19 lnsulation systems fot low tempercture tanks
This still continues, with allsides frequently claiming victory. lt is not unusual for composite systems using PUF for the inner layers and cellular glass for the outer layer to be specified. An area where particularcare is needed is in the fitting of insula-
tion to valves and other fittings. Infrared thermography discussed in Section 19.10 is a useful tool for inspecting low temperature pipe insulation systems. The identification of the causes of failure or lack of longevity of such pipe insulation systems is not always straightforward. A refinery in the UK had problems of early breakdown of its carefully installed cellular, glass-based insulation. The cause ofthis problem was eventuallyhaced to the fact that the workforce fre-
quently used the larger insulated pipes as walkways to the more inaccessible parts ofthe site. The brittle cellularglass was
no match for the workers' boots!
19.8 Heat leak calculations The basis of the heat leak calculations is quite straightforward. Itisonlyin the detail that the subject becomes a little more interesting. The tank insulation system is divided up into the areas where similar materials or combinations of materials have been used. For a typical full containment tank these would be:
. . .
The central area ofthe tank base
.
The upper tank wall where no thermal protection has been installed
.
The oerioheral area ofthe tank base The lower tank wall where thermal orotection has been installed
The tank roof
19.8.1 Basic calculation methods The basic equation to calculate the heatflux through a particular component to be adopted where a slngle insulation material is used is: H=k
xAxAT
iL
equ 19.1
where:
H k
= =
A = AT = L =
heat flux through component (W) thermal conductivity of the insulation material
(W/m'K) area of component (m2) hot to cold face temperature range ('K)
thickness of component (m)
equ 19.2
wnere:
U R1 & R2 etc
thermal transmittance (Wm'? 'K)
=
thermal resistance ofthe various insulation components (m2'K iV) - calculated from equation '19.3 below equ 19.3
Hence the equation to calculate the heat flux through a multilayer component is: H=U
xAxAT
equ 19.4
It is not normal to include surface resistance in these calculations.
396 STORAGE TANKS & EQUIPMENT
The way out of this apparent impasse is to assume a temperature profile for the various interfaces within the system, use this to calculate the mean temperatures of the individual layers and this in turn to obtain the k values of the various materials to perform the initial calculation. The results of this calculation allow the interface temperatures to be recalculated and the k values to be revised. Asecond calculation is carried out and the results ofthis allow a further revision of the k values. A couple offurther iterations should show temperature values at the interfaces converging and this should be sufficient. Figure 19.23 shows a numerical example of this Drocess. This is a tedious calculation to carryoutbyhand and is ideallysuited to Excel spreadsheets.
19.8.2 Thermal conductivity values Initially, the source of the thermal conductivity values (or K values) to be used in the basic calculations is the manufacture's technical literature. Mosi low temperature tank designers will produce their own detailed technical specifications for the supply and installation of the different parts of the insulation system. Within such specifications it would be unusual if there were not some means of confirming the K values of the materials. This could take the form of regular samples being taken from the place of production, be this a factory for materials such as slab stock PU F, PVC foam, cellular glass, glass fibre or mineral wool, or the construction site for such materials as perlite or site expanded PUF.
These samples would be sent to an agreed laboratory where the K values would be verified at the appropriate temperature (or range of tem peratures). lt is usual for the eventual owners of the tank to witness this testing, either themselves or via their hired engineering or inspection companies. As has been mentioned earlier, the measurement of low temperature insulation properties is noteasyand should be leftto those skilled and experienced in this work. Note that this testing is usually conducted in air.
For porous materials such as perlite, glass fibre or mineral wool, the vapour within which the insulation material is operaf ing wiil have a significant effect on lts effective K value. Thus the test results which are based on air as the interstitial gas will have to be adjusted to account for the presence of a different gas. This is discussed in Section 19.8.3. made from different materials, such as is the case for resilient blankets supported by being impaled on pins, the effective K valueto be used in the calculations must be adjusted to take account of the short circuiting effect ofthe pins. Asuitable calculation method for making this adjustment is given in section ,q3 of
Reference 19.3.
=
R=L/K
insulation material within the multi-layersystem must be known before the thermal resistance of that layer can be calculated, and the thermal resistance is required to establish the mean remperarure.
