ML17321A100

Dry Film Coating Thicknesses of Thermo-Lag 330-1 Subliming Matl Applied to Steel Hatch Covers for 3 H Fire Rating.

ML17321A100
Person / Time
Site:	Cook
Issue date:	02/15/1984
From:	WESSON & ASSOCIATES, INC.
To:
Shared Package
ML17321A099	List: ML17321A098 ML17321A100 ML17334A528
References
NUDOCS 8406210368
Download: ML17321A100 (47)

Text

DRY FILM COATING THICKNESSES OF THERMO-LAG 330-1 SUBLIMING MATERIAL APPLIED TO STEEL HATCH COVERS FOR THREE HOUR FIRE RATING Prepared by Wesson and Associates, Inc.

P. 0. Box 1082 Norman, Oklahoma 73070 2i February 15, 1984 PDR ADQCK 05000315 PDR WESSON AND ASSOCIATES, INC.

DRY FILM COATING THICKNESSES OF THERMO-LAG 330-1 SUBLIMING MATERIAL APPLIED TO STEEL HATCH COVERS FOR THREE HOUR FIRE RATING I .. INTRODUCTION The purpose of this report is to present the calculated minimum dry film coating thicknesses of, THERMO-LAG 330-1 Subliming Material applied to steel hatch covers which are provided in the floors of nuclear power facilities. These hatch covers are used to seal the connecting passageways between two floor levels or between a floor and a vault beneath the floor in these nuclear facilities.

The minimum dry film coating thicknesses have been calculated to provide a three hour fire rating when subjected to the fire condition specified by ASTM E-ll9 Test Method. In this test procedure, the time-averaged incident heat flux for a three-hour exposure is 42,000 Btu per hour per square foot.

The basis for the determination of the minimum dry film coating thicknesses for the THERMO-LAG 330-1 Subliming Material applied to flat plates is an engineering correlation developed from experimental results from fire tests conducted on the material. The experimental data used in the correlation include the results of fire tests con-ducted by'everal independent organizations such as the Underwriter's Laboratory', the U. S.'epartment of Transportation (Federal Railrbad Administration), Mobil Oil Corporation and Wesson and Associates, Inc.

The minimum dry film coating thicknesses have been calculated for hatch covers having thicknesses of 3/16, 3/8 and 1/2 inches, fire ex-posure from one side only and simultaneous fire exposure on both sides 0

of the hatch cover. The temperature rises used included 250 F for LESSON AND ASSOCIATES, INC.

0 personnel considerations and 930 F for maintaining the structural integrity of the steel.

II. ENGINEERING CORRELATION FOR FLAT STEEL PLATES The thermal performance characteristics of fireproofing materials such as THERMO-LAG 330-1 Subliming Material, THERMO-LAG 290 Subliming Material, CHARTEK 59, KOROTHERM and PYROCRETE 102 have been found to correlate as:

t = a function of (T,QT, W, F) where t = fire exposure time, minutes T = dry film coating thickness of fireproofing material,'

inches 0

QT = temperature rise of the protected metal substrate, F W = effective heat capacity of protected metal substrate, pounds per square foot of exposed area F = total incident heat flux, thousands of Btu per hour per square foot.

Experimental fire test data expressing the fire exposure time as a function of the fireproofing material coating thickness, the temperature rise of the protected metal substrate, the weight of the protected metal substrate'nd the total incident heat flux have been used to develop r

empirical engineering correlations for the THERMO-LAG 330-1 Subliming Material applied to various substrates such as flat steel plates, pipes and structural steel members (I-beams). The range of total incident heat fluxes have varied from a low of 10,000 Btu per hour per square foot to a high of about 95,000 Btu per hour per square foot. The coating thicknesses have varied from a low of 0.125 inches to a high of about 1.25 inches.

The engineering correlation expressing the exposure time of a protected steel plate as a function of the coating thickness, temperature rise of the metal substrate, the effective heat capacity of the protected sub-strate and the total incident heat flux is WESSON AND ASSOCIATES, INC.

1.3356 t = 23.002 (T) (QT) (W) /(F) (2) where W iF expressed in pounds per square foot of protected flat plate exposed tb the incident heat flux. The engineering correlation is presented in Figure l. As noted on this figure, if the protected sub-strate is exposed to a fire condition on both sides, the effective heat capacity, W, is based on one-half the thickness of the plate.

4 III. INCIDENT HEAT FLUXES FROM FIRES The incident heat fluxes from the ASTM E-119 Test Method and from flammable liquid spill fires are widely used for the determination of the required fireproofing coating thicknesses for various protected steel substrates.

ASTM E-119 Test Method:

The ASTM E-119 Test Method uses a specific time-temperature relation-s ship for testing the fire resistive. capabilities of various fireproofing materials. The required time-temperature relationship is presented in Figure 2. As shown the internal air temperature of the the test set-up starts at the prevailing ambient air temperature, reaches a temperature of 1700 0 F at the end of the first hour of exposure, a temperature of 0

1850 F at the end of the second hour of exposure and a temperature of of about 1950 0 F at the end of the third hour of exposure. This speci-fied time-temperature profile does not represent the time-temperature profile for a typical hydrocarbon spill fire wherein the temperature 0

within the flame zone may reach a temperature of about 2200 F or more within a very short period of time (one to three minutes). Thus, the ASTM E-119 Test Method does not truly represent the environment of a flammable hydrocarbon fire.

.The time-heat flux relationship corresponding to the time-temperature profile specified by the ASTM E-119 Test Method is presented in Figure 3.

As indicated on this figure, the integrated time-averaged incident heat fluxes are 24,500 Btu per hour per square foot for the first hour of exposure, 34,500 Btu per hour per square foot for two hours of exposure WESSON AND ASSOCIATES, INC.

4 V

FIGURE 1: CORRELATION OF THE THERtRL PERFORMANCE CHARACTERISTICS OF THERMO-LAG 330-1 SUBLIMING MATERIAL APPLIED TO STEEL PLATES AND PIPES NOTE: For angles or plate exposed on both sides, use W ~ @eight of metal sub-strate, lbs/sq.ft of exposed area based on one-half leg or plate thickness 0

~ Test points for 300 F metal substrate 200 0

~ Test points for 500 F metal substrate 0

Test points for 800 F metal substrate I ~ ~ I I 100 I'I ~ I I I I ~ ~ ~

I ~ ' ~ Is I

~ ~ '

~

~

~

I II I

~

~ C'I~

O'II ' >>

o 10 E 23o 002 (T) (QT) (W) /(F)

T ~ THERMO-LAG 330-1 Coating thickness, inches DT ~ Temperature Rise of Metal Substrate, 0F Weight of Metal Substrate, lbs/sq.ft of exposed surface Fs Total Incident Heat Flux, thousands of Btu/hr-sq.ft 0.1 10 10 (T) (DT) (W) /(F) 4 WZSSON AND ASSOCXATES INC.

FIGURE 2: ASTM E-119 TEST METHOD TIME-TEMPERATURE PROFILE 2900 ~ ~

~ ~ ~

~ ~ ~

~ ~

II

~

'.I s.".>>I Iee.'~ III', Ie l I

!: >> .(.

- i )ii'! I!  ! .'I .'ll:

~ Ill 2400 ,'.: I'.."I '.l.l ilal I(ll j:aj ~ ~ s ~ ~ ~

I (st 0 .... s... .I I~

.". ~

I I. >>

.Is):II e a

~ .".. II ..I I I ~l I. ~

j. ~

'.ll 'lls I'e I j II I

jl.':::I e I (jll I ~ a I

':' 'I' I'I'IIIII

"'i" -: .'ll' I)( )ij >> ) Ie e" ~ ~ il) ' ~ " -'

I" >>I! so'. sl('.::>> '.: ':" I: .I .~ la

" a I~ ~ I

~ I

>>:i I'::I(I "'=. ll

~ ~

-I:

~ ~ e I ( ) ( ~ ~ ~ i)'ll

'll; ":I III "I: I' s' ' ill ~ Ii ' (

I (II "el ) I" "'. I' i'i I,',

1600

':: ..'.j I

"'ii'I' j I I

' I I Ij I lj

>>II':

~

I

'i

( ) !I)l le):

I I !I i j)II I ~ I I

el)

"Ii ',' 'Ijj j' I Ii I I:I:..

~ ~

II

'I I i )i I I:II I! I II !i !i! il j j ij I Iij ~ ~

I ~ a I ~ I ~ e I. )

j ~

Ij ~ I

'j

~

.'I I <'"'I'>> I ~ ~

i': ll ill I

I ~

Ijj i.j '(el I 'I 'III sill (I ' " ' "'i'I >> ":>>I ."", "II:s'I 'll I" ~

1200 " ' >>I:I'il I

'I

~ :  :  ::.; .:il !I:

II'!ilI I (I'. ilji I ill >>" ~ ~ (II ~ . I. .>>...e...l .I...ee ~...I

~

j, s II '

. ~ Is).."...~...I.... Ill...) '( e...ll.

