The processes evaluated in this work are divided into two scales corresponding to the electrolyser power of two demonstration projects. The first scale corresponds to the 1 MW_el BioCat plant in Avedøre (Denmark), based on biological methanation (Electrochaea.dk ApS, 2014). The second scale refers to the 6 MW_el Audi e-gas project in Werlte (Germany), which operates alkaline electrolysers with cooled, fixed-bed methanation reactors (Götz et al., 2016). Furthermore, two methanation technologies are compared, namely bubbling fluidised bed methanation and biological methanation in a stirred bubble column.

As electrolyser technologies, proton exchange membrane, as well as high-temperature electrolysis by solid oxide electrolysis cells are considered. Based on the available electrolysis power and subsequent H₂-feed, a CO₂ feed stream is calculated according to the respective needs of the methanation technology. For the cases including the SOE technology, the hydrogen output of the electrolyser is kept the same as in the PEM-cases. Therefore, plant costs are comparable to the PEM-BFB case.

An overview on the considered cases is given in Table 1.

The combination of biological methanation with solid oxide electrolysis is not considered in this work since the temperature level of excess heat from biological methanation is too low for integration with high temperature electrolysis. Such a process would either require the addition of an external high temperature heat source or result in a decreased electrical efficiency of the SOE.

The two catalytic processes were furthermore evaluated for the case of direct methanation of biogas. In this case, the electrolyser power is fixed to the respective CO₂-case and a biogas-feed containing the necessary CO₂ is provided to the process. The raw biogas is assumed to have a methane content of 60 vol.-%. The cases with biogas as feed are marked with “bg.”

The comparison of the cost structure of the three processes is performed by determining the operational cost (OPEX) and capital cost. Based on the technical performance and size of the units, cost functions for the units are applied, which were obtained from Ulrich and Vasudevan (Ulrich and Vasudevan, 2004) and already part of previous work (Witte et al., 2018a). These functions have the following structure: C B M = C P ⋅ F B M , with  F B M = f ( F M ,   F p ) (9)

The bare module cost C B M consists of the purchased equipment cost C P , which is multiplied by a bare module factor F B M . This factor is a function of the material factor F M and the pressure factor F p , which are given in literature and consider the cost increase by the use of special materials and elevated pressures. The accuracy of such cost evaluations is within the range of ±20% (Ulrich and Vasudevan, 2004).

In order to obtain the investment cost C k i n v for a unit k , the cost for piping, instrumentation and further surrounding equipment have to be considered. Such costs are approximated by a factor of the bare module cost, as shown in Eq. 10. The plant cost factors F P C considered in this work are summarised in Supplementary Table S2 (Garrett, 1989; Witte et al., 2018a). C k i n v = C B M , k ( 1 + ∑ F P C ) ⋅ F C E P C I (10)

The investment cost also include a factor F_CEPCI , which quantifies the cost increase of equipment (Ulrich and Vasudevan, 2004) according to Eq. 11. It consists of the Chemical Engineering plant cost index (CEPCI, I ₂₀₁₉), listed in Supplementary Table S3. The reference value I _ref(2004) was 400 in 2004 (Ulrich and Vasudevan, 2004). F C E P C I =   I 2019 I r e f ( 2004 ) (11)

The annualised capital expenses (CAPEX, C C A P E X ) are calculated, based on the annual interest rate i and the plant lifetime a , both shown in Supplementary Table S3: C C A P E X =   C t o t i n v ( 1 + i ) a ⋅ i ( 1 + i ) a − 1 (12)

The cost for the nutrient solution for biological methanation was estimated based on a total nutrient concentration of 2 g/L of slurry (Rachbauer et al., 2016; Rusmanis et al., 2019). In case of an artificial nutrient solution fed to the reactor, this growth medium has to be mixed from a variety of chemicals at an estimated cost of around 20 €/kg. We therefore estimate the cost for growth media provision to around 40 €/m³. Alternatively, the reactor can also be supplied with digester efflux, which provides all necessary nutrients (Ullrich and Lemmer, 2019). Nevertheless, also the integration of such a waste stream and its conditioning is associated with certain costs, which may be in a similar range as shown above.

