Renewable electrified cement production: flexibility and economic competitiveness

Renewable electrified cement production: flexibility and economic competitiveness Sebastian Quevedo Parra, Matteo C. Romano Politecnico di Milano, Department of Energy Supplementary Information Contents S1 Description of the Aspen Plus process model for the operation of the reference cement plant .... 1 Aspen Plus process model ........................................................................................................ 2 Stream table: Reference plant ................................................................................................. 14 S2 Stream table of electrified cement plants ............................................................................... 17 eC – pK plant .......................................................................................................................... 17 OC – HK plant ......................................................................................................................... 19 eC – afK plant ......................................................................................................................... 22 Flexible eC – afK plant ............................................................................................................. 25 Continuous operation ......................................................................................................... 26 Minimum load operation ..................................................................................................... 28 Maximum load operation ..................................................................................................... 30 S3 CO2 compression system for the flexible eC – afK design. ........................................................ 32 S4 MILP optimization model ....................................................................................................... 34 S5 Benchmark cement plants decarbonized with CCS technologies ............................................ 39 Post-combustion capture (PCC) system with amine-based solvents ........................................ 39 Oxyfuel .................................................................................................................................. 40 S6 Solvent-based PCC system Capex in the long-term ................................................................. 42 S7 Summary tables: Electricity Supply Mix .................................................................................. 44 MILP model results – Italy ....................................................................................................... 44 MILP model results – Denmark ................................................................................................ 45 MILP model results – India ...................................................................................................... 46 MILP model results – Egypt ..................................................................................................... 47 S8 Analysis of system behavior ................................................................................................... 48 Short-term – Inflexible plants .................................................................................................. 50 Short-term – Flexible plants .................................................................................................... 51 Long-term – Inflexible plants ................................................................................................... 52 Long-term Flexible plants ....................................................................................................... 53 S8.1 Renewable Energy Generation Profiles ............................................................................. 55 Italy .................................................................................................................................... 55 Denmark ............................................................................................................................ 56 India ................................................................................................................................... 57 Egypt .................................................................................................................................. 58 S8.2 Energy storage ................................................................................................................ 59 Italy .................................................................................................................................... 59 Denmark ............................................................................................................................ 61 India ................................................................................................................................... 63 Egypt .................................................................................................................................. 65 Calcined meal storage SOC ................................................................................................. 67 S8.3 Grid electricity demand ................................................................................................... 69 Italy .................................................................................................................................... 69 Denmark ............................................................................................................................ 70 India ................................................................................................................................... 71 Egypt .................................................................................................................................. 72 S8.5 Renewable electricity curtailed ....................................................................................... 73 Italy .................................................................................................................................... 73 Denmark ............................................................................................................................ 74 India ................................................................................................................................... 75 Egypt .................................................................................................................................. 76 S9 Capex breakdown .................................................................................................................. 77 Italy ....................................................................................................................................... 77 Denmark ................................................................................................................................ 78 India ...................................................................................................................................... 79 Egypt ..................................................................................................................................... 80 S10 Sensitivity of CAC to cost variations of selected parameters .................................................. 81 Short-term horizon ................................................................................................................. 81 Long-term horizon .................................................................................................................. 82 S11 Linear regression model of inflexible and flexible LCOE ......................................................... 83 S12 Additional long-term capex uncertainty for CCS benchmark plants ....................................... 84 References ................................................................................................................................ 85 1 S1 Description of the Aspen Plus process model for the operation of the reference cement plant Figure 1 shows a simplified diagram of the reference process, which was based on the work conducted on the framework of the CEMCAP project (Campanari, et al., 2016). The plant consists of a dry-kiln design, featuring a clinker capacity of approximately 3,000 t/d, which is representative of a European cement plant size. This capacity corresponds to a yearly clinker production of 1 Mt (assuming a run time of >330 days per year) or a cement production of 1.36 Mt per year (using a clinker/cement factor of 0.737). Figure 1: Simplified diagram of the reference cement plant. Five-stage cyclone preheaters are considered the benchmark technique for a dry-process plants. The number of cyclones usually varies between 4 – 6 depending on the raw meal moisture content at the inlet and the plant’s throughput. The raw meal (kiln feed) descends through successive stages for preheating before entering the calciner at temperatures around 750 – 800 °C. Exhaust gas from fossil 2 fuel combustion and process emissions flow counter-currently supplying the heat. Each cyclone stage consists of a connecting duct for material-gas contact and heat transfer, and the cyclone itself for separating raw meal from flue gas using centrifugal force. Calcination, the decomposition of calcium carbonate (CaCO3, main ingredient in the raw meal) into calcium oxide (CaO) and carbon dioxide (CO2), primarily occurs in the calciner, operating as an entrainment reactor using the flue gas from the rotary kiln. Hot tertiary air is diverted to the calciner from the clinker cooler via the tertiary air duct to support fuel combustion. 