Equilibrium model to gasification

[Rodolfo Rodrigues]

Dear Cantera users,

My name's Rodolfo Rodrigues and I'm a PhD candidate in Chemical Engineering at UFRGS, Brazil. I'm studying multipurpose and small-scale biomass gasification systems so that I've found Cantera when I was searching for OpenFOAM applications for gasification process.

By now I'm learning to use Cantera in equilibrium calculations. In this sense there are 2 articles related to gasification: Baratieri et al (2008) [DOI:10.1016/j.biortech.2008.01.006] and Baggio et al (2009) [DOI:10.1016/j.enconman.2009.03.004]

Therefore I built a matlab script by making use of objects Reservoir, Reactor, and MassFlowController to reproduce Baratieri's data. I used 3 main devices: 2 Reservoirs (feedstock and air stream) and 1 Reactor (gasifier) besides 2 MassFlowControllers (fuel to gasifier, air to gasifier). My doubts are about the structure, if it's the best for a steady-state case, and how the LHV fuel is inserted to the Reactor device.

Below it's my preliminary M file:

% rgasifier.m

% 

% Rodolfo Rodrigues - LPR/UFRGS

% rod...@enq.ufrgs.br

%

% 2009/11/26

%% initial data

% biomass (fuel)

% ultimate analysis (%wt)

massfrac = [0.452 0.054 0.39 0.001 0]; % C H O N S

moisture = 0.094; % H2O

ash = 0.009*[0.5 0.5]; % CaO SiO2

LHV = 16.4e6; % J/kg

    % massfrac + moisture + ash = 1

    xmass = [massfrac moisture ash]./sum([massfrac moisture ash]);

    Mw=[12.011 1.00794 15.9994 14.00674 32.066 2*1.00794+15.9994 ...

        40.078+15.999456 28.0855+2*15.9994]; % kg/kmol

    xmol = xmass./Mw/sum(xmass./Mw);

    

    % standard enthalpy of formation

    HfCO2 = -393.777/44; % J/kmol/(kg/kmol) = J/kg

    HfH2O = -241.988/18; % J/kmol/(kg/kmol) = J/kg 

    HfN2 = 0; HfO2 = 0; 

    HfSO2 = -297.012/64; % J/kmol/(kg/kmol) = J/kg 

F_f = 100; % kg/h

T_f = 300.0; % temperature (K)

P_f = OneAtm; % pressure (Pa)

H_f = xmass(1)*HfCO2 + (0.5*xmass(2) ... % enthalpy (J/kg)

    + xmass(6))*HfH2O + 0.5*xmass(4)*HfN2 + xmass(5)*HfSO2 + LHV ...

    - (xmass(1) + 0.25*xmass(2) - 0.5*xmass(3) + xmass(5))*HfO2;

% air (oxidant)

phi = 0.4; % equivalent ratio

%% load components (Baratieri et al, 2008)

gas = IdealGasMix('baratieri2008.cti');

nsp = nSpecies(gas); 

% find component indices

iC = speciesIndex(gas,'C');

iH = speciesIndex(gas,'H');

iO = speciesIndex(gas,'O');

iN = speciesIndex(gas,'N');

iS = speciesIndex(gas,'S');

iH2O = speciesIndex(gas,'H2O');

iCaO = speciesIndex(gas,'CaO');

iSiO2 = speciesIndex(gas,'SiO2');

iO2 = speciesIndex(gas,'O2');

iN2 = speciesIndex(gas,'N2');

% preallocate variable 

x = zeros(nsp,1);

% attribute values

x(iC,1) = xmol(1); 

x(iH,1) = xmol(2); 

x(iO,1) = xmol(3);

x(iN,1) = xmol(4); 

x(iS,1) = xmol(5); 

x(iH2O,1) = xmol(6);

x(iCaO,1) = xmol(7); 

x(iSiO2,1) = xmol(8);

%% build devices

% 1. fuel

set(gas,'T',T_f,'P',P_f,'X',x); % LHV, where?

fuel_in = Reservoir(gas);

fuel_Mw = meanMolarMass(gas);

fuel_Fw = F_f/3600; % kg/s

% 2. air

set(air);

air_in = Reservoir(air);

air_Mw = meanMolarMass(air);

air_Fw = 4.76*fuel_Fw/fuel_Mw*phi*(molfrac2(1) + 0.25*molfrac2(2) ...

