Power cycles integration in concentrated solar power plants with energy storage based on calcium looping

1 Power cycles integration in concentrated solar power plants with energy 1 storage based on calcium looping 2 3 C. Ortiz a, *, R. Chacartegui b, J.M. Valverde a, A. Alovisio c, J.A. Becerra b 4 5 a Faculty of Physics, University of Seville, Avenida Reina Mercedes s/n, 41012 Sevilla, Spain 6 b Energy Engineering Department, University of Seville, Camino de los Descubrimientos s/n, 41092 Sevilla, Spain 7 c Politecnico di Torino, Department of Energy Engineering, Corso Duca degli Abruzzi 24, 10129 Torino, Italy 8 9 * Corresponding author. Tel.: +34 655783930 10 E-mail address: [email protected] 11 12 Abstract 13 Efficient, low-cost and environmentally friendly storage of thermal energy stands as a main 14 challenge for large scale deployment of solar energy. This work explores the integration into 15 concentrated solar power plants of the calcium looping process based upon the reversible 16 carbonation/calcination of calcium oxide for thermochemical energy storage. An efficient 17 concentrated solar power-calcium looping integration would allow storing energy in the long 18 term by calcination of calcium carbonate thus overcoming the hurdle of variable power 19 generation from Solar. After calcination, the stored products of the reaction (calcium oxide and 20 carbon dioxide) are brought together in a carbonator reactor whereby the high temperature 21 exothermic reaction releases the stored energy for efficient power production when needed. 22 This work analyses several power cycle configurations with the main goal of optimizing the 23 performance of the overall system integration. Possible integration schemes are proposed in 24 which power production is carried out directly (using a closed carbon dioxide Brayton power 25 cycle) or indirectly (by means of a steam reheat Rankine cycle or a supercritical carbon dioxide 26 Brayton cycle). The results obtained show that the highest plant efficiencies (up to 45-46%) are 27 achievable using a closed carbon dioxide Brayton power cycle. 28 Keywords 29 Global warming, Renewable energies, Concentrated Solar Power (CSP), Thermochemical 30 energy storage (TCES), Calcium looping (CaL), Power cycles, Supercritical CO2 power cycle. 31 32 1. Introduction 33 The commercial expansion of renewable energy technologies is an urgent need to limit global 34 warming to “well below” 2.0°C (by 2100) and pursue 1.5°C above pre-industrial levels as was 35 at Paris COP21 Conference [1]. In particular, Concentrated Solar Power (CSP) should play a 36 leading role within the new energy landscape as it lends itself to potentially cheap storage of 37 energy in the form of heat [2]. Thus, efficient and affordable thermal energy storage systems 38 must be developed in order to decouple production and demand [3], which would allow a deep 39 penetration of solar energy power generation into the grid. 40 41 In recent years a large number of potential thermal storage technologies for medium to high 42 temperature CSP systems have been proposed [4] based upon three main concepts: i) sensible 43 Thermal Energy Storage (TES), such as direct steam storage [5] or molten salt systems [6,7]; 44 ii) latent heat storage using Phase Change Materials (PCMs), on which Zalba et al. [8] published 45 a comprehensive review of materials and applications; and iii) Thermochemical Energy Storage 46 (TCES). Regarding to TCES, a large number of potential systems [9], experimental research 47 2 under practical conditions [10] and TCES reactor designs [11] can be found in literature. 48 Essentially, TCES consists of using the heat obtained from an external source, such as CSP, to 49 drive an endothermic reaction. When energy is needed, the separately stored by-products of the 50 reaction are brought together at the necessary conditions for the reverse exothermic reaction to 51 occur, which releases the previously used heat for power production. The main advantages of 52 TCES as compared to TES and PCMs are the considerably high energy density attainable, 53 which is well above the energy density of molten salts currently used in commercial plants (~ 54 0.5 GJ/m3) [12], and the possibility of storing energy in the long term [9]. An extended review 55 on long-term solar heat storage can be found in ref. [13]. Moreover, in addition to the 56 chemically stored heat, sensible heat stored in the reaction by-products is also usable. 57 58 The focus of the present manuscript is on TCES in CSP tower plants. In order to achieve an 59 efficient and cost-effective thermochemical storage process, a proper selection of the reversible 60 reaction is a crucial issue. Among the possibilities explored for TCES in CSP tower plants at 61 large scale, one of the most promising systems is the Calcium Looping (CaL) process, which 62 relies on the carbonation-calcination reaction of calcium oxide (CaO) (Eq. (1)) [14]. The use of 63 several CaO precursors for TCES in CSP plants has been analysed in [15]. 