Where the insulation material has been penetrated by items
Where more than one material is used the following method taken from section A3 oI Reference 79.3 is used: U = 1/ (R1+ R2 + R3+.....+Rn)
l\4ost thermal insulation materials have thermal conductivities which change with temperature. This is illustrated for cellular glass by Figure 19.6. Hence the mean temperature of a layer of
Certain insulation materials have thermal properties which change with time. PUF used as external insulation may display an increase in its K value as the original foaming gas within the cells is progressively replaced byair.ltis importantthatin these circumstances a suitably aged property is used in the calculation.
19.8.3 The influence of different interstitial gases The various equations which allow the K values of the vapours ofthe various low temDerature oroducts to be calculated at dif-
19 lnsulatjon systems for low temperature tanks
A€Bumptions for the exercise: Outer surface tamperature lnner sutfsoa ternperature Intef,stltialgas Installed perlle density lnstallsd glass fibre d€n8ity Section through wa[:
+40'c
To Ti
-165 'C Methane 60 kglm3 24 relm3
Matedal Thickne6s(mm) 600 5 Sleel linet 960 Psdite 240 Glasslibre 10 Inside 90/6 ni inner tank Assumed thermalgradienl ('C) Outside Conc€te
Conc|Eie,/perlite P€rlite/gla8s tibre Glass fibr€/product TRIAL No 1 Calculate K values: Concrete
Perlite
fibre step
step 3
Total h€at f,ux (tdal 1) Matedal
.
Glass fib|e
1
148 108
313.0000
310.256
313,0000
310.7349
174,7136 174.5*1 108,0000 108-0000
value Themalresistance(rn2 "K iv) AT 2.0985 1.6 0.3750 600 960 0.0401584 23.9054 133.771407 6S.1301 240 0.0194273 12.3538 36.6341 m5.0000 Total
Thidmess (mm) K
ConaGte Perlite
Figure
3{X}
KS1 8.252384i| R y 0.137422 0.0401584 Wm'K Kp calculate K value of interstitial gas (lQ2) 40'K ATfg 0.0133521 dm"K l{gz fuctor ftom Figure 19.25 1.455 f calculale K value of glass fibr€ 0-0194273 Kg
1
FlI(
3'13.0000 310.C015 117.1301 108.0000
313
calculate K value of petlib (KP)
step 2
Heat
CK)
Taken as 1.60 Wm'K tiroughout calqiation calculate K value of hterstitial gas (f€1) 155 'K ATp 0.0243158 w,lm'K
step l
step 2
Gla6s
h thermalcalculation
lgnore in thermalcahulalion
40 30 -125 -'165
Air/concrelg
'
lgnor€
(dmz)
5.5959
9.23 An sxample of a multilayer ln6ulation @mponent cslcllation - page
t STORAGE TANKS & EQUIFII'ENT 397
19 lnsulation systems for low temperdturc tanks
TRIAL No 2 Calcxrlate K values:
Concrets
Pedite
step
Taken as 1.60 Wm'K thrcughout calculation calculate K value of intersfital gas (lg1)
1
ATp 133.7714'K Kgl 0.0266994 w/m'K calculale K value of pertite (Kp) R 7.87Tt517 y 0.1374?2. Kp o.o428oa w/m'K calculale K value of inteGfitial gas (Kg2) Arfu 69.1901 'K Kg2 0.01499s6 w/m'K factor from Figure 9.2S t 1.455 calculate K value of glass fibre Kfs 0.0218187
st€p 2
Glass
fib|€
.l
step
step 2
1
step 3
Total heat flux (biat 1 )
Material Concret€ Perlfte cfass fibr6
Heat Ftux
Kvalue Thermal resistance(mr.XLn) at 600 1.8 0.3750 2,2744 960 0.042808 22-4287 136.012036 240 O.O21B|B7 10.999g 66.7,13€ Toiat $.9005 208.0000
Thlckness (mm)
(dmz)
6.0650
TRIAL No 3 Calculale K values:
Conqeie
Pedite
step
Taken as 1.60 w/m'K throughout c€lculalion calculate K value ot interstitiat gas (Kg1)
1
ATp 136.0i20 'K lQl 0.0285485 w/m'K calculate K value of pedite (Kp) R 7.9020321 y 0.137422. Kp 0.0426242 wtm'K calculale K value of interstitial gas (Kg2) ATfs 66.7136 'K Kg2 0.0148s87 dm'K
step 2
Glass
fibr€ step
1
step 2
factor from Figure 1 9.25 1.455 calculate K value of glass jibre
step 3
Kfg
0.02i6194
Total heat flux (trlat 1)
Matedal Concrete pedite Glass fibre
Kvalue The.malresistance(flf "tgu/) AT 600 1.6 0.3750 2.2611 960 0.0426242 Z2.Sn4 135.8027i18 240 0.0216194 11.'1011 66.936.1
Thickness (mm)
Total H€at Flux
Figure 19.23 An example
(dm2)
6.0297
ofa multi-tayer insulation component calcutation - page 2
398 STORAGE TANKS & EQUIPMENT
33.9986
205.0000
1
'efent temperatures are given in Figure '19.24. The source of tis data is Reference 19.4.
