In industry, lower cost is more important than slightly better efficiency (part 1)
In industry, lower cost is more important than slightly better efficiency (part 1)
Lower-CapEx yet electrolyzers will deliver lower cost hydrogen – even if they have worse efficiency (but not if they are unreliable)
In industry and commercial sales, higher efficiency factories that cost more CapEx will lose out to lower efficiency low-CapEx factories.
Combining this with real-world grid fees, less efficient lower CapEx electrolyzers will always produce hydrogen for lower cost than a more expensive and slightly more efficient electrolyzer. Obviously the scale of CapEx difference vs efficiency is important here, but until we step-change gains in efficiency instead of incremental gains, we are in the paradigm of “lower CapEx = lower cost, efficiency is less important.”
Caveat: Decreasing reliability to achieve lower CapEx is a bad tradeoff that will result in failure to deliver H2 and significant losses from lost sales and lower equipment utilization. Reliability is mandatory for hydrogen project success.
The implication for OEMs and equipment investors is that electrolyzer developers should focus on total installed cost savings over the core electrolyzer component costs and over efficiency.
The implication for developers and project investors is to focus on projects with opportunities for grid service (more on that below) to achieve lower power price and more likely 45V adherence.
Higher CapEx of the factory for better efficiency output is rolled into the cost of the output – eroding and often overriding any efficiency gains
Nearly everyone assumes that more efficiency always leads to lower costs. When I try to explain that more efficient rarely leads to lower cost, it typically takes a few hours of meetings and conversations to get to the “ah-ha” moment where everyone in a room understands that efficiency gains only lead to lower costs if the efficiency gains can be made without having higher cost materials or equipment/factory costs.
Why? Better efficiency usually comes at a higher up-front cost. That higher initial CapEx is expected to be paid back over time. Without sufficient savings from higher efficiency, the higher up-front cost can’t be paid back – and the resulting output is higher cost despite using less resources to make that output. Paying back the initial factory costs is expensive, and higher factory costs for more efficient operations won’t always be paid back – resulting in higher cost products. Many new technologies can squeeze more efficiency, but not enough to offset higher upfront cost.
For electrolyzers, efficiency means we’re talking about less energy used per kg of hydrogen produced.
Last month I recommended to a potential client that they look at ITM Power’s skid-mounted electrolyzers -coming in at an estimated installed cost of $1.5M per MW of electrolyzer compared to $2.5M for any other Western-made electrolyzer. I was surprised to hear the client viewed the slightly lower efficiency of the ITM system as a dealbreaker – the 50% lower CapEx wasn’t enticing because it was a relative 5% less efficient than their preferred electrolyzer. So I put together the following comparison:
The chart above shows that the lowest price of hydrogen is achievable with a low-CapEx electrolyzer that is slightly less efficient compared to a high CapEx system with better efficiency. Always on grid connected (column 1) is rarely optimal owing to high power costs, Onsite renewables with optimized grid backup (column 2) is lowest cost and 10% less expensive with a less efficient but cost effective electrolyzer, and curtailed power (column 3) is unreasonable with current electrolyzer costs.
Realistic power costs are key here
Realistic power costs are key here. A grid-connected electrolyzer that operates 24/7 will have to pay high grid fees – the $0.08/kwh I show is extremely low for current industrial prices in any state that would pay extra for electrolytic H2. The bars here should be higher in most projects owing to higher electrical costs. I see a lot of papers and comparisons than somehow don’t realize that connecting to the grid and having guaranteed power has a high fee associated with it.
The second scenario assumes an electrolyzer co-located with dedicated renewables, but connected to the grid to support the grid as well as to draw from it. By shutting down the electrolyzer during peak power demand, such as when people get home from work, the grid fees an electrolyzer will pay will lower dramatically. So our renewable plant may only operate 50% of the time, but the average power cost is much lower than our always-on grid connected system. As a result, the cost of hydrogen is much lower for 50% uptime with low CapEx electrolyzers.
