Format And Type Fonts CHEMICAL ENGINEERING TRANSACTIONS VOL. 45, 2015 A publication of The Italian Association of Chemical Engineering www.aidic.it/cet Guest Editors: Petar Sabev Varbanov, Jiří Jaromír Klemeš, Sharifah Rafidah Wan Alwi, Jun Yow Yong, Xia Liu Copyright © 2015, AIDIC Servizi S.r.l., ISBN 978-88-95608-36-5; ISSN 2283-9216 DOI: 10.3303/CET1545050 Please cite this article as: Ishak S.A., Hashim H., Muis Z.A., 2015, Optimal low carbon cement production cost via co- processing and carbon capture and storage, Chemical Engineering Transactions, 45, 295-300 DOI:10.3303/CET1545050 295 Optimal Low Carbon Cement Production Cost via Co-Processing and Carbon Capture and Storage Siti A. Ishak*, Haslenda Hashim, Zarina Ab. Muis Process Systems Engineering Centre (PROSPECT), Faculty of Chemical Engineering, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia siti.aktar@gmail.com Cement production is recorded to have released about 5 % of current global man-made CO2 emissions. They came from the burning of fossil fuel in kiln, electricity usage from grinding of raw and finished materials, and from the calcination of main raw material; limestone to produce clinker. This paper objective is to find the best minimized solution of cement production cost while reducing CO2 emission and without compromising the quality of cement product. This is achieved by developing mathematical optimisation model that will be executed using General Algebraic Modelling System (GAMS). Some of data used in this research will be costs to install the technologies, costs of raw materials and fuels, properties of raw materials, CO2 emission improvements from CO2 reduction technologies and others. CO2 reduction technologies for cement industry considered in this study are co-firing (co-process) and carbon capture and storage. This paper also discusses the best combination of energy efficient technologies to meet the CO2 reduction target and product specification. 1. Introduction Global warming causes by the greenhouse effect have been on an increase motion from past decades paralleling the increase of industrial activities globally. CO2 is emitted from both manufacturing process and also from the chemical transformation of the raw materials in cement industry. As clinker production produces 90 % of cement plant carbon emission; with 40 % from fuels burning and 50 % from chemical reaction around cement kiln (Benhelal et al., 2013), measures to reduce carbon emissions should be focused around it. One of the most common mitigation methods is co-processing. Co-processing aims to exchange carbon-based thermal sources with greener thermal sources (less carbonemissions), such as natural gas, biomass or biogenic fuels. Wennersten et al. (2015) predict that the development of alternative fuels is still marginal in the future as opposed to fossil fuels and discussed the possibility of adapting carbon capture and storage (CCS); a technology that comprises of three stages: carbon capture, transport and storage. One of the more prominent stages consists of carbon capture stage. Carbon can be captured from products or gas streams in three ways: post combustion, oxy-fuel, and pre combustion. From an economical point of view, implementation of CCS is limited due to the addition of capital investment while environmentally; CCS is highly favourable as they are capable in reducing up to 74 % CO2 emission (Barker et al., 2009). Other possibility was suggested by Gul et al. (2014) where locally available mineral material (LAMM) that possess cementitious properties was added to cement clinker to reduce energy consumption. The result shows that replacement of 15 % clinker with LAMM reduces the thermal energy consumption for clinkerisation and simultaneously the CO2 emissions. An optimisation model was proposed by Kookos et al. (2011) to find the minimized cement manufacturing costs with the implementation of co-processing. The study, however, only discusses the possibility of co- processing in cement plant without exploring other mitigation methods. To improve this, additional mitigation method; CCS, is included in the model to study the effects of implementing one or more mitigation methods on various parameters such as production costs and carbon reduction. 