[Stoves] Simple change to the WBT Re: Thermometers for WBT

Mon Jan 7 14:21:12 CST 2013

Dear Paul and All Water Boilers

>I appreciate and understand Crispin's comments.   First a question, and
then a reply.

>Question:  Does this make any difference?  I think it does.  I have
conducted enough WBT with the waiting for the final degree (local boiling
point) to be reached to know that there there can be even minutes of
waiting.   And it is especially long if the fire is low/small/weak.

There are two things at play in your 'moment of delay'. The first is the
unknown amount of heat invested water that has not yet boiled but which did
not change temperature. The second is what the 'tax' on the heat invested is
in terms of losses from the pot that are not being quantified.

You might recall dimly a few years ago Grant Ballard-Tremeer and I were
discussing something he had noticed which was a 'reproducible effect' in the
improvement of the 'thermal efficiency' of simmering a pot with the lid on
by taking the lid off, all other things being equal. The comment at the time
was that there is heat passing through the pot lid that is not being
measured.

The point is that there is only one loss of heat from a pot that is counted:
missing water. Other radiative, conductive and convective losses are
ignored. When you have a lid off the pot, you add a 4th loss which is
radiation from the water (which has a very high emissivity even though it is
transparent). I have for years recommended using lids on the pots when
assessing stove performance for a number of reasons but the fixed loss is
the one I will concentrate on today because it is relevant.

When the fire is 'small/weak' and the lid is off, then the losses nearly
match the heat input and the energy gain in the water and material of the
pot is slow. Thus the addition of subtraction of 100 watts of firepower
makes an inordinate difference to the 'time to boil'. When the fire is much
larger than the loss, there is no nearly no impact. This means that the
metric 'time to boil' is strongly, (grossly?) affected by firepower when it
is well matched to a cooking pots with lids on. The loss of additional heat
from a pot is about 80 watts or so. If that 80 watts is 50% of the heat
being gained in the pot, then removing the lid cripples the stoves and
greatly extends the time needed to bring the pot to a boil.

Reply:  Why not define the "End-of-test boiling point" to be either 1 or 2
degrees C less than the local boiling point?  

Well, the real question should be, "What are you trying to assess about the
stove's performance?"  The time to boil is not a particularly informative
metric. In fact it can be calculated from the average thermal efficiency and
the firepower but that is not the point. First ask, "What are we trying to
understand about the stove's performance?"

We as designers want to know the effective heat transfer efficiency from
fuel burning to hop pot. That is a valuable number telling us something
about the stove structure. When we want to change the structure, we want a
very accurate determination of the performance on that score because we want
to know if we made a difference or not, then track down why that difference
occurred. If you take a 'time to boil' metric it is so full of variation
that it is not helpful in most cases. If the fire was completely
controllable (as with an LPG or kerosene stove) and there was no leftover
char, it might be useful. But the question it is answering is not really
about boiling, it is about heat transfer. If the heat transfer efficiency is
mis-reported by a large variation in the amount of heat contained in the pot
of water, then it wrecks what would otherwise be a valuable number.

>So when the test pot water temperature reaches that number and remains
steady (or increases) for 10 seconds, the test is over.     

Taking a reasonable large variation in water temperature, say from 25 to 65
degrees, you can get a thermal efficiency with quite a bit of accuracy
(better than 10% and in many cases 5%) and with good precision (better than
6% and in many cases better than 3%). If you made a change to a system that
improved the thermal efficiency by 10% and use a test (one that crosses the
boiling point) you can have errors of 25%.  Why complicate things with
problems like that?

Using a pot that is already boiling is also pretty accurate. If you have a
lid on to minimise unmeasured losses and taking the evaporation rate as the
dependent variable, burning a given mass of fuel will provide an accurate
heat transfer efficiency (fire to pot, not hot gas to pot which is always a
higher number). In other words the fire-to-pot scenario is not the heat
transfer efficiency (which is hard to measure) but a proxy for the heat
transfer. We are interested in both but can only really access one of them
so we use it.