        - 0.5*molfrac2(3) + molfrac2(5))/3600; % kg/s

    

% 3. gasifier

set(gas);

gasifier = Reservoir(gas);

%gasifier = Reactor(gas);

%% build flowsheet

% build connectors

fuel2gas  = MassFlowController(fuel_in,  gasifier);

    setMassFlowRate(fuel2gas, fuel_Fw);

air2gas   = MassFlowController(air_in,   gasifier);

    setMassFlowRate(air2gas, air_Fw);

process = ReactorNet(gasifier);

equilibrate(process,'TP')

% 

% END

I will really appreciate any help.

Best regards,

Rodolfo

--
Rodolfo Rodrigues, MSc Eng
LPR/DEQUI/UFRGS
GIMSCOP/DEQUI/UFRGS
Email:  rod...@enq.ufrgs.br
           rod...@chengineer.com

[Daniel]

Hi,

I have been thinking of doing this my self. I just not got around to
do it.

What we could do is share some findings and ideas.

Here is my view of it:

The gasification of wood is very complex so I was thinking of just do
the wood gasifiaction phase and the combustion phase.
In other words. skip all non-wood chemicals... e.g. sulphur is not
natural in the wood. You can have some sulphur in the wood but only if
the trees grow in a sulphur rich land.

This phase should be under 400 degC. I think the process starts at
about 300 degC.


I am not sure if it is good enough to start with the amount C, H, O2,
and N2,N etc in the fuel.

Since the process start with solid fuel I was thinking of starting
with cellulose and lignin etc.
cellolose is C6H12O5 I think... and then set up the reaction
mechanism for each substance in the wood.

So first step is from e.g. 20 degC to 300 or 400 degC - the
gasification.

Second step is combustion part 1
Set the - combustion reactions up to about 700-850 degC where CO etc
is created
Third step
From I think 820- CO burns to CO2

Then at I think 1100 degC you get the Thermal NOx when N2 burns with O
and O2. (if you have any oxygen left in the reactor that is)

Finaly I was also thinking of simplifying the whole thing and use
drywood only.
The only water that will be in the reaction is the hydrogen that burns
with oxygen.
Then when the whole thing is ready a "water" variable can be added
that specify the moisture content of the wood that will slow down the
reactions since some energy in the gasification part is needed to boil
of the water from the fuel.


The Ash can be excluded in this part I think. It is about 2 % in wood
and in straw and ricehusk it goes up to 20%.
The Ash do not burn so in order to keep it simple I would skip that
initially.
It is part of the wood from start to end but very few reactions happen
with the ash. (Unless the temperature is too high so the ash melt etc.
but that has to be treated separately)




Daniel

On Dec 3, 1:16 pm, Rodolfo Rodrigues <rodolfo....@gmail.com> wrote:
> Dear Cantera users,
>
> My name's Rodolfo Rodrigues and I'm a PhD candidate in Chemical Engineering
> at UFRGS, Brazil. I'm studying multipurpose and small-scale biomass
> gasification systems so that I've found Cantera when I was searching for
> OpenFOAM applications for gasification process.
>
> By now I'm learning to use Cantera in equilibrium calculations. In this
> sense there are 2 articles related to gasification: Baratieri et al (2008)
> [DOI:10.1016/j.biortech.2008.01.006] and Baggio et al (2009)
> [DOI:10.1016/j.enconman.2009.03.004]
>
> Therefore I built a matlab script by making use of objects Reservoir,
> Reactor, and MassFlowController to reproduce Baratieri's data. I used 3 main
> devices: 2 Reservoirs (feedstock and air stream) and 1 Reactor (gasifier)
> besides 2 MassFlowControllers (fuel to gasifier, air to gasifier). My doubts
> are about the structure, if it's the best for a steady-state case, and how
> the LHV fuel is inserted to the Reactor device.
>
> Below it's my preliminary M file:
>
> % rgasifier.m
> %
> % Rodolfo Rodrigues - LPR/UFRGS

> % rodo...@enq.ufrgs.br

> Email:  rodo...@enq.ufrgs.br
>            rodo...@chengineer.com

[emil]

Hello Rudolfo,

nice to hear something is going on in biomass research.
I could understand your script and saw no obvious flaw.

