[Stoves] TLUD Test with Switchgrass Pellets

Fri Nov 18 21:15:29 CST 2011

Crispin and list: 

1. I think we are getting closer together. But some differences remain. Before getting into them, is your "grasifier" design now the same as shown 20 months ago at: http://www.bioenergylists.org/en/crispin_25-kw-grasifier 
I had not previously read this description. What do you consider the differences from other TLUDs? Yours looks like pretty much like designs we talked about 15 years ago. Stated differently - what have you optimized? 
. 
2. One main difference we still have is the use of a lid. I think we agree that one would want to use a lid for normal cooking tasks - to achieve higher efficiency. But the formula you are using here gives higher efficiency for higher water evaporation. 
I found this report from Aprovecho quite informative on the really high heat losses due to evaporation - and how they can be stifled with a lid: 
www. aprovecho .org/lab/pubs/rl/perf-stud/doc/61/raw 
I stand by my assertion that if you want to measure/calculate high efficiency using the water boiling test technique you are using, the numbers will look better (higher) without a lid. There are stove operations where you want to remove water from a pot - and one would certainly remove the lid to accomplish that. 
Even though we normally would want a lid, I think the test is valuable without a lid in comparing stoves and transfer of heat to the pot. If I want to skew the results in my favor, I would certainly avoid a lid. Same for picking a color of the pot (much smaller effect). 
There are parts of your argument on lids I don't follow. But most interesting is this next apparent after-thought comment - which I don't know what to do with: 
[It occurs to me that an understanding that the lid coming off would increase the thermal efficiency may be rooted in the error in the WBT3 that calculates an ‘efficiency’ for simmering which is itself a conceptual error. Increasing the evaporation during a simmering test appears to increase the system efficiency which is impossible. It is an artefact of the math error. Any WBT3 efficiency number is incorrect for this reason. Only the high power heating efficiency is valid.] 
You also have comments below about influencing the computation using skirts, etc. How does your last sentence improve things. I want to emphasize that my only initial complaint was that your concluding the test gave 22% (now less than 20%) was being terribly unfair to all char-making stoves. This last comment by you is one that needs further exploration by all of us. 

Few more inserted comments below [Labeled RWL2] To get at differences in definitions, I have bold-underlined. every use of the term " efficiency" 
----- Original Message -----
From: "Crispin Pemberton-Pigott" <crispinpigott at gmail.com> 
To: "Stoves" <stoves at lists.bioenergylists.org> 
Sent: Thursday, November 17, 2011 8:39:11 PM 
Subject: Re: [Stoves] TLUD Test with Switchgrass Pellets 

Dear Ron 

I am answering your post first because of its currency but I am not forgetting that I need to respond to Paal and Paul from the weekend and Vetle’s last very information-filled message. I have been extremely busy over the past week preparing for the COP 17 demo and beating Old Man Winter with some brickwork and also today’s demonstration of three burning modes and fuels for the visiting delegation from China (which went well). 

> I like the computations you have provided, but think there are additional efficiencies needing computation. Yours below gives zero benefit to the (very sizeable) charcoal production.- which obviously will entail some howls from those (like me) interested in Biochar. 

I went through the spreadsheet first and it seems you have captured the numbers well. The column D numbers are only very slightly different from the ones I used and they are because I had to round slightly to give the mass data. 

I do immediately have a question as to whether the ‘ efficiency of char production ’ has anything to do with ‘ thermal efficiency . Yes they can bot he viewed as efficiencies but they are not related, of course. 

[RWL2a: I disagree. They are intimately related. Especially in the revised formulas, which you seem to have agreed with..] 

> First, I think this sort of efficiency measurement should be done without a pot lid. 

Why? 

[RWL2b: All this on "lids" already responded to above.] 

> The 22% number which is now measuring something about the tightness of the lid (and not related to the stove) should jump to above 25%, I guess. 

Of this I am absolutely sure: tightness of a lid fit has nothing to do with heating efficiency of boiling temperature or anything other than limiting heat loss. The ‘tight lid’ theory is an old saw that does not bear repeating. In fact I recall a discussion on this list in which I showed that putting a very heavy 1 kg perfectly fitting gas tight lid on a pot compared with a loose lumpy lid was the equivalent of moving the stove 5 floors up a building. It is irrelevant, literally. The pot was a good quality stainless steel one with a good lid fit (loos but sealing well). 

> I will send an Excel spread sheet to Erin and Tom on this – so others can play with it. 

I received it, thanks. 