*
o.oos6r
r
u'"tito'Lt'Jli 5
9l7 t 1o'
lT'l''
7
nT
.-,e6rr,
t'
lr
rrr,.1-.""" \Ttll
--
^T
L
19.8.4 Galculation of the hot face temperature
The actual maximum temperatures to which the tank roof and walls will be subjected influenced by the local solar radiation maximum levels, the attitude ofthe surface in question, the prevailing weatherconditions (clear orcloudy skies)and the nature of the external surfaces. Data and calculation methods allowing the actual maximum temperatures to be calculated for any set of circumstances are given in section AG ot Reference 19 3
uuttilo'[t.ll' o.oorot . I 258'10
For the tank base. two possibilities exist:
)r'
.
-i--L'r',
level. Purchasing expensive energy, be it electrical, steam or heated brine to boil off more product than is necessary incurring further costs in terms of re-liquefaction or product loss to atmosphere is clearly a nonsense. Consequently a design hot face value as low as possible is used. -5 'C is a not unusualvalue to use in the calculations in these circumstances.
T,-T, fK)
:
g u
For tanks built on the ground, the hot face temperature is based on the operational settings of the base heating control system. lt is clearly unwise to have the base heating
system maintaining the base temperature at too high a
Tl
t
lnsulation systems for low temperaturc tanks
The design specification will require a certain maximum heat leak into the tank contents as described in Section 19.7 Ofren the only data given is the maximum design ambient temperature, which is a shade temperature at the tank's geographic loca on.
Fomula for themal conductivity In wm"K o oorou
I
T+12CK)
re 1 L24 Formulae for calculating the K values of various gases at
d
ifferent
Forcalculating the Kvalue ofperlite with the air replaced byone cf these gases at a particular set oftemperature conditions, the 'ollowino formula can be used:
xo=Kn(-v)+
/^,+o-1f
equ 19.5
.
For tanks built on elevated foundations, a hot face design temperature equal to the maximum shade temperature would seem to be a sensible choice. There is perhaps a case for using a lower temperature. Experience suggests that the space beneath the base slab of such tanks is a cold place to be, even on hot days.
vhere:
Kp =
thermal conductivity of perlite in the replace' ment interstitial gas (Wm 'K)
Ks =
thermal conductivity of the interstitial gasatthe appropriate temPerature (Wm "K)
R" = y = p =
0.114i Ks+ 3.608
Acommon wayforthe tank maximum heat leakto be specifled to the tank and insulation system designer is to express it in terms of the escape to atmosphere of a percentage of the full
3.9x10'3xp087
tank contents perday. Hence for a large LNG tank we may see:
installed perlite density (kgim3)
For calculating the K value of glass fibre or mineral wool, the graph shown in Figure 19.25 can be used. (Both equation 19.5 and Figure '19.25 are attributed to the late Dr lan Leadley of Whessoe.)
19.8.5 Overall heat leak
"The maximum heat leak shall not exceed 0.05% of the full tank contents per day on the assumption that the tank contents are considered to be pure methane." The latter requirement to consider the tank contents as a pure product is to avoid the complication of working out the latent heat ofthe LNG which may have a range ofcompositions and to
lhermal conduclivity glass fibre gas = factor x average conductivity of
t
Figurc 19.25 K value ofglass fibre of mineralwool
STORAGE TANKS & EQUIPMENT 399
19 Insulation systems for low temperature tanks
avoid subsequent contentious arguments. lt is normal to make the same form of wording for any tank containing a mixed product, expressing the permitted heat leak in terms ofa percentage ofthe major constituent. The following points are worth bearing in mind:
.
