proceedings of the 2016 international conference on ...ieomsociety.org/ieom_2016/pdfs/155.pdf......

Proceedings of the 2016 International Conference on Industrial Engineering and Operations Management Kuala Lumpur, Malaysia, March 8-10, 2016

Thermochemical Processing of Biomass: Graphic User Interface Models for Costing and Material Balances

Calculations Rois Fatoni

Chemical Engineering Department Universitas Muhammadiyah Surakarta

Jawa Tengah, Indonesia [email protected]

Abdulhalim Abdulrazik, Zhengkai Tu, and Ali Elkamel Department of Chemical Engineering

University of Waterloo Waterloo, Ontario, Canada

Abstract—This paper presents models for the calculations of the costs involved and the general material balances during the conversion of the biomass into the final products. These models deal primarily with the conversion processes through thermochemical processing of biomass including gasification, combustion and pyrolysis. The models were developed in the MATLAB environment, all with Graphic User Interface (GUI), meaning that they interact with the user through windows and dialogs instead of through user-input commands. The user only needs to insert values in the model inputs and click a button or two, and the models will be able to generate the results. The developed models can save time and cost to understand and design biomass utilizing systems that deal with thermochemical routes.

Keywords— Graphic User Interface Model; biomass; bio-products; combustion; gasification; pyrolysis

I. INTRODUCTION

Increasing awareness and demands for sustainable products have gradually shifted conventional practices and utilizations of non-renewable resources into renewable resources including biomass. Biomass feedstocks are flexible and sharing similar conversions as fossil feedstocks do [1], have emerged as an important feedstock for many industrial processes to produce bio-based products or bio-products that range from chemicals, energy and materials. In this paper, models have been established and are able to calculate costs and the general material balances that involve in the processing of the biomass into those bio-products. Even though biomass conversion routes can be generally classified into thermochemical, chemical and biochemical, as reported by [2], this paper only discuss the practicality of the developed models to deal with thermochemical pathways which include gasification, combustion and pyrolysis.

Biomass combustion is a thermal process that converts biomass source entirely to carbon dioxide and water vapor, thus precluding conversions to intermediate fuels or chemicals [3]. As the main purpose of combustion is to convert chemical energy stored in the fuel into electrical energy, reference [4] highlighted three key components exist in a power plant that included i) a mean of converting fuel to heat, ii) a mean of converting heat into mechanical energy, and iii) a mean to convert mechanical energy into electrical energy. Biomass gasification meanwhile is the conversion of biomass feedstocks by partial oxidation into gaseous product called as synthesis gas or also known as syngas, containing primarily of hydrogen and carbon monoxide, with lesser amounts of carbon dioxide, water, methane, higher hydrocarbons and nitrogen [5]. Gasification is considered one of the most efficient ways of converting the energy embedded in biomass, and it is becoming one of the best technological alternatives for solid wastes reuse [6]. Finally, pyrolysis is the conversion of biomass feedstocks to liquid (also known as bio-oil or bio-crude), solid (bio-char) and gaseous (bio-gas) fractions in the absence of air. This thermochemical conversion process can be divided into fast pyrolysis and slow pyrolysis, which the difference between two is depending on how fast biomass feedstocks are heated relative to the pyrolysis reaction time and the residence time [7].

II. BASIC ABOUT THE MODELS

The models were developed in the MATLAB’s environment, all with the Graphic User Interface (GUI), meaning that they interact with the user through windows and dialogs, instead of through user-input commands. In this way, the user does not

557© IEOM Society International

Proceedings of the 2016 International ConfKuala Lumpur, Malaysia, March 8-10, 2016

need to know much about MTALAB to usetwo, and the models will be able to generate

A screenshot of one model is depicted iThe user simply needs to enter the informatwill then calculate the results and display the

This section was written to provide a briof user who do not has an experience runnincontains the code, and the ‘.fig file’ which cand Combustion.fig) into the same directorGUIDE’, as shown by Fig. 2. After that, in This step is shown by Fig. 3.

Fig. 2. Opening the .fig file

ference on Industrial Engineering and Operations Manag6

e the model. He or she just needs to type in the model inpthe results.

n Fig. 1, which clearly shows that the model has a straightion into all the white ‘boxes’, and the click on the ‘Calcem in the gray ‘boxes’.