62% of the total plant fuel input is consumed in the calciner, achieving a calcination level of about 94%. In the rotary kiln, the remaining calcination and formation of clinker phases take place. It represents the central part of the heating process. The kiln is a steel tube, typically 50 – 80 m long with a diameter between 3 – 7 m, inclined between 3 – 4 % towards the burner (kiln’s outlet), and rotating at 1.3 to 3.5 revolutions per minute. It is lined with refractory bricks to withstand the high temperatures, which can reach up to 2,000°C in the gas phase and 1,450°C for the material near the flame. Gas residence time in the kiln is approximately 2 to 4 seconds at temperatures exceeding 1,200°C, while solid material takes 20 to 40 minutes to pass through. Hot clinker exiting the kiln is cooled in a grate cooler using a cross-flow heat exchange mechanism. Cooling air flows through the clinker from below. The cooler generates secondary combustion air, which flows through the kiln hood to the rotary kiln, and tertiary air, which flows via a connection on the hot part of the cooler or kiln hood to the calciner. Moreover, the reference plant has an electricity demand of 131.7 kWh/tclk for auxiliary electric consumption, which includes kiln gear, fans, and conveyors, while 50% is associated to the mill drives (IEAGHG, 2013). It is assumed that plant modifications in the electrified alternatives will have negligible impact on the electricity consumption of these auxiliary components. Aspen Plus process model A simplified representation of the process model developed in Aspen Plus V10, aligned with the CEMCAP reference cement plant, is presented in Figure 2. The model is constructed using modular components that reflect key unit operations of a modern cement facility, allowing for the simulation of complex mass and energy flows. Each module resolves material and energy balances based on predefined thermodynamic methods. In this study, the Peng–Robinson property package was selected, given its suitability for high-temperature systems and the material properties characteristic of clinker production. The purpose of the model is not to reproduce the detailed kinetics or mineralogical evolution of clinker formation, but rather to generate reliable estimates of overall mass and energy balance figures under representative operating conditions. These conditions are derived from literature based on industrial BAT practice (Campanari, et al., 2016; Schorcht, Kourti, Scalet, Roudier, & Delgado Sancho, 2013). Accordingly, essential process parameters, such as cyclone efficiency, degree of calcination, and heat losses, are specified as explicit input parameters. For example, cyclones are modeled as solid-gas separators under the assumption of thermal equilibrium between phases at the outlet. Efficiencies, heat losses, and false air ingress values are imposed externally, as detailed in 3 4 Table 1, with the latter defined for each preheater stage in line with CEMCAP reference data (Campanari, et al., 2016). Since MgCO3 typically decomposes between 700 – 750 °C, a complete conversion is assumed prior to the calciner (Bhatty, Miller, & Kosmatka, 2011). Notably, the model does not account for pressure losses in the preheater stages, in keeping with its simplified scope. Figure 2: Simplified description of the reference plant Aspen model 5 Table 1: Characteristics of the Aspen simulation components for the preheating tower Process Aspen component Specifications Input Cyclone 1 C1: Cyclone – Solid separator Solids separation efficiency 95% Cyclone 2 C2: Cyclone – Solid separator Solids separation efficiency 86% Cyclone 3 C3: Cyclone – Solid separator Solids separation efficiency 86% Cyclone 4 C4: Cyclone – Solid separator Solids separation efficiency 86% Cyclone 5 C5: Cyclone – Solid separator Solids separation efficiency 75% Heat loss HL-1: Heater Duty 19 kJ/kgclk False air ingress (C1 – C4) False air: Material streams % of inlet gas 1.6% False air ingress (C5) False air: Material stream % of inlet gas 0.5% MgCO3 decomposition R-MgO: RStoic reactor Reaction 1 Conversion MgCO3 𝑀𝑔𝐶𝑂3→𝑀𝑔𝑂+𝐶𝑂2 100% The calciner is modeled using two distinct reactor blocks: C-Calciner, which simulates fuel combustion, and R-Calciner, which represents the subsequent chemical reactions, as outlined in Table 3 summarizes the key input parameters used in the rotary kiln process model. The complex interplay of decomposition and sintering reactions within the kiln is simulated with three RStoic reactor blocks arranged in series (R-Kiln), along with a fourth block dedicated to fuel combustion (CKiln). The demand for primary and transport air is met by supplying 1.74 Nm3 of ambient air per kilogram of coal, while 11.98 Nm3 of false air ingress per kilogram of clinker is assumed to enter between the inlet and outlet ends of the rotary kiln. Table 2. Coal is burned using oxygen supplied by the kiln’s flue gas and hot tertiary air from the clinker cooler, accounting for upstream heat losses, assuming complete combustion of the fuel. A small quantity of ambient-temperature primary air is also introduced to support the combustion process. The heat released from fuel oxidation in the C-Calciner is used to drive the reactions occurring in the R-Calciner. The main purpose of the calciner is the decomposition of CaCO3 into CaO and CO2, consuming the majority of the heat demand. A conversion rate of 94% is used to be consistent with the CEMCAP model. In addition, belite (C2S) formation (reaction 4) is expected to initiate at the operating temperatures of the calciner, along with the formation of precursor phases for tetracalcium aluminoferrite (C4AF) (Aïtcin & Flatt, 2015). Reactions 5 and 6 are incorporated to capture the interaction of these intermediate mineral phases and adjust the calciner’s outlet composition in accordance with the reference process, using design specifications summarized in 6 Table 6. Table 3 summarizes the key input parameters used in the rotary kiln process model. The complex interplay of decomposition and sintering reactions within the kiln is simulated with three RStoic reactor blocks arranged in series (R-Kiln), along with a fourth block dedicated to fuel combustion (CKiln). The demand for primary and transport air is met by supplying 1.74 Nm3 of ambient air per kilogram of coal, while 11.98 Nm3 of false air ingress per kilogram of clinker is assumed to enter between the inlet and outlet ends of the rotary kiln. Table 2: Characteristics of the aspen simulation components for the calciner Process Aspen component Specifications Input Fuel combustion C-Calciner: RStoic Reaction 1 Conversion C 𝐶+𝑂2→𝐶𝑂2 100% Reaction 2 Conversion CaCO3 𝐶𝑎𝐶𝑂3→𝐶𝑎𝑂+ 𝐶𝑂2 100% Reaction 3 Conversion S 𝑆+𝑂2→𝑆𝑂2 100% Reaction 4 Conversion H2 𝐻2+0.5𝑂2→𝐻2𝑂 100% Reaction 5 Conversion CO 𝐶𝑂+0.5𝑂2→𝐶𝑂2 100% Reaction 6 Conversion SO2 𝐶𝑎𝑂+𝑆𝑂2+0.5𝑂2→𝐶𝑎𝑆𝑂4 100% Primary air: Material stream Ratio 0.23 Nm3/kgcoal Chemical reactions R-Calciner: RStoic Reaction 1 Conversion CaCO3 𝐶𝑎𝐶𝑂3→𝐶𝑎𝑂+ 𝐶𝑂2 94% Reaction 2 Conversion MgCO3 𝑀𝑔𝐶𝑂3→𝑀𝑔𝑂+ 𝐶𝑂2 100% Reaction 3 Conversion SO2 𝐶𝑎𝑂+𝑆𝑂2+0.5𝑂2→𝐶𝑎𝑆𝑂4 100% Reaction 4 Conversion SiO2 2𝐶𝑎𝑂+𝑆𝑖𝑂2→2𝐶2𝑆 Design spec 4 Reaction 5 Conversion C3S 𝐶3𝑆→𝑆𝑖𝑂2+3𝐶𝑎𝑂 Design spec 5 Reaction 6 Conversion C4AF 𝐶4𝐴𝐹→4𝐶𝑎𝑂+𝐴𝑙2𝑂3+𝐹𝑒2𝑂3 Design spec 6 Heat loss HL-2: Heater Duty 95.6 kJ/kgclk Tertiary air inlet temperature HX-2: Heater Outlet temperature 1,050 °C The sequence of reactor blocks reflects the expected reaction pathway, beginning with the decomposition of residual CaCO3 and MgCO3, and followed by the formation of key clinker phases: alite (C3S), belite (C2S), tricalcium aluminate (C3A), and tetracalcium aluminoferrite (C4AF). Fuel requirements are determined based on the enthalpic demands of the associated chemical 13 achieved by slightly reducing the proportions of C3A and C4AF, while still remaining consistent with typical clinker compositions. Table 9: Comparison of key parameters between the CEMCAP plant and the reference model developed for this work Parameter Unit Modeled plant CEMCAP plant Variation Clinker production t/h 117.1 117.6 -0.4% Clinker composition CAO %wt 0.80 0.80 0 %wt. CASO4 %wt 0.24 0.20 0.04 %wt. SIO2 %wt 0.83 0.00 0.83 %wt. MgO %wt 1.12 1.10 0.02 %wt. C3S %wt 65.39 64.50 0.85 %wt. C2S %wt 14.00 14.00 0 %wt. C3A %wt 8.29 9.80 -1.51%wt. C4AF %wt 9.33 9.60 -0.47 %wt. Fuel consumption MJ/kgclk 3.2 3.2 0.25% Fuel share to rotary kiln % 37.9 38.0 -0.26% Direct CO2 emissions kgCO2/tclk 859.6 863.1 -0.4% The characteristics of the streams from the modeled reference cement plant and the CEMCAP process are provided below. 14 Stream table: Reference plant Figure 3: Simplified diagram of the reference cement plant 15 Table 10: Stream table of the simulated reference cement plant Stream Phase Temp. Mass Flows Mole Flows Molar composition, vol. Mass composition, %wt. °C kg/s kmol/s Ar CO2 H2O N2 O2 CaO CaCO3 CaSO4 SiO2 Al2O3 Fe2O3 MgCO3 MgO C3S C2S C3A C4AF 1 Moisture 60 0.2 0.01 0% 0% 100% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 1 Solid 60 54.2 0.59 0% 0% 0% 0% 0% 0% 0% 79.31% 0% 13.81% 3.34% 2.01% 1.53% 0% 0% 0% 0% 2 Gas 309 67.0 2.05 0.71% 31.04% 5% 59.47% 3.79% 0.00% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 2 Solid 309 3.1 0.03 0% 0% 0% 0% 0% 0% 0.04% 79.26% 0% 13.81% 3.34% 2.02% 1.49% 0.02% 0% 0% 0% 3 Solid 741 63.0 0.73 0% 0% 0% 0% 0% 0% 9.47% 65.65% 0.03% 14.21% 3.59% 2.17% 0.00% 0.81% 0.87% 2.90% 0.15% 4 Solid 309 61.0 0.66 0% 0% 0% 0% 0% 0% 0.04% 79.26% 0% 13.81% 3.34% 2.02% 1.49% 0.02% 0% 0.01% 0% 5 Solid 477 61.1 0.67 0% 0% 0% 0% 0% 0% 0.26% 79.01% 0% 13.84% 3.35% 2.02% 1.28% 0.12% 0.02% 0.08% 0% 6 Solid 613 61.2 0.68 0% 0% 0% 0% 0% 0% 1.62% 77.50% 0.01% 13.97% 3.40% 2.05% 0.00% 0.75% 0.15% 0.50% 0.03% 7 Gas 15 0.7 0.02 0.92% 0.03% 1.03% 77.28% 20.73% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 8 Gas 1,078 16.9 0.55 0.85% 19.76% 6.29% 71.20% 1.91% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 8 Solid 1,078 2.9 0.01 0% 0% 0% 0% 0% 0% 0.80% 0% 0.24% 0.83% 0% 0% 0% 1.12% 65.39% 14.00% 8.29% 9 Gas 860 62.8 1.90 0.70% 32.89% 4.85% 58.86% 2.71% 0.00% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 9 Solid 860 12.3 0.17 0% 0% 0% 0% 0% 0% 48.64% 6.50% 0.16% 15.41% 4.52% 2.75% 0% 1.09% 4.47% 14.91% 0.78% 10 Gas 741 63.7 1.93 0.70% 32.37% 4.79% 59.15% 2.99% 0.00% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 10 Solid 741 10.5 0.12 0% 0% 0% 0% 0% 0% 9.47% 65.65% 0.03% 14.21% 3.59% 2.17% 0.00% 0.81% 0.87% 2.90% 0.15% 11 Gas 613 65.0 1.97 0.70% 32.18% 4.70% 59.15% 3.26% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 11 Solid 613 10.0 0.11 0% 0% 0% 0% 0% 0% 1.62% 77.50% 0.01% 13.97% 3.40% 2.05% 0% 0.75% 0.15% 0.50% 0.03% 12 Gas 477 65.9 2.01 0.71% 31.67% 4.65% 59.44% 3.53% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 12 Solid 477 9.9 0.11 0% 0% 0% 0% 0% 0% 0.26% 79.01% 0% 13.84% 3.35% 2.02% 1.28% 0.12% 0.02% 0.08% 0% 13 Fuel 60 2.4 0.20 Type of fuel: Coal, LHV: 27.15 MJ/kg, Mass composition: C= 69.00%, H= 4.00%, N=0.48%, S= 0.50%, O=9.00%, Moi= 0.5%, Ash= 16.5% 14 Gas 860 62.8 1.90 0.70% 32.89% 4.85% 