Christoph
Hank
ab,
Svenja
Gelpke
a,
Andrea
Schnabl†
a,
Robin J.
White
*a,
Johannes
Full
a,
Nikolai
Wiebe
a,
Tom
Smolinka
a,
Achim
Schaadt
a,
Hans-Martin
Henning
a and
Christopher
Hebling
a
aFraunhofer Institute for Solar Energy Systems ISE, Heidenhofstraße 2, 79110 Freiburg, Germany. E-mail: robin.white@ise.fraunhofer.de
bKarlsruhe Institute of Technology (KIT), Engelbert-Arnold-Straße 4, 76131 Karlsruhe, Germany
First published on 14th March 2018
The synthesis of sustainable methanol based on renewable electricity generation, sustainable hydrogen (H2) and recycled carbon dioxide (CO2) represents an interesting sustainable solution to integrated renewable energy storage and platform chemical production. However, the business case for this electricity based product (denoted hereafter as eMeOH) under current market conditions (e.g. vs. conventional fossil methanol (fMeOH) production cost) and the appropriate implementation scenarios to increase platform attractiveness and adoption have to be highlighted. The aim of the following study was to perform a dynamic simulation and calculation of the cost of eMeOH production (where electricity is generated at a wind park in Germany), with comparison made to grid connected scenarios. Consideration of these scenarios is made with particular respect to the German energy market and potential for the reduction in fees/taxes (i.e. for electrolyser systems). This evaluation and indeed the results can be viewed in light of European Union efforts to support the implementation of such technologies. In this context, CO2 is sourced from EU relevant sources, namely a biogas or ammonia plant, the latter profiting from the resulting credit arising from CO2 certificate trading. Variation in electricity cost and the CO2 certificate price (in the presented sensitivity study) demonstrate a high cost reduction potential. Under the energy market conditions of Germany it is found that eMeOH production costs vary between €608 and 1453 per tonne based on a purely grid driven scenario, whilst a purely wind park supplied scenario results in €1028–1067 per tonne. The reported results indicate that the eMeOH production cost in Germany is still above the present (although variable) market price, with the economical evaluation indicating that electrolyser and H2 storage represent the lion share of investment and operational cost. Substitution of fMeOH results in CO2 avoidance costs between €365 and 430 per tonne of CO2eq avoided for green methanol produced in Germany. The presented assessment indicates that the eMeOH production cost in Germany will reach market parity in ca. 2030–2035 with the price for the avoidance of CO2eq turning from a cost to a benefit at around the same time. Optimistically, the cost is predominantly influenced by rapidly reducing renewable electricity costs as is already the case in South American and Arabic countries offering the potential for methanol production at a cost of <€500 per tonne.
There has been much interest in the literature and at the industrial level regarding the interface between RE storage and chemical/fuel production schemes such as Carbon Capture & Utilisation (CCU) and Emissions-to-Liquids.8 Currently in Germany, this concept is generally referred to as “Power-to-X” (PtX). X equals gas (H2 or CH4 – Power-to-Gas (PtG)), chemicals and liquid energy carriers (e.g. MeOH, Fischer–Tropsch diesel – Power-to-Liquid (PtL)). Regarding PtL, this scheme is based on the recycling of CO2 from industrial exhaust gases or biomass plants and its hydrogenation (with H2 produced from RE powered H2O electrolysis) to produce long lasting chemical molecules (Fig. 1). Taking MeOH as the desired liquid product, this C1 alcohol offers numerous potential benefits:9
Fig. 1 The production of methanol based on renewable energy, water electrolysis and recycled carbon dioxide, commonly referred to as the “Power-to-Liquid” (PtL) process chain. |
(1) An industrially established synthesis under moderate conditions such as T < 270 °C and p < 80 bar.
(2) A relatively high energy density (at ambient conditions; 16.9 MJ L−1) and suitability as a storage molecule for RE (i.e. via 12.6 wt% (H2)).
(3) A highly versatile platform molecule (e.g. olefins and acetic acid) for higher fuel production.6,10
(4) High volume, existing market (ca. 75 Mt a−1; 2016) and a high ratio of market price to production cost in the case of conventional fossil production.11
However, the extremely low price of CH4 (the main basis for current industrial MeOH production via reforming and syngas conversion6) makes adoption and market entry somewhat difficult for PtL.12
The option to recycle CO2 is attractive as it could reduce industrial costs related to CO2-certification, whilst holistically elevating CO2 from a liability to a potential commercial asset.5 It is important to note that currently the capture of industrial CO2 and its utilisation (CCU) in chemical processes is not included in the EU Emission Trading System (EU ETS)‡ and, for this reason, does not lead to CO2 certificate savings (i.e. not as for Carbon Capture and Storage (CCS)).12,13 For the current energy system, the regulatory context is extremely important to support the economical implementation of CCU.13–15 In January 2017 the European Court of Justice pronounced that in the case of CCU the transferred CO2 is not to be seen as an emission, as it is not released into the atmosphere.16,17 Hereby the emitter (i.e. a lime producing company implementing CCU for the production of precipitated calcium carbonate) is not obliged to purchase CO2 certificates. In this ruling, the court annulled the second sentence of Article 49(1) of the monitoring regulation with reference to the existence of sufficient means of monitoring for the avoidance of feared CO2 loopholes. This case is foreseen as a critical legal precedent for any CCU project which, after careful assessment of its CO2 mitigation potential, is found to result in reduced CO2 emissions. Therefore a comprehensive LCA is mandatory. Therefore, in this study it is assumed that imminent legislation will consider CCU as a part of the EU ETS market resulting in a credit due to the saving of CO2 certificates. For any large scale implementation of CO2 converting PtL processes it is important to shift focus towards biomass based CO2 sources. Only these (besides the still cost intensive direct air capture methods) enable the recycling of CO2 from distributed emitters. For the following economic evaluation we consider an industrial/fossil and a biogenic source of CO2, namely an ammonia and a biogas plant.