64 65 CaO(s)+CO2(g) ⇄CaCO3(s) Hr0 = -178 kJ/mol (1) 66 Generally, the CaL process would begin with the decomposition of calcium carbonate (CaCO3) 67 particulate solids in a calcination reactor (calciner) yielding CaO and CO2 as by-products. A 68 high energy input is necessary to rise the solids stream temperature up to the value required for 69 the reaction to occur at a sufficiently fast rate and to carry out the endothermic calcination 70 reaction [16]. Thus, the optimum calcination temperature is essentially determined by the 71 composition of the gas in the calcination environment [17]. Once the sensible heat from the 72 calciner outlet streams (CaO and CO2 streams) is recovered, these products are separately 73 stored. Storage conditions and time are flexible and could be accommodated to energy demand 74 and environmental circumstances. When needed, the CaO and CO2 products are circulated into 75 a carbonator reactor, where energy is recovered from the carbonation reaction. 76 77 A great benefit of the CaL process is the low price (~10$/ton), wide availability and 78 harmlessness towards the environment of natural limestone and dolomites to be used as CaO 79 precursor [18]. However, a usually claimed drawback of the CaL process is the marked 80 deactivation of CaO derived from these natural minerals with the number of 81 carbonation/calcination cycles. CaO deactivation is indeed particularly relevant when the CaL 82 process is used for CO2 capture [19,20] under conditions that necessary involve regeneration of 83 CaO by high temperature (around 950°C) calcination at high CO2 partial pressure and 84 carbonation at low CO2 partial pressure (~0.15 bar). Nevertheless, CaL conditions to achieve a 85 high global efficiency for TCES and electricity generation in CSP plants are radically different 86 to those corresponding to its application for CO2 capture [21]. In the CSP-CaL integration, 87 carbonation would be carried out under high CO2 partial pressure and high temperature (around 88 or above 850ºC) whereas calcination would be ideally performed at relatively low temperature 89 (~700ºC) under a gas easily separable from CO2 such as Helium [17] or superheated steam [22]. 90 Under these conditions, CaO derived from natural limestone and dolomite may exhibit a high 91 value of the residual conversion [21]. 92 93 In addition to enhancing solar energy storage capacity, advanced high efficiency CSP-TES- 94 power cycle integrations should be developed exploiting energy storage conditions to achieve 95 3 a significant improvement of CSP plant performance. Integration of power cycles in 96 commercial CSP tower plants with thermal storage in the form of sensible heat using molten 97 salts is limited by a maximum temperature achievable around 550-600ºC. This limitation is 98 mainly imposed by the degradation of molten salts at higher temperatures [6,7]. In recent years, 99 molten alkali carbonates salts (MACs) have been investigated for energy storage. According to 100 Frangini et al. [23], temperature stability of additives limits the applicability of oxidizing MAC 101 salts at temperatures below 650 ºC. On the other hand, thermal radiation losses at the open focal 102 point [24] adds a further temperature limitation in currently CSP plants. This implies that most 103 of the commercial CSP tower plants currently under operation are based in Rankine cycle 104 process [25,26]. Peak solar to electricity conversion efficiencies in these commercial CSP tower 105 plants are around 25-30%, with an annual solar-to-electricity conversion efficiency lower than 106 20% [27]. At this regard, Liu et al. [28] presents current annual efficiencies as a function of 107 solar technology used: 13-15% for parabolic trough, 14-18% for tower ad 9-13% for Fresnel. 108 On the other hand, the European Academies Scientific Advisory Council [29] shows the 109 difference of annual solar to electricity efficiencies between conceptual (around 22-28%) and 110 industrial (around 14-18%) status. 111 112 This manuscript analyses several integration schemes to use the CaL process for TCES in CSP 113 plants. Integration models aimed at similar goals have been already investigated by other 114 authors. Tregambi et al. [30] proposed a scheme whereby calcination in the CaL process is 115 assisted by CSP for CO2 capture in a coal fired power plant. Edwards et al. [31] investigated a 116 CSP-CaL integration in which the heat produced in the carbonator reactor is used for power 117 generation through a CO2/air open cycle. This configuration assumes that the CO2 stream 118 entering into the carbonator reacts completely with the CaO solids to produce CaCO3. However, 119 attending to the reaction equilibrium, carbonation will be ceased when the CO2 partial pressure 120 in the carbonator reactor reaches the equilibrium partial pressure (see Equation 8). Thus, CO2 121 in the carbonator effluent gas will be unavoidably released to the environment in a CO2/air open 122 cycle at a concentration depending on the carbonator temperature. In order to guarantee the 123 absence of CO2 emissions, alternative power cycles must be employed. In this regard, 124 Chacartegui et al. [32] have recently proposed a CSP-CaL integration wherein the TCES system 125 is integrated with a closed CO2 power cycle directly coupled to the carbonator following a 126 pinch-analysis methodology [33]. In the discharge operation the circulating CO2 passes directly 127 to the carbonator and power turbine. The present manuscript explores the integration with the 128 TCES core system of alternative direct and indirect cycles (steam turbine, closed Brayton CO2 129 and indirect-supercritical CO2) for relevant CSP-CaL integration conditions. The obtained 130 results show that the highest efficiencies are achieved using direct cycles, potentially reaching 131 global power efficiencies above 44%. 