Tube Heaters: Still
Tube Heaters: Still
Tube Heaters: Still
TUBESTILL HEATERS
-
When pipe or tubestills were first built , the important part that radia ¬
tion plays in any open-fired heater was not appreciated , and the first
stills were purely convection types, consisting of a bank of closely spaced
tubes with a combustion space below or to the side. Coke deposited in
the tubes that were exposed to radiation and the remaining tubes absorbed
.
little heat When overheated tubes were removed from the still, other
tubes began to overheat, and as the number of radiant tubes was reduced ,
the situation was aggravated . Finally , designers realized that the
radiation from the flame cloud must be distributed over a large surface,
and the modern radiant type of still has been the result. If exceedingly
mild rates of heating are required , the still may consist almost entirely of
tubes that are exposed to radiation.
Shell or batch stills are no longer used except when a single still may be
used to rerun or distill several small stocks such as “ slops, ” special
naphthas, petrolatum solutions, etc.
Types of Stills. Most modern stills are built with two distinct heating
sections: a radiant section which can receive heat directly from the flame ,
and a convection section which recovers heat from the hot gases traveling
.
to the stack Figure 18-1 indicates diagrammatically the arrangement
of tubes and the direction of liquid flow. Names are not standardized
but the several stills of Fig. 18-1 might be called :
( а) Large box-type (g ) Large isoflow (Petrochem )
(б ) Separate-convection ( Lummus) ( ft ) Small isoflow ( Petrochem)
(c) Down-convection (i) Equiflux (UOP)
(d) Straight-up ( Born ) ( j) Double-upfired ( UOP)
(e) A-frame ( Kellogg) (k ) Radiant wall (Selas)
( /) Circular (DeFlorez)
Although all of the stills of Fig. 18-1 are useful foF many services, there
are certain features in which each excels or may be deficient.
1. Flame Impingement. Too large a flame can cause impingement or
overheating of certain tubes in nearly all of the stills, but tubes that are
vulnerable are those just above the bridgewall in (c) down-convection ,
585
586 PETROLEUM REFINERY ENGINEERING
.
and the lowest tubes of ( f ) double-upfired or (d) straight-up The diffi¬
culty disappears in stills ( d) and ( f ) when the capacity is large, i.e., the
stills can be wider. Note the black tubes in Fig. 18-1.
2. Hot Tubes. The rate of heat absorption tends to be high at the
entrance to the convection section because heat is delivered by both
radiation and convection [ black tubes in types (a) , (c) , (d) , (e) , and ( j)].
The feed , because it is cooler, is sometimes introduced through these
tubes (a screen bank) , and in some instances steam is superheated or water
is heated in the screen bank. Likewise, the tubes situated closest to the
flame receive the largest amount of heat [black tubes in (d) , ( f ) , ( g ) y (h),
and ( j ) ] .
3. OH Firing . Oil fuel tends to produce larger flames especially
because larger burners are used. Stills (a) , (5) , (c) , and (e) are best
suited for oil firing although the larger siseB of the other stills encounter
.
little difficulty
.
4. Heat Distribution One phase of poor heat distribution has been
mentioned in item (2) above; but in addition the rate of heating varies
widely in different parts of the boxes or cavities of such stills as (a) and
(c) , although most of the other vertical stills can attain a fairly uniform
distribution of heat.
-
5. Two-coil Heating. Still (c) is not suited for two coil flow although
by experiment it is possible to use it for two streams. Symmetrical
two-coil arrangements can be devised for most of the other stills, but
if a different rate of heat input is needed in each coil, the stills with a
center bridgewall are most effective.
6. Control of Rates. Stills (t) and ( k ) are eminently suitable for precise
control of the rate of heat input and for operation at exceedingly high
(1000 to 1500°F) temperatures. Very low rates of heat absorption can
be attained in the circular types (/ ) , ( g), and (h) , and in other types if
enough radiant surface is provided.
.
7 Capacity. Stills (c), (d), (h), (a ), and (k ) can be built for the
lowest capacities, and types (a) , (6) , (e), and ( j ) are best suited for large
capacities.