Regarding the synthesis of fossil based methanol (denoted hereafter as fMeOH), three key stoichiometric equations are important (eqn (1)–(3)). The reforming and gas cleaning involved in the syngas preparation give rise to high specific CO2 equivalent emissions. These range from 0.50 to 0.77 t(CO2eq) t(fMeOH)−1, ref. 18 and 19 respectively, including direct specific CO2 emissions of 0.24 t(CO2dir) t(fMeOH)−1.20 These emissions are coupled with a very high environmental impact as a consequence of fossil fuel exploration and processing.21 Positively, the synthesis of electricity based methanol (denoted hereafter as eMeOH) based on the direct hydrogenation of a pure CO2 feed is also known (eqn (1)).22,23 According to the equilibrium reaction, the production of 1 metric tonne of MeOH (CH3OH) requires 1.370 kg(CO2) and 188 kg(H2) are converted (producing 558 kg(H2O) as a side product) (N.B. this value ignores CO2 and H2 consumption via the Reverse Water Gas Shift reaction (eqn (3)) which becomes increasingly favoured with rising reaction temperature). This equilibrium reaction occurs under exothermal reaction conditions and as a consequence is favoured at low T and because the volume reducing reaction requires high p. Efficient compression of the feed stream(s) as well as integration of the exothermal reaction heat are important factors determining the overall process efficiency.
CO2 + 3H2 ⇌ CH3OH + H2O, ΔHR(298 K,50 bar) = −40.9 kJ mol−1 | (1) |
CO + 2H2 ⇌ CH3OH, ΔHR(298 K,50 bar) = −90.7 kJ mol−1 | (2) |
CO2 + H2 ⇌ CO + H2O, ΔHR(298 K,50 bar) = +42.0 kJ mol−1 | (3) |
The commercial demonstration of PtL is exemplified by the George Olah Renewable Methanol Plant (Carbon Recycling International, Iceland), operating at 5 million litres per year production. Plant operation here is based on a very low electricity cost due to inexpensive renewable geothermal power and associated low cost H2/CO2 sourcing.24
One significant distinction and possible advantage of developing CH3OH production based on direct CO2 hydrogenation is the avoidance of the cost intensive syngas production step (e.g. CH4 reforming at T > 700 °C). It typically accounts for ca. 60% of total plant investment.22 Whilst replacement of syngas production with CO2 and H2 sources can be relatively capital intensive, depending on the overall process conditions, investment in such a scheme can result in a “multi-benefit” system: MeOH production, recycling of CO2 emissions, system services for the electrical power grid and indirect avoidance of fossil CO2 emissions. In this regard, a number of recent reports have sought to evaluate the cost-effectiveness of PtL, focusing predominantly on evaluation of eMeOH synthesis with little consideration given to the preceding process steps in the process chain.3,11,25–34 Therefore, it is the aim of the following work to investigate different electricity and CO2 purchase options, the fundamental influence of total and partially dynamic system operation and, in turn, the impact on eMeOH production cost, H2 storage requirements and associated investment costs. Furthermore, a complementary parametric sensitivity study is presented, highlighting the most important factors concerning eMeOH production cost (e.g. in a future RE system). Concerning the environmental impacts of PtL-schemes, the economic evaluation presented is concluded with the CO2 avoidance cost – i.e. an important key performance indicator when evaluating technologies for CO2 emission mitigation.
(I) Electricity as the main power source for the process is generated from renewables on-site or purchased from the grid.
(II) H2, the first educt for eMeOH synthesis, is obtained via H2O electrolysis or industrial streams providing high concentrations of H2 (e.g. coke oven gas from steel mills). The PtM process is investigated with the intent to minimise CO2 emissions. For this purpose, H2 used in hydrogenation should not be generated (in-)directly from fossil resources (e.g. steam reforming of CH4 and H2O electrolysis powered by the electricity grid mix). Please note that a valuable by-product of H2O electrolysis is high purity O2, which will be discussed in the following evaluation.
(III) CO2, the second educt for eMeOH production, is captured from industrial processes, biogas or even ambient air.35–38 Source dependent purification of the CO2 is necessary.
(IV) Compression of the feed H2 and CO2 and the associated energy demand. This will be source dependent.
(V) Temporary storage, the size of which will be related to the fluctuation of the electricity production and the range of the systems' dynamics.
(VI) Catalytic conversion of H2 and CO2 to MeOH is based on established technology. Recent studies focus on higher efficiency and dynamic operation.23,31,39,40
(VII) Synthesis is followed by a purification step based here on distillation.
The PtM process steps that are varied within this study are marked in orange (i.e. power supply, CO2 source, H2 storage, and MeOH synthesis; Fig. 1). The non-variable components of the PtM process steps are marked in blue (i.e. polymer electrolyte membrane electrolysis (PEMEL), compression and pressure levels, and distillation).
