Energy and Management: X Integration of gas switching combustion and membrane reactors for exceeding 50% e ﬃ ciency in ﬂ exible IGCC plants with near-zero CO 2 emissions

Thermal power plants face substantial challenges to remain competitive in energy systems with high shares of variable renewables, especially in ﬂ exible integrated gasi ﬁ cation combined cycles (IGCC). This study addresses this challenge through the integration of Gas Switching Combustion (GSC) and Membrane Assisted Water Gas Shift (MAWGS) reactors in an IGCC plant for ﬂ exible electricity and/or H 2 production with inherent CO 2 capture. When electricity prices are high, H 2 from the MAWGS reactor is used for added ﬁ ring after the GSC reactors to reach the high turbine inlet temperature of the H-class gas turbine. In periods of low electricity prices, the turbine operates at 10% of its rated power to satisfy the internal electricity demand, while a large portion of the syngas heating value is extracted as H 2 in the MAWGS reactor and sold to the market. This product ﬂ exibility allows the in ﬂ exible process units such as gasi ﬁ cation, gas treating, air separation unit and CO 2 compression, transport, and storage to operate continuously, while the plant supplies variable power output. Two con ﬁ gurations of the GSC-MAWGS plant are presented. The base con ﬁ guration achieves 47.2% electric e ﬃ ciency and 56.6% equivalent hydrogen production e ﬃ ciency with 94.8 – 95.6% CO 2 capture. An advanced scheme using the GSC reduction gases for coal-water slurry preheating and pre-gasi ﬁ cation reached an electric e ﬃ ciency of 50.3%, hydrogen e ﬃ ciency of 62.4%, and CO 2 capture ratio of 98.1 – 99.5%. The e ﬃ ciency is 8.4%-points higher than the pre-combustion CO 2 capture benchmark and only 1.9%-points below the unabated IGCC benchmark.


Introduction
As highlighted by the Intergovernmental Panel on Climate Change [1], Carbon Capture and Storage (CCS) will play a vital role in reaching the climate change targets of restricting global warming to 1.5°C above pre-industrial levels. Based on [2], CCS will be responsible for approximately 9% of the cumulative emissions reduction until 2050, amounting to approximately 2.8 billion tonnes per annum of CO 2 stored with around half of that amount originated in the power sector. Furthermore, CCS offers the possibility to balance the mitigation of carbonaceous emissions with economic growth. As pointed out in [3], the availability of low carbon emission energy solutions such as thermal power plants with CCS can result in electricity costs up to 62% lower than using renewables alone. Therefore, CCS, as a proven and wellunderstood technology, is pivotal to enable a transition to a low emissions economy for countries currently reliant on carbon-intensive electricity generation systems.
Amongst the different technologies available for carbon sequestration in thermal power plants, chemical looping combustion (CLC) proposed by Ishida et al. [4] promises high degrees of CO 2 capture and attractive economics [5]. This technology consists of carrying out the combustion of fuel by reducing a metallic oxygen carrier in a fuel reactor, which is later transported to an air reactor, subsequently reacting with oxygen and releasing heat utilized in a power cycle. Thus, an inherent CO 2 separation is achieved, minimizing the energy penalty of CO 2 capture relative to conventional abatement strategies [6]. However, progress on scale-up of dual interconnected fluidized bed reactors of gas-fuelled CLC systems at pressurized conditions has been slow [7]. High-pressure operation is a requisite for high power cycle efficiency. To overcome this problem, while ensuring high load flexibility of the power cycle, the gas switching reactor concept operating with fluidized beds was introduced in previous work by the authors [8]. This concept keeps the oxygen carrier in a single reactor where it is sequentially exposed to air and fuel streams through a valve switching mechanism, avoiding the need for solids circulation. A small decrease in capture ratio is observed resulting from the undesired mixing of outlet streams during valve switch from reduction to oxidation stage [9]. The thermodynamic efficiency of CCS plants based on GSC is noticeably hampered if the oxygen carrier material temperature limit is low [10]. Although several studies assume a maximum value of up to 1200°C [11][12][13], it is still substantially below the firing temperatures achievable by modern gas turbine technologies [14]. Our previous work [13] showed that a significant efficiency benefit is attained when by-passing a portion of the syngas fuel to an extra combustor, which enables an increase in the turbine inlet temperature (TIT) values beyond CLC limits. This TIT increase can also be achieved using natural gas instead, raising the temperature of the oxidation stage outlet stream from the GSC cluster to the Combustor Outlet Temperature (COT) of the modern gas turbine. Nonetheless, a substantial reduction of the carbon capture rate resulted from either of these strategies.
From the variety of thermal power plants using solids fuels, Integrated Gasification Combined Cycles (IGCC) have the potential to reach the highest efficiencies, due to the high temperatures achieved through a combined (Brayton + Rankine) power cycle which lead to a higher Carnot efficiency relative to the standalone Rankine power cycle of Pulverized Coal Boilers (PCB). This effect is more marked as advanced gas turbine technologies with firing temperatures above 1500°C are widely deployed, and operational difficulties are overcome. These advances result in efficiencies exceeding the most modern ultra-supercritical boilers [15], which require special materials (nickel-based alloys) that result in comparatively higher costs. Additionally, IGCC plants present the lowest environmental footprint due to the possibility of removing harmful contaminants concentrated in the small flow rate of syngas generated after gasification [16]. In parallel to this, the concentrated CO 2 at high pressure contained in the syngas allows for easier and less energy-intensive CO 2 capture compared to plants such as coal-fired boilers where post-combustion removal systems from the CO 2 diluted, low-pressure exhaust stream are applied. In particular, hightemperature syngas desulphurization or hot gas clean up (HGCU), appears to be a compelling technology to further boost IGCC power plant efficiencies [17], eliminating the energy penalty associated with large temperature swings of cooling syngas to ambient conditions for treating and subsequent reheating of the fuel before firing in a Gas Turbine (GT). Zinc Oxide (ZnO) sorbents show favourable thermodynamics to reduce sulphur components in the syngas to ppm levels [18] over a wide range of operating temperatures.
In addition to high efficiency, low cost, and low environmental impact, an increasingly important requirement of thermal power plants is the balancing of the fluctuating power output of cheap variable renewable energy (VRE) in the form of wind and solar power. The fluctuating generation profile of wind and solar is incompatible with baseload power generation, creating an optimal power mix composed of VRE backed up primarily by mid-load power plants that supply electricity during times of limited wind and sun [19]. The capital underutilization inherent to this power mix imposes substantial system-level costs, also known as profile costs [19]. These costs increase sharply with the capital cost of the underutilized capital, which is problematic for capital-intensive clean power plants such as nuclear, biomass, and CCS. For this reason, CCS and variable renewables are generally seen as competitors rather than complements [20].
Aside from the economic challenges related to the deployment of CCS plants as mid-load generators to balance VRE, significant technical constraints also exist. In particular, the IGCC power plants targeted in this study are highly inflexible due to the gasification train consisting of Reaction°A mbient/reference conditions R Retentate P Permeate several process units in series designed to operate in continuous fullload mode. In addition, intermittent CO 2 influxes into downstream CO 2 transport and storage infrastructure also poses technical challenges [21]. The original contribution of the present study is the development of a novel IGCC configuration that overcomes these technical and economic challenges related to CCS power plants in power systems with increasing shares of VRE. The key feature of this plant is the steadystate utilization of the gasification train and all equipment related to CO 2 compression, transport, and storage, despite flexible power production for balancing VRE. In this way, the coal-fired power plant proposed in this study will achieve similarly large system-level benefits to the natural gas-fired gas switching reforming plant for flexible power and hydrogen production previously evaluated by one of the authors [22]. In addition, the proposed power plant configuration facilitates added firing after the GSC reactors to maximize efficiency without compromising CO 2 avoidance.
The next section outlines the novel flexible power and hydrogen concept proposed in this study. Subsequently, the methodology for integrated reactor and power plant modelling is described, and the detailed flexible plant layout is provided together with suitable benchmark plants. Then, the technical performance of the proposed flexible GSC-IGCC power and hydrogen configuration is presented both in power and hydrogen production modes. Thermodynamic efficiencies and CO 2 emissions are benchmarked consistently against an unabated IGCC plant and a pre-combustion CO 2 capture IGCC plant, which utilize a modern H-class GT in the power cycle and HGCU for contaminant removal, as opposed to F-class turbines and cold gas clean-up used in prior studies with IGCC plants with and without CCS [16,23,24]. In addition, an advanced heat integration configuration is also investigated, enabling net electric efficiencies exceeding 50% with CO 2 capture above 98%, while preserving a high level of flexibility. Finally, the main technology gaps that need to be closed to realize the promising performance of the plants investigated in this study are discussed, and conclusions are drawn to guide future work.

