13 replies to this topic

[jimmason]

i think we've likely confused most with all the mass and energy balance work of late.  here's one of the main takeaways from it for me.   it showed in very clear terms why it is so hard to get biomass gasification to work (or why it is so easy to have it fail).  the numbers were revelatory for me.


the main problem is this:
there is significantly more tar gas produced in pyrolysis than can be fully burned in combustion and reduced over the amount of available char.  we calculate the excess tar gas to be 2.2x what is needed for stoich combustion to feed reduction and consume all the char.  the rest of the tar gas has to be thermally cracked.  this involves both plumbing the tar to where adequate temps are, as well as maintaining adequate temps and residence times once it is there.  both are non-trivial problems with many dependencies.


thus the problem is much more difficult than just well mixing and burning the tar gasses and air.  we are attempting to set up an internal thermal cracking reactor, with ever changing reactor dimensions (fuel void), that is formed by the highly endothermic "vessel walls" of char, which is fighting us at every step.  this is not an attractive context in which to be attempting a robust thermal cracking device.

if the world was redesigned for gasifiers, biomass would not average 20% fixed carbon and 80% volatiles.  that 80% portion becoming tar gasses is the killer.  this is what we are always fighting until we give up and go to charcoal or coal as a fuel.  

biomass does vary a bit in its composition.  i find it informative to look at the varying volatile to fixed carbon ratios in various fuels in the phyllis database http://www.ecn.nl/phyllis/single.html

but in general, it all stays way above the desired ratios of volatiles to fixed carbon.  way too much tar gas to char produced.

the details of all the mass energy work that informed this summary is in the science section on the wiki here: http://gekgasifier.p...-and-Combustion

jim

[Steve Unruh]

Jeez Jim Mason can we retitle this:
"Why Biomass is such an interesting challenge to harness" ?

If these ratios are true than we now have an additional hydrocarbon-tar fuel heating source to do something with. An opportunity challenge, that our clever brains can peruse. Utilizing this additional heating ability in woodburning stoves has resulted in cleaner output emissions and a reduced fuel in consumption. Yes,yes, I know a gasifier is NOT a heat producing stove but the same opportunity exists.
Maybe instead of internally trying to recycle the tars, collect them hot and in an external forced air burner oxidize them and put this high quality heat to some useful work.
Regards
SteveU.

[jimmason]

sorry, i wasn't trying to say things are impossible.  i'm just trying to better define the problem.

yes, there are thigns we can do with the tar gas other than crack it.  for instance, you can run a flare on top of the reactor where you burn off some amount of the tar gas.  varied dynamically, you could go between a regular tar cracking downdraft, or the equivalent of a charcoal gasifier with insitu charcoal making.  the flare heat can do all the things we need heat for around a gasifier.  the v3.0 gek has a bung where you can install the flare here for such purposes.  or, just to run as an updraft to make char.

but such a constant flare off shouldn't be necessary. isn't necessary.  i think that once we create an internal open space for combustion and tar cracking, with reasonable residence time, we can handle the issues internally.  this is the point of the triple hips type of design.  it gets the combustion and cracking out of the char bed, thus doesn't lose temp against the char, and increases the residence time for the cracking.

short of that, the usual solution will continue to work-- refining nozzle size so that the blast is fully penetrating of the bed.  this of course requires specific configuration for the fuel, and all the usual dependencies towards that.


jim

[Dutch John]

jimmason said:

  the rest of the tar gas has to be thermally cracked.  this involves both plumbing the tar to where adequate temps are, as well as maintaining adequate temps and residence times once it is there.  both are non-trivial problems with many dependencies.

Could this be the reason why the industry gave up on a tar cracking producers and moved to "tarproducers" with excessive tar removal downstream? More steps, but easier to control, despite a complex and perhaps less efficiënt system?