For the membrane unit, a module geometry of 1 m length and 4″ diameter is considered. This corresponds to a commercial biogas upgrading membrane. A 2″ version of such a module was tested in previous work for its capability for hydrogen recycle in Power-to-Gas applications (Gantenbein et al., 2021). The separation area of this module was estimated to be around 65.12 m², which corresponds to a fourfold area of the 2” module (Witte, 2018). The module consists of a cartridge fixed into a tubular, pressure-resistant case, which itself provides all the connections to the plant. The cartridge contains the hollow fibres and is sealed with a resin. Due to ageing effects, the cartridges have to be replaced after a certain time interval. The replacement cost were obtained from Ulrich and Vasudevan (Ulrich and Vasudevan, 2004), which indicate 50 $/m² in 2004. As the cartridges also maintain pressure and require specialised sealing, we assumed a twice as high cost for the membrane material. Considering the CEPCI, this cost is 126 €/m², which is in a similar range as a quote obtained by a supplier. This cost refers to the replacement cost of the membrane cartridge, which does not include further installations. We consider this replacement cost as operational expenses, analogously to sorbent and catalyst replacement cost. For the initial installation cost of the membrane modules, we consider the same amount, but include also the plant cost factors.

For the two methanation reactor types, a detailed cost analysis was performed, based on the reactor vessel itself, as well as the auxiliary units required for operation and heat utilisation. For the fluidised bed system, this includes the cooling cycle, the internal heat transfer tubes and a vessel for catalyst activation as already used by Witte et al. (Witte et al., 2018a). The biological system also includes an external cooling system, internal cooling coils, which also serve as baffles, as well as the multi-stage stirring system.

The high-temperature BFB methanation processes presented here allow two ways of heat utilisation: The pre-heating of the feed gases using the off-heat of the water condenser and in the case of high-temperature electrolysis the steam generation via the use of off-heat from the reactor. The second option was justified via pinch point analysis, as shown in Supplementary Figure S1. The cold streams (blue) correspond to the heat requirement of the feed water to the electrolyser. Water enters the plant at 20°C and is heated to 193°C, which corresponds to the vaporisation temperature at the operation pressure of 15 bara. The heat requirement in the evaporator is indicated by the horizontal part. The steam is then superheated to the operation temperature of the electrolyser. The last 50 K of temperature increase have to be provided by an electrical heater, whose energy demand is indicated by a horizontal shift of the curves. This point also corresponds to the minimal temperature approach of the hot and cold streams, which was set to 50 K. The hot streams correspond to the combined heat provision by the product gas streams from the electrolyser (hydrogen and oxygen). Specialised heat exchanger equipment is necessary to valorise the heat of both gases efficiently. Concepts for such equipment are under development (Flaviana et al., 2016).

The hot streams (red) correspond to the product gas streams cooled to 60°C. In the first step, the hot product gases are cooled to 287°C, where the freed enthalpy is used to superheat the incoming steam to 750°C. The reactor provides heat at a temperature level of 287°C, which is sufficient in quality and quantity to provide heat to the evaporator. Subsequently, the product gases are further cooled by the incoming water feed. Part of the steam remains in the hydrogen product gas, as it is not fully converted in the electrolyser. This steam is condensed at a temperature level of 193°C, visible as small horizontal part in the hot composite curve (red). The remaining cooling duty can be provided by air cooling, which is not further discussed in this work.

The reactor analysed in this work is cooled by a thermo-oil system and internal heat exchangers. This cooling system is directly connected to the evaporator providing an efficient heat transfer. The remaining heat after the water evaporation can be further utilised at a temperature level of around 280°C. Other concepts combining fixed bed methanation with high-temperature electrolysis make use of direct, internal steam generation in the reactor (Giglio et al., 2018; Dannesboe et al., 2020). In order to provide an efficient heat transfer, the water has to be kept in a boiling state. Depending on the reactor temperature, this requires high operation pressures, which increases equipment cost substantially and decreases cost efficiency.

As shown in this pinch point analysis, the heat provided by the off-gas streams and the reactor is sufficient to produce and superheat the steam required for the high-temperature electrolysis. Additional electrical heating is necessary to reach isothermal conditions in the stack. Nevertheless, the corresponding electrical energy demand represents only a small share of the total electrolyser power.