.
8. Stacks Stills (a), (5), (c) , and ( i ) require tall stacks, but scarcely
any stack (except to release the gaB at a high level ) is needed with the
other stills because the hot furnace itself produces a draft.
9. Cost . The roof constructions of types (a) , ( b), and (c) are expensive.
Walls not covered with tubes [see (a), ( b ) , (c), and (e)] tend to overheat
and must be of heavier construction. Suspended panel walls are usually
cheaper than brick walls, but suspended tile walls are very expensive.
The large framework of still (t ) is expensive, and the numerous burners
of type (k ) add to its cost. The A-frame still (e) conserves on structural
steel.
TUBESTHL HEATERS 687
In - t - In
j/orA
To slack
CtdOOO
> ofloao
[
_ OddoooaM r
6 0 0 9 0 00 //7 -
saaoe
7000 60 /
,
/
OoOoaooO
oaooeoo hi »
06006 ( U f l f l B f t 6 M J»
- f
-- ~<l!roVBVoUojdp 0 & 5Voo 5Ttf 3rIJ
_ VTJTtf
- ' WWJytfWAftf .
VW&WftftftP Vf . O O C* O o o a 6 0 -
’ "
'
onV M;t£*
/ S
0
\ |
1' /
/' Burners
\
1 Out
A
•
>
1 t~ t /:
^
• /mers Burners
Burners _r
fa) W
/ wooo
I To
stock
7b
stocA
°9£0?0&0°0 M a«
.
o r (i o f i o o fl a c a o o U 6
*I * IO T (
6000
0 06 0 -
OOOQO
0601
09 00
00060
ODD
O 0060
*
s _ 6« •» •
6
661
6 *0
// a
°9
; 8|
|| WA o
. o / { o 1#°“ ” 0
° « T!
a
* *
Burners £ -r-
To
stock
:
•i
\ Burners
OTV
!
.-k
'-
!J
]
§7 Yl
Burners
7o slock To stock I
Air
I Air preheat In To stack
oa oo
00 00c
r r tn
ooooo
aa oc
tI t f
i
i I i
U L u u (J L
Burners Burners tE2 ltjO *"
( f)
tEzatjO at^ ts M/ /
fa ) Burners
To stock
Burners
//3 - >6 o O i
06000
I ~tn To stack
0060
/ °S
/:If \I 6 0600
0600
06Q D O
°
I
/A
006 0 DO 6 0C
6600
60 0
O O> 660
OOOo
» 0660
nra
6 660
'fcy! 6
O
6
0 0 0 6«
\ ••••
0 O
o
a
f
11 l![i \
0 4 9
®
T O
OOfl
g * J
OOOoc
oooo
oa
°ooooo
£ ft I I
0uivm lj: a//
6
Burnt turners
Uu
,
O O O O jfl — ^
0)
I 7b
stock
Burners m
FIG. 18-1. Basic types of pipestill heaters (see text for meaning of letters) .
588 PETROLEUM REFINERY ENGINEERING
In the isoflow stills ( g ) and ( h ) , finned tubes are used in the convection
section. This greatly reduces the amount of tubing required and result*
in rates of heat transfer in the convection section as large as or sometimes
larger than in the radiant section. In still ( fc) the upper ends of the tubes
are finned and thus there is no distinct convection section. The Selas
still (k ) employs ceramic burner cups,1 spaced at about 30 in., which cover
both walls.
Gases are cooled by tubes on the walls, and this tends to cause a down ¬
ward gas flow behind the tubes. Velocities as high as 10 ft per sec have
been recorded, u and it is becoming evident that much heat is transferred
in the radiant section by convection as well as by radiation . The opti¬
mum space between the wall and the tube is about one tube diameter. IO
IN Petrochem heaters ( Fig. 18-1g and h ) about 13 per cent of the heat
absorbed by the radiant tubes is delivered by convection and by internal
recirculation.
