ML19283C947

Submits Progress Rept for 840924-1005 Re NRC Ongoing Insp of Const Assessment Program.Certification Packages of Certified HVAC Level II & III Inspectors & Completed HVAC Reinsp Packages Being Reviewed

ML19283C947
Person / Time
Site:	Braidwood
Issue date:	10/10/1984
From:	Norelius C NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
To:	James Keppler NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III)
Shared Package
ML19283C948	List: ML19283C947
References
NUDOCS 8411260530
Download: ML19283C947 (2)

Text

.

BUOYANCY, TRANSPORT AND HEAD LOSS CHARACTERISTICS OF FIBERGLASS INSULATION FOR NU. CLEAR CONTAIl%IENT AREAS MANUFACTURED BY INSULATION TECHNOLOGY INCORPORATED By

-Dominique N. Brocard Sponsored By Insulation Technology Incorporated g]

ALDEN RESEARCH LABORATORY

-~1 WORCESTER POLYTECHNIC INSTITUTE 8411280530 841115 SU8d 50-84/M528F kb-9 CF May 1984

• p

,==*-*e e*

4

,g, 6

BUOYANCY, TRANSPORT AND HEAD LOSS CHARACTERISTICS OF FIBERGLASS INSULATION FOR NUCLEAR CONTAINMENT AREAS MANUFACTURED BY INSULATION TECHNOLOGY INCORPORATED e

By Dominique N. Brocard Sponsored By Insulation Technology Incorporated s

George E. Hecker, Director Alden Research Laboratory Worcester Polytechnic Institute Holden, Massachusetts

i ABSTRACT In the event of a loss of coolant accident (LOCA) in a nuclear power plant, it is possible that insulation for pipes, steam generators, or other items inside the containment building could be dislodged and disintegrated by the high energy break jet.

This insulation debris could affect the recirculation of water from the sump of the emergency core cooling system (ECCS) by collecting on the screen surrounding this sump.

In order to help assess the potential for sump screen blockage and its effects on flow through the screen, studies of the buoyancy, transport and head loss characteristics of fiberglass insulation products manufactured by Insulation Technology Incorporated were conducted.

The tests were aimed at determining the length of time that insulation could remain floating, the flow velocity needed to transport insulation to the screen and the head loss incurred by water flowing through insulation collected on the screen.

Three different insulation products were tested, referred to as Products NC Type 1, 2 and 3.

In general it was found that the insulation tested sunk rapidly in warm water (zl20*F), although, in one case, an undamaged insulation panel remained partially afloat because of a trapped air bubble.

The transpcrt tests showed that iflow velocities of 0.8 to 1.2 ft/sec were needed to transport as-fabricated samples to the screen, velocities of 0.20 to 0.65 ft/sec being needed for smaller pieces of inner insulation, including shreds.

The head losses, or pressure differences, needed to drive flows of water through beds of insulation against a screen similar to those used in ECCS's were measured for shreds (1/2" x 1/2" x 1/8"), fragments (l" x 1" x 1/8"), and as-fabricated fiberglass insulation (without covers). The test variables were the bed thickness, which was varied from 1/8" to 10",

and the approach velocity, which was varied frcm 0.1 to 0.5 ft/sec.

Regression formulae were derived to allow estimation of head losses for any bed thickness and approach velocity in the range of the tests.

ii TABLE OF CONTENTS Page Number ABSTRACT i

1.0 INTRODUCTION

1 2.0 INSULATION TESTED 3

3.0 BUOYANCY TEETE 7

3.1 Methodology 7

3.2 Results 7

4.0 TRANSPORT TESTS 9

4.1 Methodology 9

4.2 Results 11 5.0 HEAD LOSS TESTS 14 5.1 Methodology 14 5.2 Results 17 5.3 Regression Analysis 21 REFERENCE 39 0

6 e

1.0 INTRODUCTION

In the event of a Loss of Coolant Accident (LOCA) in a nuclear power plant, it is possible that insulation material attached to pipes, steam generators, or other items in the containment building would be dislo,dged and disintegrated due to the impingement of the vapor jet from the break.

A concern with loose insulation material is its effect on the recirculation of water by the Emergency Core Cooling S stem (ECCS).

To prevent trash from being entrained into the suction fluw of t.im pumps, the sumps are enclosed by a trasn rack and fine screen upon which insulation material could collect, disturbing the flow to the inlet. Partial blockage of the racks and screens could cause a loss of head which could decrease the pump flowrate or possibly lead to insufficient available NPSH.

In crder to help assess the potential for sump screen blockage and its effect on flow through the screen, studies cf the buoyancy, transport and head loss characteristics of non-encapsulated fibrous insulation were conducted at the Alden Eesearch Laboratory (AF1) of Worcester Polytechnic Institute (WPI) for the U.S. Nuclear Pegulatory Commission (1).

Three types of insulation pillows were tested using mineral wool or fiberglass layers enclosed in covers of various types.

Insulation Technology, Incorporated, manufacturers /fabricatcrs of fiberglass insula

'n blankets which can be used in nuclear generating stations containmr + b lings, requested that AFL perforn buoyancy, transport and head less tes:

, their insulatien similar to the tests perfomed on the other insulation samples.

Specifically, the tests conducted were aimed at determining:

1.

The buoyancy characteristics of the insulation, or the length of time insulation would float while sprayed by a fine mist of water; and 2.

The transport characteristics of the insulation, or the flow velocity needed to initiate the motion of sunken insulation fragments of various sizes; and

2 3.

The head loss characteristics of the insulation, or the pressure differ-ence needed to drive different flowrates of water through beds of insu-lation, either undamaged or disintegrated (torn up) in fragments or shreds, against screens similar to those used around ECCS sumps.

e 9

S D

3 2.0 INSULATION TESTED Three different types of products of Insulation Technology Incorporated were tested.

These insulations, referred to as Product NC Type 1,

2 and 3,

are described below and in Figures 1 to 3.

Product NC Type 1 consists of a cylindrical pipe insulation cover r.ade of e.etage densa t.y c,i 5 IL/ft

~iuwul npe ina:, ale mu -

t w= q wo.

un ou covered with a 17.8 oz/sq yd Alpha Maritex fiberglass cloth (Product

1. 2025/9383) double stitched all around with fiberglass thread (Alpha Product D-24).

As s'..own on Figure 1,

the pipe cover tested was made of two halves stitched together along one side and fastened by Velcro along the other side.

The sample tested was 36 inches long with an inside diameter of 12 inches and a thickness of 1.5 inch.

Product NC Type 2 consists of pancis made of "Knauf Elevated Temperature Panel" fiberglass with fibers oriented longitudinally and with a density of 2.4 lb/ft covered with the same 17.8 oz/sq yd Alpha Maritex fiberglass cloth (Product #2025/9383) double stitched all around with fiberglass thread (Alpha Product D-24).

A stainless steel mesh can be added to one side of the panels.

As shown in Figure 2, the samples tested measured 24 inches by 24 inches by 3 inches thick.

Product NC Type 3 consists of panels made of "Knauf Elevated Temperature Flex Wrap" fiberglass with fibers oriented transversally and faced with a fiber-glass scrim cloth.

The density of the fiberglass is 2.8 lb/ft and it is

~

covered with the same 17.8 oz/nq yd Alpha Maritex fiberglass cloth (Product

1. 2025/9383) double stitched all around with fiberglass thread (Alpha Product D-24).

A stainless steel mesh can be added to one side of the panels.

As shown in Figure 3,

the samples tested measured 24 inches by 24 inches by 3 inches thick.

