ML20125B651

Expansion of Drywell Containment Vessel

ML20125B651
Person / Time
Site:	Oyster Creek
Issue date:	12/31/1979
From:	JERSEY CENTRAL POWER & LIGHT CO.
To:
Shared Package
ML20125B634	List: ML20125B632 ML20125B640 ML20125B643 ML20125B646 ML20125B648 ML20125B651 ML20125B653 ML20125B657
References
TASK-03-02, TASK-03-03.A, TASK-03-07.B, TASK-03-07.D, TASK-3-2, TASK-3-3.A, TASK-3-7.B, TASK-3-7.D, TASK-RR NUDOCS 7912190683
Download: ML20125B651 (41)

Text

.-- e ...n . . +. . . , . , , . . . . ~ . . . . . .n-..., --,

'. e ,

W I

,a :_  ;

t k

. Docket No. 50-219 9

l, . .

~

b 4

ATTACHMENT H OYSTER CREEK NUCLEAR GENERATING STATION EXPANSION'0F THE DRYWELL CONTAINMENT VESSEL

- 90000184

~

L t

i December, 1979

'79121'90 [kf

c.

$ OYSTER CREEK NUCLEAR POWER PLANT EXPASSION OF THE '

h[iI DRYWELL' CONTAINMENT VESSEL'

. CONTENTS SECTION 1 Scope 2 Purpose 2.1 Necessity for Separation ~of Ste'el and Concrete 2.2' Methods of Separation 3 Material Criteria 4 Material Selection and Testing .

4.1 Available Materials 4.2 Standard Fire-Bar

.({

4.3 Development and Preliminary Testing.

4.4 Commercial Practicability Evaluation 4.5 Final Development Tests 4.6 Rebound Evaluation 4.7 Susceptibility to Damage 4.8 Protection and Bond Breaker 4.9 Conclusion 5 Determination of Required Thickness  ; the Material lh 5.1 Concentrated Load Test of Vessel k' '

5.2 Preliminary Vessel Expansion Procedure

))p 5.3 Material Thickness

-g 5.4 Vessel to Wall Distance j

90000185 L'

s- s

, 3 rg '6' Application- of. Compressible Material 4..g .:. .

-6.1 Lower Region ,

6.2- Access for Application to the vessel 1,'

I- ,- .

i 6'. 3 - Application 6.4 Bond Breaker D D 6.5 acpairs ,

l 7 Production Inspection .j 7.1 Density ,

7.2 Thickness .

8 Production Sampling and Testing r 8.1 Samples 8.2 Tests 8.3 Off-Wall Samples 8.4 Penetrometer ,

9 Evaluation of Production Testing 10 hequired vessel Expansion Conditions .

10.1 Maximum , Vessel Temperature change 4

. 10.2 Allowance for Rebound .

,10.3 Internal Vessel Pressure Criteria 10.4 Evaluation of External Pressure

, 10.5 Selection of Internal Vessel Pressure I

. 10.6 Expansion Due to Pressure 90000186 ,

. 1 e g 4

, . . . - . . - a , , , , , - , . . . ,. ,v., . . + , . .- ,

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

J- i '

10.7e Conclusion e hfG l11" .: Selection of: Methodiof ' Expansion 12- Expansion < Equipment and Arrangement h #-

12.1 Pressurizing

^

12.2 Heating 12.3 Monitoring' Temperature. .

12.4 ' Monitoring Expansion-13 Preparation for Expansion Operation 14 ' Expansion, Operation Procedure 14.1 Pressure Test 14.2 Pressure 14.3 Heating

( .

14.4 Cool Down 14.5 Initial Computer Program t

15 March 3 ' 6, 1967 Expanssx1 Operation 15.1 Required Conditions .

15.2 Pressure Test ,

15.2.1 Pressure ' f

-)

, 15.2.2 Temperature ,

i 15.2.3 Expansion 15.2.4 Leakage I

.15.3 Heating with 480 kw Capacity 15.3.1 Temperature -

kh

,- 15.3.2 Pressure 90000187 t, l

%; I

s

~." - i

{

15. 3. 3 E::pr.nsion -

If 15.3.4 other cbservations -

15.4. Heating with L 720 Joe capacity ,

15.4.1. TemperatureE .

, 15.4.2 Pressure 15.4.3 Expansion 15.4.4 other Observations

'16 conclusions from the March 3 - ,6, 1967 Operation 16.1 Moisture in the Lining 16.2 Heat Loss through Vessel Head j 16.3 Voltage 16.4 Conclusions

( , 15.5 . Revised Computer Program 17 March 10 - 12, 1967 Expansion Operation .

17.1 Required conditions

'17.2 Heating with 960 kw Capacity

, 17.2.1 ' Temperature 17.2.2 Pressure

, 4 17.2.3 Expansion 17.2.4 other Cbservations ,

d 1

"$ T Nkh

. =;

90000188 4 6

h

y. . (-

.: W .

, J' OYSTER' CREEK NUCLEAR P'ONER' PLANT EXPAES0.hOh^THE

'g rp rm lg m -g: . .

DRWELL CONTAniMENT VESSEL : Jg  :

j g

'll SCOPE ,

-+

The purpose, ma terial L s' election, testing, .- in-

=tallctionl and the compressing. of the inelastic compress-- '

ible mctorial applied to thef drywell- containment . vessel ,

to provido for expansion of the vessel'are' described herein. 5

2. PURPOSE

• 2.1, Necessity for Seoaration of Steel and Concrete , s

~

Containment and radiation. shielding'of the Oyster ,

Crcok Picnt are provided by separate structures: Containment, ,

the " pressure' suppression"' system, is provided by steel' press-ure vessels; . a "drywell", housing the reactor andL recircula-tion . system and consisting of a 70 foot diameter sphere'sur-

.- . mounted by a'33 foot diameter cylinder, and.an interconnected dm. absorption chamber containingLthe, water' volume:necessary for pressure suppression. Radiation shielding is provided.by.a -

concrete wall outside the drywell vessel with a minimum thick--

noss of 4 feet 6 inches. . The pressur,e, 62 psig, associated with the postulated men' requires a rela tively thin walled .

(about 3/4 inch) steel vessel. The vessel has,relatively -

little capability to resist' concentrated jet forces or' the im-pact of missiles. jiuch loads are , however, readily accepted by the massive concrete shield. Accordingly, the space between the steel drywell vessel and the concrete shield outside must be sufficiently small that, although local yielding of - the steel vessel may' occur under concentrated forces, yielding - '

to the extent causing rupture would be prevented. Some space must be.provided since, if it is to function asia pressure- I container, the vessel must be allowed to expand when in its 1

. stressed condition. The vessel is also' subject to thermal' l

expansion since it will be exposed to opera' ting and accident i temperatures significantly higher than an ambient temperature l practical to maintainLduring construction.

' ~

@ .. 1. 90000189 Y,

,* - eaye- -- .* * -e e+-

2.2~ Methods of1 Seenration  ;

f49 The1 maximum acceptabie : space, in the order of 3 inches, _ (further defined below) precludes the use of a con- ,

. ventional . forming _ system f or the ' inner f ace . of the concrete wall. That face of the concre aall could be formed with n material spaced away from tro vessel and suitable to'rc-main permanently in place - (the many penetration pipes pro-jocting from the vessel would' prevent slipping forms out). pro- '

vided its support was' independent of the vessel or if supports betwoon vessel and form were removable. Alternatively, the space' could be temporarily filled with a material which could, after placement of concrete, be removed by being broken up and extracted, or by being melted and drained off, or by being sublimated.

Of these methods, those involving the c1 caring ,

of the space .af ter complete construction of the 80 foot high multiple curved concrete surface are open to question as to

, whether the material was completely removed. If the material

. were to be removed in steps as the construction of the shield advanced or' if an independently supported form were used, the procedure would be open to question as to whether the come

_ pb ted portion of the expansion space in the lo'wer region had ,

.( become obstructed during progress of construction above. In- .

cither case, the limited width of the space and its curvature and penetrations precludo subsequent inspection to establish I whether' or not the required clearance exists continuously.\

/ I Another approach to the problem would be to fill ~ '

the space permanently with a material having suf ficient com-pressibility to pe,rmit the expected vessel movement. The com-pression characteristics of a material to be left in place l

, between the vessel and the concrete would necessarily be such that. it would not deflect significantly under the fluid pres- ,

,sure of concrete.

The fluid concrete pr,e,ssure can be con-trolled by limiting the rate of placement of the concrete but l a practical lower limit would be about 3 psi,. Commercially l available compressible materials or practical modifications ]

of them having such stiffness would require a pressure of at  !

least 10 psi to be compressed to the extent equal to the ex-pected vessel expansion and would of course impose that press-ure as a reaction on the vessel. While this thin-walled large-diameter pressure vessel' is capable of resisting a much larger '

bursting pressure, it cannot sustain an external' pressure of )

this magnitude unless the external pressure is balanced by a l (fp, positive internal pressure.

o ] "oE S kuu

'Ed o .

