[Stoves] WBT disagrements

Thu Jan 11 06:04:56 CST 2018

Dear Kirk,

“I had to fight tooth and nail with you and Xavier to get tid-bits of information.“

I don’t feel it is very fair. I tried my best to answer all your questions, at length, and provide you information. As I told you first when you replied to me off-list, I very much prefer to discuss on-list, as we can all respond and clarify the parts that are not clear. It is good that you posted this message to the List.

Can you please quote where you think I was not willingly providing information to you?

This was my answer to your questions off-list:

“Are the sensors, filters, and computer graphs included in the disagreements ?  That is a yes or no question.

No, they are not included in the disagreements by the studies.

They are nevertheless included in the disagreements by Crispin. Note that Crispin criticizes the Aprovecho equipment, the PEMS and LEMS, he says it is not reliable, and he brought evidence on the List.

Your question is normally, answered. Do you think it is answered? If not, why?“

“According to these two sentences, you are including the sensors and filters (which measure the CO, CO2, PM and PM 2.5), in the WBT dispute.“

No, I didn’t include the sensors and filters which measure the CO, CO2, PM and PM 2.5.

I included the CO, CO2, PM and PM 2.5 results. It is not the same.

So, the sensors, filters, and computer graphs are not criticized by the studies that say the WBT is unreliable. They do not say the unreliability comes from that. The unreliability comes from the metrics, statistical approach, calculations. That does not mean the sensors, filters, and computer graphs have been proven reliable. They have simply not been the concern of the studies. 

Kirk: you cannot isolate the instruments from the WBT.

What testing protocol does one use the reliable instruments with? The WBT, or any other testing protocol.

One cannot use the instruments alone, can he/she? One has to follow a procedure.

You could for example ask: « is the water included in the disagreements? Is the pot? ».

No, like the instruments, they are not. But they are integrate to the protocol. They cannot be dissociated.

The CO, CO2, PM and PM 2.5 results are obtained by using the sensors and filters (not disputed) with the metrics and calculations and statistical approach (disputed).

When you mix a healthy ingredient with a poisonous ingredient, you get a poisonous recipe.

(sensors and filters) X (unreliable metrics and calculations and statistical approach) = unreliable results.

It is the same for a car. A car dashboard may work well, as well as the steering wheel and the clutch. But if there is a major engine failure, the whole car doesn’t start. It will not matter that some part of the car work well: the whole system is not reliable. “

I don’t see what else I could have done to explain. I also provided you again all the links to the studies about the WBT issues. You really have to read them.

Kirk, you said:

“I also believe in different standards for different purposes.“

An utterly unreliable standard serves no purpose. A broken tool serves no purpose at all. It can bring no good. As Philip says, a clock that doesn’t tell the time doesn’t help anyone.

“Perfection is not always needed.”

We are not talking about being perfect. The WBT is not “not perfect”: it is absolutely broken. No testing protocol is perfect, but the WBT is the only one being that unreliable.

“The WBT does very well for what I need.  It tells me if a change in the stove is helpful or not.  Whether or not it is perfect science is not important for my interests.“

It actually isn’t helping you, even if it is not easy to see.

It is actually really important for your interests, if you want to make stoves that perform well.

“I use it because it is available for me in a lab, whereas the other protocols are not.”

They are available, right now. They have always been. They are available here:

https://drive.google.com/drive/folders/0B5rmmRmIsdlnQlRQX3A1cXVOQ3M?usp=sharing 

You asked me what were the labs performing the alternative tests, and at what costs, as I said to you, I am collecting the information and will get back to you.

Why do you think they are not available to you where you are?

Like a WBT, it doesn’t need much equipment to perform a CSI test. You could perform it at home probably. 

If I don’t forget anything, I think the equipment you need is:

o   A thermocouple connected to a measuring device for determining the water temperature in the cooking vessel

o   A ruler or tape measure for measuring the dimensions of the cooking vessel

o   A scale of 30 kg capacity with 1 gram resolution for measuring the masses of the cooking vessels and water loads

o   A stopwatch

What do you think is currently preventing you from doing a CSI test?

Best,

Xavier

De : Stoves [mailto:stoves-bounces at lists.bioenergylists.org] De la part de Kirk H.
Envoyé : mardi 9 janvier 2018 20:12
À : Discussion of biomass cooking stoves
Objet : Re: [Stoves] WBT disagrements

Crispin,

I think your question was appropriate and the answers were hopefully helpful.

No, you don’t deserve credit for being helpful.  I asked my question and instead of answering it you projected your agenda into me and proceeded to answer your own projections, not my question.  Without an answer from you I had to make a guess.  You didn’t like my guess and became angry and in that anger actually gave me a partial answer.  You proceeded to push your agenda using me as a pawn, but not fully answering my question.  I had to fight tooth and nail with you and Xavier to get tid-bits of information.  Finally I had enough tid-bits of information to assemble a coherent answer.  The credit is mine for fighting for an answer.

Kirk H.