Regards,
DJ

[jimmason]

ok, so i changed the title to "so challenging".  people were getting concerned and not seeing the intended irony.

no, there is no loss of faith here.  i remain fully convinced that small scale biomass gasification is completely solvable problem.  and the solution can be of a nature that will enable a broad, distributed, individual scale power solution at low cost and ease of operation.  we are well on the way to doing it.  the base is now in place.  the iterative design cycles are gong to come progressively faster as we go forward.  we are about to reap the benefits of collaborative efforts and network effects.  i thank you all for being here for the adventure.


my note was only trying to highlight a very important and improved definition of the problem.  tar gas produced in pyrolysis is over 2x what we can directly burn.  i'm surprised it has not been pointed out more clearly before.  many of course have known this before, but i don't remember the specific numbers being much highlighted.  i found the numbers to be highly revelatory.

the truth is, by volume, we are in actuality running more of a tar thermal cracking process than we are a tar combustion process.

this might be why the tar reburn and multipoint air inject open core designs from mukunda and cpc et al have not met their performance promises.  we cannot, in fact, simply reburn the tar.  if we were to fully burn off all the tar, we would completely consume our char and end up with over half flue gasses still mixed with our reduction gasses.  we cannot completely burn all the tar.  there is too much in the system to burn and reduce.  we have to do something else with it.

this i think points strongly in the direction of open internal combustion chambers with good mixing and adequate residence time for cracking.  engage this as a cracking problem as much as a good combustion problem.  these two processes actually call for quite different environments.

so don't worry.  we're going to solve all this.  we have more than enough tools and knowledge at hand to do so.  we've made tremendous progress to date and much more is on the way.  much of it is already sitting on the floor of our shop and soon to be released.  much is likely also sitting on the floors of your shops and minds, soon also to see the light of day.  there are an abundance of solutions.

so let's enjoy the new turns of realization in the process.  the end solution(s) will only be much richer as a result.

jim

[jimmason]

Dutch John said:

Could this be the reason why the industry gave up on a tar cracking producers and moved to "tarproducers" with excessive tar removal downstream? More steps, but easier to control, despite a complex and perhaps less efficiënt system?

Regards,
DJ

yes, this has certainly been the industry solution.  give up and add complex catalytic cracking and condensation cycle systems on the downstream, usually at a cost and complexity that sink the entire project.

i agree this route is very temping, but there is no reason we cannot solve the problem with purely thermal means, and better housekeeping of our gasses and solids.  i find the 3D thermo-mechanical puzzle to be solvable without introducing a new ancillary puzzle(s).

j

[JayAlchemAPL]

There is a lot of energy in the tars, once they are cracked into smaller hydrocarbon chains, there often times could be an instantaneous -c free radical at the end which goes back (depending on location and residence time) into the exothermic chain reaction to then add to thermal capacity of the system. Most tars crack below their boiling points, and the ones that crack after their boiling points, crack soon thereafter. I believe that once the whole system is up to temperature there is enough energy in a cracked tar to crack enough tars, however maybe never to full completion due to equilibrium constants and competition with water.

[JayMart]

I do not mean to be hijacking a thread here. My discussion does follow this line of thought. I want to post it here for discussion before moving to the wiki as a more permanent resource.

The process of gasification is not a single process. It is a series of 3 distinct chemical processes in addition to a significant amount of both mass and heat transfer. I will try to focus on the the three chemical processes (which are highly driven by the mass and thermal transfer processes).

The first chemical process is pyrolysis - the breaking down of the longer chain hyrdocarbons in the biomass (wood) structure into shorter chains that are volatized (made into a gas). These chains must be volatized before they can enter the second chemical process, combustion. The process of pyrolysis is endothermic, i.e. it requires heat to make it work. In most gasifiers this heat (energy) comes from the combustion process. By tieing these two processes so closely together, it can be difficult to control both chemical processes. This is probably where the moisture content of fuel has its largest impact on the performance of the system. Water has a high heat capacity that sucks the energy out of the system, not allowing the fuel entering into pyrolysis process to achieve as high a temperature. Pyrolysis of hemicellulose begins at 440F and contineus through 620F, cellulose begins at 620F and continues through 710F, and the last component of biomasss, lignin begins at 480F and continues to 930F.

The pyrolysis of cellulose is a complicated series of reactions that cleave the molecules into gaseous fragments and condensation reactions that produce char (one of the end products we want for reduction later on). One very important observation of the pyrolysis of cellulose is that higher temperatures favor the generation of more volatiles while lower temperature favor char production. The pyrolysis of cellulose, on average, will produce 66% tar by weight, 11% water, 5% char, and 6% carbon dioxide. The remainder are a bunch of longer chain hydrocarbons including methanol and acetic acid.