132 133 2. CSP-CaL integration model 134 135 In this section the main aspects of the CSP-CaL integration model based on mass and energy 136 balance in heat exchangers, solid reservoirs, CO2 storage tank and reactors are summarized. 137 The interested reader is referred to [32] where the model is described in detail. Moreover, the 138 main CSP-CaL model simulation results are analysed as a previous step to discuss the power 139 cycle integration. 140 2.1. Model description 141 142 4 Figure 1 shows a schematic representation of the CSP-CaL integration model. The process starts 143 in the solar receiver, where solar energy input is used to carry out the calcination of CaCO3 144 (endothermic reaction). Currently commercial CSP tower systems would allow achieving 145 temperatures in the range of 700-900ºC, which are high enough to drive limestone calcination 146 in short residence times [21] using a solar calciner reactor among those already proposed in the 147 literature [34]. Thus, Meier et al. [35] have developed a solar multi-tube rotary kiln prototype 148 for carrying out the calcination reaction at temperatures up to 1100ºC. Once calcination takes 149 place, the released CO2 is sent to a storage tank after being cooled and compressed whereas the 150 CaO stream is transported to a solids reservoir. Both streams exiting the calciner at high 151 temperature are passed through a heat exchanger network to extract their sensible heat as a 152 previous step to storage at ambient temperature. This is a main advantage of the CSP-CaL 153 integration over current state of the art sensible heat storage using molten salts, whose 154 temperature must be kept above ~200°C to avoid solidification [36]. In order to use reasonably 155 sized CO2 storage volumes a minimum pressure of 75 bar is needed to store CO2 storage under 156 supercritical conditions (considering storage at ambient temperature) [32]. The high 157 compression ratio from calciner to storage conditions (1:75) requires the use of intercooling 158 compression to minimize the efficiency penalty. Solids transport can be carried out by means 159 of pneumatic conveying, an already mature technology to transport high temperature granular 160 solids [37]. For Ca based particles and a typical transport length of 200 m, an energy 161 consumption of 20 MJ/ton has been used in the CSP-CaL integration model [32]. 162 163 When power is needed the energy stored is released in the carbonator through the exothermic 164 carbonation reaction. According to thermodynamic reaction equilibrium carbonation can be 165 carried out at high temperature (>850ºC) under high CO2 partial pressure [38]. This would allow 166 a highly efficient generation of electricity thus overcoming temperature limits (T~550-600°C) 167 in currently commercial CSP plants with thermal storage in molten salts. Solids exiting the 168 carbonator are passed through another heat exchanger network to preheat the CaO and CO2 169 streams entering the carbonator. After the storage step, the CO2 stream is expanded to the 170 selected carbonator pressure, which must be below the storage pressure in order to use the 171 commercial fluidized bed technology. As can be seen in Figure 1, compression-expansion 172 process of CO2 before and after the storage step resembles a compressed air energy storage 173 (CAES) system [4]. Thus, besides of sensible and thermochemical energy storage, the 174 integration compresses orates energy storage also in the form of compressed gas with a round 175 trip efficiency of about 67% using a compression-expansion train (see [32] for further details). 176 177 178 Figure 1: CSP-CaL integration scheme 179 180 5 As can be seen in Figure 1, only a fraction of the total flow rate of CaO entering into the 181 carbonator (𝐹𝑅,𝑐𝑟𝑏) reacts with CO2 to produce CaCO3 (𝐹𝐶𝑎𝐶𝑂3,𝑐𝑟𝑏), remaining the rest as 182 unreacted CaO (𝐹𝐶𝑎𝑂,𝑢𝑛𝑟). The average CaO conversion (or activity) X determines the amount 183 of CaO converted to CaCO3 in the carbonator (𝑋=𝐹𝐶𝑎𝐶𝑂3,𝑐𝑟𝑏/𝐹𝑅,𝑐𝑟𝑏). On the other hand, the 184 carbonated particles entering into the calciner reactor are assumed to achieve a complete 185 decomposition, yielding one mole of CO2 (𝐹𝐶𝑂2,𝑐𝑙𝑐,𝑜𝑢𝑡) and one mole of regenerated CaO 186 (𝐹𝐶𝑎𝑂,𝑐𝑙𝑐) for each mole of CaCO3 (𝐹𝐶𝑎𝐶𝑂3,𝑐𝑙𝑐) according to equation (1). 187 188 The streams circulating in either the calciner or carbonator sides are decoupled. Thus, the solar 189 calciner only works in the daytime and under clear sky conditions whereas the carbonator 190 reactor must operate on demand during a 24h period, which requires a properly storage vessel 191 sizing. More sophisticated control strategies should be devised within a framework of long- 192 period control to ensure steady operation over time lags beyond 24h. Such control should rely 193 on meteorological forecasts and the power load curve. Thus, in order to guarantee a steady-state 194 operation, the mass-balance equation: 195 196 ∫𝐹𝐶𝑎𝐶𝑂3,𝑐𝑙𝑐(𝑡) 𝑑𝑡 24ℎ= ∫𝐹𝐶𝑎𝐶𝑂3,𝑐𝑟𝑏(𝑡) 𝑑𝑡 24ℎ (2) 197 must be satisfied. An average daytime period (∆𝑡𝑠𝑢𝑛) is assumed during which solar irradiation 198 is sufficiently intense to attain full calcination. In this case Equation (2) can be derived to obtain: 199 200 𝐹𝐶𝑎𝐶𝑂3,𝑐𝑙𝑐             ∙∆𝑡𝑠𝑢𝑛 =𝐹𝐶𝑎𝐶𝑂3,𝑐𝑟𝑏             ∙24 (3) For energy balance, the first thermodynamics law is applied to the carbonator and calciner 201 reactors: 202 203 ∑𝐹𝑖,𝑜𝑢𝑡ℎ𝑖,𝑜𝑢𝑡 𝑖−∑𝐹𝑖,𝑖𝑛ℎ𝑖,𝑖𝑛 𝑖=Φ−𝑊󰇗 (4) 𝐹𝑖,𝑜𝑢𝑡 −𝐹𝑖,𝑖𝑛 =𝜉𝜈𝑖 (5) where 𝜉 denotes the extent of reaction per unit time. Arranging and considering that output 204 conditions are reactor conditions, it is: 205 206 𝜉∆𝐻𝑅(𝑇𝑟𝑒𝑎𝑐𝑡)+∑𝐹𝑖,𝑖𝑛 𝑖(ℎ𝑖,𝑟𝑒𝑎𝑐𝑡 −ℎ𝑖,𝑖𝑛)=Φ−𝑊󰇗 (6) with 207 ∆𝐻𝑅(𝑇𝑟𝑒𝑎𝑐𝑡)=∑𝜈𝑖ℎ𝑖,𝑇 𝑖=∆𝐻𝑅 0+∑ 𝜈𝑖∫ 𝑐𝑝,𝑖𝑑𝑇 𝑇𝑟𝑒𝑎𝑐𝑡 𝑟𝑒𝑓 𝑖 (7) being the reaction enthalpy change at the reaction temperature. 