58.86% 2.71% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 14 Solid 860 50.2 0.70 0% 0% 0% 0% 0% 0% 48.64% 6.50% 0.16% 15.41% 4.52% 2.75% 0% 1.09% 4.47% 14.91% 0.78% 15 Solid 860 38.0 0.53 0% 0% 0% 0% 0% 0% 48.64% 6.50% 0.16% 15.41% 4.52% 2.75% 0% 1.09% 4.47% 14.91% 0.78% 16 Fuel 60 1.5 0.12 Type of fuel: Coal, LHV: 27.15 MJ/kg, Mass composition: C= 69.00%, H= 4.00%, N=0.48%, S= 0.50%, O=9.00%, Moi= 0.5%, Ash= 16.5% 17 Gas 15 73.7 2.55 0.92% 0.03% 1.03% 77.28% 20.73% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 18 Gas 1,137 26.2 0.91 0.92% 0.03% 1.03% 77.28% 20.73% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 0% 18 Solid 1,137 0.7 0.00 0% 0% 0% 0% 0% 0% 0.80% 0.00% 0.24% 0.83% 0.00% 0.00% 0.00% 1.12% 65.39% 14.00% 8.29% 19 Gas 1,137 10.8 0.37 0.92% 0.03% 1.03% 77.28% 20.73% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 19 Solid 1,137 0.3 0.00 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.80% 0.00% 0.24% 0.83% 0.00% 0.00% 0.00% 1.12% 65.39% 14.00% 8.29% 20 Gas 15 3.3 0.11 0.92% 0.03% 1.03% 77.28% 20.73% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 20 Solid 115 32.5 0.15 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.80% 0.00% 0.24% 0.83% 0.00% 0.00% 0.00% 1.12% 65.39% 14.00% 8.29% 21 Solid 300 36.6 1.27 0.92% 0.03% 1.03% 77.28% 20.73% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 22 Gas 300 1.0 0.00 0.00% 0.00% 0.00% 0.00% 0.00% 0.00% 0.80% 0.00% 0.24% 0.83% 0.00% 0.00% 0.00% 1.12% 65.39% 14.00% 8.29% 16 Table 11: Stream table of the CEMCAP reference cement plant (Campanari, et al., 2016) 17 S2 Stream table of electrified cement plants eC – pK plant Figure 4: Simplified diagram of the eC - pK plant - direct electrification of the calciner via resistive elements or inductive heating + plasma burners in the rotary kiln. 18 Table 12: Stream table of the eC - pK plant Stream Phase Temp. Mass Flows Mole Flows Molar composition, vol. Mass composition, %wt. °C kg/h kmol/h Ar CO2 H2O N2 O2 CaO CaCO3 CaSO4 SiO2 Al2O3 Fe2O3 MgCO3 MgO C3S C2S C3A C4AF 1 Moisture 60 566 31 - - 100% - - - - - - - - - - - - - - 1 Solid 60 201,080 2,183 - - - - - - 79.31% - 13.81% 3.34% 2.01% 1.53% - - - - - 2 Moisture 60 107 6 - - 100% - - - - - - - - - - - - - - 2 Solid 60 38,205 415 - - - - - - 79.31% - 13.81% 3.34% 2.01% 1.53% - - - - - 3 Moisture 60 458 25 - - 100% - - - - - - - - - - - - - - 3 Solid 60 162,875 1,768 - - - - - - 79.31% - 13.81% 3.34% 2.01% 1.53% - - - - - 4 Gas 442 63,004 1,440 0.01% 98.62% 0.42% 0.75% 0.20% - - - - - - - - - - - - 4 Solid 442 2,162 24 - - - - - 0.04% 79.34% - 13.83% 3.34% 2.02% 1.30% 0.11% 0.00% 0.01% 0.00% 0.00% 5 Gas 283 133,888 4,588 0.89% 2.60% 1.56% 74.87% 20.09% - - - - - - - - - - - - 5 Solid 283 9,217 100 - - - - - 0.01% 79.27% - 13.80% 3.34% 2.01% 1.30% 0.11% 0.10% 0.02% 0.01% 0.02% 6 Solid 757 37,145 411 - - - - - 1.86% 77.19% - 13.94% 3.41% 2.06% - 0.75% 0.17% 0.56% 0.03% 0.03% 7 Solid 750 164,331 1,728 - - - - - 0.28% 74.16% - 12.91% 3.12% 1.88% - 0.77% 4.56% 1.01% 0.61% 0.69% 8 Solid 920 137,879 1,918 - - - - - 49.61% 6.51% - 14.46% 4.61% 2.79% - 1.11% 4.48% 14.93% 0.73% 0.77% 9 Gas 865 62,408 1,420 0.00% 99.50% 0.00% 0.38% 0.10% - - - - - - - - - - - - 9 Solid 865 1,393 19 - - - - - 49.61% 6.51% - 14.46% 4.61% 2.79% - 1.11% 4.48% 14.93% 0.73% 0.77% 10 Gas 1,139 58,355 1,976 0.88% 4.57% 0.99% 73.77% 19.79% - - - - - - - - - - - - 10 Solid 1,139 9,997 51 - - - - - 3.86% - - - - - - 1.14% 63.01% 14.00% 8.46% 9.53% 11 Gas 15 39,026 1,353 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 12 Gas 2,781 54,407 1,886 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 13 Gas 1,137 13,575 471 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 13 Solid 1,137 337 2 - - - - - 3.86% - - - - - - 1.14% 63.01% 14.00% 8.46% 9.53% 14 Gas 1,137 118,465 4,106 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 14 Solid 1,137 3,102 16 - - - - - 3.86% - - - - - - 1.14% 63.01% 14.00% 8.46% 9.53% 15 Gas 300 133,205 4,617 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 15 Solid 300 3,752 19 - - - - - 3.86% - - - - - - 1.14% 63.01% 14.00% 8.46% 9.53% 16 Gas 1,050 73,207 2,537 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 16 Solid 1,050 1,917 10 - - - - - 3.86% - - - - - - 1.14% 63.01% 14.00% 8.46% 9.53% 17 Solid 112 117,080 592 - - - - - 3.86% - - - - - - 1.14% 63.01% 14.00% 8.46% 9.53% 19 OC – HK plant Figure 5: Simplified diagram of the OC - HK plant - Oxyfuel combustion of alternative fuels in calciner + indirect electrification via H2 combustion in the rotary kiln. O2 supply from ASU is preheated to 150 °C before mixing with O2 sourced from the SOEC using available heat from the hot CO2 exit stream. 20 Table 13: Stream table of the OC - HK plant Stream Phase Temp. Mass Flows Mole Flows Molar composition, %vol. Mass composition, %wt. °C kg/h kmol/h Ar CO2 H2O N2 O2 H2 CaO CaCO3 CaSO4 SiO2 Al2O3 Fe2O3 MgCO3 MgO C3S C2S C3A C4AF 1 Moisture 60 566 31 - - 100% - - - - - - - - - - - - - - - 1 Solid 60 199,652 2,168 - - - - - - - 79.31% - 13.81% 3.34% 2.01% 1.53% - - - - - 2 Moisture 60 328 18 - - 100% - - - - - - - - - - - - - - - 2 Solid 60 115,778 1,257 - - - - - - - 79.31% - 13.81% 3.34% 2.01% 1.53% - - - - - 3 Moisture 60 238 13 - - 100% - - - - - - - - - - - - - - - 3 Solid 60 83,874 911 - - - - - - - 79.31% - 13.81% 3.34% 2.01% 1.53% - - - - - 4 Gas 433 147,275 3,983 0.04% 69.40% 22.57% 5.24% 2.75% - - - - - - - - - - - - - 4 Solid 433 6,597 72 - - - - - - 0.41% 78.79% - 13.84% 3.35% 2.02% 1.29% 0.12% 0.04% 0.13% - 0.01% 5 Gas 265 52,962 2,013 0.73% 5.15% 27.92% 61.67% 4.51% - - - - - - - - - - - - - 5 Solid 265 4,750 52 - - - - - - 0.00% 79.21% - 13.79% 3.34% 2.01% 1.30% 0.11% 0.16% 0.03% 0.02% 0.02% 6 Solid 807 152,169 1,798 - - - - - - 14.19% 58.81% 0.01% 14.29% 3.74% 2.25% - 0.84% 1.22% 4.32% 0.16% 0.18% 7 Solid 750 87,378 901 - - - - - - 0.14% 71.82% 0.00% 12.51% 3.02% 1.82% - 0.78% 6.67% 1.42% 0.85% 0.96% 8 Solid 920 135,851 1,897 - - - - - - 49.29% 6.54% 0.03% 15.19% 4.66% 2.81% - 1.10% 4.24% 14.99% 0.55% 0.62% 9 Gas 920 145,520 3,926 0.04% 69.90% 22.43% 4.95% 2.69% - - - - - - - - - - - - - 9 Solid 920 179,720 2,510 - - - - - - 49.29% 6.54% 0.03% 15.19% 4.66% 2.81% - 1.10% 4.24% 14.99% 0.55% 0.62% 10 Gas 433 28,244 764 0.04% 69.40% 22.57% 5.24% 2.75% - - - - - - - - - - - - - 10 Solid 433 1,265 14 - - - - - - 0.41% 78.79% - 13.84% 3.35% 2.02% 1.29% 0.12% 0.04% 0.13% - 0.01% 11 Solid 60 19,539 1,678 Type of fuel: Alternative fuel, LHV: 17.8 MJ/kg, Mass composition: C= 49.2%, H= 5.6%, N= 3.1%, S= 0.05%, O= 21.3%, Moi= 15.1%, Ash= 5.6% 12 Gas 15 25,344 797 - - - 4.70% 95.30% - - - - - - - - - - - - - 13 Gas 1,139 51,811 1,975 0.74% 4.52% 27.78% 62.46% 4.50% - - - - - - - - - - - - - 13 Solid 1,139 8,892 42 - - - - - - 1.34% - 0.03% - - - - 1.13% 65.67% 14.01% 8.38% 9.44% 14 Gas 1,137 44,262 1,534 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - - 14 Solid 1,137 1,099 5 - - - - - - 1.34% - 0.03% - - - - 1.13% 65.67% 14.01% 8.38% 9.44% 15 Gas 1,137 88,292 3,060 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - - 15 Solid 1,137 2,312 11 - - - - - - 1.34% - 0.03% - - - - 1.13% 65.67% 14.01% 8.38% 9.44% 16 Gas 300 132,692 4,599 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - - 16 Solid 300 3,737 18 - - - - - - 1.34% - 0.03% - - - - 1.13% 65.67% 14.01% 8.38% 9.44% 21 17 Gas 1,050 88,292 3,060 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - - 17 Solid 1,050 2,312 11 - - - - - - 1.34% - 0.03% - - - - 1.13% 65.67% 14.01% 8.38% 9.44% 18 Gas 681 88,292 3,060 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - - 18 Solid 681 2,312 11 - - - - - - 1.34% - 0.03% - - - - 1.13% 65.67% 14.01% 8.38% 9.44% 19 Gas 750 9,599 533 - - 100% - - - - - - - - - - - - - - - 20 Gas 750 1,842 533 - - 9.01% - - 90.99% - - - - - - - - - - - - 21 Gas 750 7,756 242 - - - - 100% - - - - - - - - - - - - - 22 Solid 114 117,004 552 - - - - - - 1.34% - 0.03% - - - - 1.13% 65.67% 14.01% 8.38% 9.44% 22 eC – afK plant Figure 6: Simplified diagram of the eC - afK plant – Direct electrification of the calciner via resistive elements or inductive heating + combustion of alternative fuels in the rotary kiln. Waste heat available from tertiary air, vent air and the hot CO2 exit stream is harnessed for steam production, to supply the majority of the regeneration heat for the solvent in the PCC system. Additional heat requirement in the PCC unit is supplied with the help of a heat pump with COP = 2. 29 18 Gas 1,050 13,234 459 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 18 Solid 1,050 347 2 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% 19 Gas 1,050 66,971 2,321 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 19 Solid 1,050 1,754 8 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% 20 Gas 300 132,910 4,607 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 20 Solid 300 3,744 18 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% 21 Gas 563 176,952 6,133 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 21 Solid 563 4,867 23 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% 22 Gas 563 22,929 795 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 22 Solid 563 631 3 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% 23 Solid 113 117,055 558 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% Heat Streams Heat MWth Q1 0 Q2 23.3 30 Maximum load operation Table 17: Stream table of the flexible eC - afK plant under maximum load operation. Stream Phase Temp. Mass Flows Mole Flows Molar composition, %vol. Mass composition, %wt. °C kg/h kmol/h Ar CO2 H2O N2 O2 CaO CaCO3 CaSO4 SiO2 Al2O3 Fe2O3 MgCO3 MgO C3S C2S C3A C4AF 1 Moisture 60 896 50 - - 100% - - - - - - - - - - - - - - 1 Solid 60 318,398 3,457 - - - - - - 79.31% - 13.81% 3.34% 2.01% 1.53% - - - - - 2 Moisture 60 654 36 - - 100% - - - - - - - - - - - - - - 2 Solid 60 232,431 2,524 - - - - - - 79.31% - 13.81% 3.34% 2.01% 1.53% - - - - - 3 Moisture 60 242 13 - - 100% - - - - - - - - - - - - - - 3 Solid 60 85,968 933 - - - - - - 79.31% - 13.81% 3.34% 2.01% 1.53% - - - - - 4 Gas 234 239,409 8,291 0.91% 0.51% 1.46% 76.57% 20.54% - - - - - - - - - - - - 4 Solid 234 13,193 143 - - - - - 0.00% 79.31% - 13.81% 3.34% 2.01% 1.49% 0.02% - 0.00% - - 5 Gas 273 98,352 2,251 0.01% 98.29% 0.61% 0.86% 0.23% - - - - - - - - - - - - 5 Solid 273 4,879 53 - - - - - 0.00% 79.40% - 13.83% 3.34% 2.02% 1.30% 0.11% - 0.00% - - 6 Solid 640 223,648 2,465 - - - - - 1.32% 77.96% - 13.94% 3.40% 2.05% - 0.75% 0.20% 0.32% 0.03% 0.03% 7 Solid 677 82,528 911 - - - - - 1.36% 78.14% - 13.99% 3.41% 2.05% - 0.75% 0.00% 0.30% - - 8 Solid 920 207,357 3,074 - - - - - 53.24% 8.95% - 16.41% 4.94% 2.98% - 1.09% 0.53% 11.69% 0.09% 0.09% 9 Gas 15 209,973 7,278 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 10 Gas 750 209,973 7,278 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 10 Solid 750 5,498 81 - - - - - 53.24% 8.95% - 16.41% 4.94% 2.98% - 1.09% 0.53% 11.69% 0.09% 0.09% 11 Solid 100 201,859 2,992 - - - - - 53.24% 8.95% - 16.41% 4.94% 2.98% - 1.09% 0.53% 11.69% 0.09% 0.09% 12 Solid 60 134,991 1,993 - - - - - 53.23% 8.95% - 16.40% 4.93% 2.98% - 1.09% 0.52% 11.72% 0.09% 0.09% 13 Gas 326 91,910 3,099 0.80% 17.22% 12.84% 67.15% 1.99% - - - - - - - - - - - - 13 Solid 326 7,657 113 - - - - - 53.10% 8.93% - 16.36% 4.92% 2.97% - 1.09% 0.69% 11.72% 0.11% 0.11% 14 Gas 835 143,060 1,955 - - - - - 47.57% 7.97% - 14.60% 4.39% 2.65% - 1.10% 7.94% 11.66% 1.00% 1.12% 15 Gas 1,139 91,482 3,084 0.79% 17.31% 12.90% 67.10% 1.90% - - - - - - - - - - - - 15 Solid 1,139 15,716 74 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% 16 Fuel 60 10,231 878 Type of fuel: Alternative fuel, LHV: 17.8 MJ/kg, Mass composition: C= 49.2%, H= 5.6%, N= 3.1%, S= 0.05%, O= 21.3%, Moi= 15.1%, Ash= 5.6% 17 Gas 15 265,246 9,194 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 31 18 Gas 1,050 25,552 886 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 18 Solid 1,050 669 3 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% 19 Gas 1,050 54,653 1,894 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 19 Solid 1,050 1,431 7 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% 20 Gas 300 132,910 4,607 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 20 Solid 300 3,744 18 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% 21 Gas 0 0 0 - - - - - - - - - - - - - - - - - 21 Solid 0 0 0 - - - - - - - - - - - - - - - - - 22 Gas 529 187,562 6,501 0.92% 0.03% 1.03% 77.28% 20.73% - - - - - - - - - - - - 22 Solid 529 5,175 25 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% 23 Solid 113 117,055 558 - - - - - 1.61% - 0.02% - - - - 1.13% 68.13% 11.21% 8.42% 9.48% Heat Streams Heat MWth Q1 6.4 Q2 23.3 32 S3 CO2 compression system for the flexible eC – afK design. To cope with the wide flowrate range in the flexible design, a conceptual system architecture for the CO2 compression train is proposed. Specifically, a single four-stage integrally geared compressor train (based on the “moderate purity” system described by (Magli, Spinelli, Fantini, Romano, & Gatti, 2022)) could be adapted to accommodate the highly variable inlet volumetric flowrate of CO2-rich gas (ranging from 6,680 Nm³/h at minimum load to 50,454 Nm³/h at full load, i.e., a 7.5:1 turndown ratio). This variation far exceeds the conventional 75 – 105% operational window of a centrifugal compressor equipped only with inlet guide vanes (IGVs) (Atlas Copco, 2024). To enable stable and efficient operation across this range without duplicating hardware, the design incorporates (i) a fullbore bypass loop around stage 1 and (ii) an extended-range variable-speed drive. Above ~45% load, all four stages operate in series, with the shaft speed proportionally reduced to maintain stable operation while preserving ≥10% surge margin. Below this threshold, the bypass valve disables stage 1, and the remaining stages are accelerated to maintain discharge pressure, leveraging a scaling relationship determined by 𝑁𝑛𝑒𝑤 𝑁𝑜𝑙𝑑 ~√𝐻𝑛𝑒𝑤 𝐻𝑜𝑙𝑑 =√4 3=1.15 (assuming quasi-constant head coefficient). At minimum flowrate, the relative volumetric flow into stage 2 drops to ~30% of its nominal value. To extend stable operation, a small warm-gas recycle loop can be introduced to increase the effective inlet flowrate and move the operating point away from the surge boundary. In parallel, a modest reduction in the target pressure ratio at low loads can further widen the surge margin. These measures, in combination, maintain operational safety while still delivering sufficient outlet pressure for downstream equipment. For instance, a reduction in discharge pressure to ~18 – 19 bar remains compatible with the design of the second compression stage described in (Magli, Spinelli, Fantini, Romano, & Gatti, 2022), which raises pressure to 42.9 bar. Although the flexible turndown strategy combining bypass and speed control is introduced as a technically viable approach, the detailed design and control optimization of such features were excluded from the techno-economic study. Moreover, it is worth highlighting that the 20 – 150% throughput limits adopted in this study are illustrative and were chosen arbitrarily. They could be refined based on a dedicated optimization of the CO₂ compression system. Once the gas exits the CPU at ≥37 bar, it enters the dense-phase region, where its specific volume changes much less with pressure and temperature. As a result, the volumetric variation due to flexible operation is significantly reduced downstream of the CPU. The remainder of the compression train consisting of a low-pressure compressor (17 → 37 bar), a high-pressure centrifugal stage (37 → 74 bar), and a supercritical CO2 pump (74 → 110 bar) is sufficient to handle the full operating range without approaching surge, choke, or stonewall conditions. In sum, the proposed solution offers a cost-effective and technically sound pathway for continuous, surge-free operation from 13% to 100% of design throughput, relying on a bypass valve, upgraded speed control, and recycle if needed. It avoids the complexity and capital cost of parallel trains while 33 fulfilling the dynamic performance requirements of the flexible, partially electrified eC – afK configuration. 34 S4 MILP optimization model A Mixed Integer Linear Programming (MILP) model was developed in GAMS to describe the mass and energy interactions of the flexible, partially electrified eC – afK plant with an hourly resolution for one year, with the goal to minimize the total annual cost (TAC) of clinker production while optimizing the renewable energy generation portfolio (including storage mechanisms). TAC is calculated based on equation (1): 𝑇𝐴𝐶= 𝑇𝑃𝐶∗ 𝐶𝑅𝐹+𝑓𝑂𝑃𝐸𝑋+𝑣𝑂𝑃𝐸𝑋 (1) Where TPC is the capex calculated following the methodology described in the main body, CRF is the capital recovery factor (as described in equation 5 of the main body), fOPEX are the corresponding fixed costs associated with the new capital investments, and vOPEX represents the operating expenditures in addition to the operation of the conventional process. To describe the behavior of the plant, linearized mass & energy balance equations were derived from the outcomes of the Aspen Plus process models for the flexible configuration under the representative operating modes (minimum, continuous, and maximum load). Figure 8 shows a simplified diagram of the flexible, partially electrified eC – afK plant, highlighting the name of the material and energy streams utilized in the linearized equations. The model is based on a constant clinker production of CLK(t) = CLK = 117 tclk/h. Hence, material balances in the clinker production section, excluding waste heat integration streams of tertiary and vent air, can all be described as a function of CLK, i.e., as constants: 𝑥𝑖(𝑡)= 𝑓𝑖(𝐶𝐿𝐾), 𝑥𝑖(𝑡)={𝑆𝑀1(𝑡),𝑆𝑀2(𝑡),𝐹𝐺1(𝑡),𝐹𝐺2(𝑡),𝐹𝐺3(𝑡),𝐴𝐹(𝑡) 𝑆𝑡(𝑡),𝐶𝑜𝑜𝑙𝐴𝑖𝑟2(𝑡),2𝐴(𝑡),3𝐴(𝑡),𝑉𝐴(𝑡)} (2) The constant steam demand in the solvent-based PCC system (QPCC(t)) is met harnessing waste available heat from the outlet streams of the air (HR1(t)) and calciner strings (HR2(t)), the kiln string (HR3(t)), a mixture of tertiary and vent air after meeting Raw Mill requirements (HR4(t)), and a heat pump (HP(t)). For simplicity, waste heat available from the system besides the heat pump was linearized as a function of the calcined meal production (thereby taking into account variations in the available heat from hot air and CO2). Equations (3) – (6) summarize the interactions governing the heat supply to the PCC system: 𝑄𝑃𝐶𝐶(𝑡)=𝑄𝑃𝐶𝐶 =23.3 [𝑀𝑊𝑡ℎ] (3) 𝑄𝑊𝐻(𝑡)=−0.085∗𝐶𝑀1(𝑡)+32.8 (4) 𝑄𝐻𝑃(𝑡)=𝑄𝑃𝐶𝐶 −𝑄𝑊𝐻(𝑡) (5) 𝐸𝐻𝑃(𝑡)=𝑄𝐻𝑃(𝑡) 𝐶𝑂𝑃 ⁄ (6) Where QPCC(t) is the heat demand for solvent regeneration, QWH(t) is the waste heat available at any given time t already discounting the Raw Mill heating needs, CM1(t) is the mass flowrate of calcined material exiting the e-calciner in tCM/h (constrained by the minimum and maximum load operating 35 conditions – equations (11) and (12)), QHP(t) is the heat demand from the heat pump at any given time t. EHP(t) corresponds to the hourly electricity demand from the heat pump, which is estimated by dividing the heat demand by an assumed COP of 2. Figure 8: Simplified diagram of the flexible, partially electrified eC - afK configuration highlighting mass & energy streams used to build the MILP model. The size of the heat pump is a decision variable, calculated by ensuring that the heat demand stays below the heat pump capacity (𝐻𝑃𝑚𝑎𝑥) at every moment: 𝑄𝐻𝑃(𝑡)≤𝐻𝑃𝑚𝑎𝑥 (7) Meanwhile, mass & energy streams in the calcined meal production section will vary depending on the availability of renewable resources. Raw meal consumption (RM(t)), CO2 emissions in the calciner string (CO2-2(t)), and in the air string (CO2 in HotAir2(t)) are all expressed as a function of calcined meal production (CM1(t)): 𝐶𝑀1(𝑡)=0.65∗𝑅𝑀(𝑡)+0.64 (8) 36 𝐶𝑂2−2(𝑡)=0.47∗𝐶𝑀1(𝑡)−0.32 (9) 𝐶𝑂2 𝐴(𝑡)=0.009∗𝐶𝑀1(𝑡)+0.017 (10) 𝐶𝑀1(𝑡)≤𝐶𝑀𝑚𝑎𝑥 (11) 𝐶𝑀1(𝑡)≥𝐶𝑀𝑚𝑖𝑛 (12) Where 𝐶𝑂2 𝐴 (t) denotes the small share of CO2 present in the air string outlet, and CMmax and CMmin represent the boundaries corresponding to the 20% and 150% of inflexible design throughput. The CPU after the calciner section is assumed to capture 100% of the CO2-rich stream (CO2-3(t)). Most of CO2 emissions in the kiln section arise from alternative fuels combustion, while a share is emitted from the remaining CaCO3 calcination. In addition, CO2 emissions from alternative fuels combustion are divided between biogenic and fossil origin. The relationships describing the constant behavior of CO2 emissions in the kiln string are summarized below: 𝐶𝑂2 𝐹𝐺(𝑡)=𝐶𝑂2 𝐹𝐺 =0.2∗𝐶𝐿𝐾 (13) 𝐶𝑂2 𝐴𝐹(𝑡)=𝐶𝑂2 𝐴𝐹 =0.10∗𝐴𝐹 (14) 𝐶𝑂2 𝑛𝑒𝑢𝑡𝑟𝑎𝑙 =𝐶𝑏𝑖𝑜∗𝐶𝑂2 𝐴𝐹 (15) 𝐶𝑂2 𝑓𝑜𝑠𝑠𝑖𝑙 =(1−𝐶𝑏𝑖𝑜)∗𝐶𝑂2 𝐴𝐹 (16) Where the term AF corresponds to the alternative fuel demand to meet the constant heat requirement in the rotary kiln, estimated from Aspen process model to be 50.6 MWth, while Cbio represents the share of biogenic content associated to the alternative fuel consumption, assumed to be 30%. Direct CO2 emissions which arise from the uncaptured CO2 in the PCC system and the air string can be computed as: 𝐶𝑂2−𝑑𝑖𝑟𝑒𝑐𝑡=(1−𝐶𝑃𝐶𝐶)∗𝐶𝑂2 𝐹𝐺 +𝐶𝑂2 𝐴−𝐶𝑂2 𝑛𝑒𝑢𝑡𝑟𝑎𝑙 (17) Where CPCC corresponds to the capture rate in the PCC unit, equal to 95%. The mass balance between the two sections (calcined meal and clinker production) is reconciled by solving the state of charge equation of the calcined meal storage unit, defined by equation (18). The size of calcined meal storage (𝑆𝑂𝐶𝐶𝑀𝑚𝑎𝑥) is a decision variable used to optimize the portfolio of the system, without any a priori size constraint. Moreover, equation (20) was included to optimize the initial storage (SOCCM0), forcing the system to be balanced by the end of the year. 