Scientists at the Massachusetts Institute of Technology (MIT) have conducted techno-economic analysis to assess the potential competition between hydrogen-fired gas plants and large scale lithium-ion storage in the role of replacing gas-fired thermal power plants in the United States, and have found that lithium-ion batteries may be the most efficient solution in terms of costs, although hydrogen may represent a viable alternative in certain cases.
According to their findings, which can be found in the paper Techno-economic analysis of balancing California’s power system on a seasonal basis: Hydrogen vs. lithium-ion batteries, published in Applied Energy, hydrogen-fired gas turbines (HFGTs) can currently be built at a cost of $1,320/kW and may replace natural gas-fired plants to balance the electricity network on a seasonal basis. Maintenance cost for HFGTs is assumed to be the same as that of gas-fired gas plants, at around $13/kW, per year, while transport costs for hydrogen are estimated at between $0.60/kg, for a 1,000km pipeline, to $2/kg for a 3,000km pipeline.
Furthermore, the MIT scientists calculated that green hydrogen can currently be produced at a cost ranging between $3 and $10 per kilogram, depending on operating conditions, while grey or blue hydrogen can be generated at a cost of $1 and $1.50 per kilogram, respectively. “Assuming 40% utilization of a proton exchange membrane (PEM) electrolyzer and an input power cost of $10/MWh, green hydrogen can be produced at $5/kg,” they specified. “Green hydrogen is currently more expensive than grey or blue hydrogen but these costs are projected to decrease on the order to 70% through 2050 as demand increases.”
As for lithium-ion storage, the academics estimated that a 100 MW/400 MWh battery may be built at a cost of $236.50/MWh.
The U.S. group compared the levelized cost of energy (LCOE) of HFGTs and large scale lithium-ion storage in a scenario in which the two sources of energy have to meet the production achieved by natural-gas-fired gas turbines in California in 2019. It also assumed that the lifetime of any installed project is 15 years and the discount rate for any installed project is 10%. “The annual fuel consumption for the HFGT is calculated as the total cost associated with purchasing hydrogen to operate the facility, while the annual fuel consumption for the lithium-ion [facility] is equal to the total cost of the electric power stored in the system,” it further explained.
The analysis showed that the competitiveness of HFGTs is strictly dependent on the heat rate of the gas power plants they should replace. The heat rate indicates a power plant's efficiency and represents the amount of energy it needs to generate a kilowatt-hour of electricity. According to the researchers, HFGTs become more competitive than lithium-ion storage when the heat rate is higher. Furthermore, they found that HFGTs powered by blue hydrogen produced with CO2 capture may be the more cost-competitive substitute for natural-gas-fired gas turbines. “We found the power prices in today’s market do not justify investment in this technology,” they concluded. “However, we noted the propensity for more extreme power price patterns within the market as the share of variable renewable energy grows within the market.“
This content is protected by copyright and may not be reused. If you want to cooperate with us and would like to reuse some of our content, please contact: editors@pv-magazine.com.
If the current heat waves and other weather-related disasters are not enough to introduce a note of desperation into this narrative, sufficient to make us realize that sustainability matters more than efficiency, then we have to accept all of the consequences, for ourselves and our heirs.
Hydrogen vs Li-Ion is a question, but it also misses the point that we must have “low cost long-term energy storage”. Also, only part of that storage must be “electric”, and there are numerous additional “electric storage” technologies including gravity (pumped hydro, stacked blocks) and “flow batteries”. The rest of the “energy storage” needed is “thermal” for buildings and processes, and it makes almost zero sense to put in expensive electric storage for that — a water tank, especially a buried water tank gaining partial geothermal advantage, is all that is needed!
Absolutely. Thermal storage will often be a much better option when it comes to diurnal or even weekly storage. Residential water heaters (resistance type) use about 3500 kWh a year. This is comparable to the annual electricity demands of an EV. How hard would it be to program water heaters to charge themselves when electricity is cheap and abundant? The energy conversion and standby energy losses of thermal storage are on the order of 10% of the the losses of electrochemical storage. Depending on scale the costs of thermal storage capacity can be over 100 times lower than electro-chemical storage.
