In February, Nine Mile Point nuclear power station in upstate New York started making its own hydrogen. Normally, this would be the point where I’d do the math and explain why this is a terrible idea, but in this case they got it right.
So, what are they doing and why?
The Nine Mile Point reactors operated by Constellation Energy are boiling-water reactors (BWR), not the more common pressurized-water reactors (PWR) in the US fleet. There’s a really old, smaller one that produced 644 MW, which was put online in 1969 and is scheduled for decommissioning in 2029. Then there’s a bigger one, 1,375 MW, which went online in 1988 and is scheduled for decommissioning in 2046. That will make them about 60 years old when they retire, which is the maximum you want to run a nuclear reactor. Interestingly, there’s actually a completely differently owned and operated 813 MW nuclear power plant on the same site that’s limping along with subsidies and the like after it was rescued from fiscal death a few years ago.
The Nine Mile Point reactors use hydrogen as a coolant in their process, about 560 kg a day. That might be true of all BWRs, or it might not. It might also be true for PWRs. I’m not close enough to the technology to know. Regardless, it’s true for the two reactors in question.
It was gray hydrogen made from natural gas undoubtedly, at about about 10 kg CO2e from upstream methane leakage and the steam reformation of natural gas process. It was being trucked in, undoubtedly in very large, pressurized, heavy tanks at probably $11 USD per kg.
That’s roughly $6,200 per day for delivered hydrogen, or about $2.3 million per year. The annual CO2e emissions were in the range of 2,000 metric tons, which is sufficient to be concerned about.
At some point Constellation and the Department of Energy got together and said, let’s do something about this. The DOE chipped in $5.8 million, Constellation put in another $8.7 to round out the $14.5 million cost, and they set up a relatively small electrolyzer on the site. It draws 1.25 MW or about 0.06% of the facility’s 2 GW daily electricity generation when both reactors are running. This is best understood as another auxiliary power draw on the facility, like running the pumps, the monitoring equipment, the electricity management equipment, and the like. It’s a very minor additional draw from produced electricity. All power plants have auxiliary power draw, although thermal power plants have much larger ones than wind or solar for obvious reasons. This is behind the meter usage of locally generated electricity, so is at the basic cost of production.
Operational and maintenance costs of US nuclear reactors, the cost of generation, are about $38 per MWh. The 1.25 MW electrolyzer draws about 30 MWh per day, so costs are in the range of $1,200 with water per day, or about $2.15 per kg ignoring capital cost depreciation. Wholesale electricity prices in New York State have been bouncing around $40 per MWh, so they are foregoing perhaps $60 in revenue per day, a rounding error.
Making their own hydrogen with their own electricity saves them about $5,000 per day or $1.8 million per year, and foregoes about $60 in revenue per day, or $22 thousand per year. That means it’s about an eight year return on the $14.5 million investment, which is nothing to write home about, but the DOE free and clear grant brings that down to 4.5 years.
That means after 2028, the hydrogen is just avoiding costs and increasing profitability very slightly. $1.8 million more in the kitty every year is good news, although the smaller reactor is being decommissioned the next year, so hydrogen demand will likely drop by a third per day. This low-carbon electricity produced hydrogen also avoids about 2,000 tons of CO2e per year, which is a good win as well.
This is a good nuclear hydrogen use case. It makes sense to make hydrogen at point of demand using low carbon electricity rather than trucking in high-carbon hydrogen. This is a good example of the hydrogen economy of the future, which is to say displacing gray and black hydrogen used in industrial processes with hydrogen electrolyzed at point of demand with appropriately scaled electrolyzers using firmed, low-carbon electricity.
Does it mean using nuclear energy to make hydrogen on site for shipping elsewhere makes sense? No, not at all. Gray hydrogen costs $0.70 to $1.60 per kg to manufacture, and yet delivered hydrogen by truck in the US is $11 per kg in bulk. Hydrogen is very expensive to store, compress, and distribute. It loves to leak, so every step of the way adds complexity and challenges. Because hydrogen is energy dense by mass but very energy diffuse by volume, it has to be compressed massively, or even liquified at -253° Celsius.
Compression for storage and distribution releases a lot of heat, requiring cooling at multiple stages, which costs money. Liquification uses about a third of the energy that’s embodied in the liquified hydrogen. Shipping it by sea, as is a touted replacement for LNG, would mean at least five times the cost per delivered unit of energy. Nonexistent pipelines for hydrogen would require three times the energy for compression as natural gas pipelines, adding billions to costs. And nuclear reactors aren’t in LNG ports that could be converted to hydrogen ports and don’t have pipelines running into the facility. Lots more costs, more steps in distribution.
