[Stoves] Improving Thermal Efficiency (TARP-VE)

[Crispin Pemberton-Pigott]

Dear Frank

 

I will use Deans rather helpful list as a starting point to expand the topic
and perhaps provide additional insight.

 

1.)    Air is very light and by volume does not hold much heat. 

 

Air has a low thermal capacity and once its temperature has dropped to near
what the pot temperature is, there is nothing to gain by holding it in
contact with the pot. If the heat needs more time to transfer into the pot,
the air velocity (gas velocity actually) should be decreased. You can check
with a thermometer to see what the exit temperature is. You will recall I
have often mentioned the chimney temperature for a chimney stove. This is
not the same thing. That has to do with draft and pulling the air
effectively. For a small cooking stove the exit temperature factored by the
excess air quantity determines the thermal efficiency. In other words, if
you have a lot of excess air cooling the gas stream, you will get a lower
exit temperature but that does not mean automatically it is getting the heat
into the pot more efficiently. Just keep that in mind.

 

.So a lot of hot air needs to contact a surface to get it hot. 

 

I believe that taken with the first line, this is a misconception. There is
as I mentioned previously, the heat transfer rate, which means heat getting
into the pot to make the water temperature rise rapidly. Then there is the
heat transfer efficiency which is the effectiveness which any parcel of hot
gas cools against the pot. It get a high transfer efficiency, you need to
give the gas time to cool against the pot. To heat the pot rapidly, you need
to move lots of gas through that space. The combination of volume x
temperature x time to cool against the pot determines two things: the
overall heat transfer efficiency, and the overall quantity of heat that is
transferred. It is important to keep in mind the difference between the rate
of heating and efficiency of transferring that heat from a parcel of gas.

 

.Slowing down the air generally decreases the heat transfer for this reason.


 

Dean is saying that the total amount of heat transferred per second (the
quantum of heat) decreases, not that the % efficiency of the transfer drops.
The efficiency of the transfer of any quantity of heat increases if you give
it more time to make the transfer. Transferring 'more heat' to a pot in 1
second does not mean it was 'more efficient'. That is the difference between
rate and effectiveness of the heat transfer.

 

2.)    The boundary layer of still air is punctured more effectively by high
velocity hot air that heats the molecules near the pot surface and replaces
them as they cool with new hot molecules.

 

There isn't really a boundary layer of 'still air' except for mathematical
convenience. The term 'layer' is used but it applies to things like aircraft
and missiles and things travelling at high speed. For convenience when
making heat transfer calculations, it is imagined that there is indeed a
'stationary layer of air' next to the surface which is only conductive - in
other words it has a heat conduction capacity and no convective movement. 

 

The standard thickness use to make such a calculation is 0.1 mm. If you are
calculating the heat transfer from a turbulent gas flow under a pot, that
theoretical 0.1 mm of still air can be considered to be 'not moving' but
this is for convenience only. It is in fact moving and even at the molecular
level, there is no stationary air. It all moves but at a varying speed.
There are plenty of charts on the internet showing the distance-velocity
relationship. The same principle applies to liquids as well, and the same
formulas are used for both (air is a 'fluid').

 

The transfer of heat from a flying missile is very much unlike the transfer
of heat to the bottom of a pot so I will explain why the 'scrubbing' idea is
misleading, even though Dean likes it. Dean, I hope you and Damon are
reading this and then you talk about it with some math on paper because it
is important to include in your new book.

 

Hot gases travelling under a pot have two effect at play: the forces driving
the gases sideways tending to make a reasonably well-mixed gas flow with the
temperature about the same everywhere, with the stream of gas being cooled
on the top side by the (relatively) cold pot and warmer underneath where it
is no cold, then re-mixing. That is one effect and the 'boundary layer' idea
comes from the similarity between this and a missile or jet plane flying
through the air with (often) clearly defined temperature and velocity
profiles.