Currently I am occupied with other tasks, but I
would like to follow your calculations. I did similar work
in biomass combustion and still have some literature around.

Could you post your mechanism file?

The way you handle H and the lower heating value seems reasonable.
Any reason you have that part with the HfO2 since it should be 0?

@Daniel: For this example it should be possible to skip ash and the
water content of the fuel. However if this goes somewhere near
praxis then water is a must and probably also ash. So if those
parameters
don't make problems they can as well stay in the code.

Keep us informed about your progress!
Regards
Emil

On Dec 5, 2:52 pm, Daniel <daniel_he...@yahoo.se> wrote:
> Hi,
>
> I have been thinking of doing this my self. I just not got around to
> do it.
>
> What we could do is share some findings and ideas.
>
> Here is my view of it:
>
> The gasification of wood is very complex so I was thinking of just do
> the wood gasifiaction phase and the combustion phase.
> In other words. skip all non-wood chemicals... e.g. sulphur is not
> natural in the wood. You can have some sulphur in the wood but only if
> the trees grow in a sulphur rich land.
>
> This phase should be under 400 degC. I think the process starts at
> about 300 degC.
>
> I am not sure if it is good enough to start with the amount C, H, O2,
> and N2,N etc in the fuel.
>
> Since the process start with solid fuel I was thinking of starting
> with cellulose and lignin etc.

> cellulose is C6H12O5 I think... and then set up the reaction

[Daniel]

Hi Emil and Rudolfo


Regarding ash and water. I think it should be in there... for sure...
but I was thinking of keeping it simple in the first phase.

Both the water and the ash have a very big impact in real world
examples.

So it may be just as easy to add those parts from start anyway.


The ash:
Since ash is a fairly undefined gray substance and it changes from
area to area were the biomass has been harvested.
On top of that, it is differences between species as well.

However, the ash will not couse any problems or take part in the
reactions if the temperature is low. When we talk about gasification
phase the temperature is generally up to 600 or 700 deg.C. Combustion
will have started but it will not be complete combustion since CO+
1/2O2 -> CO2 will not happen with full efficiency at high speed(less
than 3-4 seconds) unless the temperature is over 800-850deg C( or in
that area) at atmospheric pressure.

So if the ash is kept cold it will not melt. If it don't melt it will
not be part of any reactions.

I sugest that if ash should be included then it shall be both in terms
of volumes as well as melting temperature.
e.g. 2 % ash- (generally less) and the melting temperature is about
850- 950 degC


The Water:
I sugest that it is treated in two ways.
1) the water that is in the fuel.
There is a number of different ways of defining the amount of water
in wood.
I find the best way is to define it as we have a kg of wood with water
% and wood%
e.g
0 % moisture content is dry wood.
60% or 65% is fresh cut.

The energy needed for combustion and gasification must be enough to
evaporate the water

2) the water that is formed from the Hydrogen that is part of the
fuel








The reason is that

[Rodolfo Rodrigues]

Daniel and Emil,

Thanks for your comments and suggestions. I am really appreciate...

@Daniel:

I am totally agree with you. A kinetic model will be better than an equilibrium model on prediction. It was my next step besides of hybrid models (an equilibrium-kinetic approach) ...  =) 

Now my main goal is to build a 2-phase equilibrium model (using Gibbs free minimization) with cantera reproducing Baratieri et al (2008). It's a reasonable first approach to a gasification system, ie a black-box model.

I have done the assumptions that biomass is consisting of only C, H, O, N, S based on a old message at Yahoo! Cantera user's group (sorry but I could not find it promptly). It is also the assumptions of the authors.

Relating to S and ash contents I think it should be more general assuming them, so that the code could be able to run for other species of biomass. I am particularly working with footwear leather waste as my biomass source.

@Emil:

I have included HfO2 and HfN2 as 0 just for didactic reasons. 

There are not kinetic mechanisms involved since that it is a nonstoichiometric model. I do not have my 'baratieri2008.cti' here but I only put 60 gaseous species + solid carbon. I will attach it in my next message.