296 2. Model Formulation 2.1 Objective function The objective is to find the minimize solution of production cost in €/t clinker that satisfy a CO2 reduction target while maintaining product quality. The cost functions in this problem include capital cost of running a fully functioned cement plant, cost of raw materials, cost of fuels, and the retrofitting cost associated with co-processing, and CCS. The objective function for this is given by       CCSc ccc Nonfossill llll Nonfossill ll Fossilk kk materialsRawj jj VCXFCImVCXFCImCmCmCFCIZ min (1) where FCI is the fixed capital cost in €/t clinker. The model involves the product of raw materials j costs; C j in €/kg of raw materials j, with the mass of raw materials j; mj in kg/t clinker produced. The product produces the cost of raw materials j used in producing t clinker. To find the cost of fuels used, there are two types of fuels involved in cement production which are fossil fuels k and alternative fuels l. Ck and Cl in the model represents cost of fuels k and cost of fuels l in €/kg of fuels while mk and ml represents mass of fuels k and mass of fuels l in kg/t clinker produced. The model also includes the retrofitting cost. Retrofitting cost in this study includes the cost in handling, storage, installation, facilities for alternative fuels l, and retrofitting costs for new CCS technologies c facilities. FCIl and FCIc is the fixed capital investment for non-fossil fuels l in €/t clinker and CCS technologies c in €/t clinker produced while VCl and VCc is the variable cost for non-fossil fuels l in €/kg of fuels and CCS technologies c in €/t clinker. Binary variables Xl and Xc are introduced for selection of fuels and CCS technologies. 2.2 Constraints The objective function is subjected to few constraints. Oxides, alkalis, sulphur and heavy metals produced in kiln came from raw materials and ash from the burning of fuels. They enter the system through raw meals and ash from fuels while sulphurs come through raw meals and the fuels themselves. The constraint is formulated as:    sHeavyMetalh h Sulphurss s Alkalidesa a Oxideso o mmmmproductionclinker (2) where mo, ma, ms and mh represents the total mass of oxides, total mass of alkalis, total mass of sulphurs and total mass of heavy metals kg/t clinker. Oxides in clinker produced are subjected to:    Nonfossill llo Fossilk kko materialsRawj jjoo mmmm ,, ,  (3) where ωo,j is the mass fraction of oxides o in raw materials j, ωo,k is the mass fraction of oxides o in fossil fuels’ ash k and ωo,l is the mass fraction of oxides o in alternative fuels’ ash l. Mass of alkalis and heavy metals are also subjected to the same formula as in Eq(3). Mass fractions are in wt.%. Sulphurs in clinker produced are subjected to the difference between sulphurs produced from raw materials, and fuels with the amount of sulphurs released during combustion as shown below: 332 80 32 80 ,, , SOfg Nonfossill lls Fossilk kks materialsRawj jjss ConcVmmmm                 (4) where ωs,j is the mass fraction of sulphurs s in raw materials j, ωs,k is the mass fraction of sulphurs s in fossil fuels k and ωs,l is the mass fraction of sulphurs s in non-fossil fuels l. The mass fractions of the sulphurs in fuels are calculated by expressing the S wt.% in fuels k and l to SO3 wt.