Wind to Ammonia- the good, the bad and the ugly

TL&DR summary: renewable electricity to produce green ammonia is an important part of a decarbonized future. But the nature of the project and the end uses of the ammonia can make the concept good, bad or really ugly and dangerous.

Several years ago German Chancellor Olaf Shulz came to Canada to sign agreements related to sourcing ammonia for use in German power production, from future wind projects on Canada's east coast. And I've been doing media interviews on this topic ever since.

I've lost track of the projects- there are many of them in Newfoundland, Nova Scotia (Cape Breton) and New Brunswick. All of them follow the same basic pattern: plans to build new wind projects, generally onshore, to use the output of those wind projects to produce green hydrogen via electrolysis, then to feed the green hydrogen- plus nitrogen from the atmosphere- to a Haber Bosch plant to make ammonia and heat. The ammonia, as a compressed or refrigerated liquefied gas, is then destined not for the displacement of black, emissive, fossil derived ammonia from its essential use as the source of substantially all the artificial nitrogen fertilizer on earth, but rather as a way to move either energy or hydrogen to "markets" which need it. These are really artificial LNG energy export projects if you think about it.

If you're new to the poison gas that feeds literally half the humans and their food animals on earth, your recommended pre-read is my article here:

https://www.dhirubhai.net/pulse/ammonia-pneumonia-paul-martin/

Ammonia is extremely toxic and corrosive to human tissues. It is a very dangerous commodity chemical made in giant quantities, which is used either directly as a fertilizer or is converted to nitrates and urea for use as a fertilizer. It is difficult to imagine a more important molecule to human success and thriving. And all of it is made by a process invented in the 1910s by Fritz Haber, who received a Nobel Price for the invention. The process was commercialized by Karl Bosch, with a catalyst innovation being key to that commercialization.

The Haber Bosch process has been refined and optimized over the past 110 years, reducing its capital intensity and increasing its energy efficiency greatly. A typical ammonia plant today costs about $1 billion, including a steam reformer to make hydrogen from natural gas and a secondary autothermal reformer to make a hydrogen/nitrogen mixture from natural gas and air. The ammonia "synloop" is a catalytic reactor where a recycle compressor whirls the gas mixture around, first through the reactor, where heat is removed by generating steam, and then through a refrigeration system to remove the product ammonia. The reaction is favoured by high pressure, and while high temperature tends to push the equilibrium away from the desired product and back to the starting materials, high temperatures are needed to achieve a reaction rate that is acceptable on catalysts that are affordable and durable. The result is a fairly low conversion per pass, which isn't a problem as the equilibrium can be continually shifted toward ammonia by removing the ammonia as it is produced.

3H2 + N2 <===> 2NH3 +heat

The CO2 produced from natural gas in making the hydrogen to feed to the process is sometimes partially captured and used to make some of the ammonia into urea. This CO2 is released back to the atmosphere again when the urea is used as a fertilizer, so no, you can't get away with calling this "carbon capture and reuse".

Ammonia production today is therefore very fossil GHG intensive, both in terms of fossil CO2 and methane emissions from the use of the feed natural gas (or sometimes coal). The CO2 emissions from ammonia production alone are similar to the emissions associated with the entire aviation industry. Furthermore, when nitrogen fertilizers are used, N2O is released by soil organisms. N2O is a powerful, persistent GHG with a global warming potential of some 266x that of CO2 on the 100 year timeframe. Needless to say, we not only need to green our ammonia production but we must become smarter and more strategic about how we use it in a decarbonized future.

Wind to Ammonia

The simpleminded notion is that you can get rid of the extensive CO2 emissions of ammonia production by making the required hydrogen from water using electrolysis, and the required nitrogen by air separation using electrically driven equipment. And the trick there isn't that it's hard to do- Haber Bosch doesn't care where the H2 and N2 come from, so is perfectly happy to be fed nice clean H2 and N2 rather than dirty stuff derived using fossil energy. No, the problem is in affording the product when you're done.

