Modelling Biomass Combustion via Species Transport
 

Hello!

I am trying to use species transport method to model biomass combustion and I encounter some doubts. I'm burning some wood chips, so this will mean that I should choose 'wood-volatiles-air'for my mixture material, right?

If this is correct, when I go to the 'materials' panel and check the mixture species, I get 5 volumetric species. Actually, what do the 5 species mean? Are they the reactants and products of the burning process? There is this 'available material' which give 'air' as a choice. Should I add air under the selected species or air has already been included in the O2, CO2 and N2 species that have been included in the '5 volumetric species'?

Since I'm burning solid wood chips, I should add some species under 'selected solid species' right? I'm not sure what solid species to add here because I can't find solid wood in the materials panel under 'fluid'. According to Fluent 6.2 user guide chapter 14, page 20, I should be able to find some solid species under the 'fluid' list. Am I doing something wrong here? Please correct me!

Sorry for asking so many things...I really hope someone can answer me! Thank you very much!

Racheal

Re: Modelling Biomass Combustion via Species Trans
 

The wood-volatiles-air would be a good first choice for a gas mixture. You should have some wood volatiles species, CxHyOz, to account for C, H, and O in the wood. The CxHyOz will react with a stoichiometric amount of O2 to produce CO2 and H2O. Nitrogen, the fifth species, won't react.

You don't need to add air. For your air steam, just specify the flow rate and then the O2 and N2 mass fractions for air.

The volatiles won't account for all of the wood mass - the rest should be char carbon. This can be released as C(s) i.e. solid carbon. C(s) can be choosen from the materials list for gases. Add another reaction for the oxidation of C(s) + O2 -> CO2.

And, you may have moisture in the wood. This can be accounted for by adding another species H2O(l) - liquid water. Release the correct amount of H2O(l) from the wood and then add another reaction H2O(l) -> H2O(g) to account for the latent heat of vaporization.

Choosing the ratios of C, H, and O in the volatiles and the latent heat of formation for the wood volatiles and representative reaction rates will allow for closure of the mass and energy balances and representative flows and temperatures.

Re: Modelling Biomass Combustion via Species Trans
 

Thank you Allan.

From your message, is it right to say that I will have to define 3 rections under the 'Reactions' panel, i.e.

1. fuel + O2 --> CO2 + H2O 2. C(s) + O2 -> CO2 3. H2O(l) -> H2O(g) Does the sequence of the reactions matter?

If I'm using Discrete Phase model as well to inject the wood chips, do I still need to enable the 'wet combustion' tab and set the evaporating species as H2O<l> since I have already define H2O(l) -> H2O(g) in 'Reactions' panel?

Thanks a lot for your answer.

*Racheal*

Re: Modelling Biomass Combustion via Species Trans
 

Yes, 3 reactions would characterize the biomass fuel.

Your "fuel" in the first equation would, of course, be the wood volatiles with some fractions of C, H, and O.

You probably realize that there are many ways to set up Fluent for biomass combustion and this is only one way. You have to be careful with all of the other parameters.

For example, in this scheme you would still enable wet combustion and have liquid water H2O(l) as the product. You then set the latent heat of vaporization in the drying stage to zero since the heat is accounted for in the psuedo reaction H2O(l) -> H2O(g).

Similarily, if you have solid carbon C(s) as the product of combustion, then the oxygen demand for DPM char combustion will be zero and the heat of combustion will also be zero as these are accounted for in a subsequent reaction.

Your DPM laws have to match what you are doing in the gas phase.

The concepts of modeling biomass combustion can be relatively simple, but keep in mind that just the mass and energy balance accounting can be a big job. I would allocate several weeks or months to check all of these things.

Good luck, Allan

Re: Modelling Biomass Combustion via Species Trans
 

Thanks very much Allan.

I have run my wood chips combustion simulation but the temperature results are much lower than expected. What could possibly be the main cause? Could it be something to do with the 'pre-exponential factor', 'activation energy' or 'temperature exponent'that I set in the 3 reactions? Is there a guideline in what range should the 'pre-exponential factor', 'activation energy' and 'temperature exponent' be for the 3 reactions?

Hoping to get some help, thanks!

*Racheal*

Re: Modelling Biomass Combustion via Species Trans
 

Do you have significant unburned fuel, wood volatiles or carbon, at the outlet of the combustion chamber?

If so, the reactions may be proceeding too slowly. Generally, to make kinetics less limiting, you could set the pre-exponential factor to about 1x10^12 and the activation energy at about 1000. The temperature exponent is usually 0.

What about the enthalpy of formation for the wood volatiles? Is the difference between that enthalpy and the enthalpies for the CO2 and H2O formed, plus that for char combustion, give you the heat of combustion for wood?