Proposed power plant concept
A simplified layout of the flexible clean power and hydrogen plant proposed in this work is provided in Fig. 1. Relative to previous work on the GSC-IGCC power plant [13], the primary modification is the addition of a membrane-assisted water-gas shift (MAWGS) reactor [25]. This reactor separates out a fraction of the syngas heating value as pure hydrogen after the gas clean-up unit. The Pd-based membrane used in this work, with very high H 2 selectivity, is alloyed with other metals to avoid surface poisoning from CO and H 2 O, and has a suitable operating window for temperatures resulting from HGCU [26]. Depending on the electricity price at the time, this hydrogen can then be used either for added firing after the GSC reactors for high efficiency power production or directly exported to the market (green diamond in Fig. 1). Other configurations, such as the three reactor chemical looping system [27], can also produce power and hydrogen, but do not offer the same level of flexibility.
When the plant is operating in hydrogen production mode, the pressure inside the membranes of the MAWGS reactor is reduced, and more steam is added to the syngas to maximize H 2 extraction. The gas turbine is ramped down to 10% of its nominal load with no added H 2 firing to enable power production from the heating value remaining in the low-grade syngas exiting the MAWGS reactor. Such a very low gas turbine load is possible because of the flameless combustion in the GSC reactors that does not introduce any NOx formation and incomplete fuel combustion issues at low turbine load.
In both these operating modes, the operation of the gasification train and the downstream CO 2 transport and storage infrastructure remains almost unchanged, avoiding the considerable technical and economic challenges with flexible CCS power production described earlier. It is also noted that a steady-state operating point anywhere between full power and full hydrogen mode would also be possible, resulting in combined power and hydrogen production.
A more advanced configuration of the plant shown in Fig. 1 is also investigated in this study where the hot GSC reduction outlet gases are used to evaporate and pre-gasify a coal slurry in a pre-gasification heat exchanger. The pre-gasified slurry (syngas with entrained coal particles) is then fed to a High Temperature Winkler (HTW) gasifier [28] at a temperature close to the gasifier operating temperature. This pre-gasification substantially reduces the heat demand in the gasifier, bringing a large increase in cold gas efficiency (CGE) and reduction in air separation unit (ASU) power consumption. The added complexity of this pre-gasification heat exchanger is to some extent mitigated by replacement of the lock hopper system for coal loading with simpler slurry pumps.

Model integration
The power plant models were elaborated in UniSim Design R451 [29] using the Peng Robinson equation of state for thermodynamic property calculation. The steam cycles employ ASME steam tables. GSC reactors were modelled in Scilab 6.0 [30], assuming the behaviour of a continuous stirred tank reactor (CSTR) to represent fluidized conditions, similarly to Cloete et al. [9]. The MAWGS reactor was also modelled in Scilab, assuming the behaviour of a plug flow reactor (PFR) with heat and mass diffusion effects. The kinetic rate of the WGS reaction was taken from Hla et al. [31], while a diffusion equation employed in Fernandez et al. [32] was used for H 2 permeation. The codes were two-way coupled to UniSim by means of a CAPE-OPEN unit operation. The GSC Scilab code solves the transient reactor profiles of temperature, flow, and composition and provides time-averaged values of temperatures and stage outlet mixing degrees to the stationary power plant model. The MAWGS Scilab model delivers the retentate and permeate reactor products provided a syngas stream. The gas turbine was modelled with the GS-code from the Energy Department of Politecnico di Milano, extensively used in the past to evaluate the performance of different power plants, e.g. [12,33,34], with the capability to accurately determine coolant flow requirements for different hot gas path conditions and compositions. The Patitug thermodynamic database from the Energy Department of the Universidad Politécnica de Madrid was used for property estimation within the Scilab codes assuming ideal gas behaviour, which is an acceptable simplification due to the high temperature and relatively low pressures encountered in these units. The change of fluid property package from one platform to another caused a relative mass & energy balance error below 0.1%. The solids properties used in the GSC reactors and HGCU model were predicted with correlations for enthalpy and specific heat obtained from data tables from [35,36].
The solving sequence to converge the MAWGS-GSC-GT loop was as follows: at a certain coal flow rate the syngas fraction to GSC and MAWGS was manipulated to reach simultaneusly the GSC averaged operating point and the nominal combustor temperature of the H-class turbine, for a given compressor air intake. Fuel stream compositions, temperature and relative flow to the cluster and extra firing chamber were delivered to the GS-code, which determined the GT net duty and actual air flow rate to the GSC cluster. Subsequenlty, the coal input in the Unisim flowsheet was manipulated to reach the same air flow to the GSC as predicted by the GT model. Since small variations in the MAWGS and GSC operating points occur, several iterations were carried out unitl both models predicted equal air flow rates to the cluster for the same fuel compositions. It is noted that the CAPE-OPEN unit operations allow to solve the MAWGS-GSC loop by exporting the product flows of each model to the stationary model, thus enabling a much faster convergence of the whole system.