𝑆𝑂𝐶𝐶𝑀 ={𝑆𝑂𝐶𝐶𝑀0 , 𝑡=1 𝑆𝑂𝐶𝐶𝑀(𝑡−1)+𝐶𝑀2(𝑡)−𝑆𝑀1(𝑡), 𝑡≠1 (18) 𝑆𝑂𝐶𝐶𝑀 ≤𝑆𝑂𝐶𝐶𝑀𝑚𝑎𝑥 (19) 𝑆𝑂𝐶𝐶𝑀(𝑡=8760)+𝐶𝑀2(𝑡=1)−𝑆𝑀1(𝑡=1)=𝑆𝑂𝐶𝐶𝑀0 (20) The electricity balance besides the heat pump is divided between the constant demand from the solvent-based PCC system (including auxiliaries and CO2 compression), calculated as a function of 37 the captured CO2 (equation (21)), and the variable demand from the e-calciner and the CPU (equations (22) and (23)). Electricity from auxiliary equipment is assumed equal to the reference process. 𝐸𝑃𝐶𝐶 =0.14∗𝐶𝑂2 𝐹𝐺 (21) 𝐸𝑒−𝑐𝑎𝑙𝑐 =0.61∗𝐶𝑀2(𝑡)−0.12 (22) 𝐸𝐶𝑃𝑈 =0.05∗𝐶𝑀2(𝑡) (23) This electricity demand is met with a combination of renewable sources (PV panels: EPV(t), Wind Turbines: EWT(t)), and the grid (Grid(t)). Excess renewable can be stored in a battery energy storage system (BESS), if installed. When the storage capacity is full, surplus renewable production is curtailed (Ew(t)), i.e., no economic benefit from selling electricity back to the grid is considered. The equations governing electricity demand and supply interactions are described below. The SOC of the BESS is calculated considering a self-discharge rate of Cdeg = 0.054%/h, and charging (ηch) and discharging (ηdc) efficiencies of 92.7%. The storage system is assumed to start the yearly simulation with the capacity depleted (i.e., SOCb0 = 0). Moreover, the balances are calculated assuming a Depth of Discharge (DoD) of 90% and an energy to power ratio (CE/P) of 4 hours. 𝐸𝑃𝑉(𝑡)+𝐸𝑊𝑇(𝑡)= 𝑃𝑉∗𝐸𝑃𝑉 1𝑀𝑊(𝑡)+𝑊𝑇∗𝐸𝑊𝑇 1𝑀𝑊(𝑡)=𝐸𝑅1(𝑡) (24) 𝐸𝑅1(𝑡)=𝐸𝑅2(𝑡)+𝑏𝑖𝑛(𝑡)+𝐸𝑤(𝑡) (25) 𝐸𝑅2(𝑡)+𝑏𝑜𝑢𝑡(𝑡)+𝐺𝑟𝑖𝑑(𝑡)=𝐸𝑡𝑜𝑡𝑎𝑙(𝑡) (26) 𝐸𝑡𝑜𝑡𝑎𝑙(𝑡)=𝐸𝑒−𝑐𝑎𝑙𝑐(𝑡)+𝐸𝐶𝑃𝑈(𝑡)+𝐸𝑃𝐶𝐶(𝑡)+𝐸𝐻𝑃(𝑡) (27) 𝑆𝑂𝐶𝑏(𝑡)={𝑆𝑂𝐶𝑏0 , 𝑡=1 𝑆𝑂𝐶𝑏(𝑡−1)∗(1−𝐶𝑑𝑒𝑔)+𝑏𝑖𝑛(𝑡)∗𝜂𝑐ℎ−𝑏𝑜𝑢𝑡(𝑡)𝜂𝑑𝑐 ⁄, 𝑡≠1 (28) 𝑆𝑂𝐶𝑏(𝑡=8760)∗(1−𝐶𝑑𝑒𝑔) +𝑏𝑖𝑛(𝑡=1)∗𝜂𝑐ℎ −𝑏𝑜𝑢𝑡(𝑡=1) 𝜂𝑑𝑐 ⁄=𝑆𝑂𝐶𝑏0 (29) 𝑆𝑂𝐶𝑏(𝑡)≤𝑆𝑂𝐶𝑏𝑚𝑎𝑥∗𝐷𝑜𝐷 (30) 𝑏𝑖𝑛 ≤𝑆𝑂𝐶𝑏𝑚𝑎𝑥 𝐶𝐸/𝑃 ⁄ (31) 𝑏𝑜𝑢𝑡 ≤𝑆𝑂𝐶𝑏𝑚𝑎𝑥 𝐶𝐸/𝑃 ⁄ (32) Where PV, and WT are decision variables corresponding to the installed capacity, in MWp, of photovoltaic panels and wind turbines, while 𝐸𝑃𝑉 1𝑀𝑊(𝑡) and 𝐸𝑊𝑇 1𝑀𝑊(𝑡) represent the hourly output of 1 MWp of installed PV and Wind power, respectively. BESS installed capacity (SOCbmax) is also used as decision variable to solve the optimization model. 38 Capex and fixed opex are based on cost assumptions from the different technologies found in the literature, scaled depending on their capacities using typical scaling factors. To incorporate the nonlinear scaling equations of equipment in the flexible section, a piecewise linear approximation formula with sos2 type variables was defined in GAMS. For a set of points x and known parameters Qi(x) representing the capacity of process i, the nonlinear formula for the capex is provided below: 𝐶𝑎𝑝𝑒𝑥𝑖(𝑥)=𝐶𝑎𝑝𝑒𝑥𝑟𝑒𝑓∗(𝑄𝑖(𝑥) 𝑄𝑟𝑒𝑓)𝛼 (33) Where Capexref represents the reference technology cost found in the literature, 𝑄ref is the capacity of the reference process expressed in thermal power or flowrate of captured CO2 depending on the unit operation, 𝑄i is the capacity of the process at the different set points in 𝑥, and 𝛼 is the scaling factor, assumed equal to 0.67. Defining the sos2 binary variables as lami(𝑥), the capex of the optimized capacity 𝑄𝑖′ can be approximated through equations (34) – (38): 𝑄𝑖′=𝑠𝑢𝑚(𝑥,𝑙𝑎𝑚𝑖(𝑥)∗𝑄𝑖(𝑥)) (34) 𝐶𝑎𝑝𝑒𝑥𝑖′=𝑠𝑢𝑚(𝑥,𝑙𝑎𝑚𝑖(𝑥)∗𝐶𝑎𝑝𝑒𝑥𝑖(𝑥)) (35) 𝑠𝑢𝑚(𝑥,𝑙𝑎𝑚𝑖(𝑥))=1 (36) 𝑄𝑖,𝑚𝑖𝑛 ′=𝑠𝑚𝑖𝑛(𝑥,𝑄𝑖(𝑥)) (37) 𝑄𝑖,𝑚𝑎𝑥 ′=𝑠𝑚𝑎𝑥(𝑥,𝑄𝑖(𝑥)) (38) With the values of 𝑄𝑖′ for the e-calciner and CPU determined with equations (39) and (40): 𝑄𝑖,𝑒−𝑐𝑎𝑙𝑐 ′=𝑄𝑒−𝑐𝑎𝑙𝑐 ≥0.95∗𝐸𝑒−𝑐𝑎𝑙𝑐(𝑡) (39) 𝑄𝑖,𝐶𝑃𝑈 ′=𝐶𝑂2−2𝑚𝑎𝑥≥𝐶𝑂2−2(𝑡) (40) 45 MILP model results – Denmark Table 21: Summary of MILP model results for short-term and long-term simulations of decarbonized plants in Denmark Short-term horizon Long-term horizon eC – afK Inflexible eC – afK Flexible eC – pK OC – HK PCC+HP Coal Oxyfuel Coal PCC+HP AF Oxyfuel AF eC – afK Inflexible eC – afK Flexible eC – pK OC – HK PCC+HP Coal Oxyfuel Coal PCC+HP AF Oxyfuel AF PV installed power MW 166 48 255 99 115 33 117 33 416 168 635 245 286 77 290 80 Wind installed power MW 177 168 268 104 121 34 124 33 153 143 233 90 105 30 107 30 Battery MWh 233 0 361 139 163 47 163 49 1,192 0 1,820 701 819 229 846 229 h 3 0 3 3 3 3 3 3 13 0 13 13 13 13 13 13 Calcined meal storage ton - 64,266 - - - - - - - 21,683 - - - - - - h - 476 - - - - - - - 161 - - - - - - Grid electricity price €/MWh 100 100 100 100 100 100 100 100 70 70 70 70 70 70 70 70 Carbon intensity of grid electricity supply kgCO2/MWh 180 180 180 180 180 180 180 180 120 120 120 120 120 120 120 120 Renewables share in electricity supply % 90% 99% 90% 90% 90% 90% 90% 90% 98% 98% 98% 98% 98% 98% 98% 98% Renewable energy curtailed % 30% 4% 29% 29% 29% 29% 30% 29% 30% 7% 30% 30% 31% 30% 31% 31% Levelized cost of renewable electricity supply €/MWh 89 54 89 89 89 89 89 89 76 43 76 76 76 76 76 76 Levelized cost of electricity supply €/MWh 90 54 90 90 90 90 90 90 76 44 76 76 76 76 76 76 Carbon intensity of electricity supply (renewables + grid) kgCO2/MWh 31 8 31 31 31 31 31 31 30 13 30 30 30 30 30 30 Scope 1 CO2 emissions kgCO2/tclk -33 -27 44 -46 43 44 -47 -52 -33 -27 44 -46 43 44 -47 -52 Scope 2 CO2 emissions kgCO2/tclk 25 6 38 15 17 5 17 5 25 11 38 15 17 5 17 5 CO2 emissions avoided (net) % 101% 103% 89% 104% 93% 94% 104% 106% 101% 102% 89% 104% 93% 94% 104% 106% 46 MILP model results – India Table 22: Summary of MILP model results for short-term and long-term simulations of decarbonized plants in India Short-term horizon Long-term horizon eC – afK Inflexible eC – afK Flexible eC – pK OC – HK PCC+HP Coal Oxyfuel Coal PCC+HP AF Oxyfuel AF eC – afK Inflexible eC – afK Flexible eC – pK OC – HK PCC+HP Coal Oxyfuel Coal PCC+HP AF Oxyfuel AF PV installed power MW 361 233 553 215 250 69 255 70 698 296 1,050 406 474 132 484 133 Wind installed power MW 234 250 358 136 157 45 161 45 29 235 43 17 20 6 20 6 Battery MWh 690 0 1,043 407 480 130 490 133 1,559 182 2,360 906 1059 294 1081 296 h 7 0 7 7 8 7 8 8 17 2 17 17 17 17 17 17 Calcined meal storage ton 0 31,973 0 0 0 0 0 0 0 61,594 0 0 0 0 0 0 h - 237 - - - - - - - 456 - - - - - - Grid electricity price €/MWh 68 68 68 68 68 68 68 68 50 49 50 50 50 50 50 50 Carbon intensity of grid electricity supply kgCO2/MWh 633 633 633 633 633 633 633 633 440 440 440 440 440 440 440 440 Renewables share in electricity supply % 90% 90% 90% 90% 90% 90% 90% 90% 98% 98% 98% 98% 98% 98% 98% 98% Renewable energy curtailed % 27% 11% 28% 27% 27% 28% 27% 27% 29% 11% 29% 29% 29% 29% 29% 29% Levelized cost of renewable electricity supply €/MWh 120 63 120 120 120 120 120 120 67 40 67 67 67 67 67 67 Levelized cost of electricity supply €/MWh 115 63 115 115 115 115 115 115 67 41 67 67 67 67 67 67 Carbon intensity of electricity supply (renewables + grid) kgCO2/MWh 86 79 86 86 86 86 86 86 43 28 43 43 43 43 43 43 Scope 1 CO2 emissions kgCO2/tclk -33 -27 44 -46 43 44 -47 -52 -32 -27 44 -46 43 44 -47 -52 Scope 2 CO2 emissions kgCO2/tclk 71 65 108 42 49 13 49 13 38 23 57 22 26 7 26 7 CO2 emissions avoided (net) % 95% 95% 80% 101% 90% 93% 100% 105% 99% 100% 87% 103% 92% 94% 103% 106% 47 MILP model results – Egypt Table 23: Summary of MILP model results for short-term and long-term simulations of decarbonized plants in Egypt Short-term horizon Long-term horizon eC – afK Inflexible eC – afK Flexible eC – pK OC – HK PCC+HP Coal Oxyfuel Coal PCC+HP AF Oxyfuel AF eC – afK Inflexible eC – afK Flexible eC – pK OC – HK PCC+HP Coal Oxyfuel Coal PCC+HP AF Oxyfuel AF PV installed power MW 101 31 154 59 70 20 71 20 252 144 378 149 174 49 176 49 Wind installed power MW 295 254 454 175 203 57 207 57 208 216 319 122 144 39 148 41 Battery MWh 121 0 188 72 83 23 85 24 753 0 1,163 443 515 148 519 143 h 1 0 1 1 1 1 1 1 8 0 8 8 8 8 8 8 Calcined meal storage ton - 8,784 - - - - - - - 18,871 - - - - - - h - 65 - - - - - - - 140 - - - - - - Grid electricity price €/MWh 34 34 34 34 34 34 34 34 24 24 24 24 24 24 24 24 Carbon intensity of grid electricity supply kgCO2/MWh 628 628 628 628 628 628 628 628 322 322 322 322 322 322 322 322 Renewables share in electricity supply % 90% 90% 90% 90% 90% 90% 90% 90% 98% 98% 98% 98% 98% 98% 98% 98% Renewable energy curtailed % 33% 10% 34% 34% 33% 33% 33% 33% 26% 12% 26% 26% 26% 26% 26% 26% Levelized cost of renewable electricity supply €/MWh 70 44 70 70 70 70 70 70 48 27 48 48 48 48 48 48 Levelized cost of electricity supply €/MWh 66 43 67 67 66 66 66 66 47 27 47 47 47 47 47 47 Carbon intensity of electricity supply (renewables + grid) kgCO2/MWh 76 71 76 76 76 76 76 76 26 18 26 26 26 26 26 267 Scope 1 CO2 emissions kgCO2/tclk -32 -27 44 -46 43 44 -47 -52 -32 -27 44 -46 43 44 -47 -52 Scope 2 CO2 emissions kgCO2/tclk 60 58 92 36 42 11 42 11 22 15 33 13 15 4 15 4 CO2 emissions avoided (net) % 96% 96% 82% 101% 90% 94% 101% 105% 101% 102% 90% 104% 93% 94% 104% 106% 48 S8 Analysis of system behavior Table 24: Installed capacity of renewable electricity generation technologies and storage capacities of the inflexible/flexible eC - afK plant assessed under different locations and time horizons. For BESS, storage duration in hours represents the additional electricity demand to operate the decarbonized system. For calcined meal (CM), storage duration in hours represents the amount of feedstock required to operate continuously the clinker production section. Short-term inflexible PV [MWp] WT [MWp] BESS [MWh] BESS [h] CM storage [ton] CM storage [h] Italy 318 164 830 9 - - Denmark 166 177 233 3 - - India 361 234 690 7 - - Egypt 101 295 121 1 - - Short-term flexible PV [MWp] WT [MWp] BESS [MWh] BESS [h] CM storage [ton] CM storage [h] Italy 200 206 0 0 36,280 269 Denmark 48 168 0 0 64,266 476 India 233 250 0 0 31,973 237 Egypt 31 254 0 0 8,784 65 Long-term inflexible PV [MWp] WT [MWp] BESS [MWh] BESS [h] CM storage [ton] CM storage [h] Italy 563 99 1,479 16 - - Denmark 