The actual heat intensity curves for several heaters are presented
by H . C. Schutt,1 when applied to cracking for the production of ethylene.
Air preheat is indicated in Fig. 18-1/ (BeFiorez) , and it can be employed
in any still that is mechanically tight from gas leakage. The recirculation
of flue gas through the combustion chamber by means of a fan is not indi¬
cated in any of the stills. The flue gas is reheated by the flame, causing a
lower flame temperature and thus a milder radiation rate. At the same
time the convection-section duty is increased , and a larger convection
section must be provided. Thus, fluegas recirculation is particularly
suited to the older convection-type stills. When it is necessary to alter
or regulate the rate of radiant absorption , no method other than flue gas -
recirculation is available except such expensive methods as the use of
large amounts of excess air, a reduction in capacity, or possibly to obtain
-
a fuel which radiates less. Flue gas-recirculation stills were widely used
during the development of thermal cracking processes ; but, now that
engineers know how to design stills for specific radiant-absorption rates,
recirculation stills are no longer installed.
-
The advantages of a symmetrical arrangement of radiant heat-absorb-
ing surface are apparent in the stills of Fig. 18-1. The ideal still will be
one in which the rate of heat absorption is the maximum (at all points)
that can be transferred to the oil without causing coking, discoloration , or
decomposition. If two coils are to be heated and properly controlled in a
1 Tate, C. C., A New Development in Radiant Heating, Oil Gas J ., May 10, 1951,
p. 106.
Reed , R. D., Radiant Heat Transfer versus Wall-totube Spacing, Western Pet.
Ref . Assoc . , Chicago Meeting, June 18, 1953.
1 Schutt and Zdonit, Designing a Tubular Pyrolysis Furnace, Oil Gas J ., May 14,
1966, p. 149.
TttnRjmu, HEATEHfl 58B
I
illicit slitt , some -
partition arrangement as in Fig. l 8 lap fir dt orj must
^
-
be provided . U is desirable to fire TOTH gas because fuel oil Sanies radiate
intensely and thus tend to " burn out ” or lUwse local overheating of
certain Luhcfl. Part of the adv & ntjigc of gas lies iu th* fact that gas
I burners rue relatively cheap and thus several gu bumam can be afforded -
%
i
m &
:W
IC;
~r
S?
mULMLJ,
.
*+
> .
F](f IM A-frarno fumh ^ o imd*r mlrucLtOD . f ^ pen hwlcr boitt ibdW ndillbl
'
^
-
rtitAHTl tends md welded rofiT «lioa htftdtf ( W . W , Kelltitffl ftf , )
reheating ths air that is used in combustion bus an effect that is the
I reverse of fiu ^gntt recirculation . It Lends io mcreaen the absorption rate
in the radiant motion and to mnrcaH the flame temperature, To be mftfll
effective, the air should be heated by only the gases that are porting to the
stack .
-
A. limiting factor in the design of ail direct fircd stills La the tendency of
590 PETROLEUM REFINERY ENGINEERING
oil to coke or decompose at the wall of the heated tube. This limits the
rate of radiant-heat absorption to 6000 to 20,000 Btu per sq ft of outside
area, depending on the operation, because if coke is formed the tube-wall
temperature rises and the tube softens and fails. An equipment that
-
somewhat overcomes this difficulty is the pebble heater.3 4 It consists
essentially of two vertical cylindrical chambers mounted one above the
other. Ceramic-ware pebbles are heated in the upper chamber by direct
contact with burning fuel , and the flue gases pass upward in counter-
current flow to the incoming pebbles. The hot pebbles flow downward
to the lower chamber where they are used to heat air or steam, crack
ethane or possibly oil or hydrocarbon gases. The cooled or spent pebbles,
along with any accumulation of carbon, are lifted by an elevator or airlift
to the heating chamber. An advantage of such a heater is the fact that
coke, etc., is deposited on the moving stream of pebbles rather than
attaching itself to a direct-fired heating surface.