90000190 1

(. t

.j IIcwever , tho'need for a continuous balancing

,,,, ' internal ~ pressuro' within the expanded vessel could be cli-  ;

'~QF

minated by creatino an air cap into which,the vessel could ji i f r o c l y e x ,r. a n d . This could be accomplished by solceting a- l cc.r.procaible material which is inelastic; such a material. 1 i could be 8;;mycasced..caq.c. hbn ' ef fect,~ simulating the con-i dition

; costulated as causing the greatest vessel expansion j i

whilo . .ain toi.ning a ,bal::ncing intornal pressure.

Th e.l.cAid_ga; _ l l y r "n1 created by the ine'a_stic l compressio.p o.f th.e.m.attria.  !

would_,rd[.or no rcsLisj:a_nce_ to , subsequent r,epetitions ,of ves_se_1 ox ,gpnion. .

3. :c,TERI AL cRITERI A l

The material to be used for an inelastic compress- j ibic lining would be required to have the following character-intics:

a. Would adhere tightly to a curved, painted steel plate surface in flat, vertical and overhead positions. -

b. Would have relatively insignificant defor- .

mation under fluid pressure of wet con-(' crete estimated at 3 psi,

c. Would be reduced in thickness inelastically by about one inch from an initial thick-ness of 2 to 3 inches under a pressure of preferable not more than ten psi. (Since

'the concrete wall must be designed for the pressure associated with this compression, a pressure approaching 62 psi would require, in effect, the cost penalty of building a redundant pressure containment vessel in re-inforced concrete)

d. Would remain dimensionally stable at the' reduced thickness without significant flaking or powdering (to avoid the obstruction of the lower region of the gap with detritus s from the lining material above. )

e. Would be unaffected by long term exposure to radiation and heat. -

- hi$-

g g {gj 3 .

e

- f. Should be on economical, commercially avail-(h; able product or, practical modification there-o f, readily applicable to the vessel.

g. Should be susceptible' to minimum damage (consistant with the other requirements) while exposed on the vessel before concrete ,

placement.

4. *ATF. RIAL SELECTION and TESTING 4.1 Availnblebaterials .

The field of commercially available insulation, fire proofing and shock absorbing packaging materials was ex:.mi ned : Plastics showed considerable elasticity and a tendancy to flow or decompose at relatively low temperature (abou t 175 F) . Fibrous blankets or boards showed excessive' ,

clacticity. Cellular inorganics (foamed glass and expanded aggregate concre te) could not be reduced to acceptable strength and were excessively friable.

4.2 Standard Fire-Bar

~

The material ultimately used in'a modified formu-lation was a combination of the fibrous and cellular' types. ,

The material is a proprietary asbestos fiber - macnesite l cement product _ applied under the trade name Fire-Bar, normally l as a spray coat over steel structures for fire protection, l by the All Purpose Firoproofing Corporation of New Hyde Park, l N. Y. The solid materials; asbestos fibers, magnesite and  !

magnesium sulphate; (roughly 75% asbestos) are premixed and, at the site of the work, are combined in a mortar mixing l machine with water and, to control density, with foam to form a clurry suitable for spray application. The foaming, agent is Aerosol PK, a proprietary product of the Mearl Corporation, Roselle Park, New Jersey.

The application of this unterial as fireproofing of structural steel on a building then under construction was observed and finished areas inspected. Although a specific compressibility was not of concern in that use, it was known that compressibility was a function of density which, through frequent periodic checks of wet density and a core-lation between wet and dry density, was the cont ~rol used to en- insure proper application. Independant laboratory fire resistance YEF tests of the material were available and showed that it had a

g 4 90000192 .

n .  :

)

c withstood exposure.resulting in. temperatures of somo D00 F

! .(j.((L 'in'the metal to which it was applied without separating from  ;

L~ the metal during the 2-1/2 hour fire test and the subse-- ,

quent'2-1/2 minute hose stream test.  ;

r 4.3' Develonment and Preliminarv Testing  ;

Certified Industrial Products, Inc. of Hillside, ,

1:ew Jersey, a firm specializing ~ in test and ' development -

of light weight ~ aggregate - concrete and insulating materials, '

was engaged to test and modify.the product.to suit the'rc-

~

quirements.. Certified's developments were-to reduce the strength of the standard product by reducing its -dry density from 12 to 8 pounds per cubic foot through an increase in .

foam content, to improve its stability by substitution of-longer fiber asbestos, and to develop a sequence of.applica- l tion requiring a base coat of 1/2 inch of standard Fire-D r followed by two coats of reduced density Fire-Bar to  !

achieve optimum adhesion to the vessel and minimum slump  ;

beforo curing on vertical surfaces. ,

4.4 Commercial Practicability Evaluation f' , The All Purposes Fire Proofing Company used the

. mofified formulation with their production equipment to spray test panels and to judge the commercial practicability -

. of the modified produc,t. They experienced no difficulty..

4.5 Final Development Tests '

The United States Testing Company was engaged to perform independant laboratory tests of the compression characteristics and stability of the test panels made with '

the production equipment. Compression characteristics; a ,

1

'CB6 reduction in thickness at the 3 psi fluid concrete pressure and reduction in thickness from 2-1/2 inches to 1-1/2 inches under 15 to 20 psi; were satisfactory. ' Inelastic compression,  ;

checked by measuring rebound, amounted to 80% of total com-

. pression; continued measurements by United States Testing over 3 months showed stability after the recovery of 20%. Material loss.after compaction was measured on panels compressed in

- a vertical position; loss was about 1% of sample weight, observation indicated loss to be occurring at the br.eak in l the samples at the perineter of the compression shoe, a l discontinuity which would not occur in service.' These tests  !

ggj,3

were completed in October of 1965.

q: '

. c.

900.00193 D"" *B 5

w e> ev .b . :o l l

. y re e 4 4*%% m y y &

Islthough the decision to use the asbestos fiber -

=g magnesite cement product was based on the tests described i 4 above, it was considered desirable to accumulate a greater -

' body of test data including compression tests at elevated temperature.

The second series of test results, also obtained by U.S. Testing, available in April 1966, demonstrated that the previousl; established compression characteristics were also applicable when tests were made on the material at 300 F.'

4.6 Rebound Evaluation The Mcarl Corporation, Foam and Chemicals Division, supplier of the proprietary foaming agent was requested to describe the action of the foaming agent with reference to the demonstrated 20 percent rebound. They deceribed their ag'ent as an aqueous protien base material capablo of entrapping air and thus imparting a cellular structure but.vihich would be absorbed by the solid materials '

, upong drying, hence would contribute nothing to the strength of the material. The cellular structure imparted by the foam would be maintained by the magnesite cement but upon (

e-being compressed would collapseirrecoverably +through failu're .

of the brittle cement walls of.the cells. The rebound was attributed to recovery of_, compression of,,th.e as_ bgg. tor fibers. They also noted that wet asbestos fibers would pack under compression and suggested addition of moisture if feasible as a means of reducing rebound.

4.7 Susceptibility to Damaae The question of susceptibi:lity of the material to damage was being considered in the course of the testing program. It had been established in the original development of the material that about 50% of the strength of the material (crystallation of the magnesium compounds) occured -

in the first 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and 100% would be developed in 7 days. .

The 50% st'rength would preclude damage by rain and sufficient- , .

drying would take place in that 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period to eliminate 't the possibility of damage due to freezing in this extremely '.

porous material.

In order to avoid the material drop off characteristic of sharp breaks within the, material,it was considered

,.mm desirable to insure that the material was firmly bonded to D the, s, teel surf ace and free to separate from the concrete D lD lD 'SI' Ih u M e M A%

., 6 90000194

u s

v

' during the planned cubsequent operation of reducing its $

(E' thickness' hy ccmpressing it 'betwoon these two. surfaces. .

[

t Furthormore,, the exterior surf ace of the compressible material, [

'being porous, required a waterproof covering to prevent I absorption of grout from the wet concrete'to be placed' I against it.

f

.+

L 4.8 Protection and Bond Breaker  %

?

Accordingly,. protective measures considered j necccuary were to cover the new work for a 24 hour2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> period 5 with a weather enclosure consisting of polyethylene sheets }

drapod over the scaffolding to permit circulation of air '

to promote drying. After the initial drying period the i polyethylene sheets would be placed against the material [

with all cages sealed by tape and would be held in place {

by insulation pins stuck on,the steel vessel before appli- [

cation of the compressible material. .

Some mechanical damage due to subsequent operations of constructing the concrete shield wall had to be expected and provisions were made in the contract for repair of such dam' age. The extent of the damage was expected to be minor since the first subsequent operation would be to erect g a curtain of reinforcing steel at the inner f ace of the - a concrete wall which would provide a " fence" between the .

material and'further operations. This assumption was valid for. vertical or near vertical surfaces of the vessel.