Fig. 1. Screenshot of the model for treatment cost

III. RUNNING THE MODELS ief overview of the model files and how they are run in M

ng the software before. In short, each model consists of twocontains the graphics. To run the model, copy the pair of ry. Then, open MATLAB, and right-click on the ‘.fig fthe GUIDE environment, click ‘Run Figure’ or press ‘Ct

Fig. 3. Running the Model

gement

puts and click a button or

htforward User Interface. culate’ button. The model

MATLAB, for the benefits o files, the ‘.m file’ which f files (e.g. Combustion.m file’ and choose ‘open in trl + T’ to run the model.



IV. DETAILS OF THE MODELS There are a total of five models: one for the calculation of the pre-treatment cost, i.e. the price of the biomass before

conversion into other products; one for the calculation of the cost of processing the biomass via the thermochemical path; and the rest three for material balances in different processes, namely gasification, combustion, and pyrolysis. Each model will be explained in details, followed by a hypothetical case study that demonstrates the sample inputs and outputs.

A. Gasification (gasification.m and gasification.fig) The model for the gasification mass and energy balance is based on a hypothetical sequential reaction network. Biomass,

assumed to be in elemental forms (thus C, H2, N2, S and O2), is fed to the gasifier together with steam. Sulfur is oxidized by oxygen to SO2 first; unreacted oxygen then oxidizes part of the carbon to CO2; lastly, the remaining carbon reacts with the steam to give syngas. The first two reactions are assumed to be complete, with 100% conversion, while the last reaction has a user defined conversion. Based on the three reactions, material balance can be performed to calculate the product composition. Once the product composition is known, energy balance is performed to find the output temperature. To be more detailed, the heat capacity of each component in the reactants and products is formulated, based [8]. With known input temperature, the enthalpy of the reactants can be calculated. Assuming no work is done, with the knowledge of heat duty, the enthalpy of the products can be found. MATLAB then uses the fzero function to find the temperature of the products iteratively.

The inputs of the model are the composition (for gasification, the elemental composition) of the biomass, mass flow rate and temperature of the input streams, conversion of the last reaction, and heat duty. The outputs are very detailed; the product composition, syngas composition and temperature as well as high heating value (HHV) are of our primary interests. For the case study, soybean stalk was fed at 25oC at a rate of 1000 kg/h to the gasifier with a stream of steam at 100oC at a rate of 1000 kg/h. Assuming adiabatic operation (0 kW heat supplied to the gasifier) and a syngas conversion of 80%, it was found that syngas can be produced at a rate of around 600 kg/h at a temperature of 880oC, with a H2 to CO ratio of 73.5 : 26.5 (in volume), and an HHV of 32,000 kJ/kg.

B. Combustion (Combustion.m and Combustion.fig) Similarly, the model for combustion material balance is based on a sequence of reactions. The reactions appear to be

analogous to those in gasification, except that the water-gas shift reaction is replaced by the combustion of hydrogen to form water. Biomass and oxygen are fed to the reactor, and the model calculates the heat output from the combustions. Note that for the combustion model, the user needs to specify the output temperature. This is due to the fact that energy is now an output of the model (while it is not for the gasification model), and the temperature needs to be specific to get rid of the degree of freedom. For the hypothetical case study, soybean stalk was again used as the biomass feedstock. The biomass and oxygen were fed both at 25oC and a rate of 1000 kg/h to the reactor. Assuming adiabatic condition, an output temperature of 25oC and completion of 80% of both the combustion of carbon and of hydrogen, the model then calculates that 4938 kW of heat can be produced.

C. Pyrolysis (Pyrolysis.m and Pyrolysis.fig) The model for pyrolysis is, on the other hand, a kinetic model. Unlike gasification and combustion which generates a

limited number of species, pyrolysis gives a large variety of products, for which a kinetic model rather than a sequential model might be more appropriate. The whole model is adopted from [9], which describes a generalized characterization of biomass and reaction framework. Since the data for morphological analysis, thus the fraction of cellulose, hemicellulose and lignin might not be available at times, a utility of ‘morphological analysis’ was added into the model. This utility calculates the C:H:O ratio from the elemental composition, and decomposes the biomass in three species P1 P2 and P3. P1 is a mixture of cellulose and hemicellulose in the ratio of 6:4, the most common ratio found in nature; P2 and P3 are both mixtures of different kinds of lignin. The biomass was characterized in these species based on elemental balances. Note that this method was limited because certain biomass, or in fact a large number of biomass cannot be characterized this way (though the material balances can still be done with a negative proportion of some species). Therefore, whenever possible it is best to directly provide the fraction of cellulose, hemicellulose and lignin.