The efficiency of a process can be increased in two main ways. The loss
through the walls and in the stack gas from the furnace can be decreased,
or heat can be saved from the products of the process. The use of air¬
cooled walls, aside from the preservation of brickwork, has often been
referred to as an economy but since hot air raises the flaine and stack
temperature, it is necessary also to add convection surface, or no economy
is effected.
The heat that is absorbed by heat exchangers is often credited as a
direct saving in heat. Such is not always the case because the efficiency
of the still decreases as the temperature of the charge stock is increased by
exchangers. As an example, consider a pipestill distillation unit that
operates with no e^ ^hangers and at a stack gas temperature of 350°F.
The stack loss is only about 8.4 per cent ( Fig. 14-2). If the charge stock
is then heated by exchangers to 300°F, the stack temperature will be about
550°F and the stack loss 14 per cent. In such a case the saving of heat by
raising the temperature of the charge stock from 100 to 300 must more
than compensate for the decrease in the efficiency of the still. A study
of the economics of the complete unit , including both the pipestill and the
exchangers, is the only way to determine the amount of heat that is actu ¬
ally saved.
RADIATION
The evaluation of the percentage of radiant absorption that will occur
in a particular pipestill, and the distribution of this radiation throughout
the furnace cavity , has always been difficult. However, in recent years,
* Norton, C. L. , Jr., Pebble Heater, Chem. & Met . Eng . , July, 1946, p. 116.
Kilpatrick et al., New Pebble Heater Process, Pet . Refiner , April, 1954, p. 171 .
A
TUBESTILL HEATERS 591
owing primarily to the efforts of Professor Wohlenberg6 of Yale and
Professor Hottel ® of the Massachusetts Institute of Technology, a
comprehensive theoretical background has been developed by which plant
performance can be predicted with some success. Still more recently
correlations of the behavior of plant stills have been made available. 71 ®
Stefan’s law of radiation is fundamental.
Q = bAT *
where A = area of radiating surface, sq ft
T = absolute temperature of the surface, ° F
Q = Btu transferred per hr
h = 1.72 X 10-9 Btu / (°F diff .)(sq ft )(hr ) at black-body condi ¬
tions
But a surface cannot radiate all this heat to another surface because the
cooler surface also radiates heat . For a small body completely sur ¬
absorbing Media, Trans . ASMS Fuel Steam Power Division, 53, 265 ( 1931) .
7
Wilson , Lobo, and Hottel , Heat Transmission in Radiant Sections of Tube Stills,
Ind . Eng . Chem. , 24, 486 (1932) .
* Lobo and Evans, Heat Transfer in the Radiant Section of Petroleum Heaters,
Trans . A .I .CtLE 35, 743 ( 1939) .
592 PETROLEUM REFINERY ENGINEERING
polished surfaces or even clean metallic rough surfaces reflect a large part
of the radiation that falls upon them. Furthermore, surfaces such as a
brick wall at incandescent temperatures appear to reflect much of the
radiation that falls upon them.
The differences in the fourth powers of the temperature are involved in
this formulation , and it becomes apparent that the higher temperature
raised to the fourth power may be so large that the lower temperature
may be neglected . In doubling the temperature, the radiation increases
sixteenfold .
Another deduction of a qualitative nature but of great value concerns
the angle of vision or sight of a point radiation source. Radiation is a
wave phenomenon, and it travels through gaseous materials in a straight
line without suffering much of a loss in intensity . It is unable to pene¬
trate opaque or dark solid materials. Thus cooling surfaces cast what
might be called “ radiation shadows * ' behind them.
Heating by radiation is practiced by allowing combustion to take place
in proximity to cooled surfaces. Radiation from flames and gases cannot
-
be easily handled by Eq. (18-1) because (1) the size of the flame cannot be
accurately determined , (2) the flame has a thickness so that radiation
from the center 6 the flame must penetrate the outer layers, and (3) the
^
luminosity of flames varies with different fuels and conditions of combus¬
,
.1
TUBE8TILL HEATERS 593
and the conditions within the flame are so difficult to evaluate that empiri¬
cal relationships have been adopted . The most important factors that
!
affect radiation from flames are (1) percentage of total heat that is
absorbed as radiant heat, (2) ratio of air to fuel , (3) arrangement and
spacing of absorbing surface, and (4) kind of fuel.