Low spot market prices are available in times of high production from RE (e.g. low or even negative residual load),43 or during periods of low energy demand (e.g. at night). Therefore, electricity purchase during such periods can aid in the reduction of the final production cost. The stock price is, however, not the biggest share since ca. 80% can be attributed to taxation and fees. Under certain circumstances (e.g. amount of energy consumed/total utilisation hours/specific electricity cost intensity), it is commonplace in Germany, as an industrial customer, to receive a tax reduction down to 15% (EEG 2017). For this study, electricity prices are calculated either with tax and fees (scenarios 1 and 3) or based on the aforementioned reduction of fees and taxes (scenarios 2 and 4). The respective reduction was calculated depending on the specific electric energy demand in the scenarios. Furthermore the calculation model differs between electric energy demand of the electrolyser and the ‘rest’ of the plant (i.e. compressors and eMeOH-plant). According to ‘Energiewirtschaftsgesetz’ (EnWG, §118 (6)) and ‘Stromsteuergesetz’ (StromStG, §9a), electrolysers are found to be exempt from the grid fee and electricity tax, respectively. The exemption from EEG apportionment for electrolysers and the energy demand of additional components in the PtM system takes effect after the first consumed GWh of electrical energy. After analysing the specific tax and fee reduction potential for each scenario, electricity prices are found to be between 3.11 and 3.28 ct€ per kWh with reduction, or 14.29–14.63 ct€ per kWh without the reduction. The partial reduction of EEG apportionment (6.88 ct€ per kWh full/0.18 ct€ per kWh reduced), grid fees (2.26–2.77 ct€ per kWh full/0.00 ct€ per kWh reduced) and electricity tax (2.05 ct€ per kWh full/0.00 ct€ per kWh reduced) is the most significant since they have the largest reduction potentials.
Another possibility for electricity purchase is based on participation in an operating reserve market (support of grid stability), where negative reserve (increase of load/H2 production) and positive reserve (decrease of load/H2 production) could both be provided. In our study, the German operating reserve market has been utilised as a model. Since market entry in this case is simplified (e.g. pooling is possible), there are more and more participants in the market, resulting in decreasing revenues. The exact economic benefit of participation in the control energy market, whilst an important global consideration, is not discussed in this study. Kopp et al. had a first evaluating insight into this topic in their work regarding the PtG plant in Mainz, Germany.44
It is important to note that PtM scenarios using grid electricity only produce eMeOH, if mainly RE is fed into the grid. The RE share will keep increasing following German/EU Climate Targets (>80% RE in 2050).44
Electricity purchase from a local onshore wind park with a rated power of 36 MW (scenarios 5 and 6): for the simulation, values at minute intervals have been applied. If insufficient wind electricity is available for H2 production (i.e. the amount required for eMeOH production), the difference is covered by either the flexible adaption of the eMeOH production rate (scenario 5) and/or an H2 storage (scenarios 5 and 6). A limited flexible operation of eMeOH synthesis is a novelty, and therefore for the purposes of this study is assumed technically feasible. For the wind electricity driven scenarios 5 and 6 an additional grid connection to acquire electricity in periods of low wind supply is not considered for the purpose of a remote renewable scenario. The levelized cost of electricity (LCOE) from onshore wind energy is assumed to be 4.4 ct€ per kWh,45 without any additional fees or taxes since wind electricity is used for personal consumption and without grid connection (as defined by the “Renewable Energy Act” (“Erneuerbare Energiengesetz” EEG) of the German Government in 2014 (§61, subsection 2, sentence no. 2)).46
Rated load, PRL [MW] | 12 |
Efficiency, ηEly [kWh per Nm3] | 4.76 |
Output pressure, PH2 [bar] | 30 |
Grid connected scenario | P RL: 40% (day), 100% (night) |
Wind electricity driven scenario | 0% < PRL < 100% |
Molecular O2 is produced as a high purity by-product and as such is included in this study as a consequence of its valorisation potential. The selling price depends on purity, with O2 purities > 99.99% (and without S or CO impurities) making it useful for many high-tech applications (e.g. medical use in hospitals, electronics industry, and oxy-fuel processing).53 If possible, O2 is utilised locally to avoid the associated energy intensive liquefaction and transport needs.3 For these reasons, the use of O2 from large electrolysers in nearby chemical/energy plants is considered the most appropriate option (e.g. in epoxide/polycarbonate production via oxidation of propylene and CO2 addition).54
In this study, two different CO2 sources are considered, namely a biogas treatment plant and an ammonia plant. In the case of biogas, CO2 is separated during a treatment process of the raw biogas. This is conventionally composed of CH4 (70–50 vol%) and CO2 (30–50 vol%), with the exact composition influenced by the biomass or waste being converted, which also dictates the presence of other compounds and impurities (e.g. N2, H2S, H2O, and organosulfur compounds). Taking Germany again as the example, the quality requirements for the feed-in of treated biogas to the gas grid are regulated.58,59 CO2 sourced from biomass is designated as biogenic.
For CO2 capture from an ammonia plant, it is assumed that CO2 is already a component of an existing ammonia production process and therefore is designated as fossil-based (from steam methane reforming as a process related step in ammonia production).60
In the first of the examined scenarios 1 and 2 (i.e. grid connected, constant supply), no storage is necessary as a consequence of a constant H2 production. The resulting H2 mass flow of 5.47 t(H2) d−1 is directly converted to MeOH. For the other scenarios, H2 storage requirements depend on the assumed dynamics of the MeOH synthesis reactor (Fig. 2). In wind powered scenarios, MeOH mass flow is a function of the actual wind power, the H2 storage pressure and the possible dynamics of the MeOH synthesis reactor (Fig. 3).