4 4.5 lb/ft FIBERGLASS INSULATION

/.FIBERGLASSCLOTH (LONGITUDINAL FIBERS) y

/

-- ^ 175"----

n ST1TCHES (ALL AROUND) 12" If e 36" LONGITUDINAL SECTION VELCR0 FASTENER / \\ \\ND \\f STITCHES TRANSVERSAL SECTION FIGURE 1 SAMPLE TESTED OF " INSULATION TECHNOLOGY" PRODUCT NC TYPE 1

5 3 2.4 lb/ft FIBERGLASS CLOTH FIBERGLASS INSULATION (LONGITUDINAL FIBERS) y-4 ir __ _= STITCHES 3u Y (ALL AROUND) = 24" = = STAINLESS STEEL MESH (OPTIONAL) --~~7 < :r.m4pga;_.

r q,

.. v.. ..,.r. - yyg,, ' ;;:.9, : < <t.lh f a b n h ! ,w,,. L,c. k&l,;M)7;47 a -.m. p .,94

1

. gg.7, , e. m y; ...,.s [ ' $. $ '. 3 .. ~.. c .s g I y 2,. -,:.,d,!',.. t,,' .y" .4 e ,._.. F ;,,y.; v .., e m,, pun y;,,:. w <- v, " ;;p %. 3 d'I,' .l. ,f -' r I 1 ..s' _~?,,.. ?! * . y. p,. ', t E: ;... - 1, R, & %? l t @ {, l J ", j

g>.- l -

p ~ o, eg.. , s 4,- f.t, g.*. _ a,? f. :.. e :; y) $;~J r 3 {.g$C : s; p-s, iq 9 . o:;wlle 'G5$ Ib.;*?b nt, ,\\ y f,jl[ [ s: y ,t. m p c.;+ - mM2 FIGURE 2 SAMPLE TESTED OF " INSULATION TECHNOLOGY" PRODUCT NC TYPE 2 b

6 2.8 lb/ft / FIBERGLASS CLOTH r+3" q FIBERGLASS INSULATION / l l (TRANSVERSE FIBERS) !llI l'll l ll jll llj/l i l I I I I,[ ll ' l A I STITCHES l 3"

l I

I l8 (ALL AROUND) j' s i i. _ _y__ e 24" / / FIBERGLASS SCRIM STAINLESS STEEL MESH (OPTIONAL) ~-.,~.-,.,n:,. m.,;.,--.-. . y:

s o ---

,.jp%;;b '"Y %HE ~ / me m 7, p y,, -- w n ..q .c ..,..c J-@@ %.:.yg~~)%I,DEEkgbi&Q; n %Q/ ,'r g$w. 1 6 7;,,,,. f,,'F;hSi?i$, i!,Qtere.H c J t ,pt-eeo s r: 'o w,, . f. ur.,S, j : ~ -, e ~,' .g. ~ ...i -- t., kDY55y[a.:h]'$h@,! $RNM)lkd[hy Ss.~O V : % r ~ L ' i $Dkd/MW'$ fO wa;wl%,q-1uw :- e Aw n 1

m.,.mv g,.. -

'cV !S'y; y L' ', h q ' '..,. ' -:a i<;fs.5,W:n l,a., 5 [',*v ~ A ;- j FIGURE 3 SAMPLE TESTED OF " INSULATION TECHNOLOGY" PRODUCT NC TYPE 3

7 3.0 BUOYANCY TESTS 3.1 Methodology Whether insulation set loose during a LOCA will be transported to the sump screen during the recirculation mode depends upon the length of time that the insulation can float on the water surface. Recirculation of emergency cooling water is initiated a car +ain +ima (on tha nraar of ?n minutas) after the start of the LOCA. Insulation still floating at that time will be readily trans-ported to the screen, while sunken insulation will require flow velocities in excess of certain threshold values to be transported to the screen. These threshold velocities depend upon material types and sizes and are determined in the transport tests. In the buoyancy tests, as-fabricated insulation samples as well as fragments thereof were placed on the surface of a large basin containing water at about 120 F and sprayed with a fine mist of the same water, simulating the con-ditions in the containment area duri ng a LOCA. The lengths of time that the test pieces remained afloat were then measured. 3.2 Results The results of the buoyancy tests are summari::ed in the following table.

8 Table 1: Buoyancy Test Results Insulation Insulation Time to Type Condition Sink Comments NC Type 1 As-Fabricated 2 min. 3" x 3" x 1.5" 20 sec. Shreds immediate NC Type 2 As-Fabricated 5 min. to over 5 hours In this test an air bubble trapped in the panel kept it partially afloat indefinitely. 3" x 3" x 3" 30 sec. to 10 min. The longer time is needed when test piece is not sprayed. Shreds immediate to 5 min. The longer time is needed when test piece is not sprayed. NC Type 3 As-Fabricated 2 min. 3" x 3" x 3" 20 sec. Shreds immediate Note: As-Fabricated refers here to the entire samples described in section 2, including covers. 4 m

9 4.0 TRANSPORT TESTS 4.1 Methodology The object of the transport tests was to determine the flow velocity needed to transport sunken insulation to the ECCS screens. The tests were conducted in a flume 6 ft wide by 6 ft deep and approximately 40 ft long. As shown sche-matically in Figure 4, water was introduced in the flume through a perforated plate, to yield a uniform velocity.11stribution. This uniformity was verified by measuring vertical and horizontal velocity profiles approximately 13 ft from the perforated plate. Further downstream the profiles become less uniform with bcundary layers developing along the bottom and side walls. At the downstream end of the flume a screen an6 trash rack similar to those used around recirculating sumps were erected vertically across the full width. The screen was made of 1/16 inch wire with a 1/4 inch mesh directly behind a standard 1 inch floor grating trash rack with its more closely spaced bars oriented horizontally. A photo of the flume showing the screen and trash rack is given in Figure 5. The flow velocity in the flume was measured using a calibrated laboratory current meter of the propeller type located 13 ft downstream of the perforated plate (where the velocity profile was measured tc be approximately uniform), on the flume centerline at approximately half the water depth. The tests were conducted with unheated water at a temperature of around 40'F, as temperature is not expected to affect threshold transport velocities. For all the tests, the test pieces, either "as-fabricated" insulation samples or portions thereof, were deposited in the flume approximately 20 ft upstream of the trash rack and screen (2 ft downstream of the velocity meter). The flow velocity was then gradually increased and the velocity at which the test piece (s) started to move was recorded. The motion of the piece (s) was

10 PE R FORATED TRASH RACK PLATE AND SCREEN [ VELOCITY j l/ METER j i \\ U i ti INITIAL 32 ; + I SAMPLE POSITION y 13' 20' 2'

• FIGURE 4 SCHEMATIC OF TEST FACILITY "g

......-s. - p k 'M ) [$d1. %. . ?Y, if. A o. eap ... /. ?- p..- -((~ h ' ' ~ w w35 us

ggaeu S

, 1r - / a . 5'. FIGURE 5 PilOTO OF TEST FACILITY LOOKING DOWNSTREAM

11 observed and recorded. If the sample stopped and judgement was passed that it would probably not move further in these conditions, the velocity was in-creased until further movement was observed. This procedure was pursued until the piece (s) reached the screen. The manner in which the piece (s) collected on the screen was recorded, particularly the velocities needed to flip the test pieces vertically against the screen were determined. ~ 4.2 Results For each type of insulation and fragments size considered, a number of repeat tests were conducted and the results of the individual tests are listed in Tables 2 and 3, thereby furnishing an indication of the variability of the phenomenon. The results reported in these tables are the flow velocities needed to i) initiate the motion of the test piece, ii) transport the test piece to the screen and iii) flip the test piece onto the

reen, as well as comments on the transport mode or other relevant cbservation.

Since Products NC Type 2 and 3 are made of fiberglass with very similar den-sities (2.4 and 2.8 lb/ft respectively) and since the orientatien of the fibers would not affect transport, these two types of samples were not tested separately and results are presented in Table 3 for both types. The transport velocities measured here are somewhat lower than those measured for the high density (11 lb/f t ) needle punched fiberglass of comparable sizes tested for the NRC (1). In particular, the intermediate size pieces (3" x 3" x 1") move at about half the flow velocity. The smallest pieces, however, (1" x 1" x 1/8") move at 0.25 to 0.4 ft/see which is comparable to the velocities found in the NRC sponsored tests.