Regarding the first of these, consider a flame burning within a furnace
! whose walls consist entirely of absorbing surface. Rays of radiation pro¬
ceed in all directions from the flame ; and with the exception of the small
amount of heat that is used in heating the intervening gas, all the heat
from the flame is absorbed by the cool surface. In this case, a large per¬
centage of the heat is transferred by radiation , although the rate of heat
absorption per square foot of surface is low. If all the absorbing surface,
except a single tube, is removed from such a furnace, the radiation strikes
the refractory walls and is reflected or reradiatfed about the furnace inte¬
rior so that the single tube receives a relatively larger amount of radiation
per unit of surface. Although the single tube absorbs heat at a high rate,
it absorbs a relatively small percentage of the total heat liberation.
The ratio of air to fuel is important mainly because it affects the flame
temperature. The larger the quantity of air (or products of combustion )
that must be heated in the flame the lower will be the resultant flame
temperature and the lower the rate of radiation . Thus large quantities of
excess air reduce the radiant absorption. For the same reason the recircu ¬
lation of flue gas reduces the flame temperature, but air preheat increases
it. The arrangement of the cooling surfaces also affects the rate of radiant
absorption. The closer the surface is to the flame the greater the amount
of radiation that falls upon it. Thus corners and ends of furnaces receive
less radiation than other surfaces ( pages 609 and 611) .
These factors have been related by Wilson, Lobo, and Hqttel:7
R = *
G VQ / aAcp
(18-2)
T
4, 200
where R —fraction of heat liberation (above 60°F) that is absorbed by
the cold surfaces in the combustion chamber
G = air-fuel ratio, lb of air per lb of fuel. For flue-gas recircula¬
tion : lb of air and recirculated gases per lb of fuel
Q = total heat developed in the flame (above 60°F) , Btu per hr.
This may consist of the net heating value of the fuel, heat in
air or in recirculated gases, and sensible heat in the fuel or
atomizing steam
Acv = area of furnace wall that has tubes mounted on it, sq ft
a = a factor by which A is multiplied to obtain the effective cool
surface a = 0.986 for two rows of tubes ( Fig. 18-3) .
,
594 PETROLEUM REFINERY ENGINEERING
0.9 &9y
0.3
&
\°
x
0.7
V
*
\
2^
?>A
* 0.6 NM
t 0.5
—
js
0.4
£ ft '
0.3 &'
0.2 x° /
0.1 ir/
/
2 3 4 5 6
Rot-fio Center-to-Center Distance
Tube Diameter
Fio. 18-3 . Distribution of radiation to tubes suspended in front of a wall. ( HotUl ,
Mech. Eng . )
The area described by Acp or aAcpis not convenient because the designer
or operator is primarily interested in the area of the tubes that transfer
the radiant heat . The wall area { AeP ) in terms of length (L , ft ) , center to-
center spacing (C , in.), and number of tubes per row ( N ) ia
-
A', = L X N X
^
and the projected area of the tubes ( A ) in terms of tube diameter ( D in.)
and number of rows ( n) is
( 18-3)
or
or
A
A =
=L X n X N X
nAav C
D
C
^
^ P °
n D
i
1
^ 4, 200
and Q (§ *;) (1 -
R
RY X
17 ,640,000
G2
(18-4 )
(§ x ) - m =j i ioi4«'
-
Equations (18-2) to (18 6) apply to any conventional box type furnace -
with radiant tubes any number of rows deep, with or without air preheat
-
or flue gas recirculation, and with any normal percentage of excess air.