416 153 1,192 13 - - India 698 29 1,559 17 - - Egypt 252 208 753 8 - - Long-term flexible PV [MWp] WT [MWp] BESS [MWh] BESS [h] CM storage [ton] CM storage [h] Italy 340 135 500 5 33,015 245 Denmark 168 143 0 0 21,683 161 India 296 235 182 2 61,594 456 Egypt 144 216 0 0 18,871 140 49 Table 25: Annual renewable electricity generation by technology and curtailment rates of the inflexible/flexible eC - afK plant assessed under different locations and time horizons. Short-term inflexible PV [MWh/y] WT [MWh/y] Total [MWh/y] Curtailed [-] Italy 529,279 435,302 964,581 21% Denmark 187,587 865,190 1,052,777 30% India 610,335 436,616 1,046,951 27% Egypt 192,878 914,807 1,107,684 33% Short-term flexible PV [MWh/y] WT [MWh/y] Total [MWh/y] Curtailed [-] Italy 332,880 546,782 879,662 14% Denmark 54,242 821,197 875,439 4% India 393,928 466,470 860,398 11% Egypt 59,200 787,664 846,864 10% Long-term inflexible PV [MWh/y] WT [MWh/y] Total [MWh/y] Curtailed [-] Italy 937,057 262,774 1,199,831 29% Denmark 470,097 747,876 1,217,973 32% India 1,166,569 52,245 1,218,814 29% Egypt 481,239 645,016 1,126,256 26% Long-term flexible PV [MWh/y] WT [MWh/y] Total [MWh/y] Curtailed [-] Italy 565,896 358,328 924,224 8% Denmark 189,847 698,995 888,842 7% India 500,441 438,482 938,923 11% Egypt 274,994 669,825 944,819 12% 50 Short-term – Inflexible plants Italy: Due to good PV and WT capacity factors, the inflexible plant in Italy favors a balanced mix of PV+BESS and WT installations. The 318 MWp of PV panels and 164 MWp of WT installed translates to a 55% annual energy generation from PV and 45% from WT, with large BESS capacity (830 MWh) to complement the intraday variability in the supply from PVs (Table 24), charged during the peak hours of sun irradiance (Figure 17). Curtailment is concentrated during the middle of the day, when the sun irradiance is higher, and more pronounced from November to May, following the seasonal variation of wind resource (Figure 31). Overall, curtailment reaches 21% of annual renewable electricity generation. Grid electricity and the BESS are used to support the shortage from renewable sources throughout the year, with emphasis in the low-wind season (June – October), when the power output in hours without sun is lower (Figure 17,Figure 27). Battery is charged with excess PV power. Backup grid electricity is concentrated in the first hours of the days while the BESS is primarily discharged during the second half of the day. Denmark: Because of the extreme contrast between the capacity factors of WT and PV panels, the optimum is found with an energy mix of 82% generation from WT vs 18% from PVs (Table 25), with 233 MWh of BESS capacity installed to support some of the intraday shortages. Charging happens primarily when the sun shines brightest (Figure 19), while discharge is used to support hours of the day with less sunlight, complemented with grid electricity backup (Figure 28). With a strong reliance on wind power output, and no long-term load-shifting storage alternative, excess renewable energy is wasted throughout the day in the months with highest wind speeds (Figure 12,Figure 32) and concentrated during the peak PV power output hours in the months with better sun irradiance (Figure 11,Figure 32). As a consequence, yearly curtailment rate reaches 30%. India: The energy mix is similar to the Italian case, slightly favoring PV power output over WT (58% vs 42% in energy generation, Table 25). Daily wind patterns are unstable, exhibiting a significant drop in the middle hours between 6 am – 6 pm (Figure 14), which are balanced with sunlight hours in the same time range (Figure 13). A 650 MWh BESS is installed to manage the strong PV fluctuations, charged in the hours of highest sunlight (between 8 am – 3 pm, Figure 21). Grid electricity helps the BESS in supporting the system’s energy demand in the early hours, when there is no PV power output (Figure 29). Its largest contribution is concentrated between September – November, months with the worst wind resource. The variability of PV power output concentrates curtailment in the middle of the day, with the exception of June and July, where excess PV is complemented with high wind speeds in the early and late hours (Figure 33). To meet the seasonal mismatch, overcapacity leads to a curtailment of 27%. Egypt: Although Egypt displays good sun availability, comparable to Italy or India, the 5% additional capacity factor for wind energy with small seasonal variation, and the higher cost on capital of battery systems needed to complement sunlight daily fluctuations, drives the optimum of the system to generate 83% of annual renewable energy from WT (Table 25). BESS capacity is limited to 121 MWh, the lowest in all locations. Wind speed patterns are unstable during the day, slowing down in the first hours and then surging to reach maximum values around 7 – 10 pm in most months, except in January and February, where there is less intraday variation (Figure 16). Because of this daily mismatch, electricity demand from the battery system and the grid are mostly clustered in the morning hours from 0 – 8 am, with some grid backup also used to supplement the lack of wind energy in the hours 51 of the afternoon. The BESS is charged primarily in the afternoon hours when the wind speeds are higher (Figure 23). Likewise, curtailment rises in the second half of the day, though, because the system is sized to meet the energy demand in the lowest wind-speed months, excess renewables are present also in the early hours (Figure 34). The period between June – August is an exception with negligible curtailment in the morning due to the low wind availability. Short-term – Flexible plants Italy: Moves to a stronger reliance on WT+CM storage, increasing the WT share in the annual energy production to 62%, from 45% in the inflexible case (Table 25). The shift is driven by a much economically attractive option of storing calcined material instead of electricity in a battery system, providing the option of long-term storage to better cope with the seasonality of wind energy, coupled with cheap grid electricity as backup. Therefore, no BESS is installed. Grid electricity is focused to support PV output during the low wind season (June – October) (Figure 27). Meanwhile, CM storage is used as intraday storage (charging with excess PV and discharging when the sun is not shining), and for load shifting, using the high-wind season between November – January to charge the system, which is depleted in July (Figure 18,Figure 25). By leveraging more stable wind energy and more cheaper and versatile CM storage, curtailment is reduced from 21% to 14% (Figure 31). Denmark: With the chance of seasonal storage, the dependence on wind energy is accentuated even further reaching 94% of yearly energy production (compared to 82% in the inflexible case), while reducing the installed capacity of both PV (-71%) and WT (-5%) (Table 24,Table 25). The storage unit is charged steadily during the high-wind-speed months, to provide load-shifting capability. Stored material is then consumed during the months with the lowest wind speeds (Figure 20). Stable daily power production from wind turbines combined with cheap long-term storage reduces the need for grid backup below the maximum constrained to the optimization model, using only 0.7% of the annual demand. Likewise, curtailment is significantly reduced to 4%. India: Despite having a low wind capacity factor (21.3%), the system decides to increase the WT installed capacity (+7%) and decrease PVs (-35%), leading to a balanced energy share now slightly tilted to wind energy (54% from WT and 46% from PV) (Table 24,Table 25). Intraday energy storage is provided by cheaper CM storage, removing the need for more expensive electrochemical BESS. In addition, CM storage now offers the possibility to exploit long-term load-shifting behavior. High and more stable winds in June and July drive the SOC to its maximum level, consuming the stored material in the subsequent months of August – October, when wind speeds drop heavily. Grid backup is utilized to complement the shortages of wind energy from September to April (Figure 29). By reducing excess peaks from PV power output, curtailment is reduced to 11%, compared to the 27% in the inflexible plant. Egypt: The possibility of cheaper storage with long-term capabilities drives the system to become even more dependent on wind energy (93% of annual energy generation) compared to the inflexible plant (Table 25). By leveraging the energy storage, excess installed capacity to meet the electricity demand is significantly reduced, leading to a curtailment rate of 10% clustered in the afternoon (Figure 34). The calcined meal storage is used mostly for daily needs, harnessing the high winds in the afternoon hours for charging, and supplying stored material in the early hours of the day (Figure 24). As a result, only 8,784 tons of storage capacity are installed. Longer-term storage is still utilized, 52 supplementing the low-wind hours in July and August, by filling the capacity in May. Besides calcined material, the grid is used as a backup to support the hours with low wind energy output, concentrated in months where wind speeds reach the minimum, or when the afternoon peaks are low. Long-term – Inflexible plants Italy: 77% increase in PV installed capacity and 40% reduction in WT installed capacity, coupled with a 78% increase in BESS, with respect to the short-term case (Table 24). With the new portfolio, 78% of the renewable energy generated comes from PV panels (up from 55%). Cheaper capital investment costs of PV panels and BESS benefit the business case of increasing the reliance on sun energy and intraday storage over installing a more balanced and seasonal electricity supply based on WT. Curtailed electricity rises from 21% in the short-term to 29% due to the overcapacity needed to compensate the more variable nature of sun irradiance, leading to a 24% increase in annual energy generation to meet the additional 8% on the renewable electricity target (Table 25). The constraint on renewable energy drives the system to lean more on BESS to complement the lack of PV output during the hours between 4 pm – 6 am (Figure 17), concentrating the small grid backup in the months of October to January, i.e., the months with worst sunlight. Denmark: Large increase of PV and BESS installed capacities of 151% and 