What about when you look at the sum of the enthalpy source from the DPM? Is that about right? Does it match up with your enthalpy balance? Mass flows correct? What about the heat capacity of your flue gases?

Re: Modelling Biomass Combustion via Species Trans
 

Dear Allan Walsh,

Here I would like to ask you some questions for biomass combustion, with a hope to activate the discusion and not to affect your previous discussion with Racheal.

My questions are in the following:

(1) Suppose that the devolatiled components consist of multicomponent gaseous products(including CH4,CO,H2,CO2,tar...), then how to implement this kind of devolatilization model which takes into account the kinetic forming rates for these different products?

(2) You mentioned 'wood volatiles species, CxHyOz, to account for C, H, and O in the wood.'. ---Do you have some suggestions on how to determine thevalues of x,y,z by using the data from proximate and elemental analysis? And in this case, how to determine further the average molecule weight for the volatile (CxHyOz)?

(3) In addition, you said '..accounted for by adding another species H2O(l) - liquid water.'. --- Let us say we have two options for the moiture evaporation. Do you prefer to realize it by using multiple surface reactions model or by activating directly the wet combustion model?

Hope to get your help.

Kino

Re: Modelling Biomass Combustion via Species Trans
 

Hi Allan and Kino,

For question (2)from Kino, I think you can refer to FLUENT 6.2 user's guide, chapter 14, page3 28.

Referring to the same page, about the heating value of the fuel, how can I utilize the information into FLUENT? Does it refer to high heating value or low heatng value? Is there a place that I can define in FLUENT this value? If there is, where?

Hope to get some help too, thanks!

*Racheal*

Re: Modelling Biomass Combustion via Species Trans
 

(1) I have used kind of a "lumped" method for wood volatiles. That is, all of the volatile gases are evolved at the same rate. Then the generic "wood volatiles" reacts to CH4, CO2, H2O, H2, CO, NH3, H2S, etc. In version 4 of Fluent, you could have the gases evolved separately, but with the variables available in the UDF for version 6, it seems to be compromised.

(2) I think Racheal answered this - for the simple case. But for a more complicated where your elemental analysis gives you C, H, O, N, S, and possibly Cl and other ash species, you can use your proximate analyis to determine how much fixed carbon you have. Then, make up a matrix of your other "volatile" gases. Some of these may be fixed if an element only goes to one species. Usually, I end up with two unknowns (or one unknown, the ratio of H2O to CO2) Then I also use the laboratory measured heating value to determine the ratio of the species that gives me the best match.

(3) I have been happy with the wet combustion model to account for moisture. Be careful though, I believe in the generic Fluent model, the moisture fraction is a volumetric fraction rather than a mass fraction!

Re: Modelling Biomass Combustion via Species Trans
 

Racheal,

I'm assuming you are using the "wood-volatiles" from Fluent database for fluids. You can select the define->materials tab to set the properties for the wood-volatiles. Change the standard state enthalpy such that when the wood-volatiles are oxidized, the difference (on a mole basis) with the standard state enthalpies will give you the heat of combustion. You may have to play with this value a bit.

Laboratory readings will normally give you the higher heating value since you will have liquid water as the product. In your reaction, I'm assuming that you have gaseous water as the product, as the temperatures are relatively high.

Re: Modelling Biomass Combustion via Species Trans
 

Hi Allan,

I didn't use the wood-volatiles from FLUENT database directly, instead I changed the molecular weight of the wood-volatiles.

The standard state enthalphy to be defined for wood-volatiles should be a negative value,right?

How can I calculate the 'difference (on a mole basis) with the standard state enthalpies'as you mentioned previously?

The heat of combustion that you mean is it refering to the 'heat reaction of burnout' of the combusting particle under DPM?

Thank you very much for your help, I really appreciate your kindness!!

Re: Modelling Biomass Combustion via Species Trans
 

Interesting discussion!

Just wonder about you answer for(3), how to determine the moisture fraction in terms of volumetric fraction? From proximate and ultimate analysis?

Re: Modelling Biomass Combustion via Species Trans
 

Yes, the standard state enthalpy for the wood volatiles should be negative - typically with a value of about: -5e09 J/kgmole.

Calculate how many moles of wood volatiles you'll per kg of wood. Then how many moles of CO2 and H2O when the volatiles are oxidized. Moles times standard state enthalpies will give you the overall enthalpy released. The enthalpy for burning volatiles plus the enthalpy for burning char carbon should be close to the measured value for your biomass.

Heat of combustion would be the complete biomass - volatiles plus char carbon.

Re: Modelling Biomass Combustion via Species Trans
 

The proximate analysis will give the moisture fraction on a mass basis. This can be converted to a volumetric basis using the particle density. I'm not sure why Fluent chose this basis for their wet combustion.