% by introducing 80 kg SO3/mol divided by 32 kg S/mol. Amount of SO3 in flue gas is calculated by the product of flue gas volumetric flow rate, Vfg in Nm 3 /t clinker with concentration of SO3 in flue gas in kg/Nm 3 . Volumetric flowrate of flue gas is calculated by assuming an ideal gas law under normal condition:    GasesFluef fg fg fg mw m V 414.22 (5) where mfg is the mass of flue gas fg in kg. mwfg is the molecular weight of flue gas fg in kg/kmol. 22.414 is the molar volume of an ideal gas under normal condition, in Nm 3 /kmol. 297 Flue gases emitted from cement plant are NO2, CO2, and SO3. The flue gases are formed via fuels combustion (CO2, NO2, and SO3), and chemical reaction from clinkerisation (CO2). Assuming that air fed for combustion consists of 76.8 % N2 and 23.2 % O2, with O2 released is controlled at a level of 10 %, and complete combustion is achieved, constraints gathered are as follows: airN mm  %8.76 2 (6)              Nonfossill ll Fossilk kkairO mStmStmm %2.23 2 (7)  %10 414.22 32 2        fgO Vm (8) where mair represents mass of air fed in kg/t clinker. O2 is calculated as the difference between O2 in fed air in kg/t clinker with O2 used for complete combustion in kg/t clinker. Stk and Stl are the stoichiometric O2 required for complete combustion of fuels k and l. Stk and Stl are calculated as per Green and Perry (2008). As stated before, CO2 emissions are contributed by the clinkerisation process and combustion of fuels. CO2 emissions from fuels are a combination of CO2 emissions from fossil fuels and CO2 emissions from non-fossil fuels. Mass of CO2 emitted from clinkerisation (mcl) and combustion (mcm) are formulated as:    materialsRawj jjMgO materialsRawj jjCaOcl mmm ,, 40 44 55 44  (9)    Nonfossill ll Fossilk kkcm CEFCEFm  (10) cmcl mmCO Total  2 (11) where ωCaO,j is the mass fraction of CaO in raw materials j and ωMgO,j is the mass fraction of MgO in raw materials j. Both 44/55 and 44/40 are introduced since CaO and MgO are used as the basis of the calculation (the same principal used to calculate S in SO3 in Eq(4)). CEFk is the carbon emission factor of fossil fuels k and CEFl is the carbon emission factor of non-fossil fuels l. αk and αl are the product between mk with binary variable Xk, and ml with binary variable Xl. Product of two continuous variables resulted in nonlinear models. In an effort to linearised nonlinear models, each continuous variable (αk and αl) are then subjected to constraints proposed by Adams et al. (2004). Summation of Eq(9) and Eq(10) generates total CO2 emissions (Eq(11)), in kg/t clinker. Heat consumed in kilns is supplied by both fuels where NCVk is the net calorific value of fossil fuels k in GJ/kg of fuels and NCVl is the net calorific value of non-fossil fuels l in GJ/kg of fuels. TED represents the thermal energy demand in GJ/t clinker. This gives:    Nonfossill ll Fossilk kk mNCVmNCVTED (12) The use of alternative fuels is subjected to permits; thermal substitution rate (TSR) as shown:  TEDTSRmNCV Nonfossill ll %  (13) Based on clinker analysis, there are assumed to be 4 major phases (p) in clinker; C3S, C2S, C3A and C4AF formed by oxides from the raw materials: 298 U p Nonfossill lop Fossilk kop materialsRawj jopLp MmBoguemBoguemBogueM    ,,, (14) where MpL; in kg/t clinker represent the lower limits for clinker phases C3S, C2S, C3A and C4AF while Mp U ; in kg/t clinker represent the upper limits. mo,j, mo,k, and mo,l are the mass of oxides in raw materials j, fossil fuels’ ash k and non-fossil fuels’ ash l. Bogue value is obtained from Kookos et al. (2011). Environmental constraint is formulated for total carbon emissions to not be more than proposed carbon emission reduction. This is achieved by:   GHGTo ta l COCOCO CCSc cc 222 %1    (15) where CO2Total is the amount of CO2 emitted gained from Eq(11). %CO2 is the emission reduction target and CO2GHG is current CO2 emission in kg/t clinker. βc is the percent reduction when CCS c is implemented. αc is the linearization variable; product of binary variable XC with CO2Total and subjected to constraints proposed by Adams et al., 2004. Fuels usage must not exceed the availability of the fuels. AXm  (16) where A is the availability of fuels in kg/y. The number of fuels mixed must not exceed the maximum amount of fuels that can be mixed. mixture fuels max lk XX (17) Xk and Xl are the binary variables for the use of fuels. The number of CCS installed must not exceed 1. 