412% respectively, while reducing WT power capacity by 14%, preserving the majority of renewable energy produced during the year based on wind (61%) instead of sunlight (39%) (Table 24,Table 25). The BESS follows the behavior of the sun, charging when irradiance is higher, and discharging when there is no sunlight, concentrated in the spring/summer months (Figure 19). Limited grid support and steep capex reduction in technology foster an over installation of renewables. Correspondingly, PV power exceeding storage capacity leads to high curtailment rates in middle of the day, reaching 30% of annual renewable energy production (Figure 32). India: Steep cost reductions in PV and BESS drive a 91% and 123% increase in installed capacity, respectively, and a reduction of WT capacity of 88%, compared to the short-term plant. As consequence, PVs represent 96% of renewable annual energy generation. With negligible influence from wind resource, excess energy used to charge the BESS and curtailment follows almost exclusively sunlight patterns (Figure 21,Figure 33), with limited support from the grid in the period of August to December (Figure 29). Reductions in curtailed energy from surplus wind production in July and June are exceeded by wasted power production from PV panels in peak hours of the day, reaching an annual rate of 29% (Figure 33). Egypt: Similar to Denmark, the system reduces WT installed capacity by 29% in favor of a 150% and 522% increase in PV and BESS capacity, respectively (Table 24). Therefore, expanding the share of energy generation from PV panels to 43%, as compared to the 17% in the short-term (Table 25). In contrast to the more skewed energy generation distribution in the short-term case, the addition of seasonally stable PV and the BESS improves relative curtailment by 7%, concentrated now more in the peak hours of sunlight instead of the late hours of excess wind energy (Figure 34). Similarly, the charging of the BESS shifts more towards the middle of the day, while the discharge concentrates in the hours of the morning to support daily periods without sunlight and scarce winds (Figure 23). Limited grid backup complements the system in these early hours (Figure 30). 53 Long-term Flexible plants Italy: Compared with the inflexible plant in the long-term, the possibility of load-shifting to months with better wind availability through the storage of calcined meal moves the optimum mix to a 61/39 split of PV/WT in terms of energy generation (Table 25). PV remains as the main source of energy supply, as a result of extremely low capex and a good capacity factor. Compared to the flexible plant in the short-term, there is a shift to cheaper PV to supply the majority of the energy demand, and the presence of now economically attractive BESS, which is charged with excess power from the PV panels in the hours with highest sun irradiance (Figure 17). In the end, the combination of PV+WT+BESS+CM storage reduces curtailment of renewable energy to 8% (Table 25, Figure 31). While calcined material is used constantly throughout the year to reduce the energy demand in hours when the sun is not shining, it is charged to its maximum capacity steadily from December to May, and depleted along the months of June to September, when the contribution from wind energy is the lowest. October, and to a less extent September, is supplemented with the available grid backup unused during the rest of the year (Figure 27). Unlike the short-term case, the charging and discharging of the calcined meal storage occurs gradually and over a longer period of time (Figure 26), supplementing the lack of grid electricity support with the use of the BESS. Denmark: Again, adding a cost-effective long-term load shifting alternative benefits the introduction of WT, bringing the optimum to a distribution of 79% energy generation from WT and 21% from PV, with no need of a BESS (Table 24,Table 25). Cheaper capex for PV panels increases the share of sunbased energy production, compared to the short-term case, reducing the need for load-shifting capability delivered by the calcined meal storage. Stable wind energy supply is used to meet the plant’s electricity demand, while the excess energy from PV is harnessed to produce additional calcined material to support the operation when there is no sunlight. High wind resource during February, March, and September drive the filling of the intermediate storage, which is primarily used from April to August, and in October (Figure 20). Curtailment is limited and concentrated in the hours of PV power output during the high-sun season (March to September). With more intraday variability compared to the short-term mix, grid electricity is used to fill the remaining 2% of annual demand, concentrated in the months of October and November. India: Cheap PV+BESS coupled with a cost-effective, long-term storage alternative leads to a balanced mix of 53/47 energy generation from PV/WT (Table 25). Both CM storage and BESS are installed, with capacities of 61,594 tons and 182 MWh, respectively. Excess calcined meal is produced all year with surplus PV output power and discharged steadily in the hours with no sunlight (Figure 22). Long-term storage capacity is charged from May to July, when the wind speeds are highest, and is used to supplement the shortage of wind energy from September to October (Figure 26). Shortage of wind resource in November is in turn supported by the use of grid electricity (Figure 29). The cost effectiveness of storing material for seasonal load shift coupled with more stable power output from wind energy outperforms a strategy of installing overcapacity of PV and battery (i.e., inflexible plant) to supplement the low-wind season. BESS, in this case, is used daily to support the afternoon hours with no sunlight (Figure 21). Similar to the short-term flexible plant, curtailment rate is reduced to 12%, mainly as a result of excess energy from PV power in the middle of the day and the strong wind seasonality in July (Figure 33). 54 Egypt: Incorporating calcined meal storage becomes more economically attractive than storing electricity in a battery system, providing additional seasonal load-shifting. Consequently, the share of PV panels in total energy generation drops from 43% in the inflexible case to 29%, replacing any need for BESS capacity with 18,871 tons of material storage. With limited access to grid electricity to support extremely low-wind hours in the morning between June and August, an optimum is found exploiting load-shifting long-term storage of calcined material instead (Figure 26). Stored calcined meal is discharged in the morning hours of the day all year as a response to the intraday fluctuations of wind resource, while charging is clustered in the hours of PV generation and the later hours of the afternoon, when wind speeds peak (Figure 24). Annual curtailed renewable electricity reaches 12%, concentrated in the middle of the day and during afternoon peaks of wind energy (Figure 34). 61 Denmark BESS charge and discharge patterns Figure 19: Hourly average profile of the BESS charging (positive values) and discharging (negative values) for the eC - afK plant under inflexible/flexible conditions and short-term/long-term scenarios. An hour of battery storage represents 92.8 and 94.3 MWhel of stored electricity for the inflexible and flexible designs, respectively. BESS, present only on inflexible cases, is charged with excess renewable energy from PV panels and discharged in the hours when sun irradiance is low. The use of the BESS is concentrated in the spring/summer months. 62 Calcined meal storage charge and discharge patterns Figure 20: Hourly average profile of the calcined meal storage charging (positive values) and discharging (negative values) for the flexible eC – afK plant in short-term/long-term scenarios. An hour of intermediate storage represents 135 tons of calcined meal. The storage behavior is dominated by the wind speed profiles in the short-term, charging during the strongest months, and discharging during the less windy period. In the long-term, addition of PV shifts the behavior, using excess power from PV panels in the middle of the day to charge. Calcined meal is discharged during the mornings and evening, concentrated in the less windy months. 63 India BESS charge and discharge patterns Figure 21: Hourly average profile of the BESS charging (positive values) and discharging (negative values) for the eC - afK plant under inflexible/flexible conditions and short-term/long-term scenarios. An hour of battery storage represents 92.8 and 94.3 MWhel of stored electricity for the inflexible and flexible designs, respectively. In all cases, the battery is charged during the hours of peak PV power output and discharged when sun irradiance is low, except under the flexible short-term case, where no BESS is installed. 64 Calcined meal storage charge and discharge patterns Figure 22: Hourly average profile of the calcined meal storage charging (positive values) and discharging (negative values) for the flexible eC – afK plant in short-term/long-term scenarios. An hour of intermediate storage represents 135 tons of calcined meal. The intermediate storage is charged during excess PV power, and discharged when sun irradiance is low. Higher wind power outputs at the earliest and latest hours of the day are noticeable during June and July, the months with strongest wind speeds. 65 Egypt BESS charge and discharge patterns Figure 23: Hourly average profile of the BESS charging (positive values) and discharging (negative values) for the eC - afK plant under inflexible/flexible conditions and short-term/long-term scenarios. An hour of battery storage represents 92.8 and 94.3 MWhel of stored electricity for the inflexible and flexible designs, respectively. BESS is only installed in inflexible plants and it is charged by excess power from PV panels during high-irradiance hours and excess power from wind turbines during high-wind speed hours in the evening of the summer season. 66 Calcined meal storage charge and discharge patterns Figure 24: Hourly average profile of the calcined meal storage charging (positive values) and discharging (negative values) for the flexible eC – afK plant in short-term/long-term scenarios. An hour of intermediate storage represents 135 tons of calcined meal. In the short-term, the charging is dominated by wind power patterns, which increase in the late hours of the day. Discharging of calcined material in the morning helps compensate lower wind speeds. In the long-term, the introduction of more PV panels increases charging in the middle of the day. Discharge concentrates in the early hours, when there is no sun and wind speeds are low, especially in June – August. 