The pyrolysis of lignin generates the gases carbon monoxide, carbon dioxide, methane and propane in percentages of 50, 10, 38, and 2 respectively, though the majority of the weight (about 55%) is converted into char.

At this point, the reaction of the gases can be separated from the char, though in the imbert type reactors, the reactions are all nearly colocated, possibly making the control of the chemical processes more difficult. The gases produced, specifically the tars, from the solid phase pryolysis must go through a gaseous phase pyrolysis to cleave the long chains into shorter ones. The activation energy to cleave a C-C bond is about 368 KJ/mole. The next higher bond energy is the C-H bond at about 410 KJ/mole. The energy needed to dissociate H20 to H and OH is 498 KJ/mole. The lowest energy chains are broken first on average since they are the easist to break. Excess water in this reaction zone continues to suck energy out of the system that could be used to cleave the longer hydrocarbon chains into shorter ones. If this zone starts to cool too much, the dissociated chains can begin to recombine.

The combustion process is defined as the combination of oxygen and fuel ultimately producing carbon dioxide and water. All of the previously described pyrolysis reactions do not require any oxygen - they are entirely driven by heat addition to the process. Though it is possible for direct combustion of longer chain hydrocarbons, the reaction favors combusting the shorter chains. Most likely, the energy released from the combustion of the shorter chains is used to pyrolyse the longer chains into shorter ones. An increased quantity of hydrogen in the gas stream tends to increase the combustion rate due to its use in the chain termination reactions.

The reduction of carbon dioxide and water by the hot char bed is the primary chemical process that generates wood gas for later combustion. The reaction of carbon and carbon dioxide to create two carbon monoxide molecules is defined by the classic Boudouard reaction. This is a highly endorthermic reaction with an equilibrium constant of 0.011 at 800K, 1.9 at 1100K and 57.1 at 1200 K. Water reacts with the carbon bed by first dissociating into H and OH and then 2 OH react with a C to create CO and H20. These OH radicals are also a common product  of the pyrolysis processes described previously. It is also possible for oxygen (02) to be directly absorbed by the activated carbon to create carbon monoxide and carbon dioxide.

The bottom line in this discussion is in terms of required chemical processes, combustion is not necessary to generate wood gas. If sufficient energy is dumped into the system to completely pyrolyze the biomass into char, carbon dioxide, water, and short chain hydrocarbons (such as methane and propane), this resulting gas stream can be directly injected into an activated hot char bed for the reduction of the water and carbon dioxide. The resulting gas stream should prove to be an effective fuel for internal cumbustion engines (ICE) or other thermal processes. The trouble with this theory is the pyrolysis and reduction processes require significant amounts of heat (energy). Without an external source, this heat must be generated by combustion. The use of heat reclaimers (such as the pyrocoil) can provide some benefit by reducing the amount of energy that needs to produced by combustion, thus leaving more energy in the wood gas stream on a weight basis of the fuel.

Some personal opinions I have at this point is that the combustion process needs to be divorced from the pyrolysis and reduction processes as much as is possible. It is the process that can be used to control the other two by controlling the amount of air that is injected into the system. The overall process should also be "observable" (i.e. as an outiside viewer we know what is going on inside) by taking temperature measurements of the processes at many different points. I think more research is needed on the rates at which these reactions occur. Do we support enough residence time for the reactions to occur? in addition to can we add the required energy to the points that need them? How much energy can we reclaim from both the hot gas stream exiting from the reduction zone and from the exhaust gas of an ICE to be reintroduced into the gasification process.

I feel that the positive results that APL has had from gasifying walnut shells results from two points: 1) their low moisture content and 2) the high lignin (and thus resulting carbon) content. Walnut shells create fewer gases when pyrolized and have more activated carbon to reduce the gases that are combusted. One option those that gasifier wood chips should consider is adding a percentage of char directly to the fuel to provide the needed additional carbon for the reduction of water and carbon dioxide.

Respectfully submitted to the masses for discussion

Jay Martin

[JDRay]

I know I'm dredging up a long-dead thread, but Jay, this is a fantastic explanation of what's going on in the wood gas production/consumption process, and probably deserves a position in a FAQ or something.  As a newbie on these boards, this is exactly the kind of information I've been looking for.