However, a splice in the wall reinforcing just above the clevation 51' - 3" floor, the delay in construction of the L concrete wall there while the floor was being built and the condition that a tangent to the vessel at this elevation was no longer near vertical all contributed to greater opportunity for mechanical damage to the compressible material u 1

in this region. Damage sustained was such that essentially all of the material to some 10 feet above the 51'-3" floor -

had to be replaced. This experience indicates that -

additional initial cost for protection against mechanical' damage, such as for a fabric reinforced sprayed-on resin shell, could probably be justified in this region.

49 conclusion

! The tests and evaluation indicated that the foamed I

asbestos fiber-magnesite cement product had the required k! compression characteiist[cs and stability, would be unaffected f"l 1

by.long term exposure to radiation and heat, would be commercially I

l' ,; 2

J

. e .

. practicable in application,.would bc1as-resistant'to weather conditions and damage as could be. expected considering the

'g~g required, compression characteristics-and could'be repaired by local replacement in'the event of-damage'. Thus no tech- -

nical objection to the material was found and it was de-cided-in October, 1965 to proceed with the inelastic com-pressible material.. approach to the providing of the de-sired expansion. space around the drywell vessel.

^

5. DETER:1IMATION OF REQUIRED THICKNESS OF THE LINING

. 5.1 Concentrated Load Test of Vessel To determine the maximum acceptable distance betwcon the steel vessel and the concrete wall, load de-

.flection tests were made by the Chicago Bridge & Iron Co.

These tests, described in the CD&I report: " Loads on

- Spherical ShcIls", August, 1964, were to determine the maxi-mum deflection consistant with no rupture but accepting yielding, e of procotype sect' ions of the vessel, under a load simulating the postulated jet force accompanying the mca.

Two of the'three tests run were terminated without

{'

rupture (due-to limitations of.the testing apparatus) at de-

. ficctions of 3.3 and 3.25 inches of deficction. The third

' test was terminated with the development of a crack.at a deficct'.on of 3.125 inches. Conservatisms in the tests in- '

.cluded: Distribution of the load over only 1.08 square feet

.of plate whereas the pipe, the failure of which is the postulated source of the jet has an area of 2.54 square feet

'in the cylinder zone of the drywell and 3.14 square feet in the.sphore; use of flat insert plates in simulating a pene-tration for the one test where a crack developed rather than plates dished in the direction of load to conform to the curvature of the main shell plate; and application of load as a static load at normal ambient temperature rather than as an impact load with steel at the higher operating tempera- l ture of the vessel. ,

. I! .

5.2 Preliminary Vessel Expansion Procedure In order to determine required rainimum thickness of the lining it was necessary to establish the extent to which it would be compressed. This amount would be determined by ' the expansion of the veshel associated with its highests

p. postulated temperature for any future operating or accident -

F condition not concu: rent with high internal pressure or by the procedure planned for expanding the vessel to create the

. . air gap if this procedure would expand the vessel to a g gpg xp'ent than the future conditions.

~

Yn "B 90000196

~ s N L%aE L~ . .

i

The simplest and therefor probnbly most reliable-g.g expansion proceduro appeared to be to release steam from a tig temporary boiler into the vessel until saturation conditions

- at a cuf ficiently high pressure to protect .the vessel against buckling prevailed and to hold this steam pressure until the vcscel metal temperature reached the steam temperature. / A suitable internal pressure was 35' psig which is associated with a saturated steam temperature of 281 F. Expansion at this condition would exceed postulated accident or operating expansion, hence it became a criterion for detennining ,

lining thickness.

5.3 Linino Thickness At the most critical location, the point on the I sphere most distant from the embedment, thermal expansion at 2810 F was expected to be 1.057 inches. Tests on the compressible material reducing its thickness this much, plus I that due to compression from the fluid concrete pressure, from j an initial . thickness of about 2 1/2 inches set the design ,

pressure which would be transmitted to the concrete wall during l

. vessel initial expansion at 20 psi. Some tolerance on thick-ness of compressible material had to be allowed; a feasible .

limit was plus or minus 1/4 inch. Since the design pressure on the wall assumed 2 1/2" minimum, a thickness of 2 3/4  !

{ inches plus or minus 1/4 was i ndicated. l l

5.4 Vessel to Wall Distance '

In considering the ac,ceptability of the 3 inch  ;

upper limit of tolerance in relation to the maximum acceptable distance between steel and concrete, it was noted that the ,

initial 3 inches would be reduced by; the compression of the material under the fluid concrete pressure, the thermal ex-pansion of the vessel in going from ambient temperature during construction to an operating temperature at which the mea

, could occur, and the fully compressed thickness of the material.

~

These conditions were expected to reduce the 3.0 inches t.o an amount adequately below the 3.125 minimum fail-ure deflection of the CB&I jet load simulation tests, particu-lary in view of the many conservatisms in those tests, It was subsequently determined by tests and measurements during the expansion operation that these con-ditions would reduce a 3.0 inch gap as follows : ,

("=""=" o m 9 g8 3 "D 1o o k LA  %@ . 90000197

I

(#Th

=~ '

Compression under~ fluid concrete pressure: .10 inch vessel thermal expansion @ 100 F: .15 Residual thickness of lining: .25 TOTAL .50 Maximum' Reduced Thickness: 2.50 inch.

6. APPL 7CATIodi oF LIMING MATERIAL .

6.1 Lower Recion ,

The lower region of the steel drywellcontainment vessel from its invert at elevation 2'-3" t.o elevation 8'-llh" is embedded inside and out, in concrete to form its permanent support. The concrete fill is continued inside up to elevation 10'-0",with a curb to elevation 12'-3".

Outside, the vessel is rest rained.by a sand filled pocket -

from. elevation 8' 11 " to elevation 12' - 3" to form a transition gone.

(_ Erection and testing of the drywell steel con- ,

tainment' vessel was completed and the first floor construction was scheduled to be started prior to the final testing and decision to proceed with. the inelastic compres'sible material. Since the vessel expansion in the lower region due to increased temperature and pressure is relatively small i (in comparison to that at the sphere-to-cylinder transition some ,60 feet above the point of embedment) the compression characteristics of a compressible material for this lower region were not as critical. Owens-Corning Fiberglas S F I Vapor Seal Duct Insulation had the required stiffness to '

resist the fluid concrete pressure; an analysis of the con-tainment vessel by the Chicago Bridge & Iron Co. demonstrated that.the load imposed on the vessel to compress the Fiberglas (about 6 psi) was acceptable in this lower region.of heavier vessel plates without any balancing interna,1 pressure. In order to proceed with the concrete construction through the first floor prior to reaching a decision on the inelastic material, this 2" Fiberglas board was applied to the vessel up to elevation 23'-6". The material was furnished with

- a factory applied laminated asphalt kraft paper waterproof exterior face, and was attached to the vessel with mastic Ihb

• and insulation pins; joints between boards, and'cdges and ,

penetrations were sealed with glass fabric reinforced mastic.

10 90000198>-

ih 1:.

.- 2

'j II;

~

6.2 Access for nonlication to the Vessel e

s It had been anticipated that the material. appli-cation vould proceed step-wise with construction of the con- y crete-shield wall. The general contractor with whom negoti-ations for the concrete work'were being conducted advised f{

that his schedule wou_ld be a month longer if the work were to proceed in_this manner. Burns and Roe's Construction '

- Management:in discussions with All Purposes Fireproofing g and a scaffolding specialist, the Chesebro Whitman Division {

of-Patent Scaffolding, developed an arrahgement Whereby '

the. material could be applied in the latter stage of the ...

Reactor' Building First Floor construction contract thus reducing i the time rcquired for cons' ruction c above by a month. .

The radial steel trusses spanning from the center drywell support pedestal to the exterior foundation walls of the building, erected above the first floor to support  ;

. 'the first floor forms, provided support and anc?crage for. -

scaffolding at a base elevated above - the floor thus avoiding significant interference with the first floor construction.