Gas switching combustion (GSC) cluster
The transient behaviour of the GSC cluster was solved in the model built in Scilab analogously to previous studies [9,13] using a stiff ode function. A CSTR model is used to represent the fluidized bed, and complete fuel conversion is assumed, given the perfect mixing resulting from fluidization and the high NiO oxygen carrier reactivity [8]. Detailed computational fluid dynamics modeling of a large scale GSC reactor has shown these assumptions to be valid when high reactor temperatures are maintained, even for the natural ore, Ilmenite, that is much less reactive than NiO [37]. The molar species and energy balances solved in the code are represented by Eq. (1) and Eq. (2), respectively.
These balances incorporate the primary hypothesis of the CSTR model, which assumes that the outlet flow of the reactor is at the same conditions (pressure, temperature, enthalpy, and composition) as the whole reactor volume. The oxygen carrier selected in this work was NiO, with the formulation taken from Abad et al. [38] and coherently with previous power generation assessments with CLC [39]. NiO is the most promising carrier due to its high oxygen carrier capacity and proven performance to fluidize at high temperatures [40]. The heterogeneous reactions considered, with a very fast reaction rate imposed by a τ value of 0.01 were: To maintain a high average oxidation temperature, while achieving a low degree of undesired outlet stream mixing during valve switch, the O 2 slip heat management strategy [37] was implemented. This reactor operation consists of a concentrated injection of air throughout the oxidation step that prevents all the oxygen from reacting, thus avoiding a sharp heat release at the beginning of the step. Furthermore, the delayed outlet valve switching strategy [9] was employed to maximize the cluster CO 2 capture in the reduction stage outlet stream. A fixed pressure drop of 0.5 bar was assumed in all the simulations Fig. 2 illustrates the reactor behaviour for a GSC cluster of one of the presented power plant configurations. The reduction and oxidation step lengths were tuned to reach similar fluidization velocities of~0.8 m/s using a cluster of 7 reactors 6 m in diameter and 12 m in height. The instantaneous outlet flows of the reactors in reduction and those in oxidation were mixed in two separate streams for feeding to the UniSim model, thus representing the cluster of dynamically operated GSC reactors as a steady state processing unit.

Membrane assisted water gas shift (MAWGS)
The MAWGS reactor was represented as a single reactor tube modelled in Scilab assuming a plug flow reactor, with mass (Eq. (8) & Eq. (9)) and energy (Eq. (10) & Eq. (11)) balances applied to each length differential of the tube to Retentate (R) and Permeate (P) sections, as shown in Fig. 3. The 1-D model is consistent with previous literature assessments of membrane technology integration in large scale IGCC power plants [41,42]. Further modelling assumptions of this unit can be found in Table 4 in the Appendix. The exothermic chemical reaction taking place in the tube length is shown in Eq. 7.
The chemical reaction rate and diffusion flux across the membrane are obtained from the expressions given in Eq. (12) & Eq. (13). The power law coefficients and kinetic and equilibrium constant for the rate of reaction were taken from Hla et al. [31], while the parameters of the diffusion expression were assumed from Fernandez et al. [32]. Several experimental studies validate the assumption of infinite H 2 perm-selectivity through the membrane [43,44] Similarly to the GSC model, the MAWGS reactor model uses a stiff ordinary differential equation (ode) solver for temperatures and species molar flows. Fig. 4 illustrates the composition in the retentate profile and H 2 production in % of the total outlet flow across the permeate side of a membrane tube for one configuration of the power plants investigated. The WGS reaction proceeds rapidly to equilibrium at the start of the reactor length. In practice, it will be advisable to implement an adiabatic WGS reactor upstream to avoid sharp temperature gradients along the membranes, but this detail was neglected here for simplicity.

Unabated IGCC
A detailed schematic of the reference unabated IGCC power plant is given in Fig. 5. Stream data can be found in Table 7 in the Appendix for the case of a 2200 K SFT target.
Coal is gasified in an entrained flow gasifier, Shell type, operating at high temperature (> 1500°C) where coal is fed via lock hoppers (using N 2 as transport gas) [23]. A gaseous quench with recirculated syngas cools the gasifier outlet to 900°C, while a syngas effluent cooler (SEC), consisting of water economizer, steam evaporator and superheater, lowers the temperature further, producing a large amount of HP superheated steam at 450°C. Oxygen with 95%mol purity is delivered to the gasifier by a high pressure pumped liquid oxygen air separation unit (HP PLOX-ASU), which is 50% integrated with the gas turbine compressor. Integration between the ASU & GT allows for operation closer to the design point of the compressor, compensating for the reduced air intake due to diluted syngas firing (which is a lower energy density fuel relative to natural gas). Since the ASU is operated at high pressure (10 bar approximately), N 2 with a purity above 98%mol is obtained at around 2.7 bar, which significantly reduces the compression duty of this stream required for subsequent syngas dilution. Higher integration of ASU and GT compressor is not recommended to avoid reliability and start-up issues [45].
After high-temperature solids removal in candle filters, the gasification island delivers syngas to a hot gas clean up unit (HGCU). This unit consists of interconnected fluidized beds with a Zinc Oxide (ZnO) desulphurization sorbent that is cyclically regenerated, following a similar approach for the modelling as Giuffrida et al. [17]. Experimental studies [46] reveal that complete H 2 S adsorption assumption is acceptable, while additional sorbents are employed to remove other pollutants (NH 3 , HCl, etc.) from the syngas stream [47]. The sorbent  regeneration is accomplished by using part of the N 2 from the ASU to provide a low O 2 concentrated stream (2%mol), reducing undesired sulphate formation in the regenerator bed. The outlet temperature of the syngas from the HGCU was fixed at 400°C, in line with the operating temperatures of the candle filters, and to limit the temperature in the downstream membrane assisted water gas shift reactor (MAWGS) for the plants with gas switching combustion (GSC) discussed in the next section. Further increases of the clean-up temperature do not show great efficiency gains, as reported by Giuffrida et al. [48]. Since the adsorbent regeneration stream is already available from the ASU at slightly pressurized conditions, the compander unit used to deliver this stream to the pressurized regenerator and subsequently expand the offgas product produces a net positive power output.
Premixed combustors used extensively with natural gas are challenging when dealing with syngas and H 2 fuels [49], resorting to diffusion flame combustors where the air is fed stoichiometrically to the fuel to maintain flame stability. Dilution is necessary to limit the adiabatic, stoichiometric flame temperature (SFT), which is directly related to NOx emissions [50]. Clean syngas is therefore subsequently diluted with the remaining N 2 from the ASU after a two-stage intercooled compressor to reach a final fuel pressure of 35 bar. However, to achieve sufficiently low SFT values, further dilution with steam from the bottoming cycle is still required as a consequence of feeding a hightemperature fuel (from HGCU) relative to the IGCC plants with lowtemperature contaminant removal and syngas saturation [23]. Two SFT values of 2200 K and 2300 K are targeted to comply with acceptable ranges for IGCC plants [49].
The diluted fuel is fed to an H-class turbine with the specifications described in [51]. The combustor outlet temperature (COT) is adjusted to reach the same turbine outlet temperature (TOT) of 641°C as the natural gas-fired reference, for all syngas fired cases. The turbine inlet temperature (TIT) is determined by the blade cooling flow model. It is assumed that, with Variable Guide Vane (VGV) closing and partial integration with the ASU, the compressor can run at the nominal pressure ratio and efficiency as when fired with natural gas. The GT exhaust is routed to a heat recovery steam generator (HSRG) with IP reheat in which steam is raised at three pressure levels. Since a substantial amount of HP steam is generated in the SEC, a large thermal duty is required for reheating, and the amount of steam produced in the IP drum is minimal. The HP evaporator of the HSRG is modelled as a once through boiler, while the condenser pressure was fixed at 4 kPa. Further details of the modelling assumptions can be found in the Appendix.