I have obtained some results using a code not so structured (using devices connections). That case I started from 'equil.m' sample adapted to a pyrolysis problem (with no air). But my doubt is related to biomass enthalpy eg, in that simple pyrolysis case we will have:

gas = IdealGasMix('myspecies.cti');

set(gas,'T',300.0,'P',OneAtm,'X','C:0.504, H:0.060, O:0.435, N:0.001'); % assuming dry biomass

equilibrate(gas,'TP')

How can I put a biomass enthalpy in the calculations since it is not the total enthalpy of initial species CHON but, it is a function of LHV?

And if you try to run my code using the devices' structure it is not consistent. I am not so confident with my implementation yet.

Regards,

Rodolfo

[Daniel]

Hi,.


I had a few notes on Spruce but I am not sure from what part of the
world the world come from so this is general data.

C: 51.06 wt.%
H: 5.75 wt.%
N: 0.11 wt.%
S: 0.01 wt.%
Cl: 0.01 wt.%

ash content 0.77 wt.%
ash analysis

CO2: 23.07 wt.%
SO3: 1.24 wt.%
Cl: 0.12 wt.%
P2O5: 3.48 wt.%
SiO2: 16.20 wt.%
Fe2O3: 0.78 wt.%
Al2O3: 0.67 wt.%
CaO: 31.95 wt.%
MgO: 3.30 wt.%
Na2O: 0.25 wt.%
K2O: 10.37 wt.%

Daniel.

[emil]

Hi all,

regarding biomass / fuel composition you might be interested in this
database

http://www.ecn.nl/phyllis/

Cheers
Emil

[Rodolfo Rodrigues]

Thank you.

I have already known ECN biomass database. That is a great database. Other good open reference is a NREL atlas:

An Atlas of Thermal Data for Biomass and Other Fuels

http://www.osti.gov/energycitations/product.biblio.jsp?osti_id=82384

Rodolfo

[Rodolfo Rodrigues]

Hi,

Starting from 'equil.m' of cantera samples I did a pyrolysis test M file using pine sawdust:

Ultimate analysis

C:  45.2 wt.%
H:    5.4 wt.%

O:  39.0 wt.%
N:    0.1 wt.%
S:    0.0 wt.%

Moisture: 9.4 wt.%

Ash: 0.9 wt.% (50 wt.% CaO, 50 wt.% SiO2)

LHV: 16.4 MJ/kg

'fig_1_original.png' shows the pyrolysis process for this biomass at 1 bar and 400-1200°C. It is the original solution from the article. My code tries to reproduce this. As you can see they are not equal. In spite of calculate the enthalpy of biomass from LHV I did not add it to the calculations ('set' function, line 64). How can I do this?

Rodolfo

[Daniel]

Nice Work,

I will look at this tonight. I invested in a new Macbook Pro 17"
yesterday :) so I have to install Cantera agian.

Anyway.

when I look at the graphs I think of a few things.(Not 100% sure of
this)

But.

I think that the reaction
2H2+O2 - 2H2O will form faster than 2C+O2 -> 2CO

So I was thinking that the amount of H2 should be less and the amount
of H2O will be higher.



Secondly
the reaction
CO + 1/2O2 -> CO2
Will happen at 850 degC and at 1200 deg C you will have much less CO.

Then again. this is just what I was expecting to see. I am a beginner
in this area.

>  fig_1_original.png
> 10KViewDownload
>
>  fig_1_cantera.png
> 9KViewDownload
>
>  baratieri2008.cti
> 44KViewDownload
>
>  pyrolysis_test.m
> 3KViewDownload

[Rodolfo Rodrigues]

Daniel,

You should take into account that figures represent pyrolysis, ie there are not O2 in the system. We are cracking the biomass to a syngas (H2+CO).

Rodolfo

[Daniel]

Rudolfo,

Yes, I agree that the lack of oxygen will generate CO and H2.

Question:
in your code under "% standard enthalpy of formation"
you do not have the data for CO.