(4) |
Recovery period | 10 years | — |
Interest rate | 2.5% p.a. | Ref. 62 |
PEMEL specific investment | 800€ kW−1 | Ref. 48, 49, 63 and 64 |
PEMEL grid connection | 66€ kW−1 | Ref. 48 |
PEMEL stack life-time | 50k hours | Ref. 64 |
PEMEL stack reconditioning cost | 2.0% of spec. invest. | Ref. 64 |
PEMEL maintenance & insurance | 2.0% of spec. invest. | Ref. 64 |
H2 storage specific investment | Depending on storage size | Ref. 65 |
H2 storage maintenance & insurance | 2.0% of spec. invest. | Ref. 62 |
MeOH synthesis specific investment | 810€ t−1 a−1 | Ref. 66 |
MeOH synthesis maintenance & insurance | 10.0% of spec. invest. | Ref. 66 |
Other installation maintenance & insurance | 2.0% of spec. invest. | Ref. 62 |
Technical staff | 14 h per week at 100€ h−1 | Own estimate |
Plant buildings and facilities | 7.0% of spec. invest. | Ref. 64 |
Engineering, planning, delivery, and setup | Inspired by internal experiences | — |
If not included in the component's investment cost, additional costs for delivery and setup are added at a proportionate share of the total investment cost (TIC). Maintenance costs are either based on literature values for the specific component or comparable processes. Not considered in this economic evaluation are costs associated with land, buildings or employees. As mentioned earlier, alongside MeOH, the by-product of H2O electrolysis, namely O2, is also considered as a valuable product in the plant and its potential sales are also included. Credit for O2 is considered at 50€ t(O2)−1,67 which is a relatively conservative value. Atsonios et al.29 and Matzen et al.30 considered a credit of ca. 75€ t(O2)−1 for their calculations (see Table 5) (NB: any possible future increase of installed electrolysis capacity may act to decrease the impact of the O2 market value, unless a large increase in oxidation chemistry is anticipated).
The supply cost of CO2 differs for different sources. From the considered biogas plant source, the CO2 production cost is assumed to be 0€ t(CO2,capt.)−1, as this is essentially a waste product from biogas cleaning and therefore, the cost for separating CO2 is assigned to CH4 production cost. Based on ammonia production, CO2 capture costs are 3€ t(CO2,capt.)−1.36,68 Furthermore, assuming that avoided CO2 emissions will also be considered within the European Emission Allowances (EUA) in the near future, a projected price of −10€ t(CO2,emitt.)−1 is obtained. These allowances are tradable and the resulting CO2 feedstock price is −7€ t(CO2)−1 (Table 3). Biogas treatment plants do not take part in the EUA trade, since the emitted CO2 is of biogenic origin.
Biogas treatment plant | Ammonia plant | |
---|---|---|
Capture cost [€ t(CO2,capt.)−1] | Excluded | 3 |
CO2 certificate price [€ t(CO2,emitt.)−1] | 0 | −10 |
CO2 feedstock price [€ t(CO2)−1] | 0 | −7 |
Output pressure [bar] | Atmosph. | Atmosph. |
Environmental burden | Biogenic origin | Fossil origin |
The operating and full load hours of the PEMEL and the MeOH synthesis reactor differ for the different scenarios. For purely wind electricity driven scenarios, the PEMEL full load hours are in the range of 3585 to 3896 h a−1 and the operating hours (the period during which the PEMEL is operational at any load factor) at 7044 h a−1 (Table 4). The costs for maintenance and replacement of the PEMEL stacks do reflect the specific operating hours and a predicted stack life-time of 50000 h. Furthermore, the yearly amortisation payments are calculated based on a linear depreciation over 10 years.
Grid connected constant supply | Grid connected electricity price driven | Wind electricity driven flexible | Wind electricity driven non-flexible | |
---|---|---|---|---|
PEM-electrolysis – production capacity [t a−1] | 1906 | 992 | 791 | 1210 |
PEM-electrolysis – energy demand [GWh a−1] | 100.80 | 55.44 | 43.02 | 46.75 |
PEM-electrolysis – operating/full load hours [h] | 8400/8400 | 8400/4620 | 7044/3585 | 7044/3896 |
Hydrogen storage – volume [m3] | — | 6 | 345 | 2840 |
Carbon capture – capturing capacity [t a−1] | 15225 | 7765 | 6392 | 9780 |
Methanol reactor – production capacity [t a−1] | 10031 | 5139 | 4188 | 6408 |
All scenarios take a PEMEL as a basis, even though the grid-connected scenarios would be able to function reasonably using an AEL electrolyser. As mentioned earlier, the TIC will likely decrease in the future as a consequence of further technology development, efficiency improvements, and more widespread use, with a target value of 440€ kW−1 considered desirable for large water electrolysis systems >10 MW.67 The 12 MW PEMEL included in this investigation has a TIC of 9.6 M€ (without a grid connection) and 10.4 M€ (with grid connection), representing ca. 60–70% of the TIC of the overall process. An exception to this observation is the wind-driven scenario using a non-flexible MeOH reactor (scenario 6), since a large H2 storage is included, representing a significant expense in itself (i.e. TIC (PEMEL) = 35%; TIC (PEMEL + H2 storage) = 79%). Mignard et al. arrived at a similar conclusion in a previous report where the electrolyser's share of the TIC is within 65–82%, by far the biggest share, followed by the H2 storage and compression cost, with a share of 9–16%.27 The impact of investment in the MeOH reactor depends on the annually produced MeOH, which is assumed in this study to be 810€ t−1 a−1.66 Therefore, the grid connected scenario without H2 storage (scenario 1) has the highest MeOH reactor cost (3.3 M€), since this scenario produces the greatest amount of product. The potential of additional cost for a flexible reactor is not considered in detail. Excluding this mentioned scenario, the MeOH synthesis reactor's influence is comparatively low. For future assessments a detailed investigation of the cost incurred by a dynamic reactor design should be included. Additionally, since methanol from syngas with high CO2 content and H2 potentially offers higher purity (free of most ketones, paraffins and other by-products) compared to methanol from syngas, a reduced demand for distillation could have a cost decreasing effect.23,69
(5) |
Currently, the MPC of all considered scenarios (Fig. 6) is above the actual market price (i.e. Methanex average 2017 (Jan to June):70 410€ t(eMeOH)−1), which has featured a low-level upward trend of +3.20€ per month over the last 2.5 years. The scenario with the lowest MPC (scenario 2B, grid connected with constant supply, fee reduction, without H2 storage, CO2 from biogas, 608€ t(eMeOH)−1) costs 1.5 times the possible market revenues, whereas the most uneconomical scenario (scenario 3A, electricity price driven, without fee reduction, with H2 storage, ammonia, 1453€ t(eMeOH)−1) is 254% above the 2017 market price level. The electricity price has a huge impact on the MPC: where a fee reduction is not possible, MPCs are in the range of 1284 to 1453€ t(eMeOH)−1. If taxation and fees do not have to be paid in full, the MPC will decrease to 608–782€ t(eMeOH)−1. The wind driven scenarios feature MPCs in the range of 1028–1067€ t(eMeOH)−1, which are characterised by large H2 storage capacities and associated high investment costs.