As shown in previous work (Witte et al., 2018a), the investment cost of Power-to-Gas systems are highly dominated by the electrolyser cost. The cases considered in this work are shown in Figures 3, 4 and summarised in Supplementary Table S5. The hydrogen path contributes to at least 50% of the total cost in case of 1 MW biological methanation, and to up to 92% in the case of BFB methanation with 6 MW_eq SOE. This extreme originates mainly in the high assumed cost for the SOE stack in the base-case. Nevertheless, even in the case of a commercial PEM electrolyser, the cost share amounts to 69% in the 1 MW and 86% in the 6 MW scale.

When a Power-to-Gas plant is in planning, two main considerations can be identified: An overall cost analysis for a newly installed plant, or a retrofit project, where the decision between methanation technologies has to be taken. In the first case, the hydrogen path has to be included in the considerations. For the second case, as the hydrogen path is identical for all methanation technologies, it can be excluded from the consideration. In some of these cases, existing electrolyser units are retrofitted with PtG equipment. It is therefore important to compare technologies based on total investment cost, as well as only based on methanation equipment cost.

The second-largest share of investment cost is the reactor vessel itself. As shown in previous work (Inkeri et al., 2018) and pilot plant projects, biological methanation requires a large reactor volume to produce utility-scale biomethane. This is also represented in the data discussed in this work, as for the operation with pressurised hydrogen and potentially corrosive slurries, i.e. CO₂ dissolved in water, a pressure resistant stainless steel tank is required. Although BFB methanation needs a smaller reactor volume, it also requires pressure and corrosion resistance at high temperatures. Nevertheless, mainly due to the size difference, the investment cost for the BFB vessel is less than a third of the cost of the bioreactor.

The data provided by the simulations of Inkeri et al. indicate a methane evolution rate of 168 vvd in the 1 MW case and 70 vvd in the 6 MW case, based on the reactor sizes given in Table 2. Other biological reactor concepts are available and under investigation (Rusmanis et al., 2019). For fixed film reactors, which include trickle bed reactors, the projects reported by Rusmanis et al. show an average methane evolution rate of around 18 vvd (Rusmanis et al., 2019). Compared to the total methane formation in CSTR systems, such rates would result in large vessel sizes. Even though these reactors are operated at atmospheric pressure, such volumes will result in high investment cost for the main reactor, exceeding the system considered in this study. The demonstration project ORBIT is a trickle bed reactor operated at pressures up to 12 bar, but also here, a methane formation rate of only 8.4 vvd is reported (Thema et al., 2021). This indicates that further development towards higher methane formation rates is necessary.

For catalytic BFB methanation, a larger main compressor is required than for biological methanation, since the membrane upgrading unit creates a recycle stream, which increases compression duty.

A further cost factor in the BFB-based processes is the heat exchanger equipment required for utilisation of the product gas stream and subsequent integration with the feed gas stream. This cost is not considered in the bioreactor, since the latter is operated at a low temperature level, which results in a low water content in the product gas. In the case with SOE electrolysis, more exchangers are required for heat management in the electrolyser. In the 1 MW-case, this doubles the cost for heat exchanger equipment compared to the PEM-BFB case, for the 6 MW-case this results even in a three-fold increase.

In addition to the CO₂ base cases, the process was furthermore simulated using biogas as carbon source. The feed stream was adjusted to match the CO₂ content with the electrolyser power. This results in a 2.5 times higher total feed flow rate than in the pure CO₂-cases due to the additional methane. This higher total flow rate causes a cost increase of the supporting units, namely the compressor and desulphurisation. The reactor cost on the other hand is almost not affected. As the absolute amount of CO₂ remains identical, the same amount of catalyst is required for the methanation unit. Furthermore, the internal heat exchanger area remains equal, as the same reaction heat has to be dissipated. The gas composition itself has therefore less influence on the reactor dimension and subsequently on the corresponding investment cost than on the supporting units.