208 6 2.2. Model results 209 210 The proposed CSP-CaL integration model has been simulated to assess the integration 211 efficiency. A sensitivity analysis has been carried out on relevant CaL cycle parameters such 212 as CaO average conversion X and carbonation equilibrium conditions [39]. Data used for the 213 reference case are reported in Table 1. 214 215 Table 1: CSP-CaL reference case simulation data 216 Net absorbed solar heat flux in calciner 100 MWt Thermal dispersions in carbonator 10 % Calciner temperature 900 °C Calciner pressure 1 bar Ambient temperature 20 °C CaO average conversion (X) 0.5 Carbonator temperature 875 °C Carbonator pressure 7 bar CO2 storage conditions 75 bar, T ambient Solid phase conveying energy consumption 20 MJ/ton Daylight hours (constant solar flux) 8h Isentropic efficiencies (compression/expansion) 0.89 217 CaO average conversion has a significant influence on the solids flow rates, storage vessels, 218 power production and consumption, and heat exchangers network configuration. Thus, a high 219 CaO conversion leads to a low fraction of unreacted CaO left, which affects relevantly the 220 plant’s performance. As the average CaO conversion increases the solids mass flow rate is 221 decreased (Figure 2), therefore energy consumption due to solids conveying is reduced. 222 223 224 Figure 2: Solids mass flow rate (left axis) and daily energy consumption (right axis) due to solids 225 conveying as a function of average CaO conversion (X) 226 7 227 Figure 3 shows the effect of the average CaO conversion on the thermal power effectively used 228 for energy storage. Keeping fixed a 100 MWth of CSP input (𝜙𝐶𝑆𝑃) into the system, the thermal 229 power used to carry out the calcination reaction (𝜙𝑐𝑎𝑙𝑐𝑖𝑛𝑎𝑡𝑖𝑜𝑛 =92 MWth ) does not depend 230 on the solids conversion in the carbonator while the rest (8 MWth) is employed to raise the 231 solids temperature before entering into the calciner. A part of the released power in the 232 carbonator 𝜙𝑟𝑒𝑙𝑒𝑎𝑠𝑒𝑑,𝑐𝑟𝑏 is used to increase the temperature of the inlet streams up to the 233 carbonation temperature, which leaves the rest of thermal energy available 𝜙𝑎𝑣𝑎𝑖𝑙𝑎𝑏𝑙𝑒 to be used 234 in the power cycle for electricity production. The difference between calcination and 235 carbonation power is due to thermal energy dispersions in the carbonator (10%). 236 237 Figure 3: Thermal power fluxes (left axis) and energy storage efficiency (right axis) as a function of 238 average CaO conversion in the carbonator. 239 240 Another relevant issue to be considered is that increasing the average CaO conversion allows 241 for an important reduction of the solids storage volumes as can be seen in Figure 4a. On the 242 other hand, a change in X does not yield a significant variation of the CO2 storage volume, 243 which is however quite sensitive to CO2 density as determined by storage pressure and 244 temperature (Figure 4b). 245 8 246 Figure 4: (a) Solids storage mass as a function of average CaO conversion. (b) CO2 storage volume as a 247 function of average CaO conversion for several storage conditions. 248 Previous works on the CaO multicycle conversion in the CaL process have been mostly focused 249 on Post-Combustion CO2 Capture (PCCC) conditions, either on laboratory-physical analysis 250 [20,40], reactor modelling [41] or process integration models [42], involving in all cases 251 carbonation under relatively low CO2 partial pressure (~ 0.15 bar) and calcination at very high 252 temperatures (~ 950ºC) under high CO2 partial pressure. Under these conditions the CO2 253 sorbent (CaO) presents a severe drop of conversion after a few cycles converging towards a 254 residual value of just about 0.07-0.08 [43,44]. Nevertheless, it is important to remark that the 255 CSP-CaL integration for thermochemical energy storage involves CaL conditions radically 256 diverse from those in the case of PCCC. Thus, thermogravimetric analysis (TGA) tests show 257 that the residual conversion of limestone derived CaO can be as large as 𝑋𝑟=0.5 for conditions 258 that correspond to the optimum CSP-CaL integration that involve carbonation at high 259 temperature under high CO2 partial pressure [21]. Moreover, according to TGA results fast 260 calcination may be achieved at a reduced temperature of just 700-725ºC under a gas which is 261 easily separable from CO2 (either He as in the TGA experiments described in [21] or 262 superheated steam [22]). Attaining such a low calcination temperature would allow the use of 263 already mature and inexpensive metallic solar receivers thus reducing technological risks. On 264 the other hand, the work of He/CO2 or H2O/CO2 separation should be also included in an 265 extended techno-economic energy analysis. 