Mile Island - Unit 2 Accident," Appendix HYD Electric Pcwer Research tluclear SafetfC-1, Duly 1979.

Institute,IS

2. Robert E. English, "Techr.ical Staff Analysis Report on Chemistry," Technical Assessment Task Force, October 1979.

3. Randall K. Cole, Jr. , "Gqnaratio of Hydrogen During the First Three Hours of the Three Mile Island Accident," Sandia Laboratory, NUREG/CR-091~.;, July 1979.

4. R. O. '.looton, R. S. Denning, P. Cybulskis, " Analysis Of the Three Mile Island Accident and Alternative Sequences,"

Battelle Colun. bus Laboratories, NUREG/CR-1219 January 1980.

077166/1

.o

l Essentially all the investigations that were performed ignore the temperature dilemma and use three basic techniques to determine the hydrogen generation at TMI:

' O 1. Oxygen depletion - Chemical analysis of the containment air space indicates that the oxygen concentration at various times following the accid (nt was less than that normally found in air. By assuming the depletion was caused by the burning of hydrogen, the amount burned can be determined.

2. Delta P - The hydrogen required to generate the pressure spike observed at TMI can be calculated theoretically since the energy released by burning is known.

! 3 Core heatup - Examination of the thermal hydraulics during the accident can be performed from the few data available to determine i the core heatup, cladding oxidation, and hydrogen generation.

The exact thermal hydraulic behavior is not well known, but the values generated using this technique agree reasonably well with the other methods.

4.3.2 Consideration of Inadequate Core Cooling The primary objective during or following any accident is to maintain adequate core cooling so that the core integrity is preserved and most of the radioactivity is maintained within the fuel rods. Additionally, significant hydrogen can only form when the zirconiumgelad fuel rods are allowed to exceed temperatures O of about 2200 F. The best approach in preventing core degradation, which results in the release of radioactivity and hydrogen generation, is to maintain adequate core cooling.

Diagrammatically, the events following a LOCA where adequate core, cooling exists are shown in figure 4-8. Dotted lines are included for processes that are not expected to occur. The diagram is modified as shown in figure 4-9 in the case of inadequate core cooling. The problem now includes a significant number of undesirable effects not considered in the design basis 4- analysis. Indeed, the key to the event is therefore the operability of the emergency core cooling systems. These systems have been evaluated conservatively to ensure their effectiveness and have received considerable attention in the past. One change made to the ECCS as a result of the TMI experience was to eliminate their actuation based.on low pressurizer water level, thus. improving reliability. TVA had identified concerns on its Bellefonte plant that, for certain small loss of coolant accidents where ECCS injection is limited because of the repressurization of the reactor coolant system, the operator might not have sufficient infor=ation to identify the appropriate corrective action. .The accident at TMI bore out tha validity of these concerns. As a result, the automatic status monitoring system which monitors components or systems important to safety, including the ECCS, is being extensively expanded. Other TVA O

4-16

FIGURE 4-8 LOCA WITH ADEQUATE CORE COOLING (See Reference 6-7)

J l : Cit /ARY CC*LAlli SfSTEM R#T'JEE l l3:EA4C=A A;'EE.iSTICSl l fAS$iLE G NERATICN l t '

lx:27S::.CCul -

4  ;

h v. ATER LEFT IN CCNTAt*WE;4T /ESSEL g l a r pg,.,r q rr-wa y l RE ACTOR VESSEL TEMPERATURE AND 8

PRESSu=E L e I l WATER BOfLC3F l l 8 8

EMERGENCY CCRE I I CCOLt.NT l g i I e 4 4 <r g g l r:E TEust ;.ruaEi ecETA -wATEa aE ACTIONS I

, g l l- g- l 8

l lr.<:EOGEN- AIR RE ACTICN L .g l

I I l r , '

l

+02 FEL C AD STATE STATE g l g

6 l 4

l 8 .

I 8

l:_.a<_utusEl . 3 l d 4 [ 4 4 l 3E SLU'.'I FISS!CN-PROCUCT GELE ASE . .. ,

CCNTA;NVENT VESSEL TO CONTALNMENT FAILURE as 1P l CC:E CIS ASSE'/3LY l WATER LECT IN REACTO9 VESSEL

<r <r h l C:45 CE54'S EMPE; ATUEE ( 5 METAL-WATER REACTIONS I d

dOA[ CE94lS STATE I l HYCoCGEN-AIR RE A'CTICN I  ;

I l 2EACTCR VESSE I TEAV GENERATION l  ;

Y EL T TH:OUGm _

WATER IN THE CCNTAINMENT VESSEL

-r 9 CC.Ri CESRIS STATEi .' .iMETAL-WATER REACTION l A .C TEvRE=:.T;RE t v

FISSION-PRODUCT lHYOROGEN! e RELEASE 13

} '

ENVIRCNS

,, { HYDROGEN-A!4 AE ACT:ONi l STE AM GENE =ATICN l

'} ,

e v

4-17

FIGURE 4-9 LOCA WITH DUGEQUATE CORE C00LII;G O l PRIMARY CCCLANT SYSTEM AUPTLRE l l 94EA< CHA ACT ERISTICS l l MISSILE G NER AT:CN l h

l COCLANT ELC'AOC AN l 4

%ATER LEFT IN CCNTA!NVENT VESSEL l CCCE PCAER CECAYl PEACTCR VESSEL TEMPERATUAE AND PRESSURE a

l WATER 80fLCFF l EMERGENCY CCRE CCOLANT 4

4, j

" E . 3 f I CBE TE'.tDE9 ATurte ' METAL-WATER REACTIO[

I I n .

I HvCROGEN- AIR RE ACTICN I

,, , a CO2 FLEL CLAD STATE STATE I

(CL A0 FAILUAE]

4 'P , r 19E SLLYdl FIS$1CN-P4CDUCT RELE ASE _ CCNIAiNviNT W5stL TO CONTAINMENT FAILURE l CCaE C'S A55EV3tvl WATER LEFT IN REACTC9 VESSEL h

l LC*E Ctsats itMWER ATURd ! !M TAL-WATER RE ACTICNS l h I I ccaf cesais STATE l lHYCsoctN-A:R RE AcrioNi .

r AcAGIUR VESSEL l STE AM GENER ATIC'J '

YELTTH:OUGH WATER IN THE CONTAINMENT VESSEL r

CCRE CEE9IS STATE j v6 TAL-WATER REACTICAl A *:0 TE9PELATURF h F'SSION-PRODUCT l HYOROGE N! -

RELEASE TO g ENVIRONS

, ImbhoGEN-Am st AttauNi  :

l STE AM GENERATICN ,' '

v 1-18 4

- - . - . _ - -- - -- - ~ . _ . . _ .

actions to prevent a similar occurrence at Sequoyah are described  !

below.

In the area of plant transient analysis and operator training, it was noted that the TMI operators were relying on procedures ,

that did not adequately address small LOCA's and were relying '

on .instrutentation (pressuriser level) which indicated adequate

, core cooling even though the core was overheating as indicated by other available instruments. Small LOCA's have not received the extensive analysis attention given larger LOCA's. As a

, result, the operators were relatively unprepared for the events  !

and conditions that occurred. To assure that this condition does not occur at Sequoyah, TVA is taking the following steps:

. 1. Improving operating procedures to assure they adequately reflect the design intent.

2. Participating in a joint effort with other utilities to study

small LOCA's, inadequate core cooling, and similar events for improved understanding of the phenomena and to determine required mitigative actions.

3 Based upon the studies outlined in 2, additional training will be given to operators to prepare them to recognize the events and take proper mitigative action.

In addition to improved procedures and additional analyses, several hardware and instrumentation changes are plannedi It .

has long been recognized that inadvertent opening of the

() pressurizer relief valves or safety valves or their failure to close could cause a small LOCA. For this reason, downstream 1

" block" valves were installed on relief valve lines and

tecperature detectors were provided on the lines to detect an 4 open relief or safety valve. At TMI, the relief valve stuck open, i but the operator, for various reasons, did not promptly identify j the problem. Sequoyan has direct, positive indication of relief

{ valve position in the control room. This will provide the i

! operator with reliable information to isolate the downstream block valve, terminating a transient if a valve should stick open.

i] (Block valves are not installed on safety valve lines since this I could prevent the safety valves from performing their essential function, and safety valves are much less prone to fail since the relief valves provide sufficient capacity for most events, j reducing the times the safety valves operate.) TVA is providing i

acoustic monitors to alert the operator to an open safety valve so that he can take appropriate steps to mitigate the event.

t Other hardware changes are planned to ensure that the operator i has adequate knowledge that inadequate core cooling may be approaching or existing. These include:

(

4-19

.m .

~ - . ~ _ , ,. .

sh ' V'? 2 Av s -,lr - o .

w N&g , i,,Z.;;Q j [&C y , ,~y ; "- h j

, , _ . _.- . . , . - , . _ , _ ,-,_ _ _ , . _ . . . . _ , . . _ _ . . _ _ . . _ _ , _ _,. , ._.,____.m. , , _ _ _ . _ .

1. The range of the incore thermoccuples has been extended to Sive the operator a positive indication of temperatures in the core region even in an event where the core is

()

'O inadequa*,ely cooled.

2. Instruments have been provided at Sequoyah which will give a direct indication to the operator when the primary system is approaching the boiling point. ,

3 Instrucents to indicate in the main control room the water level in the reactor pressure vessel are being provided at Sequoyah and will be installed by January 1981.

4. Sa pling capability of the primary system coolant will be provided. This capability will allow detection of failed fuel (4-6, 4-7).

DA03: USAHYD.P2 V

=

I I,_ ')

1 4-20 l l

I 5.0 Efsk Analysis of Present Segucyah Capability for Key Accident Sequences v

5.1 Cencral Descriptien There are several potential accident sequences which could lead to plant damage. Most of theae accidents do not invc1ve risk to the p u'.lic . For this study, the important accidents are enes involving core damage. These events have a relatively low probability of occurrence but have relatively severe consequences. In this study, a risk evaluation of the present Sequoyah design was made by ccmparison with the reference plant (Surry Nuclear Plant) of the WASH-1400 study. Vhere applicable, the Surry results have been adjusted to Sequoyah assumptions in order to provide the greatest parity for ccmparing the results of the plant studies. The results of this risk evaluation provide an indication of the relative risk from the Sequoyah Plant and which accidents should be used as a basis for evaluation of mitigation techniques for degraded core phenomena such as hydrogen evolution, rapid energy releases, etc. These results are discussed in the remainder of this section. For an understanding of the nethods used to arrive at risk in L' ASH-1400 and this study rarer to Appendix B.

The dcminant accident sequences resulting in core melt and containment failure for the WASH-1400 (Surry) study are shown in table 5.1.

Sequoyah dominant sequences are shown in table 5.2. A comparison cf the Surry-Sequoyah release category failure probabilities is shown graphically in figure 5-1.

5.2 Analysis Assumptions The initiating event probabilities used in KASH-1400 for the large and small LOCA's were used in the Sequoyah analysis. Thg6 probability of interfacing check valve failure was reduced from 1X10 per reactor year to 1X10 ~ per reactor year at both Surry and Sequoyah to account for improvements in monitoring procedures instituted since the WASH-1400 study was performed. The frequency of loss of offsite power coupled with g e failure of the diesel generators to function was taken as 1X10 per reactor year at Sequoyah. The frequency of loss of offsite power at Sequoyah was based on data specific to the Sequoyah plant.

For large LOCA sequences and transient sequences, the probability of containment failure due to a steam explosion was taken to be 1X10-3 for both Surry and Sequoyah. For small LOCA sequences, the probability~gor containment failure due to a steam explosion was takeg to be 1X10 for bcth Surry and Sequoyah. WASH-1400 used used 1X10~

for the probability of failure due to explosion for Surry, but recent infcrmation has indicated the steam explosion risk is much less than believed in 1975. For containment leakage, the_ probability of ecntainment isolation failure was taken as 1X10 for both Surry and Scquoyah. These assumptions are summarized in table 5-3 Symbols are explained in table 5-4.

For Sequoyah, failure of the reactor protection system scram function 5-1 i

TAl:11 5-1 m if0BABILITIES OF D0.' LINA'?T ACCIDE';T ST.QUT.NCES RR k'Asi!-1400 (ADJUSTED)

RELEASE CATECORY .

I::itiating 1 2 3 4 5 6 7 Eunt AB - a AD - a ACD - 0 AB - c AD - c ABjgjg

- y 10

• 2x}g 9 1x10 11 AD4x10 - S~8 2x10~8 1-10*12 1x10]

AF - a AB - 6 AH - a ADF - c AH - c 4x10 it 1x10~9 A!!3x10

- S~' 1x10~10 lx10 0 1x10~11 A ACD - a AlIF - y AF - 6 ATIF - c-5x10 12 2x10 il 1x10 s jxio 16 AG - a AG - 6 9x10 12 gx30 9 3.0x10 ~8 2.7x10 16 2.4x10~8 2.2x10 s '5.9x10"$ 3.8x10 s 3.0x10"?

SB-a t SB-y t SD-a SCD-g1 t St DF - c SD-c 1

4x10 20 1

3x10~10 1x10 S1 5x10 H - @' 3x10~20 3x10 6 3x10~13 S1 CD - a SB-6 t St il - a SD-$

t SB-c t S tli - c 7x10~13 1x10~30 3x10~20 6x10 $ 2x10~S 3x101 St SF-a 1 StHF SF-6 t St HF - 3 6x10 71

~

3x10 ~12 3x108 4x10 16 SG-a t SG-6 t -

3x10 12 3x10 s

- - - ~~

6.9x10~30 6.8x108 6.1x10 s 1.2x10 s- 6.2x10 s 6.0x10 6.0x10~6 S2B-" SB-Y 2 SD-a 2 . SDG-g2 2 S2B-c SD-c 2

lx10 12 lx10 8 9x10~10 1x10 S2D 2x10- @8 8x10'S 9x10 ~8 SF-a S217 S2-a S2 CD - c S2 !! - C 2

S2H - $s -

lx10 11 2x10" J0 6x10~10 1x10 2x10 s 6x108 S2 S2 DC - a SgB - 6 S2F - 6 S2 1E - C 1x10 8

~

2x10 12 4x10 10 1x10 ?

'. SC-a 2 SC-6 2

9x10 12 2x10 ~C SC-a 2 SC-6 2 -

2x10 30 9x10 s 2.0x10 ~8

~

2.3x10 a 2.3x10 ?

~

2.3x10'C 2.5x10"? 2.5x10 ? 2.0x107 5-2

i TABLE 5-1 (Continued) 9 PROMMLITIES OF DOMINANT ACCIEENT SEQUENCES TCR k'.\S!!-1400 (ADJUSTLD)

RELEASE CATEconY, fr.itiating 1 2 3 4 5 6 7 hent 1MLd - u TMLd - y TML - a TML - c 3x10~10

~

7x10 ? 6x10 10 TML 3x10 - @to 6x10 T!!L$~

? - c 6x10~8 T T'!LB -6 TlQ-a TI:Q-g Tl:Q - ;8 2x10 ~8 3x10~10 3x10 10 3x10

• TclQ - a T10!Q ,c 1x10~10 1x10 8 3.0x10'? 3.0x10'8 3.0x10 ?

~ -

4.0x10 s 3,oxto ? 1,oxios 1,oxio ?

RC - a RC - y~11 R-a R-c ~

2x10~13 3x10 1x10 ~10 1x10 ?

R PJ - 6 1x10~11 RC-6 1x10 12 O

~

5.2x10 12 5.0x10 11 1.1x10~10 1.1x10 10 1.0x10 ~8 1.0x108 1.0x10 I V V 4x10~8 4.0x10~9 4.0x10 s 4.0x10'S 4.0x10~10

~ ~ ~

10TAL 3.3x10'? 3.3x108 2.7x10~8 3.1x10 I 4.5x10 ? 4.5x10~8 3.9x10 t t

O 5-3

T/.HLE 5-2 FROBABILITIES OF Ir031NANT ACCIDD.T SEQUENCES FOR SEQUGYAll 0(~N Release Categories init i it ir.g .

Event 1 2 3 4 5

/.G -

• AlIF - y,6 AG - y,6 Ah - y,6 6.0x10~11 3.0x10 s 6.0x10 s 4.Cx10'?

A /.!T - a AD - y,6 3.0x10 11 1.9x10~7 8 7.0x10 s 7.3x10 s 6.7x10~7 4.0x10~9 3.9x10 SG-a S i lE - y,6 SgG - y,6 S FG - S St D - y,6

- ~

1.8x10~11 9.0x10 s 1.8x10 ? 5.9x10 12 1.05x10~8 St S alg - a SgCFG - SgH - y,6 4,gx30 7 9.0x10 12 3.1x10 12 1.7x10 ~7 1.5x10 ~8 1.1x10 s 1.1x10 ~7 2.0x10~7 i S2 D - y,6 O'V SC-"

2 S2 1T - Y,6 S2 G - Y,6 S2 CDF - S 6.0x10 11 3.0x10~7 6.0x10~7 5.9x10~11 3.5x10 ~8 S2 1:F - a S2 DF - y,6 S2 CIE - St il - Y,6 3.0x10 11 3,2x30 s 3.2x10 11 1.6x108 S2 S2 CDF - a S2 CDi - y,6 S2 FG - S S2 L - Y,6

~

5.9x10~13 5.9x10~9 _. 2.0x10~11 3.0x10 ?

S2 CHF - y,6 . SCFG-S 2 SD-2 3.2x10 ~9 1.0x10~11 3.5x10 s S2 HF - S 3.0x10 ~9 7

4.0x10 s 3.9x10 ~7 6.9x10~7 6.1x10 6.1x10 s TD2 B

- Y,6 1.0x10~8 TE2-a b

TA L - Y,6 TB2-@

3 1.0x10~56 0.0x10 ~9 1.0x10 s 3 T T LDO - y,6

[V B 2.4y'0~9 5-4 .

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

i  !

TABLE 5-2 (Continacd) l l

TROEASII.ITIES OF DU!il!! ANT ACCICE!.T SEQUE!!CES FCR SEQU0YA!! l Release Categories Initiating .

Event 1 2 3 4 5 T

l TL-S 3

T G - y,6 3 )

8.0x10~11 1.0x10 ~8 l i

TB ID ~ 0 B W - Y,6 2.4x10 ta 5.4x10~7 l l

i l

1.0x10~7 1.0x106 1.2x10 ~7 1.8x10~7 1.6x10~7 ]

! l

l

~

l V V -

l l

4 x 10 s

+ .

4x10~9 4x10 s 4x10~9 4x10~10 l TOTAL 1.6x10~7 1.5x10 ~8 1.1x10 ~8 1.0x10 ~8 9.9x10~8 l

9 4

~

4 1

l 1

i e

1 i

i i

l i

i

}

O 1 5-5

-% + c%3',i".'8 l' 4" p , g wh%%*h ,,

e - n-m-- --, - , ~, y_ e-w w w r , .r v.. , . _ _ _ _ - -m

--. O. O. . . . .. .--. . . . . . . . . . . . ,i.

RISK COMPARISON S'URRY VS. SEQUDYNI 1

FIGURE 5-1 WASH-1400(ADJUSTED)

SEQUOYNI

_4 lo Y

• 4 M

g I-B

• Q 10 ~5-

. Ed '

d m -

i

! h d

i

!e 10-6.

t 7 7 m ' -

i 4

41 l:^a. f[4 u

i 10 -7

_ m.1 b ,

RC 1 RC 2 RC 3 RC 4 RC 5 RC 6 RC 7

i i RISK CATEGORIES i

TABLE 5-3 Failure Probabilities Used for the Sequoyah Analysis A S1 S2 TA TB i

-4 -4 -3 -6 IE 1.0 x 10 3 0 x 10 1.0 x 10 1.0 x 10 3*

-6 -6 -6 B 1.0 x 10 1.0 x 10 1.0 x 10

-6 31 1.0 x 10

-6 -6 B2 1.0 x 10 1.0 x 10

-5 -5 -5 C 1 7 x 10 1 7 x 10 1.7 x 10 1 9 x 10

~ ~

-3 D 3 5 x 10~ 3 5 x 10 3 5 x 10

-4 E 1.0 : 10 F 3 3 x 10 -3 3 3 x 10 -3 3 3 x 10-3 G 6.0 x 10-N 6.0 x 10-N 6.0 x 10-N H 4.6 x 10-3 1.6 x 10-3 1.6 x 10-3 A K 3 5 x 10-5 3 5 x 10-5 3 5 x 10-5 3 5 x 10-5

() -4 -3 -4 L 3 0 x 10 8.0 x 10 1.0 x 10

-6 -6 M 1.0 x 10 1.0 x 10 0

-3 -3 2 3 x 10 2 3 x 10

-5 -5 P _ 4 3 x 10 4 3 x 10 Q

-5 5.2 x 10

-3 -3 -3 X 1.8 x 10 1.8 x 10 1.8 x 10

-6 -6 -6 Z 1.0 x 10 1.0 x 10 1.0 x 10

-3 -3 -3 CF** 1.7 x 10 1 7 x 10 1.7 x 10

-4 -4 -4 HF"' 3 0 x 10 3 0 x to 3 0 x 10 CHF** 8 3 x 10-6 3 2 x 10-6 3 2 x 10-6 m

)

l 5-7

)

i i

-..=l.O j?& w

, - . . . ~ . , ; < ;...

l

TABLE 5-3 (contin.)

Centainment failure nodes p.

LA -3 -4 -4 -4 -4 c4 1.0 x 10 1.0 x 10 1.0 x 10 1.0 x 10 1.0 x 10

-2 -2 -2 -2 -2 g 1.0 x 10 1.0 x 10 1.0 x 10 1.0 x 10 1.0 x 10 6,6 99 99 99 99 99

1. For the T BB 2 sequences, TB = 1.0

• • CF, CHF, and HF are con =on mode failures whose probabilities of occurrence are unique and are not independent.

~

A

~

m N._

58

TABLE 5-4 Explanation of Symbols

(}

A - Large LOCA (> 6")

IE - Initiating Event B - Loss of all electric power B1 - Loss of direct current power B2 - Loss of alternating current power j, C - Containment spray injection i

D - Emergency core injection 2

E - - Emergency cooling functionability i F - Containment spray recirculation 0 -- Containment heat removal (ERCW, CCW)

. H - Emergency core recirculation l K - Reactor protection system L - Auxiliary Feedwater System M - Secondary side relief valves 0 - Containment emergency safety features (C, F, G)

. P - Primary side relier valves (fail to open)

Q - Primary side relief valves (fail to close)

R - Reactor pressure vessel rupture S1 -

Small LOCA (6"5 x < 2")

S2 -

Small LOCA (6 2")

TA - Transient, loss of all ac power TB - Transient, loss of main feedwater system

~

j V - Interfacing check valve failure q

X -

Air return fans Ice condenser Z -

c4 - Containment failure due to steam explosion

/3 - Containment failure due to containment leakage di - Containment failure due to hydrogen burn 6 - Containment failure due to overpressurization g _

Containment failure due to meltthrough of the containment basemat 4

~

O

!' 5-9 i

i r

e- e e---< , s. . v-m i ---e- - - - _-- w,w,.-r-.- >,.r., mww---. , m,,e --m~,-,--y , y -.,, n ., ~ , + r-'N'*"*r' ='"'eee'*?"

was not assumed to initiate a core melt sequence based on analyses performed by Westinghouse.

53 ccminant Sequences The more important Sequoyah accident sequences from a risk standpoint were fcund to be:

1. A loss of main feedwater transient (T) followed by failure of the de power system (B ) followed by containment failure due to hydrogen burning or ove)rpressurization.

2. Small LOCA's (S followed by failure of the containmant heat removal system (G)rop)llowed by containment failure due to hydrogen burning or overpressurization.

3 Small LOCA's (S ) foll wed by failure of the emergency core 2

cooling recirculation system (H) and the containment spray recirculation (F) followed by containment failure due to hydrogen burning or overpressurization.

4. Small LOCA's (S,) followed by failure of the emergency core core injection system (D7 followed by containment failure due to hydrogen burning or overpressurization.

5. Small LOCA's (S2 ) f 11 wed by failure of the emergency core recirculation system (H) followed by containment failure due to hydrogen burning or overpressurization.

e~s 5.4 Sequoyah and Surry Comparisons A comparison of Sequoyah to Surry indicates that the Sequoyah accident sequences are relatively less probable for high consequence categories (risk categories 1, 2, and 3) and relatively more probable for low consequence categories. The major factors in this difference are the containment failure probabilities for hydrogen burn and overpressurization. Sequoyah has an ice condenser and, as such, has a smaller free volume than the Surry dry containment. Similarly, the Surry containment is capable of withstanding a higher internal pressure than is Sequoyah's. The Surry assumption for probability of containment failure given a hydrogen burn or overpressurization varied from 0.1 to 9.9 The Sequoyah evaluation assumed a containment failure probability of 0.99 for sequences including hydrogen burn or overpressurization. The hydrogen burn and the overpressurization are the dominant containment failure modes in release categories 4 and 5 and thus have the largest single influence on the probabilities in this category being higher than for Surry.

The Sequoyah analyses of the dominant sequences and their contribution to the overall risk were performed based on the results of a limited number of detailed evaluations of accident sequences and other qualitative studies.

G C' 5_10

.. . . .- =~  :~~' '

i

The analysis indicated that the risk associated with Sequoyah is similar to the risk associated with the Surry plant. The ice condenser containcent does not appear in this abbreviated study to te either significantly better or worse than the dry containment for either design basis events or events that result in core degradation.

l l

I I

't l

DA02;CUTCY.01 4

9 1

1 I

h i

h 5-n 1

i i

+ = - - - -> .- .<w.,.y.,_, y w, .. . 3wm,, . ,,_ ,_ _ _ , ,___ __ .

. , . $ h _-

e.. ,..

6.0 Accidents Beyond the Design Bases The accidents considered by this report for the Sequoyah different plant involve chsnges from design basis considerations in the area of:

phenomenology inherent in the accident; modifications to the analytical tools because of the new phenomena present; analyses of the plant response to more severe events; and an assessment of theEach of th risk present during degraded scenarios.for the events identified as key 6.1 Phencmenological Consequences of Key Accident Scenarios The key accident sequences that have been identified result in several phenomena which are more severe than previously considered in designa These pheno =ena include production of noncondensible basis events. "

gas significantly beyond that normally considered in plant desi6n, i .y the possibility of hydrogen deflagration (burning), and theIn addition, the w possibility of hydrogen detonation (explosion). energy release different.

Large release rates occur late in the transient, and, in some cases, they exceed the rates during the blowdown period atThe the beginning of the accident. accidents leading to large consider two types of scenarios:

quantities of hydrogen; and accidents leading to large quantities of hydrogen and followed by total core melt with the subsequent production of concrete decomposition gases.

(1)

The basic sequences chosen for additional censideration are:

a large pipe break in the primary coolant system with a c reactor coolant system break with a concurrent loss of high pressure injection; (3) a small reactor coolant system break with the loss of all ECCS due to an inability to recirculate water from the emergency sump after the refueling water storage tank has been emptied; and (4) a transient caused by the loss of the main condenser, loss of all auxiliary feedwater, and loss of high pressure injection.

Using the designation provided in the Reactor' Safety Study, WASH-1400, .

Case 1, is designated AD, Case 2 is2 S D, Case 3 is a modification of the S D sequence, and Cass 4 was selected by TVA due to the probabil$tyofoccurrence. The mass and energy releases and hydrogen generation rates for Case 4 are essentially the same as Each the event WASH-1400 designated TMLB' which is a loss of all ac power.

of these events, if allowed to proceed A wide time unchecked, results inbyathe span is covered meltdown of the reactor. core.The large break proceeds to core melt very choice of these events.

rapidly while in the Case 4 transient; core melt does not occur for several hours after the initiation of the event.

