Electrolysis systems — which use electricity to split water molecules into hydrogen and oxygen — have not yet proven that they can work effectively with variable renewable energy, according to new analysis from research house BloombergNEF (BNEF).
To produce green hydrogen, electrolysers need to be supplied with renewable energy and in for most projects, this means wind or solar power, or both — energy sources that are highly variable, being dependent on the weather and the time of day.
Electrolysers must therefore be highly flexible — able to produce hydrogen when the electricity input is high or low, and be able to ramp up or down accordingly —and the more hours an electrolyser is in operation, the lower the levelised cost of green hydrogen over the course of a project’s lifetime.
Thus, for green hydrogen to be affordable, “electrolysis systems need to be compatible with intermittent renewables — a feature the century-old technology has yet to fully demonstrate”, writes BNEF’s San Francisco-based hydrogen analyst, Xiaoting Wang, in a note to subscribers entitled Cheap Green Hydrogen Requires Much Nimbler Electrolyzers.
“Historically, electrolyzers were designed to work with stable electricity from the grid,” she observes. “When switching to green hydrogen production, the electricity must be sourced from renewables, either virtually in a grid-connected system, or physically in an off-grid system. Even for grid-connected solutions, regulations such as those in the EU and US may mandate time-matching between power generation and power consumption.
“Electrolysis systems must thus have considerable flexibility — meaning they can start quickly, ramp up/down quickly, and function within a wide working range — to make the most use of fluctuating renewable power to produce hydrogen. That’s a feature for which test protocols have yet to emerge.”
The comments take account of the fact that an electrolyser’s working range is of fundamental importance to green hydrogen developers.
Essentially, the working range means the amount of electricity input required for the system to safely produce hydrogen, measured as a percentage of an electrolyser’s nameplate capacity.
Assuming electrolysis efficiency does not vary as input power changes, if a 10MW electrolyser can only produce hydrogen safely when the electricity input is equal or greater than 1MW it would have a minimum working load of 10%, and therefore an operational range of 10-100%. In short, it can only safely operate when being supplied with 10-100% of its nameplate capacity.
“Although most manufacturers promise minimum working loads in the 10-30% range, commissioned green hydrogen projects suggest the actual products could underperform,” Wang writes.
And as she points out, there are no protocols in place to test minimum working loads, meaning that manufacturers’ stated working ranges cannot be objectively proven.
Wang has previously highlighted the concern that, in an electrolyser stack, oxygen is produced at the anode and hydrogen is made at the cathode, with a membrane sitting between them to prevent the two gases mixing. When hydrogen and oxygen mix, they can form an explosive mixture.
However, because hydrogen molecules are so small, some can still cross the membrane — but this is not a problem if the concentration of hydrogen in water is below 4%.
However, when the amount of electricity entering the electrolyser falls below a certain level, the proportion of hydrogen to oxygen increases, meaning that the concentrations of H2 in the O2 rises to dangerous levels, and the electrolyser must be switched off for safety reasons.
As previous reported, all the electrolysers supplied to the world’s largest operational green H2 project — Sinopec’s 260MW Kuqa facility in northwest China — turned out to have more limited working ranges than expected by the manufacturers.
Cockerill Jingli, Longi and Peric had all promised minimum working loads of 30%, but the actual figures proved to be higher than 50%.
“For off-grid electrolyzers with a narrower working range, renewable electricity would be wasted,” Wang explains in her analyst note. “A developer can introduce batteries to avoid or reduce power wasting, but in either scenario, green hydrogen production becomes more expensive.”
When using BNEF’s in-house Hydrogen Electrolyzer Optimization Model, an electrolysis system in California that can only operate with a 60-100% working range would produce a levelized cost of hydrogen that is 15% more expensive than a same-sized system with a 10-100% working range, with their power solutions separately optimised.
If the 60-100% system was designed to have the same power solution as the 10-100% case, assuming the design is based on overoptimistic assumptions, the premium would be even more.
“The cost increase would be more significant for locations where renewables have a less stable [renewables] output than California,” Wang adds.
A chart accompanying the analyst note shows that the levelized cost of green hydrogen in an off-grid electrolysis system in California would be just over $4.6/kg when using Thyssenkrupp or Nel electrolysers that promise a 10% minimum working load. But a Siemens electrolyser with a stated range of 40-100% would result in a levelised cost of a little over $4.8/kg.
(A version of this article first appeared in Recharge's sister publication Hydrogen Insight)