266 267 Carbonator conditions (pressure and temperature) are highly relevant for the global CSP-CaL 268 power cycle integration. Carbonator pressure is selected by considering the most favorable 269 conditions for the CaL-power cycle integration, i.e. a fluidized bed reactor operated under 270 atmospheric pressure if an indirect power cycle is integrated and a pressurized fluidized bed 271 reactor for direct integration with a power cycle, in order to achieve the higher integration 272 performance. On the other hand, increasing the carbonator temperature (𝑇𝑐𝑎𝑟𝑏) leads to higher 273 power cycle efficiencies and therefore enhances the CSP-CaL-power cycle integration 274 performance. However, the maximum temperature in the carbonator is limited by the 275 thermodynamic equilibrium of the carbonation/calcination reaction. Thus, for a given CO2 276 partial pressure in the carbonator there is a maximum carbonator temperature above which the 277 carbonation reaction is not thermodynamically favourable. According to thermochemical data 278 [38], the CO2 partial pressure for the reaction to be at equilibrium at a given temperature T(K) 279 is given by: 280 9 281 𝑃𝑒𝑞(𝑏𝑎𝑟)=𝑃·𝑦𝑒𝑞 =[4.137 107exp(−20474 𝑇)] (8) 282 In Eq. (8), 𝑦𝑒𝑞 is the fraction of CO2 in the carbonation environment. For a fixed carbonator 283 temperature, there is a minimum carbonator pressure below which the CO2 partial pressure is 284 insufficient for carbonation to occur. Figure 5 shows the minimum carbonator pressure as a 285 function of reactor temperature to carry out carbonation and for different CO2 fractions 286 (equation (8)). It is clear that operating under pure CO2 (𝑦𝐶𝑂2,𝑖𝑛 =1) allows working under 287 higher temperatures and low carbonator total pressures. 288 289 Figure 5: Minimum carbonator pressure as a function of carbonator temperature 𝑇𝑐𝑎𝑟𝑏 for several CO2 290 volume concentrations at the carbonator inlet (yCO2,in) 291 292 3. CSP-CaL-power cycle integration 293 294 This section is devoted to the study of several power cycle integrations into the CaL based CSP 295 storage system. Power cycles are classified in two categories: power cycles with direct 296 integration (CO2 regenerative Brayton cycle) and power cycles with indirect integration 297 (Rankine Reheat cycle and supercritical CO2 recompression cycle). 298 299 3.1. Direct integration 300 301 In power cycles with direct integration the heat transfer fluid used in the carbonator is sent 302 directly into a gas turbine. In the following a CO2 closed Brayton cycle is analysed. 303 304 16 416 Figure 11: Efficiency of the CaL-Rankine integration as a function of carbonator pressure and for diverse 417 temperatures. Note that a minimum carbonator pressure is required according to thermochemical 418 equilibrium as temperature is increased. 419 420 A sensitivity analysis has been carried out in which the main Rankine cycle parameters have 421 been tuned. As can be seen in Figure 12, the global integration efficiency is promoted by 422 increasing live steam conditions (pressure (Pvv) and temperature (Tvv)). It may be also seen that 423 efficiency is enhanced as the reheat temperature is increase. 424 425 Figure 12: Global integration efficiency as a function of (LEFT) steam turbine inlet and (RIGHT) reheat 426 temperature conditions 427 As seen in Figure 10, the preheat water of the Rankine cycle is heated by the exhaust CO2 428 stream from the carbonator in a heat recovery steam generator (HRSG) until super-heated status 429 is reached. One key parameters in Rankine power cycles is the HRSG efficiency, which can be 430 analysed from the pinch point value across the steam production process. Figure 13 shows that 431 lower values of the pinch point (higher HRSG efficiency) causes an increase in the global cycle 432 efficiency. 433 17 434 Figure 13: Global integration efficiency as a function of minimum temperature approach in HRSG 435 436 b Supercritical CO2 (sCO2) recompression cycle 437 438 The supercritical CO2 (sCO2) Brayton cycle, which was originally introduced by Feher [48], 439 has emerged in the last years as a promising technique for high-efficiency power production. It 440 basically consists of a closed-loop Brayton cycle that operates entirely above CO2 critical 441 pressure (73.77 bar and 30.98ºC) and presents a high drop in compressibility, which brings 442 about a similar reduction in compression work while the turbine operates with CO2 in a close 443 to ideal behaviour. Among different layouts proposed for sCO2 cycles, a recompression scheme 444 seems to be the highest efficiency cycle [49], which is thus the one used in the present study. 445 Figure 14 shows the recompression cycle model. An important feature of the regeneration 446 process in the sCO2 Brayton cycle is that the specific heat of the cold side is 2-3 times higher 447 than the hot side. Thus, the CO2 stream is split (stream 5b in Figure 14a) to compensate for the 448 specific heat difference in the low temperature recuperator, which maximizes the heat 449 recuperation. 450 451 CSP-sCO2 integration models have been already developed to increase the CSP power plant 452 performance. Thus, Chacartegui et al. [50] compared the integration of supercritical and 453 transcritical carbon dioxide cycles as well as a combined cycle composed by a topping Brayton 454 CO2 cycle and a bottoming Organic Rankine Cycle. Iverson et al. [51] presents the behaviour 455 of Brayton cycle turbomachinery including a data set for stable supercritical CO2 Brayton cycle 456 operation. Moreover, Ma et al. [52] analyses the integration of sCO2 power cycles by 457 considering sensible heat storage (thermocline system). One of the most important advantages 458 of the sCO2 Brayton cycle is its compact turbomachinery, albeit it is still under development 459 [53]. A sCO2 technology review is presented in [45], from which values on turbomachinery 460 efficiency and pressure drops are taken in the present work. Thus, the recompression sCO2 cycle 461 has been simulated using data specified in table 3. 