Eastern Canada has some excellent wind resources, both onshore and offshore in various locations. However, onshore wind is only about 35% capacity factor, meaning that a 1 MW turbine doesn't make 1 MW x 24 hrs/d x 365d/yr = 8760 MWh of electricity per year, it makes about 3060 or so. But sometimes it makes 1 MWh per hour, and sometimes it makes zero, and sometimes it makes 0.5 MWh/hr and so on. The output of wind varies greatly with time over the year.

If you were to pair a 1 MW wind turbine with a 1 MW electrolyzer, the electrolyzer would make 35% as much hydrogen as if you fed the electrolyzer steady using the grid. And grids in eastern Canada are not clean and renewable. Nova Scotia's grid is, for instance, about 50% fossil fired on average, with a significant fraction of that energy coming from coal.

If electrolyzers were free, or at least very cheap, it might still look appealing. Sadly, they're not cheap, and they're unlikely to get as cheap in the future as some imagine.

https://www.dhirubhai.net/pulse/scaling-lesson-2-water-electrolysis-paul-martin/

You could of course pair TWO 1 MW wind turbines with your 1 MW electrolyzer, but two things happen: 1) sometimes you would be generating 2 MWh per hour and able to consume only 1 MWh, with the other 1 MWh being surplus to your needs, and 2) you would still have periods when both are generating no power. Sadly, if wind is a major source of power to the grid, when your project generates more power than it needs, surely other wind projects connected to the grid would do the same, making your power surplus to the grid's needs as well. It's not looking good for you to have a market for your surplus power. The result is, sadly, expensive hydrogen. Spending $2 million per MW for electrolysis capacity and feeding it only 35% capacity factor power, means that your capital intensity per kg of H2 is too high. And if you raise capacity factor by building more wind, the cost of every kWh increases because a fraction of those new kWh will be curtailed. The result is still expensive hydrogen. You can fool around with optimizations between cost of capital and cost of electricity, but what you really need is more green electricity- cheap, and not correlated with the wind.

Hybrid Power

The obvious solution is to add some renewable power of another kind. Solar for instance, is very cheap per kWh, and getting cheaper still. Unfortunately, the annualized capacity factor for solar in eastern Canada is very poor- less than 14%, relative to 30% in places such as Australia.

Hydropower is of course available with much greater capacity factors, and Canada is rich in hydropower resources. Sadly, the hydropower in eastern Canada is largely centred in Labrador, which is a considerable distance from any of these proposed ammonia projects. And the infrastructure to move that power to markets, is not focused on moving it to Newfoundland and Nova Scotia, but to export markets like the USA. While there is a 500 MW interconnector between Labrador's large hydro projects and the island of Newfoundland, clearly this interconnector cannot provide firm power to more than a single 500 MW electrolyzer. To my knowledge, none of the proposed projects is proposing, as part of the project, to build larger interconnectors to Labrador.

Wider Power Networks

Of course if you can connect your project to wind resources far enough away that the wind there blows when the wind nearby isn't blowing, you can increase capacity factors further. However, the projects in question seem to largely focus on highly localized wind projects, nearby to the electrolyzers and Haber Bosch plants. Again, long distance interconnectors to distant wind resources don't seem to be part of the design of these projects.

The "three pillars" of green hydrogen production are as follows: if you want your hydrogen to be green your renewable power must be:

1) New renewables, additive to the grid. Taking power from wind projects already built, just makes the average kWh on the grid, dirtier. The result is that your hydrogen isn't clean.

2) Proximity: you might be able to do a deal to buy power from Labrador or Quebec, but if there are no cables to carry the power from these places to your location, you're not fooling anybody

3) Temporal Matching: the green power your electrolyzer uses, must be produced WHEN your electrolyzer is consuming power. There's no point in saying that you're buying power from solar when your electrolyzer is running at night

And no, any electricity coming back to you out of storage batteries or pumped hydro projects is going to be too expensive to waste on making hydrogen. You might need storage to keep your plant from depolarizing when the power is off, and to keep certain parts of the Haber Bosch plant running in an idle state, but you won't be making hydrogen when the wind isn't blowing and the sun isn't shining.

