5 replies to this topic

[jimmason]

one of the main issues in the mass balance work we've been doing over in the wiki is to figure out how much tar we have coming out of pyrolysis in excess of what we can actually combust and not overwhelm our char.  there's much more volatile than fixed carbon in biomass.  thus there is much more tar gas that comes out after pyrolysis than char.  the ratio is about 20% char to 80% tar gas by weight.  

if we burn all that tar gas, we will have more combustion product (co2 and h2o) than char available to reduce it.  we can only combust a fraction of the tar gas, and thus have to crack the rest thermally if we hope to not have tar in the output gas.  this is the core of the tar problem in a biomass gasifier.  a coal gasifier has a much better starting point, as coal is about the exact opposite of biomass, with an average of 80% C to 20% volatiles.


but how much excess tar do we actually have in a biomass gasifier?  how big is the problem in actual numbers?  we're trying to do the math to figure it out here:  http://gekgasifier.p...ifier-Mass-Flow

the mass flow modelling deals with more things than this question, but the tar question specifically is excepted below.  we need help confirming and/or fixing this.  i am not overly confident of my figuring here.


-----------------------------------------


the ratio of c to tar gas after pyrolysis by weight is 1:4.

each c is 12 mass.  each tar molecule is CH2O or 12 + 2 + 16 = 30mass

so if each c is 12 and each tar is 30 by weight, then the molar ratio of C:Tar after pyrolysis is  1/4 = 12/30x or  1:1.6

combusting tar gas is 1:1 tar to oxygen, or  CH2O + O2 = CO2 + H2O.  or 30 + 32 = 18 + 44 = 62.  for every tar molecule we combust with oxygen, we now have two molecules of combustion product.

so our original C:Tar of 1:1.6 is now C:CombustionProduct 1:2(1.6) or 1:3.2


the reduction reactions are C + CO2 = 2CO  and C + H2O = CO + H2.  thus we need one C for every one combustion product molecule to be reduced (whether h2o or co2) 1C + 1CombustionProduct = 2 syngas out


as the reduction reactions need 1 C for every 1 Combustion Product, but our C:CombustionProduct ratio is actually 1:3.2 we have 2.2x more tar gas than we we can directly combust to create the needed combustion products for reduction.  this then, is the amount of extra tar that we need to crack thermally.  2.2x times what we can combust.  ouch!


i am not terribly confident of my figuring here.  can the chemists among us confirm what i did here?

[DanielChisholm]

Jim edited the "Modelling-Gasifier-Mass-Flow" wiki page, with a number of questions and comments.  I'll slowly move them from the wiki to here, and after we've discussed/answered them I'll update the wiki page.

Firstly,

Quote

- canonical woodgas = 28 (much the same as air; SWAGing 20% CO(=28), 10% H2 (=2), 15% CO2(=44), 55% N2(=28))
(daniel, i think this is a bit  pessimistic.  woodgas is better approximated to easy to work with figures as 20% CO, 20%H2, 3%CH4, 7%CO2, and 50%N2.  this is a bit high for hydrogen(really 16-18%), and a bit low for nitrogen.  the co2 should be down in this realm, but is often higher on gasifiers that take off the syngas at high temp before reduction is completed.)

I agree that it is on the pessimistic (conservative?) side.  The only fault I could find with your nice-round-numbers gas is that it might be a bit optimistic, energy-wise.  Methane is a really, really rich gas, volume-wise - one percent of methane adds about 7.5% to the heating value of producer gas.

Later today I'll figure out the energy value of the gas I suggested, and the one you suggest, and also a typical 5-6kJ/Nm3 gas modelled by the DTU "gasifier.exe" program (in the meantime, I've got a lengthy email that I need to write....!)

[jimmason]

DanielChisholm said:

Jim edited the "Modelling-Gasifier-Mass-Flow"
I agree that it is on the pessimistic (conservative?) side.  The only fault I could find with your nice-round-numbers gas is that it might be a bit optimistic, energy-wise.  Methane is a really, really rich gas, volume-wise - one percent of methane adds about 7.5% to the heating value of producer gas.

ok, then reduce the methane a little and up the co2 a little, but keeping the two equaling 10%.  this is likely a little more accurate anyway.  so then we have  the following for a roundish but accurateish def of woodgas composition. . .

20% CO, 20%H2, 2%CH4, 8%CO2, and 50%N2.

[DanielChisholm]

Here is a screenshot of a run I did with DTU's "gasifier.exe" program.  As you can see it's a pretty "vanilla" sort of run - no preheat of anything (air, steam, fuel), and a realistic moisture content.  I chose CH4=2.0vol%.  I also chose "our" biomass formula of C(1)H(1.4)O(0.6), instead of the program's default (not that it really matters very much either way).

As you can see, this predicts a cold gas efficiency of 75% (a good, respectable efficiency), a gas heating value of 5388 kJ/Nm3 (a respectable 145 BTU/ft3).  Predicted dry gas composition is (vol%) CO=21.0%, CO2=11/5%, CH4=2.0%(input), H2=18.7% (and by process of elimination, presumably N2=46.8%).

[jimmason]

bump bump

[HarryN]

If there is interest by anyone, I would like to work on attempting to model this process with two slightly different assumptions:
- Instead of air, using 90% O2 / 10% N2  (because it is becoming relatively inexpensive to do this now)
- Possibly capturing liquids with boiling points in the range of naptha / gasoline / diesel fuel rather than burning them as tar.

The reasons for my interest in this are some potential benefits, including:
- a much more energy dense gas
- capture useful bio liquids rather than burn them
- reduce NOx levels in the gasifier + genset exhaust
- shift the wood gas to a lower H2 level, which would allow higher compression engines without pre-ignition.
- It is easier to clean up smaller flow rates of wood gas and engine exhaust than large dilute volumes.

Thanks

Harry