462 Table 4: Input data parameters for the sCO2 cycle [45] 463 𝜂𝑐 (%) 85 𝜂𝑡 (%) 90 𝜀𝑟𝑒𝑐 (%) 95 ∆𝑃𝑅,ℎ𝑜𝑡 (%) 0.5 18 ∆𝑃𝑅,𝑐𝑜𝑙𝑑 (%) 1.5 ∆𝑃𝑅,𝐻𝐸 (%) 0.5 464 Figure 14a shows the recompression sCO2 Brayton scheme proposed. Thermodynamic 465 parameters of the streams involved in the cycle are shown in Figure 14b in a temperature (T)- 466 entropy (S) diagram. Main sCO2 cycle simulation results are shown in Tables 5 and 6. A cycle 467 efficiency of around 41% is obtained from this configuration, which is in agreement with results 468 from previous works [45,54]. 469 470 471 472 473 Figure 14: (a) Base recompressionsCO2 Brayton cycle layout. (b) Temperature-entropy 474 diagram 475 476 Once the recompression sCO2 cycle is analysed and benchmarked it is integrated in the CSP- 477 CaL configuration as shown in Figure 15. A global integration efficiency close to 32% is 478 achieved, although it must be taken into account that a large amount of energy linked to the 479 cooling process before the compression stage is not used. This suggests that a bottoming cycle 480 could serve to improve the cycle performance [50]. 481 19 Table 5: Main sCO2 cycle simulation results 482 𝑄𝑖𝑛𝑝𝑢𝑡 50 MWth 𝑊𝑐𝑜𝑚𝑝,1 7.14 MWe 𝑊𝑐𝑜𝑚𝑝,2 5.74 MWe 𝑊𝑡𝑢𝑟𝑏 33.38 MWe 𝜂𝑐𝑦𝑐𝑙𝑒 41.01 % 𝜙𝑐𝑦𝑐𝑙𝑒 69% 483 Table 6: Stream data for sCO2 recompression cycle 484 𝑆𝑡𝑟𝑒𝑎𝑚 𝑚󰇗(𝑘𝑔/𝑠) 𝑇 (º𝐶) 𝑃 (𝑏𝑎𝑟) 1 223.6 650 213.9 2 223.6 521.1 78 3 223.6 200.2 77.61 4 223.6 117.2 77.22 5 158.8 35 75 5b 64.85 117.2 77.22 6 158.8 113.1 225 7 158.8 166.0 223.9 8 64.85 229.8 221.6 9 223.6 183.1 221.6 10 223.6 471 218.3 485 486 487 Figure 15: CSP-CaL- sCO2 integration scheme and main simulation results 488 20 Results from a sensitivity analysis are shown in Figure 16. It is observed that the recuperation 489 process in the sCO2 Brayton cycle greatly influences the thermal efficiency since CO2 490 properties are very sensitive to pressure and temperature near the critical point. Therefore, the 491 hot and cold sides in the regenerator are strongly unbalanced. As can be seen in Figure 16, by 492 increasing the turbomachinery efficiency (which depends upon further technology 493 development) the global cycle performance is significantly enhanced. 494 495 496 Figure 16: Sensitivity analysis results of the CSP-CaL- sCO2 integration. 497 498 c Combined cycle 499 500 The combined cycle is based on the integration of two subsystems consisting of a gas turbine 501 (Brayton cycle) and a steam turbine (Rankine cycle), which leads to an improvement of 502 efficiency due to the synergy of both cycles [55]. 503 504 A number of integrated solar combined cycle (ISCC) systems have been proposed to improve 505 the power plant efficiency [56,57]. ISCCS power plants currently in operation employ the 506 parabolic trough concentrator technology. Further work is still needed to advance in the 507 technological readiness of solar tower – ISCC power plants [58]. ISCC cycles operate using a 508 solar-fuel combination [59,60], with the gas turbine being fuelled by a non-solar source (based 509 on fossil or renewable fuel) due to the temperature limitation in CSP power plants imposed by 510 degradation of molten salts and thermal radiation losses at the focal point. Solar power share in 511 ISCC power plants is on average below 34% [58]. Compared with the solar-only power plants, 512 ISCC plants exhibit several advantages such as higher solar-to-electricity conversion 513 performance. Moreover, thermal inefficiency associated with the daily start-up and shutdown 514 of the steam turbine can be avoided [61]. Another configuration proposed in a recent work [62] 515 evaluates a combined cycle based on a closed Brayton and organic Rankine cycle for solar 516 power tower plants by means of energy and exergy analysis, showing that higher performance 517 than using steam and supercritical CO2 cycles can be achieved. 518 519 0.29 0.3 0.31 0.32 0.33 0.34 -35% -25% -15% -5% 5% 15% 25% 35% h_integration Variation P[05] T[05] dT_min_HE PR_c X_spl eta_turb eta_comp ϵ_rec ΔP_rec_h ΔP_rec_c ΔP_HE T[01] 21 Figure 17 shows the global cycle integration proposed by considering a combined cycle for 520 power production. The combined cycle involves a hybrid direct-indirect power cycle 521 integration with the CSP-CaL system. The CO2 stream exiting the carbonator is expanded in a 522 gas turbine as a previous step for transferring heat to steam cycle through a HRSG. Main 523 simulation results are shown also in Figure 17. 524 525 526 Figure 17: CSP-CaL- CC integration scheme and main simulation results 527 528 Figure 18 shows the global cycle performance as a function of the carbonator pressure (or, 529 equivalently, the inlet turbine pressure) for different values of the turbine outlet pressure. As 530 can be seen, a higher performance is obtained by decreasing the outlet turbine pressure, reaching 531 a maximum value of 40.4% for operation under an inlet/outlet turbine pressure ratio of 3.6/1. 532 In order to simplify the heat exchanger network, an atmospheric outlet turbine pressure will be 533 next considered. 534 535 Figure 18: Efficiency of the CSP-CaL-CC integration as a function of the carbonator pressure for several 536 values of the Brayton turbine outlet pressure. 