Sent from Mail <https://go.microsoft.com/fwlink/?LinkId=550986>  for Windows 10

From: Crispin Pemberton-Pigott <mailto:crispinpigott at outlook.com> 
Sent: Tuesday, January 9, 2018 2:13 AM
To: Kirk H. <mailto:stoves at lists.bioenergylists.org> ; Discussion of biomass cooking <mailto:stoves at lists.bioenergylists.org>  stoves
Subject: Re: [Stoves] WBT disagrements

Dear Kirk

I think your question was appropriate and the answers were hopefully helpful. There are many lab based systems of ‎measurement and sometimes only a few measurements are necessary to make great progress in improving an existing product. 

Something that can be separated ‎is this: Aprovecho is not 'the WBT'. What they do and how they do it is not dependent on using the WBT which is merely a fixed test sequence with a set of measurements and calculations. 

Philip's comment about the WBT telling you whether or not the stove is improved is most pertinent. If the test didn't tell you how it performed the first time, getting a different wrong answer later may, or may not, be helpful. This is the crux of the problem. 

Tuning a stove requires making refined measurements, not generalizations. A series of small, say, 3% improvements can result in a 12-15% improvement in fuel consumption or power change time, or turn down ratio. If the test cannot reliably tell you the change ‎for better or worse is 1% vs 4%, the results are guiding you by chance. 

A very carefully done experiment must, not should, deliver reproducible results‎. Prof Lloyd has I think been the most specific on this point, much more than most commenters. As an experimentalist not familiar with small stoves, he was expecting that stove performance tests would deliver the same sort of replicability as other physics experiments. The WBT doesn't do that because of the conceptual errors embedded in what it measures, when, and how it calculates the outputs. 

On the face of it, why should the calculations make a result variable? Dean once posted here that if there is a mistake that is applied to every test, then it is not important. 

The difficulty is that the mistakes manifest errors unequally in different stove types, additional to the variability created by conceptual errors. Prof Lloyd was forced to abandon the WBT in favour of a method without those errors in order to continue his work.  

Thank you for participating offlist in the discussions. ‎Xavier are I are discussing the preparation of a spreadsheet with a number of tabs, one each for different versions of the WBT and CCT. Putting in the lab info from a test will create copies of the test on each tab with the different calculations so the outputs can be compared. I think we will be able to find 12 versions of the calculations, maybe 15.

We can also prepare one tab without the errors. This will be the formulas to be applied that are developed from first principles as per the HTP/CSI where the requested metric is calculated from only the necessary measurements made to achieve it. 

Given the limitation of the test sequence itself,, high, high, low power, it would give 'an answer' that could be used to compare the performance of two slightly different versions of a stove, or different fuels, pots, ambient temperature and so on. 

In South Africa, any kerosene stove tested in Johannesburg or similar altitude must also have an emissions test conducted at a coastal location. Similarly in reverse. The performance on such a pair of tests must be highly reproducible. 

Such a comparison sheet can be very helpful for convincing the casual user that only the corrected version should be relied upon. Failing that, we must retire, as there are always going to be those who will not make the effort to understand the technicalities of the field in which they choose to dabble. 

After that comes the issue of contextuality. If the WBT with 'standard wood'‎ is used to develop a stove that will be used to simmer soup burning dung, there is no hope. It would be like perfecting a gasoline engine for 89 octane fuel then operating it with diesel oil. 

All stoves are used in some context so the assessment will be most relevant, even for internal metrics for sub-systems, if the test conditions are relevant to the expected pattern of use. The WBT is advertised as a 'cooking simulation'. That's OK. Pick and report. 

Regards 

Crispin 

All,

I recently asked a question about whether the disagreement with the WBT included the sensors, filters and computer graphing as well as the water boiling portion of the overall test.  I have received several responses from Crispin and Xavier (some off list).  From all that was said by them I have assembled an answer:  The sensors, filters and computer read-outs are part of the WBT, but not part of the disagreement with the WBT.  So this disagreement is with a portion of the WBT test, not all of it.  Also, I believe that Crispin has a second disagreement that questions whether some sensors and setups are able to provide accurate read-outs.  I believe that this is a legitimate concern for scientific study, however I also believe in different standards for different purposes.  Perfection is not always needed.

This is an acceptable answer for me.  It is not pro or con to the WBT or any protocol.  My question was intentionally neutral.  I just wanted to know the extent of the disagreement.  

My position on the WBT remains neutral.  I use it because it is available for me in a lab, whereas the other protocols are not.  The WBT does very well for what I need.  It tells me if a change in the stove is helpful or not.  Whether or not it is perfect science is not important for my interests.  

I believe that Aprovecho (ARC) plays an important part for wood stove development and education and I remain a supporter.  

Kirk H.

Sent from Mail <https://eur02.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgo.microsoft.com%2Ffwlink%2F%3FLinkId%3D550986&data=02%7C01%7C%7Cae298e4088bd4d17bcf608d5570c4bc2%7C84df9e7fe9f640afb435aaaaaaaaaaaa%7C1%7C0%7C636510632671079450&sdata=sRvB55V8H7%2BEg%2Ble5sqvvCUuMYdSNAqb9HGdcp%2FP9yM%3D&reserved=0>  for Windows 10

---
L'absence de virus dans ce courrier électronique a été vérifiée par le logiciel antivirus Avast.
https://www.avast.com/antivirus
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bioenergylists.org/pipermail/stoves_lists.bioenergylists.org/attachments/20180111/5e08c48d/attachment.html>