6.1.1 Hydrogen and Other Noncondensible Gas Production _

Hydrocen gas is generated in the core melt sequences In an inadequatefromcoreboth cladding oxidation and vessel oxidation.

coolir.i; event ( AD, S D, TMLB' , etc.), the cladding oxidizes rapidlyduetotheh$ghtemperaturespresentuntilsufficientsteam 6-1 .

. 9 t 2w= M- ? (( _

is no longer available for the reaction. Various researchers esti= ate as cuch as 60-80 percent of the core may oxidize before melting occurs. The for:ation of eutectics between the cladding and the fuel

( cceplicates the physics of the process and leads to uncertainties in the total production. Reactions of this magnitude would produce sufficient hydrogen to approach explosive concentrations in the containment at Scquoyah (approximately 1200-1800 pounds hydrogen).

i Core celt gases are also generated by the decomposition of limestone in the presence of the molten steel, fuel, cladding, and control rod material from the ccre. Both carbon dioxide and the flammable carbon j onoxide are produced during the decomposition process. Decay heat providee energy to maintain the core debris at elevated temperatures

during this period.

6.1.2 Hydroren Ccmbustion Hydrogen is a flan =able gas that when oxidized can release large amounts of energy. For each pound burned, approximately 61,084 Btu J of heat is produced. Hydrogen can burn in two ways, depending on the ccncentration present. The first is known as deflagration which is a slow burn characterized by low velocities of flame propagation.

The deflagration concentration is approximately 4 percent hydrogen by volume. Above this concentration, the hydrogen will burn upward from the point of ignition due to the buoyancy forces present as the hot gases evolve during the burn. Once concentrations of 8 percent and above are reached, the flame will propagate in both upward and downward directions due to the turbulence present. It is estimated that 50 percent of the hydrogen will burn in a 5.6 percent hydrogen

() mixture, whereas nearly 100 percent of the hydrogen may be consumed in the deflagration of a 10 percent hydrogen mixture. A flammability limit chart fer hydrogen, air, and steam mixtures is shown in figure

] 6-1. It should be noted that the initial te=perature and pressure of the mixture affects the flammability limits.

Once a cencentration of 18 percent is reached, the potential for detonation or explosion exists. The reaction begins as a deflagration but transitions to a detonation when shock waves, formed during the turbulent burning, reflect from containment walls and reinforce the i burning process until the reaction proceeds at the speed of sound or above in the mixture. Substantial energy can exist in the shock wave which impacts surfaces in the containment. Fortunately, large cetal-water reactions are required before this volumetric

, concentration limit is reached at Sequoyah. ,

Several modes of burning and their icpact on the Sequoyah design are discussed below:

1. Slow Burning During Release

, The hydrogen can burn in air if both hydrogen and oxygen have sufficiently high concentrations (greater than 4-6 percent for hydrogen and 5 percent for oxygen). Disregarding the dilution

, effect of steam, at least 12-18 percent zirconium water reaction is required to reach this high hydrogen concentration.

6-2 ,

, _ . - . -- . _ _ . - _ _ _ _ _ . . _ . . _ _ _ _ ~ - - _ _

FIGURE 6-1 Q

Flammobility Limits goo o 75 F - O psig 10% reaction container

---{>---- 300 F - O psig g ambient temperature 216 F

--G--- 300 F -100 psig 80

\ -

20 10% reoction container

! ombient 280 F 100 % reaction -_

co 100*/ reottion container gm n 2l6 F

• 60 ' , , \'j 40

\

^f /

/f-Assumed 6

p' / / / / / detonation limiM V

,g iI

\%-60'E N

O' 4

/ /\ s - o Flammobility limits f \

///

.up

\

y . ,. -

\,

~~

20 '~~ 80 10 0

$00 80 60 40 20 0 Hydrogen, v/o

)

6-3 '

• * * "v-+-- -e---- - _, ,, _ _

J If the hydrogen is reseased slowly from the reactor coolant system (over a period of about 30 minutes for a 25 percent metal-water i reaction) and is burned as released or periodically burned as the containment hydrogen concentration exceeds 4 percent, the resulting heat released to the containment could be removed through the centainment sprays and heat exchangers and by transfer of heat to the structural heat sinks. Burning the amount of l

hydrogen estimated to have been burned at TMI would add less than 30 million Btu's to the containment. This amount could be easily handled in a 30-minute period by the containment sprays and heat sinks without a significant increase in containment pressure.

2. Combustion of Hydrogen (Deflagration) After Accumulation

] Energy released by burning hydrogen is deposited in the containment atmosphere. If released over a short period of time,

{ the temperature of the air in containment will be greatly increased, resulting in increased containment pressure. The containcent temperature and pressure then drop as heat is transferred to metal and concrete in the containment and also removed by the containment spray system.

3 Sporadic, Localized Hydrogen Burning or Detonation l If hydrogen concentrates in localized areas of containment and i burns / detonates locally on a periodic basis, a condition somewhere

! between 2 and 4 results where small pressure / temperature spikes

. occur, decaying to a steady state until the next event. This

! condition might occur if there is inadequate mixing, or if a

^

periodic'or concinuous ignition source exists.

i

4. Hydrogen Detonation i If hydrogen is available in sufficient concentration (15-20

percent) with air, a detonation may occur. A detonation is characterized by extremely rapid combustion which takes place within a high velocity shock wave. The possibility of occurrence and the nature of a detonation is affected by geometry, presence of ignition sources and diluting media (such as steam). At TMI, at least a rapid combustion (deflagration) occurred. The exact l

effect of a detonation on the Sequoyah containment is unknown.

Since it would depend greatly on the lccation and total quantity of hydrogen involved. The potential exists fer containment i

overstressing.

l

! 6.1.3 Core Melt Energy Consideration Due to Steam There are three basic periods of steam generation to consider in degraded core accident sequences. These periods are: primary system blowdown; core slump into water remaining in the bottom of the reactor vessel; and reactor vessel melt through into water in the lower

reactor cavity. The blowdown period mass and energy releases for the large ,or small break sequences are essentially the same for both design basis and extended design basis accident sequences. These 6-4 -

l

. i I

relea:as are controlled by the break size and the initial reactor coolant system pressure and temperature. Active escrgency core cooling systems have little impact on this portion of the mass and energy releases. For a transient case such TMLB' , the blowdown period (3

(_) occurs in intermittent steps. The reactor coolant system heats up and in the process increases in pressure until the relief valves on the pressuriser open. The loss of mass out the relief lines drops the reactor vessel temperature and pressure until the valves close and then the process begins again. Decay heat is the driving mechanism for this event. The blowdown period for a large break lasts around 30 seconds. Small break blowdowns last longer, but the initial rapid depressurisation of the reactor coolant system occurs in a relatively short period of time. The blowdown portion of the TMLB' event lasts several hours.

Based en the MARCH calculations described in section 6.2.2, core slump into the bottom of the vessel occurs in 1200 seconds for a large break LOCA, 7000 seconds for a small break LOCA , and 13,400 seconds for a TMLB' transient. Only small amounts of water remain in the bottom of the reactor vessel at the time of core slump and containment ccnditions are not appreciably affected due to this part of the transient. Shortly after core slump, reactor vessel meltthrough occurs. When the molten mass of fuel, reactor internals, and reactor vessel metal falls into the pool of water standing in the reactor cavity, large quantities of steam are generated. The energy release rates due to vessel meltthrough are on the same order of magnitude as a large break blowdown. This energy addition was considered in the present study.

fN Steam explosions, while a potential source of large amounts of energy, were not considered. The physical phenomena associated with such explosions are complex and are not yet completely understood.

Minimal experimental data exists, particularly as it relates to the behavior of molten UO 2 (fuel) in water. The WASH-1400 conclusion that the probability of a UO,-water reaction resulting in a steam explosion causing containment failure was low has been used as a basis for not considering the event until there is more understanding of the processes involved. Recent studies also sapport this conclusion.

6.2 Analytical Tools Section 4.2.2 discussed analytical tools used to evt.luate current design basis accidents. The following is a discussion of the techniques which are available for analyzing events beyond the design basis.

6.2.1 Limitations of Present Computer Codes Computational software is not yet available to accurately model the dynamics of severely degraded accidents. For example, primary system codes do not model culticomponent mixtur es, nor do containment codes consider concentrations of individual gas species where the species concentration is modified by chemical reaction as in the case of hydrogen burning. Most primary system and containment computer codes have previsions for modeling some of these processes, but not in a l

6-5

=

l l

best estimate manner. Some of these limitations are discussed below, since they influenced techniques used in this study to approximate the accident behaviors.

f

\_/ 6.2.1.1 Primary System Analysis The analysis of the key accident sequences selected is considerably more co= plex than the analysis of present design basis accidents since the computational tools for the severely degraded accidents are not yet fully developed. As centioned previously, the current primary system thermal hydraulic codes do not model multicomponent two phase flow and therefore cannot track the evolution and distribution of significant quantities of hydrogen and other noncondensibles in the pri=ary system. For the degraded conditions that result in the key sequences described, the hydrogen production could be large, and in the cast of small break accidents may remain in the primary system for extended periods of time.

Provided the hydrogen gas does become trapped in the primary system, redistribution of the gas may occur such that heat transfer in the steam generators is influenced. Heat transfer is impaired by mechanisms associated with the presence of the noncendensible gas in the condensation boundary layer located at the steam generator tube surface. For events where steam generator function has already been lost or impaired, this will be of little consequence.

Alterations in the primary system void fraction distribution and flow regices are also possible, depending on the quantities of noncondensible gas present. Both characteristics of the flow have the potential to reduce core heat transfer. Modifications to the

() advanced thermal hydraulic code TRAC (6-1) are underway to add a noncondensible gas field for the study of these effects. Appropriate changes to the metal-water reaction model to include the production of hydrogen and consumption of steam by the oxidation process will be necessary. TVA has also been encouraging the Electric Power Research Institute to codify their RETRAN (6-2) code to include these effects through our participation in the RETRAN Utility working group.

The extension is being considered for a futu.re version.

Unfortunately, a best estimate hydrogen release for containment studies was therefore not available. Hydrogen production rates from the MARCH code were used instead with the assumption that the break size was large enough to transfer all hydrogen from the primary system to the containment as it was produced.

6.2.1.2 Containment Codes Accidents that involve a degraded core require the consideration of many factors that are also not presently included in the containment codes. These factors include the addition of large quantities of noncondensibles at very high temperatures to the containment a tmosphere , the addition of energy due to hydrogen or carbon monoxide burning, and the change in masses of atmospheric constituents due to burning (i.e., 2H2+O2 = 2Hp 0). Experiments which determine ,

ice condenser performance parameters have not been perforced for the case of v'ery high temperatures and large quantities of noncendensibles G

6-6 ,

f associated with degraded core scenarios. When burning occurs, containment temperatures exceed the limits of the table functions that existing codes use to evaluate the properties of steam.

Modifications to the codes are also required to allow modeling of features used in various degraded core mitigation concepts. Features such as opening or closing a vent as a function of time, and calculation of pressure drops in gravel bed filtration systems are not presently in the available codes.

1 6.2.2. The MARCH Computer Code The Battelle Columbus Laboratories are currently developing a computer ,

tool to study light water reactor meltdown accident response l characteristics (MARCH) that overcomes limitations of the primary j system and containment codes described above (6-3). Although considerable work has been expended on the understanding of breeder

reactor hydrodynamic disassembly and energetic meltdown accident .

4 characteristics, as exemplified by the computer codes SAS3A (6-4),

and SIMMER (6-5) used in breeder accident modeling; comparatively t

little effort has been expended on the development of light water i reactor (LWR) meltdown codes. Alternatively, light water reactor research has concentrated on prevention and mitigation of accidents l before the comparatively benign meltdown occurs, thereby reducing the risk of core melt to an extremely low level. HowcVer, the experience gained in the breeder studies may be applied to the LWR 3 meltdown accident. In addition, some LWR specific reseach has been done and several codes have been used to study aspects of the meltdown l p problem. NURLOC (6-6,6-7) and CHEMLOC (6-8,6-7) are two such codes l \s,), used to. model the fuel rod heatup, oxidation, and melting process.

i Also the BOIL computer code was written for the reactor safety study (WASH-1400) to determine the time required for fuel failure, melting, and reactor vessel meltthrough. BOIL and a concrete interacton code, INTER, are immediate predecessors to MARCH and are employed as _

subroutines within the code.

l The MARCH computer models contain sufficien,t detail to follow an accident into the phase where considerable hydrogen is produced by

- cladding heatup and reaction. Models permit parametric studies of

core celt enco=pasning slumping of the fuel bundles followed by l pressure vessel meltthrough. Although the thermal hydraulic models do cover most of the accident physics, they are not as detailed as

! the models in the RELAP (6-10) blowdown code and other vendor codes

} used for less severe accidents. No comparisons have been made to

. BEACON (6-9), the current best estimate containment code, or the RELAP primary system code. MARCH functions in many ways as a simplified 4

coupling of these two codes with the necessary extensions to incorporate core melt. Because of the lack of comparisons for MARCH

results and the fact that the code has not been made publicly j available by its authors, the conservatism present in the code is ,

not known. Several of the meltdown models may be nontechanistic and l

t i C) 6-7

-. , ,n -- - , - a.--------a, ,,-. ,,-----,..,,.en -_. . ,-,.-,.-,,nc,, . - . , , , . , , _ - . _,nn., n-7,-,-,.. -,.y-n,,-r. -,

- -. ._ - . - - - . . = _ _ - . - . . - --- - . -

)

therefore may not reflect the actual core behavior. Examination of the code details will be required in future efforts before a final i design is produced based en MARCH hydrogen production estimates.

Areas needing further development or experimentation include: the

,. melting and slumping of the reactor core in the vessel, the degree of metal-water reaction during the core melt process (both invessel

and exvessel), the amount of core internal structure and vessel material that participates in the melt, and the formation, quenching, j and coolability of the debris bed. The MARCH subroutines are i described in Appendix C.

9 6.2 3 Physical Analysis of Present Sequoyah Capability for Accident j Sequences

! The accident sequences discussed in section 6.1 have been applied j to the Sequoyah design to determine the design margins available in I the containment. For events where the capacity of the containment l is exceeded, mitigating devices are examined to determine this effectiveness in ensuring containment integrity. The containment

analyses have been based on studies done by Battelle Columbus Laboratories, using the MARCH computer code, for a Sandia study of y

Sequoyah. The noncendensible gas production is discussed followed by the expected containment response in the following section.

t 6.2 3 1 MARCH Poncondensibles

The MARCH computer code has been applied to several key sequences l used in this report. The computer results have been provided by

' Battelle via Westinghouse and were used in this study because of the

limitations of other codes available to TVA at the c.ime of this j study. MARCH is not presently released for public use so that data was limited to the studies provided by the indirect channel.

Steam production for these cases is shown for the AD, S D,2 and TMLB' accidents in figures (6-2 through 6-4). The hydrogen generation is shown in figures 6-5 through 6-7. Since the cases all lead to complete ,

core celt, both carbon dioxide and carbon ronoxide evolve from concrete interaction with the molten core debris. These gases are

plotted in figures 6-8 through 6-10 , where the time for their evolution is seen not to begin until late in the transient as expected. The noncondensible gases and steam act as forcing functions

for the containment pressure as described below. >

j

^

6.2 3 2 Containment Response Analyses of containment pressure and temperature transients from degraded core accidents were performed using hand calculations. Hand calculations were used because of a lack of computer codes in the j industry with features that allow detailed consideration of

containment phenomena occurring for the events under study. The .

calculations were conducted on a time hietory basis using the l principle of conservation of mass and energy and the ideal gas law.

l The release rates of steam, hydrogen, carbon monoxide, and carbon l O f

! 68

_ _ ~ _ _ _ _ - _ _ _ . _ . _ . . . _ _ . _ _ _ _ _ . _ , _ _ - _ _ , _ , _ _

Bequaga!I E U hragen @tubg

/ns o

(

O

.....................................p..................... .

. . . . . . . . . . . . . . p. . . . ' . . . . . . . . . . . ....................

......... .- d

.................4....................q....................(...................t................... 4....................

.......................................).......................................&....................

..................;...................q...................;...................<...................i....................

~

. .........t,.................. 4....................<...................i........

...................;........ ..........p...................;...................t...................l................... .

t . . . .

................. ........g...................

4..........

...................t................... 4.................... .

4 Q

..................g..........

...........................g.......................................g...................

-.g

..................J............ .......t................... 4....................g...................{................... .

...................g.......... . .

..............................i.................... ....................g...................

m M

............ .......(...................................... 4........................................ .-

H  :

...................r............  :  :  :  :  :

. o rn v4  :.

.....................,................. 3 .......................................

:  :  : oa g to e

........... ......q.......................................g... ...................................q...................

. W U p v .

. . e t3 H

......... ......... ................... t .. ...................................p................... . ,rq N

. . . . H

-4  :  :  : E-4 O .

m O

. . n

_o

......... ................................. ...............:........................................ .,4 M

..................... . . . ..............{................... ....................

O . . . .

....... g ...............,.

.......)... ....................................)...................

,4  :  : -

M p

y  :  :  :  :

.es g

A e N

........ g O gi ..............  ;...................t...................

k M  !.

._.m o

..4 .

-4

-e ye c.,

C Ad M  :.

~

(}x ot v . .

, ,a ,_4 02T 0 0T 01 0'D OT 0d 00 on (m) ssm 6-9 -

c c

agoragen e,uo.g y

mqungaq 5 /

C v

.......................................p...................,........

..........t...................l,................... .-4

..................;,..................g...................(...................(..................4....................

7

..........t................... 4................... 1................... 4....................

...... ............;......... .........p...................;..................

6..........~... .... ............. ....

4....................

..................;......... .........g..................

4...................

. . 4 I

4 -

N. -  :  :  :

--p

....................a....................

................... ................... ....................i ........... .......

..................{...................g..................

4.

..................g..................

. 4.................... -

................... . . . . . . . . . . . . . . . . . . ..g...................

.. ..........................................................g .

-. m

-__o_...o,..oo~~.-o.o~..oo~~.---. .-

v

.................. ..~.........y~~.{oo--.o

.o.

F-.4  :  :

Om O I O @

oo

\.-

y

..................t...............................

................... ................... . g g

-  :  :  :  :  : 5

:  :  :  : g ts Q

g ...... ..

.........i....................i...................:..................i...................

> g w M

,eq H

r-

-4  :  : .

'4 O  !. i. .

O i

.i

. o

_Q

................ ..4 o  :  :  :  : -

....................).........

........................)...................

a r c. :u  :  :  :  ; -

v j ......... ................... ................... ................... ........... ...... .-

\.

m e .4 As  :

u H M i i i

......... t q s, .........j...................j...................!...................!.................. ~. *

.d d M i.

i.

i t

,-*'**'N . . . . .

[a, g . . . . .

, O, 021 0'OT O'9 0'9 OT 02 0'0 ph (mq0 ssvn 6-1o

9equnga17 Eighrngen 9tuby so

-\ -O s ......

................................)...................)....................t....................

..~..a... ....... .e 4.,..................g.................. 4....................g.... .

4.................... .

.................... ................... ,......................................6....................

..................;...................g..................4....................t.-~~~.--.~.-i...--.o.-~~~.

1......................................................................... .......................................

...........t....................g..................t..................~g-~~~.-----4.--~~~~~~.-

.................. ..................)...................;...................i...................&...................

..................;............. ...g...................;...................(...................t....................

I t. .

. .e 1

.g

_o

..................i....................................... .......................................i .................

........................................ ....................t...................................... 3....... ..........

..................{

......................................g...................q...................g...................t..............

................... .......................................t....................t...................

m

.D H ................. ................... ................... ................... ................... ................... .

W ..................!...................!...................'...................i....................,...................  ! E A r . o .,

.i

-4 .I .

o i i /'  :.  :

O b.

..................{.......................................g.......................................q................... . (/) ;.-

. _ v E

.e . .

s

. t.

p

? . .

f..

3 ___4 ........ ......................... .....g..................... . . . . . . . . . . . . . . . . . . . .g................... .

o i i

E-4 C)  :.

p . .

i ~ . . .

l .

. . se

.i o

.e H- ...... .

N. . . .

l r.*G ........ p .........;...................t....................g...................t....................

. ~.

O . . -

l ......... Q .........

. ga .

1..................... ...................I...................

w s y< .

, A-

• w

. ....... . g O y ............:...................t................... .

p d .

g:$

M

.I 4 w-o <;.j  :.  :  :,

g

,d y

.e Ad

. n  :.

. . . ce

/~g .

O

( ) 6 6 6 6 6 , g 021 0'01 0'9 0'9 OT 02 0'0 on (mq0 ssm 4  %

, ! />-11

&cquagq Egbragen 9tubg o

V .

,,-4

. 4....................:...................)...................:...................

...................;...... ...........q................... ...................<...................;...................

r .

~e

. g fg- .

...................g............................

.........j......................................g...................

...................{...................s.............

.  ; ...4.......p...........q................4......................

e-4 .

...........t................................... ............ca .......... .....................t...................

....... .s .-. m

~

ed W

................. ..,................... ....... ..........{.......................................

v M .

}- .

........................................ .................. ....... m ...................................................

n

. Q f . .

. . . . O u%,

\~  !

O \O

.................4.....................!...................N........*.*..........i...................q!...................

4 W

N/ pa

.. Ea .

f'

---4

. . ,.a .

w O ......... ..........g...................t..... .

.............. [q. . , . . . . . . . . . . . . . . . . . .p...................

re';

g O i i jM

. o i EH

.o O  :

b  : -ar- .

. . . . e

?

_o

.......... ....... .H My -

A o . . . . . . . . .. ..................................

O t . . . .

. ...... O i

.g...................)........................................).................

......... M . . . . . . . . . .

7 4 . . . . .

a H  :  :  :  :.

........ 3 ......... ................... ....................,. .......................................

I O f  :  :  : .

......... O * .............................. .................... ...................

b.s O&

W_

O  :

a .

-e se SQ c.i .

g G  :  !  ! .

6 s i ' '

i 4 6

- O g

OT088 0'0038 0'0003 0'096T 0*00CT 0*000 0'O l

1 l

(m0 ssm 1

l 6-12 .

Dequaga!7 7E1ghragen @tubg o

,V

\ O

. ,4

...................{...................g..................4....................q..................4....................._

4....................{...................J....................{.................. 4................... :

4......... .......g..................4....................q..................4....................

...................;...................p.....

...........l....................p...................:...................

...................;...................q............

.....4.............44..1....:..............4....................

p . .

. . ~ . .

. . p . .

:. m ; 8  :

. to . p  :.

to l. N,

. ~

.e

+ N i

O

.............(................... - y(

..................g...................;...................g......... .. ... ..;... m -

.....................t....................;....................,......... ......... ........ ......;................... -

a .

..................J....................g.................. 4............. .....g...

> ...........{...................

.....................g. ~ ..... ...... ..............--to-......-~. -. - -. - - . o .....t-~~~..-~...

m p -4 . .

...................;.......................................{.............

.t................... ......... ..........g..................... . . . . . . . . . . . . . . . . . . . .g.................,_. .

O e

:  :  :  : O e 0 O A

sm- ..................q......................................q..................................... r.,

4....................

+ .  : .

vp

. _ A tv .

c.,,

(14 H

!. Q .........

.........}...................[.....................................i...................

~

pq y

0 $ $.

i i.

- $4 i . . .

O . . .

s .

6 . . . . on t-.- .

o

-4 .........

cx .........

p cc ......... m 0

I, ........ () .........................................................................................

...... .. s ..............................,........................................,...................

e .N

~

e .

u. O u

, e .

. MO o i i i i -

' e. m 1

m . . .

. . . \

D . . . . .

O

^

- 000cc 0Odaa 0 009a 0 Od6T O'00ST 0 Ob9 0'O  !

(mq0 ssvn 6-13 l l

-- I

Ocquagal7 Eigbragen @tuby o-sV o

- , -t

................. 4 ..................q...................;..................

. 4................... 4.................... .

..................;..... .............g...................;..................

4................... 4....................

..........;...................g...................;...................t................... 4....................

. . . g -

:  : 4

................ .;...................g...................;...................)..[.3.................................. .-

:4 8 : .

. . A .

@  !  !  !  !$ ~! -

.......q...................;...................{..n...........g.;................... .

o!

O. .

.e M  !

-o

.p

. .................... ....................t....................

...... ............ ................... ........................................ ...................i....................

.................. 4................... .................. 4...................g...................;................... .

. < .s

_. y)

W . . . .

-4 .

3

....................g..................... . . . . . . . . . . . . . . . . . . . .g........................................p................... -

O&

Q

. . . E

. O O \D s . . .

..................g..................................... ...................g................... U3 r-4

--4

.  :. 4.................... .

U~

. - 6 3

MH o

• -4 N

.........;.................. 3.......................................t................... . ,=q

}- l l  :  : . *-4 8

s -4  :.  :

- 6-4

O i

! i

. . . e

.i o

a

.4, gt3 i

g .........

A .................................................... . ....... . . . . . . .............

, o .o. . . . . -

3 LM .........

Q h ......... ...................................... 4....................................... .

4 4

......... 8 A. v

.........!...................Ip...................i...................i g

y t...................

5 br

• b .........

p",

t5 . . .

1 r1  :.  :  : . .

O L#.. O . . . .

, W O  :  :  :

' ~

' 2 MO  !  :  !  ; -

1 aO Cfe

:  : - 1 I

.p, Ae  !.  !.  !  !. .

. . . . 1

. l cg . . .

. M ,

l s

o g

l- 0'0060 0'0930 0'0093 0'006T O'00ST O'099 0*0 9

(mq0 ssyn l 4

6-14

e -

e ecquaga17 ,mgbragen stubg u  :  : - v-4 O

....................................c.,- . .. . . . . . . . . . . . . . . . . . . ..................... .

m ...................

x . .

x . ................. 4.....................: . . . . . . . . . . , . . . . a..

. .m. . .

o:  : .

o:

................ ............................................ . . . . . . .* . . > . . i. . . . -(). ....................;....................

,4 . . .

o: .

.  :  : A: .

t .

s

o. .

ro s .

.................o. 4....................t....................(.............s3

.... ....... ...........{...................

s . .

4 .

s g .

. s .

s .

. . s .

. . . s .

s

.....;...................g.............................s g.........t...................

v ...................>........... .

s .

. s .

. . . . s s.

g.

fa==g F-"4  !. l

.s

.g

. . . . . s H  :  :.  :.  :  : s m . . .

s O (d

\ H  :  :  : .

s

• ce

% i i  !  : i \ US

................\.........

g

._gO v(f2ca

.-...........\ O 4 ...................s.............................................................................. g ......

y H O ......... .........i....................i...................i....................:;............'.s,...

4 "

g C)  :

s.

E-*

H ........ ......... ...................

. . . . . . . . . . . . . . . . . . . . , ....................................9 .

--.-4 .

t .

~

g O

M. .

........ rs .........g...................;...................g...................;.................\

p Q . . . . 1. '

M

........ w o  :  :  :

.........;................... .................. 4...................:.................

l MA A  :.  :.

o O

l . .

p3 p M en

......... ..........;...................p........................................>...................

4 4w  :  :  :  : .

a  :

aoM E-+  :  :  : -

n  :.  :  :.  :

Q  %

1 MrO owr4 .

1

......... 4 W

0 y e.,

...................{.................................

4....................

_, #4 .

o  :.  :  :.  : -

r. e4.

s~ .

AXo  :

. i O

(j i

= i.

i* i. i*

a

- 1""'

O, 0 , ,

0 0 1 OTT O'OT O'9 0'9 07 02 0'0 '

I

,m (mm) ssu 6-15 l 1

t

.. . . ~ . . . . . . . . . . . . . . - .~ .. . . _ ..... ... .....-. .- ...

P u o a ')

99 S2D 10NC01 1 T DENSIBLE GAS O

v( --

.f.

s

~

DATTELLE RESULTS  !!  !  !  !  !  !  !!! U o

MARCH COIIPUTES CODE, i i i .i i d ARBC.D Di. Q

....OX1.

.... DE!

CAR 130N EON 0XIDE - DIOXIDE .i: :i .

i.

i i

i i.

i i

i p.!