Decompression for use has an inverse problem, in that when you decompress the gas it gets colder, a lot colder. You have to warm it up, and do that carefully, as hydrogen is a gas which loves to burn in the presence of oxygen. It will ignite in a much greater range of ratios to oxygen than natural gas, 4%-75% compared to 5%-15%, and ignites with a much lower spark temperature. Hydrogen nozzles freeze to fuel cell cars all the time as a result, and the suckers … er… innovators with fuel cell vehicles have lots of how-to guidance on the web about what to do and not to do. (Don’t: pour water on it, use a lighter on it, use an electric heater on it, jiggle it, force it, put your bare skin on the nozzle or metal of the car near the nozzle. Do: wait several minutes and try again.)
This is all technically viable and done every day, it’s just complex and costly, so it’s only done where it’s absolutely essential.
A bigger electrolysis plant at a nuclear facility would have a much larger number of components as a result of the above. The $2.15 per kg production costs would not necessarily go up, but a full scale industrial electrolysis plant is a big operation with much higher capital costs than the $14.5 million for the 1.25 MW electrolyzer that was fit for purpose for hydrogen for plant cooling. The payback period would be much longer.
And what industrial consumers of hydrogen are near enough to nuclear power plants to make it worth while? The biggest consumers of hydrogen today are fossil fuel refineries, about 50 million tons of it annually. They use it mostly for desulphurizing crude oil. Nuclear plants aren’t usually built next to oil refineries, and as peak oil demand arrives, likely in the second half of this decade, high-sulphur crude will be first off the market, as there’s a lot of low-sulphur crude available that’s cheaper to refine.
The next biggest demand center is fertilizer production, about 30 million tons annually. Once again, not typically next to nuclear plants. It’s diminishing volumes after that, and once again, not typically near nuclear plants which were built to transmit electricity dozens to hundreds of miles to demand centers over much cheaper electricity transmission distribution systems.
And that $11 per kg delivered, or perhaps $6-8 per kg in much bigger volumes, prices hydrogen out of the hydrogen for energy market. In my projection, only steel is a growth market for hydrogen demand, and that’s not as big as many projects as we’ll be scrapping things like the 3 million miles of fossil fuel pipelines in the US alone and feeding them into electric steel minimills for new steel a lot more in the future.
Building a new nuclear power plant in conjunction with a new industrial hydrogen facility multiplies the risks of the two technical build-outs, and is unlikely to pencil out when contingencies are applied. That’s not stopping the overlapping nuclear and hydrogen-for-energy types from advocating for the combination.
As a reminder, Oxford and IT school of Copenhagen professor, globally consulted megaproject expert and author of How Big Things Get Done Bent Flyvbjerg’s work on megaprojects finds normal-scale nuclear generation construction projects have lots of fat-tailed risks which means that they regress to the tail instead of the mean, going vastly over budget much more often than wind and solar. In his 16,000+ megaprojects database, nuclear power construction is 3rd from the worst, beaten only by Olympics and nuclear storage projects. Meanwhile, solar, wind and transmission projects are 1st, 2nd and 4th at the top of the list of project categories which complete on budget. As we discussed recently, it’s very good news that three of the four key technologies required for decarbonizing energy globally are low risk.
Electrolyzer plants barely exist, so as first of a kinds they are also high risk. And small modular nuclear reactors are also first of a kinds, unproven economically, and unlikely to achieve the cost and schedule benefits the nuclear community is hoping for, and if that actually occurs it will be in the 2040s, not this decade or next, so too late to be a significant contributor.
What does make sense is to decarbonize the grid, mostly with new wind and solar generation, add HVDC transmission as necessary to get the electricity around cheaply, and add grid storage like closed loop pumped hydro to bridge remaining weak spots. Then use the decarbonized electricity in appropriately scaled and engineered electrolyzers at the points of consumption, for example ammonia-fertilizer manufacturing plants. The price point will still be lower than delivering hydrogen around the place in most cases.
In other words, while the Nine Mile Point nuclear power hydrogen electrolysis deployment is a good example of what to do, it does not mean that purple hydrogen from nuclear power is going to be a significant part of decarbonizing hydrogen, and certainly not a pathway for hydrogen for energy.
Have a tip for CleanTechnica? Want to advertise? Want to suggest a guest for our CleanTech Talk podcast? Contact us here.
Latest CleanTechnica TV Video
CleanTechnica uses affiliate links. See our policy here.