> That SS shows that about 50% of the initial energy remains in the produced char 

I am not immediately clear what SS means. I agree that the remaining char contains roughly 50% of the initial energy available in the fuel. I also immediately caution everyone to realise and remember and repeat that this is a highly conditioned statement. It is only true because of the moisture content of that particular fuel with that particular analysis. I have to revise the calculation because the ask content is apparently 3% or so (confirmed by Meredith from REAP Canada this morning. We were on a tour of switchgrass production and a pelleting plant of vast capacity.) 

[RWL2c: Sorry; "SS" stood for "Spread sheet". We are agreeing on the 50% statement.] 

>… I say that makes the overall stove efficiency , making no other changes in Crispin's work, about 75% (or 25% "loss" – not 75% loss). 

Well we have to be careful what is claimed here. The boiling efficiency of the pot from beginning to end was indeed 22% of the available heat. It was not 22% of the heat offered to the pot. That is for sure. It is also not correct to make the calculation of the char-subtracted value without pointing out the implications of a slightly different moisture content because in these cases where the char mass is high, a change in moisture gives a disproportionately large increase in the apparent efficiency . In fact, I tossed in the 22% because I did not have a high heat efficiency which I would much have preferred. I will try to get one in the next few days if anyone thinks it is important. It is certainly much higher than 22%. 

[RWL2d: I am confused. Obviously moisture content is important in efficiencies - but we have small moisture here and everything can be calculated. I look forward to differences you find with a different " high heat efficiency" ]] 

> That is, I claim that just adding the boiling and charcoal-making efficiencies together is valid. The missing 25% is energy that was not captured usefully. 

The char-making efficiency is related to the theoretical maximum char possible and has nothing to do with heat transfer efficiency . I am surprise you would want to add them together. 

[RWL2e: We will have to agree to disagree. The amount of char has an enormous influence on what you are measuring and calculating. ] 

The engineering approach would be to take the heat offered to the pot and evaluate how much of it was retained within the pot in the form of hotter water or evaporated water. That is not the heat transfer efficiency , it is the system efficiency . The system efficiency heat plus the heat passing into the pot and lost through radiation, conduction and convection added together would constitute the total heat transferred. No one on this list that I have ever read (which is not the whole history) has offered a heat transfer efficiency for consideration. Maybe we should give some thought to that. 

[RWL2f: Our talk about skirts has this in mind. Besides radiation, conduction, and convection, we have evaporation "losses" (and evaporation might be the desired end).] 

> But, even without adding the charcoal – making efficiency, one can approximately double (as opposed to the above-stated tripling) Crispin's calculated 22% “boil-away” efficiencies by dividing the useful output by the net energy consumed in the boiling task, not by the total amount available . 

Correct. 

Heat retained by the pot (as described above as system efficiency heat ) H 1 divided by the Heat potential, minus heat value of the char remaining H 2 = system efficiency . 

H 1 / H 2 x 100 = Eff % 

[RWL2g: I gave the different efficiencies numerical labels. I think this is Eff3. I don't see calling H1 "heat retained" - as it is mostly a measure of heat in the evaporated water. H2 ("Heat potential") seems to be the initial energy minus the energy in the char.] 

To repeat - I calculate that about half the initial available switchgrass energy is still available in the produced char, and that this MUST be accounted for when comparing a pyrolysis stove with a combusting stove. 

I do not see that it gets special treatment. The stove type is not affecting ordinary engineering calculations. If a stove make 5 g of [RWL: "or"?] 500 g of char, determining the system efficiency requires that the heat value of the remaining char be deducted. 

[RWL2h: But the whole point of this exchange is that you did not initially make this eff4 computation - which approximately doubles the number from 20-25% up to (now) 57% . I think it is very misleading to talk about 20-25% efficiency for this (or any charcoal-making) stove.] 

When it comes to evaluation the performance of a stove in the field, in other words a practical evaluation of fuel consumption, one has to consider whether or not the char can be ‘used’. It might be for fuel or agriculture. If it cannot be re-used in the same stove, it is lost to the stove and fuel and is a loss. I have no problem with this being reflected in a note about the performance. But I think it is a stretch to say that from and energy use point of view, a stove that produces char is ‘twice as efficient because it does not burn the whol fuel’ which is what I understand from your calculation where you add the % char to the system efficiency . 

[RWL2i: Make that "thrice" rather than "twice". We are disagreeing also on the word "system". I am not disagreeing on reporting the 20-25% number. I just am saying one has to report something on the char-making efficiency as well - even if the char is thrown away (which is not very likely). I do not propose to add "%char" to anything. I do say that eff2 = %energy in the char" is meaningful and can be added. If one were doing this only for later use of the char, it would be THE key number. The first work on char-making stoves was to avoid the common horrible loss in making char in remote pits - not for placing char in the ground.] 

> This comes closer to the efficiency on the water boiling side, but still does not fully capture the value of the char. 