+--

. . 4 o .

ty m .

m &

P-4 . . .

a . . .

H  %'  :  :  :  :  :  :  : :  :  :  :  :  :  :  : : 6v m _ ........................ . ............. ...........................................

. H (Q . .

. e4 LI

.j .

4 A . .

Q .

. . . . 4-*

+-f V_ .

......../

. p .

, p' :CARIO. N M. ON.bXI.DE

.r

.s . . .

.i . . . .

1 o . . .

CQ

........................a..............

. . ....e....

.*==*. ====* -===...==.m

=. . . ..... = ......../

. 1 .. =.=.=*= .=.e.

...e=

. ==

• ==

/. . . . .

i. . . . .

. . . . . i . . .

,/ . . . . .

,,e . . .

o.

o_ . . .

8 10 10' 10" TIME (seconds)

FIGUlm 6-9

I s

9equngal7 iBgbrugen 9tuby c

O

.  : - H

...............a..

. . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6...................

m O

H

N

~

...............A.:..- .....................,..................... . . . . . . . . . . . . . . . . . . ..............O,.....:....................

. . ~

. Q .

o. . . 4

...............m.,....................{......................................

4.......).....g........................

a:

O

. I o  :.

. I .

I

, ..................,................... ................... .......g.... g ......................... -

6

. . . . s .

's s  !.

=

4 s

. . . . i s .

.................A 4................... .

4....................t......

4....................

.  %.s -

g

. g g

w. .

. . s g

. s

...................)...................s...................)..................

6........ .

. . ........)..........%........ s

. s

. s .

. . . g n H .

.e

\

t -

.  :  :.  :.  :. ^

m F-  :  :. -

'd co

'-  :  :  :  : ~:

oa C A F--

o8E

. g.vs W$

w ...................  !...................i..................i...................i...................!...................

:  : o O .................

4...................

4....................................... ...................

H g

O i

.........i.......................................i....................!................... ,

5 H j  !  !  !

h O ---- ---

---

i--------- ---------i.---- --- --------- i- ------------- -- i.--- -----

Mn

......... .........t...................;...................g...................;................... _

g .

M

~

~

O  :  :  :  :  :

........ e--e .........;...................:...................:...................:................... .

,A A . ,

A .

L.--4 O

o [

g; o a ,  : ,

A .........y...................p..................,...................>................... .

.-4 . . . .

4 H  :  :  :  :

m bo  :  :  :  :

M mZ  :  :  :  :

MMO i i i i -

........ p.o oM  :

.........g..................

4........................................{...................

W~ Z .

m o  :  :  :  : -

p .

c-4 .

Ado .

s  :.  :.  :.  :  : e J

4 O

s s s s s 4

021 0'OT 01 0'9 OT 02 0'0

, m (mm) ssm 6-17

,, -,e r - - . - , - , , , - . , _ - - - - , + - - - -

dioxide for various accidents provided in the MARCH code were reviewed, and the S D case was chosen for detailed analysis. The 2

S_P case is a small break LOCA with loss of high pressure injection.

This case was believed to provide a general picture of the conditions Os that would exist as a result of any accident involving a degraded core. Amounts of hydrogen were allowed to accumulate corresponding to a 20 percent Zr-H2O reaction and a 60 percent Zr-H,0 reaction.

' When the hydrogen reaches these limits, it was assumed to burn in five seconds, and the resulting containment temperatures and pressures were calculated. The containment response was also calculated for the S2D case assuminC that the hydrogen was burned as it was produced. This case was analyzed from the initiation of hydrogen production until core melt.

The analyses were performed using the following assumptions:

i r 1. Only the upper compartment was modeled. This assumption was made primarily to simplify the calculations; however, it is believed to be justified. First, for the four major accident sequences studied, large amounts of ice (over one million pounds) remain in the ice condenser at the time of core melt and will be capable of removing significant amounts of energy released by the burning of hydrogen in the lower compartment. The inclusion of the lower compartment and ice condenser in the model would result in lower overall pressures inside containment. However, it is not believed that this represents an undue amount of conservatism.

2. The containment spray system was included (which is consistent with the S D sequence). One train of spray was modeled Qith a

() 2 flow rate of 4750 gpm. For the rapid burn cases, the entire spray flow was assumed to vaporize due to atmospheric temperatures that were much ' higher than the saturation temperature of steam associated with the contain=ent pressure at each point in time.

For the burn as generated case, the mass of spray water vaporized .

was determined based on saturation pressure in the containment.

The partial pressure of steam produced by vaporizing the spray water was considered. -

3 Passive heat sinks were neglected. For periods when rapid burning of large quantities of hydrogen occurs the passive heat sinks have a minimal impact on the peak containment pressure or tempera ture. The heat sinks will affect the duration of a transient. The impact of the heat sinks on burning small quantities of hydrogen can be ignored due to the effectiveness of the spray system in removing the energy released during these periods.

4. Hydrogen was uniformly distributed throughout the entire containment on a volume fraction basis. A hydrogen collection-system is provided at Sequoyah which prevents the accumulation of hydrogen in the steam generator and pressurizer enclosures, the reactor cavity, and the containment dome. This system is 6-18 e

ccnnected to the air return fans which circulate ccpious quantities of air throughout the entire containment snd provide for excellent mixing.

p

\_) 5. For cases where hydrogen was allowed to accumulate prior to ignition, a flate speed of approximately 10 ft/see was used during the burn period.

6. Initial conditions (prior to the generation of any hydrogen) were taken from LOTIC runs performed by Offshore Power Systems. This provided initial temperatures, pressures, and distributions of steam and air throughout the containment.

m

%.}

h j/ \

L]

6-19 ,

6.2.3.2 1 Twenty Percent Zr-H.,0 Reaction - Rapid Burn A metal water reaction oxidizing 20 percent of the zirconium cladding d produces 420 pounds of hydrogen. In the S D sequence, this hydrogen 2

is released to the containment as produced. All 420. pounds will be in containment at 4500 seconds after the start of the accident.

A rapid burn of accumulated hydrogen results in a containment pressure of 54.3 psia and an atmospheric temperature of 960 F.

The containment pressure and temperature profiles for this case are provided in figures 6-11 and 6-12, 6.2 3.2.2 Sixty Percent Zr-H 0 Reacticn - Rapid Burn A rapid burn of 1325 pounds of hydrogen produces a peak containment pressure of 126 psia with a concurrent te=perature of 2713 F. This quantity of hydrogen, corresponding to approximately 62 percent metal-water reaction, is released at 6970 seconds after the start of the accident in the Sp D sequence. Figures 6-13 and 6-14 provide the pressure and temperature profiles for this case.

6.2 3.2 3 Burn Hydrogen as Generated In this case the containment pressure stayed below the design pressure of 26.4 psia from the start of the accident until 9651 seconds. During this period of time hydrogen is produced as the core is uncovered. At about 7000 seconds, the production of hydrogen drops sharply as the water in the reactor vessel has all boiled off.

htile significant quantities of hydrogen are generated, they are J released steadily over time and the spray system is capable of removing the heat of combustion without exceeding the containment design pressure. For tne period of time between 7000 and 9651 I

seconds, the core continues to heat up due to the decay .of fission products and at the end of this period the core melts through the bottom of the reactor vessel and falls into the reactor cavity.

The water in the reactor cavity oxidizes the remaining cladding.

This occurs with extreme rapidity and the hydrogen associated with i

the remaining 30 percent of the zirecnium cladding is generated in

) less than 10 seconds. The burning of this release produces a containment pressure of 80 psia (see figures 6-15 and 6-16).

A detailed compdter analysis of each transient would be preferable to the hand calculations used; however, these results represent a reasonable and conservative approach to the problem based on the information and methods presently available, j

6 3 Plant Capabilities to Withstand Extended Design Basis Events The Sequoyah containment systems have built-in margins which would allow the plant to withstand events more challenging than those O

6-20

,wv. ,v - . - ,-- , , -x..a y,. -,7 -- ,,, ,, v n- . , , -e-n- y .,

Brquagal7 i9ghragen 9tung O.

C .

O C

-4

e

. O.

H Oc-e r ,. .

. O 4

F- ...........:...........:........... ,...........:........... '............:...........!.....

......... H c

H  :.  :.  :.  :.  :  :

.y O .

Z O

m A- p .C

w. .......... ...................... . ........... ........... ........... ............e..... O - ..-....

-C

.  :.  :  :  ;  ;  : Z A O .v

\s Z O.

Z -_ .-4O o .........

O^

.  :  :.  :  :.  :  :. A

. U} _vM

.cf JM ko Op H O Am

• l  :  :  :  :  : O o e

j. g t- ,-

: U2 CN! U2

- Ocf Oe

_ow. U2 i

3 v -

. . c 2 .

H u N-3 . .

. O ,m eq

. . . . . o

....... ............. .............. . . . . . . . . . . . ................................ ................... . oC E-4 p- l l l  :  :  :  :  :

w-1 . . . . .

~

1 . . .

! 44  :  :  :  :  :

s

- O.

. w -

p_ _O

............ Q

. . . . . _w '

2 H  :

3 . . . . .

. O.

a . . . . .

. cJ

. ...... .. . . . . . . . . . . . .. . . . . . . . . . .. . . . . . .. . . . . . .. . . . . . . . . . . .............. ...................... ........... _o

. C O w.

3 . . . . .

= . . . .

O.

. . O O

3,

. Q

_v

- o.

2

.c

. .g C.2

-w

(v\

1 0'99 0'00 0'99 0'09 0'91 0'03 0'90 0'0S 0'93 0'03 0 9T

('eIsd) HHDSSSHd j 6-21 .

9equagalT IB U brugen 9tuby o

a . . . . .

a~

n v

. O.

ca

.H

• o

+  :  :.  :.  : .

v H  :.  : .

--4  :  :  :  :  :  : .O O  :  :  : .  :  :

4

..... ......... H c

y Z

v.

. . g4 o

.  !  :  :  :  :  : Ar

........... ....................... ........... ..... O

............: *4 AO . . . . . . . . - ---C v

N s Z

y. . . . . . .

Is l

H .

i _

- O.

O

. . . . . . . . . . . ...................................... ps -4 q d o Q

L u}

C.

,- ein A

w 7 . . . . .

O cu 84 .

o m oO

)

[:Q i i  !  !  !  ! i mam '

n)

) ..........

........................ ......... s O.J1w?

O c U I:Jm L( ~

5 Ne e4 .

w s

'-  :.  :.  :.  :  : 4 E=.

O. ,H

. oq

.. .................. . ......................... ..................................... ......................... .o O

,4

-4  :  :.  :.  :  :.  :  :  :.  :

- O.

H

........................ _oC

,, .~

, -4 . . .

_ w .

W .

-.-4  :.  :  :

H .

. O.

44

. . . . cu

............. ....................... .o

. C w .

o

--4 .

O O.

o

............................................................ .O t

C.4-l .

. O -

. . . . . c3 8

g g<\ .

. . . . ,.e e h ,

0'090T O'098 0*099 0'094 0'090 0'099 0'099 0099 0'093 0'09T O'09 (4) aanasaanal 6-22

9equaga17 Mghrngen 9tuhg

<o

\._)

.e m

. o c

. . o.

c. .

. c.

_. c m.

O x

o g o

. m v ........... ........... ........... O .........

-.c

:...........:  :  :  :..... 4* A o o

. \4 Z m ,

Fy . . .

. ., - o.

. c,a.

Z o r ......... 1........

...........l...........

g g ......... _. C ca m

. y) o M

. . . w

. > . &.?,

o g Om u.< . -

(f) 34 m o m o. o a,

- O Ao y

_. t%

o M e v rc O

y H

. . . . r=.

C. p.q H

cb N

cs c

1 . . . . . .

V4

~ . . . . .

M .

O.

4 b-

................... ........................................................................... . . . . . . . . . . . . c

. o H  :  :  :

4 . . . . . . . .

. . . . O.

e

. . eq g ...... ............. ...

...... ......................... . . . . . . . . . . .......... . 6 o

O c

. . . . O.

. o

1. .......... ...................... ............................ . ....

........... .......... . ...... . _. 6 o

. . e

. o.

. . . e

. e

:.  :  :.  :  :  :  :  :. cm l O, m )

,o l O'OTT 0 93T 0 9TT O'10T 020 0 00 0 09 0'99 0*FP 020 0'03 l

(EIsd) 3HOSSHHd l

l 6-23 -

Ocqungaly Egbragen Stuby o.

m O

. . ca s

V e

. o.

. . o

. c, c

-4  :.  :.  :  :  :  :  : o O

....................... ........ ... ............ ........... .............................. ......... -. O

. . a

.c to Z.

o

: A > ej

.  :.  :. *4 A w

O cm c

- cio r ............ . ..... .. ............. . . . . . . . . . . ............................................

rd. o= Z

-.ame,o w

ea ,cn M<

4 .

o g .

g

:  :.  :  :  :  :. . C\2 % o.

. . ca o la c so

_a u m s

.  :  :  :.  :.  :.  : _o F--A e

. . . . . . . H w

,e;

. .m

. . . . c

. . N o

. . c t .

F- .

-4 .- .

. ~. . .

,9 .

o.

w

........................ atq

~- .

o

--A . .

--a

. O

r. 4 ............... ...... ........................ .................... .

• --4 L  :  :  :.  :  :. .  :.  : - C3 o

--_4 . . .

O o.

. . o o.

. e

. . . . . a o

e O'0902 0'0dO2 0'0$C3 0'0Y02 0*0d4T 0*0dM O'Od21 0'038 0'Oh9 0'092 0'09 i 4 SHfilVHHdHS1 6-24 I

(...

e- -

e

7,uw . , e ,. .w' ' eF,

* .-

s 4

, - , _ _ _ _ -s-, - - - - -

Dequagah Egbragen 9tubg O, y .oO 2

,u~ - a

_g g ,

w ,

o.

-g o  : .

cn

.g c4 - -

L v I .

M *C i

W L

dT Cn ..

• c cn = -

g g

bg . .

. . . . .  : . . .. . .. t. o o ggg @n p $o"g .

\

g@

d S. E g . . g;m t Y -c -

o

,_ g E-dk Z

--' u - L

=,Mot4 p I.

y in B a L

Z q

4 .

H _

Z ... . .. .. . .. . .

I o

O ,

Lg o .

I

\

e

-g l I l . L 1

o g

C) o o'ii o o'ot o de o da o d4 0 d9 o-do 0-dt a-$c oda 0 01 (utsd) 3HflSS3Hd 1

n m

6-25

,_ , , . _ , . ..-+-h-i

- w 'g---,- + - - - + -v--r-i ge- --

s--

w m*--,q y 7 -+7 me m . m---ywme yrv--r: ,w -

Dequngaly Eigbrugen 9tuby v 9b

_Hw Se o

................................................................................................ ......... . d

. Q g . . .

H  :

--4  :.  !  :.  :.  :.  :.  :.  :

O .

e

. . . w.

o.

m w- . .  : .

(* .

o

>-r>4

. . . . . . . . . . . .- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ... o c-

. o. 4

.... .... ..... ........ ............ ........ m om j

4 .

. O m l N

o J ........

e@

l 981

_o nm o v

- -4 o  :  :  :  :  :  :  :

6-  :  :.  :  :.

- 4 dp

.  :  : l r-ri

:  :.  :.  :.  :.  : C:1 7 g. m A o . . . a l dO aN M  !

X$

H EH mM  :  :  :

.........................................i............:............:...........:....... o mE-4 A  :  :  :.  :.  :  :.  :

c

. . ~. . . .

h. . . . . . .

F.-

,- o

o. . . . . .

g . . . . . . .

r ,. tn e A  :.  :.  :.  :  :  :  :

. o.

H  :  :  :.  :. *

l.  :.  :

..... H.

9 I

M .

l l

l

. . . . . . . l

. . . . \

>-.-4 .

\

. . . . . . 1 44 O.

. . ........ . . . . . . . . . ............. O 1

C4 g .

D I

\

. . . 1 O

4

. . s

o. ,

.......... .......... ........... .... ...... ........... ........... ........... .......... ........... . . . . . . . . . . . . o 4

.  :  :.  :  :.  :.  :.  :  : c2 ,

c3 vj .

n O'0903 0 099T 0~003T O*09TO O'093T O'090T O'099 0'099 0'00T- 0'003 0~09 (a.) aana.vaaanaa.

6-26 '

censidered in the original design. Some of the more useful for dealing with hydrogen are discussed below:

/ 6.3.1 containnent Structure The centainment design pressure is 26.4 psia. However, there

~

are very large margins in this value. The containment steel is'not expected to yield until a pressure greater than 37.4 psia is reached, and the centainment ultimate strength corresponds to a pressure of 57.9 psia. The containment can withstand pressure more than three times design without a significant probability of

-failure.

6.3.2 Ice condenser The actual ice loading at Sequ 6

y h was 2.97 x 10 6 pounds compared with a value of 2.45 x 10 pounds used in ghe design analyses. This represents an added heat sink of 130 X 10 Btu's available to accommodate heat from hydrogen burn effects.

6.3 3 containment Spray Systems The Sequoyah containment is equipped with two spray systems composed of four independent subsystems. Each subsystem has its own piping, pumps, and headers. The spray system is capable of injection 13,500 gpm. The analysis of the current design basis events used a spray flow rate of 6750 gpm. The evaluations of the events studied used a spray rate of 4750 gpm.

O DA03:NSAHY1.PB l

. OL 6-27 ,

1 i

d

7.0 '41tigation Concepts

(-)x Eased on our knowledge of hydrogen and other core damage-related phenomena and concepts proposed by TVA and others, the following ccncepts were identified as those having the most promise and-being northy of study for mitigating the effects of hydrogen and containment overpressurization:

1. Inerted. containment

2. Halen injection 3 Ignition sources

4. Augmented cooling

5. Filtered vented centainment

6. Additional containment

7. -Coupled containment

8. Oxygen removal by chemical reactant 9 Passive enclosure

10. Structural reinforcement

11. Combinations of the above Ccncepts 1, 2, 3, and 8 provide some protection from the effects of hydrogen combustion. Concepts 4, 5, 6, 7, 9, and 10 provide seme -protection of the containment due to the effects of overpressure.

These concepts were given a screening evaluation in terms of adaptability to existing plants and plants under construction, safety, and feasibility. This preliminary screening resulted in the

() elimination of oxygen removal by chemical means, passive enclosure and structural reinforcement concepts for being infeasible, impractical, or-ineffective for Sequoyah. The preliminary study also concluded that for a number of accident -sequences, many of the concepts under consideration may not be sufficient to cope with all of the effects of core degradation and would have to be considered in I

combination with other mitigation techniques. Such combinations would consist of one of the concepts which provid_e protection from hydrogen combustion and one of the concepts which provide protection from containment overpressure.

7.1 Inerted Centainment To prevent hydrogen combustion, prior to reactor startup and operation, containment atmosphere would be diluted with nitrogen until the oxygen concentration is reduced to less than 4 percent by volume.

The nitrogen could be introduced into containment through existing ventilation and purging ducts. Nitrogen would either be supplied l by-an outside vendor or a moderate sized manufacturing facility ensite. All areas in containment must be provided with life support equipment and gas monitors. Modifications such as replacing air with nitrogen in the compressed air system will be required to eliminate oxygen sources in the containment. The systea should be designed to rapidly reduce the containment oxygen concentration to minimize plant downtime.

( .

7-1 .

n.

'7 2' Halen In.iection i

Halen is a fire suppresstnt that in sufficient quantities may prevent burning and detonation of hydrogen in the centainment. This concept

! O consists of : pray headers at various locations in containment fed ty pressurized tanks located outside containment. Halon would be i injected into the containment over a short time interval upcn j incicaticn of a failure which may lead to significant hydrogen 2

generation.or upon detection of a significant concentration of hydrogen in centainnent. The effect would be to create an inerted atmosphere postaccident. In contrast to oxygen depletion, such as by dilution with nitrogen, the halen apparently absorbs the burn energy rapidly by chemical processes, thus preventing the flace front from propagating.

7.3 Icnitien sources This concept 5 aimed at controlling hydrogen buildup within containment. ' hydrogen is released to the containment slowly and 4 burned at rele vely Icw concentrations, it can be controlled without severl. 'verpressurizing the containment (which cculd occur

{ if. it is allowec co build up in large amounts and is ignited at randem). The concept consists of ignition sources located throughout i containtent in areas where hydrogen would potentially accumulate and additional hydrogen monitors to provide indication of the buildup.

Upon indication, the igniters would be activated. Continued i conitoring would provide indicaticn of the need to deactivate the igniters-(i.e., if hydrogen buildup is too rapid). Since the ignition system will be designed to burn hydrogen in low concentrations, this will require an open flame or its equivalent.

O 7.4 Augtented cooling Additional containment cooling would help remove heat from the containment, thus reducing the containment pressure. There are several potential ways to increase the cooling capacity of the

containment such as additional containment sprays, additional air coolers, and other heat sinks. Because of containment space limitations, passive heat sinks could be considered only in conjunction with other concepts such as an additional containment volute. Uhile additional active cooling may not be sufficient for

, mitigation of large energy releases of short duration, it may be satisfactory for energy removal under slow controlled hydrogen burning.

7.5 Filtered Vented containment This concept provides containment pressure relief through a vent by flow of noncendensibles, steam, and energy. The flow is vented through a cechanism such as a suppression pool or gravel bed to remove heat and steam. The flow proceeds through a sand filter for additional removal of moisture and radioactive materials and through 4

additional filters such as particulate and/or charcoal for high efficiency renoval of radionuclides. Remaining noncondensible gases i and radicantive noble gases are discharged to the atmosphere via a t e stack-t'o' provide elevated release and dispersion. The design must k

I 7-2 ,

prcvide for capture and postaccident cleanup of radioactive liquids and filtcrs and for removal of fission product decay heat. Hydrogen 73 must be prevented from uncontrolled burning within the vent system.

(\~T This ccncept allows controlled, reduced releases to the environment to prevcnt an uncontrolled release that could occur following containment failure.

7.6 Additional containment This ccncept provides containment overpressure protection with additional volume and heat sinks to hold some of the noncondensible gases, steam, and energy generated from an event involving core damage. The additional containment volume would be coupled to the existing containments via a large vent whi'h could be opened during an accident if conditions warranted. One additional containment would serve all units at a plant sgte., 3A preliminarg estimate of the size required is between 1.5 X 10 ft and 3.5 X 10 free volume.

7.7 coupled centainment The coupled containment concept is identical tc the proceding concept except that it makes use of the second nuclear unit's containment to provide the additional volume for overpressure protection (all TVA nuclear plants have tuo ce more units per site). For pressure suppression centainments, the second unit may provide a large amount of additional cooling.

7.8 0xygen Removal by Reaction While in theory this concept has merit, our preliminary ' investigation found no chemical which would permanently remove the oxygen or remove it quickly and reliably. It appears that extensive research is needed to advance this concept to a stage permitting a conceptual design.

Thus, this concept was not considered further in this study.

7.9 Passive Enclosure .

The passive enclosure concept has previously been considered as an alternate containment concept in which the entire reactor coolant system is encased in a guard pipe / chamber arrangement. For use as a mitigation technique of degraded core events, the device would be designed to withstand extremely high pressures. It would provide a nini-containment around the entire pritary system and be designed to retard melt through. Cooling would be incorporated. This concept; however, is impractical for current TVA plants and was eliminated frca further consideration.

7.10 Passive Structural This ccncept may be used in limited cases. The purpose is to enhance the ccntainment load-bearing capability by structural reinforcement.

A review of TVA plants indicated that significant strengthening cannot

<3 be achieved within practical means. No further consideration of

(,) reinforcement was made in this study.

7-3 ,

i 7.11 Other Concepts Other concepts nay exist in addition to those listed above. It is 3

believed that others worthy of consideration would employ one or more of the principles involved in the concepts identified abo e; therefcre, those described are representative of the proposed citigative concepts.

l 1

9 ,

i l

i l

i 1

7-4 ,

8.0 Risk Assessment of Mitigation Concepts

,s A preliminary risk assessment of selected events was performed to

!w ,) provide an approximate quantification of the reduction in risk associated with the addition of various mitigation concepts for accidents resulting in reactor core degradation. This assessment does not quantify the change in total risk to the public, only the chance in the probability of containment failure for the selected events. Also, the assessment did not take into account that the additica of mitigation concepts could potentially increase the risx to the public for design basis events (e.g., if the operator inadvertently opened valves in the line to the filter bed used in the filtered vented containment concept, the dose to the public would be increased).

Based on analyses of selected events beyond the design basis LOCA that threaten containment integrity, event trees were prepared using methods similar to those used in WASH-1400. The event trees were prepared for the present plant design and then redone to account for the benefits of each concept or combination of concepts. These were used to estimate how frequently and by what mode the containment would fail (lcakage, hydrogen burn, steam explosion, noncondensible gas productien). Table 8-1 summarizes the reduction in containment failure due to various mitigative strategies for a particular small LOCA event (S D). The numbers in this table are not representative of large LOCAS s and transients and may not be representative of other small LOCA's. For these other events, the numbers may change in absolute value and/or relative to one another. This is because the

() ccnsequences of different accidents vary in both rate and magnitude (as described in section 6.2 3.1 for the sequences used in this study) such that the different citigative strategies have varying degrees of benefit. In addition, there are other important accident sequences that may lead to containment failure (e.g., failure to close all isolation valves or interfacing check valve failure) that de not involve either hydrogen burning or overpressure. None of the mitigations studied here could reduce the tisk from those sequences. Under no circumstances should the containment failure probability reductions presented in Table 8-1 be generalized to imply any similar overall risk reduction for Sequoyah.

Concepts which employ containment inerting involve the additional risk of having an accident because of reduced access to centainment for inspection, maintenance, and repair. While this risk is difficult to quantify, a small increase in the probability of failure of a safety system may outweigh the potential benefits from including hydrogen mitigation in the plant design. Effects of this type are not reflected in the results in Table 8-1.

(

8-1

TABLE 8-1 m

lv)

REDUCTION IN PROBABILITY OF CO?iTAIN"ENT FAILURE BY VARIOUS MITIGATION STRATEGIES FOR A SMALL LOCA

/

EVENT RESULTING IN CORE DAMAGE (S2 D) ,-

,,._ s.K c .

ReductionEffectivenessforthePEobabilith_[of Containten' Failure 2 v.

Partial Core Damage T 2 b '

,? .

, Core Melt \..tg v,

--A sg.

. s-Hydrogen Burn'.3, *

' Hydrogen Burn Generated :%2,All Generated All Miti;ation Strategy ' 4 Failure - Failures 1 Failure Failures

~.. ..

a Ignition Sources 6 Halen Injection ' . .C' ' -_ _ _ . % 70 .., 70 .N.?. ~ NN 7 7

~70 g 70 ' fs 75 11 Nitrogen Inerting C 100 10'0 9 100 20 Filtered Vented ?2 /

Containment  %, [ 15 4'[.15 (m) Additional Large G. f -x. '.i 9 8 Containcent ,. .

E' [_..r: 34 'l' cy, 34 16 15 Coupled Containnent Filtered Vetted g 30 30 5 5 Containcent and W Q

Inerting

\\

Large Contain:ent and g \l00, 100 100 87 Inerting.

va

,[

, 100 100

~

100 n 96

,G,.

g _ g.

1.

Theseevent.

other reductions are for the S 2D case only and cannot be applied to any 2.

The numbers represent the percentage or times, given that the event occurs, that the particular mitigation strategy would prevent containtent failure.

3 Partial core damage represents a level of damage where it is possible to terminate the event if cooling water can be restored to the core. For the Sp D case, this da= age level corresponds to 60-65 parcent clad oxidation; rdr other events, this level may be much lower. Once the partial core damage upper bound is reached, the event proceeds to core melt with extreme rapidity.

4.