From what I've read so far, it seems like your suggestion to add a certain amount of char to the fuel has merit.  I presume that, as the fuel (mix) makes its way to the pyrolisis chamber, the char would be on hand to support the immediate cracking of the early-onset tars.  As pyrolisis continues, the volume of char would increase as the volume of cellulose decreases, increasing the tar cracking capacity as the amount of tar to be cracked actually decreased.  In a perfect system, I think you'd only want to add char to the first bit of fuel fed into the system, letting the natural build-up of char as pyrolisis continues suffice for the rest of the requirement.

If I have all this right, maybe just starting the system with a blend of char and fuel in the burner (a handful) would do the trick.

Thoughts?

Thanks again for a great explanation of the chemical processes.

[HarryN]

Hi, I have ready this thread about 10x, learning more each pass through it.

The thing that strikes me, is similar to others, the shortage of fixed C.  One obvious, but perhaps not a desired approach to increase the fixed C ratio, is to add some coal to the incoming feed stream.

Even if we don't stick to this approach long term, it might be useful for experimental data.

[jimmason]

HarryN said:

Hi, I have ready this thread about 10x, learning more each pass through it.

The thing that strikes me, is similar to others, the shortage of fixed C.  One obvious, but perhaps not a desired approach to increase the fixed C ratio, is to add some coal to the incoming feed stream.

yes, you can do this, or run partial charcoal, or choose a biomass with a better fixed carbon to volatile ratio.  nut shells and grasses are some fuels with better ratios for gasification.  but the grasses usually have more minerals, so the clinker problem is much worse.

jay martin's comment above is fabulous i agree.

and generally noticing the issues that folllow from the above noted ratio are critical to understanding the problem of biomass gasification.  i'm not sure why it has not gotten more notice in the literature.  geo homsy figured it out in our shop one night on the white board and it hit me like a load of bricks.  many issues found new light.

still, we can't do much to solve the ratio, but we can change the the manner in which the volatiles go to tars, and control for conditions that produce the easiest tars to crack downstream.  this is lots of what is going on with the externally driven pyrolysis.  it is forcing a longer and lower temp pyrolysis to get less refractory tars, which are easier to crack in the hearth.

next will be separating the combustion from the char bed so we are no longer fighting the char during cracking.

j

[Ecnerwal]

A sketchy sketch of an idea for "separating tar combustion", and also possibly shifting more heat up the reactor. There are probably problems with it, but it's a start.

While it might seem to increase bridging potential, it can also be used as an agitator/stirrer, if the penetration through the top permits motion (rotary, up and down, or both.) The cone portion I'd see as perforated stainless to permit gas but not large solids in. The pipe might be copper if it does not get too hot, as a way of conducting more heat up-reactor, and I've also sketched in a heat pipe as a more aggressive way of doing that (especially if the pipe needs to be steel or stainless to stand up) Internally, might turn the air in an elbow at the top to make it similar to the swirl combustor - might also be bring the pipe down to the bottom and put on an elbow there, giving more preheat to the air as it goes through the flame, since plumbing this into the normal air preheat path adds more fuel blockage potential than a direct air feed from the top does. Would depend in part on how much the flame is pulled "down-reactor" (ie, actual testing required.)

Cone is closed on top, open on bottom.
Don't know if just having the perforated space plus an air feed would do it, or if some sort of ejector/mixer which seems tricky to manage at such low pressures would be required to get a burn. Don't know if the thing would burn nicely or be prone to explosive puffing.  Exact size and position also going to need a good deal of testing. Might be useless. Please pick it apart, or direct me to where it's been tried and failed...

Attached Files

•  Combustor2.jpg   4.58K   39 downloads

[fuel spiller]

Ecnerwal, I have seen this twice on the net as I lurke. One is a very small unit with a rotary ss cone. 3gI Gosslin grateless gasifier.
He used wood pellets to run a 6hp motor.

The second was a design that I think was given out open source
DOBSON GASIFIER Drawing on Google image will get you there (sorry, I'm bad a saving image sources)

[fuel spiller]

A pic of the Dobson

Attached Files

•  a_Dobson_gasifier_in_stars2.jpg   41.69K   23 downloads