Access the full text.
Sign up today, get DeepDyve free for 14 days.
Electrolysers: AEC, AEM, PEM and SOE for hydrogen (and syngas) production © 2021 sbh4 GmbH AEC AEM PEM SOE Notes: - In the AEC, AEM and PEM, + – + – + – + – O H O H O H Airplus O H (plus CO) 2 2 2 2 2 2 2 2 lye or water flow from the electrolyser cell with the oxygen and/or hydrogen – – + 2– OH OH O gases. These liquids are mixed and recirculated to the electrolyser. - Air is used to purge the SOE anode to avoid oxygen accumulation which may present a hazard at the high operating temperature. - Bipolar plates made of stainless steel (titanium for PEM) are used to stack H O H O H O Air H O as 2 2 2 2 adjacent cells in each as water as water as water steam electrolyser type. (plus CO ) Alkaline Electrolysis Cell Anion Exchange Membrane/ Polymer Electrolyte Membrane/ Solid Oxide Electrolysis Cell AEC Alkaline Electrolyte Membrane Proton Exchange Membrane SOE/SOEC AEM PEM/PEMEC Electrode material – Cathode: Ni, Co or Fe – Cathode: Ni / Ni alloys – Cathode: Pt/Pd – Cathode: Ni – Anode: Ni – Anode: Fe, Ni, Co oxides – Anode: lrO /RuO – Anode: La/Sr/MnO (LSM) 2 2 or La/Sr/Co/FeO (LSCF) Electrolyte Lye: 25–30% Potassium Anion Exchange ionomer Fluoropolymer ionomer Zirconium Oxide with ~8% Hydroxide solution in water (e.g. AS-4) (eg Nafion, a DuPont brand) Yttrium Oxide 100% electrical power 100% electricaI power Energy source 100% electricaI power ~25% heat from steam, ~75% electrical power 2 2 2 2 Current density Up to 0.5 A/cm 0.2 – 1 A/cm Up to 3 A/cm Up to 0.5 A/cm Hydrogen Hydrogen Hydrogen (or syngas if fed Hydrogen or syngas Hydrogen with steam and CO ) product Up to 35 bar H , 1 bar 0 Gas outlet pressure Up to 40 bar Up to 40 bar Close to atmospheric 2 2 Cell temperature ~80 °C ~60 °C ~60 °C ~750 to 850 °C Keywords: green hydrogen; microgrid; stand-alone power system; SAPS; hydrogen storage; electrolysis; fuel cells; AEC; PEM; AEM; SOEC Introduction provide sustainable electricity supply for such remote off- grid applications. Renewable electricity production, mainly from solar Due to the different frequencies associated with the and wind, increased considerably in the last two dec- volatility of solar, wind and hydro electricity generation, ades worldwide. Due to the increased serial production different energy-storage technologies are necessary for of solar panels and wind turbines, the investment costs different timescales. Battery-based electricity storage for these has plunged by >90%. In addition, innovative is typically only viable for several hours. In general, bat- small-scale hydro turbines are gaining momentum. teries are not appropriate to overcome extended periods Start-ups like Smart Hydro, Suneco Hydro and Energy of windless, sunless ‘dark doldrums’ that persist in many Systems & Designs offer turbines in the range of 2–10 locations over the course of a year. kW. Some of them do not even need a weir or dam to On the other hand, green hydrogen can be a clean en- work. These turbines compliment established players ergy carrier with seasonal storage potential and can gen- like Voith Hydro, which offers turbines with rated power erate electricity on demand. Depending on the required capacities of 50 kW–1.2 MW. But electricity from solar, storage size, different hydrogen storage are favourable. wind and hydro are subject to daily, weekly and seasonal fluctuations. Off-grid electricity supplies in remote communities or mine sites, which have a decent electricity demand 1 Off-grid power supply based on but are not connected to the main electricity grid due to hydrogen-storage solutions their remoteness, were generally driven by diesel gensets 1.1 Off-grid mine sites in the past. The transportation costs of fuel to these re- In 2016, a behind-the-meter microgrid energy-storage mote locations are exorbitantly high. Hence, a combin- system was implemented at the Raglan Nickel mine in nor - ation of wind turbines, solar panels and potentially also thern Canada Fig. 1 . Electricity for the mine is provided small-scale hydropower might be economically viable to 30% KOH (aq) Anode Diaphragm Cathode 30% KOH (aq) Anode Membrane Cathode 1% KOH (aq) Anode Membrane Cathode Anode Electrolyte Cathode Downloaded from https://academic.oup.com/ce/article/5/3/441/6333454 by DeepDyve user on 03 August 2021 Borm and Harrison | 443 Fig. 1: The wilderness of northern Canada