All failures occurring as a result of the S D 2 event (e.g., overpressure due to noncendensible gas production).

g 1

L 1

8-2 3

9.0 Mitirative Ccnceptual Designs Each of the seven mitigations described in section 7.0 was icplacented r,

in a conceptual design by various architect-engineering and TVA design

( ,

organizations based on the preliminary design criteria summarized

'- in Appendix D. These criteria include performance criteria, envirentental conditions, operating modes, seismic criteria, passive vs. active components, and safety-related versus nonsafety-related qualification. It should be ecphasized that none of the designs were implemented as safety-grade for this study except where required to prevent degradation of current safety systems (e.g. , contianment isolation valves). In the absence of any clear regulatory directives and due to the brief duration of this study, this decision was made purposefully to simplify the mitigation designs and not escalate their cost. More rigorous critiera would have a major impact on cost.

The designs based on these criteria included enough detail to evaluate feasibility, safety, scheduling, and cost. The results of these evaluations are summarized in this section.

9.1 Filtered Vented Containment 9.1.1 General Description The filtered vented containment (FVC) system will allow high tecperature and pressure gases to be vented from either containment after an accident through iodine and heat re= oval subsystems and a particulate removal subsystem before being released to the a tmosphere, es The system, as shown in figures 9-1 and 9-2, will penetrate each l centainment structure with two 36-inch lines. Each line has three principal isolation valves: one inside the primary containment, one in the annulus, and one outside the concrete secondary containment.

Each line also has a rupture disk which opens at 35 psia. The two lines from each contianment tie into a 54-inch header in an underground valve pit. Each line will have an isolation valve to isolate the containments from each other an,d to prevent contamination of the lines not in use.

A cc: mon 54-inch line will be routed to a second underground valve pit where it will split into two 54-inch discharge headers, each with its own isolation valve. One of these 54-inch headers, the pool vent gas discharge header, is employed when the affected containment presure is at or above 5 psig, and releases the gas under a pool of chemically treated water through a distribution network.

The other 54-inch header, or the pool bypass vent gas header, discharges the gas into the air space over the pool and is used when the affected centainment pressure is approximately 5 psig or below.

The pool of chemically treated water, located in the lower porticq of the filtration building, will serve as the primary steam condensing, heat removal, and iodine removal mechanism during use of the system. It is also the source of water for the iodine removal

~

g- spray subsystem, which is located at the top of the air space above

! the pool. When the vented filtered cotnainment system is required to operate, a vertical turbine pump will dit;harge the chemically 9-1 -

i I

.2

- . . _ _2. . . . . . . . . . , . , .

,l 'a  !{l-l TF.

> ~:6

/~N, ' Ii; y 5 !.

s t,v) .- ?, P;j

..'..  ; ' l', 2, a

..:.4

~' 114:; lr/ ' '. ;~

4

. ci;....:

gL

-'{

:. $ '

!a S. h

=...

e o rj '.)

-- sas9, .

i e Q _

' ! j 8,_ i 1

1. - 12:;[

g.!. Ji

' . ((I ' j s

Df '

,J.Q l ---

1

.g.g

- ~-

.! i  ;

7.;p hi7 l q  ; .g g

3

' 1:-

. Q <-4 ,t h.  :

4 .

t- I

+

17.im

.. 5 :],,, l r-.qm t: m-- .-

l tt ,

1 w

!!! /s j i t id-t 1 :y--:-

.X.uwgr: h,u

, 0: O.7. m!=,-- w!.'i I' - rni f .I u.1 r.

~:

)

a ou: Laa: g

-= ij -

! , c:i -

~

gD.C:

C i i O'

'9 (2l}?  !

pmQ 1

i

- m s  !

3 ,

6M ' sh, nl el

.. - D .

p}

h.

i  :

ms q)  !

s C (,

a gm l

g-

. -4 -1 s s [ '

f -

. I: si * -v -41 l if {F ll

,1

• I g 'Q s i<

ff

.p

,v t 1

-1 *

-l a-

2 y c

. w.,

r.

i,qi F

3 s i,s t<~S ,r., f. 4

(.

.i i i.

s> a: j g g.

w g,  ;; g a .

o .

J_

d ,

1 if P it l

l l l l  !!! .

p{

y j:  % o n , ,  :

3 u s p ., 0 u

e a

u !. . I i it-j%l p l

6--

ri , - - , .

.O_O ....._. if O0 I!

la .O >- C.Q .- W ..)._D_.1; j t..(... ys g .O-p;3 g

,. " ohia

.,9 _

,1_ ,

I

-ep . e , ;'g O s. a.-.

=

j m' e-g e, , -

Le---!73--

,-- ..g m1

(-.  ! ,,,,,w!q (.... m, .

g

_ .t 1 .. m .. .

I

,, i ;. '

5 8 I! ' I "I

,O:- +. ovv* (d'i .

I

/

,E I .@'ic--: y Q:#r"e 2 . .

.O m.. ..,i .O ..__> i ,

t, , i,^ .O...m ,l .O. .w......i sa j l 4 j j 3 '% "f ]1  ;

' !. )

i . .

nca.z 91 fs 9-2 ,

a , .

I a, n-i i t,

'

,: .n n  : <.,.:

/ \

( ; * 'l:b.'

\/ -

l

<.,! Q1 i .il d

- BM 5 . ;b. l' llj l. y:' [

q! . -Q.

' _J._e 7 .. I t= l('),i e 9;n1

.i ,O..,, -w.')-------'i g.._ . I

- , ef t

gg E i.rit , , .

D--':i_

f

_ _ 3_{_,i _FO_ F_

r :-

yg O-p' 1_llg , -

14 y

g' IE-Fi) _

. -. 7. , ,

l I

,s_. w g. h L' ..i 6"--@h p" -

I j r - --fLM.i. I r---4 7 8 f I

.Q i l en g}{,il -

l 1

.I a e

, .i *e..

y

• c-

, .i y

. ..  ?

'...'.___"4 . .:_ .. . _.{h

. . . __,_q, ,, ,, , , g g l

: A @ '5 ild ;!  !. :

'D l

i Dil .w d n:2 in~ _

ri *'

j l

l sm ftiJ = !! q 8 a, 8, ,, ._._

Q.

i {. .. c...eg L- r1 i,2 h.

l s  ! g l-

!.  !  !! I ]ify.j! 9 a

n 4,w ,j -

L_. y pg =7 b.- h [j 2

• s

. s o

• C L b <

.I!

_ i i

i s .

--C. s

' I j r.m i!

e...,

. , - . ,-m y a 1-  ;;

• e i 31 y' .

V- 8)

B 1 I

- q'M

" "~

' E

,:- r EN

,@j ,,

.O "O. V .

N,-d)i4e ! '

, y f '

9

,..s._ ,

~

Hz ., g . ~ . _ . .

_ n I .

c._ss  ;*

E ;I' t

y l

r 8 i  !

