The Australian report Comparison of dispatchable renewable electricity options does the very useful service of quantifying the energy storage landscape in dollars and cents. It reaches many interesting conclusions, not the least of which is that hydrogen, and by explicit extension, ammonia, is the key option for long-cycle storage. And while the study’s focus is Australia, “with costs in AUD and based on Australian conditions,” its lead author says that “much of the information and many of its findings are expected to hold independent of jurisdiction.”
The comprehensive, 188-page document was commissioned by the Australian Renewable Energy Agency (ARENA) and produced by the consulting firm ITP Energised Group, the Institute for Sustainable Futures at the University of Technology Sydney, and the consulting firm ITK Consulting. It was released on October 30, 2018.
(Note: To foster readability, all cost figures in this post will be given in the original Australian dollars (AUD$) only. Multiply AUD$ figures by 0.72 to derive USD$ at today’s exchange rate.)
Energy Storage Essentials
One of the report’s strengths is its presentation of the evolutionary processes currently at work within our electricity system. (Readers who are already well versed in the basics of the transition can skip to the next section.) A foundational driver of change is the displacement of fossil fuels by renewables. An unyielding grid imperative is that electricity demand must be matched at every second with supply, no matter how quickly or substantially demand goes up or down. Demand variability is characterized by patterns on three distinct time scales: second-by-second as electricity consumers turn electrical devices on and off; hour-by-hour as the level of activity ramps up and down over the course of a day; and day-by-day as seasonal changes call for more or less heat and light in the built environment. Fossil generation, with its reliance on large-scale plants, is not especially adept at responding to variations in demand, but over time utilities and grid management organizations have developed methods for meeting them.
Wind and solar photovoltaics, the two most important workhorses in the renewable generation portfolio, represent a step in the wrong direction in responding to demand variability. They are able to generate only as much electricity as sun and/or wind conditions allow at a given moment in time. There’s no dial to generate more electricity and there is a strong economic disincentive to generate less via output curtailment. More precisely, they are neither dispatchable nor firm. (Per the report: “dispatchable generators are those that can raise or lower power output on command from the system operator or facility owner;” while “firm” means a “constant level of power output that a generator can legally or commercially guarantee for a specified time interval.”)
The lack of dispatchability and firmness certainly constitute significant drawbacks to wind and solar PV, but both technologies have arrived at a point of economic advantage not just within the club of sustainable options, but relative even to legacy generation technologies. As the report says, “generation from wind or photovoltaic (PV) systems [is] now the cheapest electricity per MWh for new build systems.” These circumstances provide plenty of motivation to develop energy storage solutions that can complement wind and solar PV. Fortunately there is a robust range of options in this area. All that remains before implementation is to determine which ones are the most affordable.
The parameter of choice for examining the affordability question is levelized cost of electricity (LCOE). LCOE combines capital and operating costs in a single measure expressed in dollars per MWh. (Using capital or operating costs separately to compare the economics of systems is not helpful given the stark differences in cost per MWh for elements such as plant construction and purchased fuel across radically dissimilar generating technologies.)
An important point made by the report is that LCOE varies not just from technology to technology, but within a given technology according to the length of the energy storage cycle. Utility-scale PV made dispatchable via the integration of battery storage, for example, has an LCOE of AUD$120/MWh if the firming commitment is for one hour, but an LCOE of AUD$435/MWh if the commitment is for 24 hours. This is driven by the fact that the full-day case involves a great quantity of expensive batteries.
In addition to wind and solar PV, the report considers generation from concentrated solar thermal, biomass, and geothermal. Storage technologies considered include batteries, pumped hydro, molten salt (in associated with concentrated solar thermal) and hydrogen. These technologies were selected because they are “considered commercially advanced.” The report acknowledges the possibility that “new technologies can and will play a role over the coming years.” Specifically mentioned in this regard are compressed air, cryogenic storage, flywheels, supercapacitors and superconducting magnetic storage, and biomass gasifiers.