Making green hydrogen is therefore more complex and expensive than you might originally think, despite wind electricity being fairly cheap- when it's available. The result seems to be structurally expensive hydrogen, if only wind and perhaps a small amount of hydro "firming" is available. And it's not possible to make cheap ammonia from expensive hydrogen.

But of course we're not done suffering yet.

Operating Haber Bosch Intermittently

Like most high pressure, high temperature processes, Haber Bosch reactor systems don't like to be interrupted. Most of the problems occur during startup and shutdown, and cycles of pressure and temperature are hard on the equipment. Certain equipment such as compressors are very intolerant of being started and stopped frequently, and they have a limited turndown.

Very smart people are working hard on making ammonia plants more easily adaptable to fluctuating operating conditions, but sadly the exact same thing happens to the ammonia produced by such plants as happens to the hydrogen to feed them: operated at low capacity factor, H-B plants make expensive ammonia, because the same capital expenditure is spread over fewer tonnes of ammonia product.

You can of course store hydrogen from your intermittent hydrogen production system in order to produce a steady stream to run a smaller H-B plant continuously. But sadly, although hydrogen storage is possible in theory, in practice its poor energy density per unit volume makes that storage very expensive. The result is, you guessed it, expensive ammonia.

Where Does This Make Sense?

There are places in the world which, unlike anywhere in Canada, have 30% solar capacity factor and 45+% offshore wind capacity factor which are actually anticorrelated in time. Deserts with an ocean to the west, where a convective wind blows in from offshore every night as the land cools down. These places, unlike eastern Canada, have access to not only cheaper kWh (because more solar means the average kWh is cheaper), but also higher capacity factors, meaning less storage and less intermittent operation.

Eastern Canada's wind resources are good, but is making a project in Canada sufficiently cheap to overcome a ~ 50% better use of capital where electrical capacity factors are higher? I sincerely doubt it. And I'd want more than the project proponents' say-so to conclude otherwise.

Use Cases- Making Good Ammonia into Bad Ammonia

Notice that so far we've been discussing ammonia for good use cases- displacing black ammonia made from fossils, in uses necessary post decarbonization, i.e. for fertilizer production and as the basis of the whole nitrogen chemicals industry.

But...what if we were to consider burning ammonia as a fuel?

Ammonia does burn, reluctantly. In fact, it's not the ammonia per se that is burning. Rather, what's happening is that ammonia is being thermally cracked in the flame to the H2 and N2 that it was made from, and the resulting H2 is burning. The result is no CO2 of course, which makes it seem favourable at first glance. That NOx emissions are inevitable, whether ammonia is burned in air or even in pure oxygen, is a major downer- requiring a NOx abatement SCR catalyst. But the real problem is that the energy efficiency of any scheme which converts electricity to hydrogen, hydrogen to ammonia, then either burns the ammonia directly or cracks the resulting ammonia back to H2 and N2 to burn the H2, is very poor indeed. The heat produced where electricity is in excess- in the Haber Bosch plant, must be put back in again to satisfy the 1st law of thermodynamics.

An extremely optimistic energy cycle chain is laid out in the figure below, assuming that the ammonia produced in eastern Canada is fed to a combined cycle power plant in Germany to make electricity. What does this Sankey diagram mean? It means "buy 10 kWh in eastern Canada, and get 2 to perhaps 3 kWh back again in Germany". That's not just pathetic- it's a dangerous distraction from real decarbonization in my opinion. You can of course make this value proposition worse: you can try to use ammonia as an engine fuel.