t- i Q s ---

i --*

5.

~~~

f jt /.a -

f . .

FICl%E f).2 f'%

1 9-3 ,

treated water to the spray headers.

4 The vented containment gas mixture will rise through the iodine removal cprays above the pool and flow upuards through a cand filter.

The cand filter uill act as a .high efficiency particulate and heat absorbing filter which has the ability to uithstand the high heat and pressuro expected after an accident.

! The sand filter is composed of layers of progressively smaller gravel

. and sand covered by two layers of larger gravel and sand. The 4

overburden of larger sand and gravel is needed to maintain the sand filter's integrity and to prevent dusting.

The vented containment gas will exit the sand filter into a plenum above, where it will be mixed with outside dilution air. The quantity of dilution air is sufficient to keep the hydrogen concentration of the vented gas below the explosive threshold.

4 When the accident occurs and core damage becomes likely, the appropriate containment isolation valves would be opened. This action would not initiate flow to the FVC system until containment pressure reached the setpoint of the inline rupture disks (35 psia). The appropriate flow path isolation valve would also be opened. The other valves in the system should all have been prealigned so that no other

. valve operation would be required. Once the containment isolation i valves have been opened, creating the possibility of FVC system

operation, the iodine recoval spray supply pump should be started.

,QV When the rupture disk finally blous, the containment vent gas will travel down the 54-inch gas supply piping, into the sparger piping, through the pool and iodine removal sprays and into the particulate sand filter. . This path of gas venting could continue until an isolation valve is closed or until the pressure in containment falls below a pressure of about 5 psia (the nominal height of the pool water plus system losses). When the vent gas supply line flow indicator indicates little or no. flow, or when the containment pressure

indicators indicate a pressure of about 5 psig, the pool discharge vent gas header isolation valve should be closed and the pool bypass vent gas header 1 solation valve should be opened. This action would transfer the system to the low capacity / pressure code of operation.

The combined containment vent gases and dilution air exit the j building through the backdraft dampers and flows through an undercrcund discharge tunnel where it is monitored for radiation levels, temperature and flou. The vented gases are released to the atmcsphere via a 400-foot freestanding vent stack.

, Two large exhaust fans are also provided to maintain either primary

, containment at a slightly negative pressure if the containment

. isolation function is degraded. Exhausted containment atmosphere

} wculd bypass the suppression pool via the pool bypass vent gas header. .It vould pass through the iodine renoval sprays above the 1

pool, through the sand filter, and be exhausted through the vent

[ stack. Auxiliary subsystems are provided to maintain proper chemistry

, in the' pool, supply building heating and ventilation, and supply.

9-4 4

7

A+$.$ , eEEv <e 1,e~

TEST TARGET (MT-3) 4.% 4 l.0 ,'#B2RM S

a E24 Hn D bb l,l p l1.8 I.25 1.4 1.6

< 6" >

MICROCOPY RESOLUTION TEST CHART

1. 4 4%

ARn/ T .h? ++

.it_ _ _ = ._.

b +%.

+8 NNNNN W&,,, h> ,

///h<#,,%,

Nv' IMAGE EVALUATION TEST TARGET (MT-3) l.0 'dDLS E4

! E!E EE l.1 g.88 Ee

. lla 1.25 1.4 1.6

< 6" >

MICROCOPY RESOLUTION TEST CHART

1. lll4 ' +%4%

• ?QV *///;/ 'A :h4 5,/ 7 L.-

(,

= =_ ,.a . u;..; -

. = _ _ .La_.

%< y i

dccentamination availability.

,3 9.1.2 gystem Operating Modes

('~')

9.1.2.1 Hich Capacity / Pressure Venting The accidents that the filtered vented containment system is designed to miti ate G will generate pressures of 35 psia or greater. At this pressure, a maximum ficw rate of 467,000 cfm could be sent to the FVC systen. There is a distinct flow path in the FVC system for handling the cctbination of high pressure and high flou capacity.

The flcu path which is utilized for cases where high capacity and pressure are not present is described in section 9 1.2.2.

After the spray system has condensed the steam, approximately 300,000 cfm of gas will leave the pool area and be forced under pressure through the sand filter directly above the pool. The intermediate sand layers will capture and retain any particulate matter in the gas stream. Upen passing through the sand and gravel layers, the gas will enter the dilution plenum. In this mode of operation, dilution fans will supply 300,000 cfm of outside air to lower the hydrcgen concentration.

The gas is forced through a bank of gravity dampers, into the discharSe tunnel. Frca there, the gas will be dischar6ed thrcugh the discharge vent stack.

9.1.2.2 Low capacity / Pressure venting

/~'N t

Uhen the pressure in the postaccident containment has receded to approximately 5 psig, it will no longer te possible or necessary to pass the gas through the pool of water. By closing the pool discharge vent gas header isolation valve and opening the pool bypass vent gas header isolation valve, the system will be able to continue venting the containment. This valve line-up puts the system in the leu capacity / pressure venting mode. In this mode of operation, the vent gas discharges across the surface of the pool's water and travels through the iodine removal sprays and on to the particulate sand filter. The spray supply pump should be in operation in this mode in crder to cool the gas and remove iodine. After leaving the water spray system, the air flow path is identical to the high pressure  ;

mcde of cperation. However, the flow rate of gas from the containment '

is substantially reduced.

i 9.1.2.3 Partial vacuum venting i

l If a ecntainment isolation function has been degraded, the partial vacuum venting mode of operation may be used. l l

In this mode, the containment vent gas flow path to the filtration j building is identical to the low capacity / pressure venting mode with respect to the valve line-up. Operation of the iodine removal spray subsystem uill probably be required, depending upcn ccnditions in the affected containment.

m Y

l l

9-5 ,

1 i

The flow of cas will te from low flow to'60,000 cfm, depending upon the extent of degradation of the cortainment isolation. The gas ficw

p) uill be draun through the sand filter by use of the vacuum exhaust

^

v fans. The intermediate sand layers will capture and retain any particulate matter in the gas stream.

4 U en passing through the sand and gravel layers, the gas will enter ti.cdilutienplenum. The dilution fans are not operating in this mode, and their discharge valves are shut. The. gas is then drawn thecush the two vacuum exhaust fans and exhausted through lhe diceharge tunnel to the discharge vent stack.

9.1 3 Evaluation 9 1.3.1 Physical Effectiveness The sand filter used in the system is of proven design and is

' It is relatively self-sealing and has the ability inherently sturdy.

to withstand shock loadings and large changes in upstream pressure i without becoming inoperative. It has a high heat retention capacity, is fire resistant, inert to chemical attack, and its efficiency improves with life (to the point of breakthrough). However, sand filters require large arca, have little design flexibility within

cost and efficiency limits, and require higher operating pressures 2

than other particulate filters. In addition, decontamination and ,

dieposal of the spent media would be difficult and expensive.

The mechanical ventilation system was selected to dilute the gases

()

4 leaving the sand filter with outside air to keep the hydrogen ,

concentration below the detonation threshold, since detonations within the filtration system are unacceptable. Dilution was selected instead of a controlled combustion because it is a more reliable method even though it increases the size of the system components. Other disadvantages of this active system are that it needs electrical power.

and periodic maintenance. In addition, dilution downstream of the sand filter does not prevent detonable mixtures from being formed in the air space between the water pool ant the sand filter. The damage potential of a detonation in this region may be significant i even though the filtration building is a relatively massive structure. Disadvantages of dilution upstream of the sand filter

uould include requiring a larger filter area and higher fan static

pressure.

.The iodine removal subsystem incluJes a water pool and a spray system uhich have the advantaces of proven effectiveness for iodine capture

~

by a buffered sodies thiosulfate spray, a proven effectiveness for steam condensatica and heat removal by a uater pool, gas stream prefiltering, and simple operation and control. The disadvantages include the need to periodically test and maintain the chemical l solution and the need for an active, very reliable. spray pump. The only other viable iodine removal system would require charcoal filters which have the relative' advantage of longer maintenance or replacement intervals and the relative disadvantage of higher initial and replacement cost, the need for cooling loth the gas upstream of the 1

, filters and the filters themselves and the need for-filter fire 9-6 ,

i. - .

e

\- protection.

If a large fr action of the noncondensible gases in the centainment were vented, excessive vacuum could result as the atmosphere is cooled and.the steam ccndenses. No provisions are included in the design for adequato backflow through the vent or for any other replacement of noncend<ensible gases.

9.1 3.2 Schedule ,

At least 42 conths are esticated for filtered vented containment engineering design and construction time, accounting for maximum '

potential benefit from overlap. This esticate includes no time for licensing or safety evaluations which could have a very significant impact on the schedule due to the controversial nature of this mitigation.

9.1 3 3 cost A total cost of is esticated for design and construction of the filtered vented containment system for Sequoyah.

9.2 Additional Containment Volume 9.2.1 General Description

'n

%)

A vent system to an additional containment volume will also allow high temperature and pressure gases to be vented from either containment after an accident. However, the gas stream will not be filtered and released from a stack as in the FVC system described in section 9 1 but will be contained within the.added building as it is vents.d.

The system rs shown in figure 9-3 will penetrate each containment structure with one 48-inch line. Each line has three isolation valves and a rupture disk as in the FVC syste=. The lines from each containment are routed un'arground and tie into a co==on 48-inch line 3

which then continues underground to the 1.5 million ft added contain ent structure. This additional volume will serve as a mass and energy sink to reduce the pressure loading on the original, vented containment structure.-

9.2.2 System Operating Mode ,

The vent to the additional containment allows both high l capacity / pressure venting and 1cw capacity / pressure venting through  !

the same flou path since no major pressure drops are encountered. .

No valve realignment is necessary since the original containment will i vent until the flou stops when the pressure are equalized. At this j time, t,ce necessary backflow will probably occur as the original i

()

containment cools further. No provision.for partial vacuum venting has been included in this design.

I

~

9-7 -

l

A3]lT O\ A_

~~

CO \ ~~A '\ V Ef\

.1 4

RUPTURE Ox I *N= xx O

• 2 48- zs

\ /

O

~

k 6 ADDITIONAL '

1.5 x 10 FT 3 20 PSIG J O FIGURE 9-3 .

~

!O .

4 98 ,

c+

9 2.3 Evaluation 9.2 3 1 Physical Effectiveness Venting to an' additional containment volume shculd be very reliable

('/)

\- using this design. Other than prealigning the containment isolation valves, no active co=ponents are required for the vent system to function during the entire course of the evcat. Increasing the volume 3

of the additional containment to three or four cillion ft shougdbe considered since the pressure reduction with the 1.5 million ft volume was not adequate for all the accident sequences evaluated.

9.2 3 2 schedule At least 36 conths are estimated for the additional contain=ent vent engineering design and' construction time accounting for maximum potential benefit from overlap. This estimate includes no time for licensing or safety evaluations.

9.2 3 3 cost A total cost of is estimated for design and construction of the additional contain:ent vent for Sequoyah.

9 3 coupled centaince_nt 9 3.1 Gener al Desuription The coupled containtent vent system will allow high temperature and

,,() pressure gases to be vented from either containment to the other after an accident. The gas stream will not be filtered and released as described in section 9.1 but will be contained as it is vented similar to the additional containcent as described in section 9.2.

The system as shown in figure 9-4 will penetrate each containment structure with one 24-inch line. Each line has three isolation valves and a rupture disk as in the FVC system. The line between the containments is routed underground. The coupled containment will serve as a tass and energy sink with both active and passive heat removal to reduce the pressure loading on the original containment experiencing the accident.

9 3 2 system operating Mode Similar to the vent to the additional containment, the vent to the coupled containment allows both high capacity / pressure venting and low capacity / pressure venting, venting until the containment pressures are equalised, and necessary backflow. Again, no provision for partial vacuum venting has been included in this design. .

9 3.3 Evaluation 9 3 3.1 Physical Effectiveness f3 Venting to the coupled containment would not be as reliabic as venting

()

to an additional containcent since core isolation valves must be opened in the vent path. In additien, operator action from both units 9-9

E

- I OO D Z l UUv __ _ _/

=

~~

C O \ ~~A' ix V \ -

I i

4 t

t t

_ _ _ 24" ,, _ _ .

vs vn vn l; vn vs vn

\  !

RUPTURE DISCS .

iO i

i .

2 FIGURE 9-4

1

! l l

l 1

l i

1 .

l

O i

i 9-10 .

t.

+

. , , - - - . . - ~ ,,, -,, ,, . --. . ...., -, , ..,-, - - .. m , , , , . . , _ , - , - . . . , - - _ - . , , . - - - . - , - , - - - - , . - _ , - - , _ ,

. . _ _ _ _ _ ~~

is required to prealign both sets of isolation valves. However, no other active components are required for venting. Increasing the vent size to 48 inches ~should be considered to improve its t

() effectiveness'.

9 3 3 2 schedule At least 15 months are estimated for the coupled containment engineering design and construction tice accounting for maximum potential benefit from overlap. This esticate includes no tim fue licensing or safety evaluations which could have a sign $ficant ap ct on the schedule.

9 3 3.3 cost ,

A total cost of is estimated for design and construct.an '

of the coupled vent for sequoyah.

9.4 Augmented containment cooling 9.4.1 General Description Austented containment cooling was acco plished in this conceptual design by containcent air cooling. It was proposed in the design that the existing upper and l'ower cocpartcent air coolers (see figur 9-5) be codified so that they would be environ =entally qualified for

. the~ accident conditions. None of the air coolers are currently

_f- claited to serve for postaccident heat removal. To qualify the

(,j coolers, new fan motors and controls would be required along with some revisions to the present ERCW control logic.

9.4.2 system operating Mode The augmented cooling from the containment air coolers will be automatically initiated upon signals of high containment pressure

and temperature after a LOCA, without waitips for indication of hydrogen generation. The coolers will constantly be removing heat at, their c2ximum rated capacity and will be available to mitigata heat added during potential hydrogen combustion.

i 9.4.3 Evaluation 9.4.3 1 Physical Effectiveness The cooling from the four upper and four lower cocpartment air coolers should be reliable since they receive trained power and are seismically qualified. However, the active components (motors, controls) may not survive in the elevated temperature following a postulated hydrogen burn. In addition, the coolers' limited short.

term heat removal capacity is not effective for the rapid energy addition rates that could result from so=e hydrogen burns.

9.4.3.2 schedule i k-r~) At least 18 mon'ths are estimated for the augmented cooling engineering 9-11 .

.g., , ,m. - # .,e,, , pyry-i y-y y_, ..--,e w 7 -y+e-a 9  %

- n . r,. .

,A

..n- :m .:,. p. . rri-

.~. . a

'"r*; ; -. _-- - -..

s .

.... , . . . . . _ .-  :-.=. . - m r. .- . ......-_...----.y ---.~

. , - .T.-............,.

- . . .-4. . . -

4.. _

p.- _

,; l e

,. i.

t.,3-s

,0 .- ., z -) .

g

,r1 g :', -

.i . .

,.U n

y

" m y 7_.g_f;.f /.J l -

,i , N2* Ir e. ' " 5'

.I.t.,/,)

m.Ds

%,% \ ,

jj a c y.g.. . r"'rr ,- . - - -(A- J m-- + -- - .TG, - H.L -

-- 6.%'

n .- 6 .' c.- :

/,j '"U;y? 7.M.M '.' yb,7 4 ,)..b-G-

. V; .:W:i b*} 20':-i.M.] QPAR G. e+s . c'  ;

t

.:j/- e.,  : t_ .L, .T " ', -

H Ov

,u.)  ;-

r

. , . ~ - . - - -

-' M..,,,t ., "- Us.v.

.v.

j n>

A p: f.: : .,,

I -- - o.. c..dl, -

e,9,a:

q n. H tt 1...,,~,,, ,

.r. , ,,, c c.

e- > -

e y ---

.i-.

i-),y.__

p.E.al q >s .a t

o ;m . :al v. -.:..

s.

T.T--' } r- 2e O o "- elf 13 le!t' ji J-'J U l ?)  ! .,,,,----O N I .,

52 % '2 F t h.

~-1 4:;

I'9 1

.a,. ' ~ ~e , i, f., {4 p fi x -.-o l1 p Jp 4 7

".1 6 # 4 e SIE l ./~i ~!

4c: 1.s -st.1 c a -.o t

- .i = c, ,: -

4- p.". . H.l' e w 9:.-lr . '

(cd -..

_1 l :.

o

.4

-> u.4 t- W

'd'.' .i ,G%. N <: c_m

. .: i.3 g

h).9 i

.T' . 8 i: to M.I e  ?*- .

n 1 qu ., CA

- .4., ,)

- ==I 4 :t y,*._( . c-- - 1  % .j '-

t

~. ! t* - l- c.~

.tt . ,l$:

1 l~

'- ,our : .,

.a l

- , 1, 4

- .4-  ::

tt a: l A vn n  %

j't) e,  ;'s 1 '

G

. ,.(- n n rg a I'1 a~n. i E[ .. .. 7  ;,_ .t.:.  %{!.5*I

~. - . . . ,

3, -

Mi dp

'7 6

')W

$ E.S 1

n y Q --

nr hj  ; ,. -- ,

. k

[

O,;yb@ ir

, t  : ,

i r x;; go . _. L -: -._. t t.

t -, S 3

,# H.i 3a, , u. ,

m

.m,,- ; 44,

; #1

1. 1  :: , aji .-^ : 4

..' f t%:..:U412,i.

2 - -

y - .

at<% r p 9 r4 - m. . l n.s :

ji1 .

s.- .n , e i s 4 -

e(

3 i t

_: t . g ar y;. ,.,{' . j4 t c.  ;-

.,,4 en ;_. .- . . : d, y:.:g, V [,. .

!5 . -:

M't. dt ...Z.@b

? 4.;

t l  : ( -

!- Q(L % U ^ ' .L m *.,O: n\5 - - - "

i - - - -

f Il !

IA $j $ " > '$ [ [y l 3h,I f:'

i. }R fps3 h.1 -

. l. ]) U- I'tJ[,

r. m q

~

.[t .

. :3

~4- ,

.. I i 1.

g.

1 I;.

wh

$ ] -

f .'4.. up ,, I h- -k. )ne.rTl--- e .- a ,a y

! l

-{!

a t of.,.,un ,, p w.q ';)

m,- i 67.r,- --  :

-; 5g 9

t Ip 4,19t,j?it UI t  ;

r.

+$j-  :

r--"'M r I;Z@,, +=

dM g&.g 5

-yd.- i @j 4)i t 1 :,rls

.. 9 m

=

1. w.i ---4 s! .d(!

i

@ s i j r, - r 1, U.T

,1* i f, . o.

.i ' Q 3 aj - r

i: h t.;

G- k., -

tt ' a i: ., t h.

9 I .ljd l ,

.I f! .)

m:

B t h*.i E l$2 '

1.u. ,

s.

w!.pA . d,- , -., x _Lf < . .,~ o N. - .

i. a n- t i t 8( t i

^~ 8WP -8 ' I>

.u, -

4- ,? i

m pc> :. d; O,j 1 1

- bq- p' mgi

• Ai 4  :. to 9 - '

(4

.i $. +)r.4 ,s "1,,

1  ?

tl :g> , U,-

3 e

$,E3 W ( 3 ,b ) A- ,

).;.

k l ' " '

It w-

.ca. t p.

~ ttt; y

2 h n% U 1 g.- L 0 gg

.L.

i 1a 1 n.. r,* -

u:

~ky ,.

t i F

b.nf *g i

l  :

s.

9 .c 4

". 4 -

q A*.  ! t a t .. u , ,t 11.' i

, w .It i

[ =

v at t.

5. : ,- ',e ,. 3 I. h a i, J. , I et

sta p j[a,kpd,

.Nk! :;;t fg U l I

}}

dg Iw $!2 h .E i fl-. 1 __ 1 c ", .

t.-

h.

i

[\

i 4

I[f.j

.,.4

\..,m, e' aIfI M. .wr.v \ 9

• %: . e..a;, wI I ;,,

2 .a fNg/ d

,. ..e

t. - _,s

j;' . .i - - [ g, yC e.m[M.)g.!.. C1s-k { N L;

<MD h M,.l...n- - --F*i1

-r y, :e:1s h; -----~

3

y.t.2p-- u nd -) -

+f.v.1

'. a or:. , G. .. 's .%9 .".;

w_ C

,.: I dp, e . /m. p ,,.1 } Q , '"

ls ? ,.I I

w

._3__ _

M 'ig, fTI yi9.'NN .- .=g =g' q r - --r' =.. e 4 .i _q4. ;

=:_.==.-- . - . . _

.,W,(?(I ..g'D I 1 , . , , , . , ,_ , _ _ . , . , . , ,

@icj 7,uTD - VV IIY , ,,,

. .! . w_ l 1 .I 9  ; i ! j..l_1 d j : t m n,s.1- .- .

,1.' h*' "V Q. .

+ -w--; -. -. - . . . , . , ,7 .4, . - - , .- , ,,e ..--,-.,~,.-,,.g.,,, ,,s,n-,,n r . , - , . . , ,--,..,w~a.,-p.~- . . . , , - - -

design, procurement, and construction time accounting for caximum potential benefit from overlap.

• e~s 9.4.3 3 cost U

A total cost of is estimated for design and c:nstruction of the augmented cooling at Sequoyah.

9.5 centainment Inerting 9 5.1 General Description The containment inerting system designed for this study prevents combustible hydrogen-air mixtures from being formed after a LOCA by

. replacing the air in the containment with nitrogen when the reactor is brought to power and by maintaining this inert at='osphere during

. normal operation. Liquid nitrogen, stored in tanks onsite, is vaporized and blown into the ontainment using the pr'esent centainment purge air supply system, which can be modified for mere efficient supply and exhaust.

The system, as shown in figure 9-6, takes advantage of the fans and supply and exhaust ducts of the present 28,000 cfm capacity purge system. Suppigingnitrogenatthisratewillreplacethe .

1.2 million ft of air in the. containment building every 45 minutes.

It was estimated that eight to ten volume changes would be necessary to reduce the remaining oxygen concentration to the design level of a percent. This would require about 140,000 gallons of liquid

(~} n_trc3en for each period of inerting so tank storage capacity of

\~/ 160,000 gallons per unit was included in the design. Vaporizing nitrog.a for the 28,000 cfm flow rate requires a heating unit of 8.5 j 1W capacity.

After vaporization, the nitrogen enters the purge air system upstream of the pair of supply fans which discharge into the containment.

The present distribution system could be modified by routing ring headers around the upper compartment and around the lower compartment inside and outside the crane wall with branch lines into compartments with poor circulation. Nitrog'en to the ice condenser compartment must be supplied at 15 F in a separate header to prevent ice subli=ation as much as possible. Provisions were made in the design

, to deinert the instrument room separately to allow freer access while

{ the rest of the contain=ent is still inerted. The present exhaust system could be modified by taking suction from the present hydrogen 4

collection system which draws air from all the tajor co=partments.

., Some =ixing fans would probably be required, and any additional j modifications could be made after an onsite survey of the containcent.

Since auto:stic control of the inerting system would be very difficult to design and operate, the design specifies canual control by the operater. Extensive information on local gas concentrations would w be provided so the operator could adjust the flow rates as necessary {

to ensure good mixing. '

, l'h

L.) - ,

i 9-13 1

,- I

. e l

,fC ,A1 ,,

g g'r}- D

' I s ',

I ' [. [ k' I 1 i inI

-M -I{.--

^

M -*

i-Ij ,'

f; p Qi)  !!!l t i..; d- !!

>; . i.  ?*

I :1 ! . ..

h.is.} - WJJQi) if4 tid :3e :

15 t; e, _

9 .'i _' .

e. -- -

fs U

E N'j@

$m- -

k E!!{d d[j .n Pf e V - :Jll NEs (!l 7:

li 4 l

~

{, Y' s 9

l '

~

v. \T I

m I .I

)

&@S 4

+ ..

s l u 3, sy @~ @T

~@gg

!!g @~ i

' $n ' @1 6

EaB 43 g]

x- g V

g"n* _c8 %

x x g _, , 4 @ - 4 '~ @

h 7:21/<W 2 / 5 f lb

--a, s e <x - - us. 3. , *

@rgg -

3. s 4 3 s p a

" ~

.I -

y e gi d g,!B #_

el na x,

w wI m i

- l ,t

@~I .

4 04 D O ,( *

\ D (- x $p.

, - N .)cq

/ /

rg ,

I I / g;) '

h-&

I ,

O _._

n -~ ~ ~

i f; .t- ,,.

.._._.m ,

,, .... a li!. h i;i ... .in.e 9-14 -

i

,=.

l

. . . ~ .

Cther nsdifications to allow effective inerting would change the present ncnessential control and service air systen to nitrogen inside

,_s ccntainnent. This vould require a separate nitrogen nakeup supply

( ) with its own 150 kW vaporizer. The four pressurizer safety and

'~'

prescurizer spray valves that require essential air could be replaced with nenair-operatcd valves.

Perscnnel access for ice condenser equipment inspection and raintenance (uhich occurs daily) and other purposes while the containment is inerted would require portable air packs. Access even uhen the containment is deinerted would require portable oxygen

.cr.iters to avoid local pocketing of nitrogen.

The design included provisions for naintaining an inert nitregen at csphere in an additional containment building of 1.5 million ft 3 as described in section 9.2. The amount of stored nitrogen would be sitilar, but the design flow rate would be cuch less since no forced dcwntime is involved. The supply, exhaust, mixing, and centrol systets would be simpler since the added containment building would not be as compartnentalized.

9.5.2 Systen Operating Modes Sitrogen to inert the containment, supplied at the maximum flow rate of 28,000 crm, will require six to eight hours to reduce the oxygen concentration below 4 percent. Deinerting with air to reach oxygen levels of 16-18 percent may take slightly less time at that flow rate. If supply and exhaust at these large flow rates is no.t allowed

(~T due to licensing restrictions (e.g., if the reacter is at power)

\> correspondingly longer times would be required. The containment instrunent room is capable of being deinerted separately in this design. A constant makeup flow of nitrogen will be required for the valve centrol system inside containment.

9.5.3 Evaluation 9.5.3.1 Physical Effectiveness -

This design has an elaborate supply, exhaust, and tonitoring system and should te able to effectively inert and deinert the containment.

The high flou rate of 28,000 cra would allcw faily rapid inerting, but nay not be practical to achieve in a nitrogen supply system.

Oxygen pockets may still exist after inerting in comparttents with poor circulation but should not be significant in the event of a hydrogen release. Nitrogen pockets remaining after deinerting could be significant to ucrker life safety. Deinerting the instrument roon separately kould not be very effective without further codifications to the room to anintain a breathable atmospacre. Any inerting systen that disturbs the ice condenser atmosphere vill increase the sublicaticn rate, but the cold air supplied by this system should have as little impact as possible. Any inerting systen that taintains the inert atnosphere during perscnnel access is very hazardous to life cs well as causing inefficient ucrking conditions and licited access duratiens (see Section 10.3.1) .

O

\J 9-15 o w

9.5,3 2 schedule gS At least 30 c.onths are esticated for the containment inerting system

(_,/ engineering design and construction time accounting for taximum potential benefit for overlap. This esticate includes no time for safety evaluations.

9.5 3.3 cost A total cost of is estimated for the design and constructicn of the ccntainment inerting system for Sequoyah and an additi cost of is estimated for the design and ccnstructicn of the inerting system for an additional containcent.

These costs include and , respcotively, for the cost of enough nitrogen to inert the two present containments and the additional containment once.

9.6 Halen Injection 9.6.1 Gene. 31 Description The containment halon injection syste: prevents cc bustible hydrogen air cixtures frc= being forced after a LOCA by adding enough Halon 1301 (C Br F3 ) posLaccident t,o occupy at least 31.4 percent of the centainment volute. Liquid halon, stored in tanks onsite, is supplied to four centainment halon ring headers and injected and vaporized through nozzles.

n, x_, To postinert the containcent would3 requir 11,000 gallons of halon 0 ft tanks, one per header. Each which is divided halon tank has its among own 750four ft 7} nitrogen accumulator at 650 psig to to supply injection pressure. The halen tank level is monitored to autccatically shut the isolation valves upon low tank level to reduce the arount of nitrogen injected into the containment.

The halon is supplied to the containcent through three individual 6-inch ring headers in the upper compartment and one in the lower cetparttent. Smaller branch lines will be provided into the various enclosed areas of the lower cocpartzent. At least one nozzle will be provided for sbout every 2000 ft #of containment free volute.

9.6.2 system ope.ating Mode The containcent isolation valves betwen the halon tanks and the ring headers would be autcastically opened upon a high containment pressure signal followed by a low reactor vessel water level signal. However, tanucl operatien of another isolation valve at each tank would be required before halon could begin to be injected. This is to reduce the pctential for inadvertent actuation. The halon would be injected in a taximum of 30 minutes after an accident, assumming it is all vaporized in the line before it passes through the nozzles. The flow .

would probably be socewhat faster in practice.

O qJ -

9-16 ,

e

9.,6.3 Evaluation 9.6.3 1 Physical Effectiveness 3 The postaccident injection rate in this design should be fast enough

(_) to effectively incrt the containcent before significant acounts of hydrcgen are released if the manual actuation by the operator'follows the auto =atic trip fairly quickly. Operation of the air return fans should pro =ote adequate mixing so that the hazards from any remaining noninerted areas are insignificant. Inadvertent actuation of the halen injection system as designed could be catastrophic. Unless sufficient heat is being added to the containment atmosphere (such as occurs following a LOCA), the heat removed to vaporize the injecte!

halon may result in a drastic depressurization of the containment.

In addition, the halon injection occurs at a higher rate when sufficient heat of vaporization is not readily available. Further analysis would be required to size any additional vacuum breakers or halen heaters that might be necessary. Inadvertent actuation when personnel are present could also be hazardcus if portable air packs were not i =ediately available. This design also suffers from the environ = ental ard cleanup problems that would result from any massive injection of halon.

9.6.3 2 schedule At least 30 =cnths are estimated for the halon injection system engineering design and construction time accounting for maximum potential benefit from overlap. This esticate includes no time for licensing er safety evaluations.

r^x .

() 9.6.3.3 cost A total cost of is esticated for design and construction of the halon injection system for Sequoyah, 9 7 Ignition sources 9 7 1 General Description _

The design for controlled burning of any hydrogen released after a l LOCA, before a detonable concentration can be reached, includes a system of ignition sources and hydrogen tonitors located throughout ,

the containment. The potential hydrogen release points were located l' and the ensuing circulation paths were analyzed. Ignition sources were then located to efficiently burn the hydrogen.

Since mixi , helps prevent locally high concentrations of hydrogen, hids in co pleting the co=bustion process, and allows the caximum u.'e of the contain=ent heat sinks, the design for controlled burning specified continuous operation of the air return fans with the '

hydrogen collection syttem after an accident. Containment sp ays would also promote good mixing as well as aiding in removal o[r heat from the burning.

The ignition system consists of 12 igniters placed in various

(~') relatively open areas within the containment as shown in figures 9-7 through 9-10. Any hydrogen release from a rupture in the reactor T

9-17 ,

.s .

. c m.

n .

v. .s -

!* a .%. ,,,,.

. .I. ...

__g, .g O

. r .- . .  ; r

'. s e

. ._c_ ._ . , ,

c n s-na.n

.u .

Jlg). i

~ ~ - - , _ , ;t7& , , r t . .,,LgrCH, .-

Lt.

~*

1.: w w .

-- -. I ,,

p.y f..:

(

21,r-

%.y .a.,! .

l .-

l. i j

h

f1 WNtn :.r N/

- ).

~

. m .

~.'.'~~;~* :Q' <-~*-~s

- .,7...

, G. .g. .

p .:.' ' %y s _ g.1. ,

n

-.G~K - . . + -

~~

ue.,.

,, . . . , .s 4

.... . _. mw . . ..

n ..s. ,- q y. ..y ,p .

n~~~.. ..

~

,,,s

?

i?. ] f c_. a l k. e..k cE w li g h. ,.

J. s

. . ~

/  %,.q ..y.m v .-

s

, , -m%g r m a~

z,, ;.~ w' ..

lip \

• N,.

.,, ,, ! ! s , .m.. .,. s t

. W',.:.;. i, . .

, ,4.

~. f s 's fl O . ?..,,,~vv_Q.e .v[. y,vV.A

.. .% LO A .. : aq:

i

. . . , , m .. .a -NN .

s . , ,

. ,ycf. ,, g 9

' s,tc /f  ;,-s a n d *- }* .

(Q.

/:2_.:{;T)=. E

_:.  :._%,. . ze--

Lt u- _ L *G ',' up_ ,- 7,,

Q ' ,

k

- . . q-.

n-,

+ .m . . . . . . , -g,? .v . ,w< %. .:..:s - , n2x,=t .c, .., . -t,. .m t

/

\ f{ ~

L.e .e n.- r-o.

n .

t

• -* \s.* . . . y - ..... . ,.; , . . ~ .

~-) .

~g'

=-%-f~N.-<>.W,.\' "=="%

^

1

! - ,sdn**- ta f, ,

. . . -. &, N'v~.'SZG%f..'  :

/

• lm *. ' h? - ' ' ' ' " '

• r.:::. ,rr \ , \. Q': ' * ., r . .. . -p :' n . N. ,.. ; .,i -t eJ 4 c

nn: \.

. \ :, . ,

... 'u_.._//

,s -

-x

,ci e: 2

/

=:a n v ,~w Y -~ j f; ,

'l,",'fQ, /.-, i s,K

% t ..N g,{, _,,,f:[$. .

. .< -=- . , y;,&s w,. .s . "Q(' w.G" =, g )N m.. '. xn e.c

q. X ,

p . u.. g,,s __

.. . , , . _-. /; ., ,_;.. n

_ _. u.

...,.s.

e. ...

.~ .m- <

. .;- %,- ,.. . .. 3 g!hf[,.. k

~ -

.g gg,,,,3

\.,,W.\./,'y,

~

Mj..: ,2, . .

s ..~, - :. . . .., . - ,

,,. . . s m., ,

v~-...~.. ,\

%*.4. .n,

.,2:i'-: ;, . x,.g.,. e . .. ~ u . / *=.a.*

y y:::.'

g .. D*,g: .

" ** e " 8 - **-- * , 2' .r.a.

g g , as si, s

,..... . . ..., ' 4' .' !. . v=~5".gY{A' s

..e.-i..- -% \

/

m'Q*.... ,, .. .. . .. ..... , .N, -

you.

.x.t .s.

s.

. ' .a ~

g jW ,

s n sw e i .i..s,.,...,,

• t sa 72s.s RAN*l,*W& ($MAe AILOf I"N N8 t *** tT'"a S h 5"ACTJe CD4* TW h ET"AM gds. CnJ. h '#*'G NA.N

~ w -~ e nn. a. ~ g---

81~ w.. ~ e .ru,, e. ciu. e ecrD - -

m-- n mn. n. e=a. e mr~* u.na mm:n au. g, - c24. - = w w.

i FIGURE 9-7. CONTAltEENT PLAff AT RPV ELEVATI0ti - IGilITER 1.0 CATIONS 9-18 ,

s ..

a, 7

%.. x%-% , . .

I * . [

, . -;' /

i Ka  %

N.

t .7 i

o

< l ,/.

f s f Ii

/ -

1 ' / _.. . ..-- k '

=

r <- :.~.:- [i .l

_. ,. s d.qi x /n._ .m .

a

/ ,

4 - t-- --

-- - -, s; -, y3

. . e

• • • "* / , ,

1n : .o.,g g'_;_

_ _ , ,C-- .g 4 ,

r: i.

y

• ,) (6.,.** . f .e a F .__ .-'

... ,i y . -. i u, . _._ .a,

- . . _ i

g. \., .i

.. .~

_a.  ;

~ s..s, i. ,.

4rl.. ,_ _ .. . _

'; --. . _ey e4'.

'a, i-

) g[ p,

\ ..

. --- i - __.f.i ( !,fl r,

i ' "

!_ l.

... . .I .

,. I i .....

d;.W t.o$..

~

.r;..:. 'il  ;%:

h.. .-

a___ r., ., . & {: d i ,.' ,

- .l .

.3. i-./4:

- , !( l 1

i . ,-

i i,

t Pg.o

• r. .. .

- *  !;>  :- ~.= >a

- p , &.,,,,,,,

i

.:.,;...;T- n... ,' .l. .N., , 3 l '., ..' 6m4 e

o

- u - x l

.i y b .., .,,;,Jp

.., $l $. E ..-.t_..,,.

~

a.

. ,p,. i

< ~ -

. ,_. A. .

i 7., ,.;S.

vp L. -

4p .

i

!c5-c><.uh

i. ....

i - m  %-- i ,. l ,

., ..,.t.- .#. ,m. . i i  ::  ::

--m - ,

O ,, i ;.

.,r.r_i 2. .

I.._. w .>,.r. ~1 -

- pi.~~-

I .u_.:.mv n- \z.a -- .:.

.i r

(

)

l ~L._J \  ;. :t

, _ 'O ,,..,,, -

l .. _._

C, --.c ' ,.I,'. . fM' ',.L....yt.-_ ,,,

, , .n A,

r--

I -- p l  %. t.,

. 3- f .

' .~'JL ,. y .y ,,

. . ,- ,: ~ --.r

- ft-t.. . ., f.e' ,y, . f' f. ': * .;f.

3't __._

l . J. ,

3 3'" "  ;  %, % . ,- M

i. .

5 : .! '- ' -  ?~.- 79 I .- &. - . ._ a ..

.t 42 ' ' . . O(iip P. T); - .- . l p:x.q), . >

."%r

-i :m. .. m

.y

. X N.

s. . .

., .U d.

e-

.a.

.(.--.

- . r .

~7

. g ,...,z. . ' - l . . .' -

. l 5 _, .s 1 i ,. .,

E I

• g * -* .-..

. f i ,.4 0' 0 0 1,

• e. %

.,..3*.. . * ..

, .i .i 6 l l l 1 w mm s .vericu ois.o,e e s=:A M'pt ClpgnMT)C.'T STE.Ane ces, ca.L, h U*'m C#PART*t247 c:, .- e o c ,.. m g - c--

e.c. c-- g . . o . cru. e ,

n g Tm- em n,.m. o e.a.

8 -m eo ua c-~ ca. g, _,m,,c__

o. c=u.

I

< FIGURE.9-8. CONTAlt; MENT ELEVATI0' WITH PRESSURIZER - IGNITER LOCATION p

J 4

[.

}- 9-19 ,

.e.y- - . . 1-...c<, , ..-..e ,.-w-. . - , ,-.--.m.< - - - . - - -..%,- . - - .. - . -4 .c, e

<~-

uq ,

~

. . _ . _ _ . 1 _ .I.

w v.M~s:::gsa .

.w,

,.7

,K Ws ,j. n w st * '- .

? t yQ'a! x k 1 ::= G$ S;M;+;w:.y;gs8 n ='~ s '

?ys%..;;..p a \ ')

G,%:rdjWas:a ..,,t

., y y ,. , W ,.m 1

l g-&

, l \: spa .s: . %;l

-, -n x. ,. . . .

~ r'.v W @.%g n<:n.6 .'~ *'h ~

dEEi -

w%

'h

. 5,n , . mn ,, . Ni .

u y's *. /

n ,/ s.nw- f,;61:n[. p +M .r:=~;=\r.\

.; g

.?,

~ , .. . , n. .. ,. ,

1.

j wf .. ..:n..;.-gs _s

'-= .r::

/ il

/

/

A , ,,,

./gyZ) 4 N~*fj h..""'5 p)f [~;r.:.,g\

P Q7 E:i!!?!

,Q))- 1-  %==lM3  :

\{

• L - ... / /t , + . =

gkbh+'b>).:'l~.h;j  :

a [{ ~' \,, n mg ,qi . . ..~. >

r .

i n . Oo

.( A . .- . \ , ;o s

-o ns  %./

j.

L'. * - - * ' -

k ',','['

- i

\\

l t

..1...-/. l m :.. ., nl

/.

\ 'y.,,./ ld; ,c'

~

\ \.N ' l~- n, ug ., .. . .. ;

xl. /

\

f . ., . : . . . . {'

.-. - .~n, /

z-j-y g ,; = = :. 'ip

!!@l !ll5Gl /

N .a.. . .g. . . . -

QZ W

~^; . . .. ~ , ~ . . . . .

w&

i

, ; e. .g . .

a: u .'<u _

. - a.. ,,

.w -J.

.ota.v-tff!R CCU;4ATWNT

!",N!m ' 'v'm?S Q .sn ex..n.ca @ sun w. exu. @ v-m eve =r-

-==r-< e smo, w . cru. @ - --:-

82 1 = c= =~ e s u,. . esu. g en ><- -

3- 8'sm sser:n cr2.

sn u. w . era. O msui:a ru - crc.

m ,,m.,

FIGURE 9-9 CONTAIRMENT PLAN AT ICE CONDENSER ELEVATION - IGNITER LOCATI0tt i

.T.3 9-20 y - . _ - - . -r

r t

. ~

%, ,h% ' %, N

, ' //, g N N \N.'4j 1 . i A g' f /  !

.;- "q:- - A%

t t 9

.- ,, 1 e rl

• . f .

j, ,

~-' t-1 l

/

[b K n g

.s,.,

.,.- ._._.__._._. tt ,

[ .~r

. yI

< { n. ,,_s, \

c

. , , . _ u-W,p=6 2 l rd}, ,ms .. . . >. . . . t -

t},

-: - u m y a.-.~,,,,.- gI . .,__,p

, , ., , .' ij -

- l' -i

,- , O]. Jv 3 l-I,'.:*'!h 1 *

[

y :r.*,.:: ^ : c-__- vm-

• h ,- ' .

t

--,,v @,

f. .; e y6 .: - . .

l-

_.. . . .T m

. c ? . :y+ l i-l

---

i. i t ._..__a, 6i e  :

l1! L' dt.. s i.

- =.,. N 7.j ,i., t $.e J.;.4..'.i..g

.L r- 'i

.- lc / i: .. w ,e! 3

? .6 :. :r '

e It t 9 , . ,

w .., -

. _ ?. , m} ;1l,7.*3 .

1,  ;, n .' .* ,

.is ,

3.

/u,{T.1, l i

. lc .3. . ..: ,) .:' Ua ;  ; g

.. ;3.,p;,s

.l.  : , . ,

' ' .;,. .i. j .si i1 *m

%, ,...f.

g i . . ,

7 ..

l.

s .,

<~... l..-- );t,(

U 'I**' - -

Ib,k.).' . ' ';~;'::'W~hM !Il,sh:..f-l- L ; 7 y }:  ? t:  ; n T lii

l.i). ( j,jt .Q. ds .. .:ifi ,

l;a :b .( ..Y k

.~

7;;;;-:;e._ _ .

i.- -l

c. ,- s

. "q \ i. ' .yM +i ./ t -

I f - - - -

. , , . , _ - h r ., l.'f'I: :-. pgm, g..WXh O.

_ +si

+': f, a ,-MN< L.

,o uw- ,i .e., v =. r,,=

3,,.

t ._ ; - 1.la j

, . . , - & 4

" ' ~ " ' ' g

p?. .. , @ a

!. .i

~.J i. m. , I .~.--. .

a , , -

, G , 9. aG:1 ,l . .

., . A. .

1 , -. . . - , n, ,, .  ::-: .:.- .:-- -

.:g ; p . m-~ ,j vi

!'t,B.]y .i .M - bI')3

I r

-g;- d '"", v tf 't.;-.,sk..4

! 'U$P(N:  ; .,j o al  !

, - ~l

. .m . . ,,n . .r~f 4 1: . .'- .. i p' :  ! ..

f[" ' i

~ ; ! l:- r-[- , , . . > ,' .,- , , -

[]

. 'l Y S)t f. - El . if -

. x r: .

- g .- '

1s. :.

.t.i

. . . . ll .op. . .. . . ' - .

b . . . . . j . 4lb..

,-l

\. . . .

p, ,.,

j.. 7* . - l l,

(..

w u- .-- . -

e -

.i .a.. ..:m . .

..g

- - . . . , . . . . . . .g

• . . ,. . ~.

m ,

scenov tu-ca n.am s_~ sms eerm en..,wa @ mu, a. au. @ v'm c"'*m" us e.ar en g m e. au. @ wecw='e" u:.u ex aren g mma.cu @ ecim co vs n i.c,2 c-mm,

_,,x m a=.

_ ,m,

m. <=u. e -= Ta* en e.-

FIGURE 940. CONTAINMENT ELEVATION WITH STEAM GENERATOR - IGNITER LOCATION

.O

'%J' 9-n ,

e , , 4, _ . , -, ~ -,- - - - , - , ~ ~ - , - -

ritar, lecp ce frem pressure-relieving devices vill cccur in the louer ecnparttent of the containment. I

ine of the ignitors are located in this pertion of the centaintent to burn the hydrogen as

(,s)

scen as it reaches a combustible level. If the hydrogen is suept into the upper compartment, three igniters in this area will burn the hydroccn in the same nanner. In addition to the 12 igniters, tuo usste gas turners uith integral igniters are 1ccated in the louce cetpartment above the pressurizer relief tank (see figure 9-11), where they vill burn hydrogen from the reacter head vent or the relief tank vent.

Igniting hydrogen-air-steam mixtures at low concentraticns of hydrogen (4-3 percent) requires a uniform, well controlled, hot ignition source. The design specified an option of continuously operated glow plugs or cyclically operated arc igniters. These should be equivalent to flate igniters and easily meet the minimum flash point temperature of nydrogen-air nixtures. The igniters will be shielded to prevent water spra fn frem impinging on them.

c.7.2 System Or. crating Modes The igniters should be started automatically upon an accident signal or nanually if the hydrogen analyzers indicate the presence cf hydro;cn belcw detonable concentrations. Hydrogen levels shculd be individually monitored at each igniter location with indication in the nain centrol room. Great care cust be taken to ensure that igniters are not operated when th< potential for hydrogen detonation exists. -

O>

