Document dYkwgzogKXY0X1o7EBD10OXGQ
Why not write specifications,
argues N J Connors, sales en
gineer, Babcock & Wilcox Co,
in words that enable compari
son? Start with terms that re
late heat release to perform
ance; gas temperatures to
VoJ. (0.000
measurement; ash to firing SaVCy',267 Cop Ift/Tir 290,000
96.000
.200 900.000
94.000
.78 790,000
ASSOMEO in all cases 90016-900? FROM rw<> 49IT u 66% FF
toor1
3*000 222 750.000
Make Your Furnace-Design]
aSSUMCO IN all cases 9OOL0-9OOF FROM FW@> 492F & 66% EPF
r.;,
Specifications Work for You
A larcc turnas* of specifications of wall construction, and knows just x 20 ft, 0 In. Hero the volume ii 54,OoJ
that are issued tail to accomplish their objective. So far as furnace designs are
how to vary.this value for furnaces of different sizes, he has no controlling
cu ft and the surface 9,600 tq it. i(v. liberate heat in this furnace at the
concerned, they surely fail if we assume that the objective it to receive bids that
factor. Almost every engineer tealiies that
some rate, that is, 704100 Btu per sq ft per hr, the Btu per cu ft is only
are truly comparable. The furnace-design problem with all
Its interrelated variables is complex and difficult at best. Probably for this rea son engineers who mite specifications are somewhat reluctant to tie the designer's hands too tightly. But they usually do take a hand in the problem, at least to the extent of specifying o value of Btu per cu ft that is not to be
the bounding surface docs not change proportionally with volume. Yet a great number of engineers continue to use Btu per cu ft--a terra that disregards this truth. Why is this?
For one thing, habits of thinking on furnace designs have been formed through a long period that knew no better guide. For a second, a large number of engineers do not appreciate
12,500. Yet in each case the tins amount of best It being presented for absorption to eoeh square foot of fur.
naee cooling surface. Does a yardstick that varies like this have much raise?
Just one more example to drive bent the point. Consider the large tunuce in the last example. Without chsngiog the dimensions, let us put e division wall down through the furnace. Tbb
exceeded. 'On some occasions they specify, as well, the maximum tempera ture of gases leaving the furnace. But they don't qualify by stating how the temperature is to be taken. Behind all this is the specification writer's desire to control something.
The furnace-design problem will probably always be a field of some dif ference of opinion, That Is no reason, though, why We should not strive fa
the danger of too much relionce on Btu per cu Ct as a measure of the furnace-design adequacy. It is sur prising how few engineers really know what happeos to this measuring slick of Btu per cu ft from size variations alone. Let's simplify the problem to the extent of removing the other vari able, namely, degree of waterwall cov
ering, and assume only completely water-cooled surfaces.
wall substantially divides the furnace into halves and receives heat on both sides. Now our surface Has chtoged to 12,000 sq ft, an increase in radiantheat obsorbing surface of more than 50%. If we introduce no more tatsi best then before, no engineer wwM have much trouble deciding which of
these two furnaces is more effective la reducing gas temperature. But out commonly used yardstick, Blu per ca
every way possible to dear up thinking on the subject. At least let's avoid use of terms that tend to further misunderstanding.
Heat Release. What the specification
Working Example. To illustrate, take o foresee of cubical shape, having a dimension of 8 ft, 0 in, on all sides, and assumed to bo watercooled on all six faces. This is not for from what
ft, which fs supposed to tell us some" thing about what Is happening in 1 furnace, fails utterly, to recognize tbn drastic change in surface. The 8l1 per cu ft is exaetty the same after id<b
writer tries to control, of Course, is the temperature of gases at those points where slagging trouble it first likely to make itself felt. When he does Ibis through the medium of a value of Btu per cu ft. not only Is he doing it in a
one might find in marine practice. 'The volume is $13 eu ft and the en velope ares 384 sq ft. If we liberate heat in this furnace at, tay, 70,000 Btu per sq ft of watercooled furnace sur face per hour, the Btu per cu ft fs
ing the division wall as It wsa beforePerhaps this Btu-per<u-ft wctknttt
can be cleared up by closely examining the diagrams above.