Why is that?

dAniel

[Rodolfo Rodrigues]

Daniel,

Only CO2, H2O, and SO2 enthalpies are involved into calculation of the
biomass enthalpy of formation. That is, if we consider:

C + O2 -> CO2

LHV(C) = Hf(C) + Hf(O2) - Hf(CO2)

so that:

Hf(C) = LHV(C) - Hf(O2) + Hf(CO2)

For biomass, we do not know Hf but LHV is available. Considering that:

[C(a)H(b)O(c)N(d)S(e) + wH2O] + (a + 0.25b - 0.5c + e)O2 ->
aCO2 + (0.5b + w)H2O + 0.5dN2 + eSO2

so that

Hf(biomass) =
LHV(biomass) + a*Hf(CO2) + (0.5b + w)*Hf(H2O) + e*Hf(SO2)


Rodolfo

On Wed, Dec 9, 2009 at 6:54 PM, Daniel <daniel...@yahoo.se> wrote:
> Rudolfo,
>
> Yes, I agree that the lack of oxygen will generate CO and H2.
>
> Question:
> in your code under "% standard enthalpy of formation"
> you do not have the data for CO.
>
> Why is that?
>

> Daniel

[Daniel]

Rodolfo,

Yes, but isn't it so that CO2 is not created directly.

I assumed that in pyrolysys it will still be:

1)C+1/2O2 ->CO
2)2CO+O2 ->2CO2

Reaction 1 will compete with

H2+1/2O2 ->H2O

Or am I missing something

Daniel




Daniel

[Rodolfo Rodrigues]

Daniel,

That is the method to calculate an formation enthalpy from LHV
demonstrates in Souza-Santos' book (Solid fuels combustion and
gasification) at page 387. Coincidentally there is a preview in Google
books:
http://books.google.com/books?id=WLfTLZzkI4sC&pg=PA387

Somehow I must have enthalpy of formation for biomass to calculate
initial conditions.

Rodolfo

[ann]

Hallo all,

Rodolfo - Thank you very much for posting your example!
I see you stated (as it is also in Bartieri) that you use in your .cti
file, apart from the gri30 mechanism, the solid C.

It seems that everybody sees this crystal clear, but I dont understand
how you did thus.. When I read Bartieri s paper I thought that he use
a heterogeneous reaction, but from your code you probably didn't.

How do you manage to do this?

Ann

On Dec 10 2009, 2:08 pm, Rodolfo Rodrigues <rodolfo....@gmail.com>
wrote:

> Daniel,
>
> That is the method to calculate an formation enthalpy from LHV
> demonstrates in Souza-Santos' book (Solid fuels combustion and
> gasification) at page 387. Coincidentally there is a preview in Google
> books:http://books.google.com/books?id=WLfTLZzkI4sC&pg=PA387
>
> Somehow I must have enthalpy of formation for biomass to calculate
> initial conditions.
>
> Rodolfo
>

[Rodolfo Rodrigues]

Ann,

Thanks for your considerations.
Take a look at Appendix A from Baratieri's paper. There, the authors
explain the applied modeling approach. Baratieri uses the Gibbs-free
minimization method to find the final component distribution. This
algorithm that is implemented in the Cantera.
Nevertheless there is already a non-solved failure at the input
enthalpy in that code.

Rodolfo

[ann]

Hi Rodolfo,

thanks for your answer. Yes, it s good that you pointed out that the
appendix contains the explanation of the "equilibrate" from cantera.

My question was more related to the way you declare the carbon C(s) in
the cti file. When I first looked at the Baratieri's paper in Table 2
(solid and gas phase) I thought they declare the C(s) as a
stoichiometric_solid with a surface_reaction label. Did, you declare C
(s) in this manner?

I am asking this because I am interesting in doing ash transformation
with Cantera, but I haven't set up a model (that works) yet - so your
example was very useful!

thanks a lot,

ann

On Jan 11, 3:41 pm, Rodolfo Rodrigues <rodolfo....@gmail.com> wrote:
> Ann,
>
> Thanks for your considerations.
> Take a look at Appendix A from Baratieri's paper. There, the authors
> explain the applied modeling approach. Baratieri uses the Gibbs-free
> minimization method to find the final component distribution. This
> algorithm that is implemented in the Cantera.
> Nevertheless there is already a non-solved failure at the input
> enthalpy in that code.
>
> Rodolfo
>

[Rodolfo Rodrigues]

Ann,

I really did not worry about the solid carbon. I've only added C(s) as
graphite carbon the same way that other gaseous compounds. I was
trying to do this simple.