't- 9.7 3 Evaluation 9.7 3.1 Physical Effectiveness The air return fans should provide mixing between the upper and lower cetparttents that is adequate for dispersion and combustion of the hydregen. Scme additional mixing fans may be required in localized dcad-ended volunes. Uith adequate cixing, the igniters as located in tPIs design should ensure that conbustica would be fairly complete for Icu hydrogen release rates and should burn most of the hydrogen for larger relcases. The interaittent energy addition that would result frc: these stepwise burns as the hydrogen reaches combustible concentraticns should be within the containment heat removal capacity. Careful hydrogen level conitoring must be performed for pecper igniter operation, and careful centrol of the flou to the waste Cas burners uould also'be required. Local flame and temperature effcets should. be considered for cach of the selected ignitor locations. Inadvertent actuatien uhen excessive hydrogen or other flattable matcrials (solvents, paints, etc.) are present could be hazardous.

9.7 3.2 Schedule 1.t Icast 18 months are estinated for the ignition system engineering design and ccnstruction time accounting for maxitum potential benefit

(~'i fron ovcriap.

LJ 9-22 ,

e

ELECTRIC l LEAD i IGNITOR k

7 = ^^

mg;

c. O 1

Q qu s /

3 1

/ -

I

\

IRON / CONSTANTAN LEAD / 316 SS SHEATH j

l X

1

!o ex ~

! v.

q r 2 x l

f PRESSURIZER l RELIEF TANK i

e I l

)

FIGURE 9-u. HYDROGEN IGNITER FOR REACTOR HEAD VENT i

!. 9-23

_,. . ~ . . _ _ , _ .

, , _ _ _ . _ , , . , , . , , , _ , , . . . . . . . . _ . , . . _ - ~ . _ . . , _ . . . . _ _ . . ,

i f

9.7.3 3 Cent i

i l A total cost.cf is esticated for design and constructic.

cf the ignitica system at Sequoyah.

l i

l

{ DA02 :!!S AHYD.M5 1

l 1

l i

i I

l i

i i

l i

i I

~

i~9 I

i I

i i

==

l 4

l i

.I I

l 9-24 ,

.. . -. ~ . - .

10.0' . Evaluation of the Alternative Concepts .

The evaluation of the alternative concepts included the potential benefits, problems, and costs. This evaluation stressed the relative

.() merits of each concept; definitive and quantitative evaluations would require additional design detail, analyses, and research. None-theless, this evaluation does provide a good comparison of relative cerit. Limitations of this study are discussed in section 12.

i 10.1 Overview No individual concept or combination of concepts is without significant drawbacks or uncertainties. However, several appear technically feasible and relatively effective at mitigating the effects of core damage.

The cencepts for hydrogen burn suppression (nitrogen inerting, halon, and ignition sources) provide varying degrees of protection from the

effects of hydrogen combustion, but provide essentially no protection from overpressure due to noncondensible gases. These concepts are discussed in section 10 3 4 The vented containment concepts (filtered venting, additional

containment, and coupled containment) provide less protection from the effects of hydrogen but provide substantial protection from overpressure due to noncondensible gases. These concepts are

• discussed in section 10.4.

10.2- Evaluation Basis ,

The following f actors were considered in judging the relative merits of each concept

j Effectiveness - The ability of the concept to accommodate the j effects of core degradation, including hydrogen generation.

! ~

Technical Feasibility - The present capability to implement each j concept, emphasising any serious technical problems.

Additional Rick - Hazards to nuclear safety that may accompany j cach concept.

Reliability - The degree of assurance that the system would be available and used when needed. This includes the inherent reliability of the system and the need (and time and information available) for operator action.

j Cost - Includes capital costs, reductions in plant availability, and operating cost. Special:research and analysis costs have not been included. These may be large'and cannot be estimated at this ,

time. \

10s3 Concepts to Prevent or Minimize Hydrogen Combustion l The three cencept s . evaluated are inerting the containment (with i

nitrogen .during operation), halon (postaccident inerting), and 1

1 10-1 l

_ . . . . . , . - __~.._.,__,__,_,- -.,_,.__.,...,_!

ignition sources (controlled burning).. The first two prevent combustion, while the latter attempts to .'emove. hydrogen to prevent the rapid codbustion of a large accumulation. These three concepts do not require major new structures, but dc require a significant

() amount of equipment. Since our evaluation indici ;es that (for events .

where containment cooling is available) the plant can withstand the releases of energy and noncendensibles (except the energy of hydrogen burning rapidly) for events up to core melt, these concepts can provide substantial protection for noncore melt accidents. These concepts do not provide any protection against overpressure failure due to the steam and gas releases during a core melt.

10 3.1 Containment Inerting with Nitrogen Effectiveness - No significant sources of oxygen into containment during operation have been identified (except the compressed air system which would be modified to eliminate this potential source).

Therefore, inerting should be fully effective at preventing postaccident hydrogen combustion. The additional partial pressure of hydrogen could lead to containment pressures above design, but still well within the capability of the containment. The increased containment pressure could lead to some increase in leakage, but the secondary containment features at Sequoyah would minimize the impact.

Technical Feasibility - Based on our experience at Browns Ferry, this concept is feasible, but technical problems exist. The compartmentalization at Sequoyah will require extensive measures to eliminate trapped pockets of oxygen. Of more concern is the problem of reduced access for inspection and maintenance. ,

("% .

(_) Additional Risk - An inerted ice condenser containment cpuses a major additional risk to the operator. Ice condenser containments require a significant increase in containment entries over other types of containments. This increase is due to both additional surveillance required by technical specifications and additional inspection and maintenance on the ice condenser system components. Technical specifications require weekly visual inspections, while current experience at Sequoyah and other ice condenser plants indicates that containment entries are required every two days to perform maintenance on the ice condenser systems, particularly the air handling units.

Based on this high rate of maintenance, entries are required daily to inspect these units. .a addition, ice weighin6 activities are expected to require several weeks per year with a crew of people inside all day. Even assuming tha'e all possible plant modifications are made to reduce access requirements for maintenance, it is estimated that personnel entries would still be required at least 16 times per year with an associated forced downtime of 15 percent.

Industrial experience has shown that personnel occupancy in inerted enclosures can be extremely hazardous. Even while deinerted, there is a risk of a pocket of nitrogen existing that could threaten workers.

In addition, the reduced access for inspection and maintenance will ,

reduce the reliability of other plant systems. Experience has l indicated ,that other (nonice condenser) maintenance will require 10-2 i

l

centainment entry on a weekly basis. This includes instru=3nt, pump, and vcive repairs. Reduced maintenance could increase the possibility or severity of accidents. Often, it is desirable to enter the containment for manual valve operations in the event retote operation 5 fails.- Reduced access will inhibit such repair activities in the

%) event of nonpipe break accidents. This additional risk appears significant.

Reliability - Since inerting wodd take place prior to going to power, after which the system is passive and the status of containment can be monitored, this concept is judged to be very reliable.

Cost' - While the capital costs are moderate, the,, cperational costs are very high due to the cost of large quantities of nitrogen, the additional downtime for inerting/deinerting, and the forced outages for increased maintenance in containment. All costs are in 1c80 dollars and for a 2-unit plant.

,CapitalCbst -

Nitrogen Cost - /yr.

Inert /Deinert Loss of Availability - /yr. (minimum)

Additional downtime would be required to perform mintenance that cannot be dcne in an inerted containment. Also, the periodic inerting and deinerting operations would increase the rate of ice loss in the ice condenser and force its earlier replacement with concurrent downtime of several months. Neither of these operating costs have been included in the estimate above but could be substantial.

.O 10 3.2 Halon Effectiveness - There is fairly good evidence that Halen 1301 can suppress hydrogen combustion. However, there is some question about its long term behavior in containment. If it decomposes or is removed by the containment sprays or ice, combustion could still take place, probably at a high hydrogen concentration.

These questions must be answered before it can be concluded that halon would be effective. (Also, see Reliability below).

Technical Feasibility - The system is very similar to installed fire suppression systems and should not cause significant technical problems except in scaling up to such a large volume with the need for uniform injection and good mixing. Since steps have to be taken to prevent inadvertent operation (which could pose a life safety and containment pressure problem), and since the operator mast have sufficient warning of a potential hydrogen release to actuate the system manually, the controls may be complicated. It is thought to be feasible.

Additional Risk - Halon can be a hazard to personnel. We judge this hazard to be smil since the probability of inadvertent actuation can be kept low.

(ym .

10-3 ,

Halon itself can causs som3 risk of containment' failure. Wa believe that the risk of containment overcooling during normal operation can be minimized. If halon were injected after a hydregen burn or when high pressure exists in containment, the additional partial pressure (even accounting for the cooling r~ - effects of halon) could cause failure due to overpressdre

%.s In addition,' the decomposition products of halon include certain acids, halogens, and other substances which have the potential for inducing cracking in stainless steel, degrading other equipment caterials, and complicating any cleanup cperations. This may ulti=ately prove to be so severe as to preclude its use.

Reliability - The system is very similar to installed fire suppression systems and should be reliable mechanically. However, since the system must be actuated by the operator, there is seme risk that the operator will fail to recognize the need' to actuate

. the system.

Cost - The principal costs are due to the halon, mechanical equipment, controls, and instrumentation.

Capital Cost -

10.3 3 Ignition sources .

Effectiveness - Ignition sources are effective only for those events where the hydrogen evolution and rate of burn will be slow s;a

~

the order of the heat removal capability of the containment sprays).

Such conditions occur when the hydrogen concentration in C,T) containment is between 4 and 8 pc. cent. Ignition would be most effective in the lower co=partment because of the availability of heat removal in the ice condenser. However, it is not certain that reliable burning can be achieved in the lower 'co=partment due to the presence of steam in large quantities and the relative lack of air. ' Above a hydrogen concentration of 8 percent, burnh occurs very rapidly, producing energy at a rate whi,ch::cannot be removed quickly by the containment cooling ~ systems -ifra large quantity of hydrogen is available.

Technical Feasibility - We believe that the igniters can be

. developed although considerable research and testing will be needed. Further analysis is required to establish how many are needed and where to locate them. Controlled ignition of the effluent from the reactor vessel head vent prior to mixing in tb -

containment may require a more sophisticated system than in open areas, but may be feasible.

Additional Risk- - The igniters could cause combustion that otherwise might take place. We believe this risk is scall since there are already mechanisms in containment which will probably cause ignition sooner or later. Local effects of controlled

+ burning could be severe.

. 10-4 e

h _.

9 F.eliability - At least some degree of operator action is required. It appears that oc= plex guidelines will be required

-for which igniters to use, when to use them, and when not to use them. The chance of operator error is therefore significant.

(h k/

Cost - The major cost items are hardware, controls, and instrumentation.

Capital Cost -

10.4 Conceots to Increase Containment Capacity for Overpressure Events The four concepts evaluated are filtered vented containment (relief through filters to atmosphere), additional containment (an added containment building),- coupled containment (connecting both containments at a 2-unit site), and additional containment ecoling (heat exchangers to cool the containment air). The first three provide protection against containment overpressure by relieving some of the gases to the environment or another structure.

The last providea protection by serving as a backup to the containmer' sprays and by removing steam to allow for more noncondensible gases.

All involve major construction and design efforts. The first three provide limited protection from the effects of hydrogen by reducing the hydrogen in containment and providing relief for slow to moderate pressure in, creases. ,

10.4.1 Filtered Vented Containment Effectiveness - Since venting is to the environment, this concept O'

s provides unlimited total relief . capacity. However, its effectiveness is limited for rapid pressure transients because of practical vent sizing which limits the rate of relief. The concept does not provide complete mitigation since deliberate releases of radionuclides are used to prevent against accidental releases due to containment failure. In fact, the release of noble gases could be nearly as severe as for ccntainment failure; the release of halogens and particulates can be reduced by about a factor of 100.

Effectiveness as a mitigator of hydrogen is limited to: those l events that are relatively slow such that operator action can take place before any major pressure increase (to reduce the amount ,

of gases inside containment); and those events that result in slow i to moderate pressure increases.

It is believed the effectiveness in practice would be reduced due  ;

to a' natural reluctance by the operator to use a system which involves  !

a deliberate radioactive release of large magnitude. (A relief duration of only a few seconds could lead to releases many times that from the worst event presently analyzed for the design basis.)

Technical Feasibility - Overall, it is felt that the system could be desi5ned and constructed to work. There are several areas where research and/or development are required, including-1

1. The connection to containment is a difficult design problem. l

(~[)

s- The princip'al concerns are modifying an existing design to l 1

l

. 10-5 -

minimi:s loads on containment and ensuring the integrity of the relier line.

2. The choice and design of the heat recoval facility (pool or

'O sand / gravel bed) will require extensive analysis and possibly some research. Concerns include structural provisions for dynamic loads, draining of the bed, and long term behavior.

3 If hydrogen and oxygen mixtures mast be handled, a hydrogen burning cha=ber or dilution system must be provided. It is not known at this time if such syste=s can be designed to function reliably.

4. The effectiveness of various types of particulate and iodine removal filters when subjected to large, high tecperature, high

. moisture content flows should be demonstrated. ,

, Additional Risks - This concept involves deliberate releases; therefore, it only reduces,rather than eliminates risk. It is estimated that the whole body dose would be in excess of 900 REM in the low population zone compared to the Sequoyah FSAR LOCA whole body dose of 2.4 REM in the LPZ. In addition, venting could result in undesirable doses to the operators in the control room. Also, the system can lead to violation of containment by inadvertent operation when not needed. There is some possibility that if the system is used for less severe events in conjunction with containment sprays, the ccatainment pressure could go sufficiently subatmospheric to threaten the cortainment structurally or the core cooling systems due to inadequace NPSH. ~

(1-) Reliability - The mechanical reliability is high because the system ,

is essentially passive. Some designs employ active filtration, heat removal, or hydrogen control co=ponents, though. However, operator action is required to control the system. Considering the impact of using the system (deliberate releases) and the short amount of ~

ti=c in some postulated events for making the decisions, it is believed that the chance of misoperation or_ delayed operation could be significant.

, Cost - The cost is largely in' constructing the facility with some maintenance cost (not estimated).

. Capital Cost - _

10.4.2 Additional Containment -

Effectivess - Since venting is to an additional containment, th'is concept provider a large but finite total relief capacity. Its effectiveness is limited for rapid pressure transients because of practical vent limits (but larger venting rates than for filtered vented contain=ent can be handled). The concept does provide co=plete mitigation for those events within its capacity since no deliberate release to the environment occurs'. ,

Effectiveness as a hydrogen citigator is limited similarly as the lo-6 ,

I

1 filtered vented system to those events allowing operator prealignment of isolation valves and involving slow to moderate pressure increases.

(It may be possible, however, to_ make the isolation valves automatic.)

s Technical Feasibility - This concept is also believed to be feasible

% and does not involve the technical problems of filtration. However, the problems of containment / vent structural attachment and the adequate treatment of the hydrogen in the effluent retain.

Additional Risks - No additional risks were identified.

Reliability - The mechanical reliability is high because the isolation valves are the only active components required. However, operator action is required to actuate the valves..which introduces come chance of delayed operation.

T

_ Cost - The cost is in constructing the vent system and the additional containment structure.

. Capital Cost -

This cost is for a 1.5 million cubic foot containment. A containment with more capacity (3:0 million cubic feet) may be required for mitigating all core melt pressure effects. It would cost ,

apprcximately .

10.4.3 couoled containment Effectiveness - Since venting is to the other reactor containment q

5-building, this concept provides a finite total relief capacity, even though some benefit may be obtained from the other unit's heat removal and pressure suppression systems. Again, the effectiveness is limited for rapid pressure transients because of vent size limitations. The coupled concept does provide complete citigation for those events within its capacity since no deliberate environmental release occurs.

Effectiveness for hydrogen mitigation is also limited to those events allowing anticipatory operator actuation of-isolation valves and t involving slow to moderate pressure increases.

Technical Feasibility - The coupled concept is believed to be feasible, but still faces the structural problem of containment / vent attachment without allowing excessive accident loads to be transmitted between units. -

Additional Risks - In principle, venting between containments should not cause proble=s in the second unit that do not potentially already exist in the first one. However, some equipment failure or misoperation in the second unit could result due to the environmental effects of the effluent from the first. This is especially likely due to the high temperatures that could result from hydrogen

? .corbustion. In any case, the ' detrimental effects of radioactive centamination and reduction in accessibility, need for cleanup, etc. ,

would be doubled.

~

U.

10-7 ,

e.

Reliability - The mechanical reliability should be high since the isolation valves are the only active cc ponents. However, core valves are involved than for venting to an additional containment and anticipatory operator action from both units is required to actuate

(~T the valves. In addition, the significant impact of using the system

(~/ (contaminating the other unit) suggests that the chance of misoperation or delayed operation could be significant.

Cost - n'hile the cost in constructing the coupled vent system is moderate, the operational cost would be very high due to the additional downtime that would be forced. The units would have to be operated in conjunct $on since one could not be at power while the containment of the other was open because of maintenance, refueling, or initial construction. This could al=ost double the expected unit downtime as well as greatly co: plicating outage planning and personnel

. requirements.

. Capital Cost -

10.4.4 Augmented Cooling Effectiveness - The practical short term heat removal capacity of air coolers or additional containment sprays is limited to the coderate energy addition rates frcm partial or intermittent hydrogen 1 burns. The air coolers or current sprays are ineffective in removing energy from rapid combustion.

Technical Feasibility - This concept is technically feasible since air cooling and sprays ~ are cur rently used for containment heat

(~T s/ removal, either normally or postaccident.

Additional Risks - No additional risks were identified.

Reliability - The mechanical reliability is fairly high. Operator actuation would not be required bsaause the systems could Laitiate automatically on accident signals since no additional risks are involved. -

. Cost - The cost is in the initial installation of the equipment.

Capital Cost -

10.5 Concepts to Mitigate the Combined Effects of Hydrogen Combustion and Containment Overpressure Concepts have been discussed in this section which could alleviate some aspects of core damage. Some concepts would prevent the failure of containcent by clicinating the rapid energy released with hydrogen burns. Others atte=pt to prevent the failure of containment due to overpressure by removing mass and energy in large amounts. No single concept studied was found capable rf preventing containment failure due tc > ith techanisms. Thus, cc=binations of the concepts were

+ evalu-t cd for mitigating the main effects of core melt.

O .

4 10 8 ,

---

v-

A review of the individual concepts show that venting to an additional containment in combination with either halon injection'or containment inerting with nitrogen may-provide co=plete protection. Each of these '

ce=binations involves =ajor construction, extre=ely high cost, and

=ay drastically affect operations.

~Effcetiveness - When co=bined with venting to an additional containment, either halon injection or inerting with nitrogen appears to be capable of preventing' containment failure due to either hydrogen combustion or overpressure. However, the additional ecntainment would also have to be inerted or provided with a halon injection system, depending on the option selected for the original containment. This is because venting from a halon-injected or an inerted containment into an sir-filled additional containment could alter the overall gas concentrations enough te make co=bustible mixtures possible.

i Technical Feasibility - There is nothing inherent in either of the co=binations which would preclude the use of the concepts in a cotbination. However, each of the concepts has serious technical problems which must be solved before they could be used, individually or in co=bination. '

Additional Risks - There are no additional risks by using these concepts in combination beyond those identified for each separately.

Reliability - The combination with nitrogen inerting should be as reliable as the individual venting to an additional containment since the inerting is judged to be very reliable. However, the combination of halon injection and venting would not be as reliable since two separate active syste=s, each actuated manually by the operator, would

(} be required to operate. .

Cost - While the capital costs for both options are similar >

( for inerting/ venting and for halon/ venting),

the opecational costs of inerting are very high as described in -

section 10.3.1.5. ,

~

6 l

l l

NSAHY6.PB O

~

10-9 ,

h.. .

11.0 Conclusions and Recommendations 11.1 Conclusions

() 11.1.1 Theco studies have shown that the overall risk to the public from the Sequcyah plant is about the same as the KRC reference plant which has a dry containment and, thus, Sequoyah presents no special risk to the public.

11.1.2 The Sequoyah plant containment was found capabic of withstanding a hydrogen burn from an amount of hydrogen generated in the oxidation of approximately 20 percent of the zircaloy fuel cladding without any additional mitigation.

11.1 3 All concepts studied in this report were found to have serious technical and/or operational problems which need extensive research and development before design work could proceed.

11.1.4 Containment inerting was found to be the most reliable of those cencepts which control or prevent hydrogen, but it is extremely expensive, would likely cause personnel deaths, and may lead to reduced maintenance and safety. The halon suppression concept has few of these problems, but has serious technical concerns with its actual perfor=ance in a postaccident containment environment. An ignition system also has some technical qu?st ions requiring resolution and is not effective for complete reaction 'T ;he zircaloy fuel cladding.

11.1.5 s Of those concepts which prevent containment overpre'sure,

(') venting to an additional containment appears to be the best. Although it is more expensive th3r. filtered venting, it is simpler and releases no radioactivity to the public. The major drawback of the filtered vented containment is a large dose to the public. The coupled containment concept is too restrictive on operation.

11.1.6 To mitigate the conseqences of an accident resulting in 100 percent circaloy clad reaction with water or one that proceeds to complete core damage, two or more of the~ concepts described in this study must be used, at least one each from hydrogen control and one from overpressure prevention.

11.2 Recommendations 11.2.1 These studies have shown that (1) Sequoyah is comparable to the best plants operating in the U.S. in terms of risk to the public, (2) Sequoyah containment can withstand a burn of hydrogen equivalent to 20-percent metal-water reution, or about the same as the NRC estimate of the hydrogen burned at Three Mile Island, and (3) all concepts of mitigation of hydrogen effects studies clearly need substantially more research and development. Based on these conclusions, we recommend that no commitments to specific mitigation devices be made at this time. Further investigation and development of the most promising concepts identified in this study or which may be proposed are recommended.

fT U

11-1 i i

-- _ - . . _. ~ _ _ -

11.2.2 In the event that a commitment to incorporate some mitigation of hydrogen effects becomes necessary, the following approach is r' commended:

e Commit to an ignition system - Although this system is not .

effective for 100-percent reaction of the zircaloy fuel cladding and does notireduce the overall risk significantly, it may handle hydregen releases up to 60 percent metal-water reaction, which is the upper range of the estimates of hydrogen generated at Three Mile Island ('iMI). This system has the advantages of short implementation time folloving further developmuntal work, low risk of jeopardizing current safety, relatively low cost, no effect on operation, and can be treated as a temporary measure until the NRC has gone to rulemaking on hydrogen.

11.2 3 In the event that a commitment to incorporate mitigation of hydrogen effects for 100-percent zircaloy cladding-water reaction and/or complete core degradation becomes necessary, the following approach is recommended:

The best mitigation requires the combination of venting with prevention of the hydrogen burn. We recommend that this option be committed to, only if no choice is given. The best concepts when considering effectiveness, life safety, operational problems, costs, and improvement of risk to the public is the l halon suppression system plus venting to an additional containment; however, the halon system has serious technical problems which may not be resolvable before design must,begin.

Therefore, at this time, nitrogen inerting plus venting to an O additional containment mast be recommended. It is reliable, but has very serious operational problems, is extremely expensive, and poses some safety problems. ,

11.2.4 As a result of our findings, we recommend that TVA discourage NRC from use of the filtered venting concept as the requirement of the rulemaking process because of the large radiation dose to the public, which may be in excess of 900 REM body dose in the low population zone a* Sequoyah.

E50243.01 a

11-2 S

// -)

APPE!! DIX A O CURREllT DESIGN AND BASES A.1 The Reactor Core, Fuel, and Coolant System The react,or core is comprised of an array of fuel assemblies which are identical in mechanical design, but different in fuel enrichment.

The core is cooled and moderated by light water at a pressure of 2250 psia in the reactor coolant system. The moderator coolant contains boron which is varied as required to control relatively slow reactivity changes including the effects of fuel burnup. The pri=ary barrier to radiation release is composed of the individual power producing fuel rods in the core. Two hundred and sixty-four fuel rods are mechanically joined in a square array to form a fuel assembly. The fuel rods are supported in intervals along with their length by grid assemblies whir maintain the lateral spacing between the rods throughout the desigr. 'fe of the assembly. The fuel rods consist of uranium dioxide cera pellets contained in slightly cold worked Zircaloy-4 tubing which is 'tgged and seal welded at the ends to enapsulate the fuel. The Zircat ,..-4 is primarily composed of circonium metal. A schematic of the fuel rod is shown in figure A-1.

The fuel pellets are right circular cylinders consisting of slightly enriched uranium dioxide powder which has been compacted by cold pressing and then sintered to the required density. The ends of each O pellet are dished slightly to allow greater axial expansion at the center of the pellets. Power is produced volumetrically by the fission reaction within individual uranium pellets. Thermal energy is transferred across the gap to the cladding that serves as the heat i transfer surface to the primary coolant. To avoid overstressing of the clad or seal welds, void volume and clearances are provided l within the rods to accommodate fission gases released from the fuel, differential thermal expansion between the clad and the fuel, and ,

l fuel density changes during burnup. At assembly, the pellets are stacked in the clad to the required fuel height, a spring is then inserted into the top end of the fuel tube, and the end plugs pressed into the ends of the tube and welded. All fuel rods are internally pressurized with helium during the welding process in order to minimize compressive clad stresses and creep due to coolant operating pressures. The fuel rods retain all fission gases and radioactive fission products unless the cladding is breached by mechanisms present during an accident. One function of the reactor coolant system is to ensure the cladding remains covered and cooled to provent clad melt during normal and accident events.

The reactor coolant system shown in figure A-2 consists of four similar heat transfer loops connected in parallel to the reactor pressure vessel. Each loop contains a reactor coolant pump, steam Generator, and associated piping and valves. In addition, the system includes a pressurizer, a pressurizer relief tank, interconnecting piping, and. instrumentation necessary for operational control. All

(]

v the above components are located in the containment building.

A-1

• i FIGUPS A-1 o

[ \

END PLUG l t/ -

( i

\

\

N l l -

\ l N l . N

,' \ '

}7- S PR I NG s s -

N N N \

\ /

6.3" p , e (TYP) - s'-

N

\

( -

\ \

\ N ir \ N n 002PELLETS

\ p \

\

\

\ s 151.6" \ \

(REF) ,qq,g, \

[

(TYP) N >

+

s x- e

\

l t' / \ *

\ g N ZlRCALOY CLAD s' '

\ I

\

\

(\ - 1

\ l N o \ \

N / \ \\

s N \

\ ) -

w -

9 .]~ I L.

- l I l SPECIFIC DIMENSIONS DEPEND ON DESIGN VARI ABLES SUCH AS PRE-PR ESSUR IZ Ail 0N. P0'wER HISTORY. AND DISCHARGE BURNUP 1

l A-2

• e-o - -

e , - -- - . --

D i'

~.

r 5

6 .,

3.

Id 2 a l- 'I

• M

> tlky.

Da 0"

< i i

.N-".-

s ta

.J y -

-A7L,w; .- ,

oa

) O ' ' %. > a - ,-.-y'h C8 ab, .

W.-y r* O' O C%,

W o a w

O4 'y to tu MC v.

u ,

W w

b

~

w yw.

pmm}rg,= -=-- _-m _.;m .r a- c, m.=.s.gx.g s .. ,,. m m ._ ,. .

.G u_t.>.~. . .:.- v. . . . . ,. - , _t

• 4'- = , y s~- n I

d. 'h.~(1.1 -

T- '

--N' *

• v .i. - - - .

l c.

%:); ' *~LT15.?i:Cu m .wn u,1Li L~ -a ;~.1-os}>~A.3 v.i _=_c i ,

ct'c~ cm tc n

.~

rf ru. :r-9 cu 3 <C3)~ W';\ Q1 ~l*G5. k'Yit;'C) T*~'--- V '

o

_.4--.m.._tM - ; f. , - - - _n ,- - . U."~.'

,;...w <

u y " -1.,, " -- : s'==

..-:::: J p;.._ . .. w

.g='._~....._..'7-

e. '

.~ , . . . . .

g L_;

y lf 3 h y

4 m'* ,

^~

J~~~~, R s- as

• e lw g  %

., 6, d -- - -l ,3 = ,s s

n--m . .

,.m . . .A; jg jyy

.w f

C3 Y0 hy, -

Q .y~~s Y l V - w a 11 Cf s T- --

U M A.11 E  % e' g T l s

- 1 e .

B e d.

-3 p

.$ a f_es. s .- ,

,i pn.W .,. ===.:n a

1=-- =~w~, m a -

l ,'./ ( % i l  ;

l' M -- j

[' <

-- a1:

1 F _ l bd A U}l Mh__y7A fM -; p (

,_a p7- m m m[ _,"m m e . N P- l 7 ;p,p... . a 4 _. s . , . .e _ . -

a d l "

I

. 9 1 5

'r

=, 'ob

• l g

., lj

l

-l, b '

8 8 a .

5

• r'.  ;

) 8 l

.._/

s

• l A-3 '

l

During operation, the RCS transfers the heat generated in the core to the steam generators where steam is produced to drive the turbine Borated decineralized light water is circulated in the

() generator.

RCS at a flow rate and temperature consistent with achieving the design reactor core thermal-hydraulic performance.

The water also I

acts as a neutron moderator and reflector and as a solvent for the I neutron absorber used in chemical shim control.

The RCS pressure boundary provides a barrier against the release of radioactivity generated within the reactor and is designed to ensure Its a high degree of integrity throughout the life of the pl. ant. ,

extent is defined as:

1. The reactor vessel including control rod drive mechanism housings.

2. The reactor coolant side of the steam generators.

3 Reactor coolant Lamps.

4. A pressurizer attached to one of the reactor coolant loops.

5 Safety and relief valves.

6. The interconnecting piping, valves, and fittings between the principle components listed above.

7 The piping, fittings, and valves leading to connecting auxiliary O or support systems up to and including the containcent isolation valves.

RCS pressure is controlled by the use of the pressurizer where water and steam are maintained in equilibrium by electrical heaters and water sprays. Steam can be formed (by the heaters) or condensed (by the pressurizer spray) to minimize pressure variations due to contraction and expansion of the reactor coolant. Three spring-loaded safety valves and two power-operated relfer valves are mounted on the pressurizer and discharge to a relief tank, where the steam is condensed and cooled by mixing with water. These valves ensure the integrity of the RCS during overpressure events. For accidents where the pressure boundary integrity is lost, emergency core cooling This systems are used to maintain co_re cooling within the RCS.

entails removal of both stored and fission product decay heat from the reactor core. Systems employed in this function include: cold leg infection accumulators; upper head injection accumulators; residual hett removal pumps and heat exchangers; centrifugal charging pumps; safety injection pumps; and indirectly the auxiliary feedwater system which maintains steam generator secondary inventory.

A.2 Containment The contalument is divided into three compartments: The lower compartment; the upper compartment; and the ice condenser compartment. The lower compartment completely encloses the reactor

{J A-4

d 5

coolant system equipment. The upper compartment contains the refueling canal, refueling equipment, and the polar crane used during refueling and maintenance operations. The upper and lower compartments are separated by the operating deck, which provides a low-leakage barrier between these two compartments. The ice condenser connects the lower compartment to the upper.

The ice condenser concept utilizes a large mass of ice to condense escaping high-energy steam from postulated loss-of-coolant accidents

- (LOCA) or steam line break accidents. The rapid condensation of steam in the ice bed keeps the maximum containment pressure relatively low while maintaining the capacity to absorb a continuing high energy input from the reactor core and reactor coolant systems. Tgeice condenser is made up of 24 individual bays which form a 300 arc inside containment. Each bay consists of three major sections: A lower plenum: a'n ice bed; and an upper plenum (figure A-3). The lower 7

!- plenum is ieciated from the lower compartment by doors in each bay that open at a differential pressure of 0.007 psi. The ice bed contains a minimum of 2.45 million pounds of ice. The ice is stacked in columns one foot in diameter and 48 feet high. The upper plenum a contains cooling units used to maintain the low ice bed temperature

, during normal plant operations. The upper plenum is separated from i

the ice bed and the upper compartment by two sets of doors that will open with a differential pressure of 0.028 psi.

In the event of a LOCA, steam pressurizes the lower compartment which opens the lower inlet doors. An air-steam mixture enters the ice bed where all the steam is condensed. The rising air then causes the top two setc of doors to open and then flows into the upper compartment. To provide maximum use of the ice bed, air return fans

are provided which circulate containment atmosphere through the ice i bed condensing steam released after the initial pressurization. When all the ice has melted, the spray system located in the upper

compartment removes the remaining energy released te the containment.

, A.3 Design Bases of Containment _

The ice condenser is designed to limit the containment pressure below the design pressure for all reactor coolant pipe break sizes up to i

and including a double-ended pipe severance of the largest main reactor coolant pipe. Characterising performance of the ice condenser requires consideration of the rate of addition of mass and energy to the containment as well as the total amounts of mass and energy added. Analyses have shown that the accident which produces the

highest blowdown rate into an ice condenser containment will result

! in the maximum containment pressure rise; that accident currently I

is the double-ended severance of one of the main reactor coolant

pipes. The design basis accident is therefore defined to be the double-ended reactor coolant loop pipe break. Postblowdown energy releases are also accommodated without exceeding containment design pressure.

A.3 1 Mass and Energy Releases O-A-5 ' ,

,-c , , , ,,,,-e-. .--e. e , ,. +,n , , - - , , - . . , ~ - . , . ,-.~,n.-- , ,,-.c, ..,., -, . , -

Ice FIGy33 g,_3 conganggy ,

O O

5 y

}

b1&D .-

. .v fj .

% 5SM$.%-=

[ 4 W%Q'f .

AIR HAN3L[yg UNIT

[

@ Wut y p'N%gw m y* '

i CRANE TERMlogggg

-- g Y AlR DIsiglBUTION

• py f.;*. ,

f CRANE gggy I,.'*  :

N,

%l i 3 k ,

[$5,k *.sl '

~~-~ %Ng ll STEAg l

% C, *:

.'s GENERATOR LATTICE Fgguz. ENCLOSU.E t y i.y..

g's a 2s Q s ;-l -: '

Q~

! f5'

'1

[ f ';. Y!!.*:4?.S '

%. ,jhI' ..

SHELL ~

j ,,- *

,l ~

Q

' I

_;_.% ' l{

5'*

WALL P44ELg [Il n - i

<ss }Y

~' l 'k STEAM GENEggggg

• hY I _.., -

yf 1

s I s~ **.

fj{,

'EU"'

v t*i Qs*s%g N s

,-l-l>Q\

/ .

/

T.' /

Q r

^ DIVIDER DECg 1

f;l

• .s: .

/l.

LOWER INLET 00cgg T,.

/

/ LOWER R>*.Ij!'a:/ SUprogy STRUCTURE FLOOg rm .

V -

l A-6 l

Following a postulated rupture of the Reactor Coolant System (RCS), i steam and water are released into the containment. Initially,

• r" the water in the RCS is subcooled at a high pressure. When the break occurs, the water par 1 through the break where a portion flashes to steam at the lowc A essure of the containment. These releases .