12 Table 2: Transport Test Results for Product NC Type 1 Velocity to Velocity to Velocity to Initiate Transport Flip Onto Test Piece Motion to Screen Screen Condition (ft/sec) (ft/sec) (ft/sec) Comments As-Fabricated 0.80 0.80 0.90 The sample, (12" pipe with 0.80 0.86 1.04 initially cover) opened and flat en bottom as

shown, m

folded in two just before starting to move Pieces of inner fiberglass (without covers) 24" x 3" x 1.5" 0.39 0.45 0.45 0.35 0.35 0.35 0.35 0.44 0.44 3" x 3" x 1.5" 0.43 0.45 0.45 Test pieces 0.45 0.45 0.45 tended to 0.36 0.48 0.48 separate in 2 or 3 layers in water. 3" x 1" x 1.5" 0.25 0.25 0.25 Test pieces 0.20 0.25 0.25 tended to 0.23 0.23 0.23 separate in 2 cr 3 layers in water. 1" x 1" x 1/8" 0.25 0.26 0.26 0.35 0.35 0.35 0.24 0.24 0.24

13 Table 3: Transport Test Results for Product NC Types 2 and 3 Velocity to Velocity to Velocity to Initiate Transport Flip Onto Test Piece Motion to Screen Screen Condition (ft/sec) (ft/sec) (ft/sec) Comments As Fabricated 0.75 0.85 1.4 Sample with (24" x 24" x 3" steel mesh ~ panels with against bottom. covers) 1.19 1.19 1.2C Cample without 1.05 1.15 1.15 steel mesh or with steel mesh up. Pieces of inner fiberglass (without covers) 24" x 3" x 3" 0.65 0.65 N/A 0.72 1.00* N/A

• Velocity needed to free samples from flume side wall against which it had moved.

3" x 3" x 3" 0.55 0.67 N/A 0.34 0.48 0.47 0.47 3" x 3" x 1" 0.47 0.47 0.47 0.47 0.50 0.50 0.30 0.37 0.37 1" x 1" x 1/8"

0. 10 0.40 0.40 O.29 0.36 0.36 0.29 0.36 0.36

14 5.0 HEAD LOSS TESTS 5.1 Methodology The tests were conducted in the set-up shown in Figure 6 and schematized in Figure 7. A 1/4 inch screen was placed at the bottom of a 4 ft vertical section of 8 inch diameter pipe in a recirculatory flow system. The flowrate was measured using a calloratea orirace ilowmeter wit.h a wetrr monumelet. Tine head loss through the insulation bed was measured between points approximately 32 inches upstream and 12 inches downstream cr the screen. Either a water manonieter or a differential pressure transducer was used for these measure-ments, depending on the magnitude of the pressure difference. For head losses less than 4 ft the water manometer was used with a measurement accuracy of approximately 10.005 ft of water. Fcr head losses over 6 ft, the pressure transducer was used with a measurement accuracy of approximately 0.02 ft of water. Tests were conducted with beds of insulation material in the following three conditions: i) "As-fabricated" insulation, without cover material, cut in 8" circles and glued to the pipe walls to nrevent leakage. Different thicknesses were cbtained by peeling the initial mats in layers or by adding several thicknesses. ii) Insulation fracments approximately 1" x 1" x 1/8" randomly accumulated on the screen. These fragments were obtained by peeling 1" x 1" pi~eces of the insulation mats in layers. iii) Insulation shreds obtained by manually tearing the 1" x 1" x 1/8" fragments described above in four, leading to pieces measuring approxi-mately 1/2" x 1/2" x 1/8".

15 ? j?$$$ . w,; vp? .u4 LU 11: &T': WaWb*b?$?E h 5,, w?% l '$h 5 guyn y gif((hgt [.[7

I j[

jk:n1 > r y l::ywel, h w;, '.,;. sa f M'- ?n k %QM)$?I k,

.; ggi-h

.S N 39 p

^ t%. '9 YI l h' b j p

.aM r>

. :5l., q $~b e;.' ^ 4 1 k)l y CMy 4- %gd.MN M"'n?Ij;*$5,- +(f - t 1 rnsm vs= Tg t!%v.. '1T'.Ny?qigr';y$uis; DNb [ h.'{ g; e s.mn ._. c q7395, FIGURE 6 RECIRCULATING FLOW LOOP USED FOR HEAD LOSS TESTS

16 g PRESSURIZATION TAP H f h Qn ' I ../ L, <r_m I e -g h T I _ 8" _ l H l E B 1 q INSULATION d FRAGMENTS N g ? ORIFICE [~'ggj] FLOW METER 7..._...a u / SCREEN f PUMP FIGURE 7 SCHEMATIC OF FLOW LOOP USED FOR HEAD LOSS TESTS

17 In all the tests, the fragments or shreds sizes were uniform and while this is obviously a limitation, as a high level of non-uniformity of sizes can be expected following a LOCA, this procedure has the advantage of providing representative data which are reproducible and which can be used to compare one inrulation material with another. b2th insulation in the three conditions described above, head loss tests were conducted with nominal bed thicknesses varying f4.>m 1/8" to 10". The nominal bed thickness is definad e- +ha un1"=a nf inan1ntinn in an-f abri cated con-ditions divided by the screen area. Actual bed thicknesses depend on whether the insulation has been torn up in fragments or shreds and on prior com-pression history. The tests were conducted for the different bed thicknesses with approach velocities increasing from the lowest value (0.1 ft/sec) to the highest (0.5 ft/sec) to avoid residual compre_ssior ef fects. 5.2 Results Measured head losses for the condj' ions and thicknesses described earlier are given in Tables 4, 5 and 6. Because Product NC Type 2 and 3 have approximate-ly the same density (respectively 2.4 and 2.8 lb/ft ), head loss tests were not conducted with fragments and shreds of both products. Only Product NC Type 3 was tested for fragments and shreds. For the as-fabricated insulation, both Product Types 2 and 3 were tested because the orientation of the fibers is different and this could be expected to affect head loss. Product Type 3, which has transverse fibers, also contains a fiberglass scrim cloth on which the insulation is glued. This scrim cloth was included in the test pieces, although its effect en the head lors is probaby small because of its light weight. For Product NC Type 2 the fibers were oriented perpendicularly to the ficw, giving Icwer head losses than Product NC Type 3, which fibers were aligned with the flew. In general, shreds (1/2" x 1/2" x 1/8") produced slightly larger head losses than the 1" x 1" x 1/8" fragments and significantly lower head losses than the

18 as-fabricated insulation, except for Product NC Type 3, which produced low as-fabricated head losses because of the flow aligned fibers. It is interesting to note that the shreds and fragments of Product NC Type 1 generated lower head losses than those of Product NC Type 3, which has a lov.cr density (2.8 lb/ft for Type 3 compared to 4.5 lb/ft for Type 1). For a given nominal thickness, E, a larger density leads to a larger mass of fiberglass

and, usually, to larger head losses.