where mining is common Source: shutterstock.com. by a wind turbine and power delivery is subject to fluc- were to produce hydrogen at a lower pressure of 10 bar. tuations in the weather conditions. In order to compen- The electricity output from the stored hydrogen is 90 kWh sate for that, the scheme uses a 350-kW rated HySTAT-60® in total and corresponds to a fuel-cell efficiency of 54%. alkaline electrolyser and a 200-kW rated HyPM® polymer electrolyte membrane (PEM) fuel cell, both from the 1.3 Off-grid consumer applications Hydrogenic range. The hydrogen storage has a capacity of 4 MWh = 120 kg . Three horizontal steel ‘bullet’ storage Different hydrogen-storage technology is applied at the re- H2 H2 vessels operate at the electrolyser outlet pressure of 10 cently launched LAVO system  that is targeted at con- bar . Due to the relatively low output pressure from that sumer markets. It integrates one 5-kW PEM fuel cell and type of electrolyser (10 bar), the working pressure differ - two 2.5-kW AEM electrolysers with a predicted stack life- ence—which defines the total hydrogen capacity—might time of ~30 000 hours. With an assumed annual utilization be only 8 bar (10 – 2 = 8 bar). Therefore, the necessary of 3000 hours, that would result in stack replacement after storage volume for this amount of energy storage is very 10 years of operation. The electrolyser produces hydrogen large, at 185 m . at a pressure of 35 bar for storage on a newly designed metal hydride system that can store ≤2.4 kg of hydrogen, which corresponds to 40 kWh of electricity assuming a 1.2 Alpine refuge huts fuel-cell efficiency of 50%. Since 2015, an integrated off-grid energy-storage system designed by PowiDian has powered the alpine refuge at Col 1.4 Can external hydrogen compression be du Palet, in the French Alps. At its heart, a 2.5-kW Anion favourable? Exchange Membrane (AEM) electrolyser from Enapter pro- duces ≤0.5 Nm /h hydrogen at 30 bar . The hydrogen, As an alternative to using a pressure vessel to store hydrogen which is mainly produced during the summer months at the electrolyser output pressure, the use of high-pressure from solar panels, is stored safely in a pressure vessel in compressed-gas hydrogen storage can be favourable for ap- a separate shed. During winter, when the solar panels are plications where space for the compressor is available and covered in snow, a 2.5-kW fuel cell delivers electricity from portability is not a priority. An example might be the in- the hydrogen on demand. Due to the increased working jection of hydrogen into the natural-gas pipeline grid for pressure difference of ≥28 bar, the storage volume for the admixing with methane. A maximum pressure of 100 bar is 5 kg hydrogen (= 166.6 kWh ) is only 2.2 m . This is consid- typically employed in transmission pipelines for natural gas. H2 erably smaller than would be required if the electrolyser Usual outlet pressures of alkaline and PEM electrolysers are Downloaded from https://academic.oup.com/ce/article/5/3/441/6333454 by DeepDyve user on 03 August 2021 444 | Clean Energy, 2021, Vol. 5, No. 3 ~30 bar. To achieve the required injection pressure, hydrogen vice versa. As an alternative scheme to overcome this low- compression is required. As the pressure ratio from 30 to 100 capacity factor, unitized regenerative fuel cells (URFCs) bar is only 3.33 (100/30 = 3.33), a single-stage piston com- could be used to operate either in fuel-cell mode to pro- pressor could be utilized. Nevertheless, even a single com- duce electricity from hydrogen or in electrolysis mode to pression unit for hydrogen adds considerable CAPEX to the produce hydrogen from electricity. Due to this flexibility, whole project. OPEX is also impacted due to the requirement a URFC could potentially achieve a much higher capacity for the maintenance of moving parts and the compressor- factor. motor power demand. Ahluwalia et al.  recently analysed different hydrogen- storage technologies. For the compressed-gas storage in 2 Reversible fuel-cell technologies for steel tubes, he found a minimum CAPEX of 516 $/kg . For off-grid applications H2 his analysis, he found the optimum with coated steel pipes 2.1 Reversible PEM-based systems (inner diameter of 585 mm) and a working pressure differ - In the 1990s, NASA conducted , for its extraterrestrial ence of 92 bar (100 – 8 = 92 bar ). External compression