(I I,;I I IIII

~ .. I I .I I I I: I! .'i.

I I I I!I ~

( )III )

I IIII I~I I I

( )) I I I I ~ ~

I (

~

I

~

I s

~ I e I ~ el

":I!i' I'lli 'a'I'I I'I'I"

~ e ~

(

i!iI. '!Ii "!I! Iil! ',i !i jii! (j ji'!! jii i '!! !I Ill ' ll 's '>>I:"(>>I '.:: (<> ' ~

'I ~

I'I ' 'Ill s I"

II III I I II(. I Ill. )II.. ~ I ~ I ll I( ()IiI'Ili I ~ I~ lljl I )

I~

I I ~ si el I'll !i!i 'jij

~

j ~ ~ I~

.... Ii). . II IIII i I 'I I Ill( ,'Ij I'j'ilj '(j: ~

.,' II 'a 'll Il!i jl >>  ! ~

. I!'I

'I. I I>>i

.( I )s( ai. I .~ . e ~ ~ ~

Ie

~,:>>:.'ll

~

I

.' il Ii )j Will,i

i 'li,'ii' ll I I I I

!) I j I I II j a

( ) ( ) ~ ~

ij ill) ij" ~~ ~

~

is

<> ~

I I Ij i ~ I I Ij ~

I jl I, I e

I e

s>>,(

I I I ) I I I i I jl)

'I e' i.,

~ ~

j>> I I I I

. I) ll ~

I I II) ~ I I) . .)ei e

~ .~ ..... I ~

I I I

~

j ~

I s

~

I e

~ I,, j .Ili II I j Iji: '>>I i I: I i  !  :.:Il I'I il

' I I I'.!. :!i :!  :  !!:i: I '>>: Ill "Ii " I!  !'! , '

~ I )

~

Is j

~ ~ I ) I'.".il li,j .

~

.  ;( Iij , ,

i.. I .. I,>> .,, -, ... ,'... II.I I . Ii,, ij" ',,:

~

, .: i  ! , "i I  ! I I I i : 'I '. ~

0 0 3 4 5 7 8 9 10 FIRE EXPOSURE TIME hours

FIGURE 3( INCIDENT HEAT FLUX LEVEL OF FIRE SPECIFIED BY ASTM E-119 TEST METHOD AS A FUNCTION OF FIRE DURATION/EXPOSURE TIME

1. I~  :: II

' ( ' ai I ~

4

~ ~ ~

4 e

4 ~

I I~

~ ~ ~

4 ~ ~

I:i li I fji i: I ~ 4 ~ I

~ ~

II ~ ! il ii'i ~ ~

j: I i'i .' iil I~ ~

I 4

~ ~ << ~

4

"~:I ~ ~

I I': ! I.:. Iaa I:i

+a I

I;

:*; ~ ~

~

'h(i I:. ~ ~ ~

II i;ij! I 60 I ~- :i:  :

~ ~ ~

j'I ::I: ! ~ ~

js!  !.Ij. ~: ~ ~ ~ ~-

~ I il 4J I:

~ ~

all. .. ~

~ ~

!.!! il I I '

'. . ll', ~ ~

~

...I

~a

I i jl~j! ~

~ ~ ~J!

". ii!: a!.Is I ~

I ljl' ~ ~

~ ~

i: j I ! I"! '.: l: ~ I il:. III . ill I 50

.'~

I ~ ~ I s

III I ~ '.. I

III ~ ~ ~ ajl; ~

(Ii !I  !:I: ~ ~

Ii.i iii'. j.

'I '.

I:,!  ::ii ~ i'l ij ;II m 'I a

ua

'I ~ I:: Jl ~

I il:. I ~

is'Ia 44 .II

~

I~ ~

I

~

I" ~

I ~ ~

I

isi !I!! . ~ III

~ 4 II:I I I ~

0 'I I I~ i: .'

I~ i:I'((I

II

~

j II ~

~ il s I: ae lJ'I

~

Ii '.I: I p ~

t t:I II I!i! :l'j! 'I .'.

~ ~ ~ ~ ~ ~ ~

~ I e ~

~

I ~

'I ~

I I~ I ~

'(I

. Iel>> (

I 40

~ ~

il !(I I'I: I

~

~ ~ ~ i:I! "~ I~ (

~

Jig! jLJ: i:JI I lil

.'ill

!ll 'I( i(:. ll. ~ ~ ~ ~

I~ I

'I III: I" i( j: il:i !': I~

~ ~ ~

-~ ~ ~ ~ ~

~ ~

I

~ ~

JjLJ:::

il;":

!:I:! ilij I~ ~

I I

';tijl . .II:.:

30

jj: l I I(J) ~ ~ ~

I~ "il  !.I

~ I  !;Il Ill( ~ ~ ~ ~ ~

'll

~ ~ I II

~ ~

ij::il'::  !:!

~ ~ ~

il :: (I ill: I I I ilail I~I~ ~

~ ~

i!jjj fj(

I "! JI  !'I ~

4 ~

I~ I I:I I ~ ~ ~

il~<< J!! i!

i a( TIME-AVERAGED INCIDENT HEAT FLUXES I ~

'!l:

~

I~ Ja

~ ~

J,lie I ~ I" 20 I!

II I I

~

@ji I

(:J.i~

~ I~ ~

Li I:: sl!

~

I:'.: .'I:

.,I

~ ia!

I: ii.!;!

~ ~

I ONE-HOUR EXPOSURE: 24,500 Btu/hr-ft2 !I Il I!,I li .:. Jsl Btu/hr-ft2

~ ~

Ji,! I ~

M J ill I TWO-HOUR EXPOSURE: 34,500 III leI'

'l

~ '!

';JI I: I ii i!!l J!: .:II I!j:

~

u

II::".'I)J

~

i u ;jj; li;I i(L II

~

10 sij(

illi il': ils( ~ I Ii ' "JI I THREE-HOUR'EXPOSURE: 42,000 Btu/hr-ft2 III'  ::I ll. !!:j III "!i. Ilj!

~ ~ I~S ~ ~

~

I ~  !

'll'l',

!!I! ls:

Ii.l

~ ~ ~ ~

jI:  :.l'j  !)IJ

~ ~

lil! !ilI ~ ~ ~

I:I Jl I la!I ('ll !jan jJi I~

~

I

~

i'  !::. !I! Ijl' I ~

sjl: ;I

~ ~

~

~

I~

0 0 60 120 180 TIME minutes

and 42,000 Btu per hour per square foot for three hours of exposure.

H drocarbon Pool Fire Total Heat Fluxes:

It must be emphasized that all liquid hydrocarbon fires do not produce the same total heat effects. As shown by Table 1, different liquid hydrocarbon flames have very different heating effects. For example, a fire involving methanol will only produce a total incident heat flux of about 12,000 Btu per hour per square .foot whereas a fire involving LPG could produce a total heat flux of about 40,000 Btu per hour per square foot for a relatively large diameter spill fire (fire diameters in excess of 30 feet) .

Since the total incident heat, flux appears as a linear term in Equation (2), it is very important to specify or know the type of fire for the determination of the required fireproofing coating thickness.

Incident Heat Flux Used in Determination of Coatin Thicknesses:

The three hour fire rating presented herein has been based on the incident heat flux level associated with a three-hour exposure to the fire specified in ASTM E-119 Test Method. The total incident heat flux used to calculate the coating thicknesses was 42,000 Btu per hour per square foot.

IV. RE UIRED THICKNESSES FOR STEEL HATCH COVERS A complete listing of the calculated coating thicknesses of THERMO-LAG 330-1 Subliming Material applied to steel hatch covers is presented in Table 2. Four cases arepresented to cover various aspects of fire exposure and temperature rises of the steel hatch covers. The covers are assumed to be exposed to a fire from one side only and also to a fire from both sides simultaneously.. The temperature rises considered 0 0 were 250 F in consideration of personnel safety'and 930 F in considera-tion of structural integrity of the covers.

It should be pointed out that the dry film coating thicknesses presented in Table 2 do not include a 10 percent aging and weathering allowance WESSON AND ASSOCIATES, INC.

TABLE 1 TYPE OF FUEL MAXIMUM HEAT TRANSFER FROM FLAMES TO COLD TARGET (BTU/HR - FT.SQ.)

