Diapositiva 1

1
ENHANCED MODELLING AND INTEGRATED SIMULATION OF
GASIFICATION AND PURIFICATION GAS UNITS TARGETED
TO CLEAN POWER PRODUCTION
Mar Pérez-Fortes, Aarón Bojarski, Sergio
Ferrer-Nadal, Georgios Kopanos, José Mª Nougués,
Enrique Velo and Luis Puigjaner CEPIMA
Group Chemical Engineering Department Universitat
Politècnica de Catalunya (Spain)
2
Brief Outline

• Tool proposed
• IGCC Modeling
• Model integration approach
• Results
• Conclusion and further work

3
1. Tool
The main objective is to provide a support tool
for optimal assessment of different gasification
plant configurations, under different input data
sets.
KEY ISSUE for IGCC plants The production of
CLEAN GAS. To achieve the required quality to
feed a gas turbine or to produce hydrogen.
4
2. IGCC Model Gasifier
1. The GASIFIER is a PRENFLO for co-gasification
of COAL and PETCOKE.

• Non-isothermal adiabatic reactor.
• Heat integration is considered.
• Different reaction steps occur sequentially.
• After the pre-treatment, feedstock enters with a
  2 in wt of moisture.
• Air Separation Unit ? Oxygen with a 85 of purity
  in.
• In a previous step to Heat Recovery Steam
  Generator, a quench gas stream cools syngas from
  1600 to 800º C.

It can be adapted, through PARAMETER
RECONCILIATION, to a fluidised bed. Changes on
Moisture, temperature, pressure and particle
diameter.
Model based on Badzioch, S., 1970 Loison, R.
1964 Usón, S., 2004 García-Labiano, F., 1995
and 1996 Govind, R., 1984. Data from Higman, C.
and Van der Burgt, M., 2003.
5
2. IGCC Model Steam/Power generation
2. HRSG. The heat from the turbine exhaust gases
is recovered here, producing steam at three
pressures (127, 65 and 6.5 bar). In the HP and IP
water steam systems, water is also heated by the
syngas refrigeration process (from 800º C to 240º
C)
3. COMBINED CYCLE. Brayton Cycle The clean gas
is subjected to a saturation and a dilution
process to reduce the NOx formation during
combustion. The air is compressed and separated
into two fluxes to the ASU and to the GT.
Rankine Cycle The generated power is the
addition of the effect of the three water steam
circuits (HP, IP and LP).
6
2. IGCC Model Pollution abatement (1)
4. VENTURI SCRUBBER AND SOUR WATER STRIPPER.
Abatement of Syngas contaminant compounds,
mainly CO2, H2S, NH3 and HCN with water. Polluted
sour water is further cleaned with steam
stripping, using two columns one for acid and
another for basic pollutants. Electrolytes
chemistry is used to model phase equilibria.
7
2. IGCC Model Pollution abatement (2)
5. COS HYDROLYSIS REACTOR. Catalysed reaction
with alumina COS H2O ? H2S CO2. r k COS,
and Arrhenius behaviour for k. Kinetic
parameters A 1.71107 m3/gcath and Ea
94733,9 J/gmol (Fiedorow, R., 1984 Huang, H.,
2006).
6. MDEA ABSORBER. It removes the H2S from syngas
by means of MethylDiEthanolAmine (MDEA) which is
further recovered in the desorption column. With
no entrance of MDEA, all the flux is recovered.
8
2. IGCC Model Pollution abatement (3)
7. CLAUS PLANT. It is used to remove sulphur
(H2S, SO2) and nitrogen (NH3) components from the
polluted gas streams by producing liquid sulphur
and N2. It consists of two parallel thermal
stages and two catalytic stages, with alumina as
catalyst. The hydrogenation, also catalysed, is
used to increase the overall sulphur recovery.
The catalytic stages are considered equilibrium
reactors. Kinetic reactions information
(expressions and parameters) from Hawboldt, K.
A., 1998 and 2000 Monnery, W.D., 2000 and 2001.
9
2. IGCC Model Overall Flowsheet (1)
Gasification, ASU and CC
Raw gas and raw coal materials
Clean syngas
10
2. IGCC Model Overall Flowsheet (2)
Gas Cleaning
Raw gas and raw coal materials
Clean syngas
11
2. IGCC Model Software tools used
The model has been developed using two main
commercial process simulation software
Aspen HYSYS
Aspen PLUS

• Electrolytes chemistry

• Unit Operation Extensions
• Reaction Extensions
• Hypothetical components

• Strippers, flash vessels
• MDEA, Sour Water treatment

• Compressors, Chemical reactors, flash vessels,
  etc
• Gasifier, Turbines

12
3. Model integration approach

• The whole model has three different parts
• AspenPlus calculations, involving ionic
  species, are used for phase equilibrium problems
  solution (for venturi scrubber, sour water
  stripper and MDEA absorber)
• ANNs are used as data driven models. To train
  each ANN, data from sensitivity analysis in
  AspenPlus are used. Key performance variables
  are selected from sensitivity analysis. Training
  is carried out using the ANN package provided
  with Matlab 6.5.
• AspenHysys is used to integrate all models
• ANN models based on AspenPlus results
• AspenHysys library models
• Gasifier and Claus plant reaction models.