In these diagrams are four furnaces assumed to be mtercoolt*
very indirect manner but ha also gets no control. Until he further tics this value Into a particular kind and extent
524100 per hour. Now take d furnace of dimensions found in large centralstation units, 45 ft, 0 In. x 60 ft. 0 in.
on all six laces. Each one is dimenrioceo. realistically for a unit of the espsciff. labeled under each sketch. Also
102 (300|
POWER Msr |H*J
Rtbe ratio of surface to volume and the been living with the comfortable feeling of radiant heat. As you increase a
ReloiBB itself. Vr/Mt $!*. Now let ua assume
Rji !) these widely varying sizes of for-
that our pet value of Btu per cu ft was somehow synonymous with a satis factory leaving temperature. For some
volume, the ratio of hounding surface area to volume diminishes. Btu per cu ft then can be used as a yardstick
||utes ate sited at the same Blu per cu one site of unit this Is true. Diagrams t If this i; the case and with ell el above and on p 102, though, show how
for comparison only if volumes are the same. Just what steps can be taken' to
(die surfaces operated under one set of. for from true it really ia over a range find out how ihis yardstick should
attfidefli si to cos), combustion and [tb costing, we know from actual high-
of sizes. A New Stendavd, A Urge measure
change for different volumesf Tiffs f* the point where opinion and guess work
nJrdty thermocouple tests that the of the misunderstanding now prevalent come into the picture to plaguo and con
Skiving temperature of the gases for in the furnace-design problem would be fuse thinking.
jlbb set of conditions is as shown above feeb sketch. It will be seen that these
removed, if a determined effort were made to (1) refrain from uslog Blu
Btu available per sq ft per hr is a fundamentally sound unit for design
I[itQperttutes vary widely even though Bin per to ft is always the same. J' Now whst we really want is to have
fjeptre) of these temperatures. This it kut is-what we think we are doing k.ipeeifying a certain Btu per cu (L {4's ice, in the diagrams above, *ksi happens to these same furnaces tad the capacity of the boilers they *7ve when we really control this exit
^operatore. ; We want to hold this temperature at J&Q F. To do this we have seen from ftbs diagrams on p 102 wo cannot profcnt ta the furnace surfaces more than IHOOO Btu per sq ft per hr. Notice *ht this does to the Btu per cu ft vatue ai worse still, what it does to the ipict(y we thought we had when we . *tt placing so much reliance on 8to per cu ft in these diagrams, l ',1a conclusion, Btu per cu ft is safe , * te only in comparisons of units of
ipprorimately the same site, ft is very ^wigttouj and misleading to use It over
range of sizes. It is dangerous j^yCsase for many years most of ua have
per cu ft and (2) adopt as e yardstick comparison*. Why not use ft? With
Btu available per square foot of exposed so many natural variables In the pul
watercooled surface per hour. Just verized coal-fired furnace problem, why
how will this come about?
go out of the way to compound diffi
The most important function of a culties by Introducing an artificial one?
tumaco is to burn the fuel end properly
Got Temperature. Occasionally engi
condition the gases before entering the neers specify limits of furnsee design
boiler. Only watercooled surface can desired by using a furnace exit-gas tem
reduce get temperature. The engineer perature as s measure of design limiL
can assure himself that the comparisons Usually when this device is employed,
he makes are true, not only between no description Is given of the method
units of approximately the same else, by which the temperature Is to be taken.
but also between units of widely varying This practice, is not conducive ta clear
sixes, if he specifies that the design thinking or; to obtaining comparable
must conform to a maximum Btu re furnace offerings.
leased ond available per square foot
And here is why. It is known from
of projected tratcrcooled-fumaca en work with high-velocity thermocouples
velope Area per hour.
over the last eight to ten years that (a
Consider for a moment how unseten- the range of temperature from 2000-
tific and wholly inadequate is the uni to 3000 F a bare metal thermocouple
versally used yardstick "Blu per cu ft" exposed to furnace gases will register
In describing a furnace design. A values of the order of 250 to 300 F
measure of the volume tells absolutely below the true gas temperature.
nothing about the kind of surface by
The curve shown In Fig. 3 gives (l)
which it fs enclosed (see above). And High-velocity thermocouple (HVT} tem
remember, only watercooled surface peratures si a base end (2} a compara
appreciably cools get by absorption tive reading of (be bare metal thermo-
Flower . Msy 194B
(301) <01