Pre-combustion CO 2 capture IGCC
A detailed schematic of this power plant model is shown in Fig. 6.   Table 7 in the Appendix. Detailed stream data for this power plant model is shown in Table 8 in the Appendix. The reference plant with CCS consists of a pre-combustion CO 2 capture IGCC power plant extensively studied in previous works [23,24,52]. However, in this model, the benefits of HGCU treating are also incorporated. The gasification island is identical to the unabated IGCC plant, where an HP-ASU unit, which is 50% integrated with the GT compressor, delivers O 2 with 95% purity to a Shell gasifier. Coal is loaded with N 2 from the high-pressure column of the ASU. After syngas quench and cooling, the raw syngas is routed to the HGCU unit operating at 400°C. After sulphur and contaminant removal, steam from the HP stage steam turbine outlet is added to reach the required H 2 O/CO ratio for water gas shift conversion of 1.9. Because of the gas clean-up at elevated temperature, part of the H 2 O is added as saturated water from the IP drum, tempering the mixture to reach the required High-Temperature Shift (HTS) inlet temperature of 300°C. After WGS reaction in the adiabatic reactor bed, the syngas is cooled down to 200°C, raising HP steam before it is fed to the Low-Temperature Shift (LTS). Equilibrium conversion is assumed in both beds, and the overall CO conversion is approximately 98%. A pressure drop of 1 bar per bed is specified. The shifted syngas is cooled down in a series of heat exchangers heating several water streams from the bottoming cycle, and further cooled down to ambient temperature. After water knock-out, the syngas is sent to an absorption column modelled with 10 equilibrium stages where Selexol removed approximately 94% of the CO 2 . The use of a physical solvent is justified because of the high partial pressure of CO 2 achieved after the shift. Relative to previous studies [52], the Selexol absorption unit is simplified as there is no need in the present configuration for selective H 2 S and CO 2 removal (due to HGCU). Instead, a single column line-up with solvent regeneration through pressure let-down is used, similarly to Arnaiz del Pozo et al. [53]. This sequential regeneration of the solvent at different pressures (7.5, 3, and 1.05 bar) reduces the CO 2 compression duty of the downstream 5-stage intercooled compressor. A CO 2 pump further increases the CO 2 stream pressure to 150 bar for transport and storage.
Accounting for the CO 2 slip from the absorption unit and the unconverted CO from the WGS, the resulting plant capture ratio is approximately 91%.
The H 2 rich syngas is then mixed with N 2 from the ASU to reduce the flame temperature and avoid NOx formation. A small portion of the H 2 corresponding to 0.9% of the coal LHV is extracted for coal drying. Since the ASU delivers N 2 at 2.6 bar, all the available N 2 (minus the amount required for sorbent regeneration in the HGCU unit and for coal loading) is compressed to 35 bar in a two-stage intercooled compressor (with no aftercooler) and used for syngas dilution. The mixed stream is routed to a saturator unit integrated with the low-temperature heat recovery units of the shift conversion, increasing the moisture in the syngas, and then it is further heated to 220°C. The syngas is subsequently fired in the H-class GT combustor. The N 2 and water added are sufficient to reach an SFT below the values targeted for the Unabated IGCC plant, due to the lower fuel temperature in this plant.
The bottoming cycle of the pre-combustion CO 2 capture IGCC plant with HGCU is identical to the Unabated IGCC reference plant, with a three pressure level with reheat HRSG. The steam cycle is consistently integrated with the steam demand and production of the WGS unit and gasification island.
Pre-combustion CO 2 capture IGCC plants also have the potential to alternate between H 2 and Power production [54,55]. However, these plants require the use of an additional Pressure Swing Adsorption unit (PSA) to purify the shifted syngas and would not be able to capitalize on the efficiency benefits of an integrated ASU. In the present work, however, this plant is designed exclusively for power generation as a benchmark of existing CCS technology.