The system efficiency is not intended to capture the value of the char because it has nothing to do with the combustion or the heat transfer. It is a completely separate subject with its own rules and understandings. It would by quite legitimate to point out how much of the carbon in the original fuel was in the char, for example, because someone might want to optimise the retention of carbon. 

[RWL2j: I still cannot fathom how you decide/believe that the (weight or energy) "value of the char ... has nothing to do with the combustion or the heat transfer". ] 

Would you agree that it is reasonable to report the amount of heat that was useful to the pot as a % of the potential of the fuel, as well as the % of the heat offered to the pot? In fairness, the combustion efficiency of the fire should be taken into account when making this calculation of system efficiency, and should be demanded if calculating the heat transfer efficiency. The heat transfer efficiency is a very narrow examination of something taking place within a chain of events – one link of it, so to speak. 

[RWL2k: Of course (re the "as well as"). This is why I wrote the note. Giving only an "efficiency" of 20% is going to drive folks away from a very promising stove type.] 

> The lost energy is about half of the non-char energy (not 3/4 of the input energy - as implied with a 22 or 25% efficiency computation). 

I did not state any losses or describe them. You have inferred that I said the char as a loss. It is left over and I said so. It might be fuel, it might be fertiliser. In the light of the comments I presented about the thermal efficiency having to overcome the carbon subtracted from the fuel-stove energy chain, I think it is relevant to present all the information so the user can interpret the test in a way that is applicable to their intended purpose. 

[RWL2l: It won't be just me that will infer something about a (single, unaccompanied) 25% efficiency statement (when others like me see it as closer to 75%). 

> To restate, with simple numbers that might result with a re-test without a pot lid, that hypothetically finds that 25% of the energy was productively used in boiling the water and also found that half the energy remained in the char: 

I do not see how taking off the lid would do anything other than increase the radiation losses (which are not counted in a standard test) and increase the evaporation (which is). The result of taking off a pot [RWL: "lid"?] ALWAYS results in a lower system efficiency number. The reason is that water has a emissivity coefficient of about 0.99, similar to black oil. If you take off the pot [RWL: "lid"?] , a lot of heat is lost by radiation to the sky/ceiling. This loss is not countable by present methods and appears to be an inefficiency, that is it lowers the system efficiency number in all cases. There is no case where you can increase the uncounted loss, and increase the system efficiency number. 

[RWL2m: All of this paragraph is contradicted by your sentence below starting: " [It occurs to me..." ] 

>… My preference: eff3 = eff1 +eff2 = eff1+ 2*eff1 = 3* eff1 = ¾ 

You cannot add efficiencies that if they are producing the same thing like heated water, but not if they do not have the same units. Steam and char are not the same units. 

[RWL2n: I didn't and wouldn't; my units were all compatible - all in energy terms.] 

> (I believe that having the same denominator when calculating efficiencies is essential. 

Exactly. The denominator of one is MJ and the denominator of the other is g. 

[RWL2o: Look again at what I did. Using grams is too trivial - being a ratio of measured quantities.] 

> Of the ¾ useful output 2/3 is in the char and 1/3 is in the boiled water; one quarter of the input energy was not captured.) 

It is important to note that char is not a ‘useful output’ from heat generation point of view. It is unburned carbon. It went in unburned and came out unburned. That is not an efficiency. 

[RWL2p: Disagree. Putting char in the ground will have to directly compete with the energy value. 

Also I think you need to rephrase this above sentence:" It went in unburned and came out unburned ". There is definitely an efficiency involved in the magnitude of this " unburning". 

Combined (EPA-like) eff4 = eff1/(1-eff2) = eff1/(1-1/2) = 2 eff1 = ½ (I think this undervalues the char – and has no theoretical basis when computing efficiencies. It does a better job of bringing in the char, but not the correct calculation. This says that about half of the available energy passing the pot is captured. It does not give enough credit to the char. 

But about half the energy offered to the pot DID enter the pot. How well it made char is irrelevant to the system efficiency . I am tired of repeating that so I will end now. 

[RWL2q: We both are tired of the repetition. However to repeat: your system efficiency is directly impacted by the magnitude of the char production. 

The spreadsheet: 

The char content was expressed (as normal) as a % of the dry fuel mass, not the wet fuel mass because if expressed as a % of the wet fuel mass, one cannot make correct comparisons between different stoves and fuels. Yes, I know people usually report the wet fuel number but it is not very helpful. 

I have corrected all the numbers relating to ash content and derived therefrom. 

It was 97 degree rise – sorry if there was some mistake. It was 3-100 heating. 

[RWL2r: For others: I (slightly erroneously) deduced that Crispin had a 96 degree rise, as it was unstated.] 

I have added the correct heat value for dry switchgrass and calculated the heat available. 