RADIANT CONVECTIVE TOTAL Methanol 5,000 7,000 12,000 Acetone 10,000 7,000 17,000 Hexane 22,500 7,000 29,500 Cyclohexane 31,000 7,000 38,000 JP-4: Small Spill Fire 23,700 7,000 30,700 JP-4: Large Spill Fire 31,000 '0,000 41,000 Benzol 39$ 000 7,000 46,000 LPG: Small Spill Fire 25,500 7,000 32$ 500 LPG: Large Spill Fire 34,500 10,000 45,500 LPG: Impinging Fire 70,000 LNG: Spill Fire on Land 45,000(Maximum) 10,000 55,000 LNG: Spill Fire on Water 45,000(Maximum) 10,000 55,000 Ethyl Mercaptan 18,800 7,000 25,800 T"Butyl Mercaptan 23,500 7,000 30,500 Ethylene 28$ 500 7>000 35,500 Buthylene 29,750 7,000 36,750 Butadiene 270500 7,000 34,500 Carbon Monoxide 4,500 7,000 11,500 Vinyl Chloride 8,500 7,000 15,500 AVESSON AND ASSOCIATES, INC.

and, therefore, represent the absolute minimum required coating thick-ness to provide the specifed fire rating. This allowance is based on long term environmental testing programs conducted by Ballistics Underwriters'aboratories, U. S. Anny Research Laboratories and commercial users in the hydrocarbon processing industry. Therefore, to provide an allowance for aging and weathering of the THERMO-LAG 330-1 Subliming Material, the coating thicknesses presented herein should be increased by at least 10 percent.

WESSON AND ASSOCIATES, INC.

TABLE 2 MINIMUM DRY FILM THICKNESSES FOR THERMO-LAG 330-1 SUBLIMING MATERIAL APPLIED TO STEEL HATCH COVERS Basis for Fire Ratin  : Three-hour exposure to fire condition specified by ASTM E-119 Test Method 2

Heat Flux = 42,000 Btu/hr-ft Hatch Cover Exposure* Dry Film Coating Thickness Condition Thickness in Inches**

inches 250 F DT 930 F ~T

0. 1875 single 1. 495 0.595

0. 375 single 1.055 0.420

0. 500 single 0.915 0.365 0.1875 double 2.110 0.840

0. 375 double 1.495 0. 595

0. 500 double 1.295 0.515

• Denotes fire from one side (single) or fire from both sides (double) .

• • Does not include any allowance for aging and weathering of material.

10 WESSON AND ASSOCIATES, INC.

4

~

~

~ ~

g Sub...itted to Butane-Propane News: April 1976 THERt'M EFFECTIVENESS OF VARIOUS FIRE RESISTANT COATINGS APPLIED TO STRUCTURAL ST"ELS EXPOSED TO DIRECT FLAYERS CONTACT AND/OR RADIATIVE HEAT FLUXES H. R. Wesson Wesson 6 Associates, inc.

P. 0. Box 1082 Norman, OK 73069 INTRODUCTION The rapidly growing acceptance of fire resistant coatings for thermal protection of structurals steels, flammable product storage tanks, pressure vessel support structure, among other applications, has placed this unique fireproofing concept in an approved posi-tion for extensive usage in the area of "exposure control" for structures that could be exposed to direct flames impingement, free burning plus pressure torching conditions, and/or prolonged periods of high intensity radiative heat fluxes. The inherent reliability and low maintenance costs for low this "passive concept" of exposure protection, together with the performance level of conventional water cooling systems under flame engulfment and/or high pressure impinging or torching type fire conditions-, have also giv .. these fireproofing coatings a very high cost-effective, or cost-benefit, characteristic for high heat intensity applications.

These type coatings are also finding applications where simultaneous low temperature (cryogenic liquid impinging conditions) and high temperature (flames contact conditions) protection is required for the structural steels in LPG, LNG, and SNG facilities.

The different types of fireproofing coatings that are commonly available, the results of. extensive fire testing on these coatings, and engineering correlations of the experimental data that can be used for determination of the required coating thicknesses for a desired period of protection in various heating environ-

. ments are presented and discussed herein.

GENERAL TYPES OF FIREPROOFING COATINGS The most c'ommonly accepted fireproofing coating materials include the following:

=

WESSON AND ASSOCIATES, INC.

~ ~ ~

~ ~

I Cement Compounds: Concrete, gunite, and similar concrete compounds provide good fire exposure protection during 'ase both direct flames contact and high intensity flames radia-tion cond':ions for extended periods of time. are In general, however, the cement compounds are quite heavy, expensive to install, in some applications are corrosive, and in general exhibit poor mechanical bonding properties between the sub-strate and the cement compound.

2. Ablative Coatings: These type coatings provide excellent fire exposure protection for structural. steels. The funda-mental principle is to apply a coating that gradually erodes due to the absorbed energy input from a fire condition. To change tne virgin solid coating into a gas composite requires heat input that would otherwise be absorbed by the structure being protected. The temperature rise of the protected struc-ture is retarded in direct proportion to the ablative coating thickness and its thermal properties. The incorporation of ceramic-like intumescents 'have resulted in a tough microporous char layer which provides additional insulating properties while most of the heat input is required for the physical transformation of the base material. The major disadvantages of these type ablative coatings appear to be the complexity of the application process and the final installed coating costs.

3. Subliming Compounds: The subliming compounds provide a pro-tected substrate temperature based on the temperature of sub-limation for each particular compound, the thickness of the coating material, the heat capacity of the substrate, the coating thermal properties, and the degree and time of heat exposure. In general,- the subliming compounds form a very tough, esthetic compound that is very tightly bonded (bonding strength of 100 psi and more) to the protected steel surface.

Another prime advantage of the subliming compounds is that they are not adversely affected by prolonged exposuresimultan- to low temperature liquids such as LNG and LPG, as well as eous exposure to such low temperature flammable liquids and resultant flames contact heating effects from liquidresulted spill fires. These advantageous thermal properties have in the use of the subliming compounds at some LNG Facilities for the protection of carbon steel structures, including the actual LNG storage tank,. that could be subject to LNG sub-mergence and/or LNG liquid spray impingement as well as direct LNG spill fire flames contact. These coatings must be surfaces applied to specified types of prime painted metal with airless spray equipment during relatively warm and dry atmospheric conditions (above 40 F and not during rains).

LESSON AND ASSOCIhTES, INC.

Department of Transportation aging and environmental tests give these type coatings a 20-year life when properly cured and the top-coat renewed every five to seven years.

4. Inturnescent Mastic Compositions: The most common of these type coatings are a modified vinyl, heavy-bodied mastic containing inorganic fibers in an aromatic solvent blend and a reinforced epoxy, two component, 100 percent solids (no solvent) spray system. In general, these type coatings react by absorbing heat in a chemical reaction which generates a foam-char system on the flames exposed side of the coating.

Additional heat input is used to drive the liberated gases through the matrix. The foam-char is also an effective thermal insulator. All of these heat absorbing and/or heat flow retarding mechanisms serve to keep the substrate below its allowable rated maximum operating temperature. The period

.thickness, the of substrate protection depends on the coating applicable thermal propert es, and the period and intensity of heat exposure. The heat capacity of the protected sub-strate also significantly affects the period of protection for a given coating thickness. Like the subliming compounds, these mastics do not suffer any adverse consequences when subjected to LPG and LNG contact, and are being used for thermal protection of steel structures associated with LNG storage tanks. One disadvantage of these type coatings appears to be the greater thickness. required for the same period of protection xn a given fire situation. For example, the published results of tests using the ASTM-K-119 Test Method indicate that using a 1000 F temperature for a 8';iF31 beam as a basis for comparison, a Q" thick coating of a typical vinyl-base type intumescent mastic will give a "two-hour" fire rating,. a 5/8" thick coating of the epoxy-based intumescent mastic will provide a "two-hour" fire rating, and a Q" thick sublimation compound coating will give a fire rating of "two and one-half" hours. Another disadvantage t'e of some of the intumescents appears to be the propensity of active ingredients to leach out over prolonged periods of exposure to outdoor environmental conditions. Once such a leaching has occurred, the protection time interval provided by such coatings is significantly reduced over the initial rating period.

As indicated above, the heat capacity of the protected sub-

'strate significantly 'affects the peri'od of protection piovided by a given coating thickness. An excellent example of this effect was given by O'ourke (1) in the 1973 Annual A.'I.Ch.E.

symposium on the fireproofing of structural steels. For ease of reference, Figure 1 presents this effect for wide flange structural steel beams.

WEssoN hND AssocIhTEs, INc.

,) I

,1 1/2 7/Z6 8WF31 STRUCTURAL BEAM C

~ R I

~10WF49 3/8 C3 A4 M

>/16 r

O 14MF228

'U O 1/4 M

3/16 REF.: O'ourke, J.F., "The Use of Xntumescent Coatings for Fire M 1/8

/ / Protection of Structural Steel" 1/16

/

20 . 40 60 80 100 120 140 160 180 200 TIME FOR PROTECTED BEAM TO REACH 1000 F-minutes FIGURE l: EFFECT OF INTUMESCENT MASTIC COATING THICKNESS ON THE EXPOSURE TIME

) IC c"PII11I ~

Unfortunately there are also a number of materials which are frequently KISUSiD as fireproofing systems. ?'.aterials which a'e

o misused for outdoor, fully exposed environmental conditions include:

Standard Thermal Insulation Systems: Conventional, so called standard insulation techniques, such as metallic-sheath covered cork, glass-wool, or aggregate systems such as vermiculite, perlite, or calcite provide excellent heat transfer protection for the flowing/stored media. However, such systems are poor fireproofing materials. Eornally the thermal insulation systems have very poor bonding properties to the base structure and are usually covered with a thin metallic-sheathing for protection of the thermal nsulation from environmental effects. Under direct flame contact, and/or high intensity radiative heat fluxes, these thin metallic coverings will quickly experience large deformations with an attendant loss of thermal protection. entrapped moisture between the thermal insulation and the steel struc-ture can provide a corrosion problem as well as generating sufficient steam pressure to actually blow large sections of the insulation system off of the protected structure under high heat flux conditions.