GSC-MAWGS IGCC with reduction gases recuperator
A detailed schematic of this plant configuration is given in Fig. 7. Stream data for the power mode case are given in Table 9 in the Appendix.
The gasification island of the GSC-MAWGS IGCC plant is similar to the reference case, but analogously to Spallina et al. [12] Fig. 6. Schematic of the pre-combustion CO 2 capture IGCC power plant. Stream data can be found in Table 8 the Appendix. lock hoppers with part of the captured CO 2 stream. The ASU consists of a standalone low-pressure cryogenic cycle, whose specific power consumption was taken from Spallina et al. [12], equal to 325 kWh/tonO 2 . The LP-ASU configuration with no integration with GT compressor was selected to increase the plant flexibility in H 2 production operation mode, and because there is no need to dilute syngas with N 2 . After hydrogen sulphide and other contaminants are removed in the HGCU unit, clean syngas is mixed with IP steam from the HP stage steam turbine outlet prior to the shift conversion. The ratio of steam to CO in the feed stream to the shift reactor was set to 1.9, to prevent catalyst carbiding [23]. A small portion of the H 2 O added to the syngas stream consists of saturated IP water from the IP drum, cooling the feed stream to the shift reactor to the extent that the maximum temperature in the membrane (outlet) does not surpass 600°C. Since it is not required to shift all the CO available in the syngas to produce enough H 2 to reach the desired combustor temperature, a split flow configuration was adopted, bypassing part of the syngas directly to the GSC cluster. In this way, the steam consumption from the bottoming cycle is reduced. The MAWGS reactor consists of a reactor vessel with a total of 6000 membrane tubes of fixed length (10 m) and diameter (0.05 m). These tubes would cost approximately 47 M€ at $5000/m 2 [56], which is estimated to be about 2.5% of the total plant cost based on an earlier economic assessment of a GSC-IGCC plant [57]. The shift reaction takes place in the reactor, diffusing the H 2 product across the membranes to the permeate side, thereby increasing the equilibrium conversion to the products (H 2 ) in the retentate. The H 2 production (permeate flow) is controlled by specifying the permeate membrane pressure and/or the split flow ratio between the syngas that is shifted and that which is directly fed to the GSC. With these operational handles, the relative heating value routed to GSC (retentate) or extra firing chamber (permeate) can be controlled.
When the plant is operated in power production mode, the membrane permeate pressure is set to approximately 2.8 bar, and around 60% of the syngas is routed to the MAWGS reactors. These values represent the optimal trade-off between the energy penalty associated with steam extraction when a higher fraction of syngas is routed to the MAWGS reactor and the H 2 recompression duty, which depends on the permeate pressure imposed. The fraction of the syngas heating value that is routed to the GSC reduction stage, together with the retentate flow, must ensure that the GSC oxidation and reduction outlet temperatures calculated by the transient cluster model are reached, whereas the H 2 produced in the permeate side allows the O 2 depleted air stream from the GSC to reach the COT of the GT. The H 2 obtained in the permeate side is cooled down in a recuperative heat exchanger and then recompressed to 35 bar (required fuel pressure in the GT combustor) in a 4 stage intercooled compressor and subsequently heated in the recuperator before being fed to the extra firing chamber. A minimum temperature approach (MITA) of 20°C was assumed in the recuperator. The MAWGS retentate, consisting of a low heating value syngas with a large amount of water and CO 2 and the bypassed syngas, is mixed and expanded to the pressure ratio delivered by the GT air compressor in a syngas expander. The syngas is then heated up in a recuperative heat exchanger before entering the GSC, using the hot gases from the reduction stage outlet. A MITA of 30°C, resulting in a heat exchanger effectiveness of approximately 93% was employed. Special materials will be required to manufacture the heat exchanger   Table 9 in the Appendix.  [58] due to the elevated hot end temperatures. This recuperator ensures that most of the sensible heat in the syngas/reduction gases stream is effectively transferred to the air stream, which will produce electricity upon expansion in the turbine. The reduced gases recuperator outlet is routed to a heat recovery unit where steam is raised. The stream is further cooled to ambient temperature and, after water removal, the CO 2 is compressed and pumped to a delivery pressure of 150 bar. Since the GSC outlet pressure is preserved (minus heat recovery pressure losses), the recompression duty is small, and only two intercooled compression stages are (plus a supercritical CO 2 pump). Hot compressed hydrogen is fed to the added combustor to increase the COT to 1648°C. It is assumed that well-distributed injection of the hydrogen fuel into the hot depleted air stream for spontaneous combustion in excess air can limit NOx formation in a similar way as a premixed combustor [50]. This avoids the need for dilution with N 2 or steam that is required when using diffusive flame combustors to avoid large NOx emissions, such as in the unabated IGCC plants. After GT expansion, the exhaust air stream is sent to an HSRG with three pressure levels and IP reheat for extra power production in a steam turbine, similar to the reference IGCC plants.
In H 2 production mode, the membrane permeate pressure is reduced to 1.6 bar, and the fraction of syngas sent to the MAWGS reactor is increased to 81%. In this mode, the GT runs at 10% load at a significantly lower pressure ratio than the nominal case. The H 2 produced is cooled down, raising steam for the bottoming cycle and compressed in the pressure stages that were used for fuel compression in the power production mode, and further compressed in two subsequent intercooled stages to a final delivery pressure of 150 bar [59].
Since the air flow rate to the oxidation step of the GSC reactors is reduced substantially due to the part-load GT operation, a larger number of the reactors are used in reduction mode (3) to maintain similar fluidization velocities in each reactor. The total number of reactors and the dimensions remain the same as in the power mode.
The depleted air stream from the GSC is expanded in the GT without added H 2 firing at a TIT around 400°C below the nominal value. The exhaust gas is used to raise steam in the HSRG for the bottoming cycle and to supply IP steam to the MAWGS reactor. The same steam turbine stage efficiencies were assumed as in the power mode in this assessment, acknowledging that appropriate technological and maintenance approaches must be taken to make bottoming cycles more flexible [60]. Regarding the heat recovery units, a conservative assumption was taken, and the heat exchanger elements were specified with the same approaches as in the power mode. The items that require a larger size in H 2 mode, such as the CO 2 compressor, H 2 compressor, and syngas expander (due to the lower pressure ratio of the GT and MAWGS permeate and H 2 delivery pressures) were redesigned to fit the process requirements, considering that, in a future economic assessment study, the largest unit in each of the two modes should be accounted for, to reflect the extra capital expenditure resulting from flexible H 2 and power production.