The methodology includes discrete cost calculations for initial collection of energy; conversion of energy to storable form; storage itself; and conversion back to electricity. Cost information was collected from “real projects” via “published sources and stakeholder consultations.” The lead author’s email announcement of the study’s release provides further description: “The study has developed an innovative subsystem based approach to modelling installed cost as a function of configuration and system size for each technology combination. This has been used to examine the LCOE of optimised systems under real input conditions as a function of duration of storage or operation.”
One of the report’s high-level findings is that “for each timescale different technologies are seen to offer the lowest cost energy.” (As a point of comparison for the figures below, a 2015 report sponsored by ARENA put the LCOE of dispatchable fossil generation technologies in Australia in a range from AUD$60/MWh to AUD$140/MWh, while variable renewable wind and solar PV have LCOEs of AUD$60/MWh and AUD$70/MWh respectively.)
At AUD$106/MWh, the system that has the lowest LCOE to meet a one-hour firming commitment of renewable electricity is wind generation with battery-based energy storage. To meet a six-hour commitment, the two lowest LCOEs are provided by biomass processed with anaerobic digestion, at AUD$107/MWh; and “grid variable renewable electricity” (i.e., electricity generated at various locations from a mix of renewable sources and delivered over the electric grid to the point of storage) paired with pumped hydro, at AUD$113/MWh. At AUD$83/MWh, the biomass-AD option has the lowest LCOE to meet a 12-hour commitment. And geothermal has the lowest LCOE to meet a 24-hour commitment, at AUD$60/MWh. For time scales between 24 and 40 hours, the LCOE prize goes to systems that integrate either wind or PV with pumped hydro, and concentrating solar thermal systems, all in the range of AUD$120/MWh to AUD$140/MWh. (Baseload geothermal that runs continuously — i.e., beyond 24 consecutive hours — rather than as a dispatchable complement to other generating technologies, is excluded from the analysis. Regardless of other considerations, geothermal is a resource whose contribution is ultimately limited by the distribution of economically exploitable geological formations.)
The report classifies anything beyond 40 hours as “very long-term storage.” And past this point the LCOE winners are all based on the energy stored in chemical bonds: biomass and hydrogen integrated with solar PV or wind. This should come as no surprise since the legacy electric grid, with its hard-won dispatchability profile, is also based substantially on the energy contained in chemical bonds. The report shows that the LCOE for biomass ranges from AUD$55/MWh to AUD$125/MWh, depending primarily on the price of the feedstock and whether the energy collection is carried out by anaerobic digestion or combustion. The LCOE for hydrogen-based dispatchable electricity systems ranges from AUD$220/MWh to AUD$280/MWh, depending on the generating technology and the length of the storage cycle.
Wide deployment of the biomass option would essentially maintain the grid status quo – i.e., a combination of baseload plants and dispatchable intermediate and peaking plants — with the substitution of woody and agricultural materials for coal and natural gas. The reason biomass, with its favorable LCOE, does not take over the electricity sector is the finite nature of the resource. The report highlights the limitation for anaerobic digestion, saying it is “likely to contribute only a small amount to national electricity demand as the available waste material is limited in supply.” While “dry biomass combustion has the potential to make a much more significant contribution,” a variety of studies from around the world have established that this resource will fall far short of meeting our demand for electricity. (See, for example, the Billion-Ton Report from the U.S. Department of Energy.)
So that leaves hydrogen as the economically advantaged, resource-unconstrained solution for long-cycle energy storage.
The report includes just one reference to ammonia as a hydrogen carrier, but it is clear and emphatic: “An alternative to [hydrogen] compression is conversion to ammonia, which has a higher energy density by volume of 6.8 MJ/litre than that of liquid hydrogen (4.8 MJ/litre), and is under physical conditions that are much easier to achieve and maintain. This is an important consideration for transport.”
We know this reference to ammonia is serious because the report’s lead author is Keith Lovegrove, Managing Director of ITP Energised. Lovegrove has had an interest in ammonia energy since at least 2008 when he co-authored a presentation for the NH3 Fuel Conference entitled Ammonia Production and Baseload Solar Power. At the 2016 edition of the Conference, he delivered a presentation entitled Japan – a future market for Australian solar ammonia.
An ARENA webpage accompanying the report contains links to supporting materials, including two spreadsheet-based calculators that allow costs of dispatchable renewable energy systems at different scales and configurations to be estimated.