The factor a may be obtained from Fig. 18 310 but for normal tube arrange¬ -
ments, such as two rows of tubes spaced at a center to center of about
twice the tube diameter, the value of a is 0.986. The relative absorption
6QOOO
\
50,000
V V 07o - Per cent excess
\ V V air
V \ \ 307o
2u 4( V
x 607o
o 1 ^000
\ 90?o\ \
V
*S 30,000
[
Q
e \ \ W
\
ZQOOO
\
\ \ \
.3
S IQOOO
——
: < IL 0 API fuel 11.5
Gosfufil 21.5
tin
.
*1
20 30 40 50 60 70
Percentage of Net Heat Developed That !$ Absorbed in
the Radiant Section
Fio. 18-4. Rate of radiant absorption. ( Btu per square foot of projected area ) for a
center -to-center tube spacing that is double the tube diameter , and for two rows of
radiant tubes [also see Eq . ( 18-6) 1. .
rates in the first three rows of tubes, wh^en only three are present , is
approximately 7:3: 1 so that the third row is of little value.
A study of this equation for the effect of the type of fuel is useful.
Tabl & 18-3 is based on the fuel analyses and data given in Chap. 14.
The radiant-absorption factors and rates of absorption per square foot of
projected area are computed for 30 per cent excess air when half of the
heat liberation is absorbed in the radiant section.
The effect of the percentage of excess air is indicated in Fig. 18^t. The
radiation rates for two common fuels, i .e., dry refinery gas (G for 30 per
10 Hottel, H . C. , Radiant Heat Transmission , AfecA , Eng . 62, 699 (1930) .
J
TUB EBTIUJ HEATERS 597
cent excess air) and heavy fuel, are showD for 0 to 90 per cent excess air.
Other fuels can be represented by lines between these two extremes by
referring to Table 18-3. This chart may be used directly in the design of
simple furnaces, but it should not be used for furnaces using air preheat or
flue-gas recirculation . Note also that it is based on a center-to-center
tube spacing that is twice the tube diameter [ Eq . (18-5) ] and on the use of
two rows of tubes. It may be applied to other situations- by means of
Eq. (18-6) . With proper burners and an experienced fireman , it is pos ¬
sible to fire with less than 25 per cent excess ait, but the average amounts
used are about 35 per cent for gaseous fuels and 50 per cent for oil fuels
( Table 14-6 ) .
If a low rate of absorption is not necessary but gas is available and must
be burned , a mixture of oil and gas may be fired together . The luminosity
or radiating power of such flames appears to be almost as good as for pure
oil flames. Hence for such mi ed flames the curves in Fig. 18-4 for oil
^
flames may be used. The radiating power of these mixed flames may be
explained by considering the carbon particles in an oil flame as radiators.
These particles also exist in the mixed flame.
Lobo and Evans8 have made a complete study of petroleum beaters
and the Wilson, Lobo, and Hottel radiation equation [Eq . (18-2) ]. They
also present a theoretical equation that appears to check well with all
types and arrangements of heaters. With regard to Eq. (18-2) , they con¬
clude that it may be used for designing the older conventional box-type
heaters (Fig. 18 1a , b , c) or other very large stills but subject to the follow¬
-
ing qualifications:
£E
--
d 3 & 'J3 4 fl P>
-
5 1 I
l>
p
1 i S3 C *= M ? t 3| £
3
s
"
& ^
^
- ix i^ o | = a -
.
-
k H 8 Jfl
a * 1s 3i * g iil
m
llllSfll
'
4 si aJiHtf
r
s : ^ = *a
y r- -. 01
2 TJ s - = -- . h 5£ o * O * * j
*
a s :J =
:
S
'
5
>, S
-
o 3
s
1* 1 r
Ij £ t.
BE
>:
y
£
Si
c
a
ftli
s^
1~
»«
V,i~ jr r ~*~
.
-' •
= ;i n i li KKS 3; . —I:
>
*S
-
2
§ *
r is s* S 8
£
wmtr
ts
3e S® JJ 'iB ii
ihl -F i?i
.
itiiri
e Si iN l - 2 3
s -f
J
rl»*il s s
,
« 1
» S 1 f s -Sc
UlfUs " s
= I l
51l s g!
.I H- U J i S
I n "S J mi
« IS i S