.
lt may be necessaryto run the in-tank pumps during the test period to ensure proper mixing. Allowance for the energy inout from this source must be made.
.
Boil off is known not to occur at a uniform rate, but rather as a series of irregular "burps". This is another reason why the
test must be conducted over a protracted period.
lt is importantto use the correct density in the calculation of
the permitted heat leak. For LNG a latent heat of 507.0 kJ/kg should be used with the pure methane density of 0.422. Using the design density of LNG (frequently given as 0.48) will give too high a value of the permitted heat leak.
.
.
Arrangements must be made to record and take account of the effects of wind and solar radiation.
These difficulties combine to make a physical heat leak test time consuming, expensive and inconclusive. To avoid this
lt is often presumed that the worst conditions pertaining at any point on the tank outer surface at any time during the whole day will persist for the full 24 hours. This has occasionally become a point of dispute beh,veen the owner and the designer, with the tank designer claiming that it represents an unnecessarily conservative interpretation.
problem area, the following procedure is often adopted:
.
The tank designer must prepare detailed heat leak calculations together with the appropriate certification (and possibly QA records if these are available at the time) to demonstrate that the materials used havethe required thermal properties.
.
The full tank contents is usually taken to mean just that, i.e. with no deduction for in-tank pump NPSH etc.
.
.
Whilst the calculations seek to cover all of the sources of possible heat leakage from tank to atmosphere, there will
These calculations and the associated documentation will be submitted to an expert third party, previously agreed by both the owner and the tank contractor, who would review the calculations and whose findings would be binding on both parties.
probably be some which have been ignored or overlooked such as the smaller connected pipe connections. To cover for these uncertainties, it is usual for the designer to aim for a calculated heat leak lower than the full target value. A not unusual starting point would be to aim for 85% of the full value in the first instance.
A well
set out heat leak calculation for a large full containment type LNG tank is shown in Figure 19.26. This makes use of a series of linked Excel soreadsheets.
19.9 Heat leak testing With the customer or his engineer setting a heat leakage requirement for the tank and the distinct possibility that at least some of the process equipment will be designed based on this figure, it would seem sensible to test the finished storage system to see that it fullllls this performance criteria. This is not as simple as it would appear for a variety of reasons:
.
. . .
lvleasuring the heat leak will require either a significant change in the tank liquid level to occur, which may take
An infrared camera will produce images which will identify areas where the heat leak is abnormal or merelydifferentfrom the surrounding areas of insulation. lt is a useful tool both at the time of tank commissioning and as an occasional maintenance device to locate any changes in the thermal insulation system and its performance, perhaps due to such time dependent phenomena as insulation material degradation or perlite settlement. The equipment is nowadays quite cheap to purchase, or there are companies who will come and perform this service.
Insulation problems from the past and their lessons
The tank must be fullor close to full at the time of the test to avoid contentious arguments revolving around the extrapolation ofthe heat leakfrom a lowerliquid levelto a fulltank.
service.
The tank must not be subject to any liquid movements during the test period.
The test must be carried out at a time when barometric Ambient temperature must be monitored throughout the test period.
.
19.10 The use of the infrared camera
some days depending upon the accuracy ofthe level measuring equipment provided, or will require the accurate measurement of the vapour flow through the vapour outlet line, something difficult and expensive to achieve. Vapour flow measurement is not a normal part of the tank instrumentation. For a large LNG tank with a specified boil off rate of less than 0.05% (a typical figure for such tanks) the level change will be of the order of 15 mm/day. To get a sensible measurement which will be sufiicient to negate any uncertainty caused by tolerances on gauging accuracy, it is clear that the test duration must run into several days.
pressure is anticipated to remain relatively constant.
.
One of the added advantages of this procedure is that in the event of a shortfall being found in the thermal insulation provided, then this can be made good prior to the tank entering service. This could perhaps be by a simple addition to the thickness ofthe insulation on the suspended deck, an action with no knock-on effects. With the physicaltest route, this pre-commissioning adjustment is not possible.
For LNG LPG and other mixed products, it is necessary to determine the composition of the liquid in the tank. This will require sampling as the iank is filled.