1 c X (18) Xc is the binary variable for the use of CCS. All the binaries are subjected to:    otherwise 0 used isy technologif 1 X (19) 2.3 Case study Cement plant with an annual capacity of 1 Mt Ordinary Portland Cement/y plant with 5-stage pre-heater pre-calciner kiln system. Raw materials options are limestone, clay, sand, iron source while fuels options are coal, petroleum coke (PC), refuse derived fuel (RDF), sewage sludge (SS), tire derived fuel (TDF), meat bone meal (MBM). CCS technologies options are post combustion and oxy-fuel. For the base plant, no alternative fuels are used. TSR value of 40 % is used. Data are gathered from Kookos et al. (2011) for chemical analysis of raw materials and fuels and fuels price, U.S. Geological Survey (2012) for clay, sand and iron source prices, Willett (2011) for limestone price, Moya et al. (2010) for plant FCI value, and Barker et al. (2009) for CCS data. The models are executed in GAMS by changing the reduction target. The optimisation stops when infeasible result is achieved indicating maximum possible CO2 reduction achieved. 3. Result and Discussion Results for this study are shown by Table 1 and Table 2. Table 1 includes the base case results (%CO2 is 0 % and TSR 0 %) and later optimisation right before the selection of CCS technologies while Table 2 shows the results of optimisation when CCS technologies selection is involved. Table 1 shows that, at base case; a cement plant with no alternative fuels nor CCS technologies installed, the production cost is € 131.266 /t clinker. The CO2 emitted is 859.984 kg CO2/t clinker, with 325 kg CO2/t clinker and 534.984 kg CO2/t clinker generated from fuels combustion and clinkerisation. Since no CCS is installed, amount of CO2 produced from cement kiln is the same with CO2 released to the atmosphere. 299 Installation of alternative fuels facilities is applied after TSR amount is set reducing the CO2 emission by 3.1 % and simultaneously increasing the production cost; € 129.389 /t clinker. As the reduction target is increase; 4 %; 5 %; 5.9702 %, the minimized costs also increase; € 130.03 /t clinker, € 131.311 /t clinker, and € 136.396 /t clinker. The cost increases because the pricier coals used to replace the higher carbon content PC increase as the CO2 reduction target increase. Large cost gap between 5 % and 5.9702 % (€ 131.311 /t clinker and € 136.396 /t clinker) reduction target as shown in Table 2 is because post combustion CCS is chosen to be installed at 5.9702 %. It can be seen that post combustion CCS is chosen when the amount of fuels needed for cement kiln has reached its maximum; 59.091 kg PC/t clinker (maximum CO2 reduction that can be achieved from co-processing). The cost to achieve CO2 reduction between 5.9702 % and 62.1862 % did not change since the amount of CO2 needed to be reduced is always fulfilled by post combustion CCS capturing abilities. Maximum amount of CO2 that can be captured by post combustion CCS is achieved around 62.1862 %, since from 62.1862 %, 62.2 %, and 63 %, manufacturing costs changes; € 136.4 /t clinker, € 136.425 /t clinker, and € 138.78 /t clinker, where amounts of coal used increase as the reduction increase. At 64 % CO2 reduction, maximum amount of fuels needed and limits for CO2 capturing abilities by post combustion CCS is reached, making oxy-fuel CCS chosen as one of the mitigation strategy. The cost is higher; € 161.852 /t clinker, since higher installation cost is needed for oxy-fuel CCS. Highest amount of CO2 that can be reduced in this study is 75 %, at € 162.43 /t clinker. 