GSC-MAWGS IGCC with slurry Pre-gasifier
A detailed schematic of this plant configuration is given in Fig. 8, while stream data for the power mode case is shown in Table 10 in the Appendix.
As explained earlier, the main difference in this plant is the use of a   Fig. 8. Schematic of the GSC-MAWGS IGCC power plant with slurry pre-gasification. Stream data can be found in Table 10 in the Appendix. slurry pre-gasification heat exchanger to increase the CGE of the gasifier. The slurry pre-gasifier is modelled in two sections. The first is a heat exchange section where the water slurry is fully evaporated. The second section consists of the pre-gasification of coal, which is divided into 10 stages, each modelled as a gasifier unit with a proportional heat input provided by the GSC reduction outlet gases. An equal volume was specified to each gasifier stage section to ensure that, due to the endothermic gasification reactions taking place in the hot end, the temperature pinch occurs in the cold end of the pre-gasifier. In this way, the zone most prominent to fouling avoids a narrow temperature delta, allowing a high degree of heat extraction from the reduction gases stream.
The pre-gasifier outlet stream (syngas with entrained coal particles) is subsequently fed to a Winkler gasifier operating at 44 bar, and the O 2 demand from the ASU is manipulated to obtain a syngas stream at around 900°C with a fixed carbon conversion reaching 97%. Because of the high-temperature pre-gasified feed and the circulating fluidized bed operation, this conversion value is achievable [61]. The operating pressure for this case was chosen to be the same as the Shell gasifier to maintain the same configuration for the MAWGS reactor. Although HTW gasifiers have up to now been demonstrated at around 30 bar, it is assumed that the scale up to a higher operating pressure is feasible due to the replacement of the dry feeding system with a slurry pump. Furthermore, the operating flexibility and fuel versatility of the HTW fluidized bed gasification will be advantageous for the pre-gasified slurry feed relative to an entrained flow gasifier.
The syngas produced in the HTW gasifier is routed to a recuperator and then to a small syngas cooler, which raises some HP steam, cooling it down prior to the HGCU. This recuperator, similar to the concept presented by Hack et al. [62], is suitable for a syngas produced in an HTW gasifier because the low operating temperature eliminates the need of any gas quench and therefore no syngas recirculation loops are present, resulting in an improved temperature profile relative to HP steam generation. The desulphurization temperature was fixed at 400°C, as in the previous cases. Approximately 63% of the syngas is fed to the MAWGS reactor after the addition of IP steam/water. The retentate stream, together with the fraction of syngas bypassed directly to the GSC, is expanded in a syngas turbine and later routed to the recuperator, which operates with a MITA of 30°C, exchanging heat with the syngas gasifier outlet. The heated fuel is then fed to the GSC reduction stage, heating the air stream delivered by the GT compressor to the GSC oxidation stage outlet temperature. The GSC reduction step outlet gases are fed to the pre-gasifier to heat the coal water slurry, operating with a temperature approach of 30°C in the cold end of the exchanger. The cooled GSC reduction outlet stream is routed to a heat recovery unit consisting of an LP economizer and evaporator, which efficiently retrieves the condensation enthalpy of water present in this stream, generating LP steam for the bottoming cycle before being sent to the CO 2 compression section. The depleted air stream from the GSC is upgraded in the added combustor and expanded in the GT in the same way as described earlier.
In H 2 production mode, the GT runs at part load with a normalized output of 10%, as described before. In this case, since the MAWGS retentate stream contains a substantial heating value due to the presence of methane, a smaller fraction of the clean syngas after desulphurization must be routed directly to the GSC (around 10%). The high fraction of methane in the retentate makes the GSC reduction reactions more endothermic, reducing the reduction step outlet temperature relative to power mode operation. This results in less heat transfer in the pre-gasifier to the coal water slurry in the hot end of the exchanger, slightly increasing the O 2 demand from the ASU and reducing the CGE by around 2%-points relative to the power mode. Nonetheless, the large amount of heating value preserved in the syngas due to the high CGE results in a comparatively more attractive H 2 production efficiency. Analogously to the case with reduction gases recuperator, the H 2 is cooled down, raising HP steam and the compressed in a 6 stage intercooled compressor to 150 bar.

Key performance indicators
A simplified representation of the power plant system is provided in Fig. 9. The plant transforms the chemical energy of a coal fuel into H 2 and/or electricity. A series of by-product energy (heat released to ambient) and material streams (captured CO 2 , exhaust air, gasifier slag etc.) are also plant outputs. Because of the large steam requirement in the WGS unit, a substantial amount of water make-up must be reintroduced into the steam cycle.
The definition of plant thermal efficiency is straightforward when the plant runs exclusively in power production mode (Eq. (14)). However, if H 2 is produced as well, two different hydrogen production efficiencies are calculated, without accounting (Eq. (15)) and with accounting (Eq. (16)) for the net power production of the plant. In Eq.
Since the reference power plant was modelled only for electricity production, the CO 2 avoidance (Eq. (17)) and specific primary energy consumption for CO 2 avoided (SPECCA) (Eq. (18)) are quantified for the GSC plants only in power mode operation.
Finally, the water consumption (dedicated to H 2 generation and syngas dilution for the unabated plant and the schemes with CCS, respectively) per unit of thermal input is also quantified (Eq. (19)).

Results and discussion
In this section, a detailed discussion of the power plant results for the reference IGCC models and the configurations with GSC cluster and MAWGS reactor, for power and H 2 production modes, is given. Firstly, the power breakdown and key performance indicators for each case are provided, after which a more in-depth analysis of the GSC cluster, MAWGS reactors, GT integration, and pre-gasifier operation is presented. Finally, the key technology gaps that must be overcome to deploy these IGCC concepts are discussed.