The scaffolding all around the vessel extended from that -

base up to the vessel flange and was cantilevered in, fol- #

t lowing the curvature of the' top hemisphere,to provide access to all pa'rts of the sphere and cylinder without support 'from p the vessel. f 6.3 Application ,

Y The mixing and foam injection was done at grade and the slurry was pumped to the point of application starting ,;

at the. top flange. The material was built up in 3 coats in y accordance with the procedure developed by certified Industrial P Products, the first coat 1/2 inch of standard Fire-Bar, ,

the balance, 2 coats of Fire-Bar of reduced density to make up the total specified thickness of 2-3/4" f 1/4 inch for y the upper hemisphere and 2-1/2 inches + 1/4 inch for the 1 lower hemisphere and the cylinder. The actual application of the material was completed in about two weeks' during-the latter part* of May and early June, 1966, and proceded -

[

without difficulty other than'some washing c ff of inadequately l protectod'new work by rain. On that occasion inspection

. of cther exposed work Which had cured 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> or more showed  !

no damage. )

• ,l

(*. ' '

90000199 6

11 4

(

i 6.4 nond Breaker d+:

== At was desirable from the schedule point of' view to apply the bond breaker as soon as possible to permit re-moval cf scaffolding. Development test samples were essentially dry in about 1 weeks but, as noted above, the cementing reaction I was completc in 7 days and was 50% complete in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />'. -

l Since' the lichtest density material which had been found to be practicabic required al5 to20 psi load to be compressed the necessary amount rather than the ideal 10 psi, there was no need to be concerned with achieving gain in strc agth l after 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. Leaving the surface unsealed to facilitate I removc1 of excess water from the material was not particularly desirable since rebound would- be less if the material were i 1

to be compressed in a moist condition. The only disadvantage i to early covering.would be the reduction in value of the

~

l damp material as insulation during the operation of expanding' l the vessel by heating. Schedule being of paramount impor-  !

tance, the bond breaker was applied af ter a minimum period )

of 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br />. Material used was Grif folyn 4 mil clear poly-

~

ethylene sheet reinforced with glass fibers, attached by i means of speed wi.shers on the nylon insulation pins previously stuck,on the vessel at 4 foot intervals each way. Seams

_ - were scaled with pressure' sensitive fabric tape. Some

( consideration was given to providing further protection from w

' ind in the form of a light wire mesh (chicken wire) however, the first areas donc demonstrated that the reinforced ~ poly-eth ylene, thoroughly taped, provided a stable 3 tight cover, and no further measure appeared to be justified. Prolonged exposure through the winter resulted in strippin'g of the

' polyethylene from the cylinder in .high winds; loss of ,

compressible material was limited to spots where insulation pins'were pulled from the vessel; the spots were patched with previously cured plugs of compressible material and the covering was' replaced as the concrete construction proceeded.

6.5 Repairs As noted above, extensive mechanical damage was sustained above the 51'-3" floor level due to the circum- -

stances of extraordinary exposure to other construction operations there. The damage involved not loss of material but a reduction in thickness from the pressure of workmen and reinforcing steel leaning against it. Repair was accomplished by stripping the compressed outer layers but leaving the base coat from which the required thickness was

@p5

~

F. again built up by spray application in layers as specified for the original work. Testing of production samples described i

below had shown greater strength than previously anticipated hence in the replacement work, wet density was hold at the 12 90000200

1 ; m .

J

)

.g .

1ower limit : of ? the range; usedl in . the original work andL '

Eg? -thicknes s" was ' increased tol the ' upper : limit. ,

7.. PRODUCTION INSPECTION

=r '

'The' United States Testing Company, having' pr$vious experience' with the, material and having an ' operation at the j

'jobsite (for concrete testing) was engaged'to provide l inspection. services. .

j

~

', 7.1 D'ensity The' development tests-had-demonstrated that the , l desired compression characteristics were achieved by the formulation producing . a material having a dry' density of 8 pounde per cubic' foot and that this dry density corres- ,

ponded to'a wet-density at the. spray' nozzle'of 29 pounds

. per, cubic' foot. Henceithe principal inspection offort'wns

. directed toward monitoring wet density and of course, the- -

, gaging of thicknes's. Density was checked approximately s

. hourly ; at' the infrequent; deviations from the acceptable 7 ,

( range of 27 to 31-1/2 pounds .per cubic foot, work was stopped l_. 'a'nd"the variables affecti,ng density, the foam and/or water con- l

,1; tent, were adjusted until satisfactory density was demonstrated by another check 7.2 _T_hickne s s Thicknesswasga'edduringthe.sprayingby.obberving g

the projection of th'e uniform length insulation pins stuck on.the. vessel at 4 foot intervals and was checked by probe by United States Testing. Deficient or excess thicknesses were corrected as the work progressed. Thickness was'sub-sequently. checked by Burns and Roe before release to the general contractor for each concrete' placement against the material; no out-of-tolerance thickness was found in these subsequent-inspections. Initally no record of actual thickness' was made other than that. it conformed to the specified tnickness within the tolerance. It was later decided,

• after about a third of the concrete wall height.had been place'd, that a knowledge 'of actual material thickness might serve some future purpose; hence a record of measured

,. . thickness at.about 3 foot intervals was kept but only at r and above elevation 47 feet. ,

1 .13 90000201 .

9

?

O

---

__' . * . . ..m., ..._.,,-e .,r,_, p-_. .w,., ,,,m..,p3,

-ar-- 5- a. -, .--,f w y-

8. PRODUCTION SN4PLING AND TESTING lI g... . There was little question that the material',

45 properly formulated and applie'd, would have sufficient )

strength to, resist the load to be applied during the placing l of wet concreto against-it without excessive deformation '

since the~ development testing had shown'that resistanc,e to compression associated with vessel expansion in the order of ,

twice.the ideal would have to be accepted. However, with ,

the concrete wall designed for a 20 psi load expected during the operation of compressing the material, some check was l indicated to insure against greater resistance to compression.  !

8.1 Samples Twenty-one samples of the material were ob-toined as the work progressed by spraying into 18 inch square y plywood boxes located on the work platform at the point of application, to the depth in the box equal to the thickness  !

being applied. The samples were taken at intervals such l that eac'h would represent the material applied to a specific l area of the vessel, and were labeled with elevation and azimuth to identify that area.

_ 8.2 Tests

(

Since the concern was to determine stress-strain characteristics at maximum strength, the compression testing of these production samples was deferred until they were thoroughly cured and dried.' Tests were performed in August,.

1966 by United States Testing using the same procedures and -

equipment previously employed in testing of the development samples.

The variation in strength between different samples was wider than expected considering that the previous development samples had been made with the same equipment; resistance of 9 of the 21 samples at 1 inch of compression exceeded the 20 psi maximum observed in the development samples and used as.a design load for the concrete wall.

8.3 Off-Wall Samples I A question of whether the samples were representa-tive of the work in place ~was largely dispelled by subsequent test of two samples cut off the wall,of the vessel. These

. two tests also demonstrated the variation shown by the (pr .

14 90000202 I n --

I-

.v

=-

g;. -

production samples and the more resistant of these two.also exceeded 'the maximum strength of the development samples.

f K

=~

R 8.4 Penetrometer ....

.  ?

The production samples, their resistance to y compression having been established by the tests, provided [

a means of calibrating a hand penetrometer which was used-

'to spot check the initially applied material and l

i 7

to check the strength of replacement material applied to E arcas where the original lining was damaged and removed.

- h w

D

9. EVALUATION OF PRODUCTION SAMPLE TESTING [

.Y u

In order to evaluate the effect of the greater indicated strength of the material, vessel expansion, ,

as it would vary with distance from the vessel embedment, I was calculated - still assudling the heating of the' vessel [

to the 281 F which would permit the relatively simple vessel' -

cxpansion by steam heating. Comparing vessel expansion at , ' , f

. sample location with strength of the sample for that location [

and discounting a:'articularly p high-strength sample because

.the material it represented was damaged and replaced, thel

~

, y maximum load on the wall was indicated to be 28 psi. l ,jj E

Althopgh 20 psi had been established as a i design criterion for the concrete wall, this criterion

• 5 was not controlling in the design. An analysis of the rein- g forcing provided to resist the loading condition at mea; &

dead and live' load, thermal gradient and jet force; showed 5 a capacity to accept a pressure of 31 psi without exceeding

{

normal. allowable stresses. (An increase in allowable stress  ;

on the order of 33 percent would be consistant with accepted practice for this temporary condition.) [

r While the numbers er'sulting from the material tests supported the preliminary-plan to expand the vessel with saturated steam, the margin for error was small and  ?

the disparity between individual test results was large. $

Under these c,ircumstances it was prudent, from the viewpoint of protecting the vessel and the concrete wall from overstress, i to consider alternate means of heating and pressurizing which y

, would control the heating to the lower temperature actually ,

required, um .

k E" -

. 90000203 4

. 15 1

• l

10. REQUIn2D VESSEL EMPD.NSION CONDITIONS

- p.=ss In order to consider means of he- .ng and l

pressurizing 'the vessel alternative to the pr. tminary selec-tion of saturated steam, it was necessary to a termine the ,

amount.cf vessel expansion required.

The minimum extent to which the drywell vessel  !

would be required to be expanded would be that expansion which would equal the thermal expansion associated with the .

conditions causing the. greatest difference in vessel temperature postulated for any future op'erating or accident condition not accompanied by concurrent high internal vessel pressure pins some allowance for possible, inaccuracies -

in assumptions.

10.1- Maximum Vessel Temocrature Chance i

conditions not concurrent with a concurrent with a high internal drywell vessel pressure such as a malfunction of the containment cooling system could cause the temperature of the drywell atmosphere to rise to such a level that it would be necessary to shut down to correct the condition.

(

thile the~ shutdown temperature had not been established it

. would be in the order of 1600 to 1700 The drywell atmosphere temperature which was expected to persist af ter the relatively rapid. (17 minute) rise and fall of temperature and pressure expected to accompany the mca fill within this range, at 162 F.