is space activities, extended research into URFCs, particu- necessary in order to achieve such a high working pres- larly for energy storage. These were based on the PEM tech- sure difference, which adds additional CAPEX and OPEX. nology. URFCs based on PEM or AEM may be an option for On the other hand, when reducing the working pressure energy storage on future Mars missions (Fig. 2) in outer- difference, also the relative storage capacity decreases space applications where payload is everything and stack proportionally, while the specific CAPEX increases accord- lifetime is not as important. ingly. The results of this simple analysis are shown in Table Later, at the beginning of the 2000s, Proton Energy, now 1 and the usage of an identical steel tube was assumed, Proton OnSite, a subsidiary of NEL, offered their reversible for the sake of simplicity. With a more sophisticated opti- product UNIGEN . Despite these initial product launches mization, the break-even point at which external hydrogen and various ongoing research projects, no commercial re- compression with a higher working pressure difference is versible PEM-based products are available for use at scale more viable than the installation of more steel tubes with today. a lower working pressure difference can be determined. If at the same time the maximum total pressure decreases, the pipeline wall thickness can be reduced, which might 2.2 Reversible AEM-based systems reduce CAPEX furthermore. As fuel cells operate at am- bient pressure, industry standard pressure regulators are The AEM technology uses a solid electrolyte. It combines utilized in all cases, as this is already done nowadays when the advantages of the PEM and alkaline cells by utilizing compressed hydrogen is shipped in bottles at 200 bar. membranes that avoid precious-metal loading. In the last decade, Giner received several grants from the US DOE to perform research into AEM systems. Results published 1.5 Remote communities recently show a good performance within the first 2000 operating hours . Long-term degradation and lifetime Currently, Horizon Power is replacing its aging diesel- expectation still need to be assessed to demonstrate that wind power-supply system in the remote community of this technology can be applied reliably at scale. Denham, Western Australia , with a solar-hydrogen- based utility. Hybrid Systems Australia was awarded the contract to build a solar farm with a rated capacity of 2.3 A high-potential candidate—solid-oxide 704 kW . During the day, excess electricity is stored cells (SOCs) as hydrogen that is produced on a 348-kW electrolyser. During times of low wind and no sun, a 100-kW fuel A SOC can operate in both fuel-cell and electrolyser modes. cell supplies electricity from the compressed hydrogen There are three main types of cell: anode-supported cells storage. Annually, 13 000 kg of hydrogen will be produced. (ASCs), metal-supported cells and electrolyte-supported This corresponds to a hydrogen production of 1185 kWh / cells. Each has distinct specifications for power density, H2 day = 35.6 kg /day, and results in a capacity factor for the lifetime and operating temperature. Since an oxygen-ion- H2 electrolyser of only 21.8%. conducting electrolyte is utilized, alternative fuels like nat- Besides the volatility of wind and solar power, the elec- ural gas, methanol and ammonia can also be used directly trolyser does not operate when the fuel cell is in use and in the fuel-cell (SOFC) mode. Table 1. Effect of hydrogen storage pressure on CAPEX cost and storage capacity Working pressure difference bar 92 52 22 8 Specific CAPEX $/kg 516 913 2158 5934 H2 Relative storage capacity % 100 56.52 23.91 8.70 Downloaded from https://academic.oup.com/ce/article/5/3/441/6333454 by DeepDyve user on 03 August 2021 Borm and Harrison | 445 Fig. 2: Rendering of a potential Mars station where URFCs have been considered Source: shutterstock.com. In the early 2010s, Versa Power Systems, now a subsid- 3 Limitations of reversible systems iary of FuelCell Energy, conducted research into reversible The spread in the operating voltage is challenging for solid-oxide fuel cells (RSOFCs). Their focus was on degrad- power electronics in reversible systems. The operating ation mechanisms, with ultra-high current densities of voltage in the electrolyser mode is twice as high as in the ASCs in