Energy breakdown and CO 2 emissions performance
The different power plant model results from an energy and CO 2 emissions perspective are provided in Table 1, for power and H 2 production operating modes: When analyzing the results from the benchmark unabated IGCC cases, a substantial efficiency enhancement with respect to past studies of IGCC plants without CCS is observed. This is partly due to the use of a highly efficient H-class gas turbine and higher steam temperatures in the bottoming cycle. Furthermore, it is also due to the added efficiency benefits of HGCU, which amounts to up to 2%-points of efficiency [17]. A large GT duty is obtained in this plant because of the large flow rate of fuel, which reduces the compressor air intake with respect to a natural gas-fired case and, consequently, the compression duty requirements. When looking at the influence of a lower SFT through a higher degree of dilution with steam, the power obtained in the topping cycle is higher for a lower SFT value, while the bottoming cycle output decreases because of the larger IP steam extraction. The results show a decrease of around 0.6%-points efficiency per 100°C lower SFT.
On the other hand, when looking at the pre-combustion CO 2 capture IGCC plant, the energy penalty associated with CCS amounts to 9.7%points. The benefits of HGCU for this plant are, to an extent, curtailed by the fact that shifted syngas cooling to ambient temperatures for CO 2 removal must still be done. Nonetheless, some IP steam savings are attained as the required steam/CO ratio in the HTS can be reached by addition of IP water (to quench the high-temperature clean syngas from the HGCU), which is less energy demanding to produce. Similarly, the removal of the sulphur compounds prior to the shift conversion simplifies the Selexol absorption train, resulting in a comparatively lower auxiliary consumption than schemes with syngas scrubbing and sour shift [23], as a lower solvent circulation rate is achieved and no LP steam is required for regeneration in the H 2 S stripper unit. Around 9% of the CO 2 generated in this plant is emitted (originating primarily from the CO slip of the WGS unit and the CO 2 not captured in the Selexol plant). Because of the lower efficiency resulting from CCS, using the same GT, around 17% higher heat input (coal) must be fed to the plant, delivering approximately 5% lower electricity output. This results in a carbon avoidance of around 2%-points below the capture rate of the plant.
When comparing the unabated IGCC power plant against the GSC-MAWGS IGCC with reduction gases recuperator, the energy penalty of CO 2 capture results in only 4.4%-points. Since O 2 from the air stream is withdrawn in the GSC cluster, and the fuel input only consists of a small H 2 flow rate, the net turbine output is smaller than in the syngas fired cases. As only a small amount of fuel is added to the hot gas, the compressor operates at the nominal air flow intake with a slightly lower pressure ratio. Furthermore, substantial auxiliary power consumption is needed for H 2 compression.
A key feature to mention about the GSC-MAWGS plant is that the large energy penalty encountered by the pre-combustion CO 2 capture IGCC plant to produce an H 2 fuel is to a great extent avoided: • Since only a fraction of the syngas is required to produce H 2 to reach COT from the GSC oxidation outlet, the steam extraction from the bottoming cycle is greatly reduced. For the pre-combustion plant, the syngas stream must be entirely shifted, needing much more steam.
• By using the recuperator and the syngas expander, the heat released in the WGS reaction is converted to power at a high efficiency. In the pre-combustion plant, the shifted syngas must be cooled to ambient temperatures, condensing most of the water for CO 2 removal through absorption, retrieving a small portion of the heat of reaction as HP steam.
• The pre-combustion plant recovers CO 2 at relatively low pressures and carries out the compression in 5 intercooled stages, while the pressurized reduction gases stream present in the GSC concepts allows to reduce the compression train size and associated auxiliary consumption quite significantly.
From a CO 2 emissions perspective, this plant achieves 3.8%-points higher capture than the pre-combustion CO 2 capture plant, with a 1.6 MJ/kgCO 2 lower SPECCA index and a CO 2 avoidance, which is around 5%-points higher.
When operating in H 2 production mode, the minimum load imposed by the GT results in a H 2 equivalent efficiency slightly below the literature values for pre-combustion CO 2 capture plants designed for low electricity production [55]. However, the range in which this plant can operate flexibly between H 2 and power production is much wider than for the plants in the aforementioned study (with ad hoc power cycles for different degrees of electricity outputs) as the GT can be ramped up safely from 10 to 100% power load. Furthermore, the high efficiency at which electricity is produced in the current reference plant tends to comparatively decrease the equivalent H 2 efficiency (Eq. (16)). If the pre-combustion CO 2 capture plant reference efficiency was used to calculate equivalent efficiency in Eq. (16), the resulting value would rise to 58.0%.
When looking at the GSC-MAWGS IGCC plant with slurry pre-gasification, it can be seen that the energy penalty is reduced to only 1.3%points, reaching a capture rated above 98% due to the elimination of coal drying requirements and CO 2 lock hopper venting of the Shell gasifier. This results in a carbon avoidance that is 8.5%-points above the reference pre-combustion plant and a minimal SPECCA Index.
In terms of H 2 production equivalent efficiency, this configuration clearly surpasses the case with the reduction gases recuperator by approximately 6%-points. Interestingly, all the heat duty invested in 10% load GT operation is sufficient to satisfy the internal electricity demand of the plant, with a negligible net power output.
Finally, it can be mentioned that the process water make up required due to the shift conversion/syngas dilution units for these novel plants is also substantially lower than for the pre-combustion benchmark, mainly due to the large water retrieval in the reduction gases condenser.

Power plant system analysis
In order to fully comprehend how these efficiency benefits arise, a more in-depth analysis of the heating value distribution for the plants with GSC and MAWGS reactors is performed.
For the case with slurry pre-gasification, the hot feed to the gasifier results in a CGE of around 100% (effectively retaining all the coal heating value in the syngas) while substantially reducing the O 2 requirement from the ASU. This appealing result is due to the use of heat in the GSC reduction outlet gases to displace fuel combustion with O 2 in the gasifier. On the other hand, for the plant with reduction gases recuperator, only approximately 80% of the fuel heat input is preserved in the syngas that is distributed in between the GSC cluster and the membrane reactor, because of the lower CGE attainable in the dry-fed entrained flow Shell gasifier. The remaining heat is downgraded to HP steam in the SEC, which can only be utilized in the bottoming cycle for electricity production. This explains the large difference in H 2 production efficiency that is obtained between the models, as depicted in Fig. 10.
Alongside this, an illustrative heat-temperature profile of the pregasification unit is provided in Fig. 11. The effect of endothermic gasification reactions prevents a temperature pinch in the hot end of the exchanger, thereby extracting more heat from the reduction gases stream.
Thus, the water evaporation and heating in the slurry can be done directly with low-grade heat as the temperature profile of the coalwater slurry conveniently matches the cooling curve of the reduced gases stream. The higher temperature of the coal slurry fed to the HTW after the pre-gasifier results in a lower gasifier volume and a smaller oxidant stream required, and consequently a substantially smaller and ASU, that ultimately leads to a lower specific plant cost.
It should be noted that, due to the high operating pressure and low temperature of gasification, the syngas obtained in the models has a high methane fraction (around 7.3%mol), which represents approximately 22% of the coal LHV. These values are consistent with the trends reported in Higman [61] for similar operating conditions. Although this species cannot be effectively transformed to H 2 in the MAWGS reactor, as the steam reforming equilibrium is inhibited at the reactor temperature [63], the presence of this high heating value component in the fuel allows to reach the required heat rate for 10% GT output operation in H 2 mode bypassing a smaller portion of syngas to the GSC. Together with the higher CGE attained in the HTW gasifier, the net overall effect is an increase of the H 2 production efficiency observed in Fig. 10, comparatively to the case with Shell gasification.