Hey! One of my main objectives is also adding ash transformation since
that's important to my analyses. So I've started from pyrolysis to try
the simplest possible in order to go to a gasification analysis. Maybe
you can cooperate. Email me in private.

Rodolfo

[Akshay Singan]

Hi Rodolfo,

I am beginning work on the underground gasification of coal, and would like to learn the use of cantera for predicting equilibrium composition. I have seen the pyrolysis test case, and would like to know how to proceed to include ash effects.

Also i d like to have control on what reaction mechanism to provide as a constraint in the gibbs minimization.

It d be very helpful for me, if you could help me in this regard and let me know how you d approached it, or point me to some resources that are for this specific problem.

Thank you

Akshay

[Rodolfo Rodrigues]

Hi Akshay,


I have two publications about equilibrium approach using Cantera for gasification problems. However, both were presented in conferences so they are very short. They are available to download so, please, take a look at them.

[1] Rodrigues, R.; Marcilio, N. R.; Trierweiler, J. O.; Godinho, M. & Pereira, A. M. S. Co-gasification of footwear leather waste and high ash coal: A thermodynamic analysis. 27th Ann. Int. Pittsburgh Coal Conf., 2010. URL: http://rodolfo.chengineer.com/rodrigues_2010.pdf

[2] Rodrigues, R.; Marcilio, N. R.; Trierweiler, J. O. & Godinho, M. Thermodynamic efficiency analysis of gasification of high ash coal and biomass. Int. Conf. Coal Sci. & Tech., 2011. URL: http://rodolfo.chengineer.com/rodrigues_2011.pdf

You can assume ash as an additional phase in the Gibbs minimization. This might work well for TP constant cases. Burcat's database has a good number of species to ash modeling.

Akshay, you have to be aware Gibbs minimization approach doesn't take into account reaction mechanism. You should just assume larger or shorter thermodynamic database insofar you need more or less detailed prediction of the equilibrium composition.


Regards,
Rodolfo

--
You received this message because you are subscribed to the Google Groups "Cantera User's Group" group.

To view this discussion on the web visit https://groups.google.com/d/msg/cantera-users/-/GhC-yGb7kw4J.

To post to this group, send email to canter...@googlegroups.com.
To unsubscribe from this group, send email to cantera-user...@googlegroups.com.
For more options, visit this group at http://groups.google.com/group/cantera-users?hl=en.

--
Rodolfo Rodrigues, MSc Eng

  PhD-student in Ch. Engineering, UFRGS

  Research scholar at ORNL

  Visiting scholar at UTK

[akshay singan]

Hi Rodolfo

Thank you for your timely reply. I was able to generate some results for coal gasification, i d like to know if i can specify the reaction data in the cti file, in the same format as given in your "gri30.cti" file.

I am also trying to compare the output gas composition at equilibrium with the output composition i got from ASPEN R GIBBS module.

Also, I was wondering if you have come across any examples for liquid phase reactions in ideal reactor configurations. We are interested in trying the same using CANTERA.

Thanks again

Akshay

Regards

--

Akshay S

[Rodolfo Rodrigues]

Hi Akshay,

You can specify the reaction data as same as you can find in "gri30.cti" file. There is information about how to work with CTI files in Cantera documentation: http://cantera.github.com/docs/sphinx/html/cti/input-files.html

And, sorry, I don't have any experience with liquid-phase reactions for your case. Maybe you could try out liquid-phase reactions in CHEMKIN format so that you might import these one to use in Cantera.

Regards,

Rodolfo

Regards

--

Akshay S

--

You received this message because you are subscribed to the Google Groups "Cantera User's Group" group.

[Ajay Singh Redhu]

Hi Ann and everyone else,

I am facing the same problem Ann talked about. How to consider C(solid) in gri3.0.yaml file. I want to get solid carbon in the products after wood pyrolysis using equilibrate function in Cantera. I tried adding C(gr) in the species of gri3.0 with the thermal properties of C(gr) same as from graphite.yaml. But when I am running the code it is showing an error asking for transport properties of C(gr). Can anybody please help me with that.

Thanks,

Ajay