Rough estimates for plant scaling are obtained by relating the capacity increase of equipment to the cost increase by an exponential law according to Eq. 13 (Tribe and Alpine, 1986): C 2 C 1 = ( 6 M W 1 M W ) α (13)

Such scaling exponents are obtained for all three processes and listed in Table 5. For the electrolyser we assumed a linear relationship of cost to the rated power, implying that there is no economy of scale applicable. Again, due to the dominance of the expenses for the H₂-path, this somewhat conservative assumption has a high influence on the scaling exponent. For the total plant, the exponent is close to unity. When only the methanation plant is considered, the exponent for the biological system is close to 0.6, which is a common rule of thumb value for the scaling of vessels and therefore reflects the dominance of the cost for the reactor vessel. The scaling exponents for the catalytic system are in the range of 0.4–0.5, which implies that there is a strong decrease in specific expenses with increase of the scale.

The annualised capital cost is calculated according to Eq. 12.

The specific investment cost for the biological methanation without the hydrogen path amount to 2975 €/kW in the 1 MW_el-case (B1) and 1538 €/kW in the 6 MW_el-case (B6), related to the energy output of the produced biomethane. Schlautmann et al. (Mörs et al., 2020; Schlautmann et al., 2021) estimate the specfic investment cost for the pilot plant in Solothurn, which corresponds to a biomethane output of 0.325 MW, to 4320 €/kW. They furthermore provide estimates for specific investment cost for plants scaled to 1 MW_biomethane and 5 MW_biomethane, which amount to 2172 €/kW and 940 €/kW. In Supplementary Figure S2, the estimates of the Store&Go project are compared to the results provided in this study. Although the estimates provided in the current study fit well within the estimates reported by Schlautmann et al., they are slightly more conservative. Especially the value for the larger plant size indicates that the cost-related scaling factor provided in this study is lower than the one by Schlautmann et al. It is difficult to assess such differences in detail, as the full cost assessment and detailed methodology of the Store&Go project is not known to the authors.

The annual operational expenses for the processes are shown in Figures 5, 6 and summarised Supplementary Table S6.

The obvious main driver for operational cost of the Power-to-Gas process is the electricity cost. In both scales, the cases PEM-Bio (B1 and B6) and PEM-BFB (C1 and C6) are comparable in terms of electricity consumption. The SOE introduced in cases C1S and C6S has a higher electrical efficiency, which can be considered as a reduction of electricity use per hydrogen produced.

The annual operation cost also includes labour and maintenance cost (O&M), which are estimated by 1.5% of the investment cost for the electrolyser and 5% of the total investment cost for the rest of the plant. This differentiation takes into account that electrolysers do not have any moving parts and require therefore less servicing. Due to the dominance of electrolyser costs, the O&M costs are also separated in two parts.

When excluding the H₂-path in this analysis, the main contribution to the OPEX is the O&M cost for the rest of the plant. The actual consumables and utilities used in the process amount only to around one half of the remaining OPEX. For the biological methanation, the largest share of its OPEX are related to the nutrient cost. In the BFB processes, the largest contribution comes from the bi-annual membrane replacement, followed by the catalyst replacement, which occurs on an annual basis. According to our assumptions, the nutrient provision cost are around 4 times higher than the annual catalyst replacement cost in the 1 MW catalytic case. For the 6 MW case the amount increases tenfold. The amount of nutrients required scales with the volume of the biological system, subsequently there is a disproportionate increase compared to the energy scale.

The stirring power required for the bioreactor amounts to 0.9–1.1% of the electrolyser power, as shown in Table 2. This corresponds to less than 1% of the total OPEX, but remains in a similar range as the cost caused by other supporting units, such as the main compressor and sorbent replacement costs.

Nevertheless, when including the H₂-path in the calculation of the pure CO₂ cases, the costs for utilities and consumables are marginal. The cost for the H₂-path, including the corresponding O&M expenses make up 76–92% of the total.

The charges for membrane replacement have a direct influence on the OPEX, however it represents only a minor part of the total OPEX. Therefore, the sensitivity towards the replacement interval is only low. An annual replacement would double the share of membrane replacement cost compared to the base-case represented in Figures 5, 6.