.Th.is temperature, 1620F, was used as a basis for determining the minimum extent to which the drywell vessel would be required to be expanded.

The most severe temper',ture difference would .

occur if the abnormal increase in temperature were to be ex-perienced during start-up from a completely cold condition at the coldest time of the year. It was considered adequately conservative to assume that the temperature of the vessel

.and surrounding concrete shield wall within the completely enclosed Reactor Building would not be lower than 50 F.

With the insulating effect of the compressible material and the air gap that will have been created before start-up, the concrete wall temperature would not have increased materially before tha assumed abnormal increase in temperature

~

at start-up, hence the vessel thermal expansion to be provided

• for must be that associated with a change in vessel teaperature dI.; from 50 F to 162 F. It would, of course, not be necessary to start the expansion operation at 500F or any specific temp-craturc but only to provide for a temperature difference of i 'the 1120F.

. 16 90000204

4

.4m.

Y$:r .

10.2 Allowance for Rebound ,

Sin'ce the selected compressible material is

~

not: perfectly-inclastic, an allowance for rebound..had to be added. As Comonstrated.bv the tests described'above, inolastic compression stabilized at 80%_ of the reduction in thickness of the material. With temperature and expansion a

. linear relation, the required expansion, in terms of tempera-ture difference, would be.112/0.80 or 140 F.

10.3 Internal Vessel Pressure criteria It was intended that the pressure to be maintained within the vessel during the expansion operation to protect it against buckling be substantially greater than the external pressure imposed on the vessel by the compress-ing of the asbestos-cement material.

The pressure would of course be limited by the. . -

design pressure for the vessel. However, the discontinuity

(~ ~

stress at the point of embedment due to internal pressure would be additive to discontinuity stresses at this po' int due to thermal expansion and to the external load to be imposed by the resistance of the lining. Hence the pressure .

should also be subject to the limitation of maintaining discontinuity stress in the embedment zone to within the allowable consistant with the criteria for the vessel, ie with ASME Section VIII and Nuclear Code Case 1272N.

Furthermore, net positive internal pressure would expand the vessel and therefore the thermal expansion need not be ,

that of the entire 140 0 F difference derived above but I some lesser amount which with the expansion due to pressure, -

would be equivalent to 140 0 F thermal. l Thus,in order to assure that external pressure did not exceed in'ternal pressure, to determine discontinuity )

stress and to evaluate the excess internal pressure available j to expand the vessel, it was necessary to evaluate the external i pressure v/hich would be expected to be imposed by the compressible material during the expansion operation.

4 .

90000205 17

. t

- 10.4 Evaluation of F.xternal Pressure

-GE= ' >  ;

=::

• External pressure would, of course, depend on ,

j the extent to'which the material would be compressed and on; its resistance to compression to that extent. ,[

\

The maximum movement of the vessel at the ,

critical' point of sphere-to-cylinder transition caused nyl

. a 1400 F_ thermal' expansion or its equivalent, taking into ", i account upward accumulation of~ expansion due to the lower region'of the vessel being embedded, was calculated ~to be  :

0.67 inch. Of'the several compression tests described earlier,

, those on the samples removed from-th'e vessel were consi'dered to be more reliable than tests on samples not actually a part' of the work and more reliable than the penetrometer readings since those were tests of a smaller area. The .

more resistant of the two off-wall samples was in the upper range of strength of all tests hence use of its compressive strength would be' conservative.

'Using that test and considering the increased resistance due to the precompres sion resulting from . fluid ,

concrete having been placed against it, its resistance to j_ compression at ht' e calculated deflection of 0.67 inch was 23 psi. ,

~

Deflection and thus load was expected to de-crease from this amou't n at the sphere-to-cylinder transi-tion to zero at the embedment.

The fiberglass applied to the vessel between e.pbedment at elevation 12'-3" and first floor at elevation 23 '-6" had a lower resistance to compression than the asbestos-cement material; in this region of expected relatively ,

small vessel deflection, the stress-strain characteristics  !

of the two materials were assumed to be the same for the purpose of evaluating exterr.al pressure. l Selection of Internal Vessel Pressure _

10.5 To select an internal pressure consistant with allowable discontinuity stress'at the embedment, the stress due to this load from the compressible material, with the other concurrent loads, other than internal pressure, was ,

combined with stresses due to a range of internal pressures. .

(;?h -

90000206 18 l

-)

l

. Frcm:this range, a value of 40 psig was found to nroduce with jgg wr the other concurrent loads,ca discontinuity stress within j z

allowable. -The other concurrent stresses included those  ?

' due to the fluid pressure-Lof concrete (this_being considered  ;

a residual stress since no opportunity prior.to formation of~the air gap had existed which would allow the deflection of the vessel induced by that load'to be relieved) :the j

, thermal expansion and the normal dead loads. '

j

'l 10.6. - Expansion Due to Pressure "

Having this selected'an isternal pressure to. ,

be used during the' expansion operation and having evaluated the external pressure, it was then possible to make a judge .

mont,as to how much expansion due to excess-internal-pressure )

would ~ be considered as making up a .part of the required {

equivalent of 140 F of thermal expansion.

j For this. vessel the' radial growth caused by  ;

one psi of internal pressure.is roughly equal to the thermal ~

radial growth due to 1 F of temperature increase or about '

0.003 inches. However,.due to such factors at the restraint- 7 on mechanical expansion imposed by the heavy plate at the k

sphere-to-cylinder transition and the uncertainty as-to the actual resistance of the compressible material, the amount of expansion due to pressure to be considered effective was arbitrarilg set at the low value of 10 psig or the equiva-  ;

lent of 10 F. Thus the temparature difference requgred to be achieved in expanding the vessel was reduced to 130 F.

i 10.7 conclusion ,

e .

i e As developed above, the conditions rcquired to be achieved in the oneration of expanding the vessel were ,

established as a 130 F increase in temperature of the metal over the c*ncrete o shield wall. temperature during the operation, ,

at a vessel internal pressure of 40 psig.

f 90000207 '

ve' ' g 19 I

I,

, -y qeg y- wt - ytM"1-T T'* * '*8'N#* " ' " *I" "'i ' * ' " " ' '* ' '"

11. SELECTION OF METIIOD OF EXPANSION Expansion temperature and pressure were established at 1300 F ove; ambient or about 180o F, and 40 psig. T'o accomplish the expansion using saturated steam at the 40 psig pressure desired would result in heating to 2370 F, the saturation temperature of steam at this pressure,i.e.more than 1000 F above the temperature actually required. -

V'  !

Pressurizing with air was_thus,indipated and various methods of heating the air.were investigated.[

It was desirable to accomplish the expan-sion operation over a weekend to minimize disruption of other construction activities To allow time within the weckend for checking, pressurizing and cooldown after- .

I heating; it was desirable to limit the heating period to l about 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br />. A preliminary calculation of the heat in-l put required to bring the drywell metal temperature to 1800 F in this period showed minimum heat required to be l in the order of 24 million BTU.

This required heat input eliminated the

(- possibility of the compression of the air providing I sufficient heat as a practicable procedure. The use of oil or, less objectionably, gas-fired heaters within i the vessel would be complicated by the necessity for

• I remote operation, operation under pressure, providing oxygen, and the handling of combustion products. Such heaters outside the vessel involved a rather extensive, ,

relatively high pressure duct system.

Electrical heating within the drywell appeared  !

to have none of these disadvantages other than that of remote operation which was least objectionable with this method and the power requirement, indicated to be in the l order of 500 kw, was well within the 2000 kw of temporary l power availaIble.

Electrical heating'of the compressed air

' atmosphere within the drywell was selected and a detailed j equipment list and operational procedure were developed.

g .

90000208 20 1

..s

12. EXPANSION EOUIPMEMT AND ARRANGEMENT ,

ts. ,

12.1 Pressurizinc 4

Compressed air was provided by a bank of three 600 cfm, engine driven portable compressors located outside the building and connected to a common 6 inch line.

. The line was extended into the vessel through an existing 10 inch vessel penetration (No X-67 ) just chove outside grade, some 17 feet above the concrete fill floor inside the drywell. Valves for bleeding off air were installed, a 3 inch at the supply and a 6 inch at another penetra-tion (No. X-3A) just below the top flange of the v7ssel.

, Pressure guages, 0-100 psig, 6 inch dial, were installed at the two bleed points and in another penetration (No . X-4 5 )

opposite the supply point at about that level.

12.2 Heating Eight General Electric duct heaters, GE No. 2A 699 G 5, 60 kw, 3 phase, 480 volt, 60 cycle, were spaced uniformly around the interior of the drywell on the

. platform provided by planking on the radial structural

_ steel' beams at elevation 23 feet, 6 inches. These were

( mounted in unistrut channel frames strapped down to the '

planking.

Four Chromalox duct heaters, Typo TDH 600, 60 kw, 3 phase, 480 volt, 60 cycle were available as spares.