To be short, the model works by solving a system of ordinary differential equation (ODE), where the rate of change of each species is a function of temperature and the concentration of reactants. A batch operation is assumed with a user-defined residence time. The solver in the MATLAB then integrates the system to get the moles and mass of all species at the end. For demonstration purposes, for the pyrolysis model, a hypothetical biomass that has slightly higher carbon content than soybean stalk was used (so that the morphological analysis utility gives positive fractions). With a C:H:O ratio of 0.5 : 0.0632 : 0.4786, the morphological analysis gives a cellulose fraction of 0.47, hemicellulose fraction of 0.32, and the rest lignin. Then,



the biomass feed rate of 1000 kg/h was assigned at 25oC. With a residence time of 2 second, the model gives an output that has a large proportion of CO, CO2, and hydroxyacethaldehyde (HMFU).

D. Pre-treatment Cost (PretreatmentCost.m and PretreatmentCost.fig) The model for calculating the pre-treatment costs was based on [10]. The pre-treatment costs consist of the production

cost (fixed and variable) at the farmers’ gates, the processing cost for pre-treatment, and lastly, the transport cost. For the production cost, it was assumed that only the data for the first 3 years (or shorter for certain cost items) would be needed; the cost in later years will be based on the cost in the 3rd year, adjusted by inflation. For the pricing, it was assumed that a relatively constant price (which has yet to be calculated) was agreed between the farmer and the pre-processing plant for the whole production period. With all the input completed, and an assumed average profit level, the price at the farmers’ gate was then calculated (iteratively if the price is to increase slightly each year.) For the pre-treatment cost, the cost in operation was straightforward since it is simply an addition of all the cost items. The cost in financing was similar, with assumed interest rate of 6%, return on equity of 17.5%, and repayment period of 10 years. For the transport cost, as from the paper, it was assumed that for all the three mode of transport, i.e. truck, rail and marine, there are a fixed part and a variable part. Thus, the cost can be correlated by

cost = C1 * distance + C2 (1) The values of two constants, i.e C1 and C2 in (1) were taken from [10], however the user is free to change them in the .m file.

E. Treatment Cost (TreatmentCost.m and TreatmentCost.fig)

The model for calculating the treatment cost was based on [11]. The model is very similar to the pre-treatment cost except that inputs and results from the material balance models might be needed, and the user now has more flexibility for the financing cost. The original model was only for costs in gasification, but in this paper, it was generalized to accommodate for all three different thermochemical processes. For the hypthetical case study, pre-entered values of all parameters were kept. Having said that, a capacity factor of 85% (or 85% of the time the plant is in operation), a biomass price of $200/tonne, and a desired profit margin of 20% were used. For the gasification pathway with syngas production rate of 0.6 tonne/h, the model resulted that the desired price for syngas needs to be $453.3/tonne; for the combustion pathway with heat output rate of 4938 kW, the heat cost needs to be more than $0.055/kWh; lastly for the pyrolysis pathway with the production rate of products assumed to be 0.3 tonne/h, the product price needs to be $906.6/tonne.

V. CONCLUSION AND FUTURE WORKS In conclusion, the user friendly GUI has been developed that enables process designer to quickly determine costs and

material balances in thermochemical processing of biomass. User may only require typical available data of received biomass such as elemental analysis for the combustion and gasification and morphological analysis for the pyrolysis. These benefits could save time and cost to understand and design those systems.

For the future works, the first area would be to integrate the cost models and material models so that a holistic model can be built to ease analysis and fine-tune the GUI. The second area, which is of great interest, is to modify the models so that they can be ready for optimizations. In theory, there are many options on this including the optimization codes can be directly added into the model considering the powerful optimizing capability of MATLAB.

ACKNOWLEDGMENT The authors would like to acknowledge the Natural Sciences and Engineering Research Council of Canada (NSERC) for

partially funding to this research.