In this study, a positive value of 50€ t(O2)−1 for selling the electrolyser by-product was used. Previously Atsonios et al. considered an O2 revenue of 87€ t(O2)−1 reflected in a decrease in eMeOH production cost (ca. 10% reduction as a consequence of O2 sales).29 Matzen et al. also included O2 sales in their process evaluation and concluded that a price of 75€ t(O2)−1 had a positive impact on the economic feasibility of their process.30 Likewise, Rivarolo et al. assumed a much higher market value of 150€ t(O2)−1.34 Compared to the 50€ t(O2)−1 used in this study, this results in a difference of 150€ t(eMeOH)−1 (1.5 t(O2) t(eMeOH)−1 produced). For further analysis a detailed evaluation of the necessary process steps for pressurizing, bottling and transportation of O2 has to be included as well as a discussion of the development of the O2 market price in the case of increasing market penetration for electrolyser technology.
Fig. 7 Sensitivity study for the wind powered, flexible synthesis based on biogas-derived CO2 (scenario 5B). |
Electricity | Electrolyser | CO2 | Storage | Methanol reactor + distillation | Source | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Source | Price [ct€ per kWh] | Type | Spec. invest [€ kW−1] | O2 price [€ tO2−1] | Source | Spec. costa [€ tCO2−1 a−1] | H2 | CO2 | Spec. investb [€ tMeOH−1] | Plant capacity [t d−1] | Pressure level [bar] | Plant depreciation time [a] | Methanol production cost [€ tMeOH−1] | |
a Without including CO2 certificate revenues. b Whole methanol plant. c Currency exchange rate in the year of the study (average). d Currency exchange rate of 2017 (average). e Calculated with the daily needed hydrogen amount for methanol production. f Outside system boundaries. g Assumed purchase price for hydrogen: 3090€ t−1. h Assumed purchase price for hydrogen: 3000€ t−1. i Conversion with low heating value (LHV), RES = renewable energy sources, CPP = coal power plant, and CHP = combined heat and power plant. | ||||||||||||||
Grid | 3.74–13.43 | PEM 12 MW | 800 | 50 | Biogas/ammonia | 0/3 | No | No | 1514 | 29 | 40 | 10 | 608–1615 | This study |
Grid | 3.67–14.13 | PEM 12 MW | 800 | 50 | Biogas/ammonia | 0/3 | 6 m3 | No | 2745 | 15 | 40 | 10 | 816–1969 | This study |
Wind onshore | 4.4 | PEM 12 MW | 800 | 50 | Biogas/ammonia | 0/3 | 345 m3 | No | 3830 | 12 | 40 | 10 | 1040–1051 | This study |
Wind onshore | 4.4 | PEM 12 MW | 800 | 50 | Biogas/ammonia | 0/3 | 2840 m3 | No | 4144 | 19 | 40 | 10 | 956–967 | This study |
Hydro power | 1.3 | — | — | — | Flue gas/atmosphere | — | — | — | — | 200 | — | 15 | 258/387 | 1999-Specht and Bandi |
Hydro power | 2.5 | — | — | — | CPP/atmosphere | — | — | — | — | 200 | — | 15/20 | 529/717 | 1999-Specht et al. |
RES | 1.5–2.2 | AEL 100/500 MW | 422–1409 | 102 | Flue gas (CPP) | — | Floating head/pressurised tanks | — | 1299–2420 | 178–372 | 50 | 15 | 498–711c | 2003-Mignard et al. |
429–613d | ||||||||||||||
Grid/RES | 4 | AEL 2 MW | 200 | Sold/used | Flue gas (CHP) | 15 | 100–900 tH2 underground gas storage | No | — | 890 | 144 | — | 555 | 2010-Clausen et al. |
Grid/CPP | 2.9–5 | AEL 140 MW | 376 | 87 | Flue gas (CPP) | 44 | Partially | — | — | 318 | 65 | 25 | 870–913 | 2015-Atsonios et al. |
Wind | 6.3 | AEL 41 MWe | 440 | 75 | Ethanol plant | — | Pressurised storage | Liquid storage | 850 | 97 | 50 | 10 | 324–781c | 2015-Matzen et al. |
389–938d | ||||||||||||||
CPP | 9.5–12.1 | —f,g | —f | —f | Flue gas (CPP) | 0 | — | — | 497 | 1320 | 76 | 20 | 724 | 2015-Perez-Fortes et al. |
—f | 9.3 | —f,h | —f | —f | Purchased | 50 | Yesf | No | 1329 | 138 | 75 | 10 | 980 | 2015-Tremel et al. |
RES | 2.6/3.3 | SOEC 50 MW | — | Used | Flue gas (CPP) | 20 | — | Buffer storage | 1000 | 96/235 | — | 20 | 450/484 | 2015-Varone et al. |
— | 5 | SOEC 20 MW | 1469–5785 (SOEC) | Sold | — | 3–10 | — | — | 6115 (SOEC) | 58 (SOEC) | 78 | 20 | 5459 (SOEC) | 2016-Rivera-Tinoco et al. |
PEM 24 MW | 600 (PEM) | 883 (PEM) | 49 (PEM) | 891 (PEM) | ||||||||||
RES | 1–5 | PEM 2190 MW | 700 | 40 | Flue gas | 40 | — | — | 1247–2495 | 5003 | — | 15 | 400–2775i | 2016-Bertau et al. |
RES + grid | 3–5 | AEL 1 MW | — | 150 | Biogas/purchase | 10 | — | — | — | — | 80 | 10 | — | 2016-Rivarolo et al. |
Due to the different parameter assumptions used in the calculation of eMeOH production costs, the results vary considerably from study to study. Therefore selection of system boundaries, system components and framework conditions is critical as they have a significant impact on the final results and as a consequence, different studies should be compared carefully.