both modes with hydrogen as fuel . Recently, fuel-cell mode, which means that the area-specific cur - FuelCell Energy received another 3-million-dollar research rent densities are doubled, while internal resistance re- grant to further develop RSOFCs. mains roughly the same. This results in a four to six times Sunfire demonstrated their reversible solid-oxide cell higher power input (AC) of the electrolyser than the power (RSOC) technology within the EU-funded GrInHy pro- output (AC) in the fuel-cell mode. The implication is that ject from 2016 to 2019. Their RSOC had several operation the ratio of the electrolyser input power to the fuel-cell modes : output power cannot be chosen freely—they are linked by this fixed ratio. input with 40-Nm /h hydrogen • electrolyser: 150-kW (AC) Putting some numbers around this point can clarify the output; case. A 10-kW electrolyser stack would only operate as • fuel cell: hydrogen fuel with 30-kW output; (DC) (AC) a 2.5-kW fuel-cell stack. When coupled to a photovol- • fuel cell: natural-gas fuel with 25 kW output. (DC) (AC) taic (PV) array with an assumed capacity factor of 20%, There are several challenges for integrated reversible PEM- 28.8 kWh of hydrogen can be produced per day (10 kW × 24 and AEM-based systems, which the RSOC does not suffer hours × 0.2 × 0.6 = 28.8). In this calculation, 60% efficiency from. The reversible AEM or PEM system must handle hu- has been assumed, meaning that 48 kWh of electricity (DC) input has been used to feed the electrolyser. An additional midified gas streams in the fuel-cell mode, but liquid water 5 kWh are necessary for the supply of the BoP (Balance will be in contact with the membrane in the electrolysis of Plant), increasing the actual power input to 11 kW mode. Each mode must therefore consider two-phase (AC) . The BoP includes pumps, valves, a control system and flow optimization. The pressure differential in the fuel- heaters—all of which consume power in addition to the cell mode is low, at 0.3 bar, but very high in electrolyser electrolyser stack itself. mode, at ≤30 bar . This places high sealing loads on the When operating as a fuel cell, the same unit could run equipment. for 6.9 hours at 2.5-kW output (28.8 kWh × 0.6/2.5 kW ). On the other hand, RSOC systems operate at am- (DC) (DC) This calculation assumes that the fuel cell is operating at bient pressure. Therefore, hydrogen generally needs to full load and 60% efficiency. Over that time duration, the be compressed after the electrolyser. A hydrogen com- fuel cell would produce 17.25 kWh of electricity from pressor may require maintenance every 4000–8000 (DC) the stored hydrogen. Due to the BoP demand and other hours. Compressor operation is especially challenging if energy-conversion losses, only 15.2 kWh are usable the power is to be derived from variable renewable en- (AC) and can be fed into the grid, reducing the effective output ergy because the compressor prefers a stable operation. power to 2.2 kW . This power output is five times smaller This constraint makes the RSOC systems less suitable (AC) than the electrolyser power input of 11 kW . The overall for off-grid applications. (AC) Downloaded from https://academic.oup.com/ce/article/5/3/441/6333454 by DeepDyve user on 03 August 2021 446 | Clean Energy, 2021, Vol. 5, No. 3 efficiency results in 31.6% due an output of 15.2 kWh systems, the two operating functions can be independ- (AC) in relation to 48 kWh of input. If the electrolyser would ently optimized to maximize the overall system perform- (DC) be fed with AC electricity from the grid, additional losses ance according to the local requirements. from the rectifier would need to be taken into account. The combined capacity factor is 48.75%, whereas the combined utilization factor is 74.5 %. This assumes an average daily Conflict of Interest operation of 11 hours in electrolyser mode—mainly in part None declared load—plus the 6.9 hours of full-load fuel-cell operation. Depending on the geographic location, the capacity References factor of the PV panels would vary and that in turn influ-  Natural Resources Canada. Glencore RAGLAN Mine Renewable ences the required fuel-cell operation time. If the above Electricity Smart-Grid Pilot Demonstration. https://www.nrcan. parameters fit the location and application, then the re- gc.ca/science-and-data/funding-partnerships/funding- versible unit may be suitable. If the above operating profile opportunities/current-investments/glencore-raglan-mine- does not fit the local requirements, then separate electro- renewable-electricity-smart-grid-pilot-demonstration/16662 lyser and fuel-cell components may be preferred. In that (16 April 2021, date last accessed).  