The line about pot lid off has been removed. [It occurs to me that an understanding that the lid coming off would increase the thermal efficiency may be rooted in the error in the WBT3 that calculates an ‘efficiency’ for simmering which is itself a conceptual error. Increasing the evaporation during a simmering test appears to increase the system efficiency which is impossible. It is an artefact of the math error. Any WBT3 efficiency number is incorrect for this reason. Only the high power heating efficiency is valid.] 

[RWL2s: This last parenthetical [...] seems to leave everything in this dialog in doubt. I hope we don't drop the topic at this point. Something has to be meaningful as it relates to total char production.] 

As the char heat content is included ‘above’ in the sheet I have removed all the lower section with estimates and the combining of thermal efficiency and char making realised. 

[RWL2t: I am going to include the original - and maybe more, as well as your new material. I will check your following.] 

The Average system efficiency is 56.66%. It also produced char which could theoretically have been 100% of the carbon and 0.176 as much ash and volatiles (which is 15% of the total mass). That char could have been 617.6 which I calculated as 1050g x 50% Carbon / 0.85 = 617.6 g. 

[RWL2u: I am surprised it got this high. But glad you are no longer talking 22.3%. With the huge swings we are seeing in the various efficiencies, I suggest calling 56.66% as 57%.] 

The char realised was 365g so it is 59% of the char that might have been produced. 

Some may jump in here with a ‘fixed carbon’ angle. Well, there is carbon in the volatile materials and it is no possible to know exactly how much because there is no exact value for what is ‘fixed’ save a different number for each and every temperature. The best we can say scientifically is that is it 59% of what might have been. 

Now, it is clear that adding 56.66% thermal efficiency to 59% char production realised does not make sense. It is not a stove with ‘115.8% efficiency’. I hope that is obvious to everyone. It is an efficient stove that makes char. If that is what you want, maybe use it. As mentioned in previous posts, is it constructed based on the formulas used to make a Vesto. The main difference between this and other TLUDs like the Anderson gasifier and Peko Pe is there is a lot more flame space which provides more draft to the secondary air which provides better mixing and turbulence. It does not have a pot skirt or even a proper top deck the size of the pot, both of which might increase the efficiency. 

[RWL2v: Much of the early reporting was for two cans of equal height. Agreed that the upper space should not be too short.] 

As I said, this is a pretty efficient little stove! It cost less than $1 to make and cooks for about 2 hours per kg. I will try to get a better number for different stages of the burn. The power drops slowly during the burn. There was no un-pyrolysed fuel at the end. 

[RWL2w: I think you did a nice job on the construction. The use of stove pipe is good - and I like the conical shape for a preheater. I am anxious to hear more about how results change with different amounts of primary air (through nails/bolts in the eight holes). 

The edited spreadsheet is attached. 

[RWL2x: and a new version, with some comments was going to be attached to this response. But I stopped went I got to this pair of lines 15 and 16 - which don't look at all right. Would you check with REAP again? Fortunately, I don't believe you used these numbers at all later, so I continued with the old-new comparison. 

	15. Ash content of the initial fuel: 	grams 	283.5 
	16. Ash content of the char: 	grams 	94.5 

RWL cont'd I think there is sufficient uncertainty now about the meaning of this test (since you are now disowning the results) that I am going to not supply a revised spreadsheet. I do have to emphasize that now everything is based on assuming that this switchgrass char has an energy content of 29.47 MJ/kg (presumably from Roger Samson and REAP). Earlier, I was assuming a considerably smaller number based in part on your then-assumed larger ash content. 

Starting with 18.70 MJ/kg for dry switchgrass, you (Crispin) also this time calculated (not assumed) an energy content of the switchgrass pellet at 10% moisture of 16.56 MJ/kg, whereas you had earlier used 16. This caused only a small change. 

To summarize, the old and new eff1's (for boiling) are about 22%. The old and new eff2's (for char production) are 50 and 62%. The old and new eff3's (sums of the previous two) are 72 and 83%. [Crispin does not accept the meaningfulness of, and did not compute, eff2 and eff3]. The old and new eff4's (the ones we seem to agree areappropriate to define the stove- and which take account of the energy in the char) are 42 and 57%. Crispin did not calculate in the last spreadsheet that the percent capture of the energy generated is about 22%/57% 

One lesson for me out of this is that the assumed energy content of the fuel and char can cause serious changes in the various efficiencies. Now it appears that the stove is capturing only about a third of the available energy (after subtracting that in the char), rather than half. 

Apologies to all - as it now appears that Crispin feels none of this is meaningful. I hope we can start with firm definitions - which should go well beyond this list. Maybe, hopefully, we did accomplish something - but I am not sure what. Ron] 

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/20111119/4c5177b5/attachment.html>