2. Refractory Protection Systems: Yost refractory materials provide excellent high temperature thermal protection in such applications as kilms, ovens, and high temperature process lines. However, these materials are often misapplied as fireproofing systems for steel structures that could be-come exposed to. flammable liquid spill fires. Host flammable liquids reach their maximum burning intensity within a few seconds and impose very high thermal gradients in the outer regions of the refractory protection systems in a short exposure period. Under large thermal gradients and the resultant high thermal stresses, most refractory materials will crack and/or spill, possibly leaving large structural sections of the basic structure completely unprotected. In general, the refractory unsubjected materials are designed to be brought up to their normal operating temperature over an extended time interval, as well as being cooled down quite slowly.

3. Intumescent Paint Compounds: These painting compounds, when to flame temperatures, puff. up to,form an air-filled ash which acts like an insulator material. Unfortun-ately their ability to intumesce is lost after short periods of exposure to outdoor environmental conditions, usually less than two years. A very serious problem in using the intumescent painting compounds for the fireproofing of exposed structural steels that could be subjected to high velocity WESSON AND ASSOCiATESO INC.

flames impingement is the extreme fragility of the air-filled ash formed by the exposure of the intumesc0nt paint to high temperatures. Experimental data have clearly shown that the gas valocities associated with Class I flarmable liquids under direct flame contact conditions are sufficient to completely destroy, or dislodge, the insulating air-filled ash layers'

~ >later of Hydration Plasters: These coatings are simply plaster compositions which undergo chemical and physical changes when exposed to high temperatures to releas water vapor. The theory is that the temperatures of the protected structure will be limited to the temperature of hydration process and that the fire energy is absorbed by the hydration process and in the vaporiz'ation of the water vapor produced by the various reactions. The materials that have been tested and reported upon in the literature have exhibited a high degree of hydroscopicity and a very limited ability to with-stand exposure to outdoor environmental conditions for even short exposure periods, less than one year. he inherent possibility of corrosion due to the water content of these coatings is a serious drawback to the use of these materials for fire protection of steel structures.

DISCUSSION OF EXPERIMENTAL DATA The principal sources of experimental data on the fire protection capabilities of the various types of fireproofing materials, other than the individual company research and develop-ment programs which 'are not normally available to the general public, are technical papers that have been presented at engine-ering conferences such as the 1973 Annual Heeting of the A.I.Ch.E.

in Philadelphia, PA (1, 2), the Fireproofing and Safety Symposium of the Vest em Research Application Center of Los Ange1es, CA, in 1971 (3), independent testing programs such as the Department of Transportation-Federal Railroad Administration LPG torching tests on coated plates and full-scale fire engulfment tests on 33,000 gallon capacity LPG tank cars filled with LPG in 1974-75 (4), and Factory Hutual Research testing reports made available to the author by a sublimation compound type coating manufacturer (5,6,7;8). All of these separate sources of experimental data

.have been utilized to form as large.a. data base as is possible

'or i technical evaluation of t&ie thermal performance character-istics and capabilities of the various fireproofing coatings.

Unfortunately, most, if notdirect data have been obtained under all, the available experimental flame contact conditions and/or EVESSON hND ASSOCIATES, INC.

iS

~ ~

e 7

(

under relatively high pressure impinging, or torching, fire conditions, and as such are not directly applicable to those conditions wherein only protection from "radiant heat fluxes" is disired, or-required. However, due to the very wide vari-ation of the types of hydrocarbon fuels in the various direct flame contact tests, and the resultant. wide variation in coating surface incident heat fluxes (from a low of 12,000 BTU/HR SQ-FT to a high of 67,200 BTU/HR SQ-FT), it has been possible to correlate the experimental data in a form that it can be used for the prediction of the required coating thickness for various types of fire conditions ranging from high pressure flames impingement to only incident radiative heat flux considerations.

Table I presents a listing of the different types of hydro-carbon fuels that have been used in the various reported testing programs and the radiative, convective, and total heat transfer rates reported in the research literature for each type of fuel.

A listing of the literature sources for these heat transfer rates is also noted on Table I. As listed in Table I, the radiative heat fluxes range for 5,000 to 39,000 BTU/HR SQ-FT depending on the fuel and fire size, and the convective heat'fluxes range from about 7,000 to 11,000 BTU/HR SQ-FT, depending on the. fire size.

A tabulation of the experimental data used in the engine-ering analyses and evaluations reported herein is presented in Table II. As shown, experimental data for a sublimation compound coating, an intumescent mastic coating, a composite system com-posed of an insulating type concrete with an exterior coating of an intumescent mastic, and an ablative type coating have'been utilized as typical examples of the various fireproofing coatings applicable for the protection of outdoor structural steels and LPG storage tanks. The fuels used in the Table II experimental results include methanol, hexane, JP-4 and LPG. The various coating thicknesses ranged from 0.125 inches to 0.750 inches.

The structural steel substrates include 5/8 inch plate (LPG storage tank shell material) and 8$ ~r31, 8VF39 and 104%49 steel beams. The exposure times for the particular steel substretes to reach 300 oF, 500 F, 800 F and/or 1000 F, as applicable, are also given. The sources of the experimental data are also listed on Table II.

DATA ANALYSES: STRUCTURAL STEEL BEAMS In order to generalize the available direct flames contact and impinging fire test data and develop a generalized engineering data correlation that can be us'ed for any type of fire heating condition, the Table II experimental data have to be expressed as LESSON hND ASSOCIhTES, ?NC.

ft

~ ~ ~ ~

TABLE I

SUMMARY

OF TOTAL CONTACT HEAT FLUXES FOR VARIOUS TYPE HYDROCARBON FLAMES MAXIMUM HEAT TRANSFER TO A COLD TARGET (BTU/HR SQ-FT)

RADIANT CONVECTIVE TOTAL Methanol 5,000 7,000 12,000 Acetone 10,000 7,000 17,000 Hexane 22,500 7,000 29,500 Cyclohexane 31,000 7,000 38,000 JP-4: Small Fires 23,700 7,000 30,700 JP-4: Large Fires 31,000 10,000 41,000 Benzol 39,000 7,000 46,000 LPG: Impinging Type Fires 64,850 Avg LPG: Small spills 25,500 7,000 32,500

REFERENCES:

Atallah, S. and Allen, D. S., "Safe Separation Distances from Liquid Fuel Fires", Fire Technolo , 1, 47 (1971). Industry",

2. Law, M., "Structural Fire Protection in the Process Buildin , 86-90 (18 July 1969).

3. Nei , D.T., Welker, J.M., and Sliepcevich, C.M., "Direct Contact Heat Transfer from Buoyant Diffusion Flames", J. Fire 6 Flammabilit 1, 289 (1970).

4. Rasbash, D.J., Rogowski, Z.E., and Stark, G.W.V.,"Properties of Fires and Liquids", Fuel, 35, (1956).

Fires",

5. Bader, B.E., "Heat Transfer in Liquid Hydrocarbon Fuel Trans-Proceedings, International Symposium for Packaging and portation of. Radioactive Materials, Sandia Corporation and U.S.

Atomic Energy Commission, SC-RR-65-98, Albuquerque, NM (12-15 January 1965).

6. Anderson, C., Townsend, W., Markland, R., and Zook, J., "Comparison of Various Thermal Systems for the Protection of Rail Cars Tested at the FRA/BRL Torching Facility", BRL Interim Memorandum Report No. 459 (December 1975), Funded under Federal Railroad Administration, DCN AR 30026/Req. 731231 WESSON hND ASSOCEhTES, INC.

TAbLE II

SUMMARY

Of EXPERIMENTAL DATA ON THERMAL PROTECTION SYSTEM EXPOSED TO DIRECT FLAMES CONTACT TYPE Of fUEL TYPE OF I OF COATING. ~ INCIDENT HEAT'. "INCHES OF COATING

" TIME FOlL SUBSTRATE TO REACH SPECIP IED SUBSTRATE THICKNESS FLUX PKR THOUDANDS TE%'ERATURE Minutes (BTU/HR SQ FT) BTU/HR SQ-FT Ol'in) 300 of 500 F 800 f 1 ~ 000 SUB LIMITATION COMPOUND Methanol  : ~ BWF39 Beam '.