 

The second effect is buoyancy, or the effect of hot gases rising if they are
in a cooler gas. This effect dominates cooking but is completely absent in a
flying missile. With aircraft, the velocities are high and there is no real
transfer of heat - the skin of the missile is in equilibrium with the air
around it and no heat transfer is taking place once that equilibrium is
established. There is no water being boiled inside the missile. The effect
of 'hot air rising' so to speak, is that hot molecules rise quite forcefully
in the gas stream impinging on the bottom of the pot and transferring heat
to it.  This takes place mostly in the 1-2mm space immediately below the
pot. The gases in this space are moving vertically, rapidly. This is not a
'boundary layer' it is just the vertical space in which the heat loss is
happening. This cooling zone was shown experimentally by Dale Andreatta, if
you recall his paper on it subject.

 

So the question is, which is the more effective force: the horizontal
movement tending to keep things well mixed, or the hot air rising that
brings hot molecules into contact with the pot? The best effect of hot air
rising would be to have all of the hottest molecules rapidly rise to the top
where the pot is, and the cooler ones to fall down, even though there is a
sideways flow which is probably pretty turbulent. Which is the stronger
effect?

 

I put this question to Prof Chris Snow from the University of London
(Mechanical Engineering Dept). He calculated for me on a single sheet of
paper a comparison between the tendency to maintain laminar flow
horizontally and the buoyancy effect tending to break that laminar flow and
send the hottest molecules to the top. We used gas speeds of 100 to 200 mm
per second as the region of interest.

 

The long and short of the calculation is that the buoyancy effect is 30
times the laminar flow effect. This means that there is nearly no effect at
all of 'insulating with a boundary layer'. Any 'insulating layer' forming is
immediately broken up by the rising hot molecules clawing their way through
the gases trying to get above the cooler ones. The movement is very rapid
and involves a lot of collisional energy transfer. The overall effect
however, is to break up any tendency to form a non-heat transferring layer
of gases, with an overpowering energy ratio of 30:1. Under a pot there is no
such layer.

 

This effect is not present on a missile or on the side of a pot. It only
applies to the underside of pots or other heated surfaces. Because the
effect is so strong, gases under a pot can be used as an insulator to stop
heat losses downwards. By deepening the space under the pot the amount of
heat transferred (by 'convective heat transfer') downwards can be limited
and the overall efficiency raised. In this case the space below the pot is
enlarged and the designer is counting on the depth to get the bottom cooler
and leave the top just as hot (because heat still rises). This design was
CFD modeled by a company in Cape Town and the design used in the BP/Arivi
paraffin stove - a novel product with a very high heat transfer efficiency.
This is one of the proofs that the buoyancy effect is stronger than the
laminar flow/insulating effect.

 

3.)    Extending the time that hot gases flow next to the pot is good. Make
the skirt length longer. Don't slow the flow.

 

This cover two points: extending the heat contact time increases heat
transfer efficiency, and possible interference with the gas flow.  Dale's
experiments were done using a Rocket Stove type product that had a simulated
fire (using gas). Dale showed and reported that the sensitivity of the
pot-skirt's radial distance was low. In other words, the gap between the pot
and the skirt was not really very important. The main problem (or benefit)
arose when the gap was so small that it began to constrict the flow of gases
and affect the way the fire burned. Rocket stoves, as we all know by now,
tend to have a high excess air ratio (typically between 800 and 2000% if you
follow their design manual - Dean, I hope you are going to fix that). The
ideal for a wood fire is about 100% so there is typically a great deal more
air going through the stove than is needed. If you put a tight skirt around
a pot and seal it to the upper surface, you can use the skirt as an air
controller. This was shown repeatedly by Aprovecho when they reported that
there were significant changes to the efficiency when making small changes
to the pot-skirt distance at the range of about 8mm. 