537 22 4. Comparative analysis on the CaL-CSP-power cycle integrations 538 In order to compare the performances of the diverse CSP-CaL-power cycle integrations a 539 sensitivity analysis has been carried out using reference parameters for which the efficiency of 540 these integrations is optimized. Main power cycle parameters for each integration scheme are 541 given in Table 7. A carbonator temperature of 875ºC has been selected, which guarantees 542 carbonation under the different carbonator pressures of each cycle. Unless otherwise indicated, 543 the values of the parameters employed for the CaL cycle are those previously specified in Table 544 1. Note that temperatures for steam turbine in the case of the combined cycle are conditioned 545 by the Brayton turbine exit and therefore the values shown in this table correspond to the 546 maximum temperatures achievable. 547 548 Table 7: Main power cycle parameters for each integration scheme 549 550 Closed CO2 Brayton Reheat Rankine sCO2 Recompression Combined cycle 𝑃𝑜𝑢𝑡,𝑡𝑢𝑟𝑏 (𝑏𝑎𝑟) 1 - - 1 𝑃𝑐𝑎𝑟𝑏 (𝑏𝑎𝑟) 3.2 1 1 3.2 𝑃𝑣𝑣 (𝑏𝑎𝑟) - 160 - 160 𝑇𝑐𝑎𝑟𝑏 (º𝐶) 875 875 875 875 𝑇𝑣𝑣 (º𝐶) - 540 - 540 𝑇𝑟𝑒ℎ𝑒𝑎𝑡 (º𝐶) - 540 - 540 ∆𝑇𝑚𝑖𝑛 - 10 10 10 𝑃[05](𝑏𝑎𝑟) - - 75 - 𝑇[05](º𝐶) - - 32 - 𝑃𝑅 3.2 - 3 3.2 𝑋 0.5 0.5 0.5 0.5 551 Figure 19 shows the global integration efficiency obtained for the different power cycles 552 analysed in this work as a function of the carbonator pressure. As can be seen, the CO2 closed 553 cycle direct integration yields the best efficiency results. Only by means of the indirect 554 integration is possible to operate the carbonator under atmospheric pressure, being the 555 efficiency hampered in this integration as the carbonator pressure is increased further. The 556 opposite trend occurs in the CO2 closed and CC power cycles. Using these power cycles, the 557 global efficiency is promoted as the carbonator pressure is increased up to a certain optimum 558 value, which is around 4.2 bar for the CO2 closed cycle and 5.1 bar for the CC cycle 559 (atmospheric turbine outlet pressure). Results show also that despite sCO2 recompression cycle 560 could be a potentially attractive choice from a thermodynamic point of view, the conservative 561 values used for the turbomachinery efficiencies (in accordance with the current state of art 562 [45,47]) prevents the CSP-CaL-sCO2 cycle integration from reaching very high global 563 efficiencies. Efficiency results are plotted in Figure 20 as a function of the CaO average 564 conversion. Generally, the enhancement of CaO conversion promotes efficiency as would be 565 expected. 566 567 23 568 569 Figure 19: Global cycle integration efficiency as a function of carbonator pressure for the different power 570 cycles coupled to the CSP-CaL system (using data showed in Table 7). 571 572 573 574 Figure 20: Global cycle integration efficiency as a function of average CaO conversion for the different 575 power cycles coupled to the CSP-CaL system (using data showed in Table 7). 576 577 Additional considerations regarding costs must be addressed to further assess the applicability 578 of these power cycle integrations in the CSP-CaL. As this technology is in an early concept 579 stage, data from prototypes or experimental installations are not available for the TCES core. 580 0.25 0.27 0.29 0.31 0.33 0.35 0.37 0.39 0.41 0.43 1.00 6.00 11.00 η integration p_carb CO2 closed Rankine Reheat sCO2 recompression Combined Cycle 0.27 0.29 0.31 0.33 0.35 0.37 0.39 0.41 0.43 0.45 0.20 0.30 0.40 0.50 0.60 0.70 0.80 η integration X CO2 closed Rankine Reheat sCO2 recompression CC regenerative 24 A detailed economic prospective analysis is under development and will be presented in next 581 works. However, some preliminary considerations can be already made by extrapolating 582 components information from other technologies. In commercial CSP plants, the power block 583 cost percentage is estimated around 32% [63] and power cycle integration has a critical 584 influence on capital investment. Some considerations may be made also on the maturity of the 585 power technologies analysed in this work. These include one full mature technology (steam 586 power cycle), two fully feasible power technologies with already available commercial 587 components (real gas CO2 closed cycle and the derived combined cycle) and a promising power 588 technology with great advances expected (supercritical CO2 cycle). 589 590 Steam power cycles fulfilling specific conditions for their integration in CSP plants (optimized 591 for complex and challenging cycle conditions) are at the commercial level [64]. For 592 conventional steam power cycles, capital and O&M costs can be estimated as 1280 $/kW and 593 5.7 $/MWh, respectively [65]. For the CO2 closed-cycle gas turbine, although not fully 594 available, main components are already usable or can be integrated from different applications. 595 Thermal turbomachinery, compressors and turbines are already in use at commercial scale using 596 as working fluid air, e.g. the Gelsenkirchen plant [66] and Oberhausen I [67], or Helium, e.g. 597 Oberhausen II [67]. In the case of CO2 as working fluid, compressors are being widely analysed 598 and tested in recent years as fundamental equipment within the Carbon Capture and Storage 599 (CCS) technology [68,69]. Thermal machinery characteristics for the pressure ratios and 600 temperatures presented in this work will be quite similar to the ones operating with air and 601 combustion gases, as shown by Najjar et al. [70] for compressor pressure ratios of 5, compressor 602 temperature inlet of 310 K and turbine temperature inlet of 1100 K. Therefore, already available 603 technologies for turbines and compressors could be employed. In the case of closed-Brayton 604 cycles, the introduction of additional heat exchangers increases capital investment [50]. 