REFERENCES [1] Prabir Basu, Biomass Gasification, Pyrolysis, and Torrefaction: Practical Design and Theory, 2nd Edition, Academic Press, New York,

2013. [2] Roberto Garcia, Roberto Garcia, Consuela Pizarro, Antonio G. Lavin, Julio L. Bueno, “Characterization of spanish biomass wastes for

energy use”, Bioresource Technology 103(2012) 249-258. [3] Spencer Abraham, Donald L. Evans and John H. Marburger III, “A report on U.S. climate change technology program: technology

options for the near and long term”, 2003. [4] Francis M. Vanek and Louis D. Albright, Energy Systems Engineering: Evaluation & Implementation, Mc Graw-Hill, 2008. [5] Jared P. Ciferno and John J. Marano, “Benchmarking biomass gasification technologies for fuels, chemicals and hydrogen

production”, A report prepared for U.S Department of Energy, 2002. [6] Maria Puig-Arnavat, Joan Carles Bruno and Alberto Coronas, “Review and analysis of biomass gasification models”, Renewable and

Sustainable Energy Reviews 14 (2010) 2841-2851



[7] Prabir Basu, Biomass Gasification and Pyrolysis; Practical Design and Theory, Elsevier Inc., UK, 2010. [8] Don W. Green and Robert H. Perry, Perry’s Chemical Engineers’ Handbook, 8th Edition, McGraw-Hill, New York, 2008. [9] Ranzi E., Cuoci A., Faravelli, T., Frassoldati A., Migliavacca G., Pierucci S., Sommariva S., “Chemical kinetics of biomass

pyrolysis”, Energy Fuels 2008, 22 (6), 4292–4300. [10] Oo A., Kelly J., Lalonde C., “Assessment of business case for purpose-grown biomass in ontario”, Western University, London,

Ontario, 2012. [11] http://biomass.ucdavis.edu/files/tools/energy-cost-calculator-gasification-power-generation-model.xls (accessed on 23rd April 2015).

BIOGRAPHY

Rois Fatoni is an Assistant Professor, and Department Chair of Chemical Engineering Program, Universitas Muhammadiyah Surakarta, INDONESIA. He earned S.T. in Chemical Engineering from Universitas Gadjah Mada, Yogyakarta, Indonesia, M.Sc. in Environmental Process Design from University of Manchaster, UK, and Ph.D in Chemical Engineering from University of Waterloo, Canada. During his PhD studies he was supervised by Professors Ali Elkamel and Leonardo Simon. His research interests include modeling, simulation, optimization, scheduling, process safety, mixture design, advanced materials, biomass, and innovative technology for medium-small scale industries. He can be reached via email at: [email protected] Abdulhalim Abdulrazik is currently a PhD candidate at University of Waterloo. He earned B.Eng in Chemical-Gas Engineering from Universiti Teknologi Malaysia, and Masters in Process Integration from Universiti Teknologi Petronas. Before joining Universiti Malaysia Pahang as a lecturer, he gained industrial experiences from oil and gas companies. His research interests include modelling and optimization of biomass and natural gas systems. He is a member of Board of Engineers Malaysia. Zhengkai Tu is a 4th year chemical engineering student at the University of Waterloo with options in chemistry, statistics and management sciences. He also works as full-time and part-time research assistant at the university. Previously, he worked at various co-operative education positions in large corporations including Hatch Ltd, Rockwell Automation and NOVA Chemicals. He is actively looking for research opportunities to develop next generation computation software based on cloud technology and parallel algorithms, and to integrate chemical and material sciences into additive manufacturing.

Ali Elkamel is Professor of Chemical Engineering at the University of Waterloo. He holds a BSc in Chemical Engineering and BSc in Mathematics from Colorado School of Mines, MS in Chemical Engineering from the University of Colorado-Boulder, and PhD in Chemical Engineering from Purdue University (West Lafayette), Indiana. His specific research interests are in computer-aided modelling, optimization and simulation with applications to energy production planning, sustainable operations and product design. He has supervised over 70 graduate students (of which 30 are PhDs) in these fields and his graduate students all obtain good jobs in the chemical process industry and in academia. He has been funded for numerous research projects from government and industry. His research output includes over 190 journal articles, 90 proceedings, over 240 conference presentations, and 30 book chapters. He is also a co-author of four books; two recent books were published by Wiley and entitled Planning of Refinery and Petrochemical Operations and Environmentally Conscious Fossil Energy Production.


proceedings of the 2016 international conference on ...ieomsociety.org/ieom_2016/pdfs/155.pdf......

Documents