In this context, CO2 sourcing differs from study to study. Typically, flue gas of a Coal Power Plant (CPP) is used. Varone et al. evaluated an integrated oxy-fuel/co-electrolysis process, where the oxy-fuel coal plant delivers the CO2 and the O2 by-product from the electrolysis is used for the oxy-fuel process.32 Such combined processes are indeed logical but have not been considered thus far by other studies. Besides CO2 captured from coal fired power plants is of fossil origin and should therefore not be part of a future energy system with the intention to provide sustainable fuels. The use of CO2 from the atmosphere has been considered by Specht et al.,25,26 and was recently revisited by the German-Swiss initiative of Audi, Sunfire and Climeworks.71 More recent studies have considered this possibility but assess the acquisition cost to be still too high (e.g. Bertau et al.3). The EUA has, except for this study, so far not been considered by other authors. Depending on the source, the assumed costs to provide CO2 lie between 0 and 50€ t(CO2)−1.
Compared to the actual MeOH market price of 410€ t(MeOH)−1 (Methanex, average 2017),70 almost all reported production costs are above the market price and the production via H2 and CO2 is as such not considered economical (when excluding additional O2 sales). The cost of producing fMeOH from conventional processes via natural gas or coal varies depending on the geographical location of the production site between 51€ t(fMeOH)−1 and 408€ t(fMeOH)−1. They are therefore less expensive than any process based on RE, H2 and CO2.6 If legislation or regulation regarding either RE or CO2 conversion is established (e.g. within the EU), it remains to be seen how long these relatively low prices will be maintained. The smallest difference compared to the actual price level was reported in 1999, by Specht and Bandi (i.e. an eMeOH production cost of 258–387€ t(eMeOH)−1).25 This report was based on an unrealistically low electricity price of 1.3 ct€ per kWh (in 1999), leading in turn to low eMeOH production costs. However, recent reports demonstrate the rapid reduction in renewable electricity costs (e.g. solar power in Dubai72,73 or Chile74,75 with prices <3 ct€ per kWh).
The highest eMeOH production costs (excluding the production via High Temperature Solid Oxide Electrolysis)33 with >2000€ t(eMeOH)−1 were reported by Bertau et al.6 In this work, eMeOH production costs depended on the electricity price and operating hours per year. With less than 2000 h a−1, the MeOH production costs rise sharply, reaching values up to 2800€ t(eMeOH)−1. Their calculations show clearly that the maximisation of operation is desirable regarding an economical production. A comprehensive PtL-overview focussing on electrofuels for the transport sector is provided by Brynolf et al.(2018).76
(6) |
The CO2 avoidance cost (CO2,AC) (eqn (6)) is expressed as the ratio of the differences between the production cost (PC) of green (‘electricity based’) and fossil based methanol (PCeMeOH and PCfMeOH respectively) and their respective global warming potentials (GWPs).
The CO2,AC can be seen as a key performance indicator when discussing the future of CCU processes and the establishment of an economy covering the majority of its C-demand with recycled carbon. It is important to note that the CO2 avoidance cost is an indicator for the connection of economic and ecological efficiency besides many other ecological parameters such as water demand and indirect land use change or occupation. For a holistic evaluation of PtM and other CCU processes and their comparison to fossil based references a multitude of economic and, even more importantly, ecological parameters have to be included in a process' evaluation.
For the production cost of green methanol the results obtained in scenarios 5B and 6B with production cost of 1028 and 1062€ t(eMeOH)−1, respectively, have been used. The production costs for fossil methanol depend on geographical location, ranging from 51€ t(fMeOH)−1 (Saudi Arabia) to 408€ t(fMeOH)−1 (Europe) and are dominated by the cost for the syngas feedstocks.77 In order to put a price on the amount of avoided CO2eq not emitted within the European region we decided to compare ‘our’ green methanol pathway (mainly dominated by the German electricity prices) with any fossil-based methanol production facility located within European borders. Therefore for this evaluation the production costs for fossil methanol are set to a ‘European level’ of 400€ t(fMeOH)−1.|| It could be argued that for other regions featuring significantly lower production costs for fossil methanol such as Saudi Arabia or China the difference between cheap fossil and cost-expensive green methanol production would increase. But likewise a lower levelised cost of renewable electricity would have a clear decreasing effect on green methanol production cost and would partly balance the cheap market conditions for the fossil production.