continue until the hCd danressurizes to the pressure in the containment (end of blowcown). At that time, the vessel is refilled by water from the ,ccumulators and Safety Injection (SI) pumps. The analysis assu' .aat the lower plenum is filled with saturated water at the end of blowdown to maximize steam releases to the containment.

Therefore, the water flowing from the aco"-ulators and SI pumps start to fill the downcomer causing a driving heao across the vessel which forces water into the hot core.

. During the reflood phase of the accident, water enters the core where a portion is converted to steam which entrains an amount of water into the hot legs at a high velocity. Water continues to enter the core and release the stored energy of the fuel and clad as the mixture height in the core increases. When the level, two feet below the

top of the core, is reached, the core is assumed to be totally

quenched, which leaves only decay heat to generate steam. This type of break is analyzed at three locations.

The location of the break can significantly change the reflood transient. It is for this reason that the hot leg, pump suction, and cold leg break locctions are analyzed. For a cold leg break, all of the fluid which le ves the core must vent through a steam generator and becomes superheated. However, relative to breaks at

() other locations, the core flooding rate (and therefore the rate of fluid leaving the core) is low because all the core vent paths include I the resistance of the reactor coolant p p. For a hot leg pipe break, the vent path resistance is relatively low, which results in a n'sh

, core flooding rate, but the majority of the fluid which exits the core bypasses the steam generators in venting to the containment.

The pump suction break combines the effects of the relatively high t

core flooding rate, as in the hot leg break, and steam generator deat addition as in the cold leg break. As a result, the pump suction breaks yield the highest energy flow rates during the postblowdown period. The spectrum of breaks analyzed includes the largest cold and hot leg breaks, reactor inlet and outlet, respectively, and range of pump suction breaks. Because of the phenomena of reflood as

discussed above, the pump suction break location is the worst case.

l This conclusion is supported by studies of smaller hot leg breaks which have been shown, on similar plants, to be less severe than the double ended hot leg. Cold leg breaks, however, are lower both in the blowdown peak and in the reflood pressure rise. Thus, an analysis of smaller pump suction breaks is representative of the spectrum of break sizes.-

The LOCA analysis calculational model is typically divided into three phases: (1) blowdown, which includes the period from accident occurrence (when the reactor is at steady state full power operation)

' to the time when zero break flow is first calculated; (2) refill,

.which is 'from the end of blowdown to the time the ECCS fills the

( Vessel lower plenum; and (3) reflood, which begins when water starts A-7 .

-e ~r -- - 4,_ .,= m- r--.c,-, - - --- y .- .- ..-,--%.-.--:.ye . - ar-- -

,-.pw. s-- ,-w. - .-

moving into the core and continues until the end of the transient.

For the pump suction break, consideration is given to a possible

! fourth phase; that is, froth boiling in the steam generator tubes gs after the core has been quenched. (See reference A-1 for a

! (_) description of the calculational model used for the mass and energy release analysis.)

A.3 2 Noncondensible Gas Procuction Another source of mass that must be considered in containment design i is the production of noncondensible flammable gases. The more severe loss-of-coolant accidents ;'stt0ated for plant design may result in the production of excess hydregen which ultimately is released to the containment air space. Depending on the final volumetric concentration of the hydrogen and the quantity of oxygen present, the gas can be a fire hazard, an explosion hazard, or too diffuse to chemically react.

The corrosion of metals in the presence of water at temperatures characteristic of light water reactors has been recognized for many years and many studies of the phenomena have been conducted.

Fortunatley, the oxidation rates at operating temperatures are extremely low and do not pose a problem for the design life of the j fuel assemblies or the reactor vessel. At elevated temperatures encountered during mitigated accidents, the corrosion rates become moderate and some damage to the fuel cladding is expected. Baker and Just defined an upper bound on the oxidation expected for the Zircaloy fuel cladding used in light water reactor systems (A-2).

() At temperatures below 2200 F, the oxidation does not degrade the cladding fission product barrier. For this reason and others, all j power reactor designs use this limit as one of the criteria for the effectiveness of the emergency core cooling sgstems. For excessively

high temperatures expected at clad melt (4000 F and above) the reaction rates for Zircaloy and water become large and,'in addition, oxidation of other metals becomes important. Since the vessel is cceprised of stainless and carbon steels, at melting temperatures these can account for thousands of pounds of hydrogen. Reaction equations for various metals (in the Arrhenf.us equation form) are shown in Table A-1. At Sequoyah, the total rirconium present for oxidation is approximately 43,204 pounds which, if oxidized, completely produces about 2000 pounds of hydrogen (figure A-4) .

The concentration by volume percent in the Sequoyah containment based on various clad reaction percentages is shown in tigure A-5. The i exothermic oxidation reaction alone accounts for 2805 Btu's per pound 4

of zirconium oxidized. Only 18 percent reaction can be accommodated 3 by dilution in the containment before the flammable limit of 4 percent by volume is reached. However, considerable oxidation must occur l before the detonation limit of 18 percent is reached. For each pound l of gircaloy oxidized, 7.88 cubic feet of hydrezen at STP is released (32 F, 14.7 paia).

Another source of hydrogen is the corrosion of zine primers and .

aluminum metal inside the containment. These sources along with sump radiolysis are slow in evolution and are important in design for long A8 '

,,.~_,-c. .-

_s u,y f- *'

l ?, > ep hg ) 4 a ,-; ,

4 ,

,e

- --- - , - - - - , -- + - - , - , , . , , ,m-na--~ , - -,n. v.---

TAB'E A-1 U, (See references 6-7 and A-4)

METAL-WATER CORROSION RATE PARAMETERS Metal A 6

B Zirconium" 333XIg2 -45,500 Stainless Steel ** 2.4 X 10 -84,300 UO2 *** 1.65 X 10 5 -50,800 W = A t exp (B/RT) where t = time (sec)

R = gas constant 1 987 cal / mole K T = temperature K 2

• W = weight (cg) zirconium rengted per em area

• • W = oxygen uptake (mg) per em grea

• • • W = oxygen absorbed (mg) per em area O(.-

N, ]

l A-9 r .

- r l

Dequagal7 iBgbrugen Stung o.

p o

, . . . . . . o

. o.

. ........ .... ...... .... ...... ........... ......... . . ......... ........... ..... ... . ........... ........... o o

......... _o o

H  :.  :.

Z Z

H  :  :  :

:  :  : o o O ..........:...........:...........:..........:...........:...........:...........:.....

o

......... _o c- -

H m .

A o vM i ,,  ! i i i

~

>tcn w gg54 -

O

~

!  !  !  ! k;

!a* 8 o o"a op M

c

t Q  :

u p d

ce, m

Ed O

O i i i i i i o W

~

4 Z i i  !  !  !  ! $oo N ZZ N

!  !  !  ! . i .

oZ

_o Cs .

o m_, .

M<

s  :.  :

~  :  :  :.  :  :.  :  :  : Ed w4

. c. o

.........:...........:..........:...........:..........:...........:.........:...........:...........:........... .o \ -

N O

va

. 04

............. ................................. on s .

H. .

o.

.......... . ............ _o w

. . 1 o.

........................ .......... _o-

- o.

V

) $ i b 4 $ $ $ $

'k V 0 0002 0'009T O'000T O'00tT 0'00M 0000T O'000 0'009 0'00F 0'003 0'0 (41) NHOOHCL1H A-10 f7

.~.

Dequngal7 Eighragen Stuby o.

o

. . . . . . O g

. o.

...... .... ........... e o.

.........1............:...........;.......,..:..........:...........:..........1...... ......... -o o

F.- .

---4 .

Z

• :  :  :  :  :  :  : o c!

.......... ........... ........... ........ . ........... ........... ........... ..... o ......... -o

.  :.  :.  :.  :  :. >-4 6

. . ,. m _ m

:  :  :  :  : A o g C i. i i

i i

. i. i. *-*

%f u) m e

v o y) .

o _o 1............:...........:..... .....:. .........:...........:..... o

....t.... *r ......... C3 A-oH C.-4 O

p4  :  :  :  :  :  : o n.

F L.4  !.  !. .!  !.  ! i i

O g

MM .

<d

:  :  : .  :  : O o I_3 i i i  ! i i N ZZ ~ o C'4

............ ...........i........................

_.C H  :  :.  :  :.  :.  :  :.  :.  :.

in

. . . . . . . s

. . A 6- .

. s

o. ,

-.1 .

m, zy

--4 .

4 .

H  :  :  :  :  :  :  :  :  : ~

._ o

o. M

_4 ..........

O s . . . . . . .

. C.

-N O.

............. . . . . . . . . . . . , .. ......... . ........... -o H

. - o.

o O ooc 'os os os os 'os os os os 'W oo

(% 10A) NOIJ,VELN20N00 "H A-11

- . - . . - - - . . - - - .. w.-

c - - - _ _ _ . - - . - . _ . - - -

term control. The corrosion reaction is similar to that discussed

above; however, it occurs at a much lower temperature characteric of the post-LOCA containment' environment. Nor= ally, this reaction O would not be rapid enough for considertaion; however, very extensive surface areas render this source i=portant.

i The production of noncondensible gas inside a pressurized water * .

i reactor (PWR) primary system occurs during normal plant operation. '

Hydregen and oxygen gas are simultaneously generated by the dissociaticn of core coolant via ionizing radiation in a PWR, Fortunately, a radiolysis back reaction operates concurrently which recombines the oxygen and hydrogen present to again form water.

Provided excess hydrogen is placed in solution in the primary system coolant, free oxygen generated will not remain because of this reccabination reaction, thereby preventing any flammable m.' :ture formatien within the reactor coolant loops. The plant cheuical and i volume control system ensures excess hydrogen is present in solution.

As long as the hydrogen remains in solution and inside the reactor I

coolant presaure boundary, it represents no hazard to the

containment. However, when the primary fluid is discharged to the containment, the dissolved hydrogen gas may come out of solution at f the lower pressure of the containment. Solubility of hydrogen in the water at the temperature and pressure characteristic of the reactor coolant system before the accident determines the maximum hydrogen that may be released. In addition, radiolytic decomposition of the water in the core and the sump is assumed to favor the dissociation reaction, producing hydrogen which separates from the coolant. The production rate depends on the quantity of fission products released to the coolant and the core power history."

gg)

(_ Appropriate pessimistic assumptions are made during analysis to bound the production of gas by this mechanism.

A.4 Computer Programs for Analysi_s ,

A.4.1 Current Thermal-Hydaulic Programs for Primary System Analysis The reactor coolant system response to accident initiating events j

i is studied using advanced thermal hydraulic computer codes. These software tools numerically approximate the continuity, momentum, and anergy equations for a series of interconnected control volumes.

Those most frequently used are the one dimensional RELAP 4(6-10) ,

RETRAN (6-2), and in the case of Sequoyah, the Wentinghouso WFLASH (4-1), and Westinghouse SATAN (4-2) codes. These programs predict the behavior of fluid systems when subcooled liquid, satuarated vapor, superheated vapor, and when the more physically co= plex two-phase

-(liquid and vapor) mixtures exist. The codes approximate the reactor system as a series of fluid control volumes and flow paths.

Homoseneity of the fluid is assumed which means that the liquid and vapor phases in a given volume are intimately mixed and flow in the same direction with the same velocity. Thermal equilibrium is also assumed between fluid phases indicating that the liquid and vapor in a volume are at the same temperature.

Some of~these codes include additional models that relax some of the i - g A-12 .

, av- i -- -- x-y-- - -m,~ ,---r - ~ -r- ~< s , e, - - - - -m,-- a n y m y ,-s-" w- o

above assumptions, particularly the assumption that the vapor and f3 liquid move at the same velocity. This is necessary for the proper

\,,/ modeling of small break loss-of-coolant accidents where the vapor produced during depressurization has sufficient time to bubble through the liquid, slipring past the slower moving fluid. The additional models only modiff the fluid solution which remains basically hemogeneous.

Currsnt research has led to the development of computer programs that solve a more complete set of equations, describing the liquid and vapor phases separately with coupling terms between phases to detail; for example, the drag of the vapor flow on the liquid. These codes include RELAPS (A-3) and the advanced multidimensional program TRAC (6-1). Thermal nonequilibrium is incorporated to accurately describe injection of cold emergency core cooling water into the primary system fluid which is at elevated temperature. These codes represent the state-of-the-art in reactor system mathematical description.

All of the above codes are single component, two-phase approximations to the fluid behavior. Vapor and liquid phases of water may be present simultaneously, but other components (e.g., nitrogen, hydrogen, and other noncendensible gases) cannot be followed using these programs. This limitation is not a problem since significant ccre oxidation followed by a large hydrogen production rate has not been predicted for the present design basis accident. Because the computer models do not determine the physics associated with the

_ presence of hydrogen and noncondensible gases in general, the

(_,s) distribution of the gas through the primary system cannot be derived directly from the vendor analyses. For the purpose of containment analysis, it is assumed that the hydrogen does not collect within the primary loop, but is released directly into the containment within two minutes of its production. This is not unreasonable in the design basis large LOCA due to the rapidity at which mass is released from l the primary system via the break. However, the single co=ponent limitation is a problem for the inadequate core cooling events.

A.4.2 Current Thermal-Hydraulic Program for Containment Analysis The Sequoyah containment was analyzed using the LOTIC (4-3) computer code. The Westinghouse LOTIC code (Long-Term Ice Condenser Code) has the capability to properly describe the postblowdown period in the ice condenser containment. Not only are the upper, lower, and ice condens'r volumes described, but also the ice condenser is divided into six circumferential sections, each with two vertical divisions.

Another significant feature of the code is the two sump configuration (active and stagnant sumps) such that the sump level increase and temperature history of the containment is accurately modeled. The code also describes the performance of the air recirculation fan in returning upper compartment air to the lower compartment. Coupling of residual and component cooling heat exchangers is provided to give i l

an accurate indication of performa1ce for these heat exchangers.

The spray heat exchanger performance is also accurately modeled in the transients. Tha basic equations used are the standard transient mass and energy balances and the equations of state used in any (s) containment transient, but appropriately coupled to the multi-volume A-13

ice ccndenser' containment. The cc.s also considers noncondensible

(} accumulator gas added to the containment and the displacement of free volume by the refueling water storage tank volume.

The LOTIC code, as do the primary system codes, uses the control volume technique to represent the physical geometry of the system.

Fundamental mass and energy equations are applied to the appropriate control volumes and solved by suitable numerical procedures. The initial conditions of the containment by compartment is specified before blowdown. Ice melt is calculated for the blowdown period based on the mass and energy released to the containment. After the RCS blowdown, the basic LOTIC code assumption is made that the total pressure in all compartments is uniform. This assumption is justified by the fact that after the initial blowodown of the RCS, the remaining mass and energy released from this system into the containment are small and very slowly changing. The resulting flow rates are unable to maintain significant pressure differentials between the containment compartments.

Tests have been performed by Westinghouse at their Waltz Mill facility to demonstrate the long term performance capability of the ice condenser. Specifically, these tests verified the ability of the ice condenser to reduce the cont'timent pressure within a few minutes following the blowdown, and have provided ice condenser performance parameters for tests simulating the long term addition of residual heat.

The thermodynamic conditions in these compartments are obtained as a function of time by the use of fundamental equations solved through numerical techniques. These equations are solved for three distinct phases of problem time. Each phase corresponds to a distinct physical characteristic of the problem. Each of these phases has a unique set of simplifying assumptions based on test results from the ice condenser test facility. These phases are the blowdown period, the depressurization period, and the long term period.

In the control volumes, which are always assumed to be saturated, steam and air are assumed to be uniformly mixed and at the control volume temperature. When the circulation fan is in operation, the fan flow and the reactor coolant system bolloff are mixed before entering the lower compartment. The air is considered a perfect gas, and the thermodynamic properties of steam are taken from the ASME steam table.

The condensation of steam is assumed to take place in a condensing node located, for the purpose of calculation, between the two control volumes in the ice atorage compartment. The exit temperature of the air leaving this node is set equal to a specified value which is equal to the temperature of the ice filled control volume of the ice storage compartment.

A-14 ,

A.S.1 Hydregen Analyzem Hydrogen detection and monitoring postaccident is essential to determine the proper time for mitigation of accumulated hydrogen.

This function is accomplished by the hydrogen analyzer system. The analyzer system is normally in a standby mode and calibrated, ready r for immediate operation. In addition, sampling may be performeu by (s-)/ removing gas from the containment and analyzing it in the chemistry lab as a backup.

The analyzers are located in the annulus and are constructed of materials compatible with the environmental conditions expected postaccident in the annulus area. Controls are located in the main control room and on a remote panel located in the auxiliary building in an area accessible postaccident. The system is designed to sample the containment atmosphere for the presence and concentration of hydrogen and is completely redundant (2 per unit). It is also designed to operate continuously during and following an accident oveg the containment conditions of 2 to 50 psig and 40 to 290 F. Concentration indicators and system alarms for both analyzer systems are located in the main control room.

The accurate measurement of hydrogen in the presence of nitrogen, oxygen, and water vapor is possible because the thermal conductivity of hydrogen is approximately seven times greater than nitrogen, oxygen, or water vapor, which have nearly identical thermal conductivities at the filament operation temperature of approximately 500 F. Hydrogen measurement is accomplished by using a thermal conductivity measurement cell and a catalytic reactor. The sample first flows through the reference section of the cell, then passes through 'ae catalytic converter where free oxygen is catalytically

,_s recombined witn hydrogen to form water vapor, and finally, flows

( ,) through the sample section of the measuring cell. The hydrogen content is indicated by the difference in thermal conductivity between the sample and reference sides of the cell.

The system currently analyzes H 2 concentrations over the range of 0 to 10 percent, with 1 percent of full scale accuracy; however, through inquiries with the manufacturer, it has been determined that the analyzer could be altered to read over the scale 0 to 25 percent wth only minor modifications.

A.5.2 Electric Hydrcgen Recombiners The accumulation of hydrogen in the primary containment building is centrolled via the hydrogen recombiners. The recombine.'s are manually initiated by the operator based on information from the hydrogen detection system. After an initial warmup time of five hours, the recombiners will begin depleting the hydrogen at a relatively slow rate. At the maximum design hydrogen concentration of 4 percent by volume, the recombiners will process 1.35 pounds hydrogen per hcur for each operating recombiner (two available). This recombination rate is approximately equal to the hydrogen production from sump and core water radiolysis in the first few hours of the accident and is sufficient to equal or exceed sump radiolysis, aluminum corrosion, and zine corrosion at times later in the accident. The short-term g

w)

A-15 -

P6w vdSMTEiN55*d SM

, . ,, . mm.

hydrogen generation by tho zirconium-water reaction in the core is accommodated by dilution into the containment volume.

Recombination of hydrogen and oxygen is accomplished by electric heating in the permanently installed recombiner units. The r"N reccabiners are sized to process a sufficient amount of containment

\s-) atmosphere such that the containment hydrogen concentration remains below flammability limits after the release of hydrogen from a 5-percent metal-watcr reaction assuming only one recombiner is functioning. Ai a result of these design parameters, each primary containment building is supplied with two such recombiners, each with a minmum capacity to process 100 SCFM of containment atmosphere even in the sc7ere post-LOCA containment environment.

Containment air is drawn into the unit by natural convection, passing first through the preheater section. This section consists of the annular space between the heater-recombination section duct and the external housing. The temperature of the incoming air is increased by heat losses from the heater section. This reduces external heat loss and results in increased efficiency of the unit. The preheated -

air passes through an orifice plant and enters the heater-recombination section which consists of a thermally insulated vertical metal duct enclosing five assemblies of metal-sheathed electrical heaters. Each heater assembly contains individual heating elements, and the operation of the unit is virtually unaffected by the failure of a few individual heating elements. The ingoming air is heated to a temperature in the range of 1150 to 1400 F, where recombination of hydrogen and oxygen occurs. The air from the heater-recombination section finally enters an exhaust section where it is mixed with cooler containment air before being discharged from the unit.

() Tests have verified that the recombination of hydrogen and oxygen in the unit is not the result of a catalytic surface effect but occurs as a result of the increased temperature of the process gases. The performance of the unit is, therefore, unaffected by fission products or other impurities which might poison a catalyst. However, should the hydrogen concentration exceed 6 percent by volume, it may be necessary to terminate recombiner operation to preclude initiation of a hydrogen deflagration by the recombiners.

A.S.3 The Hydrogen Purge System The hydrogen purge system is a backup to the hydrogen recombiners should they be unavailable. The system consists of two subsystems; an exhaust subsystem and a dilutent air supply.

The hydrogen purge exhaust subsystem consists of a single penetration in the primary containment wall equipped with two normally closed, remote manually operated isolation valves, one on either side of the centainment wall; one pneumatically operated annulus purge exhaust valve located within the annulus; and two 1/2-inch leakoff nipples located between the outboard isolation valve and the annulus purge exhaust valve. With the containment isolation valves open and the annulus purge exhaust valve closed, a flow path is established from the primary containment through the leakoffs and into the annulus,

\_/

A-16 f

c

which will pan 't purging of the containment for hydrogen control subsequent to a LOCA. The impetus for flow is provided by the differential pressure between the primary containment and annulus.

If the concentration cannot be maintained below 4 percent through

. N(" -

the leakoff path, the annulus purge valve will be opened to supply diluted air for a minimum time sufficient to maintain the hydrogen concentr,ation below the flammable limit. The containment effluent, purged for hydrogen control, will flow directly to the annulus where it will mix with the annulus atmosphere and be filtered by the air cleanup system prior to discharge to the outside environment.

Dilution air flow is introduced into the containment from the service air system. The service air system has provisions enabling it to receive diesel power. The dilution supply subsystem consists of a single 2-inch penetration in the primary containment wall equipped with provisions for containment isolation. The inboard containment isolation feature is a check valve located in the primary containment. The outboard containment isolation feature is a double 0-ring sealed flange located in the auxiliary building. A pressure hose will be required to provide a flow path from a service air flange. Operation of this subsystem is accomplished by removing the flange and coupling a service air hose to the pipe penetration. The system is sized to provide 60 SCFM of dilution air at a service air pressure of 60 psig.

A.S.4 Containment Mixing System

!!ixing of the containment atmosphere to minimize localized concentration of hydrogen is accomplished by the containment air

~ return fan system. The associated ductwork consists of: two 12-inch y,) ducts (one connected to each air return fan housing) which draw air in from the containment dome region; one 8-inch duct which circles the containment removing air from accumulator rooms and other dead-ended spaces and terminates at each air return fan housing; one 8-inch duct which circles the ;rane wall, removing air from steam generator and pressurizer compartments and terminates at each air return fan housing; two 8-inch pipes (one connected to each air return fan housing) which remove air from above the spent fuel pool; and a main duct between the upper and lower compartmeEt through the divider deck that includes the nonreturn dampers.

The ductwork described above is embedded in concrete, where possible, to prevent damage from buildup of pressure during a LOCA. Ductwork not protected by embedment is designed to withstand the LOCA environment. Rapid pressure buildup in ductwork in the upper containment compartment is precluded by nonreturn dampers which prevent the high pressure LOCA effluent flowing from lower to upper cocpartment.

U A-17 ,

- 6 3+ ~ e d e wsom.wevaW4 . ~:

s L . . w. . . _ .w -

u.

APPENDIX B RISK ANALYSIS BACKGROUND AND

^x IDENTIFICATION OF ACCIDENT SEOUEMCES

{d The safety of nuclear power plants can be addressed in terms of risk.

Risk of a specific event is defined as the product of (1) the probability of occurrence for that event and (2) the consequences of that event. Both entities must be considered simultaneously to obtain the overall plant safety.

The first extensive evaluation of nuclear power plant risk was released in 1975 by a group headed by Dr. Norman Rasmussen of MIT.

Known as the Rasmussen Report or the Reactor Safety Study, WASH-1400 was the first concentrated effort to provide an analytical assessment of the safety of a nuclear power plant using probabilistic methodologies. The WASH-1400 study addressed plant, system, and component failures, radiological assessments of accidents, and comparisons of risks associated with other common risks (such as airline accidents, natural phenomenon, etc.). The basic concept used in the WASH-1400 study was to estimate the risk of a nuclear power plant through an extensive evaluation of key and dominant sequences.

Key sequences are plant failures which are representative of many other combinations of system and containment failure states. Dominant sequences are those sequences which are considered to have the highest probability of occurrence.

The UASH-1400 study provided an estimate of plant risks based on r')

(_/ determining the probability for a particular sequence of system failures given an initiating event and then determining the consequences of such a failure sequence. The initiating events chosen for WASH-1400 consisted of large and small loss-of-coolant accidents, normal operating transients, and some special case failures. The study selected the Surry Nuclear Plant as the representative pressurized water reactor design. The Surry plant uses a Westinghouse nuclear steam supply system enclosed in a dry containment.

For each initiating event a decision was made as to what safety systems would be required to mitigate the consequences of the initiating event. An event tree was then developed for each initiating event in order to determine all the possible system failure scquences which could occur (see figure B-1 for an example) . Many potential sequences in the event tree were eliminated from detailed conaideration because the failure of certain systems would make immaterial the success of other syst m3 needed later in the sequence.

Each of the safety systems identified for use in mitigating the initiating events were then qualitatively and quantitatively analyzed in order to obtain an estimate of the probability that the system would fail to accomplish its intended function. These system failures will, in general, be independent (dependence being accounted for in the construction of the event tree).

Having found the probability of a particular sequence, the appropriate g~;)

(_ containment failure mode must be determined for each sequence. When B-1 ,

FJ4 LVLNT 1 ALE - LA%E LOCA (O > 6")

. A 2 0 I C 9 L 5 f 6 8

1 44 AF s

AF4 Ag l  ! Ang I g ,y Al l '

Al4 I ALF Ab i '

13G

! ADF

, 10

' #CG

, - ACF

' ACFG A CM j ' ACNG I ACNF ACL l

' Att3 f ACtf

, ACD I ' ACOG ACDF AA

' AIG m1F

' AIF G

, A tu I ' AIM 3 I AttF q ~ Alt

^

i ditG I

AIEF

(' I 480 fIDG I AICF

, Att AICG AICF AICFG AltN 1

AICMS I

11CMF

, A1CT I AtttG I

AIC(F

, AtCD

} AICDG I

AIC0F A A3

, Al AIG

, 12 F AZFG

, AZW AIMF

, AIL Alt f

, AfD AICF

, AZC AICG A!CF

' j AZCFG

, arc. \

' s' Alter

_ (j) t A2Cf l A:Co A!CCF LIG i

ricunt B-1 B-2

a particular containment failure mode was selected for a sequence (a sequence may have more than one possible ! allure mode), the h probability of that containment failure mode was estimated based on quantitative and qualitative analysis.

The next step was to estimate the magnitude of the consequences which could result from nuclear plant accidents. WASH-1400 did this by defining nin* release categories for a PWR uhich distributed by nagnitude the fission product releases of the accidents and by assigning each key sequence to one of the nine categories. Category one was defined to have the most severe fission product releases and category nine the least severe. Categories one through seven were 1 core melt sequences.

The selection of a release category for a particular Sequoyah accident sequence was made using WASH-1400 as a guide. However, the uncertainties in the analysis techniques, computer codes, and data

. require that the possibility be considered that a sequence has been placed in an incorrect category. The WASH-1400 study accounted for this possibility through the use of sequence smoothing. This method was also used in the Sequoyah analysis. Smoothing is a method for incorporation of possible release variations, and it assumes that j the release category chosen for a particular sequence was the correct j choice with a 78 percent probability. The method further assumes 4

ff the sequence was placed in the incorrect category, it most likely was a borderline case and the correct category would be the category i:anediately before and immediately after the one originally _ chosen

] with a 10 percent probability. A 1 percent probability exists that l

the sequence actually belongs in a release category two before or two after the one chosen.

I All relevant sequences were then compared to the key sequences in order to determine in which category they were to be placed. The probability of each sequence was calculated by multiplying the initiating event probability by the system failurc probabilities by the containment failure mode probability. By summing the probabilitics in et- release category, one can obtain an estimate of risk of the plant; that is, the probability of the various i sequences (occurrence) in that release category (consequence). WASH-1400 carried the analysis all the way to the ultimate consequence  !

by analyzing the effects of the release categories on a representative i population around the nuclear plant and estbating- the human and environmental impacts due to the fission product releases.

l j The objective of the Sequoyah analysis was to determine a reasonable  !

i estimate for the risk associated with reactor core damage and core j celt accidents at Sequoyah. ' As in WASH-1400, noncore damage sequences

, were considered as inconsequential to the risk of the plant and i therefore ignored. The risk analysis for Sequoyah was carried to I

! the point of-estimating the total probability of accidents in each release category.

l A format. similar to the WASH-1400 (Surry) study was used to establish '

the key and dominant sequences for the Sequoyah Nuclear Plant. '

4 Similar initiating events (large and small LOCA, etc.) were used for f l

B-3

Sequoyah as were used for the WASH-1400 study. However, the transients evaluated for Sequoyah were the loss of normal feedwater and the loss of ac power. The Sequoyah event trees showing the possible failure sequences for each of the initiating events are shown in figures B-1 through B-5 Table 5-4 provides an explanation

( ]) of the symbols and provides failure probabilities for the containment failure modes and those systems affected by the study. The assumption was made for Sequoyah that if the core melted, the containment would fail either by steam explosion (<x), hydrogen burn (3), containment leakage @), or containment overpressurization ($). Basemat penetration (G) was not considered. These assumptions resulted in reducing the release categories to five for this study. Sequoyah dominant sequences and their probabilities for each release category are shown in table 5-2.

The Sequoyah analysis was based on a limited number of detailed evaluations of accident sequences and other qualitative evaluations.

Three detailed evaluations were obtained from the Battelle, Columbus Laboratories and these were:

1. A large LOCA followed by failure of the emergency core injection system (AD).

2. A small LOCA less than 2 inches followed by failure of the emergency core injection system (S2D).

3 A transient which was loss of all alternating current power followed by loss of the auxiliary feedwater system (TMLB').

TVA used these three sequences to perform a detailed evaluation of

{s'-}/ a fourth sequence: a small LOCA followed by the failure of the emergency core recirculation system (S2H).

J B-4 ,

v... . . - - . . . , _ . -- -

,,g  ;

o m '. -

, - , .n .. .-. . . . . _ ._ - . - - - - -,. -.. .-,2. .

3g 2 & i & s & h F 4

, le SeG

< v I' t

, l S im3 SenF I

a r I0 1 l SgD$

1 ' Si tF J

r II I SgCG r I I II

, SgCfG i

, IgC4 4 '

l SgCnG I

SgtnF Si t0 i ,

51 5:14

( Se tF

$ 31FG 1, , SgIN

, I '

$sth6

) I Si ttJ

! , si t0 I '

t il?$

I

! latDF i

I I II SgICG

, $ gICF ,

8 l SgICf G

, $ 3tCH l _l 3:ICH3 I

SgICHF j SgICO sit I SgtG I

SgtF SgIC

, f irt -

l '

$ 3f tG

' I sptr i - SgtrG jl , sse SgFt J

i

, Sgl

{

• Sg2G I S II g

3, I

.i $glfG

, sgFn l $ 3ZhG e

, C Si lhF

, 5:20 Ig2cG

$g20F

, Sg!C Si2C3

$g2CF M S gICFG r-- -- S lCN i I '

Sg2CNG 1

I

-- $gtChr IglCD

, $ git I '

5 girl I

5,7tF

! $ 3l&C i- , sgra

• i i

ricuRE D-2 B-5

. . _.. , _. ,_ . _ c ; . a; ;; 4 a.m ,,_.,_.,.__m.,,z_,m.;_.__.

' -[hh3E'M h 2, 4,mdJ-h .: - , C '

L- - 4 ' J:-, , -? d 99- .-

Q4 L M 41 th( WA.L L'i.A (0 ( 2* )

Sg I O E I b 4 9 8 8 0 83

'll s,*

ses A s SG

'Q' I 9,88 v

SgC6 Sgtf Sgtng

~

BytB Sgf B6 i

yer yte 34 6

• St g

• Bg08 s yst

-- - sgars

• -r-c ::: Synet

, se S git Sg ste Bgatr

_ Sg5Cf6 93tLn9 96 53 rtar r-C : ".

i

(')

\ r-C ':::'. SystLP

'r-C :::.

SgS9 r-C::1 r-C:l1 i

S 6 i

GEE Sg, St 1338 5 ,7 3

_ spas I "#

Egfes

,96 opt agts

-C::::',

f _r-C ;','; ,

5 (86

,,n-13 4

.r-C l:::',, , ,,

i -

a Q '

spst r-C P' e

inutt B-3 B-6

O SON EVa4T TREE - TRANSIENT (LOSS OF AC POWER)

TA K B2 L If P Q

[ TA f TQA TAH

- - TAli0

- --.- TAMP

-~

Tf,L

~

O -_

-- T A'9

--. T L A P

.=

TBA2

-- TA2 39 Tg ------- T GA2P g---- TAK i

_ _ _ _ T AKQ

-- T KA ?

i TAKB2 l- TAKD29

--- T A K C 2P C' FIGURE B-4 l

1 B-7  ;

SQH EVENT TREE - TRANSIENT (LOSS OF MAltl FEEDWATER) .

rs L) T8 K L M B2 UI P Q D 0 Tg j T9 B i TgQ0

[

- TgQD  ;.

- TgQD0 l l ._ TBM '

T BM0

. T MD6 T BMD0 i TB!;P i

= TgMP0 {

q TBL TBLO l

k r -= Tgl0 l'

- ---- - - - - - - - 3 TglD0 [

TBL."0 j

- TBDg l

~~

_~L----- -

73gig f.

TBP OI  !!

- TgBlM ll

- I- -

- T DBiMP i g -= TB Bil  ;

I -

TBI B LP j'

. _ , __f- -

B2 B Tg I - Tg82P i-TBK TgK0 I TDKP

- TBKP0 TgKBl

-- T r.B B l P TBKB2 (J's TBKB2P FIGURE B-s B8

~

- O>

\--

APPENDIX C MARCH COMPUTER CODE C.1 General Description 4

The MARCH modules necessary to cover the accident physics are shown in figure C-1 and described below.

Subroutine BOIL is a modification of the BOIL program which was written for the reactor safety study (WASH-1400). The subroutine calculates primary system mass, energy balances, and primary system heat transfer. In combination with subroutine PRIMP, which calculates primary system pressure variation and leakage rates, the subroutine solves the necessary equations of energy and continuity to simulate a one volume RELAP type model for the core. A level swell or bubble size model is included in the core heat transfer nodes. Core fuel elements transfer their heat to the water and thereby, as an option,

, to a steam generator. In areas where only steam is present, convection to the steam, as well as radiation to the upper internal structures and to the water below, is modeled. Hydrogen is generated by oxidation of the hot cladding with steam, consuming the steam in the reaction. MARCH appropriately limits this reaction based on the availability of steam. Core slumping is calculated once the melting

()

point of the fuel is reached. Cladding metal and relocation is ngt 3

treated separately from core melt. A melting temperature of 4130 F is used. Three options are included for the slumping behavior, however, due to large uncertainties it is not known whether some of these are truly mechanistic. Other models which are simplified include the critical flow discharge model that permits only single phase steam or water to be discharged but not to a two-phase mixture.

i Subroutine MACE determines the containment response to energy and mass released from the primary system. Ice beds, containment fans, sprays, and air cooler unit models are included in this subroutine.

Concrete decomposition and hydrogen burning effects are considered in the containment model. MACE functions as a simplified BEACON or LOTIC code. - -

Additional coding is used to determine meltdown specific phenomena.

Subroutine HEAD determines the time of failure of the vessel due to meltthrough once fuel assembly failure and slumping has occurred.

Following vessel melt, molten debris is expected to drop into the water remaining in the reactor cavity and to boil any fluid p"esent.

. Subroutine HOTDROP determines the boilaway rate. Finally, sut outine INTIR calculates the decomposition and penetration of the cont nment basemat by the molten core debris.

Many of the physical problems modeled by MARCH have been found to be phenomenologically complex during reactor accident research. The s-5 C-1

• FIGURE C-1 MARCH Input Reactor Description Engineering Safety Features Initial Conditions Modeling Options 1

y .

PRDt? 3 L= OIL +

EXITQ I E ECCHX PRIMP -+ HEAD +

f- ECC BURN

% MACE +

CCOL Y

CSHX HOTDROP +

SPRAY U

INTER +

A

' r - - - - ' - - - - - - - "A A

b