4OO STORAGE TANKS & EQUIPMENT
19.1 1
19.11.'l Base insulation failure Two LNG tanks belonging to GAZ Metropolitan in lMontreal, Canada, had been in continuous satisfactory service until July 1990, when instrumentation in the tank base ofone of the tanks began to show evidence of cold spots. After double checking and adding new thermocouples, the problem persisted and it
became obvious that the tank required to be taken out of Following decommissioning it was found that the cellular glass base insulation was the subject of massive cracking and mechanical breakdown. This damage eventually necessitated the lifting ofthe innertank andthe complete replacement of the cellularglass base insula-
tion, an expensive and time consuming process.
An investigation into the cause of this base insulation failure was carried out and this is reported in Reference 19.5. The investigation revealed that in July 1990, blasting work had been carried out within 200 m ofthe two tanks. The tank nearest to the blasting was full of Iiquid at the time and undamaged. The tankfurthest awayfrom the blasting location was fllled to 20% of
Figure
'1
gi
ai
$$ 9!
E]II
htil r
tsl
r'1
n 9-E
:
E=s' =s' i6 ':+ E
! ! 5! Ed i.E
SeS9qqq Fq$$ECH
ls
Ei *seRsBg *6bb{oo
page
t
sn
Fl
Fl fl
19 lnsulation systems for low tempenture tanks
b
!
i
+
gl9 383 bbb
teF
lf;lFtF|F|nt : I
i b
nHB
ri
=!5! F: ;:
I FI ;;-Ec 13
F" i!
ll ks 9r,) ** 'i- lEl fi'= [l $B €HElll
€; s*:$€
it
:E
3
ffiHuffi
_
$tiiEiiigiiitiiui,u
9.26 A typlcal heat leak calculalion for a laBe LNG tank
STORAGE TANKS & EQUIPMENT 401
a
€
f
19 lnsulation systems for low tempenturc tanks
!
t I ! I I t
6
!
!;€ B
! T
I
t I t tz E
{ l!
g
t
.*-t €t
iie
I .
EE
d
!rii ! i9
.6 6i 6
iE€€€
: EB€
iEIiS
EEi!E i3"r,r.i
Ei: ili
I !
tEt
r
is:
! dd
g3 ,t;
,
i
EEE
*"_" ! +" oi *E *i t *" *'
E
r
* ia
!ll
-.8
.:
t! ,t
5
I
-''
i
!
[:l
,!*
;3
Er
I
t-+ rtl
q E-,' I (t
85
tl t;
rR i !E
t -t t. B
I 3
E .l
5E 5
IE L fB
rl -r;i
B.T
I
c
a,Il s Ei B fii I TE H €€ {3 :
{ 5* E;
!l
lx
T q 5i E!5 g tl
H H
g
fl
Figure 19-26 A typical heat leak celculation for a laEe LNG tank - page 2
402 STORAGE TANKS & EQUIPMENT
t !
I
.t RE
i;' 8X dd
dd
EE
ij
It
t it 8E -s {: i
3 t* 1E 5
l!
i{
se EC
Ei
€f iitE
: ii
E !F
E
l? "q I It
3
!
j- E * tl
.:
,l
ar
F
rId
ilii ;ii
*rttt*r :i;:;;;
lit"
I r
$g313l3
-E
33 i
gE€gEgI 3; l_
,.,,",. f rEi .!tE;ft
G
I
g!:
ffis EEI
3t
33
5x s!
EE,s3;
,3 !: ri'
il i!'!.s 595$ir ii'i:
:! s:; jI ; ,;i ii i i, ii-'e
*r :;u : i:t
;e
c r:i::iii ffi { .;t!; r l* i* j
;
E
I3
I
!
! E
9 !
I lr I ! f I3
Figure 1 9-26 A typical heat leak calcutation for a large LNG tank - page J
:
1
-q E- ir
ri: 8!:
ir! i
3
i !
e
! I "5
g g
t
I I
*
f
t
E
?
!
a
:
i I
9 lnsulation systems for low tempenture tanks
$ =,,!
u ' $ $E ! I .E.liii i
EHtili r Ir gii:- tl ii
I
=
:,
i i i "
I lli 'f€ i
s$
ral l-l ;5!Fl l"l '6 i '
I
ii!l'i
;
r
!
+E E I : ,. E t:.' :q;9 !i
att t!
STORAGE TANKS & EQUIPMENT 403
1
I
iE
r3:!:!i
tt
€
!