Lower CapEx creates more savings here. The penalty of less operation is that the CapEx has to be spread out over fewer kg of H2 made and sold. If the CapEx is lower, then the cost stack contribution of CapEx from low capacity factor is lower. Hence lower CapEx systems are at a comparative advantage when they don’t have high capacity factors.
There is a limit here - no matter how low power prices get, low utilization electrolyzers will have high prices until electrolyzer costs come down further.
Grid services will be critical to achieve these cost structures
Most electrolyzer projects will do “grid services.” While this is a complicated host of activities, it boils down to “turn electrolyzer off when the grid is congested, ramp the electrolyzer in specific ways when the grid is on shoulders.” There are many words people throw around for this. Demand response. Virtual power plant. What it results in, however, is significantly lower average power costs. Looking at the chart above, the conclusion that every project developer needs to take away is “my local utility needs to be a partner in developing my project in order to achieve the lowest power price and thus lowest cost H2.”
Grid services requirements are limited – but will rise significantly as more renewables are on a grid
The market for grid services isn’t unlimited, and electrolyzers are in direct competition with batteries. We aren’t going to see 1GW of grid services for electrolyzers in most locations – because that is roughly the power needs of an entire megacity.
The grid services option is a project strategy – not an industry one. That being said, as more renewables enter the grid there will be more need for grid services and demand response. In some locations – especially in developing areas outside the US- grid response will provide immense opportunity for hydrogen production as the cost of electrolyzers comes down.
Storage costs make “Always on” and “50% uptime” closer in price
No buyer wants to take hydrogen only when the weather suits production. Extra hydrogen storage is necessary to assure hydrogen supply when the hydrogen is not being produced around the clock. Functionally, add $1.00/kg for storage and other equipment[1] at the 50% uptime price for extra hydrogen storage during downtime. In nearly all circumstances lower operating time will result in lower power cost, achieving lower cost of delivered hydrogen.
45V makes 50% uptime scenario significantly better
Projects are rushing to start construction in the next two years to hit 45V. The 90% uptime scenario will not be 45V compliant in any commercially viable scenario owing to price competition with data centers.
In other words, accounting for H2 storage and 45V, a low CapEx electrolyzer with lower uptime will have delivered hydrogen costs around $1-$3/kg lower than a 24/7 electrolyzer that pays full grid fees.
These prices are far above gray hydrogen prices in the US for the next decade – making electrolytic H2 commercially viable only where supported or required
Gray hydrogen in the US is $0.85/kg for an old steam methane reformer that has already been fully paid for. A new plant can make gray H2 for $1.50/kg or less. Until the total installed cost of electrolyzers is around $500/kw we will not see electrolytic H2 compete directly with gray H2 in the US – incentives are required. The higher cost of gas in other parts of the world can change this relationship – especially in places where the natural gas grid doesn’t exist or can’t expand.
We see this “cost beats efficiency” everywhere
Fertilizer production allows us to grow food. Fertilizer plants[2] production could use a ruthenium-based catalyst to convert natural gas to ammonia fertilizer. This would achieve better efficiency. Instead, all fertilizer plants use an iron based catalyst that is much less expensive. We’re talking $5M for the iron catalyst instead of $100M for the more efficient ruthenium catalyst.
Despite gas prices historically being volatile, more scarce, and much more expensive than today, there was no price regime in which the more efficient process resulted in the more efficient process resulting in less expensive fertilizer.
For nearly everything we buy, there exists a more efficient process that is more expensive. Cost wins, full stop.
Don’t forget reliability
ITM has produced and fielded electrolyzers. No other western low-cost electrolyzer has. Use untested companies at your peril.
[1] Other equipment includes having a compressor that is only at 50% capacity factor – and thus doubles the CapEx contribution to the cost stack
[2] Fertilizer plants is a simplification to mean “haber bosch plants to produce ammonia which is then used in fertilizer”