Eight American Coolair Corporation fan units, Model No. C36M, propeller type, 36 inch diameter, for V-belt drive were positioned in the drywell, one behbad each heater. Each fan was driven by a General Electric motor, 5 hp, 460 volt, 3 phase, 60 cycle, 184T frame size, Class B insulation, ball bearings, open drip proof, for belt drive, GE 700 Super Standard Line , SF 1.15. The standard grease for the motors was replaced with Dow DC-44 silicone grease to allow operation in the high temp-erature environment; the standard lubricant for the fans was suitable for.up to 2120 F. The f an and drive units, were also strapped down to the planking.

y=?

<8 -

90000209 21

The fans and heaters were oriented so that the discharge would impingo obliquely on the vessel wall,

.= rhis to promote circulation.and to avoid having the motor 4r f or the ne::t-in-linc fan in the path of 'the air stream directly off -the heater.

The power cabics from all heaters and fans were routed to one existing 16 inch diameter vessel ponctrution (ma. X-66) and through two 3 inch conduits welded into the penetration closure plate. Vessel leak tightness was maintained through use of siliconc putty injected into Crouse Hinds Type EYS sealing fittings on the conduits. .

A separate fused disconnect switch was installed outside the vessel for cach heater and for each fan drive since it was anticipated that if might be necessary to operate the fans without the heaters in the event of uneven temperature distribution or to disconnect individual fans and heaters, in case of malfunctions, -

without shutting down the en' tire operation.  ;

12.3 Monitorinc Temocrature A General Electric Multipoint Temperature Recorder was used to monitor vessel metal and inside air k temperatures. Twenty-nine copper constantan thermocouples l were placed inside the drywell, twenty four taped to the j metal surface a'nd insulated from the drywell atmosphere, five suspen'ded in exposed locations to read air temperature. .

Those to read air temperature were located at three elevations, two at elevation 23 feet, two at 64 feet and one at 85 feet.

Those to read metal temperatures were placed in 8 levels with levels spaced at 8 to 12 feet aprat, 3 per level at 120 degrees apart with those in each level rotated 30 degrees from those in the level below to form a uniform spirai cattern over the entire surface of the v'essel exclusive of t>a 1.e ad .

The wires from all thermocouples were routed to the vessel penetration used to bring out the pcwer cables and were brought through in a separate conduit welded into the penetration c losure plate and sealed in the same manner as for the power cables.

90000210 G. .

22

- ~ - -

+ 9

12.4 Monitorinc Excansion .

,=

=" The existing vessel penetrations, being pipes welded into .the vessel and extending out through the con-crete shield wall through sleeves (thus not in contact with.the concrete) provided accessible pointsLat'which '.

vessel . pansion and contraction could be measured.

Fourtecn dial indicator extensometers were bracketed to the outside face of the concrete shield wall;

.two at'encia of-seven vessel penetrations to monitor horizontal and vertical components of expansion and contrac-  ;

, tion of-the vessel. Flat plates were ta~cked to the penetra-

. tion pipes to provide-true horizontal and vertical surfaces  :

to contact the extensometers. i The seven penetration pipes used were selected  ;

to provide n range of elevations and locations as tabulated l below:

Penetration No. Elevation Azimuth X-66 27' 386 i g- X-2A 27' 1710-15' 1

s_ X-32 36' 125 l X-36 44' 2900 X-63 62' 340 X-12A 62' 240 0 X-19 86' 1900 The temperatures at the outside ends of tr e penetrations were observed, by means of insulated the. mo-meters taped to the pipe. These, averaged with the tempera-tures at the inside end (the vessel temperature), would pr' ovide a basis for correction of the indicated vessel move- -

ment to compensate for thermal expansion of the penetration pipes.

90000211

=.

JnEE r -

23

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

_y-g--.*.ew%v-,vvuwwm,,.,.,.w..ww,, ,S. --

O 13.

PREPARATION FOR EXPhNSION OPERATION

~.... The personnel airlock which had been removed i:.

a to allow use of the larger concentric equipment hatch for con-struction purposes, was reinctalled. The top head gasketed man-hole cover was bolted in place. Several penetration pipes had been opened, the principal ones being the main steam, feedwater and

~ vacuum breaker connections; except for main steam and feed- g

~

wcter, the penetrations were closed to maintain pressure and a reasonable degree of leak tightness by rowelding on l,[

the original caps'. To avoid the extensive welding operation of replacing the dished head closures for-main steam and feedwater, a stiffened 3/4 inch flat plate with 3/16 inch -

rubber gasket was placed against the inside end of each 30 and 36 inch penetration pipe and was held in place by a

1 inch threaded rod extended out through the penetration to an assembly of 6 inch steel channels spanning the diameter of the pipe at its outside end. In the subsequent pressur-izing aperation the substitute closure plate proved to be structurally adequate but it was apparently not possible to ' apply suf ficient seating pressure on the gasket; one -

of the four blew out at below 5 psig and all 4 were replaced by seal welding the plate to the penetration pipe.

_ }

Penetration closures were intended to be y structurally adequate for the pressure and without gross leaks which would significantly reduce internal pressure, however minor leaks.would not affect the planned operation hence inspection before pressurizing was limited to a  ;

visual check. 3 The equipment and thermocouples were installed ii in the arrangements described above and were individually 1 checked for operation and the thermocouples were calibrated. s Both inner and outer gasketed airlock doors were closed. ,

Ik

14. EXPANSION OPERATION PROCEDURE  ! i i{

14.1  !

Pressure Test 1

Before starting the heating operation it ;5 was desirable to obtain, by pressure test, assurance that  !

74 the replaced penetration and access closures wer- safe and , free of gross leaks,hence the first step in the opera-tion was intended to be to pressurize without heat, l[l j;

l 24 90000212  :

initially to 10 psig at which point the replaced closures m, would be inspected and soap bubble tested and then, if '

(s# satisfactory, to increase the pressure to 50 psig, 25 percent over the pressure intended to be maintained l during heating,as an overload test. The pressure was to I be increased to the 50 psig in increments of 10'psig,.' hold'ing at each increment to monitor expansion reflected by exten-

, someter nndings to assure, before , proceeding further, that the vessel was free to expand in response to the internal pressure.

14.2 Pressure .

l A satisfactory overload test would be follo'wed by biceding air to reduce pressure to 30 psig anticipating that expansion of the air due to heating would increase i pressure to a level approaching the intended 40 psig with '

final adjustment to that pressure to be made by pumping '

or bleeding off as required. I 14.3 Heating With the heaters and fans in operation it .

was intended to interupt the operation once to experimen-t

(, tally determine the metal temperature below the desired maximum at which heat input should be cut off in order that continued heat transfer from the air to the metal without further heat input would result in equilibrium at the desired maximum metal temperature. The monitoring of the representative range of temperatures at 15 minute inter-vals would indicate any necessity'to correct uneven tempera-ture distribution; it was expected that operating the fans with6ut the heaters would do this. The monitoring of expansion by local reading of the extensometers initially at 1/2 hour intervals and subsequently hourly (or more frequently if abnormal) would show up deviations from the predicted pattern if and as they developed, to permit evaluation as to whether holding for further investigation or shutdown of the opera'-

tion was indicated.

90000213 4

O lb. .

25 e-

3 14.4 Cool Down 4

[i[l Upon reaching the desired temperature it would be' necessary to cool down with pressure 'on until the vessel had co'ntracted sufficiently (approximately 20 percent of expansion) to assure that the rebound of the

• compressible material'would not impose an unbalanced

• external pressure on the vessel.

14.5 Initial Computer Program On the basis of the rated output of the heaters selected, the total weight of material to be heated and assumed conductivities based on the character of the sur-rounding materials, a computer program to predict the '

time-temperature behavior of the vessel during heatup was developed and printed out for 10 minute intervals for use as a guide to progress.

The ini'tial computation indicated that the {

heating 0

of the vessel 130 F 0 to the expected. requirement of I 180 F would be accomplished in 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br />.

15. MARCH 3-6, 1967, EXPANSION OPERATION s

15.1 Recuired conditions l

, Preparations were completed at 8:20 pm Friday, March 3, at which time the concrete shield wall temperature was observed at mid thickness and over a rangs of elevations and was found to be at an essentially uniform temperature of 43 F. With the required increase established as 130 F, the objective of the expansion operation was to obtain.a vessel metal temperature of 1730 F at an internal pressure of 40 psig. i 15.2 Pressure Test (March 3, 8:20 pm to March 4, 1:00 pm) 90000214 (px .

26

15.2.1 Pressure

,; Pumping started at 8:20 pm but was

'a= shut off at 10:00 pm when the pressure, then less than 5 psig, was lost due to the forcing out of the gasket on the temporary closure of one of the main steam and reactor feedwater penetrations. The gaskets on all four were then removed and the closure plates at the inside ends of

, the penetration pipes were seal welded to the pipes.