Differences in fiber characteristics or bindcr: bct.;;cn rr duct: !!C *yp: 1 cr.d 3 probably accourt for the observed inverse trend. For the as-f abricated insulation, the trend of increasing head loss with increasing density is restor-:d. O

19 Table 4: Head Loss Test Results for Product NC Type 1 Head Loss (ft of water) Velocity Fragment Size Thickness (inches) (ft/sec) (inches) 0.125 0.25 0.5 1.0 5.0 10.0 0.1 1/2 x 1/2 x 1/8 0 0.005 0.002, 0.015 0.102 0.225 1 x 1 x 1/8 0.001 0.001 0.002 0.010 0.082 0.131 As Fabricated 0.015 0.048 0.120 0.324 1.21 3.12 0.2 1/2 x 1/J x 1/8 0.002 0.009 0.015 0.065 0.552 1,66 1 x 1 x 1/8 0.003 0.004 2.013 0.048 0.382 1.35 As Fabricated 0.038 0.113 0.282 0.517 2.91 7.70 0.3 1/2 x 1/2 x 1/8 0.006 0.020 0.057 0.187 1.44 4.16 1 x 1 x 1/8 0.009 0.022 0.033 0.140 1.03 3.75 As Fabricated 0.092 0.210 " 525 0.888 4.66 13.52 0.4 1/2 x 1/2 x 1/8 0.015 0.037 0.108 0.385 2.77 7.60 1 x 1 x 1/8 0.019 0.039 0.084 0.309 2.U8 7.27 As Fabricated 0.138 0.313 0.841 1.312 6.55 19.94 0.5 1/2 x 1/2 x 1/8 0.030 0.080 0.192 0.650 4.66 11.73 1 x 1 x 1/8 0.041 0.074 0.171 0.538 3.49 11.24 As Fabricated 0.193 0.446 1.145 1.757 8.68 28.06 O

20 Table 5: Head Loss Test kesults for Product NC Type 2 ~ Head Loss (ft of water) Velocity Fragment Size inickness (inches) (ft/sec) (inches) 0.125 0.25 0.5 1.0 5.0 10.0 0.1 Ac Fabricated 0.025 0.037 0.100 0.155 0.22 2.47 0.2 As Fabricated 0.065 0.092 0.238 0.330 2.79 6.46 U.3 As Fabricated 0.115 0.176 0.405 0.565 4.92 11.43 0.4 As Fabricated 0.180 0.282 0.615 0.847 7.48 16.86 0.5 As Fabricated 0.305 0.391 0.840 1.135 9.75 22.85 Table 6: Head Loss Test Results for Product NC Type 3 Head Loss (ft of water) Velocity Fragment Size Thickness (inches) (ft/sec) (inches) 0.125 0.25 0.5 1.0 5.0 10.0 0.1 1/2 x I/2 x 1/8 0.001 0 0.002 0.023 0.227 0.775 1 x 1 x 1/8 0.001 0.002 0.004 0.023 0.162 0.26 As Fabricated 0.080 0.170 0.775 1.41 0.2 1/2 x 1/2 x 1/8 0.005 0.014 0.022 0.124 1.105 3.78 1 x 1 x 1/8 0.005 0.013 0.025 0.083 0.821 1.57 As Fab ricated 0.175 0.366 1.72 2.E8 0.3 1/2 x 1/2 x 1/8 0.020 0.044 0.062 0.316 2.79 8.09 1 x 1 x 1/8 0.020 0.037 0.065 0.237 2.16 3.46 As Fabricated 0.280 0.664 2.77 5.07 0.4 1/2 x 1/2 x 1/8 0.030 0.085 0.122 0.598 4.84 13.36 ~ 1 x 1 x 1/8 0.035 0.088 0.137 0.468 4.09 6.22 As Fabricated 0.407 0.948 4.15 9.31 0.5 1/2 x 1/2 x 1/8 0.055 0.139 0.202 0.936 7.55 20.35 1 x 1 x 1/8 0.065 0.143 0.265 0.788 6.53 16.30 As Fabricated 0.530 1.26 6.91 15.16

21 5.3 Regression Analysis In order to allow estimations of head loss for different bed thicknesses and flow velocities than measured, best fit expressions are established for the head loss data. Multiple linear regression analyses were performed using the following general form for the head loss versus bed thickness and approach velocity: U LH a aE V (1) in which AH = head loss E = bed thickness = volume of insulation in as-fabricated state divided by screen area V= approach flow velocity Eq. (1) can be transformed to in AH = a + b in E + c in V (2) giving a linear expression upon which multiple least square regression can be perf ormed using standard techniques. A measure of the error associated with the regression formula can be obtained by calculating the standard deviation of the data set with respect to the regression formula: T. (in LH - In LH )2 M R c= (3) n-1 in which AH is the measured head loss, AH is the head loss given by the g g regressicn formula and n is the number of measured data. For data normally distributed about their mean, which can be assumed here as a first

22 approximation, 95 percent of the data points can be expected to fall within 2c of the regression formula: In AH - 2c < In AH < In SH + 20 (4) p g g or, exponentiating each term All < AH < AH e (5) 20 M R e i By writing e'o = 1 + 100' 9' AH - ( ) % < SH < AH + it (6) R 1 M R 1 + 100 giving upper and lower percent limits within which 95 percent of the data is contained. It can be noted that for small i/100 compared to 1, the upper and lower percent limits are approximately equal. The regression formulae and error bands for all the material types and sites tcsted are given in Table 7 These formulae are compared to the actual data on plots of head loss versus approach velocity and bed thickness and on plots of measured versus regression head losses in Figures 8 to 21. In general, exponents above 1 for the thickness, E, in the regression formulae indicate compressibility ef fects (1). The formulae in Table 7 show that the accumulated fragments and shreds exhibit significant compressibility effects, with exponents on the order of 1.3 to 1.4. This is expectable, as cutting up the fiberglass in fragments cr shreds amcunts to fluffing it up, increasing its volume, and once accumulated on the screen these fragments or shreds become very prone to compression. The as-fabricated insulation, on the other hand, exhibits very little corpressibility effects, with exponents close to 1.

1 23 For the as-fabricated insulation the exponent of the velocity term, V, is, in all three cases, on the order of 1.3 to 1.4 indicating the beginning of turbulence effects in the flow through the insulation (1). For the fragments and shreds, the exponent of the velocity is larger, on the order of 2.4, reflecting the combined effects of compressibility and turbulence. e a

24 Table 7: Head Loss Regression. Formulae Error Band Insulation Insulation Containing 95% Type Condition Regression Formula of Data Type 1 1/2" x 1/2" x 1/8" AH = 85.9 V

• E*

-46% to +86% ~ 1" x 1" x 1/8" AH = 80.1 V

• E*

-42% to +71% As-Fabricated AH = 77.3 V

• E*

-30% to +42% Type 2 As-Fabricated AH = 62.8 V

• C*

-35% to +53% Type 3 1/2" x 1/2" x 1/8" AH = 175 V~ E* -50% to +99% 1" x 1" x 1/8" uH = 108 V* E* -35% to +53% As-Fabricated AH = 34.6 V E* -21% to 27% 1) Fegression Formulae give AH in ft of water with V in ft/sec and E in ft; valid for V < 0.5 ft/sec and E < 0.83 ft. e 4

25 10 10 l l g _l l 1 4 i l i 6 i 1 g i 6 i l 6 6 4 4 2 2 1 1 0 E = 10" 10 10 e g e 6 6 5" 4 o 4 5 f Q 2 D 2 f 2 c. 0 S 0 o-3 10 10 3 B a e o-e O e V 1" [ 4 V 4 o? m m 3 2 2 y 0 1/2" b 10~ l o, / ~ o M 10 I e ,/p /,[ > / e A 1/4a / e y 0 e ql/N' /l',' l 4 1 4 i 1/8" o / 2 -/ b / 2 v /A o v o / 10 -2 -2L 10 e / e i e d),/ e o 4 4 _' 2 -A O 2 O 10

• O

~ ~ ~ 10 2 4 5 B 2 4 6 8 2 2 4 6 8 -1 0 1 -1 0 10 10 10 10 10 E = BED THICKNESS (INCHES) V = VELOCITY (FT/SEC) FIGURE 8 MEASURED HEAD LOSSES COMPARED TO REGRESSION FORMULA FOR 1/2" x 1/2" x 1/8" SHREDS OF PRODUCT NC TYPE 1

26 2 10 ig iiig iiig 3, i i_ i i i g i i ii i e i 6 s a 4 SIZE: 1/2 X 1/2 X 1/8 ,e' / 2 l 2 FITTING FORMULA 85.9V.m E*#1 1 95 % OF DATA WITHIN /,' 10 e' N / [ A H - 46% < A H < AH + 86% ,/ j 8 0 4 ,/ 9,/ g U e' di , [ /,' y 2 / S 10 i , '%'/ B p 6 ,/off,,' \\ 95% data limits u. v 4 / i i y 2 / oo' O / a _g w 10 / A/ B / @O',/ f B w 4 ,/ f,/ Q O' w L ,/ g,/ 1 2 ~ / op of 10 * ,I y',,/ 8 i 6 ,' j o, O s' / s' / i i l /oA 2 / i 10 "! l' ~ 2 4 68 2 4 68 2 4 68 2 4 68 2 4 SB -3 -2 -1 0 1 2 10 10 10 10 10 10 ~ REGRESSION AH (FT FOR WATER) FIGURE 9 MEASURED VERSUS REGRESSION HEAD LOSSES FOR 1/2" x 1/2" x 1/8" SHREDS OF PRODUCT NC TYPE 1