 

What was happening was the skirt was choking the fire effectively enough to
reduce the excess air without interfering with the combustion efficiency -
might even have improved it. Peter Scott and others tried a lot of different
gaps and found that at certain critical gaps the stove worked most
effectively. This is not the effect of 'scraping stationary air' off the
pot, it is the effect of controlling the largely uncontrolled air flow
through a Rocket stove. You can accomplish the same thing by reducing the
size of the fuel hole, adding preheating & descending counter-flow primary
air (as per the Lion Stove) or choking the combustion chamber by reducing
its diameter near the top, or reducing the pot-stove gap (lowering the pot).
Any of these has the same effect - to limit the air flow to what is needed.
Lowering the excess air flow also increases the gas temperature leading to a
higher heat transfer efficiency and thus a higher heat transfer rate for any
given fuel burn rate.

 

Dale's work clearly showed that the pot-skirt gap could be 10 or 20mm and
have nearly no effect in the heat transfer. If the gap was wider, the flow
rate was slower. If it was narrower, the flow was faster. The overall
difference was hardly detectable, unless the gap constricted the whole gas
path. If someone tells you that the skirt has to be a very particular gap,
you can rest assured that the stove already has too much excess air passing
through it and the skirt is being used to compensate for that design fault.

 

4.)    The most effective heat transfer technique is to decrease the channel
gap until velocity of very hot gases starts to diminish.

 

Correct. I reiterated this in detail above. If the stove already has the
correct air flow then the skirt can be run to that limit. If you have a
stove with high excess air (an O2 level in the exhaust above say, 13%) you
have to take a good look at what to do about getting it reduced. A tight
skirt is not a good choice.

 

5.)    A smaller fire creates less hot gasses that can flow successfully
through a quite narrow channel (5mm-6mm) resulting in generally higher
theoretical thermal efficiency.

 

Dean is confirming the point that a high heat transfer efficiency can be
obtained by having a longer 'residence time' in contact with the pot. The
more time, the better the transfer. When designing an air-to-air heat
exchanger or a water boiler, the residence time is a major consideration.
The physical orientation also plays a role. If one were to compare a series
of horizontal channels like a stack of dinner plates with a vertical array
like a set of dishes in a dishwasher, there is a different ratio of forces
at play. If the surface area is limited (a pot) and the surface is vertical,
a small gap might be most efficient at transferring heat, but not if the
velocity is so high as to create a 'gain' for laminar flow ad a 'loss' for
buoyancy. Speed it up enough and the hot gases will zip through the laminar
flow in the centre of the gap. We induce this effect deliberately on the
SeTAR BLDD Coal stove to use combustion gases insulate the combustion
channel walls without resorting to an insulating liner as Charlie has just
done. We do it by spinning the gases to create a hot central punch-through
combustion zone. In that state, the buoyancy effect is almost nil.

 

6.)    Larger diameter pots have an advantage because narrow channel gaps
add up to bigger constant cross sectional areas. 

 

Here Dean is referring to a gap under the pot - the 'narrow channel' being
the gap between the (horizontal) bottom of the pot and the top of the stove
body. As described above, CFD modeling shows this is not the case. It is not
necessarily more efficient to have a small gap based on the theory that the
velocity is higher and therefore there is a 'scrubbing' action taking place.
(The capacity to disrupt the gases near the pot is already completely
overwhelmed by the buoyancy effect.) 

 

There are in any case two additional problems with that idea. The first is
that the velocity is not constant in a constant area profile because the gas
is cooling and shrinking in volume. If one were to hold that a 'constant
velocity gives maximum heat transfer' one would make a profile that took
into consideration the temperature of the gases as they cooled under the
pot. The gap would be much smaller at the periphery of the pot (about ½ the
'constant area' value) and it would not be a straight tapering cone.

 

But suppose we used a straight tapering cone with a gap based on the
combustion chamber area and the pot diameter. The calculation is done in
degrees Kelvin in order to use the universal gas law which holds that there
is a linear relationship between temperature an volume. 

 

Let's use typical numbers: Input 1073 degrees, output 548 degrees, 10 cm
combustion round chamber, 30 cm diameter pot.