4. Conclusions In conclusion, highest emission reduction that can be achieved is 75 % with cost of € 162.43 /t clinker. The cost is higher than base case since in order to maintain the quality of clinker produced while submitting to carbon reduction target, various facilities have to be installed. This study can be further explored by considering the effect of installing CCS to the feeding materials. In this study, the effect from CCS is only in respect to the amount CO2 reduction, while in real practice, installing CCS led to the change of air fed affecting the amount of flue gases emitted from cement kiln. Table 1: Results for base case scenario and after optimisation without CCS technologies selection Only fossil II III IV Reduction target (%CO2) 0 0.031 0.04 0.05 TSR (%) 0 0.4 0.4 0.4 Cost (€/t clinker) 131.266 129.389 130.03 131.311 CO2 produced (kg CO2/t clinker) 859.984 833.324 825.585 816.985 Fossil fuels 325 194.408 186.003 178.75 Nonfossil fuels 0 104 104 104 Clinkerization 534.984 534.916 535.582 534.235 CO2 released (kg CO2/t clinker) 859.984 833.324 825.585 816.985 Raw materials used (kg/t clinker) 1,486.807 1,486.079 1,486.404 1,484.746 Limestone 1,314.526 1,314.461 1,316.352 1,308.336 Clay 0 0 0 29.492 Sand 161.397 160.139 157.422 131.355 Iron source 10.884 11.479 12.63 15.563 Fuels used (kg/t clinker) 98.485 99.931 102.988 105.625 Fossil fuels 98.485 59.306 62.363 65 Coal 0 2.368 35.99 65 Petroleum coke 98.485 56.938 26.373 0 Non fossil fuels 0 40.625 40.625 40.625 TDF 0 40.625 40.625 40.625 300 Table 2: Results for optimisation with CCS technologies selection V VI VII VIII IX X Reduction target (%CO2) 0.059702 0.621862 0.622 0.63 0.64 0.75 TSR (%) 0.4 0.4 0.4 0.4 0.4 0.4 Cost (€/t clinker) 136.396 136.4 136.425 138.78 161.852 162.43 CO2 produced (kg CO2/t clinker) 833.87 833.827 833.523 815.882 833.87 826.908 Fossil fuels 195 194.954 194.624 178.75 195 187.439 Nonfossil fuels 104 104 104 104 104 104 Clinkerisation 534.87 534.873 534.899 533.132 534.87 535.468 CO2 captured (kg CO2/t clinker) 508.661 508.634 508.449 497.688 617.064 611.912 CO2 released (kg CO2/t clinker) 325.209 325.193 325.074 318.194 216.806 214.996 Raw materials used (kg/t clinker) 1,486.055 1,486.057 1,486.07 1,483.635 1,486.055 1,486.348 Limestone 1,314.328 1,314.338 1,314.412 1,302.801 1,314.328 1,316.029 Clay 0 0 0 46.412 0 0 Sand 160.33 160.315 160.209 117.745 160.33 157.886 Iron source 11.397 11.404 11.449 16.677 11.397 12.433 Fuels used (kg/t clinker) 99.716 99.733 99.853 105.625 99.716 102.466 Fossil fuels 59.091 59.108 59.228 65 59.091 61.841 Coal 0 0.184 1.506 65 0 30.243 Petroleum coke 59.091 58.924 57.722 0 59.091 31.598 Non fossil fuels 40.625 40.625 40.625 40.625 40.625 40.625 TDF 40.625 40.625 40.625 40.625 40.625 40.625 Post combustion ◌ ◌ ◌ ◌ ● ● Oxy-fuel combustion ● ● ● ● ◌ ◌ Acknowledgment Special thanks to Universiti Teknologi Malaysia (UTM) and the Ministry of Higher Education of Malaysia. This research is funded by GUP research grant Vot number Q.J130000.2544.06H73. References Adams W.P., Forrester R.J., Glover F.W., 2004, Comparisons and enhancement strategies for linearizing mixed 0-1 quadratic programs, Discrete Optimization, 1(2), 99 – 120. Barker D.J., Turner S.A., Napier-Moore P.A., Clark M., Davison J.E., 2009, CO2 Capture in the Cement Industry, Energy Procedia, 1(1), 87 – 94. Benhelal E., Zahedi G., Shamsaei E., Bahadori A., 2013, Global strategies and potentials to curb CO2 emissions in cement industry, Journal of Cleaner Production, 51, 142 – 161. Gul S., Saeed F., Amin N.U., 2014, Conversion of Low Cost Mineral Material to Pozzolana and its Impact on Cement Parameters and Cost of Production, Chemical Engineering Transactions, 39, 1225 – 1230. Kookos I.K., Pontikes Y., Angelopoulos G.N., Lyberatos G., 2011, Classical and alternative fuel mix optimization in cement production using mathematical programming, Fuel, 90(3), 1277 – 1284. Moya J.A., Pardo N., Mercier A., 2010, Energy Efficiency and CO2 Emissions: Prospective Scenarios for the Cement Industry, Publications Office of the European Union, Luxembourg. Green D.W., Perry R.H., 2008, Perry’s Chemical Engineers’ Handbook, 8th ed, McGraw Hill, NY, USA. U.S. Geological Survey, 2012, Mineral commodity summaries 2012, U.S. Department of the Interior, US Geological Survey, USA. Wennersten R., Sun Q., Li H., 2015, The future potential for Carbon Capture and Storage in climate change mitigation – an overview from perspectives of technology, economy and risk, Journal of Cleaner Production, 103, 724 – 736. Willett J.C., 2011, 2010 Minerals Yearbook: STONE, CRUSHED, US Department of the Interior, US Geological Survey, USA.