27 2 2 10 10 i i , i i i i g i i , i 6 6 4 4 2 2 10 10 9 B 0-6 6 4 o j o 4 65' b Q 2 4,\\ 2 D / 10 9 e-A S 10 3 8 sk D 6 y 1 6 4 4 e- !S 7 2 o 2 ~ 1/2" b 10~ 10 / I s 7 a yf4 U 6 6 jf 4 0 ~ I 4 -O 1/8" 2 2 10' O 10 -V B 8 6 / 6 4 A / 4 A 2 O 2 -0 !t il t i i 1 -3 -3 1 t I I f I 10 10 2~ 4 6 e 2 4 6 e 2 2 4 6 e ~ 10 10 10 10 10 E = BED THICKNESS (INCHES) V = VELOCITY (FT/SEC) FIGURE 10 MEASURED HEAD LOSSES COMPARED TO REGRESSION FORMULA FOR 1" x 1" x 1/8" FRAGMENTS OF PRODUCT NC TYPE 1

28 2 10 8 ' 3 '/' ' l 'l 'l B 8 8 5 4 SIZE: 1" X 1" X 1/8 " j E 34 ,/ ,f 2 FITTING FORMULA 80.IV.2 2 l e' 10 95 2 0F DATA WITHIN 1 e ,/ ~ [ A H - 42% < A H < AH + 71% e'o ,/ 4 w s e s ,/ 2 ,/ 4 b 10 B s H S e' 95% data limits b 4 0,' ,! W s' 3 2 j o' oW 10 3h ,/ p'O;e y e-m 6 s 4 / / h W I 2 f, -2 10 l B S t ,/ O' 4b 2 ;' O' / i i i Il il -3 mr i t tI t i t i t I t i i i # 2 4 SB 2 4 SB 2 4 SB 2 4 SB 2 4 SB -3 -2 -1 0 1 2 10 10 10 10 10 10 REGRESSION AH (FT OF WATER) FIGURE 11 MEASURED VERSUS REGRESSION HEAD LOSSES FOR 1" x 1" x 1/8" FRAGMENTS OF PRODUCT NC TYPE 1

29 10 10 l l 8 -l 'l 3 3 8 s 0 4 4 p O E = 10" 2 / 2 1 1 5" 10 10 O ~ g B 6 Y 6 O 4 o? O 4 a: a w 4 / Q 2 2 l" ~ o S 10 O' 10 1/2" - [ s 6 / 4 A 1/4" - 4 //p/f -t;[/ / 2 1jgo 2 10

• 10~

~ I B B / D 6 6 I 4 a 4 o 2 2 O O 10 10~ ~ 8* B 6 6 4 4 2 2 10 10~ ~ 2 4 68 2 4 68 2 2 4 5 8 -1 0 1 -1 0 10 10 10 10 10 E . BED THICKNESS CINCHES) V = VELOCITY (FT/SEC) FIGURE 12 MEASURED HEAD LOSSES COMPARED TO REGRESSION FORMULA FOR AS-FABRICATED (WITHOUT COVER) PRODUCT NC TYPE 1

30 2 10 i , i i e i i i,. i g ig iig iiig g i i i i ,/ - 6 ', ',/ - 4 SIZE: AS FABRICATED 1 / E.12 ,g 2 FITTING FORMULA 77. SV.35 1 if</ 1 95 % OF DATA WITHIN ,'/ ,,Pf' 10 8 ~ A H - 30% < A H < AH + 42% /' [' 4 w 2 ,'g, ' 5% data limits - >~ d M,,' 9 0 g O ,b g W 6 s b 4 ' W,' <3 2 e' 8,'/ / 10,' ~ ,p 7 m 6 ,' '/ ,!, O' h 4-s w I l j y / -2, ',e 10 B s 6 l s' ~ 4 j j 2j l l

l

-3f,e t i t t i 1 t , i g i i ri r i l l gg 2 4 68 2 4 68 2 4 68 2 4 68 2 4 68 -3 -2 -1 0 1 2 10 10 10 10 10 10 REGRESSION AH (FT OF WATER) FIGURE 13 MEASURED VERSUS REGRESSION HEAD LOSSES FOR AS-FABRICATED (WITHOUT COVER) PRODUCT NC TYPE 1

31 10 10 l l B -l 8 6 e 4 4 O E " 10" 2 E) 2 W 10 10 5" B B 6 6 4 4 0: Lo e Q 2 O' 3 2 1" og 4 '/b O' O D V x D w O T ~ T Vp 1/2" [ e B 6 S 7 g / v / N O / / A V, O p / 3 2 / '7 2 17g. O V O ~ U h 10~ k 10 6 # A B B O/ Il 6 4 S I 4 g 4 3 0 0 2 2 ~ ~ 10 10 B-BH e eL 4 4 2 2 t-10 * ~ ~ 10 2 4 68 2 4 5 B 2 2 4 5 B -1 0 1 -1 0 10 10 10 10 10 E = BED THICKNESS (INCHES) V = VELOCITY (F SEC) FIGURE 14 MEASURED HEAD LOSSES COMPARED TO REGRESSION FORMULA FOR AS-FABRICATED (WITHOUT COVER) PRODUCT NC TYPE 2

32 2 10, i i( ig ig i,,i i i i i i i i i i i i i i 6 ,' ] - 4 SIZE: AS FABRICATED ,/// - i E. oS ,/,,/ 1 2 FITTING FORMULA 62.8V 37 3 sf/ s O 95 % OF DATA WITHIN l ,/ ~ 8 A H - 35% < A H < AH

• 53%

g 4 j,,/ / 2 / e' 95% data limits u. 10 D / Y U W 6 s' p' ~ b 4 P i f,, 2 p o 3 / ul 10 8 , o', m 6 i 4 ,i,$,s', Q p ,'q/ 2 ,',/,i 10 ,/,/,! ~ g 6 / i 4 / / i i i 2 ,i j 10~ /' / ' I I I I 2 4 60 2 4 68 2 4 68 2 4 68 2 4 68 -3 -2 -1 0 1 2 10 10 10 10 10 10 REGRESSION AH (FT OF WATER) FIGURE 15 MEASURED VERSUS REGRESSION HEAD LOSSES FOR AS-FABRICATED (WITHOUT COVER) PRODUCT NC TYPE 2

33 10 10 ~ e B S e 4 4 / E = 10" 2 2 10 / 10 e 4/ / e [%" ~~ 5 c, B O e [. A 4 O 4 v m 2 2

ks 2

5 m 10 0 10 1" a B O B T [ e e 2 4 4 / O',/ 7' 1/2" a / o a 2 o 2 / O //v 0 _3 / _t v 0 1/4" Lu 10 / V/ / 10 I g s s Q O' 6 - D S 7 O D/ I 4 - p e 4- / 2 D Of 2 ,/ O / d / L / / 10' 10 ' h f O 8 6 / 6 4w 4 2 O 2 4 l -3 9 ~3 i f I I I I f I 2 4 5 8 2 4 S 8 2 2 4 5 8 ~ ~ 10 10 10 10 10 E = BED THICKNESS CINCHES) V = VELOCITY (FT/SEC) FIGURE 16 MEASURED HEAD LOSSES COMPARED TO REGRESSION FORMULA FOR 1/2" x 1/2" x 1/8" SHREDS OF PRODUCT NC TYPE 3