Enapter. Hydrogen Seasonal Storage—Electrolyser Use Cases. case, each of those units can be freely adapted to be op- https://www.enapter.com/use-cases/hydrogen-seasonal- timized to the requirements of the local climate and the storage (16 April 2021, date last accessed). power-demand pattern.  LAVO. LAVO Hydrogen Battery. https://lavo.com.au/lavo- PEM fuel cells, for mobility applications, are optimized hydrogen-battery/ (16 April 2021, date last accessed). for a fast start-up and a high power output. Their expected  Ahluwalia, RK, Papadias DD, Peng J-K, et al. System level ana- lifetime is ~10 000 hours. On the other hand, an electro- lysis of hydrogen storage options. In: U.S. DOE Hydrogen and lyser stack is expected to have a lifetime of ≥60 000 hours Fuel Cells Program, Annual Merit Review and Peer Evaluation Meeting, Arlington, VA, 29 April–1 May 2019. to be able to be economically viable.  Carroll D. Horizon Looks to Future With Green Hydrogen Microgrid. A PEM- or AEM-based reversible system could poten- PV Magazine, November 27, 2020. https://www.pv-magazine. tially be used for an off-grid energy-storage application. com/2020/11/27/horizon-looks-to-futur e-with-gr een- The benefit would be that when hydrogen storage is in- hydrogen-microgrid/ (16 April 2021, date last accessed). corporated, the system could have a higher energy-storage  Horizon Power. Hybrid Systems Awarded Australian-First capacity than currently available battery technologies. In Renewable Hydrogen Microgrid Project. December 17, 2020. https:// this application, due to the maximum stack lifetime, sev- www.horizonpower.com.au/our-community/news-events/ news/hybrid-systems-awarded-australian-first-renewable- eral replacements would be necessary over the course of hydrogen-microgrid-project/ (16 April 2021, date last 20 years (175 200 hours) of operation, which is the typical accessed). lifetime expectation for an infrastructure project.  Bents DJ, Scullin VJ, Chang B-J, et al. Hydrogen-Oxygen PEM When considering the cost of multiple stack replace- Regenerative Fuel Cell Energy Storage System. Technical Report ments, the economic benefit of an URFC compared to a NASA/TM—2005–213381. Hanover, MD: NASA Center for similar system based on separate devices (an electrolyser Aerospace Information, 2005. and a fuel cell) become questionable. The economics of  Porter S. UNIGEN® Regenerative Fuel Cell for Uninterruptible Power Supply. FY 2003 Progress Report, Hydrogen, Fuel Cells, and integrating the electrolyser and a fuel cell are also com- Infrastructure Technologies. Wallingford, CT: Proton Energy promised because of the different BoP components, such Systems, 2004. as power electronics or gas-cleaning units, are necessary  Xu H. High-efficiency reversible alkaline membrane fuel for each operating mode. cells, In: 2020 DOE Hydrogen and Fuel Cells Annual Merit Review, 2020-05-21. https://www.hydrogen.energy.gov/pdfs/review20/ fc315_xu_2020_o.pdf. (25 June 2021, date last accessed). 4 Summary  Tang E, Wood T, Benhaddad S, et al. Advanced Materials for For many applications, separate electrolyser and fuel-cell RSOFC Dual Operation with Low Degradation. Report on Contract Number DE-EE0000464. Littleton, CO: Versa Power Systems, units are more likely to fit the application rather than Inc., 2012. integrating both functions into one unit because, in the  Schwarze K, Posdziech O, Kroop S, et al. Green industrial single unit, there is a fixed ratio of the electrolyser and the hydrogen via reversible high-temperature electrolysis. ECS fuel-cell capacities. Transactions, 2017, 78:2943. A single unit might save some CAPEX on day one, but  Borm O. Steam electrolysis as the core technology for sector long-term system performance is generally much more coupling in the energy transition. In: International Conference important than initial CAPEX and, when using separate on Electrolysis, Copenhagen, Denmark, 12–15 June 2017.
Clean Energy – Oxford University Press
Published: Sep 1, 2021
Access the full text.
Sign up today, get DeepDyve free for 14 days.