The total OPEX for catalytic methanation operated with biogas are almost twice as high as in the corresponding base cases. The additional cost originates from the price paid for the raw biogas, which can be considered as an opportunity cost for biogas upgrading through an alternative method. Apart from the increase of the total OPEX, the influence on the OPEX related to the methanation equipment is also visible: Due to higher gas throughput, the electricity consumption of the compressor increases, as well as the amount of sorbent required each year. Additionally, the membrane replacement cost approximately doubles. Although roughly the same amount of CO₂ and H₂ has to be separated from the product gas stream a larger amount of membrane modules is required. The increased number of modules compensates for the higher total gas flow rate, as the permeation rate of CO₂ and H₂ remains constant.

The energy content of the biogas injected to the grid is determined according to Eq. 14 from the higher heating values of the species CH₄ and H₂. E b i o m e t h a n e =   V ˙ b i o m e t h a n e ( x C H 4 H H V C H 4 + x H 2 H H V H 2 ) (14)

The total revenues from biogas sales are then obtained by multiplication with the biogas selling price. The heat revenues are calculated by considering the high and mid temperature level off-heat. In order to account for heat losses in the heat extraction equipment, we assumed that only 70% of the heat produced can be sold. The specific production cost per kWh of gas injected to the grid are calculated by dividing the total annual cost of the plant (OPEX + CAPEX) by the biogas sales. The values obtained for the different processes is summarised in Supplementary Table S7.

The absolute production cost per kWh of biomethane injected to the grid are indicated in Figure 7. The values correspond to the base-case scenario represented by the input parameters in Supplementary Table S3. For this case, neither of the pure CO₂ processes can be operated economically. Using biogas as feed reduces production cost by around 26–39%. This causes the two PEM-based catalytic processes to reach a profitable range (cases C1bg and C6bg) at a production cost of 10.2 and 9.5 €-ct./kWh. The process including high-temperature electrolysis is right at the profit margin at a scale of 6 MW.

As we have no full model for biological methanation available, the production cost for biological methanation (cases B1 and B2) using biogas were projected as shown in Figure 7. The analysis of the investment and operation expenses showed that for the catalytic cases, the change of the feed gas to biogas does hardly affect the cost related to the reactor but increases the expenditures for auxiliary units. We therefore assume that approximately the same cost decrease occurs for the biological system as for the catalytic technology. Even though there is a significant step towards profitability possible, the margin is not yet reached under base-case assumptions.

The 1 MW-scale SOE-BFB process has the highest production cost, mostly caused by high investment cost for the SOE stack. The cost decreases with scale, but remains the highest among the technologies. When switching to biogas methanation, the production cost decreases by 39%, which is the highest relative decrease among the technologies. This is a result of the higher electrical efficiency of the SOE stack, which reduces the relative OPEX due to lower expenses for electricity while still producing the same amount of biomethane.

For both feed gases, the 6 MW PEM-BFB case indicates the lowest production cost, which is a result of scaling effects, lower-cost electrolyser technology, as well as a small reactor size. Therefore, switching the technology from biological methanation to BFB-based technology can lead to a cost reduction of 17% (1 MW) and 19% (6 MW) in the case of pure CO₂ methanation. When using biogas, this benefit even increases to 24% (1 MW) and 26% (6 MW).

A further decrease of production cost is achieved by valorizing the process heat. For the PEM-BFB cases a cost reduction of around 5%, and for the SOE-BFB cases a reduction of around 2% is possible. The latter cases do have a limited heat output due to the heat requirement of the SOE. Furthermore, when biogas is used only very limited heat sales are possible, as part of it is redirected to feed water evaporation. The valorization of the oxygen produced in the electrolysis would be another opportunity to decrease cost. However, this is challenging to realize at most sites and is therefore not considered in this work.