34 2 10, i i i i y i i_ ig i) i i i i i i ei i i i e 4 SIZE: 1/2"X 1/2'1/B' ,/ E * '8 ,/,/O,/ 2 FITTING FORMULA l'/ 5V,41 2 1 95 % OF DATA WITHIN / / 10 O/ ~ e-AH - 50% <aH < AH + 99% 4 /o/ / W O' i 2 l / / / 0 u. 10 O B /O H S t- / s' 95% data limits / / / O 4 'O/ s' ,' ? Oj 2 / / / s C -1 O' / w 10 / O. x g-i 5 6N ,s' o,! O, / 4 / Of,/ j (w / O/ s e' Of O/ 1 'o 2- / p Q' i 10~ g( ,i // ,! l s' 6 4 s' ,/ / c' o 2 l-10'b' 2 4 5B 2 4 00 2 4 08 2 4 SB 2 4 50 -3 -2 -1 0 1 2 10 10 10 10 10 10 REGRESSION AH (FT OF WATER) FIGURE 17 MEASURED VERSUS REGRESSION HEAD LOSSES FOR 1/2" x 1/2" x 1/8" SHREDS OF PRODUCT NC TYPE 3

35 10 10 j Il B -l i i i g I i i i 6-6 4 4 2 2 E = 10"~ 10 / - 10 B B 0 V 8 Of5" 6 [3 / 4 G 4

• y' l f/

2 m L Q 2 ,,,/ A w/ / a / A' B 1" [ e 8 4 / 4 1/2" cn /b/ 7/ h o .J 2 ,/ 2 f v,[ ~ h 1/4" o O O O D' ~ I 10 ' - j/ 7 f 10 I e g D N 6 0 1/8" 6O II I 4 4 7 O Y 2 -O 2 O ~- 10 'r-10~ B 0 6 S 4 O 4 4 2 2 -3 t I I I I t ! I -3 1 il I i i 10 10 m 2 4 ee 2 4 ea 2 2 4 ee -1 0 1 -1 0 10 10 10 10 10 E = BED THICKNESS CINCHES) V - VELOCITY CFT/SEC) FIGURE 18 MEASURED HEAD LOSSES COMPARED TO REGRESSION FORMULA FOR 1" x 1" x 1/8" FRAGMENTS OF PRODUCT NC TYPE 3

36 2 10 'l l 'l 'l '/' '- 3 0 3 e ',/ - 4 SIZE: 1"X 1"X 1/B' //- ,e' f / '/,,/ 2 FITTING FORMULA 109V. 40 E.29 1 2 1 95 : OF DATA WITHIN ,0, q j,/,s g A H - 35% < A H < AH + 53% ,. < 0 gy l / / p l ?p, / W 4 n i 'i i 2 N, ' O/ / 3 O' l i g sb / O j ,e 95% data limits g .e ' 0,/ s ,/ i i / / <Q 2 / / O' a _g w 10 / / O 2 B j O/ O' ] e h 4f j l E 2 Mkf/ l O' /,/, l .g 10 B i i lhe',' g / 4 j / / / ~ 'Il 3 I I ' ' 10 6 I ' ' 2 4 5B 2 4 SB 2 4 68 2 4 SB 2 4 08 -3 -2 -1 0 1 2 10 10 10 10 10 10 REGRESSION AH (FT OF WATER) FIGURE 19 MEA 3URED VERSUS REGRESSION HEAD LOSSES FOR 1" x 1" x 1/8" FR AGMENTS OF PRODUCT NC TYPE 3

37' 10 10 l g i i i 6 l i i i i i i ! g e e 4 4 2 2 0 V E = 10" 10 10 BL B O 5" ~ e c 8 j 4 ,9 4 D* b Vb 2 4 2 S 10 10 B- /o. o, 8 / ,6 Y/ l ~ B ~ ,b 1/2" ~ B ~ 4 ,/M,!y', y/ 4 k / h k ] 2-,/////,/ / 2 s ~ h 10~ / 10 j/ e s Se,/ ,/ e n = 4 f / 2 _/ g 10 J 10~ ~ B-B 6 6 4 4 2 2 ~ ~ 10 10 2 4 5 8 2 4 8 8 2 2 4 0 0 -1 0 1 -1 0 10 10 10 10 10 E = BED THICKNESS (INCHES) V = VELOCITY (FT/SEC' FIGURE 20 MEASURED HEAD LOSSES COMPARED TO REGRESSION FORMULA FOR AS-FABRICATED (WITHOUT COVER) PRODUCT NC TYPE 3

38 2 10 i i iiig i i ii i iiig g i i i i e ,o 4 SIZE: AS FABRICATED / U E 88 FITTING FORMULA 34.SV 30 1 2 1 95 % OF DATA WITHIN 8 ~ A H - 21% < A H < AH + 27% ,d e' ~ hj?' 9 4 s'h U 0 e S 10 95% data limits e g H S O 4 s'A z 2 O _1 e e e' w 10 f B w S ,li,' k 4 ,s',',' W ,'e I ,s',' 2 i. 10' ~~ B ,/s e i,' ' 4 ,s',' 2 / Il 10-f r I f i f l I I f ! I I I I I l 2 4 68 2 4 5B 2 4 SB 2 4 58 2 4 5B -3 -2 -1 0 1 2 10 10 10 10 10 10 REGRESSION AH CFT OF WATER) FIGURE 21 MEASURED VERSUS RFGRESSION HEAD LOSSES FOR AS-FABRICATED (WITHOUT COVER) PRODUCT NC TYPE 3

39 REFERENCE 1.

Brocard, D.N.,

"B uoy anc',', Transport and Head Loss of Fibrous Reactor Insulation," Nuclear Regulatory Commission Report NUREG/CR-2982, November 1982. S 9 0

_a . s 2.s. .G -.. e. ...e k ' i .,i 4 t A WORCESTER POLYTECHNIC l INSTITUTE m 4 O O ^9 q e e 9 4 9 ALDEN RESEARCH LABORATORY HOLDEN, MASSACHUSETTS 01520 Se

~., " NUCLEAR GRADE" INSULATION BLANKETS INSULATION TECHNOLOGY, INC. i M S. =mm-Insulation Technology Incorporated

M 9 mmm, insulation Technology incorporated " NUCLEAR GRADE" INSULATION BLANKETS INSULATION TECHNOLOGY, INC. REACTJ"wust. .ma.4ro, l a IIT .d WK A w. c?n 6 Q... i; p._ A s -:-.:..are [L ' q - a gi c[.. m. ,'i i casov or o r en no i / .uo!,,in99,,j,,9% : ),,' g,g;(,'1 +; i 91.autw =a - ~ ~ InInhipis7. 4.M. ,e; m M 9 s, s*i} t sa. k ?. h k 555> "' 'N

&%'

's:0, iv at arseven earssser d'"Es gy 7

s- ~~ fig TYPE 2 S'[g,*,g','" 22 cr ervas W / , rw; A .xxx rarrw.a - \\ . ;s'~.ril/;;", I s .;%+.*L f (,., N M ut,> g afd % TYPE }' k 962i., ,3DN f m,c, l x u=~su s 7d l - 8'rw"'"'"' ""** ),$,*g.g. sr*= or'ait '~ ,T.'3l".'01 ".' " '*" w 5731 WEST HOWARD STREET

• NILES,IL 60648 312-6471500. (Outside lilinois) 1400-323-3936. TELEX: 467998

dTt ,9 =mmer InsulationTechnologyincorporated A PRESENTATION OF " NUCLEAR GRADE" INSULAIICil BLANKETS by INSULATION TECHNOLOGY, INC. Insulation Technology, Inc., (I.T.I.) has developed a " Nuclear Grade" insul-ation system for use in Light Water, Nuclear Steam Supply Systems, including Boiling Water Reactors and Pressurized Water Reactors used th roughout. the world today. This-system combines the finest features of the I.T.I. blanket / pad type insulations, which are prominently sold to a vast number-of Utility and Petro-Chemical companies for various service requirements, with the latest technological advancements required for Nuclear plant construction and operation. In 1982, the Electric Power Research Institute (EPRI) conducted surveys, testing and analysis concerning the in-service perf ormance of Reflective, Metal Encapsulated, Mass Type and other insulations used to insulate the "NSSS" of numerous domestic power plants. This report, CONTROL -OF CONTAINMENT AIR TEMPERATURE: AN INDUSTRY SURVEY AND INSULATION TEST, literally identified a need to upgrade the approach previously taken to insulate the reactors, piping and-equipment in nuclear service.