~ e 0,150 12,000 0 '125 48 u

Hexane BMF39 Beam 0;150 29,500 0 ~ 0051 15 II llexane BWF39 Beam 0.250 29,500 0.0085 '17 I

Hexane 1OMF49 Beam 0.150 29 >500 0 '051 13>5 34

~

Methanol 10MF49 Beam 0, l50 12,000 0 0125 48 105 n llexane IOWF49 Beam 0.217 29,500 0.0074 '128 u Hexane 10MF49 Scam ~ 0,200 29,500 0,0068 120 II 5/8"

>I LYG Press. Plate 0.125 64,850 0.00l93 7>5 14.5 25.5 LPC Presa ~ 5/8" Plate 0 ~ 187 64>850 0 '029 14 24 48 I>

0 LPC Press, 5/8" Plate 0.250 '4,650 0 '038 22 38>S 64 JP 4 5/8" Pla te 0.125 32>500 0.0038 17.4 33 60

>I JP 4 5/8" Pla te 0.250 32 >500 0 '077 49,2 70 6 141 >2 QtTUMESCKNT. MA$TIC IC llexane BWF31 Beam 0.125 30>700'. 0.0041 35 II Hexane 8MF3 1 Beam 0.250 30,700 ~ 0 ~ 0081 64 u llcxane 8'WF31 Seam 0.500 0.0162 120 n 30,700'0, Hexane IOMF49 Beam 0.125 700: 0 '041 45 l>

I>

Hexane 10WF49 Boom ,0.250 30,700 0.0081 73 ~

Hexane 10WF49 Beam 0. 500 30,700 0,0162 132 COMPOSITE SYSTEM ".CON+AD llexane 8WF31 Beam 0,250 30,700 0 '081 50 CRETE + 1 /8" INTUMKSCENT Hexane BMF31 Beam 0 ~ 500 30,700 O,ol 62 85

?QSTIC TQP COATING Hexane 8MF31 Boom 0> 750 30,700 0,0244 125 i+IATIVKCOATING;," LPG Pool Pire 5/8" Plate 0 ~ 125 32,500 0.00385 ~ 12 19 42

>I L'PG Pool Pire 5/8" Plate Oe250 32 >500 0.00760 27 41 9S R FERENCES!

li Anderson, C., Tovnsend> W., Markland> R., and Kook> J"Comparison t the FRA/BRL Torching Facility", BRL Interim Memorandum of Various Thermal Systems for the Protection of Rail Cars Tested

~ Rcport No. 4S9 (Decembet 1975) ~ Funded Undet the Federal Railroad Admin~

istration, DCN AR 30026/Rcq. 731231 2~ Concerning Fire Protective Coatings, A Svmssry of ~ Symposium Presented at the A. I.Ch.E. Meeting in Philadelphia, PA (November 1973) ~

3 ~ Fcldmsn > R., "Fire Retardsncy and Heat Transfer Transmission Control Using Applied Materials" ~ Presented to the Fireproofing and Safety Symposium, Western Research Application Center, Los Angeles > CA (May 1971) ~

4, O'Rourl,c, J.F ~ ~ "The Use of Intumcscent Coatings for Fire protection of Structural Steel", Presented at the. Annual Meeting of the A I ~ Ch E > Philadelphia > PA (November 1973) ~

5> TSI> INC. > Tcchnical Note No. 75120> "Thermo-Lag Subliming System for Extended Fitc Resistance of LPC Stotade Tanks" ~ Januar< 197S ~

the exposure time required to reach a preselected temperature level as a function of the coating thickness, incident heat flux and substrate heat capacity for each particular correlation type of

'coatihg and metallic substrate. Figure 2 presents a of the Figure data for an intumescent coating applied to a variety of structural beams sizes. As shown, the time required

~

for structural steel beams to reach the design limiting tempera-ture of 1000 F can be expressed as a function of (T) (W) '(F),

0.5 where

T Fireproofing coating thickness in inches W Weight of the structural steel beams in lbs/ft F = Total incident heat flux in thousands of BTU/hr sq-ft The Figure 2 correlations have cons dered a fireproofing coating thickness range of 0. 125 inches to 0. 500 inches, structural beam sizes from 8WF31 to 14WF228, and a total incident heat flux of 29,500 BTU/hr sq-ft as being applicable to the ASTM-E-119 flames exposure test method.

The different data correlations shown for the intumescent .

mastic coatings and the sublimation compound coatings adequately illustrate the very significant effect of the coating thermal properties on a generalized engineering correlation. If, or when, sufficient data on the "energy absorption rates" of the various type coatings become available, it should be possible to express the individual data correlations as a single generalized correlation of the type:

b c d e a function of (T, AT, F, W, E )

a where, t Flames exposure time T = Fireproofing coating thickness AT = Temperature rise of structural beam substrate F Total incid nt heat flux W Weight of beam per linear foot exposed to flames heating Coating energy absorption rate.

DATA ANALYSES: LPG STORAGE TANKS Due to the large scale engulfment fire tests and plate torching tests conducted by the Department of .ransportation-Federal Rail-road Administration on full scale 33,000 gallon capacity LPG rail-cars filled with LPG product, and the possible application of these data for fireproofing of other type flammable product storage tanks, particular attention has been given to the Table II experi-AVESSON AND ASSOCIhTES, INC.

~ IQ lVO

~ s,p r ~ ~

~ ~

'EGEND

~ ..-.":"-'... 08WF31 Beams Covered with Intumescent Mastic (1) 010WF49 Beams Covered with Intumescent Mastic (1)

~14WF228 Beams Covered with Intumescent Mastic (1)

O8WF39 Beams Covered with Subliming Compound (5) 10WF49 Beams Covered with Subliming Compound (6)

-'T. ~ Fireproof coating thickness inches

-'.:-: ':: '..!:.-':- F ~ Incident heat flux, thousands of BTU/hr sq-ft Weight per foot of length for Steel Beams, 1bs/ft 00

~<:.:: Time for Beams to 8: .: reach 1000 oF:

1 "A ---,- Coating, th Subliming

i. ': '

': -':-j e

.I. '

"." Time

.1000 for'eams to reach F: Intumescent 80 O I

60 A ~, ~ ~

C/1 C

Time for Beams to reach 40 -. "300 F: Subliming Compound

~ ~ ~ ~ e ~ . Coatings I ~

~ (

~:--:

~ ~

':::.-"-- '::. ~ ~

I 20 I' I '