605 Therefore, technologies for the conditions presented in this work are already available, and 606 costs for the fully developed technology could be expected as similar to the ones in the range 607 between current gas turbines and combined cycles. For a 100 MWe power block estimated 608 capital and total O&M costs for an open gas turbine combined cycle (considering only the 609 power block) are around 660 $/kW and 2.2 $/MWh, respectively [65,71] whereas for an air- 610 open Brayton cycle they can be estimated as 1026 $/kW and 3.42 $/MWh respectively [72]. 611 Finally, in the case of supercritical CO2 technology, in spite of that sCO2 cycle is a non-mature 612 technology, in a project under development granted by US DOE through the Sunshot initiative 613 [73] a power plant investment cost of 1200 $/kW at the commercial stage is assumed [74,75]. 614 615 Regarding the CaL thermochemical energy storage system, it implies an intrinsic benefit 616 regarding life cycle cost and system sustainability as it is based on the use of low price, non- 617 toxic and widely available natural CaO precursors such as limestone and dolomite. According 618 to [76], the use of the CaL process for TCES would make it possible to achieve a thermal storage 619 cost lower than 15$/kWht. This preliminary approach shows the potential of these integrations. 620 Detailed and fully developed life cycle and economic analysis are under development and will 621 be the subject of future works. 622 5. Conclusions 623 This manuscript analyses several CSP tower plant integration schemes with thermochemical 624 energy storage (TCES) using the Calcium-Looping (CaL) cycle. The work is focused on 625 assessing the power production cycle. The CSP-CaL integration yields high temperatures 626 (above 850ºC) at the power cycle inlet, which allows using high efficiency power cycles 627 employed in fuel based power plants (or combined CSP-fuel power plants). Thus, the CSP-CaL 628 integration achieves high density/long term storage capacity and lends itself for the integration 629 25 of higher performance power cycles as compared with the current state of the art in commercial 630 CSP plants. 631 632 In regards to direct-indirect cycles integration, results show that higher performance is achieved 633 for direct integration. On the other hand, higher efficiencies are attained as the average CaO 634 conversion is increased. Among the power cycles analysed in the present work, the CO2 closed 635 Brayton cycle shows the best overall performance, reaching efficiencies potentially above 44- 636 45% (including penalty for solids conveying) if the carbonator is operated at temperatures 637 around 950ºC and under pressures about 3.5 bar for atmospheric pressure at Brayton turbine 638 outlet. Importantly, carbonation conditions in this integration allows for high values of the 639 residual conversion of CaO derived from natural minerals such as limestone and dolomite as 640 recently demonstrated by thermogravimetric studies. The wide availability, abundance, lack of 641 corrosiveness, non-toxicity and cheapness (~10$/ton) of these natural minerals makes the 642 proposed integration an attractive technology for large-scale storage of solar energy and highly 643 efficient grid-level power production. 644 645 Notation 646 647 𝑐𝑝,𝑖 Specific heat, kJ/(kmol·K) 𝑃𝑜𝑢𝑡,𝑡𝑢𝑟𝑏 outlet turbine pressure, bar 𝑑𝑇_min _𝐻𝐸 Minimum temperature approach, ºC 𝑦𝑒𝑞 equilibrium fraction of CO2 in the carbonator 𝐹𝑖 molar flow rate of component i, kmol/s 𝑇 Temperature, K 𝐹𝐶𝑎𝐶𝑂3 molar flow rate of CaCO3 𝑇𝑐𝑎𝑟𝑏 Carbonator temperature, K 𝐹𝐶𝑎𝐶𝑂3,𝑐𝑎𝑟𝑏 molar flow rate of CaCO3 (carbonator side) 𝑇𝑖𝑛,𝐶𝑂2 CO2 temperature at carbonator inlet, ºC 𝐹𝐶𝑎𝐶𝑂3,𝑐𝑙𝑐 molar flow rate of CaCO3 (calciner side) 𝑇𝑖𝑛,𝐶𝑎𝑂 CaO temperature at carbonator inlet, ºC 𝐹𝐶𝑎𝑂,𝑐𝑟𝑏 molar flow rate of CaO (carbonator side) 𝑇𝑟𝑒ℎ𝑒𝑎𝑡 Reheat temperature (Rankine cycle), ºC 𝐹𝐶𝑎𝑂,𝑐𝑙𝑐 molar flow rate of regenerated sorbent 𝑇𝑣𝑣 Live steam temperature, ºC 𝐹𝐶𝑎𝑂,𝑢𝑛𝑟,𝑐𝑎𝑟𝑏 molar flow rate of unreacted CaO (carbonator side) 𝑊𝑐𝑜𝑚𝑝,1 compressor 1 power, sCO2, MW 𝐹𝐶𝑎𝑂,𝑢𝑛𝑟,𝑐𝑙𝑐 molar flow rate of unreacted CaO (calciner side) 𝑊𝑐𝑜𝑚𝑝,2 compressor 2 power, sCO2, MW 𝐹𝐶𝑂2,𝑐𝑙𝑐,𝑜𝑢𝑡 molar flow rate of CO2 at calciner outlet 𝑊𝑡𝑢𝑟𝑏 turbine power, sCO2, MW 𝑊󰇗 mechanical power, kW 𝐹𝑅,𝑐𝑎𝑟𝑏 recirculating molar flow rate (carbonator side) 𝑋 CaO conversion 𝑋 𝐹𝑅,𝑐𝑙𝑐 recirculating molar flow rate (calciner side) 𝑋𝑠𝑙𝑝 Split factor, sCO2 ℎ𝑖 Enthalpy, kJ/kmol 𝜂𝑐 isentropic compressor efficiency HXA solid-solid heat exchanger 𝜂𝑡 isentropic turbine efficiency HXB gas-solid heat exchanger 𝜂𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑡𝑖𝑜𝑛 Global integration performance HXE gas-solid heat exchanger 𝜂𝑠𝑡𝑜𝑟𝑎𝑔𝑒 storage performance HXF gas-solid heat exchanger 𝜂𝑐𝑦𝑐𝑙𝑒 power cycle performance HXI gas-solid heat exchanger 𝜂𝑐𝑐 Combined cycle performance HXG gas-solid heat exchanger 𝜙𝑐𝑦𝑐𝑙𝑒 power cycle practicability 𝑚󰇗𝐶𝑂2,𝑐𝑟𝑏 CO2 mass flow rate through carbonator ∆𝑡𝑠𝑢𝑛 average daytime period 𝑚󰇗𝑠𝑜𝑙𝑖𝑑𝑠 solids mass flow rate, kg/s ∆𝐻𝑅(𝑇𝑟𝑒𝑎𝑐𝑡) heat of reaction at the reactor temperature 𝑃𝑐𝑎𝑟𝑏 absolute carbonator pressure, bar 𝜉 extent of reaction per unit time