:
!-
I
rl is
i
:
!E! i
3s 3iE i: t::i ': !i i i ,.r i i c;
ii
rrM trrrt
_
page 4
9939:
9.26 A typical h6at leak calculalion for a large LNG tank
I
€ Itr'"t FTE€
"r 5gr;
6E;: :;:t
,,. , .,,,. + i,rt -!ij;Ff ii
€r EE;E-;E EgIFI:: l5 f fi l;El i
g
i :ii:3|i ,. i; s;aa;ce i! i: ,iij;ii
19 lnsulation systems for low temperature tanks
Figure
404 STORAGE TANKS & EQUIPMENT
l-l ie tBt
f-t :
:3
-'
j
l.t -ri lEt
lgql5l
s
a
r^
!F
t*
;-;
it
t T
I
I
3i*$
Fl* BEg:
:
: iii i I
E
E itt
!ls
i:ii
lxi 6 |i
I
! !! ;i
'; g
I I -!
E
I E
I
I I t
Figure
1
iE"
":lYtYtl r ss;sss; ! N
:; v:
.g
l€ EFi 6
tt
-5.9
i is €;t
"E
I
Ei3 i* : : E E*f
FC
;::;.S srl:t
gE
+e
ii
FJE:.;
:;:*E
E
E
EcEsEEs 3 €, ,,,, lEl€ +i €ggE i 9rl ,
l +! :j E
'
IE ;'6 "{
E cE s ! F Er i; E €! i i 3 Eg E;
d'
€
! .p
$
; E -5
E
!8 5*E
€€€ .E
I
I I
{
TE
9 E$;€;
g 1€$iE r i _c a; E c
I
3
t
* 6 I
19 Insulation systems for low tempemture tanks
!b
;*.
e+-
tt 6 lelq: a .r t ;3 ei t;.'i c .r g 3 I! € € € I *s -." > t _E E l'l b € ' -i = -n- -r'l ; E' s I tl8lt l lB EF!E ^r E : iI os ';6qE t; -3[ 5l'lE 6l"l -:fr s Y IJH l-i 3 E -: B ; A\b i 5 s ti ! I t: -€- EEI 9 FtdlJ e :lF
i! P3
_!Er#Ft €3 €i lii t i
P
E
i* E!E! {s :i i; E !.9 ES r ii ii Eg F€€
{5
9.26 A typical heat leak calculalion for a large LNG tank - page 5
STORAGE TANKS & EQUIPMENT 405
1
;-q
gE
knks
:>p
96N;
,'
FF
.3
ir
AE
a f,
a
E
*i
B os
i
i4;;
=oo
9ss essr
.. ;'
i.i : -
te€lEs E; EBdr Eiriss
.'
;
Ei EqE ; eE
!
F B 5
t g g : €
=
i
€i i
*€'ee+F :;
q
E:
P
i -d
9r
€!
t8
b&
-d 5i
it*
I
tl I-I E €I ] tt
3$i-" .E^qq a-dd
6
--E o ll!.!.ix ;'.d Eg i 3I t. e s i . i i gi i x 3 r i i ! ; i 5 5 i F F E Fg
B 9 699
o
E Ef ;E ; $ E FE€EE gI g €r
.i E
I
9.26 A typical heat le€k calculation for a large LNG tank - page 6
!i ; E ii
3:"
3q -!3 EE
a5t 6<
- 8 *^ri -a 3
_€ P y .. ..
_a
;
;$ ;8
E E r'6 FI E"
19 lnsulation systems for low temperature
Figura
406 STORAGE TANKS & EOUIPMENT
;
8
z
s
f
g
z
"s9 aE.!
tt 'Eg E5 Ed E;
t :tE !ET 3
a6'
EE
g
9E
g
i! td
Ps
*t
E
::
19
eAEBg 4oooo
l4oooo
lnsulation systems for low temperature tanks
i
E
5s
E
3
E
tl
€sEAsg i :i r i:---tiisBEs *€Ei:"'"=: giE9qEsg
:i El EeEsE
3
€ii'ln;F :: 55
tt l-l
€is -""' i B5;:3 :
tl
qi
9r
.j
rE E*
I llE
,l-
:; : s-o
&
I
lil $ i ] O;E iFliFlsFFl+Fl I ],8
HEffiffi I ll?iX "s. r;di ;;;;r
.6!
riEie$t$$3 -Et Sigl!33339
Figure 1 9.26 A typical heat leak calculation for a large LNG tank - page 7
STORAGE TANKS & EQUIPMENT 407
19 lnsulation systems for low tempercture tanks
! 6E
;-s 9d
E;>>=== T9
d7:gq
E,o66eFi. _ltn*p ot,t* 6-
!6
j!