Key technology gaps
The realization of the promising results reported in this study requires further demonstration and scale-up of several process components. Four key technology gaps and associated risk mitigation strategies are briefly discussed below, in addition to potential challenges with flexible operation.
GSC reactors. Due to the simple standalone bubbling fluidized beds employed as GSC reactors, there is little doubt that the concept can operate successfully at large scale under pressurized conditions. Labscale pressurized experiments have not encountered serious technical challenges [64,65]. The key uncertainty is related to the maximum achievable operating temperature. The oxygen carrier material, reactor body, and downstream valves and filters must all be able to withstand the maximum reactor operating temperature (assumed to be 1200°C in this study). Lower GSC operating temperatures will reduce the plant efficiency in power mode as more hydrogen needs to be extracted from Heat transferred (%) coal slurry reduction gases carbon conversion Fig. 11. Pre-gasifier temperature profile for the GSC-MAWGS IGCC plant in power production mode. the syngas stream and recompressed for injection into the added combustor, and more steam must be extracted from the steam cycle for the WGS reaction. MAWGS reactor. Pd-based hydrogen perm-selective membranes are an established technology and have been demonstrated for MAWGS reactors [66], but long-term durability along with long-term cost reduction potential remain uncertain. Alternatively, the conventional combination of WGS and PSA can be used, where the shifted syngas must be cooled before producing a pressurized H 2 stream and a lowpressure depleted syngas stream. This option is expected to reduce efficiency for power production because of the poorer integration potential of the heat in the syngas stream and the additional heat released by the WGS reaction. For H 2 production mode, however, conventional WGS and PSA could be attractive because the large H 2 stream is produced at elevated pressure.
Added combustor. Controlled and low-emission spontaneous combustion of hydrogen injected into the hot depleted air stream from the GSC reactors needs to be demonstrated. If multiple fuel injectors are used to evenly distribute the fuel in the oxidant stream, this combustor could achieve low-NOx combustion performance similar to premixed combustors. The present study assumes that such combustion can be completed using pure hydrogen as fuel. If future studies indicate a necessity for fuel dilution, N 2 from the ASU could be used as a sweep gas in the MAWGS reactor to allow for higher membrane permeate pressures, thereby minimizing efficiency losses associated with fuel dilution. From an experimental point of view, high-temperature air combustion has been studied for many years, mainly for application in furnaces. However, high flame stability and pollutant reductions also make it an interesting application for gas turbines [67].
Pre-gasifier. As discussed earlier, the pre-gasifier adds significant gains in efficiency and CO 2 capture ratio and is also expected to considerably reduce plant capital costs. However, it does introduce significant uncertainty. Fouling is likely to be high in the tubes carrying the slurry, although the potential cost increases from a reduced heat transfer coefficient due to fouling should be minimized by a relatively large temperature difference driving heat exchange. The rate of the endothermic gasification reactions must also be high enough so that the ash melting point is not crossed. If this is not possible, more water could be added to the slurry, leading to a minor efficiency penalty but also to less methane formation that will allow more H 2 extraction when the plant is operated in H 2 production mode. Furthermore, the unknown size of the coal particles after slurry water evaporation can significantly affect the gasifier performance, although the flexibility of fluidized bed gasifiers mitigate this risk. Understanding these effects will require separate experimental demonstration activities. Some experimental studies have already been done, showing that it is possible to vaporize the slurry without coal particle agglomeration [68].
System flexibility. Operation as a fully flexible load-following plant could be practically challenging due to the reduction in the gas turbine pressure ratio under part-load operation. This reduced pressure changes the duties of the syngas expander and CO 2 compressor and will also affect the heat transfer characteristics in the recuperator or slurry pre-gasifier. In addition, flexible operation of the bottoming cycle could pose challenges due to the need to supply steam at constant pressure to the gasification section, despite the large change in load between power mode and H 2 production mode. A practically simpler option would be to design the plant for binary power or hydrogen production by including a small added gas turbine for operation only during H 2 production mode with a TIT that can be achieved without added firing and the same pressure ratio as the large H 2 -class turbine. In this way, the system pressure remains constant when switching between power and H 2 modes. This turbine can be sized such that remaining heat sources only supply process steam, avoiding challenges with steam cycle integration. The cost of this added turbine should be comparable to the avoided cost of oversizing the syngas expander and CO 2 compressors for part-load operation of the H-class turbine at a lower pressure ratio.
Prior to the development of these technological enablers, an ex ante economic assessment should be carried out to weigh the benefits of the novel configurations proposed relative to established, de-risked CCS technologies such as the pre-combustion CO 2 capture IGCC plant  presented in this work.

Summary and conclusions
In this work, a flexible plant configuration for H 2 and power production based on the integration of a GSC cluster with a MAWGS reactor has been proposed. An advanced scheme with coal water slurry pre-gasification using the high-temperature GSC reduction stage outlet was also developed. Both plants were consistently benchmarked against an unabated IGCC plant and a pre-combustion CO 2 capture IGCC plant utilizing modern gas turbine technology and HGCU for syngas treating. The benefits derived from this technological feature are to a degree limited in the unabated IGCC plant due to the need to extract steam from the bottoming cycle to limit NOx emissions, while in the precombustion plant the efficiency gain is reduced due to the ambient temperature CO 2 extraction in the Selexol unit. From an electrical efficiency point of view, the GSC-MAWGS plant with reduction gases recuperator to preheat the syngas routed to the GSC was capable of reducing the energy penalty due to CCS by 5.3%-points relative to the pre-combustion benchmark, while the GSC-MAWGS IGCC plant with slurry pre-gasification achieved an 8.4%-points reduction. Alongside this, the novel plants reached CO 2 avoidance rates up to 8.5%-points above the reference CCS technology.
Besides these attractive performance figures, the proposed configuration can operate to produce H 2 as an energy vector in times of low electricity prices, where the demand is satisfied by existing renewable energy infrastructure. This is done by ramping down the GT to 10% of its rated power and bypassing most of the syngas fuel to the MAWGS reactor, while the traditionally inflexible units (gasification, ASU, etc.) can operate continuously. Low part-load operation of the GT for long  periods is acceptable as harmful emissions are avoided due to the flameless combustion taking place in the GSC, therefore allowing a higher degree of flexibility output between power and H 2 . The equivalent H 2 production efficiency is in line with the values reached in pre-combustion H 2 production IGCC plants designed for low electricity outputs [55] for the GSC-MAWGS IGCC plant with the reduction gases recuperator (56.6%), but increase substantially to 62.5% for the pre-gasification case. This is achieved because the reduction gases of the GSC effectively transfer sensible heat to the gasification feed, minimizing the oxidant requirements from the ASU. In this case, the gross electricity output of the combined cycle matched the auxiliary consumption demand of the plant.
In conclusion, the present work outlines two efficient IGCC schemes with CCS, using a modern H-class GT, which overcome the inflexibility feature of traditional IGCC 'baseload' power plants by using H 2 as an energy storage vector. Such a plant can simultaneously balance variable renewables and produce cost-effective hydrogen for decarbonizing other sectors like industry, transport, and heat. To realize this promising performance, further development and demonstration efforts are required for GSC and MAWGS reactors, the added combustor, and the        slurry pre-gasifier. As future work, a rigorous economic and system integration study should be completed to quantify the potential reductions in total energy system costs and emissions that can be realized by this flexible plant.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.