13
3. Model integration approach
Parameters Estimation / Optimization
MatLab
Solver SQP/GA

• It allows for treating any simulation flowsheet
  as an implicit function
• Y f (x)
• Y (simulation output / calculated values)
• f (actual simulation models, not directly
  available for modification)
• x (simulation input / user entered values)

14
4. Results single models
Elcogas and simulation data comparison, for each
pollution abatement model.

• Comparing venturi scrubber and sour water
  stripper simulation values with real data
• The model shows good agreement with industrial
  data.

15
4. Results single models
Elcogas and simulation data comparison, for each
pollution abatement model.

• MDEA absorber. The real and the predicted
  composition for the main components of the clean
  gas show good agreement.
• Claus plant, liquid sulfur real and predicted
  values are in good agreement (3,113 and 2,810
  kg/h, respectively).The main difference between
  the predicted and the real stream is the CO
  composition is in the recycle gas.

16
4. Results Global flowsheet
The main input parameters, inlet conditions from
the real plant and directly introduced in the
plant modeling, for different raw materials
Feedstock composition
Gasifier
17


4. Results Global flowsheet
The main outlet parameters of the IGCC plant
the produced electricity by the model has a good
agreement with the real produced power.
18
5. Conclusion and Results Discussion

• Differences found between simulated and real
  data may be caused by a combined effect from
  several simplifications in which this model
  relies on.
• The pyrolisis model that estimates the
  production of char, nitrogen and sulphur
  compounds is based on experimental correlations.
• Char combustion and gasification reactions are
  also based on experimental correlations. However,
  these correlations have been taken from the
  literature and do not exactly correspond to the
  actual raw material mixtures.
• ANN results are limited to an interval of
  variation of gases composition.
• Also, combustion of the clean gas (gas turbine
  model) is modeled with a Gibbs equilibrium
  reactor.

19
6. Further work

• ANN variables selection will be done on the
  basis of global sensitivity indexes. Allowing for
  emphasizing modelling on most sensitive
  variables.
• A more detailed model of the gasifier is under
  study.
• Literature data is being searched to obtain
  better experimental correlations to model current
  raw materials behaviour.

20
Acknowledgements Financial support received from
the European Community projects
(MRTN-CT-2004-512233 RFC-CR-04006), the
Generalitat de Catalunya with the European Social
Fund (FI grant) and the Ministerio de Educación y
Ciencia (FPU grant) is fully appreciated. ELCOGAS
IGCC power plant provision of data for validation
purposes is acknowledged with thanks.
Thanks for your attention!
21
Questions?
Contact emails luis.puigjaner_at_upc.edu mar.perez-f
ortes_at_upc.edu
22
ENHANCED MODELLING AND INTEGRATED SIMULATION OF
GASIFICATION AND PURIFICATION GAS UNITS TARGETED
TO CLEAN POWER PRODUCTION
Mar Pérez-Fortes, Aarón Bojarski, Sergio
Ferrer-Nadal, Georgios Kopanos, José Mª Nougués,
Enrique Velo and Luis Puigjaner CEPIMA
Group Chemical Engineering Department Universitat
Politècnica de Catalunya (Spain)
23
3. IGCC
The model has been implemented using two main
softwares
Modelisation
The model has been implemented using two main
softwares
24
WHAT WE ARE GOING TO DISCUSS
2. THE SOFTWARE TOOL Commercial softwares used
to develop the platform.

• THE IGCC PLANT
• Modeled units with general explanations.

RESULTS, DISCUSSION AND CONCLUSIONS WITH FUTURE
WORK.
2
25
4. RESULTS
LCA
26
The model has been implemented using two main
softwares
Punto que hace referencia a el CO2 removal,
siguiendo con la simulación de IGCC
Punto que hace referencia a el CO2 removal,
siguiendo con la simulación de IGCC
The model has been implemented using two main
softwares
3. IGCC
The model has been implemented using two main
softwares
17