An additional factor supporting thermal storage has to do with the wholesale price profile we can expect in grids with ultra-high penetrations of wind and solar – i.e. 95% of electricity supplied by renewables. We can expect daily, weekly and seasonal patterns but I want to focus on the weekly pattern because I think it could end up being the most consistent. You never know what you’re going to get day to day but there’s a lot more consistency when you looks at weekly data.
We use sophisticated neural network tools to predict load but the basic aim of these tools is to match up historical temperature and wind speed data against historical load data. If load was 10 GW on a Friday at 1 PM when the temperature was 75 F and windspeed was X you can expect the load to be similar on a Tuesday at 1 PM with similar weather conditions.
A key part of these models has to do with putting weekdays in one pile and weekends/holidays in another. You don’t want to mix up these types of days up because the load patterns are much different. Weekends typically show 7 to 10% lower loads compared to weekdays. The lower loads on weekends result in wholesale electricity prices which are around 15% lower due to the drop in demand.
In an ultra RE grid we’re likely to have wholesale electricity prices which are dramatically lower than what we see currently. I haven’t modeled this but I’m guessing 30% lower prices on weekdays could become a thing. I’ve spoken to several energy experts about this idea and, while speculative, we could end up seeing a much larger price differential than this. What’s the best way to capture the lower electricity prices on the weekends and spread them out over the week? I think thermal storage is the winner hands down and then some.
This is highly speculative but I’ve always had a soft spot for thermochemical cycles. There are thousands of different thermochemical cycles. Some of them produce hydrogen. Others extract CO2 from the atmosphere. Still others make ammonia.
When it comes to producing hydrogen you want to operate your re-finery as much of the time as possible. I know a couple of people in the hydrogen space. These 40% estimates in this MIT paper for PEM utilization are well off the mark. The capacity factor should be able to reach up much closer to 80 or 90%. That said, it seems to me a thermochemical cycle could also achieve an 80 to 90% capacity factor but if you build these plants such that they had a thermal reservoir which could preferentially charge up on the weekends you’d be able to dramatically reduce your energy costs relative to a PEM production method.
I’m not saying this is an economically efficient idea. I just think it’s kinda cool. There are thermochemical cycles where 90% of the energy demand is thermal and only 10% of the demands are for pumps, compressors and such. I’m thinking… Move the thermal energy charging to the weekends and run your plant as much as possible. PEM hydrogen production doesn’t have this flexibility. I’m not saying PEM is going to lose or that thermochemical hydrogen is going to win. As I said… I just think this idea is kinda cool.
$236.50Mwh? At present they are hoping to get the price at best per a kw down to $40 while at present they are around $100+ per a kW.. at 1,000 kW in a mw that is still $40,000 per a MWh .. I’m assuming a typo? Because at $236.50 that wouldn’t even cover the resources for said battery before any refining let alone building it.
There are some serious problems with the MIT calculations. Taking a price of $10/MWh looks like direct connection to PV – which is fine. The 40% capacity factor is likewise consistent with what you could expect from an over-sized PV plant connected to a PEM set-up.
A PEM electrolyser will use around 55kWh to produce 1kg of H2. This puts the non-CAPEX/energy-only price of H2 from a PEM at around $0.55/kg. Consensus amongst electrolyser manufacturers is that CAPEX adds 20% to the energy only price. So a CAPEX + energy price for H2 from PEMs looks like $0.7/kg. This is a very very long way from what MIT is claiming. I suspect that they have assumed a $100/MWH energy price – which would fit with what they are saying.
However, there is a problem with $100/MWh – because utility-scale PV in sunny places such as Calif (or Spain) have LCOEs circa $10/MWh. & projects are being developed in Spain with prices that reflect this reality.
If the above is not the case, then it is difficult to see how MIT get from an energy-only price of $0.55/kg to $5/kg.