P

~ ~ ~ ~

~ ~ ~

A ~

~

I t:I

~ ~ ~~ ~ ~ ~ ~ ~

I I ~ ~ ~ ~

'4

~~~~ ~~ ~ ~ ~ ~ ~ ~ ~

I ~ ~ ~ ~

I l ~

I e

01 02

. 05 '

'"lo .20 .40 (T)(W) F FIGURE 2: CORRELATION OF THE THERHAL CKARACTERISTICS OF DIFFERENT TYPE FIREPROOF COATING FOR STRUCTURAL STEEL BEPJ'iS Ie ~ .el I

~ ~ ill<< ' .'I'.< ~ t w

~ I'

mental data relating to this DOT/FRA testing program (4).

~

However, before presenting the results of the data. analyses of

~

the DOT/FRA LPG railcar test programs, it may be of interest

~

~

to note a few of the characteristics associated with LPG

~ ~

~

storage tank fire hazards.

is important to realize that past fire experience Itwater shows that cooling of LPG tanks is not totally effective for the protection of such tanks when the tanks are exposed to full engulfment and/or torching fire conditions, especially when the impinging fire is on the LPG tank vapor space. It is equally important to realize that the newlv developed "passive fireproofing" cannot delay LPG tank BLEVE (Boiling Liquid Expanding Vapor Ex-plosion) for an indefinite time period. conomic considerations, as well as design and system applications considerations, dictate that practical tine exposure limits must be established for these "passive", or fireproofing, protection systems. These exposure limits are influenced by the following considerations:

1. "credible" amount of fuel available to be burned.

2.

3.

The A "credible" rate of fuel release of fire condition(s) to be if a spill fire is involved.

considered. For example, iX Type the downwind distance of flammable vapor-air mixture is to be limited, then the LPG spill surface area must be controlled.

This may require impounding of the spilled LPG at the LPG tank area, or close by, with a resultant possibility of spill fire flames impingement, or high intensity radiant heat fluxes, directly upon the LPG tank.

4. The availability and/or response time for emergency counter-actions such as manual shut-off of flow control valves, time for setting up remote cooling water monitors, time for local Fire Departments to respond, etc.

The failure of an LPG tank exposed to a fire situation is directly related to the tank's steel structural strength char-acteristics as a function of tank shell temperature. In general, the strength of LPG tank steel materials increases as the steel 0 temperature increases to a temperature range of from 600 to 800 F.

Somewhere in the range of 650 to 850 F, depending on the particular.

steel being considered, the strength starts to decrease. At a steel temperature of about 1000 F, the burst strength of an LPG tank will be reduced to about 300 psig internal tank pressure.

'At about 1100 F, the burst strength can'b'e as low as .200 p'sig.

Thus, prolonged exposure to fire heating conditions can reduce the burst pressure capabilities of an LPG tank from the normal range of about lOCO to 1250 psig at anbient temperature conditions to 200 psig, or lower, during a fire situation. Then, depending on the exposure tine, the steel temperature, the relief valve setting'and capacities, and the amount of LPG in the tank, a BLEVE condition could result.

WESSON hND ASSOCIATES, INC.

~

~ Th'e energy stored in an LPG tank, or any pressure vessel for that matter, due to internal pressurization is proportional

'-to the volume available for product vapors and the amount of

.: energy available for release per unit time. A generally accepted method for calculation of the net amount of energy available is to equate the relief valve set pressure to a calculated equivalent o8 TNT per cubic foot of tank volume. This can be done using the relationship:

F Lbs of TNT = 0.00135 V P

P lnP P

a where, V Volume of LPG tank, cubic feet P = LPG tank pressure relief valve set point, psia P Ambient pressure, psia.

The val'ue thus derived 'for a particular tank's TNT equivalent is useful in estimating the over-pressures resulting from a BLEVE condition.

The damage potential of a TNT explosion as a function of the separation distance from the explosion source point can be estimated from the maximum overpressure at the point of interest. Assuming a cylindrical charge of TNT, the maximum overpressure can be estimated from the relationship, Pm = Po 11.34 185.9 19210 Z2 Z3 where, P = maximum overpressure, psi P = Ambient pressure, psia Z ~ 3.967 R/(TJ)

R Distance from explosion source, feet h' TNT equivalent weight, lbs.

The assumption of a cylindrical charge of TNT in Equation 2 gives a conservative value for the overpressures as compared to those for a rectangular charge of TNT. However, the normal configuration of an LPG storage tank dictates the use of the cylindrical shape charge. The variation'f maximum overpressure

~ .

with distance for several TNT equivalent weights has been generated from the Equation 2 and these results are presented in Figure 3.

A cross-plot of Figure 3 is presented in Figure 4 and is somewhat more convenient to use for the estimation of the damage potential due to an LPG tank BLEVE. For reference purposes, the maximum overpressure from a 250 psig LPG tank BLEVE condition is indicated LESSON hND ASSOCIhTES~ INC.

4 e I I~

~ ~ ~

Il l I ~

I ~I I l Damage t

~ ~ ~

~ ~

Steel S!

j ~~~

Id oo- 'l ~ ~ W ~

~, I

~~ I lI I l I l l I ~

F-

~ ~ l ~ t ~, 'I I ~ ~ '

i ~ k t+ 'L. ~ 1 I ~

e'r-cen a tall 0

-! l a. tais.<

l

-l-7 res, .

I i';i I -.i"

,fun .Da in ge:

H4 ei OV Q ~ ~ ~ ~ I i'I !'

4 ~

0 V

~ ~ I

.,: ~ I I-:: ~

. i ~ '..

V I 0

Q ~I =-,.: .bio'erate!da a~e t.'o itpood Xv Xs I V

~

w V -'Xxa~me.l& .concrete 3Qock- '-

CII V pJVIE l=':

- I-'

~

~

~ ~ l I

~ ~ i ~

I  ! I~

I ~ I ~

I.er-:.+.a s s-.window

~

I I I I- ~

i ' i ~~ .-- -'-':

~

~ l l

~ ~ ~ ~ ~ ~

~

)it

~

~~

~ ~ ~ , ~

~~~

~ I << Cncl ~ VI VI r n IV Olio t e VI <<WCIII Sa VI

/dO Distance from Explosion Source, feet FIGURE 3: DPJ'!AGE POTENTIAL'S FRO~! TNT EXPLOS IONS