The electricity cost states the largest influence on the operation cost of the processes. The base-case used above is based on an electricity price of 5 €-cts./kWh. This value is comparable to the production cost of wind and large-scale photovoltaic electricity (Bauer et al., 2017). Figure 8 indicates the production cost for each process as a function of the electricity price. The profit margin corresponds to the selling price of renewable gas indicated at 0.11 €/kWh_biomethane. As shown in Figure 7 for the pure CO₂ cases, the 6 MW PEM-BFB case shows the lowest production cost and therefore reaches profitability at the highest electricity cost of approximately 3.2 €-cts./kWh_el. The corresponding 1 MW-case with the second-lowest production cost reaches profitability at 2.2 €-cts./kWh_el, whereas the larger-scale biological methanation process requires an electricity price as low as 1.2 €-cts./kWh_el to reach the profit margin. The 1 MW-scale biological process only reaches a profitable range when electricity is provided at 0.3 €-cts./kWh_el. The performance of the SOE-based processes is inferior, as no profitability is reached, even if no electricity cost is considered. Nevertheless, it is notable, that the SOE-based processes are less sensitive towards electricity cost, i.e. the production cost do not rise as fast with the electricity price as in the PEM-based cases. This is a direct consequence of the higher efficiency of the SOE stack and the heat integration with the BFB reactor.

Biogas as feedstock increases the total amount of methane produced per kWh of electricity. Generally, the processes are therefore less sensitive towards the electricity price than in the case of pure CO₂ methanation. This can be seen in Figure 8 by a decreased slope of the corresponding lines. The 6 MW-scale PEM-BFB process reaches profitability at 7.4 €-cts./kWh_el and at 6.2 €-cts./kWh_el in the 1 MW case. Lower, but still above the pure CO₂ processes, the methanation with high temperature electrolysis reaches the margin: At 5.4 €-cts./kWh_el for the 6 MW case and 3.5 €-cts./kWh_el for the 1 MW scale, respectively.

In order to make the SOE-based processes profitable, the investment cost and subsequent CAPEX for the SOE stack have to decrease considerably. Schmidt et al. (Schmidt et al., 2017) estimate the cost for the SOE technology to be in a range of 3000–1900 €/kW_el in the year 2030, depending on R&D funding and production scale. This is still around three times as high as their estimate for PEM electrolysers. Figure 9 shows the influence of the investment cost of the SOE stack on the production cost of biomethane. Electricity prices of 5 €-cts./kWh_el or 2.5 €-cts./kWh_el are taken into account.

Depending on the feed gas, two further cost ranges can be identified: The case for pure CO₂ methanation and the case for direct methanation of biogas. As in the latter a larger amount of methane is fed to the grid based on the same nominal power of the electrolyser, biogas methanation has a higher profitability range and allows for higher investment cost for the electrolyser. For the same reason, the production cost for direct methanation is less sensitive towards the electrolyser investment cost and the corresponding lines appear with a reduced slope in Figure 9.

As shown in Figure 7, the two SOE-based processes for CO₂ methanation exhibit production costs of 19.4 €-cts./kWh_el for the 1 MW-scale and 17.4 €-cts./kWh_el for the 6 MW-scale respectively. To reach a profitable range, the production cost has to be reduced to below 11 €-cts./kWh_el. As shown in Figure 9, for the higher electricity price this will be at investment costs lower than 500 €/kW_el for the 1 MW case and lower than 960 €/kW_el for the 6 MW-scale. This is a very ambitious cost decrease, nevertheless, under favourable circumstances, a capital cost lower than 1000 €/kW_el in future is considered possible (Schmidt et al., 2017; Adam Christensen, 2020). Literature indicates that capital cost for SOEs may adjust to a similar range as PEM or AEC technology (Schmidt et al., 2017; Buttler and Spliethoff, 2018). A halving of the electricity cost to 2.5 €-cts./kWh_el results in an increase of the allowed investment cost in the 1 and 6 MW case of approximately 2.6-fold to 1260 €/kW and 1.9-fold to 1840 €/kW, respectively.

Due to the reduced production expenses for direct methanation, the maximum allowed electrolyser investment increases to 2260 €/kW_el (1 MW_el) and 2930 €/kWh_el (6 MW_el) for the two scales based on an electricity price of 5 €-cts./kWh_el. Using the lower electricity cost, this numbers increase to 3150 €/kW_el and 3820 €/kW_el. This indicates, that even with moderate improvements in the investment cost of the SOE, profitability of the process can be achieved.

Again, the profitability range can be further extended by also including heat utilisation and oxygen sales. For the heat utilisation, a production cost decrease of around 2% is possible in the pure CO₂ case.