>Latent heat of 2257 Joules per gram is a significant amount when compared
with the energy to raise one degree when well below the boiling point.

And that is my point. Why introduce an error that is several times larger
than the thing we are trying to measure?

Paul, if you run the heating test to near the boiling point you start to
lose water mass and that complicates things. There is simply no need. The
heat transfer efficiency at different pot temperatures has been studied
closely by the Indians and there is nothing to worry about. The heat
transfer efficiency from 30-65 is extremely safe in that virtually no water
evaporates except in ridiculous circumstances (those 'racing stoves' we had
years ago). Over 70 C the water starts to disappear which makes testing more
complicated. 

A very simple and very accurate and very precise test of thermal efficiency
is to heat a closed pot or water on a fire from below 30 to above 60 C,
marking the time interval and the mass of fuel burned. If you replace the
pot with another and carry on for as long as you like, you will end up with
a total energy released figure (char compensated if appropriate) and a total
mass of water heated a known number of degrees. If you ignore the charcoal,
you get the system efficiency in terms of fuel consumed and work done.

Both are valuable and repeatable metrics.

Regards

Crispin

As Crispin points out from Prof. Lloyd's work, even the position of the
temperature probe makes a difference.  

On 1/6/2013 6:45 PM, Crispin Pemberton-Pigott wrote:

Dear Tom

Something further on the subject, Tom, is the work by Prof Philip Lloyd in
Cape Town on water boiling tests (meaning the boiling of water, not WBT's).

He noticed that as the temperature of water goes up at a steady rate, you
can project the time at which the pot should boil, but the targeted time is
always missed - the temperature levels off just as the moment of boiling
arrives. 

I believe that the cause is the investment of energy in the water at 100
degrees without reaching the point of evaporation which is 100 C + 2257
Joules per gram.

It only take 244 Joules to raise water from 0 to 100 C. That means water
which is in the pot and heated to 100 and 'halfway' to becoming steam holds
quite a bit of energy that is not being measured by the thermometer.
Consider the error involved:

Water at 90 C has a total enthalpy of about 1520 Joules/g. Between 90 and
100 + half way to becoming steam is another 1170 Joules. But only 42 of them
show up as 'a temperature increase'.

Thus as the boiling point is reached, a lot of energy disappears into the
mass of the water and does not change the temperature much. This is easily
seen on a temperature:time plot.

For this reason if you wanted to determine, for example, the heat transfer
efficiency of a stove design working with a certain fuel, pot size and
firepower, it should be done without crossing the boiling point.  Either it
should be measured when the water is below perhaps 70 C (as per SeTAR and
Indian methods) or when it has a fully developed rolling boil (as
recommended by Piet Visser). Trying to determine the heat transfer
efficiency or a proxy of it while crossing the boiling point pretty much
guarantees a large error because it is impossible to tell how much energy
has entered the water at the boiling point.

While this observation has been called 'speculative physics' that fact
remains that using a thermometer/thermocouple to get stove metrics is not
quite as simple as it first appears.

Using a clock at the same time one can get Time to Boil, with a scale: Fuel
to Boil, ditto for different pot sizes, but not get an accurate measure of
thermal efficiency, heat transfer rate or true energy content of the pot.
One of the proofs of this error is the calculation of the apparent energy
efficiency during the heating of a pot water and the same calculation to a
boiling pot. The latter efficiency is always a lower number. If you
calculate the difference, it is the heat invested in the water at 100 C
without boiling it away.

When it comes to temperature, measure with care!

Regards

Crispin

_______________________________________________
Stoves mailing list

to Send a Message to the list, use the email address
stoves at lists.bioenergylists.org

to UNSUBSCRIBE or Change your List Settings use the web page
http://lists.bioenergylists.org/mailman/listinfo/stoves_lists.bioenergylists
.org

for more Biomass Cooking Stoves,  News and Information see our web site:
http://www.bioenergylists.org/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20130107/9ae4b944/attachment.html>