~~~

PROPS FPLOSS (Ref. 6-3)

C-2 ,

'_ _ . . , ; : c,:. . . : ' ' -'- . .: s. . -

12:~ '

melting of the core to a pool of steel, uranium, cladding, control rod material in the possible presence of some remaining water is a complex heat transfer problem. If the fuel is able to heat to

() extremely high temperatures in the range of 5000 to 5500 F (note that some literature suggests the fuel may not melt until this temperature is achieved) the stainless steel and other metals may begin to boil forming a frothing metallic pool.

Boiling of the steel would raise the pool level and increase heat transfer from the sides of the molten debris. This could delay vessel melt-through, possible radiating more heat from the outer vessel walls to the reactor cavity liquid, boiling a larger quantity of this liquid, and therefore reducing the probability of steam explosion following vessel failure. It is also unlikely that symmetrical

, melting of the core, vessel, and other components will occur. A more likely scenario would be that portions of the core may melt, dripping into the remaining vessel water causing a slower boiling of this water than predicted by MARCH. It should be noted, in summary, that uncertainties exist in the calculational models employed in this code, that the models may be too conservative in some areas, that some phanomena may not be treated properly, but that the overall code is a significant extension to the current analytical tools and that no othc' codes are currently available to study the total core melt sequence.

C.2 MARCH Code Check A simplified inhouse computer code named H2 GEN was created to examine adiabatic heatup of the Sequoyah core fuel assemblies following a O major accident with inadequate core cooling. Adiabatic heating of the fuel, once uncovered, results in the most conservative rate of temperature increase. Fuel rod oxidation rates and consequently hydrogen rates are maximized, allowing an upper bound on total hydrogen production to be estimated. Comparison of H2 GEN results-with MARCH output can be used to determine the conservatism in the more advanced code.

H2 GEN considers positional variation in the average power vi core fuel assemblies. Higher average power assemblies generate more heat following uncovery and achieve larger hydrogen production rates earlier in the transient. Axial variation in the individual rod power has been neglected in H2 GEN. Core boildown is not treated explicitly, but is entered as user input from emergency core cooling system studies. The fuel assemblies are modeled as axial segments with twelve nodes per rod. Heat transfer to the core mixture adequately cools a node until it is uncovered, at which time heat transfer from the node is terminated and heatup begins. The program tracks the oxide interface as the cladding reacts with the steam present. Once i

all the cladding in a rod has corroded, no more hydrogen is generated l by that rod. Several models are included as a user option to j approximate.the oxidation rate including the more conservative Baker-Just correlation. Unlimited quantities of steam are assumed available for cladding oxidation unlike the physical event where oxidation of the lower rations of cladding may consume all steam available, or

. water bollaway rates, influenced by radiant heat transfer, may not c-3 .

. .- . - = .- . .- .

be sufficient to oxidize cladding at the rates predicted. Rod heatup is based on volumetric heat generation at ANS decay heat power levels with no excess conservatism included. Thermal energy is also supplied

()

}'

by the oxidation reaction in each node where oxidation is in progress. This contribution can be very significant at large oxidation rates.

Comparison of H2 GEN results with MARCH output for small break i inadequate core cooling events demonstrates that for the portion of

the transient immediately following core uncovery, the MARCH results

! are similar but slightly below the strictly adiabatic H2 GEN results.

Deviation is noted in later periods as the adiabatic generation accelerates in the presence of unlimited steam, whereas the MARCH results exhibit steam limiting. The H2 GEN hydrogen production terminates after all the zircaloy cladding is oxidized, while MARCH hydrogen generation exceeds the maximum cladding reaction for some i events. However, this is not unanticipated, since MARCH also esticates the quantity of hydrogen produced by the steel corrosion reaction. MARCH appears to be adequately conservative over the

[ inadequate cooling transient based on these limited studies.

1 1

i O

.i 9

i I

i 1

i 1'

. ()

C-4

- - . . . _ . - ~

_- _ ~ _- _

APPE!! DIX D D DETAILED DESIGN CRITERIA h

D.1 Filtered Vented Containment Ue pecpcse tuo design cases for evaluation:

I Ccre celt event (large LOCA, no hydrogen burn)

II Cegraded core event (emall LOCA, partial hydrogen burn)

CASE I_

1. Maximum vent volume flow rate - 400,000 cfm at 600 F, 34 psia. This lasts approximately 600 seconds, then drops.

2. Vent Temperature - 600 F - Design Temperature - 750 F 3 Vent Driving Pressure - 35 psia - Design Pressure - 55 psia

4. Constituents - Air - Swept Out (88,000 lb) after peak

, presssure H - 8% 3,400 lb 2

has swept most air Steam - 47% 450,000 lb out

(,O) CO 3

_ - 23% 35,000 lb CO 2 - 22% 128,000 lb Average Density 0.04 lb/ft 3

5. Radioactivity - see Table D-1 -

6. Duration - 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> 6

Total Heat Input - 600 x 10 Btu

7. Operator will cycle valves - no auto on pressure.

8. Vent size - approximately 7.5 ft - Vent flow rate is to be centrolling factor.

CASE II

1. Maximum volume flow rate - 467,000 cfm at 700 F and 55 psia. This lasts approximately 100 seconds. The flow rates are much less at other times, eg 20,000 cfm for previcus vent periods.

2. ' Vent temperature - 700 F Design Temperature - 750 F D-1 *

~

. TABLE D-1 CASE I FISSIC'T FRODUCT P2 LEASES TO CCNTAIIMENT ATMOSFHE?2 DURING CORE MELT ACCIDETr

=

8 Number Name Source (Curies x 10 )

1 KR-85 .006 2 KR-85* .26 i 3 KR-87 .52 4 KR-88 .76 5 SR-89 .12

6 SR-90 .0057 7 SR-91 14 8 Y-90 .00068 9 Y-91 -

.018 10 ZR-95 .021 11 ZR-97 021 12 NB-95 .021 13 M0-99 .13 c 14 TC-99* .11 15 RU-103 .08 16 RU-105 .046 17 RU-106 .015 18 RH-105 .046 19 TE-129 .28 20 TE-129* .10 21 TE-131 .15 22 TE-132 1.2 23 I-131 .85

_..~~ 24 MI-131 .85 1

25 I-132 1.2 j 26 .i I-133 1.7

27 I-134 2.0 28 I-135 -

1.5 29 MI-135 1.5 30 XE-133 1.7 31 XE-135 .26 32 CS-134 .017 33 CS-136 .06 34 CS-137 .058 35 BA-140 18 36 LA-140 .021 37 CE-141 .021 38 CE-143 .020 39 CE-144 014 40 PR-143 .020 41 ND-147 .0078 42 PM-147 .0022 43 PM-149 .0052 44 PU-238 .000013

_ 45 PU-239 .0000013 1 of 1 D-2

3 Vent Driving Pressure - 55 psia (design pressure). Initiate venting at 35 psia.

4. Constituents - several hydrogen burns frcm 5 percent to 4 percent until last big burst:

Constituent After First Burn After Last Burn N 72% 55%

2 0 15% 6%

2 H 5% 4%

2 Steam 8% 35%

Minimum vent open or driving pressure 35 psia.

5. Radioactivity - see Table D-2.

6. Duration - 10 to 12 burns (ventings) over 2 -3 hours. Total 6

heat input - 200 x 10 Stu over 2-3 hours.

7. Operator will cycle valves - no auto on pressure.

8. Vent size - approximately 10 ft - Vent flow rate is to be controlling factor.

(qI

• 9 Vent line length - 500 ft total run - assumed 15 ft below grade for yard pipe run.

10. Decentamination Factor - 100 for particulates ~

100 for iodine 1 for noble gases

~

11. Condensate - should be collected as part of conceptual design of filtered vent system.

12. Seismic - to withstand OBE (0.09g).

13 Tornado - no pressure or missile requirements.

14. Eedundant - no redundancy requirements (except isolation valves).

15. Qualification - Mechanical Quality Group "C"

16. Passive versus active - design should be passive as far as possible. Active components can be used when necessary or required to minimize sizes of components.

17. Cooling Uater - 5,000 gpm of 100 F cooling water available if necessary.

q L.

D-3 '_ .

TABLE D-2 CASE II z:.

FISSION PRODUCT P2 LEASES TO CONTAI!UEIT DURING DEGRADED CORE ACCIDENT (CURIES) ll O

PRIMARY UUCLIDE CONTAIIREIT

=

1 I-131 7.685E 06

. 2 I-132 9.222E 06 4 3 I-133 ' 1.680E 07 4 I-134 1.077E 07 5 I-135 1.443E 07 6 I-131* 8.774E 05 7 I-132 1.053E 06 8 I-133*- 1.918E 06 9 I-134* 1.230E C6 10 I-135* 1.647E 06 3.481F 07 i

(]) 11 ER- 85M 12 KR- 85 9.989E 05 13 KR- 87 4.898E 07 14 KR- 88 8.858E 07

, 15 KR- 89 4.045E 03 16 XE-131M -

9.211E 05 17 XE-133M 5.107E 06 3

18 XE-133 2.021E 08 19 XE-135M 6.570E 06 20 XE-135 -

5.223E 07 4

2.997E 04

~

21 XE-137 22 XE-138 .1.719E 07 1-i .

I i"

1 of 1 D-4

16. Isolation - three remote manual valves, one inside, one in annulus, and cne cutside secondary containment - folloued by a 35 1. sis rupture disk outside accondary containment.

19 Vent stack - 400 ft. height O

t f

D2 Additional Containment Volume

1. Design temperature - 750 F

2. Design pressure - 55 psia 3 Vent size - 48 in.

4. Additic.Til containment volume - 1.5 x 10 6 ft.3

5. Additienal containment design pressure - 20 psig

6. Vent line - below yard piping (approximately 15 ft below grade) 7 Sei mic - to withstand OBE

8. Redundant - no redundancy cequirenents (except isolation valves) 9 Isolation - three remote manual valves, one inside, one in annulus, and one outside secondary containment - followed by a 35 psia rupture disk outside secondary containnent r"5 10. Actuation - operatcr will cycle valves - no automatic on U pressure D.3 coupled containment

1. Vent line design temperature - 750 F

2. Vent line design pressure - 55 psia 3 Vent size - 24 in 4 Vent line - below yard piping (approximately 15 feet below grade)

5. Seismic - to withstand JOBE

6. Redundant - no redundancy requirenents (except isolation valves) 7 Isolation - three remote manual valves, one inside, one in annulus, and one outside secondary containment - followed by a rupture disk outside secondary containment

8. Actuation - operator will cycle valves - no automatic on pressure i

D-5 ,

wm . .-

- ..u. ~ . . . a.: w .

D.4 *u mented Centaintent Cooling

1. Environnental design temperature - 342 F 0

' 2. Heat renoval rate - 9 x 10 Btu /hr 3 Scistic - Category I(L)

4. Redundant - no redundancy requirements

5. Actuation - autcmatic on high containment pressure and tcaperature D.5 Containv.ent Inerting

1. Maximum 0 f r inerting - 4 percent 2

2. Minimum 0 f r deinerting - 16 percent 2

3 Nitrosen supply - present 24-inch purge 4 Nitrogen exhaust - present 24-inch purge

5. Seistic (inside containment) - Category I(L)

6. Ecdundant - no redundancy requirements D.6 Halen Injection

() _

1. Minicum halon for inerting - 3 percent

2. Halon supply - 3 ring headers it ?per cccpartment, one ring header in lower compartment, at least one spray nozzle every 2,000 ft3 of free volume 3 Minicum halon discharge pressure - 260 psig

4. Seismic (inside containment) - Category I(L) 5 Redundancy - no redundancy requirements

6. Containment isolation - standard

7. Actuation - containment isolatien valves open a2tematically on hi 6h ccntainnent pressure followed by low re actor vessel level; nanual isolation at halon tank D.7 Icnition Sotrees

1. Hydre;en release rates - see Table D-3

2. Hydrogen release locations - stall /large LOCA; pressurizer relief tank; reactor vessel head vent (assume it can be Gli D-6

.: 4hea p m ablh .-

TABLE D-3 S

7g

- HYDROGEN RELEASE i)

TD'I MASS TDiPERATUPS (EEC) (1b,/SEC) P 0.0 0.0 61.24 3430. 0.0 61.24 3804. 0.0413 66.56 4116. 0.2600 1582.29 4423. o.7400 795.45 4752. 1.0700 771.47 5700, 0.4300 661.53 6330. 0.2233 555.39 6648. 0.1600 535.22 6960. 0.1167 519.43 8070. 0.0367 519.43

, _s 9252. 0.0043 . 519.43

's )

_ 9384. 0.0 61.40 9552. 0.0 61.24 9630. 0.1667 226.68 9636. 0.0 61.24 9651 o.6667 61.24

~

9654. 39.0667 4251.30 9660. 100.20 1670.01 9660.1 0.0 550.42 9672. 0.0 61.24 22374. 0.0 61.24 24690. 0.0267 2676.15 30138. 0.0417 .2678.20 31608. 0.0450 2681.12 35202. 0.0533 2684.05 35208. 0.0617 2681.12

-y ,'

1 of 2 '

D-7 .

6 TABLE D-3 (cont) i i

O .

42403. 0.0717 2675.27 46CO2. 0.0800 2675.27 4

49614. 0.0 61.24 6.0 x 1010 0.0 61.24 i

l 1 I i  :

f i

i l

I O

1 l

> i I 1

i O

2 of 2 D8 -

l - . . _ - . . . - - - _ . _ - - . _ . . _ . - - , - - ... - -

i i i i

I routed or burned as desired) 3 Mixing - at least one air return fan is running  !

4.

Scirmic - Category I(L) i S. F.edundant - no redundancy requirecents l 6. Actuation - actuate at low concentrations (approximately 4".) of H 2 and monitor controlled burning i

i b i DA02:APPDCY.01 i  !

i ,

i i

t e

i

^

. O 4

i

)

l l

l i

l l

l O

D-) ,

d I.IST OF REFERENCES O 3-1 Rae==eien N., et e1., "Reeeter Sefety Stuey - im Aeeeesment of Accident Risks in U.S. Cocmercial Nuclear Power Plants," ,

WASH-1400, NUREG-75/014 (October'1975).

4-l' Esposito, V. J.; Kesavan, K.; and B. A. Maul, "WFLASH-A FORTRAN-IV Computer Program for Simulation of Transients in a Multi-Loop PWR," WCAP-8261, Revision 1 (July 1974) 2 (Honproprietary).

I 4-2 Bordelon, F. M., et al., CSATAN-VI Program: Comprehensive

, Space-Time Dependent Analysis of Loss-of-Coolant," WCAP-8302 '

-(June 1974) (Proprietary).

4-3 Griham, N. P.; Colenbrander, H.G.C., "Long Term Ice Condenser

Containment Code - LOTIC Code," WCAP-8354-P-A (July 1974)

(Proprietary), WCAP-8355-A (Nonproprietary) (July 1974).

4-4 " Analysis of Three Mile Island - Unit 2 Accident," Appendix HYD, Nuclear Safety Analysis Center, EPRI Report NSAC-1 (July 1979).

4-5 Memorandum from W. Butler, Chief of the Containment Systems Branch, to R. Tedesco, Assistant Director for Reactor Safety,

] and entitled "Three Mile Island, Unit 2: Analysis and Evaluation 1 of Selected Containment Related Issues," (April 1979).

O a-6 re e11. o. o . et e1.. "rvi Nue1eer er 8ce= Review" (81ee 8 ex)

(May 1979).

4-7 Beasley, E. G., et al., "TVA Nuclear Program Review: Sequoyah j Nuclear Plant and the Report of the President's Commi.3sion on the Accident at Three Mile Island" (Green Book) (November 1979).

6-1 Jackson, J. F., et al., " TRAC-Pl

An Advanced Best Estimate Computer Program for PWR LOCA Analysis," NUREG/CR-0063 (Jun