;P E:Eg{io -+lrE i:55 "qEg'::PF
E;E! ;.s1 F6dotsiF
Figufe 19.26 A typ cal heat leak calculation for a Lafge LNG tank - page
408 STORAGE TANKS & EQUIPMENT
I
.^o
q -3 js
vx =n
E
I
I !
€
{i 6lF !tF et ;_EF 9:iE
-d6 PF :-s.n
UPS
E* :*a
s
a
r
19
Insulation systems for low temperaturc tanks
:s capacity and was damaged as described. The investigation :ound that the partiallyfilled tank had a higher natural frequency
.
Exposed to frequent waterfall events, due to concentrated rainfall from the roof and shell.
:han the fuller tank and was thus more susceptible to damage iom the blasfinduced ground motions. -lhe lesson from this incident is clearly that any blasting on the same, or adjacentsites should be avoided, and ifthis is not possible, then detailed investigations should be carried out to deiermine the susceptibility of tanks with brittle base insulation naierials to damage.
.
Complications to detailed design caused by holding-down anchors.
.
The correct selection of materials, detailed design and careful installation together with regular inspection all have their part to play in this area.
19.1 1.4
Perlite settlement
19.11.2 External vapour sealing A number of low temoerature tanks with external thermal insuation on their shells, and in some cases also on their roofs, rave come to grief over the years. The reasons for these failJres are usually associated with poor performance of their ex:ernal vapour sealing arrangements. This has allowed moisiure-laden air to invade the insulation material and form ice /vithin the insulation or on the tank shell beneath the insulation. The damage can manifest itself immediately following the tank cool down or following several years in service.
Higher than anticipated heat in leak and consequent product Soiloff, the appearance of external condensation orofice spots or (in at least one case) sudden failure and collapse ofallor part of the shell insulation are the usual signs.
The lesson here is that the external vapour seal and its ong-term abilityto keep atmospheric moisture out ofthe insulation material is viialto the survival ofthe insulation system. Corr.ect material selection, sDecification and installation are all aclivities, which will help to ensure that the required performance and service life are obtained.
19,11.3 Bottom corners The bottom corner of tanks where the wall insulation is on the outer surface ofthe shell and the base insulation is beneath the iank bottom, have on occasions given rise to problems. Again, the cause is moisture ingress and the reason is poorwaterand vapour sealing materials and details. This is a difficult area of the insulation system to design for, for a number of reasons:
.
Large radial thermal movements caused by tank contraction.
.
High shell line loadings, requiring materials with good, mechanical and thermal DroDerties.
Double walled tianks using perlite insulation have on occasions had a history of poor performance. Aring of condensation orice at. or close to the top ofthe outer shell, is an indication of excessive perlite settlement. The reasons for this can be:
.
Lack of, or inadequate, or ineffective vibration ofthe perlite during its site expansion and insiallation.
.
The provision of insufficient hoppervolume atthe top corner
of the tank.
.
On at least one occasion, the location of a large diesel-powered generatoradjacentto the tank following perlite installa-
tion. The use of experienced perlite installation companies using ap-
propriate methods and equipment will help to avoid this problem and its solution, expensive in service topping up of the perlite.
19.12 References
19.1
Foamglas@ lndustrial lnsulation Handbook, Pittsburgh
Corning NV Waterloo, Belgium.
19.2
Research into the structural integrity of LNG tanks, D. Neville and G. White, British Gas Engineering Research Station. LNG 9. October 1989.
19.3
The lnternational Heating and Ventilating Guide, Chartered Institution of Building Services.
19.4 The Handbook of Cryogenic Engineering,
fhe J.G.
Weisland ll, Taylor & Francis, London, 1998.
19.5
Damage to base ofLNG tanks from blast loadings - A case study, R. Tinawi, A. Filiatrank, C Dor6, Journal of Performance of Constructed Facilities. Vol 7. No 3. Auqust 1993.
STORAGE TANKS & EOUIPMENT 409