67 Calcined meal storage SOC Short-term comparison Figure 25: State of Charge of the calcined meal storage unit for the flexible eC - afK plants under short-term scenario. Left axis shows the storage in tons, right axis shows the number of days of kiln operation at constant load. 68 Long-term comparison Figure 26: State of Charge of the calcined meal storage unit for the flexible eC - afK plants under long-term scenario. Left axis shows the storage in tons, right axis shows the number of days of kiln operation at constant load. 69 S8.3 Grid electricity demand Italy Figure 27: Hourly average of grid electricity consumption for the eC - afK inflexible/flexible plants in the short-term/longterm scenarios. 70 Denmark Figure 28: Hourly average of grid electricity consumption for the eC - afK inflexible/flexible plants in the short-term/longterm scenarios. 77 S9 Capex breakdown Italy Figure 35: Capex breakdown for the assessed plants located in the south of Italy under both time scenarios (ST: short-term; LT: long-term). Renewable installed capacities (PV, WT, BESS) are represented in green patterns. 78 Denmark Figure 36: Capex breakdown for the assessed plants located in Denmark under both time scenarios (ST: short-term; LT: long-term). Renewable installed capacities (PV, WT, BESS) are represented in green patterns. 79 India Figure 37: Capex breakdown for the assessed plants located in the northwest of India under both time scenarios (ST: shortterm; LT: long-term). Renewable installed capacities (PV, WT, BESS) are represented in green patterns. 80 Egypt Figure 38: Capex breakdown for the assessed plants located in the north of Egypt under both time scenarios (ST: shortterm; LT: long-term). Renewable installed capacities (PV, WT, BESS) are represented in green patterns. 81 S10 Sensitivity of CAC to cost variations of selected parameters Short-term horizon Figure 39: Sensitivity of the CAC of all assessed plants to a ±50% variation in the cost of fuel, capex repayment for renewable electricity production (includes: PV panels, wind turbines, and BESS), capex repayment for other equipment (includes: calcined meal storage, e-calciner, CPU, amine-based PCC system, heat pump, plasma burners, ASU, SOEC, retrofits), and T&S in the short-term scenario. CCS benchmarks assessed assuming 100% AF combustion. Error bars represent the deviation between locations. 82 Long-term horizon Figure 40: Sensitivity of the CAC of all assessed plants to a ±50% variation in the cost of fuel, capex repayment for renewable electricity production (includes: PV panels, wind turbines, and BESS), capex repayment for other equipment (includes: calcined meal storage, e-calciner, CPU, MEA-based PCC system, heat pump, plasma burners, ASU, SOEC, retrofits), and T&S in the long-term scenario. CCS benchmarks assessed assuming 100% AF combustion. Error bars represent the deviation between locations. 83 S11 Linear regression model of inflexible and flexible LCOE A least squared regression model was used to approximate a linear function between the LCOE of inflexible and flexible plants (Figure 41). The model was built using the calculated figures from the four locations under the two time horizons, including the evaluation of plants located in Egypt and India with a WACC consistent with developed economies, i.e., 6%. Because the model does not incorporate variations in climate conditions, different renewable share constraints (between time horizons), or grid electricity prices across locations, some scenarios deviate from the baseline (green dashed line). For example, the LCOE of a flexible plant in Egypt under long-term conditions is calculated at 27.4 €/MWh, while the model overestimates a value of 29.4 €/MWh. Meanwhile, the flexible LCOE of Italy in the long-term is 36.0 €/MWh, which the models underestimates at 34.3 €/MWh. Therefore, two boundary curves have been included by shifting the baseline’s intercept +75% in the upper case and -75% in the lower case. These margins were incorporated as an uncertainty range in the estimation of the inflexible/flexible competitive regions. Figure 41: Linear trend, estimated with least square method, of the relationship between the LCOE of inflexible and flexible plants. Upper (+) and lower (-) ranges where calculated by moving the trendline ± 75% from the intercept to cover the whole range of LCOE pairs and show the impact of variability in the estimation of competitive regions. 84 S12 Additional long-term capex uncertainty for CCS benchmark plants An optimistic scenario for the PCC system assuming a LR = 17% was included in the assessment of competitive regions, corresponding to the higher end of the learning rates obtained for postcombustion FGD systems (IEAGHG, 2021). With a LR = 17%, the specific capex of the PCC system is further reduced to 1.4 [M€/(tCO2/h)], improving the cost-effectiveness of the PCC benchmark case and the eC – afK plants. Improved capital cost estimates for solvent-based PCC technologies in power generation and gas processing applications have provided a more robust basis for updating investment assumptions applicable to the cement industry. Notably, figures reported in the recent DoE–NETL study represent a 205% increase over previous estimates from earlier techno-economic assessments (Gardarsdottir et al., 2019), once adjusted to 2022 values using CEPCI scaling and an exchange rate of 0.9 €/US$. While the updated figures correspond to a higher CO2 capture rate of 95%, compared to 90% in earlier studies, they nevertheless reflect a significant upward revision in expected capital expenditures for amine-based capture systems in cement applications. By contrast, no updated capex estimates for Oxyfuel cement plants have been published to date, and the first full-scale industrial demonstrations are expected to enter operation in the short-tomedium term horizon. In light of this uncertainty and given that the reference capex for Oxyfuel systems used in this work is derived from the same techno-economic study as the earlier benchmark for solvent-based PCC, a 50% increase in Oxyfuel capital costs was assumed for the competitiveness region comparative analysis. This adjustment allows for a more nuanced and comprehensive evaluation of the economic competitiveness of electrified cement plants relative to Oxyfuel systems. 85 References Aïtcin , P.-C., & Flatt, R. (2015). Science and Technology of Concrete Admixtures. Woodhead Publishing. Atlas Copco. (2024). Driving Centrifugal Compressor Technology. Retrieved from https://www.atlascopco.com/content/dam/atlas-copco/compressor-technique/gas-andprocess/documents/driving-centrifugal-compressor-technology-new.pdf Barlow, H., & Shahi, S. S. (2024). State of the art: CCS Technologies 2024. Global CC Institute. Retrieved from https://www.globalccsinstitute.com/wp-content/uploads/2025/08/CCS-Technologies-Compendium-202408-16-1-1.pdf Bhatty, J., Miller, F., & Kosmatka, S. (2011). Innovations in Portland Cement Manufacturing (2nd ed.). Portland Cement Association. Campanari, S., Cinti, G., Consonni, S., Fleiger, K., Gatti, M., Hoppe, H., . . . Voldsund, M. (2016). Design and performance of CEMCAP cement plant without CO2 capture. De Lena, E., Spinelli, M., Gatti, M., Scaccabarozzi, R., Campanari, S., Consonni, S., . . . Romano, M. C. (2019). Techno-economic analysis of calcium looping processes for low CO2 emission cement plants. International Journal of Greenhouse Gas Control, 82, 224-260. doi:https://doi.org/10.1016/j.ijggc.2019.01.005 De Lena, E., Spinelli, M., Martinez, I., Gatti, M., Scaccabarozzi, R., Cinti, G., & Romano, M. C. (2017). Process integration study of tail-end Ca-Looping process for CO2 capture in cement plants. International Journal of Greenhouse Gas Control, 67, 71-92. doi:https://doi.org/10.1016/j.ijggc.2017.10.005 ECRA. (2012). ECRA CCS Project-Report on Phase III. European Cement Research Academy. Retrieved from https://api.ecra-online.org/fileadmin/files/ECRA_Technical_Report_CCS_Phase_III.pdf GCCA. (2024). Cement Industry Net Zero Progress Report 2024/25. Retrieved from https://gccassociation.org/wpcontent/uploads/2024/11/GCCA-Cement-Industry-Progress-Report-202425.pdf Heidelberg Materials. (2022). Capital allocation alignment with carbon reduction targets. Retrieved from https://www.heidelbergmaterials.com/sites/default/files/202211/Capital%20allocation%20alignment%20with%20carbon%20reduction%20targets%20- %20CA100%2B%20Net%20Zero%20Company%20Benchmark.pdf Holcim. (2024). 2024 Integrated Annual Report. Retrieved from https://www.holcim.com/sites/holcim/files/docs/28022025-finance-holcim-fy-2024-report-fullen.pdf#page=64 IEA. (2025). CCUS Projects Database. Retrieved from https://www.iea.org/data-and-statistics/data-product/ccusprojects-database IEAGHG. (2013). Deployment of CCS in the Cement Industry. Retrieved from https://ieaghg.org/publications/deployment-of-ccs-in-the-cement-industry/ IEAGHG. (2021). Towards improved guidelines for cost evaluation of carbon capture and storage 2021-TR05. Retrieved from www.ieaghg.org Magli, F., Spinelli, M., Fantini, M., Romano, M. C., & Gatti, M. (2022). Techno-economic optimization and offdesign analysis of CO2 capture purification units for cement plants with oxyfuel-based CO2 capture. International Journal of Greenhouse Gas Control, 115. doi:https://doi.org/10.1016/j.ijggc.2022.103591 86 Mastorakos, E., Massias, A., Goussis, D., Tsakiroglou, C., Burganos, V., & Payatakes, A. (1999). CFD predictions for cement kilns including flame modelling, heat transfer and clinker chemistry. Applied Mathematical Modelling, 23(1), 55-76. doi:https://doi.org/10.1016/S0307-904X(98)10053-7 Schorcht, F., Kourti, I., Scalet, B. M., Roudier, S., & Delgado Sancho, L. (2013). Best Available Techniques (BAT) Reference Document for the Production of Cement, Lime and Magnesium Oxide. doi:doi:10.2788/12850 Voldsund, M., Gardarsdottir, S. O., De Lena, E., Perez-Calvo, J.-F., Jamali, A., Berstad, D., . . . Jordal, K. (2019). Comparison of Technologies for CO2 Capture from Cement Production—Part 1: Technical Evaluation. Energies, 12(3), 559. doi:https://doi.org/10.3390/en12030559