The pressure test was restarted at 2:40 am, March 4; with the three 600 cfm compressors in operation, pressure increased at a rate-of 6 to 7 psig per hour. The full overload test pressure of 50 psig was reached at 11:00 am and was reduced by bleeding off to 30 psig at 1:00.pm, 15.2.2 Temoerature Minor increases in temperature were, obtained from he warm (98 0 F) air introduced over the 8 +

hour. pumping period; air from 50 to 54* F, metal from 40 to 47 0 F.

- The bleeding operation (expansion of

( the remaining atmosphere) resulted in decreases in temper-ature, air to 390 F, metal to 430 F. Pressure was reduced at a rate of 10 psig per hour.

15.2.3 Expansion Extensometers were read and recorded at 5 psig increments and decrements of pressure (approxi-mately 45 minute intervals). Movements corresponded reasonnbly well with predictions except for vertical move-ment at elevations 62'-0" where 0.12 inch vertical move-ment at 40 psig was calculated. Penetration No. X-12A at azimuth 2400 showed essentially no vertical movement s while Penetration No. X-63 showed 0.32 inch at this pressure.

Another inspection of the space between Penetration X-12A and its sleeve through the concrete wall assured that the pipe was completely clear of restraint from the wall.

The other observation points, particularly the one above, Penetration X-19 at elevation 86', confirmed general vessel movement in accordance with predb dons. No explanation was apparent other than the general observation that these penetrations were immediatly below the 2-5/8 inch thick e';

,, 90000215

- - 4

• .o sphere to cylinder transition whichi as a restraint, would 2

([j locally. influence the curvature of the. expanding vessel. ,

To attribute'the cpparently. abnormal readings to response  ;

of the vessel discontinuities to pressure loading, requires  ;

that the different size and different reinforcing details  ;

for these two penetrations, both at elevation 62'-0", ex-plains their different. response. The 9' length of pipe ,

between' vessel and gauge point as a lever arm would magnify the effect of a change in vessel curvature perhaps enough ,

to obscure actual ~ vessel movement. ,

Since'the. vessel itself had previously been successfully overload tested, it was considered enough, for the purpose of this' operation, to verify that "abnonnal" readings were not due*to restraint imposed by the wall; the other observation point readings and the inspection established

-reasonable assurance of no restraint hence the pressure test was continued to completion. Movements of these points during the subsequent heating operation were nore in line

'with predictions thus providing a further indication that the " abnormal" readings were characteristic of the manner in which the vessel responded locally to pressure.

15.2.4 Leakage

(

The penetrations which had had test caps're-moved and replaced or modified were soap bubble tested starting at 10 psig. Small leaks were observed in all of the replaced vacuum breaker to vent pipe penetration closure welds and in the weld of the conduit for thermocouple leads through the cap on Penetration X-66, at 10 psig. At 25 psig a small leak developed in one of the main steam scal welds made at the beginning of the test and in the other main steam and one feedwater at 40 psig. The leaking welds were inspected and judged to be of no consequence to the expansion operation.

15.3 Heating with 480 kw Cacacity (March 4, 1:12 pm to March 5, 12:40 pm)

- 15.3.1 Temperature All heaters (8 at 60 kw) and fans were started at 1:12 pm and were all operated continuously over this period.

Initial conditions were air at 39 2, metal at 43 F. Metal i% temperature increased at 3 F per hour until early 28 90000216

March 5 (about 1 am) then dropped to 1 F per hour. Heat-

, g. ing was stopped at 12:40 pm, March 5 to install additional

";~

heaters since the rate of progress indicated an impractical length of time to complete. Conditions at shutdown were air at 107 F, metal at 92 0 F.

15.3.2 Pressure Compressors were not operated during this period.. The pressure increase due to heating over the roughly 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and 68 0 F change in air temperature was from the initial 30 psig to 35 psig. To shutdown the operation for installation of additional heaters it was not considered necessary to cool down before reducing '

pressure since metal temperature was only 520 F above the initial condition hence vessel co' traction due to pressure reduction and the expected tempe ature drop during blow-down would exceed rebound of the lining material. Pressure was reduced from 35 to O psig between 12:43 and 2:00 pm.

15.3.3 Expansion Extensometers were read and recorded ,

~

at intervals of 1/2 hour during the 1st 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> of the

(' heating operation and hourly thereafter. Movements corresponded reasonable well with predictLons, including those at the

. pointe at ele.vation 62' Where the apparently abnormal . movement' during the pressure test was previously observed, thus rein-forcing'the conclusion that the previous readings were charac-teristic of the manner in which the vessel responded locally to pressure, as noted above under " Pressure Test".

15.3.4 Other Observations Snow began.to fall on the exposed top head at about midnight, roughly the time when rate of tempera-ture increase dropped from 3 0 F per hour to 1 F per hour.

Arrangements were made to cover the head in the morning and tarpaulins over planking were in place at 10 am, March 5.

15.4 Heating with 720 kw Capacity (March 5, 12 :40 pm to March 6, 8 00 am)

The operation was, suspended for 4h-hours while

. grq. the 4 spare 60 kw Chromalox heaters were installed at alter- -

# nate locations of the eight 60 kw GE heaters.

90000217 29

15.4.1 Temperature hh:.

Me al temperature at 5:15 pm when heating a

was resumed was 840 F, a drop of 80F 'during the shut-down. Considering this temperature level, the rapid pressure increase to be expected, and the slow temperature increase,  !

it was not considered necessary to await pressure build- j up before starting heating. The 920 F maximum metal ]

temperature reached before shutdown was recovered in 3-1/4 hours (pressure at that time had reached 25 psig' and air

-inside the vessel was at 122 F.) After compre'ssors were off  ;

(input of relatively cold air stopped),' rate of temperature increase was 4 F per hour briefly, until 11 pm When it dropped to 2 to 2-1/2 F per hour. Heating was interrupted for one i hour, 4 am to 5 am, to repair a smoking connection on the

1. 1 main power cable.

At 8:00 am Monday, March 6, the ,

metal temperature was at 1120 F and was increasing at a i rate of 2o F per hour. The extent of time which would be needed to complete the operation at this rate and the consequent continued suspension of other construction activ-ities coupled with the changes which exper'ience had shown would improve the rate and which could be made if the -

q operation were to be interrupted led to the conclusion that l the operation should be suspended at that time and be resumed the following weekend. l l

15.4.2 Pressure Compressors were started at 5:15 pm and, in. order to observe What improvement in rate of temperature increase the additional heaters would produce, the input of relatively cold air was stopped at 8: 30 pm When the pressure reached 25 psig. If the rate justified optimism about completing this weekend, pumping would be resumed to reach 35 psig. Pumping was not resumed. Pressure had reached 26 psig When the operation was discontinued at 8:40 am March 6. It was not considered necessary to cool down before reducing pres'sure; the metal temperature was now 72 F above initial ccnditions; vessel contraction due to pressure reduction and temperature drop during blowdown would exceed rebound of the compressible material.

l 15.4.3 Expansion gg; Extensometers were read and recorded houx1y; movements corresponded reasonably well with predictions.

90000218

115.4.4 Other observations '

k.,<..

G' -

LVol'tage was1 observed to be. extremely low: readings atftime.were: .

421-

-March 5',7 9 pm ,

valts-

'. 10 pm 421.5 volts.

2 . 11 pm; 430 volts .

March-6,- 2 am 400 volts' .. . .

3 am- 400 volts

'S:30 am .'420 volts .

'l

16. CONCLUSIONS FROM'THE MARCH 3-6, 1967 ' OPERATION

.i

< The failure to achieve the' predicted rate of

( progress, heatup to'173 F in 14 hours1.62037e-4 days <br />0.00389 hours <br />2.314815e-5 weeks <br />5.327e-6 months <br />, was concluded to be basically attributable to moisture in the compressiblei material and to the wetting of the exposed' top. head ~ causing greater'than anticipated heat loss.

16.1- MoistureLin the comoressible Material

'( The condition of excessive moisture in the .

compressible material became apparent when compression, forced water out through the sleeves around several of the pene-trations. . Probable sources of moisture'were: mixing water from the spray application sealed in by the polyethylene covering; rain to which-the upper areas of the lining were  !

exposed when the covering was damaged; drainage entering'the H material at the top flange of the vessel and; locally at the contr61 rod drive: penetrations, the wet drilling operation performed to clear the space around those penetrations Other than to assure no penetrati'on of grout during the concreting operation, no particular effort had been made to exclude ,

• . moisture since its effect on compression characteristics would be a moderate improvement through a slight reduction in' strength ~and a lesser rebound.

The moist condition of the material,- contributing t'o less of-heat by decreasing. the value of the asbestos lining as an insulator and by absorption of latent and sensible heat,

,was.not susceptible to change in any reasonable time hence the= alternative of increasing heat input was indicated. Four

,eg 60 kw Chromalox heaters in addition tothe'4 already on the 41" .

job as spares-were availab.lein time to .be used the next weekend.

