This page is supplementary material to the post “Electricity in Texas part II – the cost of a 100% renewable grid” on the Climate etc. web site.
Hydroelectric or pumped hydro generation may satisfy a small fraction of the tier 2 storage requirement, but Texas has very little of either.
The tier 2 renewable (synthetic) methane storage process in the grid hourly simulation was chosen primarily because all the components have already been used at industrial scale, costs are known and the end-to-end process meets the requirements for tier 2 storage. It is thus a viable baseline solution.
In the tier 2 storage process, spare renewable generation drives water electrolysers which produce hydrogen and oxygen at separate electrodes. The oxygen is discarded into the atmosphere. The hydrogen is combined with carbon dioxide from a source such as cement production, to produce methane (same chemical formula as natural gas) and water. This is called the Sabatier process. The methane is stored in salt caverns, partially-depleted gas fields or in surface storage tanks. One large-scale implementation of renewable methane via electrolysis was the 6 MW ETOGAS project producing low-carbon methane fuel (20gm/km) for Audi “A3-tron g” cars.
For the 100% renewable solution proposed the grid hourly simulation shows that renewable generation and tier 1 battery storage leave gaps in supply of 6%. At these times the stored renewable (synthetic) methane drives gas turbine generators which fill the gap in supply.
Renewable methane storage is very cheap, but the process proposed is inefficient, with a round trip power to gas to power efficiency of only 34%. That figure includes the backup gas turbine generation. Fortunately only 6% of supply goes through this process so the low efficiency is not a show stopper when costing the overall 100% renewable solution.
The renewable over-generation to drive electrolysis is included in the “Renewable Generation” (top) section of the cost model spreadsheet.
One alternative process is to store and use hydrogen directly rather than converting it to methane and to use hydrogen fuel cells instead of gas turbine generation as back up. The USA already has 1,600 miles of hydrogen pipelines and three salt caverns are used to store hydrogen. Another potentially much cheaper solution is mixed gas electrolysis which can be up to 70% efficient (similar to pumped hydro) and therefore requires many fewer electrolysers. Both are described more fully in the part 1 article.
Tier 2 electrolyser 2030-2040 high LCOE – $1,200 / kW, $12.1 / MWh
Because electrolysis uses only surplus electricity after batteries are full, 40 GW capacity of electrolysers are required to provide an average of 7 GW of electrolysis.
The FCH JU March 2015 document “Commercialisation of Energy Storage in Europe” page 51 figure 19, gives costs for PEM (proton exchange membrane – currently the most efficient technology) electrolysers. The current costs are $2.1-2.6 / W and 2030 costs are given as $0.3-1.45 / W.
Germany is heavily promoting renewable gas as a cheap long-term storage solution for renewable grids. At a recent presentation at Imperial College London, in a verbal answer to a question from me, Dr. Sigmar Bräuninger, Research Manager in the BASF Electrochemical Processes Group said the German industry consensus was for a likely future cost of €0.7 / W ($0.8 / W) for PEM electrolysers.
On balance, $1.2 / W has been used as the conservative high estimate. This equates to $12.1 / MWh of supply. The renewables generation to drive electrolysis is included in the separate renewable over-generation item in the cost model spreadsheet.
At present the grid hourly simulation assumes that electrolysers will be in use only when the tier 1 battery storage is full. Thus the electrolyser load factor is only 18%. The capital cost element of the electrolyser LCOE contribution might be lowered if electrolysis had a higher priority than filling battery storage. This might reduce the overall LCOE across all generation a little, but requires a change to the grid hourly simulation which will have to wait for now. A more even use of electrolysers would lower the renewable gas LCOE contribution by increasing the electrolyser load factor and thus reducing the required capacity .
Clearly there is plenty of scope for technology enhancements to reduce the cost estimates. Mixed renewable gas electrolysis (of water and carbon dioxide) is one such possibility. The higher round-trip efficiency obtained (70% instead of 34% for the more mature renewable methane solution as described above) enables the electrolyser capacity to be reduced significantly.
Tier 2 electrolyser 2030-2040 low LCOE – $400 / kW, $4.0 / MWh
The low 2030 figure in the FCH JU document above is $316 / KW and the slightly higher cost of $400 / kW has been assumed which translates to a $4.0 / MWh low LCOE. Electricity to drive them is included in the renewables over-generation item in the cost model spreadsheet.
Sabatier process, negligible LCOE
Conversion from hydrogen and carbon dioxide to methane and water takes place in a large reaction vessel with a catalyst. It generates heat which must be disposed of but is otherwise low tech and expected to be cheap compared to electrolysis.
Tier 2 storage cavern $250 to $300 per MWH capital cost, LCOE $1.0, $1.2 / MWH
The cost of renewable gas storage in salt caverns or partially-depleted gas fields is $250 to $300 per MWh of storage capacity. This is based on figure 6 of the FCH JU document above. The raw figure is $200 per MWh of hydrogen storage capacity. However, methane is assumed here, and the low figure is increased to $250 using the 43% to 34% round trip efficiency ratio of hydrogen to methane storage.
$300 per MWh is assumed as a high figure. Neither is critical as the contribution to overall electricity LCOE is small ($1.0 to $1.2 / MWh).