~~ssaaraPwal

~~~0 g! g%

~ kgHIIM~SE ggli~%%gSEHS~MlgWISltllllWgH~~INISMIIW~INIMWWRRsaaalMN~~WW gggplF

'l

%%ll5 I

, $ 88%5

'WSI I, gl SIC IllBlBR II Fi 1

I W ~

)IIIII' I 8

~ d goal~

t gl ill - ~ s ~ . i.h I

t

~

II 0

II

It

" on Figure 4.do not should be noted that the Figure 4 damage account for "projectile" damage that might potentials result from an LPG tank BLEVE condition.

There are, numerous examples in the literature of the con-sequences of LPG tank fires and BLEVE conditions. However, the most common and frequent cause of major tank failures appears to be from safety relief flare fires burning for prolonged periods of time above the tank's vapor space and/or impingement on the vapor space of adjacent tankage. A review of the literature, available test reports and published articles indicate the following facts:

l. Most engulfment hour fires of fire exposure.

exhaust the tank contents within one

2. Thermal coatings that are approved by nationally recognized and independent testing and/or fire rating agencies are ~

available for fire rating under direct flames contact con-ditions for in excess of a two-hour exposure period.

3. A good medium response time for a City Fire Department and set-up for application of cooling water for LPG storage tanks is about 15 to 20 minutes.

4. The medium time to BLEVE for an unprotected tank is about 14 minutes (somewhat less than the medium response time for the City Fire Department).

5. Safety relief valve fires can be extinguished by cooling of

the tank contents to below that pressure level at which the safety relief valve will open.

6. None of the conventional standard insulation systems now available will withstand all design requirementg and keep the LPG tank vapor space temperature below 120 F temperature is about that for 250/225 psig relief valve setting.

7. Excess flow valves cannot be depended upon alone to stop the flow of fuel due to possible restrictions in the supply lines and leak rates well below that necessary for excess. flow valve operation.

8'. A "passive" thermal protection system (a system that does not require the actuation of protective equipment or manpower response) is just as important a tank design feature as the safety relief valve.

LESSON hND ASSOCIhTES, INC.

9. A "passive" thermal coating that affords at least one-hour of protection should be applied to all LPG tankage to allow firemen to initiate application of supplemental cooling water.

10. Automatic fire, or heat actuated, valves are commercially available 'and are highly reliable. Such valves should be installed in all liquid transfer lines and should be of the full internal type.

As a result of the large number of LPG tank fires and/or BLEVE's that have occurred and are still occurring in this country, and perhaps due in part to some identification of the types of fires that cause such incidents, the DOT/FRA sponsored a research and full scale fire testing program on full size, and filled, 33,000 gallon capacity LPG railroad tank cars. This testing included environmental tests, one-fifth scale preliminary fire tests, full scale spill f re engulfment tests on 33,000 gallon tank cars, and high pressure flame impinging (torching) fire tests on sample size LPG tankthermal material plates protected with most, Some of the if not all protection available, systems failed protection systems.

during environmental tests, others failed during the one-fifth scale tests, and others successfully completed all the required tests. Since the high pressure LPG impinging fire tests r'esulted in the most severe, but realistic and possible, fire heating rates (up to 67,200 BTU/hr sq-ft incident heat fluxes), coating erosion conditions, and coating thermal stress rates and levels, the remainder of this paper will be devoted to the general analysis of the two highest performance level systems resulting from the DOT/FRA

~

(4) experimental testing programs, an ablative type coating and a sublimation compound coating.

From the former an'alyses discussed for structural steel beams, it appeared that the data obtained from the sample plate torching tests should correlate in the form of, c,d t = a function of (Ta , Fb , BT , M )

where, t = Plate exposure time, minutes

'T = Thermal coating thickness, inches 0

4T = Steel plate substrate temperature rise, F F

I Total incident heat flux, thousands of BTU/hr sq-ft W = Steel plate weight per unit area exposed to flames heating, lb's/sq-ft a,b,c,d = Correlating coefficients.

Figure 5 presents the correlating results for the ablative coating and the sublimation compound coating experimental results obtained from the DOT/FRA torching tests on 5/8" thick steel plate samples LESSON hND ASSOClATES) INC.

J S.

LEGEND

'TIME FOR 5/8" PLATE TO REACH 800 F

": 8TDK FOR 5/8" PLATE TO REACH 500 OF

&TINE FOR 5/8" PLATE TO REACH 300 F OPEN POINTS, SUBLDfATION COaiPOUND COATING

~

SOLID POINTS: ABLATIVE COATING I, ~

I i

~ ~ ~

I ' ~

4J 200 6 t

r

// / / /

/ // /

C4 100 o 80 A

// B / /

/ /

/ //

CQ 60

/

0 g',m'

/'O

/ /

/ / //

(

.001 20 i'.

......../. ~

.002

/

/-"- /

~

.004

~ ~

I I

~ i

.01

~ ~

INCHES OF COATING/THOUSANDS OF BTU/HR SQ-FT INCIDENT HEAT FLUX

~ ~

.02 t

~

I

~

.04

~

7 FIGURE 5: CORRELATION OF DOT/FRA LPG TORCHING TESTS RESULTS ON 5/8" THICK LPG TAhK PLATE l'fATERIAL tie!aclttimt>>, "a "I'.)>>tei

in the form of plate exposure time expressed as a function of the

'oating thickness divided by the total incident heat, flux wigh

. 's the metal plate substrate temperatures of 300, 500, and 800 F a correlating parameter. The five test points shown in Table II for the subt.imation compound type coating resulted in an ex-cellent linear correlation for the Figure 5 log-log type of presentation. The two experimental test points (at each of the three noted plate temperatures) for the ablative type coating shown in Table II and the relative locations with respect to the sublimation compound coating correlations for each temper-ature, indicate a linear correlation for the ablative type coating that has the same slope as that of the sublimation com-pound type coating'. You might recall that this characteristic was not true for a comparison of the sublimation compound coating and the intumescent mastic coatings for steel structural beams, wherein the slopes were quite different.

A close examination of the Figure 5 data correlations indicates two important features; one, the parallelism of the linear lines

.shown for the 300, 500, and 800 oF plate temperatures indicated that it should be possible to collapse the three lines to a single line correlation incorporating plate temperature rise as a general correlating parameter and, two, the sublimation compound coating, taking a given plate temperature rise at a given period of exposure, has a higher thermal performance capability than does the ablative coating, using the required coating thickness as a measure of the coating thermal performance capabilities. For example, for a two-hour exposure at an incident heat flux of 30,000 BTU/hr sq-ft (this heat flux could come from any type of fire situation: direct flames contact, flames impingement under pressure, or only rad-iative heat loads) and a limiting plate substrate temperature of 800 oF, the sublimation. compound coating requires only 66/. of the thickness required by the ablative coating (0.180 inches versus 0.273 inches).

If we make an assumption similar to that utilized for the Figure 2 general correlation for structural steel beams wherein it is assumed that the metal substrate heat capacity can be correlated as the beam weight per linear foot, it should be possible to obtain a completely generalized correlation for the sublimation compound coati-. g when applied to metal plate substrates.

As is shomx by Figure 6, such a correlation is possible, and correlates all the Table.II'test data for 5/8".thick steel plate quite well. As shown, the exposure time can be expressed as a general function of the sublimation compound ccating thickness times tho substrate te-...perature rise to an exponent of 0.70

,times the metal plate substrate weight in lbs per sq-ft of exposed surface area to an 0.50 exponent divided by the total incident heat flux in thousands of BTU/hr sq-ft. Thus, the Figure 6 LESSON AND ASSOCLAYES% INC.

lo t o ~

~ ~

I ' I

~ ~

I I r

~ ~~

~ ~

I

~ ~ ~ I ~ ~

~ . . . o ~ 0 ~ ~ ~ ~

~ ~ ~ ~ )

...I '

0 TEST POINTS FOR 300 F METAL SUBSTRATE

~

~ -' '.- OTEST POIXi'TS FOR 500 F ~iiETAL SUBSTRATE

~

00, ': ~

'. 4TEST POINTS FOR 800 F iiIETAL SUBSTRATE ~ ~ ~ ~ ~ ~

~

~, I I ' ~~ ~ ~ ~ ~

~

~ ~ ~

c-.. [ ~ ~ ~ ~

~

~ ~ ~ ~

~ ~

~ ~

1

~ ~ ~ ~~ ~ ~ ~ ~ ~ s o I

~

L ~ o [ o ~ ~ ~ ~ ~ ~ ~ ~ I ~ ' ~

200 I I ~~I

~

~

~ ~ ~

[ ~ ~ ~

I

~

l ~ ~

~ ~

L I,~,

I ~ ~ t' o

~ 1 ~ i ~ . ~ ~

~ ~

I ~ I ~

I "I l l ~ ~

ll[

I ~ '

f

"}

o

~ I

' ~ ~

100 ~ ~ ~ ~ ~ ~

I' I ~ I I

o I

0 ~ ~

I ~ ~ I ~

~ ~

~ ~

I o ~ ~

60 ~ ~ \ ~

1

~ ~ ~ ~ ~ ~

0 e ~

r- -

~ ~

~

I .

~

~

~ o ~

Po 20 o ~

0 Thermal Coating Thickness, inches Temperature Rise of Metal Substrate, OF 10 Xncident.Heat Flux, Thousands of BTU/HR SQ-FT Weight of Metal Substrate Plate, abs/sq-ft of plate surface area exposed to heat flux

~ ~

~, l ~ ~

~ I ~

~

1 ~

I ) ~ ~

~ ~ ~

~ ~ ~ o ~ ~ ~ ~ ~

~ ~ ~ ~

I ~

2 ~

o 0.5 j (F) ~ t ~ ~

o

~ o 0.1 1.0 10 20 XGURE 6: GENERALIZED CORRELATION FOR THE THERK%L EFFECTIVENESS OF TIIE SUBLI 1ITATION CO.PO&i'D COATING APPLIED TO A SUBSTRATE OF CARBON. STEEL PI XTE

~

~

~o ~

~

~; .o ~ ~ I i I

c orrelation can be used for engineering design purposes for the determination of the required sublimation compound coating thick-ness for any given fire situation, given metal plate substrate "thickness', and specified allowable substrate temperature.

The parallelism of the Figure 5 correlations for the sub-limation compound coating and the ablative coating rate" also suggests that a parameter expressing the "energy absorption of the two type coating could be used to make the Figure 6 generalized applicable for. both type coating. However, this has 'orrelation not been done as yet due to a lack of knowledge on the exact energy absorption characteristics of the two coatings, but can be done once this characteristic is defined.

To illustrate the potential usage for the Figure 6 data correlation, let us assume that we wish to thermally protect the roof of a particular product storage tank from the thermal radiation due of an adjoining tank fire situation for a period of one-hour. Typical numbers applicable to such a situation

. would be as follows:

l. Incident radiant heat flux: 12,500 BTU/hr sq-ft

2. Roof thickness: 0.250 inches of carbon steel plate (10.2 lbs/sq-ft)

3. Design allowable roof temperature: 350 F (70 F ambient)

4. Protect with sublimation compound. coating.

From Figure 6 at 60-minutes Elapsed Exposure Time, we read a figure of 2.0. Thus,

2. 0 ~ (T) (4T) (W) (F) or T = 2.0 (12.5)/(280) '10.2)