Note that the 70% LHV conversion efficiency for electrolysis can be improved a bit- to perhaps 83% LHV (100% HHV) efficiency if you use solid oxide electrolyzers (SOECs) which electrolyze steam rather than water. That allows the recovery of the 6.1 kWh/kg H2 heat of condensation of water vapour, using some of the byproduct heat of ammonia production for a useful purpose. However, the more direct heat integration you do, the less well these systems handle intermittent energy feeding. And SOECs, operating at temperatures above 800 C, tend to be a poor fit with fluctuating electrical supply as well. While solid oxide fuelcells are established technology, demonstrated at some scale, SOECs are of lower technology readiness level (TRL) than water electrolysis and hence aren't really available at this time, at scale, to use in ammonia production. Perhaps that will change in the future.

EDIT: here's another energy efficiency chain, prepared by Michael Liebreich:

(this chart is a screen grab from Michael's excellent talk here: https://www.youtube.com/watch?v=w0Q9cuF8zKg&t=2121s)

Ammonia as a Shipping Bunker Fuel

We should be horrified by the prospect of using a toxic gas as a shipping fuel. This notion violates the 1st principle of safety in design- it shouldn't merit a moment's additional consideration. While we do carry about 8% of the world's current ammonia production from place to place by ship, using ammonia as a shipping fuel is a whole different matter. Aboard a ship at sea, there's literally nowhere to run. And if you have a leak, you haven't just got a toxic emission to worry about- you have a gas which is very corrosive to human tissues while also being extremely toxic to sea life. The very best thing we can say about ammonia as a shipping fuel is that it might be cheaper, at scale, than a vastly safer option such as methanol, because methanol production from green CO2 and green hydrogen would need not just excellent wind+sun resources, but also a source of biogenic CO2. Finding those things in the same place is not going to be easy.

What's Really Going On Here?

The eastern Canadian wind to ammonia projects are numerous and some of them seem quite serious, spending considerable money on engineering studies, environmental assessments and the like. So what's really going on, if these projects are unlikely- for structural reasons- to make inexpensive ammonia?

Seriously, your guess is as good as mine!

Of course the first thing that comes to mind is subsidy grifting. Proponents of such projects are no doubt looking to significant investment tax credits from the Canadian government, which might effectively pay up to 40% of their capital cost. Why that would be in Canada's national interest isn't clear, but it would certainly help them reduce their high capital intensity to have Canadian taxpayers pay for 40% of the cost.

In addition, there may be other sources of either cheap funds- low interest or interest free loans, other government programs which cover R&D expenses etc. Yet more Canadian money going into the projects- largely public money. For what benefit?

Then there are subsidies expected on the offtake end, i.e. German or EU programs to cover the cost of lower carbon fuels of "nonbiological origin".

That explains why the focus of these projects is never to displace black ammonia use in Canada, most of which is made from natural gas in Alberta and used on fields in Alberta and Saskatchewan. Rather, the focus is always on wasting ammonia as a fuel, either directly or as a means to transport hydrogen- which itself is infeasible to ship for reasons that innovation will not be able to contend with.

Note that it is also possible, though I have no direct proof of any of this, that some of these projects are less than earnest efforts. Some of them may be boondoggles of the sort that eastern Canada has long been familiar with. Megaprojects promising jobs, jobs jobs...but ultimately siphoning public funds, roping in credulous investors, and ending up producing nothing.

What Should We Do Instead?

It seems obvious that if we really cared about mitigating climate change, we'd be focusing on greening up Nova Scotia's dirty grid, using those excellent wind resources along with ever-cheaper batteries to displace coal and then natural gas from the grid. That makes abundant sense. But of course, that would mean Nova Scotians would need to pay for the projects, rather than having the dreamy potential of having rich Germans to pay for an expensive export commodity. Decarbonization at home is for some reason considerably less sexy than the fantasy that someone else will pay more to decarbonize their grid instead, out of desperation I guess. And to me, that's just really sad.

Disclaimer: this article has been written by a human, and humans are known to make mistakes. If you can point out to me where I've gone wrong, with good references, I will be happy to correct my errors.

If however you don't like the article because I've taken a dump on your pet idea, please feel free to contact my employer, Spitfire Research Inc., who will tell you to piss off and write your own article.