31 90000219 a - . . , _ . , ---e

16.2 IIeat Loss Throuch Vessel Head Eh

'*" The risk of greater heat loss through the top head due to its becoming a wet surface, as it.did in the . snow of ' March 4 and 5, had been eliminated by covering it with tarpaulins; the condition could be further impr'oved .

by insulation.

16.3 Voltage Another contributing factor to the slow rate of heatup was the effective operation of the heaters at less than rated capacity due to low (400 volt) voltage.

This condition could be improved by changing the trans-former taps.

16.4 Conclusions It' was expected that the operation could be completed in the next weekend in view of the conditions that: 1) the 17 hour1.967593e-4 days <br />0.00472 hours <br />2.810847e-5 weeks <br />6.4685e-6 months <br /> pressure test phase had been completed and need not be repeuted 2) the total heater capacity wa's doubled from the original installation by the addition of eight 60 kw heaters 3) the top vessel head, previously

( exposed, was insulated and protected from the weather 4)

~

voltage to the heaters would be up from 400 volts to 480 volts 5) an air gap had been formed between the lining and the concrete which would provide additional insulation '

in the early phase until expansion to that extent was re- ,

c'ove red .

16.5 Revised Comeuter Program To determine whether this expectation was justified, the computer program of time-temperature perfor-mance was revised to conform to the experience curve de-veloped with the original 480 kw heating capacity. The re-vised computer program was run for the new 960 kw heating capacity. The new run predicted 42 hours4.861111e-4 days <br />0.0117 hours <br />6.944444e-5 weeks <br />1.5981e-5 months <br /> to reach a metal' temperature of 173 F, or well within a weekend. Further-more, the 42 hour4.861111e-4 days <br />0.0117 hours <br />6.944444e-5 weeks <br />1.5981e-5 months <br /> time was an outside limit, barring major disruptions, since it took into account no improvements other than the additional heaters and assumed a 45 F starting metal temperature whereas actual starting tempera-ture would be found to be 58 F.

,=. .

32 90000220

o

• :L i

>= 17. MARCH 10-12, 1967 EXPANSION. OPERATION

.G _

17.1 Recuired Conditions 5

Preparations for resumption of the '

p, heating operation were completed at 3:00 pm, Friday,  ;

- March 10, at which time the concrete shield wall temperature $

was observed and again found to be 430 F hence the objective ' I remained to obtain a vessel metal temperature of 1730 F at , ,

h an internal pressure of 40 psig. [ [

(

17.2 Heating with 960 kw Capacity )

)

Temperature i

17.2.1 i It was not considered necessary to delay start of heating for the pressure build-up since i the vessel had already been expanded to the extent associated i with conditions of 112 F and'26 psig. All heaters, sixteen at 60 kw and fans were started at 3'pm March 10. ,

Initial air and metal-temperature was 58 0F. f Initial heat up rate was about 10 5 F i

_ per hour metal temperature increase which produced a rather wide range of metal temperature, about 450F;. highest

~

1 d

at the heater level and decreasing upward. The heaters were turned of f for 20 minutes between 6 and .7 pm with fans .  ;

remaining on to promote circulation and reduce the metal temperature variation. The result was a rapid drop in 1 air temperature with only a 50F reduction in the spread.

Since,the variation was fairly uniform and temperature within each level of thermocouples reasonably close, the j effort was abandoned with the reservation that it would  :

be pursued again if the variation did not narrow in the course of heating.

The metal temperature had reached 1000 F, 120 F below the maximum reached previously, when the pressure build-up was to 32 psig at 8:18 pm.

After passing the maximum conditions attained the previous week, heat up rate dropped gradually to about 3 F per hour. ,

69 .

90000221 33 ,

e. .

One pair of heaters .was out of service 67 . 'between 3:20 am and 4:15 am March 11 to replace a ourned out fuse. Voltage, initially 480, had increased'to 500;

' transformer taps were dropped;.vol.tage at-5:00 am was back to 480. .

i Main breaker, thus all heaters, went out at 5:25 am March 11 due to.short in'No. 1 main switch.-

^

A brief electrical fire required the No.- 1 bank- (half the. .

heaters) to stay out temporarily for repairs. No. 2. bank was

.back on at 5.:40 am: 2 . pairs o f - h eat er s _in No .1 bank were back on at' 7:50 ~ and all heaters were in service at 10:30 .

am. Probable cause was moisture being forced out of

• the drywell inside the cable insulation; the condition

-became so pronounced that by 8:00 am water was dripping out at the switches. Metal temperature,-at 1440F at 5:25 am, dropped and did not recover to that level until about 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> later. Temperature rate of increase returned to 30F per hour when all' heaters were back in service.

With average metal temperature I

at about 170 F at 6:27 pm March 11 the temperature variation was still about 280 F. All heaters were then

(_ shut off for 45 minutes. With f ans left on during this ,'

period the spread was reduced to about 18 0 F.with the 2-5/8 " thick plate at the sphere to cylinder transit' ion being low at hbout 1650F. It was considered desirable to.get the transition up to the minimum objective of 173 0F if this could be accomplished with overall average not more than about 180 0F.

Heating was resumed at 7:10 pm.

With the operation approaching completion, the concrete shield wall mid-thickness temperature was checked at 8:30 '

pm and was found to be still at the starting condition of 143 0 F. Overall average metal temperature reached 180.5 7 .F at 9:53 pm at which time the transition was at 1720F.

Heating was then terminated.

Cool down and pressure reduction were accomplished as described below under " Pressure". Tempera-ture conditions at the time the airlock was opened were air and metal at 140 0 F.

90000222 6.2b 6

L 34

17.2.2 Pressure ,

47a

== Compressor.s were started at 4:15 pm March 10; pumping was stopped when pressure reached 32 psig 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> later. Pressure increased with temperature at a much lower rate than previously experienced, possibly ' ,

due to the additional leakage through the cables between.

vire and insulation; increase was only 2 psig while air temperature had increased front 58 to 206 0 F. Pumping was resumed at 3:30 pm March 11 when metal temperature was at.159 F and stopped when 38 psig was reached at 4 pm.

Further pumpi.ng was not required; temperature increase caused pressure to ultimately override the intended 40 psig by 2 psig. This moderate excess was accepted to avoid the temperature drop which would accompany bleeding.

Af ter the termination of heating at 9:53 pm, pressure

- reduction was deferred for an initial cool down period, then performed in steps ' alternating with cool down and gauging of contraction to assure adequate internal vessel pressure to balance any external pressure due to rebound of the material. When temperature was down from 180.5 to 1680 F pressure was reduced to 30 psig. When temperature was down to 1600F pressure was reduced to 20 psig. At 12 :10 am March 12 -average temperature was ?.550F '

('

or approximately 20% less than maximum temperature increase.

The extensometer readings showed a range of 28 to 38 percent reduction from maximum expansion due to combined cooling to 1550 F and pressure reduction to 20 psig indicating no further need for concern regarding rebound of this material since it had been established as not exceeding 20%. Pressure reduction to O psig followed and the airlock was opened at 1:25 am, March 12.

17.2.3 Exoansion Extensometers were read and recorded hourly. As maximums were approa'ched a pattern someWhat

. different from the prediction developed in that, while horizontal movement was in good agreement,.the upward accumulation of expansion expected due to the embedment of the lower region was at all points less than predicted. (There-fore vessel discontinuity stress at embedment would have been less'than calculated and load on the concrete wall would have been more uniformly distributed and with a lower maximum)

All gauge locations had been expected to rise, actually the one near the_ vessel equator moved essentially horizontally,

,those 10 feet below the equator moved' out and slightly .

33 90000223 1

i

, 1

~

6:un while the high guage points moved up and out but I with vertical component less than predicted, since the

}.g;. calculated rise assumed a contribution from expansion below the equator. ,

The gauge. readings were continued ,

through the cool-down - pressure reduction phase and were : '

used to advantage to confirm vessel contraction to safe '

l

~

limits before final pressure reduction as described under " Pressure" above.

~

Maximum measured vessel expansions compared with the calculated expansions at each gauging point are tabulated below:

Drvwell Movement at Time of Maximum Averace Temnerature,&

Pressure, 180.50F: 42 esic: 9:45 cm, March 11, 1967.

Movement (inches)

Penet. Calculated 1 Measured No. Elev. Azi Hor. Vert. Tccal Hor. Vert.l Total'

• ** l

• V H ,

I' X-66 27' 360 .42 .22 .48 .46 -01 .46 X-2A' ,27' 171 -15' .42 .22 ' .48 .40 .05' .40 X-32 36' 125 .42 .34- .54 .43 .02 .43  !

e I

X-36 44' 2900 .43 .44 .61 .38 .20 .43 0

!i X-63 62' 340 .31 .66 .73 .32 .50 .59 X-12A 62' 240 .31 .66 .73 .36 .10 .37 X-19 86' 1900 .21 .93 .95 .31 .52 .61 A A A

• (Hor 2 + Vert 2 )h

• • Dial readings corrected for 140 0 F avg. penetration pipe temp.

cm .

~

33 90000224 9