Regarding the GWP for the green methanol production the values are based on a life cycle assessment based on hydrogen from a wind-powered PEM electrolyser in combination with CO2 from a biogas upgrading plant** (representing scenarios 5B and 6B). The LCA was performed using the Umberto NXT Universal software with the Ecoinvent-v3.3 database. With a GWP of 506 kg(CO2eq) t(eMeOH)−1 during the production phase and an uptake (‘molecular binding’) of 1374 kg(CO2) t(eMeOH)−1†† the PtM process can be seen as ‘netto-negative’ (506–1347 = −868 kg(CO2eq) t(eMeOH)−1) in terms of CO2eq-emissions when evaluated by a cradle-to-gate approach. Expanding the system boundaries by containing also the utilisation of the produced methanol (cradle-to-grave, C2G),‡‡ all C-content will be oxygenated (1388 kg(CO2eq) t(MeOH)−1) again and released into the environment. Utilisation could be in the MeOH-form as a partial fuel substitute or further processed to downstream derivatives such as DME or OME3–5. In this case the resulting cradle-to-grave emissions would add up to 520 kg(CO2eq) t(eMeOH)−1.§§
As a reference for the GWP of a European based fossil methanol production process we used the Ecoinvent process ‘methanol production [GLO] from natural gas’.¶¶ Where possible the contained activities have been adjusted to an assumed methanol production in Europe. This accounted for the supply of natural gas for syngas production and the market for electricity as well as the provision of tap water. The adjusted EU-based fossil methanol production results in slightly higher specific CO2eq-emissions (cradle-to-gate: 526 kg(CO2eq) t(fMeOH)−1 (global) vs. 623 kg(CO2eq) t(fMeOH)−1 (EU-based)) which mainly arise due to the transport intensive provision for the syngas production with natural gas/process heat in Europe (GWP100a: +18.3 and +33.4% resp.). Concurrently the electricity supply shows a lower GWP (GWP100a: −36.0%) presumably due to higher efficiencies in fossil electricity production and higher shares of renewables within the EU. For fMeOH also a complete oxygenation of the C-content is assumed for the utilization phase (1388 kg(CO2eq) t(fMeOH)−1) resulting in cradle-to-grave-CO2eq-emissions of 2011 kg(CO2eq) t(fMeOH)−1.
Based on these values the CO2 avoidance costs are 421€ t(CO2)−1 (scenario 5B) and 444€ t(CO2eq)−1 (scenario 6B) giving emitted CO2 significantly higher prices than the recent prices envisioned by the European Emission Trading Scheme (EU-ETS) with CO2-certificate prices of ca. 5€ t(CO2eq)−1 for the last 12 months (Sept 16 to Aug 17).||||
Fig. 8 contrasts the resulting development of eMeOH with an increasing methanol market price, demonstrating a prediction for market parity in ca. the year 2032. At around the same time the CO2 avoidance cost will turn negative (based on the assumptions made regarding investment and operational costs) which can be interpreted as the market offering an economic incentive to the industry for utilising their CO2 emissions instead of emitting them. These points of time fit the published results for the development of the market demand for electrolyser technology showing a clear market launch from the year 2032.78,79 Furthermore a future increase of the GWP for fMeOH production could be assumed due to the necessity of a more intensive exploration of fossil resources and longer transport distances. This would have an additional lowering effect on CO2 avoidance cost, which was not considered here.
Fig. 8 Estimation for the future development of green methanol production cost and CO2 avoidance cost in the case of fossil methanol substitution. |
Under the evaluated process conditions in this study, the production of eMeOH in Germany is currently not economically competitive, if one presupposes the competition with fossil based large capacity methanol production and its current low price. The investment and operational costs are currently inhibitory and therefore, the eMeOH production price exceeds the expected revenues. Currently, conventional large scale methanol production based on CH4 reforming has a very low price due to very inexpensive fossil supply (e.g. from increased fracking in the USA and increased production of fMeOH from coal in China). A high degree of capacity utilisation is advantageous and favourable for low eMeOH production cost. Therefore maximised plant operation (e.g. in terms of hours) is also to be highlighted as an important factor in achieving eMeOH economic feasibility.
The presented sensitivity study demonstrates that a decrease in the specific investment costs of the PEMEL leads to a major economic efficiency improvement for the described PtM process. Heading towards the target value for specific electrolyser investment (target value: 440€ kW−1) is in general an important step towards the economic feasibility of PtL-systems. The low-emission eMeOH production scenario (wind energy driven) had a higher production cost due to the additional requirement for H2 storage (i.e. 2nd largest share of investment costs). A reduction in specific H2 storage size and costs would therefore be beneficial. The possibility of underground H2 storage in salt caverns (e.g. in northern Germany) depends on local conditions but offers a promising way for mitigating the strong interlinkage between H2 storage volume and its associated investment costs.64 In the long-term it will be necessary to offer possibilities for a more dynamic operation of the synthesis step to reduce the storage demand for H2.
To successfully introduce any PtL process, it is initially important to think of potential synergies: the methanol production of Carbon Recycling International is underpinned by low-cost geothermal electricity and CO2 from a geothermal power plant (located adjacent to the plant). Thyssenkrupp AG is aiming to significantly cut their CO2 emissions by utilising the waste gases from steel production (coupled with additional H2 production) to generate side streams of fuels and chemicals.80 The presented PtL scenarios are all on a small-scale compared to fossil based plants on the Mt-scale. Future large-scale PtL-plants, coupled with existing chemical or heavy industry, would also have a number of advantages: instant access to low-cost grid electricity, availability of (highly concentrated) CO2 streams (which still need purification) or unused H2 in off-gases, if necessary access to HT-process heat/steam and last but not least, the vast cost-reducing effect of a PtL plant scale-up.
Regarding policy instruments to support the introduction of economically viable PtL schemes, amendment of general policy frameworks and taxation systems (e.g. special grid fees for energy storage and grid-supporting technologies, and an improved European Emission Trading System) would be beneficial. A modified taxation system for CO2 emissions could generate a CO2 certificate price and market conditions appropriate for an industrial business case for CCU. Another approach could be a tax reduction for “renewable fuels”. The gradual reduction of fossil fuel subsidies would also aid in the long term attractiveness of PtM schemes, particularly as the amount of the RE share in the electric grid increases. The proposed ‘double counting’*** of 2nd generation biofuels according to the revision (iLUC Directive 2015/1513/EC81) of the RE directive RED (2009/28/EC82) is a first step when also including CO2-based fuels. Regarding fossil-based production, environmental costs and impacts are generally not internalised and therefore factoring environmental burdens would lead to a higher price of fMeOH. One example is the introduced foundation for climate protection and CO2-compensation (Foundation for Climate Protection and Carbon Offset (Stiftung Klimaschutz und CO2-Kompensation, Klik)),83 following the revision of Swiss CO2-law. It commits producers and distributors of fossil fuels to provide compensation for the environmental impacts of their products and in turn provides investment in funding programs focusing on Swiss GHG emission reduction.