T ~ 0.152 inches of sublimation compound coating.

Based on the preceeding discussions and engineering data correlations, it can be concluded that LPG tankage can be thermally protected with a "passive" fireproofing coating system that following performance capabilities:

exhibits'he

l. The'assiv'e

. erature to thermal coating must keep the LPG tank steel temp-below 800 oF, for a period of ..two-hours when;,the tank is not more than 807. full of liquid product, and the tank to direct flames impingement from a spill fire below is'exposed the LPG tank having the following characteristics:

a. Incident heat flux of from 40,000 to 50,000 BTU/hr sq-ft WESSON hND ASSOCIATES, INC.

~ ~ j

b. Flame velocity on the order of 100 ft/sec

-"--"- c. Distance from spill surface to LPG tank bottom is 3-ft or less.

2. The thermal protective coating should be durable in the intended exposed environmental service conditions for a period of 20 years, with the top coat renewal being at least five to seven years. During this service period it should not dust, flake, chip, crack, or spall off during normal service conditions.

3. During fire conditions, the residual coating should not spall from the thermal shock due to supplemental water stream cooling.

4. The thermal coating materials should be non-toxic and entirely non-flammable.

5. The material should not contain any asbestos.

6. The material should not be corrosive to structural steels.

7. The materials should be resistant to chemical spills and fumes from those chemicals normally associated with petro-leum and petrochemical processing and storage plants.

The materials should be applicable with airless spray equip-ment and the coating should cure within a maximum time period of three days, at 75 oF and 50% relative humidity.

9. The material should have a bonding strength of not less than 100 psi.

10. When used for protection of low temperature flammable liquid storage or transfer lines l,'such as LPG or LNG), submergence and/or liquid spray contact with the stored product should not result in any adverse consequences on the fireproofing capabilities of the coating. Further, the coating should be able to withstand simultaneous exposure to the low temp-erature liquids and direct flames contact conditions without loss of protective capabilities.

CONCLUSIONS Eased upon the experimental data, data analyses, and dis-cussions presented herein, it can be concluded that:

It is possible to generalize the experimental data obtained from specific rating tests on specified structural substrates with specified coating thicknesses exposed to direct flame WVEssoN AM) AssocIhTEsi INc.

contact fire conditions into generalized engineering cor-

---

'elations for each type of steel substrate and coating which express the protection time as a direct function of the

....-':- "'coating thickness, substrate temperature rise, substrate heat capacity, and total incident heat fluxes. These engineering correlations can then be used for the determin-ation of the required type of coating thickness for a given substrate, given substrate design temperature and given sub-strate heat capacity under any type of fire heating condition (flame contact, impinging flames, and/or flames radiation).

Based on the experimental data presented in'this paper, and now available in the research literature, the type coating gives a superior fireproofing performance, sublimation'ompound as measured by the thickness of coating required with all other applicable parameters held constant, than any other fireproofing coating analyzed in this paper.

REFERENCES O'ourke, J.F., "The Use of Intumescent Coatings for Fire Protection of Structural Steel", Presented at the 1973 Annual Meeting of the A.I.Ch.E. in Philadelphia, PA (November 14, 1973).

2. Kayser, J.N., "Tests of Fireproofing Materials for Structural Steel for Refineries and Chemical Plants", Presented at the .

1973 Annual Meeting of the A.I.Ch.E. in Philadelphia, PA (November 14, 1973) .

Feldman, R., "Fire Retardancy and Heat Transmission Control Using Applied Materials", Presented to the Fireproofing and Safety Symposium, >lestern Research Application Center, Low Angeles, CA (May 27, 1971).

4. Ballistic Research Laboratories, Aberdeen Proving Ground, YD (Anderson, C., Townsend, V., Markland, R., and Zook, J.),

"Comparison of Various Thermal Systems for the Protection of Railcars Tested at the FRA/BRL Torching Facility", Interim Report No. 459 to the Department of Transportation, Federal Railroad Administration, Uashington, DC {December 1975).

Factory Mutual Research, "Fire Endurance Test on Steel Columns Protected with Thermo-Lag 330-1 Coating", Report to TSI, Inc.,

St. Louis, MO (November 6, 1972).

Factory Mutual Research, "ASTM E119 Fire Endurance Test (Modified)

Structural'Steel Colure Protected by Thermo-Lag 330-1 Coa'ting-Design CT-36", Report to TSI, Inc., St. Louis, MO (April 1974).

Factory Mutual Research, "Exploratory Fire Endurance Fire Test on Structural Steel Column with Thermo-Lag 330-1 Coating".

Report to TSI, Inc., St. Louis, YO (November 30, 1973).

8. Factory Mutual Research, "Exploratory Fire Endurance Test on Structural Steel Column with Thermo-Lag 330-1 Coating", Report to TSI, Inc., St. Louis, MO (November 30, 1973).

Q ESSON hND ASSOCIhTES, INC.

, ~ t v'/AA &(le

~ g-gqypz jgpTs Ago oN g

r~- .p~r ~

II I

I/

I I

DIA IL II II II

c. wA>PLE g NN'K C?

E-E 5-Li II II )~~L llfj El Q,25'-IO I II I II I II ZZ AR~ 8" I.

0 F ZAIRE H< FLO-TWO +<Gl,'P

'R K F, PM/Q 1-0" I 2-34-3+ ' 2- $ + $ 6 -naoroHDDIPf

>q'i I Kt 625'-Io 2

I ii 8)

C'AR SgX q TA LZe P EVAL L 5-4 (I

WAI I H OL g5 ONK RAc>w slX) Q S E.CTlo(4 6 I:.

lilt 1BCSP Llk R FIPazotlE 4- ll8LT I PIRE ~Kg- llNI'IR

l. ~F-

SIZE (WIDTH) I II SINGLE LEAF STEEL

'2 DIAMOND PATTERN PLATE

'ED VINYL Reinforced for 150 lbs. per square foot live load.

GRIP I

I-I I

FLUSH LIFT I

STEEL J 6 HANDLE STRAP J Z I' STAINLESS STEEL ANCHOR RCO VINYL General Contractor, Please Note:

Be careful not to rack or twist frame ~e~ setting unit.

K IJ IJ J 4 --)

U --- BRASSSLAM LOCK WITH HANDLE I I GRIP Block up and shim the frame if necessary to be sure II I door rests evenly on frame all around.

IJ a 0 A. I A. AUTOMATIC NOLO.

N I I OPEN ARM

'I( I I IYES" DIAMOND II II II PLATE COVER I

II II II tuj OR SION BARS

/ FaCtOQ FiniSh: Steel red oxide primer Hardware II I d~Gr J's HINOC TORSI~

cadmium plated steel BARS SECTION B-B Manufacturers of Doors for Special Services THE BILCO COMPANY (Cover in Open Position) P

~ Hcw Haven, Connecticut 06505 PLAN. VIEW ftuufTITT HFE IIITII LEICTI REMOVABLE Ql/4"XS"XS" STEEL 2'4"x2'4" KEY WRENCH ANGLE FRAME fIIE Ql ltilT l I/4" STEEL DIAMOND PLATE COVER TGRSIGN5 g 7i5 ilITg Q2 2'4" x 2W'gal 0 < gQ iiST:i-a x 3J+I 3'4" x 3'4" SLAM LOCK I'2.8 IJ g 53. i N

GI I ~

lilt'TEEL SIZE (WIDTH) ilia lT H NGE ARCH'T. OR ENG'R MASONRY OPENING ANCHOR a e

F PURCHASE ORDER ATE PROJECT GEN'L. CONTRACTOR PURCHASER This drawing is the property of The Bilco Company K HlH665 and incorporates specifications and patented designs BILCO REPRESEHTATIV 0 HaOOP&l~

in which The Bilco Company has proprietary rights and, accordingly, is not to bc reproduced without DWG. NO. DATE

<4 BLED, AlAW the express written consent of The Bilco Company.

~ ~s ~  % a ERAR

i)'L~};p%S

+plCAN ELfci AMERICAN ELECTRIC PO'WER SERVICE CORPORATION OH ER Systsw May 15, 1984 suszEcTi D. C. Cook Nuclear Plant Fire Rated Floor Hatches RFC's 01-2676 and 02-2692 V. Del Favero F. S. Van Pelt JI.

g In response to Item 4, "Provide any alternatives to the insulation or compensatory measures that may be available",

of the NRC letter to Mr. Dolan dated April 4, 1984, the following measures were considered:

Provide a vertical fire rated enclosure above the hatch. This is not possible due to the limited space and close proximity of electrical cabinets which require access for maintenance and operation.

2. Laying a fire rated blanket above door. This impedes the operation of the hatch, and creates a personnel safety problem.

3 ~ Provide a vertical fire rated enclosure below hatch. This is not possible due to many interferences with cables, conduit, troughs, and cabinets.

4 ~ Add a horizontal fire rated panel below hatch.

There is an interference with the access ladder and a personnel safety problem of access to the hatch.

5. Replace hatch with a fire rated hatch. No prefabricated fire rated floor hatch is available.

We have contacted The Bilco Company about design and testing of a fire rated floor hatch. (See.

attached communications)

V. Del Favero VDF:b cc:: ST Fox W Q<~

~ use a CVCVraa

AMERICAN ELECTRIC POWER Service Corporation AEP 1 Riverside Plaza P.O. Box 16631

$614) 223-1000 Columbus, Ohio 43216-6631 March 13 1984 Robert Lyons, President The Bilco Company P 0 Box 1203 New Haven, Connecticut 06505 RE: D. C. Cook Nuclear Plant bc=0-4aeo *

Dear Mr. Lyons:

In a recent telephone conversation you may recall our request that The Bilco Company submit a quotation for furnishing a 2'-6" x 3'-0" floor hatch bearing an Underwriters "A" label.

It is understood that you do not manufacture a U.L. rated floor hatch, however, as AEP anticipates the likelihood that retro-fitting of several Bilco installations in the subject plant may be required, we need to make allowance for such a contingency.

If this request is agreeable to you may we suggest that your quotation also include the cost of one submission to U.L. for testing and labeling and a separate price for each successive U.L.

application as may be necessary.

At a future date, if AEP becomes committed to the replacement of hatches as referred to above, the program will probably be industry wide and these rated hatches will be in demand.

As a long standing purchaser of many of your products, we hope that you will be able to furnish us with the desired pricing data.

If you require any further information, please don't hesitate to contact us.

Your early response will"be greatly appreciated.

Very truly yours, A. C. Macksoud Chief Architect ACM: b

THE BILCO COMPANY P.O. BOX 1203 NEW HAVEN, CT 065O5 March 21 ~ 1984 Mr. h. C. Macksoud Chief Architect American Electric Power Service Corporation 1 Riveraide Plaza P ~ 0 ~ Box 16631 Columbus, Ohio 43216%631 RE: D. C ~ Cook Nuclear Plant tg 4 200 A

Dear Mr Macksoud:

~

Thank you for your letter of March 13, 1984 concerning your require-ments for a floor door to carry an Underwriters "A" label .

'Re have contacted both Underwriters Laboratories and Factory Mutual Engineering Division with requests for costs to fire test one of our single leaf J-3 doors, size 2'6" x 3'0", and also one of our J<<4 doors, size 5'0" x 5'0", in double leaf design.

Just as soon as I receive some information from them I hope I will be better able to answer your letter and I will be in touch with you at that time Yours truly, THE M . ANY

\

i p Robert: on Q

RJL:wfg RECEfVFD MAR 2 3 1984 ArcRtechu.d S

~tt~'lANI AND MAIH OCFCc 07 waseca cvaccg wc'avcu rnklklcl 'Til IIT Ihh'I Q'1 e%t'1

e I: *4