Thus, I.T.I. developed a solution to this industry problem and can now furnish a completely tested and qualified system for your use.

The f ollowing is a general description of the product-known as " Nuclear Grade" insulation blankets by I.T.I, with supplemental data on testing and other analysis completed f or your review. Additional information is available relative to industry experience and vendor qualification from Insulation Technology, Inc. upon request. INSULATION TECHNOLOGY INCORPORATED August 1984 5731 WEST HOWARD STREET. NILES,IL 60648 312-6471500. (Outside lilinois) 1-800-323 3936. TELEX: 467938

fTt e-

==m-insulation Technology incorporated DESCRIPTION . NUCLEAR GRADE BLANKE'IS ----- TYPE I, TYPE II, & TYPE III TYPE I Type I material is made of high temperature, pre-molded fiberglass primarily designed f or-pipe applications in nuclear service. It is available-in standard lengths of three feet and in thicknesses to six inches without double lavering, which minimizes the cost.of-modules requiring thicknesses in excess of two and one half inches. The construction of the insulation modules is unique, as each half section of pipe covering is encased in a high temperature fiberglass cloth j acket, sewn with fiberglass th read and connected with a FULL LENGTH, FIBERGLASS CLOTH HINGE. Book and Loop (Velcro) flap f asteners are then utilized to secure the insulation to the pipe. The use of th e pre-molded fiberglass -mate rial offers more rigidity than the standard batts mechanically contoured to fit the pipe outer diameter -which-equates to greater durability during construction and in-service or operating life. TYPE II Type II material is made of high temperature fiberglass panels. The utilization of the more flexible fiberglass is necessary in the construction of most typical applications in a Nuclear plant. It is not always necessary to conform to the contour of a particular piece of equipment, yet the intent during our design stace is to " wrap" the equipment and fittings to minimize convection and physical size within the containment or other area. With th e Type II mate rial, we use high tempe rature fiberglass cloth sewn with fiberglass thread and attached with Book and Loop (Velcro) flaps and f asteners-to connect assemblies requiring more than one panel. Additionally, any support steel or lattice work is designed and furnished to support the insulation where required. Type II material is the most ef ficient blanket type material available in the industry today, as the major emphasis is on thermal pe rf ormance and nwrt taitorh of insulation modules to meet the "real world" requirement of tight and secure fit up in the field. Our experience in providing these specialty items for many years affords I.T.I. the luxury of prior exposure and knowledge. 5731 WEST HOWARD STREET. NILES,IL 60648 312-6471500. (Outside lilinois) 1-800-323-3936. TE LEX: 467998

."Tt .O: , insulation Technology Incorporated PAGE 2 nESCRIPTION T'IPE III Type III materials consist of fiberglass panels with fibers oriented transversally and faced with a fiberglass scrim cloth. This material is intended for equipment with large radius-bends such as Reactor Coolant Loops, Recirc. Loops, Piping 20" diameter and above, etc. It is available in single layer thicknesses to four inches and increments of.five inches. With the Type III material, we use high temperature fiberglass cloth sewn with fiberglass thread and attach Book and Loop (Velcro) flaps and f asteners to connect assemblies. The purpose of this colaposition is to ensure the stability of the fiberglass within the glass cloth modules during in-service inspection procedures, after initial heat loading, and also during the modification and application processes of construction. We invite your additional inquiries concerning every aspect of our " Nuclear Grade" insulation materials. As all applications vary to some degree, we are prepared to custom design our product to meet any condition within your nuclear environment. We are anxious to hear of those specific needs and we stand ready to serve you. 5731 WEST HOWARD STREET. NILES,IL 60648 312-6471500. (Outside Illinois) 1800-323 3936 TELEX: 467998

C7"* M.+ be s ".~s,.~'., ~ )O t n. :_:.,,. - k \\k M'f r

f..

L 4 h. 1 .: 8.:_. _ is '

.x_

]- )) \\ \\k 1 Q:.:.:;. _ r Y::~. i, T@. !f 5'f'i ~ L" e ** f@* 99 f ggt oc*' goo $ f a.,, .= O r:7.~...:.w_.. s 1 m .r %s NG TYPE I o HIGH TEMPERATURE FIBERGLASS CLOTH HIGH TEMPERATURE PRE M,OLDED 99%* ees #'@O

REACTOR VESSEL I I INSULATICN # -p> y l d + I. m- / [ Y l STEEL LINER i \\ ( t 1 ("p' 'r3 I L 3 e

k

• INSULATION COVER MAY

~[ CONSIST OF ONE OR TWO LAYERS OF INSULATION f' 23h l21 12 23 23 23 22 23 23 24 PA DS ( OPTIONAL 1 WITH .. l / T BOND BRICK APPLICATION it it 21 30 30 31 ~ J( M 31 30 31 73 d 32 33 34 34 3f' 34 33 1,! 33 Il 't 'N g 'l ' ,E 4 ,, 'C 'r STAINLESS STEEL WIRE MESH ( OPTIONAL ) i HIGH TEMPERATURE FIBERGLASS INSULATION [HIGH TEMPERATURE FlBERGLASS CLOTH TYPICAL REACTOR PRESSURE [ ALL SIDES 7 HOOK FASTENER I X / NG TYPE H* FLAP W/ LOOP f g;, FASTENER e s - / .p / .::t.%. EDE'ip N w WIRE MESH w,. y______ J .6 ,. y qp DOUBLE SEWN SEAMS A-LOOP SEAM D E Tall FASTENER eethB

f:}.YY%,....'Q:. bl.' 'h, W_, y.... - - >::. N.Q... %.' h_.. ,j ' 5' 'h. s... ..s.'M...'.%'h. .y% y. -s I T HIGH TEMPERATURE LOOP FASTENER FIBERGLASS CLOTH HOOK FASTENER HIGH TEMPERATURE FIBERGLASS PIPE INSULATION OR FLEXIBLE FIBERGLASS INSULATION (OPTlONAL) j.t:i- ~ -g o .. j,:..:p _ .- ~ 5 ' T$.h. .$n

i.. --

y*.

• I.'

.f

• f
• r4 h

.., ~ I-C :- l ***:*:H: \\ N

.:..a;).. :2f.Q 1

x ... t N .n N ', ii:: ' ~: :- .. :. l..' ';;:'i.: N c: x Q.1 ...g., 1 FULL LENGTH FABRIC HINGE ( OPTION AL) ELBOW NG TYPE IE CONSTRUCTION M

N- -.~g N*. ' .Q... '?';'s.'7 g. N .s -t.;g.. h.w. '

. g..

.~ ~ f. s. ~..:.;. ,.%; $. Q~... m ,.0- .--> g. . ~ 4 f J % g, k..,..f ~ ' ;. ., ' ?:.*%. D( p, !? N ' e '..,

l..

.; Q'~

• .;.., y, p q,.

e

/ . <;LQ t ' / .%.? 4 s-i g s '. '. g i Ng\\ ) ,. h.:;q.'.' _ ~ d s/ N} @d. s > g ALL AROUND STAINLESS STEEL WIRE MESH (OPTIONAL) HIGH TEMPERATURE FlBERGLASS INSULATION HIGH TEMPERATURE FIBERGLASS CLOTH

?// .= f "I /[ / " 7. w,. ...........S... 4. - p:.... t, . %~. ~. ,......c. h -.Y [M D* L.t[.M.,,g!.h:. 5.- . :.3y.

2..y p.

...n N. ;g D"... J ,a* l . =.. a.... .:.... ~ f. _ ~..... _ s ............t.*f......... ..:..... 1.... .........w... . ~.. 4 ..? -....w =. L_ e. .e 1 ~.. + ....,.%. a., w :$......*. .,6, c.