Re: Modelling Biomass Combustion via Species Trans
 

Dear Allan Walsh,

I am very happy to get your professional advices on biomass thermal conversion. Thanks very mcuh. Here I still have some questions for your first/second answers to my post. My questions are as follows:

(1)As we know, when using Fluent6.2, the built-in devolatilization model will be used. So, does the 'lumped' kinetics you mentioned denote the kinetics from the Fluent's devolatilization model? Or, is it your own model? If it is from Fluent built-in model, then how will you distribute the devolatied 'volatiles' into other gaseous species (CO,CO2,tar,....)?

(2)Racheal directed us to UG(pg14-28) for the CxHyOz; however,it is just assumed that the Molucular Weight is to be 100.00. This numerical value is convenient only for treatment purpose;furthermore, the separation treatment for the 'volatile' and 'char carbon' is not provded on pg14-28 . In fact, in realistic cases, the MW of 'volatile' may be other values(instead of 100.00), say 78.00. So, my question is how to determine the average MW for the 'volatile'? What is the calculation prinple or method?

(3) You proposed a very meaning method by using the proximate and ultimate analysis as well as the measured Higher Heating Value(HHV). Then get other components through matrix conversion? Am I right? I cannot understand '....I end up with two unknowns (or one unknown, the ratio of H2O to CO2)..'. Could you give more explanation on this? It will be grealty appriacted if you could guide us with some references for the treatment.

Re: Modelling Biomass Combustion via Species Trans
 

Kino,

(1) The questions asked in this forum seem to fall into two catagories. The first are from students or casual users. For these, using the Fluent scheme for wood-volatiles-air is probably adequate. The second group are from more serious users like academics or industrial users. For this group, in my opinion, the Fluent built-in combustion models are inadequate. My replies to Rachael were assuming that she was interested at more than just the basic level. This may cause confusion for others. But, yes, I am using combustion laws customized with UDFs. I can't share my exact models but can describe a general approach.

(2) I'm a little confused by this very basic question. If you look in the Fluent database for fluids-wood volatiles it gives a MW more like 30 kg/kg-moles. But, for CxHyOz just use x=12, y=1, z=16. Then make sure the subsequent reactions balance. This is just basic chemical engineering. I explained before how to use the fuel ultimate and proximate analysis to get the volatiles elemental composition.

(3) For a general reference on the way I have handled biomass combustion, search for a paper I gave at the 2007 INFUB conference in Porto, Portugal. I can't give you a copy of the spread-sheet that I use, but the principal that you use is to assume certain products from volatilization of biomass - CO, CO2, H2O, CH4, etc. Then, based on how much of each element C,H,O etc. that is in the biomass (based on lab measurements or literature values) you'll develop a matrix. With a bit of book keeping and assumptions, you should be able to get a decent mass and energy balance. Again, I'm assuming that anyone who has to solve this type of problem for work or a research project will have several months to develop it.

Re: Modelling Biomass ... Kino
 

Sorry, it was the 2006 INFUB conference. The years go by!

Re: Modelling Biomass Combustion via Species Trans
 

Kino,

For your question (2),I have considered the problem of having MW = 100, the way I see it (not sure if I'm right) is:

For the case of having MW = 100, you will get a set of (A, B, C, D) for the reaction A CxHyOz + B O2 -> C H2O + D CO2. Say now you use CH(y/x)O(z/x), apparently you will get a different value for MW. This is give you another set of A, B, C, D for the reaction. Therefore, for each different value of MW you use, you will end up defining different coefficient for your reaction in the 'reaction' panel in your 'mixture material'(which in this case is wood-volatiles).

Therefore, I think both ways are correct. Not sure if I'm right, what do you think? Allan, what is your opinion?

Thanks! I really learnt a lot from this discussion!!

Re: Modelling Biomass Combustion via Species Trans
 

Hi Allan,

I'm really happy that what I have got for wood-volatile standard state enthalpy is of the same order of magnitude as yours!

But I want to make sure about something that you mentioned: 'the measured value for your biomass' do you mean the LHV for the fuel? I supposed it is since we are considering the reaction 'wood-volatiles(g) + O2(g) --> CO2 (g) + H2O (g)' which H2O is in gaseous form.

What I have done is: (1) wood-volatiles + O2 --> CO2 + H2O ---- get heat of reaction, I call this H1

(2) char + O2 --> CO2 --- get heat of reaction, I call this H2

therefore, LHV = H1 + H2, which the only unknown in this equation is the standard state enthalpy of wood-volatiles.

I didn't get your last line about heat of combustion clearly. What is known here and what is the unknown? It seems like just the same as the calculation on top. Please correct me if I am wrong.

Thanks a lot!

Re: Modelling Biomass Combustion via Species Trans
 

Hi there, I don't see how can the mass basis value can be converted into volumetric basis. I suppose the particle density is a constant value, so I may just end up with the same fraction. Please help me to gain a better understanding!