 

Initial cross sectional area  = 100 cm2 at 1073 degrees

Final cross sectional area = 548/1073 x 100 = 51 cm2 (This compensates for
the shrinkage in gas volume) 

Initial pot clearance at the combustion chamber lip = 100 cm2/(10 cm x π) =
3.2 cm vertically

Pot clearance at the lip = 51 cm2/(30cm x π) = 0.54 cm vertically

 

Compare this with an 'equal area' calculation of  100 cm2/(30 cm x π) = 1.1
cm for the vertical gap at the pot edge. Quite a difference.

 

If the theory of scraping gases off holds, then the drop in the height at
the outer edge of the pot should be calculated considering the drop in the
gas temperature. The modeling of the Arivi and in fact the proof with the
physical design shows that none of this applies - not Larry's constant cross
section nor my constant velocity. In fact the best heat transfer occurred
when most of the gap for the Arivi was deeper than the nominal diameter of
the combustion chamber! Why? Because buoyancy overcomes all need to disrupt
laminar flow. The cooler gases acting as an insulator gave more net gain
than loss for 'architectural' reasons. I hasten to add that the gap at the
outer edge of the pot was quite small - about 8mm. 

 

Some of you may have seen the chart of thermal efficiency I circulated a few
weeks ago. The highest heat transfer efficiency (93%) was achieved with
quite a large gap, low velocity and small flame. Why so high? Lots of
contact time under a horizontal pot at a very low speed with insulating cool
gases under the pot, trapped by a metal shield. Interesting. No disrupting
from speedy gases taking place there.

 

Larry's rule of thumb, when beginning a design, is to maintain equal cross
sectional area throughout the entire stove/pot. 

 

There is no scientific basis for this rule of thumb, even though I like
Larry a lot. I like Sam Baldwin too - both are nice guys. The idea is wrong.
There is nothing to support the idea that the cross sectional area through a
stove should be constant. Stoves built to this plan and which have an air
entry at the bottom and an exit at the top almost invariably have far too
much excess air passing through the stove and have an overall cooking
efficiency topping out at about 27%. Peter Scott's rather nice bread baking
stove built in Lesotho on that basis were immediately improved by dropping
the constant area idea and going with a constant velocity. It worked not
because of any pot scraping effect, but because the heat transfer is mostly
on vertical surfaces and not so dominantly from the bottom. I believe part
of the improvement resulted from accidentally decreasing the excess air at
the same time by choking.

 

The added control for secondary air purposes of having the restriction on
the air flow being at the entry point cannot be over-emphasized. We were
discussing here a few days ago the need for secondary air injection with the
correct amount of draft (I think it was here). This is best accomplished by
having plentiful draft available at the injection point (suction) rather
than having a choke point at the exit with the secondary air port under
slightly positive pressure. In the latter case one is very dependent on
getting the hole sizes exactly right, and even then, it only holds for a
certain fire power level. Choke on inlet, pull secondary air in. COI-PSA.

 

Sam Baldwin has a chart on page 48 of his book (Biomass Stoves:) showing
firepower/channel gap/length of channel gap/thermal efficiency.

 

That if course only holds for those stoves operating on that fuel under that
power level. There is no general case one can make on the subject, just as
there is no such thing as 'typical emissions' of PM from fuels. (PM is a
result of the stove's ability to deal with the fuel, or not.)

 

7.) Use radiation, too!

 

If heat is being lost to the stove body and not recovered (recycled) or
otherwise desired, it is lost. Calculations shown here on this list
regarding the radiant wire gauze that Paul Olivier tried indicate quite
clearly that the maximum one gets from radiant heat (in that example) is low
- about 14%. If you are only getting 5% then you have something to gain. If
you are losing 10% because you have high excess air, a little radiation is
not going to save your design. If you are losing 50% from high excess air,
getting 100% of the available radiation is not gonna save you. I have a heat
exchanger at the house which is 92-93% efficient and it has nearly no
radiant heat involved at all. Grab it if you can; don't sweat the small
stuff.

 

Best regards to all

Crispin

 

PS Buy a combustion analyser, a thermometer and a scale

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20120604/b4ed5655/attachment.html>