The presented analysis highlights that the CO2 avoidance costs for the evaluated wind electricity based scenarios are currently in the range of 421–444€ t(CO2eq)−1 avoided and strongly depend on the process and market conditions within the selected scenarios. The process conditions (electrolyser efficiency, synthesis pressure, source of CO2, etc.) not only influence eMeOH production cost but also the results for the CO2eq footprint and thus the CO2 avoidance cost. It has been shown that for future technological optimisation of some CCU key components (resulting in the improvement of central economic process parameters) even for cost intensive 100% RE based set-ups the green methanol production cost can be strongly decreased (543€ t(eMeOH)−1, −47%, scenario 5A) within the next two decades bringing market competitiveness with fossil based methanol within the realms of possibility. In the same time frame the CO2 avoidance cost will feature a clear drop (−124%) resulting in negative values and through this provide an incentive for industries to reinterpret their CO2 emissions from waste to feedstock.
It is important to consider that the CO2 avoidance cost should be used only as one key-criterion for the selection among the variety of possible CCU processes.84 In combination with the specific CO2 avoidance potential of each PtL-/CCU-technology, the most cost- and eco-efficient routes should therefore be selected. In this context, a detailed Life Cycle Assessment (LCA) is mandatory to identify the specific CO2 avoidance cost and potential. Positively, the analysed technology will be supported by (and also support) the rapidly growing and intended expansion of RE (e.g. in Germany and the EU in general) over the next few decades. This growth will necessitate consideration of technologies such as PtL, further supported in terms of economic readiness by the ever reducing RE price per ct€ per kWh. PtX technologies will be one important pillar of the integrated-energy concept (‘sector-coupling’) within the European markets of electricity, heat and mobility. Further postponement of integrating renewable energies in the mobility sector and chemical industries will further impede the climate protection programs.79
Our results demonstrate that off-grid wind electricity is promising for eMeOH production, but in the case of an isolated wind energy system and without large H2 storage it cannot ensure a high degree of capacity utilisation. For realisation of an off-grid renewable PtL process, the coupling of wind power with photovoltaic or solar thermal electricity generation and/or additional baseload supplies (e.g. biogas plants or combined heat and power plants based on biomass, offering also a biogenic CO2 point source) would lead to higher capacity utilisation and a reduced storage demand. Thinking globally, relocation to regions with higher solar irradiance (e.g. the Maghreb region, Australia or south western parts of North America) and/or reliable wind power (e.g. Chile, Peru, South Africa, or Scandinavia) offers the possibility of a low cost RE supply. Low cost RE electricity of around 2.5 ct€ per kWhel (reflecting recent price bids for PV electricity72–75) and a higher degree of capacity utilisation due to a less fluctuating RE potential MPC of ∼500€ t(eMeOH)−1 and below are in a realistic range. Regions with a current and prospectively low RE production cost provide opportunities for the further development of a business case for PtL (and PtX) technology within the coming decade. Further installation of pilot plants and the associated operational experiences, coupled with expected economies of scale, will lay a path to improved PtL economics, even in countries where RE electricity costs are currently prohibitive.
Footnotes |
† Current address: University of Oxford, Department of Engineering Science, Parks Road, Oxford, OX1 3PJ, UK. |
‡ Annex I of the EU Directive 2003/87/EC underlines that “under the EU ETS rules for any other transfer” than in the case of long-term geological storage “of CO2 out of the installation, no subtraction of CO2 from the installation's emissions is allowed”. |
§ Listed examples for equivalent (eq) emissions: CO2eq: climate change, CFC-11eq: stratospheric ozone depletion, and SO2eq: acidification. |
¶ In this evaluation: methanol production based on syngas from steam-reforming of natural gas. |
|| In the case of assuming lower fossil MPC for example for any production facility in Saudi Arabia or China, the green MPC would also decrease presumably because of lower renewable electricity generation cost (PV electricity in Abu Dhabi @ 2.45 ct€ per kWh; in Chile @ 2.91 ct€ per kWh; and levelised cost of electricity for large PV plants within the G20 states below 4.10 ct€ per kWh). |
** To be reported in a forthcoming article. |
†† 41 kg of CO2 are vented during the synthesis step. Therefore 1388-41 = 1347 kg of CO2 per t(eMeOH) are ‘bound’ in the product. |
‡‡ The cradle-to-grave assessment does not consider impacts resulting from transportation of the produced methanol and any recycling of the infrastructure. However it is assumed that impacts resulting from these phases in the eMeOH and the fMeOH are in a comparable range. |
§§ GWP100a, method CML 2001, allocation cut-off. |
¶¶ Original ecoinvent activity methanol production [GLO]: GWP100a: 52545 kg(CO2eq); method CML 2001, allocation cut-off. |
|||| EU CO2 Emission Allowances (Sept 16 to Aug 17) according to the European Energy Exchange. |
*** Double counting of advanced/second generation biofuels: regarding the extension of the Renewable Energy Directive (RED) 2009/28/EC (28.04.2015) member States must ensure that 10% of the final energy consumption in the transport sector is provided by renewable sources. The share provided with advanced/second and/or third generation biofuels is double weighted. |
This journal is © The Royal Society of Chemistry 2018 |