• ...p*

m...%

s. j .h; y ~~z:' 1-

7. a.

...:., ~. >... '...'... - :. : !!: **,'.!.!;* ~:,:.y M.. :s 7 ..??..:. .M.i.h...... * ( s

I ::. :'.

~

w. w......

s - w m,. .a ..........ss:*m... e.....

.=.

a .......*(... =. .....ss -:Q., '**!. ?:: -

• d *A

. k MM, -M s G......=..... M:.m.. s

• ~.. -. = -.

.'.....:::::* n.;*... 3:. \\ r.;;........:...:.:;;. ~ A,. r=.= w...... -.-..;.,a =.

• 't

..........J*"*"*.....N. .- ~ - - - = --

• . ::*..<... f;;.

\\> %s = ..s...; .....d:[ x = s a...,.. ..,m.~. ..,........,..a.se REACTOR REClRCULATION COOLANT PUMP

.T. =mmer insulationTechnologyincorporated A PRESENTATION OF " NUCLEAR GRADE" INSULATION BLANKETS by INSULATION TECHNOLOGY, INC. Insulation Technology, Inc., (I.T.I.) has developed-a " Nuclear Grade" insulation system f or use in Light Water, Nuclear Steam Supply Systems, including Boiling Water-Reactors and Pressurized Water Reactors used th roughout the world today. This system combines the finest features of the I.T.I. blanket / pad type insulations, which are prominently sold to a vast number of Utility and Petro-Chemical companies for various service requirements, with the latest technological advancements required for Nuclear plant construction and operation. .In 1982, the Electric Power Research Institute (EPRI) conducted surveys, testing and e. n aly s i s concerning the in-service-performance of Reflective, Metal Encapsulated, Mass Type and other insulations used to insulate the "NSSS" of numerous domestic power plants. This report, CONTROL. OF CONTAINMENT AIR TEMPERATURE: AN INDUSTRY SURVEY AND INSULATION TEST, literally identified a need to upgrade the approach previously taken to insulate the reactors, piping and equipment in nuclear service.

Thus, I.T.I. developed a-solution to-this industry problem and can now furnish a completely tested and qualified system for your use.

The following is a general description of the prod.uct known as " Nuclear Grade" insulation blankets by I.T.I, with supplemental data on testing and other analysis completed for your review. Additional information is available relative to industry experience and vendor qualification from Insulation Technology, Inc. upon request. INSULATION TECHNOLOGY INCORPOPATED August 1984 5731 WEST HOWARD STREET. NILES,lL 60648 312-647-1500. (Outside Illinois) 1800-323 3936. TELEX: 467998

m O = der losulationTechnologyincorporated TESTING 5731 WEST HOWARD STREET. NILES,lL 60648 312-647-1500. (Outside lilinoist 1800-323-3936. TELEX; 467998

."Tt O Insulation Technology incorporated INTRODUCTION Insulation Tecnnology, Inc. has completed an extensive test program for " Nuclear G ra de " and is pleased to furnish the specific certified test results and other relevant data sheets to our clients. The tests that are required by the nuclear industry are

varied, due to the unique design or eacn power plant.

Therefore, InsuAation Technology, Inc. has utilizec tne requirements or several utilities, consulting engineering firms and the N.R.C. to determine the extent of testing applicable to qualify these materials for use in the nuclear environment. We have selecr eo the folluwing tests for inclusion in this

booklet, as they represent necessa ry qualifications for use in your nuclear service.

Furthermore, we have developed a summa ry for

eacn, wnich we feel is representative or tne product's integrity anc competitive aavantage, when app 2icaole.

1. " Buoyancy, Transport and Head Loss Characteristics of Fiberglass Insulation for Nuclear Containment Areas". The slynificance or this test is increasingly important to the owners or nuclear power plants, as in the course or upgrading product qualification to meet the safety criterie during a loss of coolant accident (LOCA), new standarcs have been set. These standaros are developed by eacn utility for the unique conditions of the nuclear plant and our goal was to test with broad parameters to meet varied needs. For

example, our original test was modeled after the N.R.C.

testing or various insulation materials insiae containment areas.

However, the testing or the " Bead Loss" f or our material per the original

criteria, was subsequently enhanced to reflect greater and/or different velocities, more representative of conditions within some or the BWRs in se rvice today.

Please review this original test and the

revision, which will illustrate our superior results to that wnich has been previously tested by others.

2. Thermal Perf ormance of " Nuclear Grade". The testing of thermal conductivity for the insu2ation material useu within " Nuclear G ra de ", is well represented by the manutacturer. Tne Knaut Fiber Glass Company has furnisnea a consiceraule amount or data relating to the thermal pe rf vrmance or tneir product and that comoAneu with our testing at Dynatecn Laucratories in Cambriuge, MA. will more than qualify our product to meet the specified thermal perf ormance criteria of nuclear power piants in se rvice today. If there are spe cial considerations for thermal perf ormance that exceed the tested perf ormance or " Nuclear Grade", we have the resources to upgrade the materials to meet your needs and provide the testing to substantiate same. 5731 WEST HOWARD STREET. NILES,lL 60648 312-6471500. (Outside filinois) 1400-323-3936. TELEX: 467998

.1T,st er she-Insulation Technology incorporated 3. Regulatory Guice 1.36 Material Compliance. Compliance with this criteria is critical to the se rvice of insuAation procucts in a nuclear environment. Reg. Guice 1.36 is clear and states that all materials used must be testea to prove the limiteu content of leacnacle enorices and halides. We have provicea tne data sneets and certificates which are required to illustrate complete compliance with this requirement. " Nuclear Grade" insulation is in compliance with this requirement. 4., ASTM E - 84 Flame Spread and Smoke Density Test. We have provicec the necessa ry certificates or compliance fur this requirement. Please see tne test section f or detallo. 5. Gamma and Neutron Irraciation Testing of InsuAation Materials for Use in Nuclear Environment. The basis for testing " Nuclear Grade" insuAation fur the effects of gamna and neutron bombarument is the determination or pnysical cnarac: eristics after irradiation, as well as the rauloactive decay characteristics, also after irradiation. Insulation Tecnnology, Inc. was compelled to meet criteria which had formerly been developed by a competitor marketing blanket type insulativn, as it seemea to be the only criteria published for use by the utility incustry. The obvious concerns are the retention of radioactivity within the procuct, f or determination or allowable radiation levels f or in-service inspections and maintenance during outages, as well as possiole air-corne contamination of fragmented irradiated

material, and the structural integrity of insuAation materials subjectec to specified raulation levels.

We have addressed these concerns and represent the resuits in the test se ction with certified test r es ul ts. " Nuclear Grace" insuAation has snown no significant deteriorization of physical cnaracteristics which would affect the thermal and/or practical functions or the

proauct, and the retention or rauloactivity

(" half life") is not detrimental to it's use in nuclear power plant se rvice. In fact, there are notable differences between the reported data publishea by the previousAy mentioned competitor and that which we have furnishea.

Again,

" Nuclear Grade" is favored on a tecnnical basis. These differences snould be considered valid and germane to the technical qualification or our product. 5731 WEST HOWARD STREET. NILES,IL60648 312-6471500. (Outside lilinois) 1800-323-3936. TELEX: 467998

- C'M 9 ime-Insulation Technology Incorporated

SUMMARY

The technical qualification or " Nuclear Grade" for use in nuclear

service, is easily estaDlished by reference to the attacned test reports and other data within this submittal.

However, shouac you require acdtional testing and/or any clarification or the

data, we urge you to contact us with your inquiries.

Insulation Tecunology, Inc. is committed to the full development of " Nuclear Graue" and will guarantee it's integrity as a fully testeo product in fu.t1 comp 2iance with all the requirements for insulation materials in nuclear power plant service. 5731 WEST HOWAFID STREET. N LES,lL 60648 312-6471500 (Outside lilinois) 1400-323-3936. TELEX: 467998

m

9.

6 InsulationTechnologyincorporated 5645 WEST HOWARD STREET

• NILES,IL 60648 31216471500

'}}