The economics of power to fuels

Does it make financial sense to construct chemical plants that use surplus electricity to make liquid and gaseous fuels? This topic is rarely discussed in the UK but is an increasing focus of interest in the rest of Europe, and particularly in Australia.  As previous posts on this web have tried to suggest, full decarbonisation is completely impossible without such ‘power to gas’ and ‘power to liquids’ (or P2x).

In this short article I look at some of the results contained in a new presentation from Lappeenranta University of Technology (LUT) in Finland.[1] This shows that some P2x products are likely to be competitive with fossil fuel variants in 2030, even before carbon taxes

 Source: Fasihi and Breyer, Synthetic Fuels and Chemicals, figures converted to US $. Market prices  converted to US $ per megawatt hour.

Source: Fasihi and Breyer, Synthetic Fuels and Chemicals, figures converted to US $. Market prices  converted to US $ per megawatt hour.

The work by Mahdi Fasihi summarises detailed investigations on the likely cost of P2x for a variety of chemical energy carriers, ranging from hydrogen to dimethyl ether, a potential diesel substitute. He and his colleague, Professor Christian Breyer, have built detailed flow charts that show how the chemical plants that make these fuels will operate, both in terms of process and thermodynamically. Standard industry software can then convert these flows into estimates of the full costs of these products.

Fasihi’s process charts assume that the hydrogen contained in fuels is entirely derived from water electrolysis. The carbon (where necessary) is shown as being distilled directly from air. Although direct air capture is still in its early infancy, LUT has been at the forefront of research into possible technologies through its involvement in the Soletair project.[2]

The cost of hydrogen produced by water electrolysis is dominated by the price of the renewable electricity used to generate it. Although the impact of the capital cost of the hydrogen generators is far from negligible, the price of electrolysers is falling very sharply as technology improves and bigger machines are built. Modern electrolysis machines are approximately 80% efficient, meaning that for every 1 unit of electricity used about 0.8 units of energy are made available in the form of hydrogen. (This ratio will improve slightly in years to come). Therefore electricity bought for €40 a megawatt hour will produce hydrogen at raw cost of €50 per MWh.

Fasihi assumes electricity costs will come down, particularly in areas of the world with the best renewable energy availability. The presentation looks in detail at the places where a combination of wind and PV will produce large amounts of electricity for very large numbers of hours per year. This measure called ‘full load hours’, with the best places offering high renewables generation for 6,000 or more hours per year, meaning that any P2x plants associated with them have a reliable source of power.

As is well known, Australia comes out well in the full load hour rankings, as do parts of Chile, including the Atacama desert, and Patagonia. Somalia has good numbers, as do Tibet and the Great Plains of the US. For us in the UK, the west coast of Scotland and the Hebridean islands also score very well. 

By 2030, parts of the world are expected to see full costs of electricity from renewables at around €17-20 per megawatt hour (about £15-18, $19-23). (This compares with wholesale prices today for all forms of electricity of about £50 in the UK). The researchers calculate that electricity in 2030 can produce hydrogen for about $41 per megawatt hour based on the likely costs of renewables.

How does this compare to the market price of hydrogen today? Hydrogen isn’t an easy commodity to price because most of the gas is made in refineries to serve the petrochemical processes there. It doesn’t change hands much. When it is traded, it is usually also shipped from its source to the customer and transport costs for hydrogen are high because it has to be shipped in liquid form at very high pressure. But, very roughly, hydrogen prices are between about $65 and $118 for a megawatt hour of energy content for traded gas.[3] Today’s hydrogen costs more to make, using natural gas as its key ingredient, than hydrogen from electrolysis will be in 2030 in large parts of the world.

Australia sees its huge resources of wind and solar as helping to build a hydrogen business, particularly for shipping to Japan. There the gas will be used for fuel cell cars, if Japan’s ambitions are successful. It probably doesn’t make sense for Australia to transport hydrogen but instead to merge the gas into ammonia (NH3, or three atoms of hydrogen and one of nitrogen). Ammonia uses much less space and doesn’t need to be heavily pressurized. It can be turned back into hydrogen gas at the destination.

Fasihi at LUT calculates that the full cost of ammonia in places such as Australia in 2030 will be about $72 a megawatt hour. This compares to $50 for ammonia delivered in bulk today.[4] But if ammonia can be cheaply converted back to hydrogen, ammonia may become the way in which hydrogen is transported. Importantly, the main Australian research organization in the energy field just demonstrated a successful trial of extracting 100% pure hydrogen from ammonia for filling up fuel cell cars.[5]

Methanol made from hydrogen and captured CO2 is almost as cheap in 2030 as this liquid would be today: $81 per megawatt hour compared to about $76 for the fossil fuel version. However at today’s CO2 prices of around $22 per tonne in Europe, synthetic methanol would be about the same price as the conventional product, which would be burdened by permit costs. About 60 million tonnes of methanol are made each year and it is one of the top five products made from oil but not used directly in cars. Hydrogen and ammonia are two of the others.

Fasihi’s numbers suggest that the difference between standard natural gas (mostly methane) and methane from power to gas processes is larger. In the table, I’ve shown today’s natural gas price in Japan, one of the world’s higher cost locations for this fuel. At $34 a megawatt hour, the price today is well below the price of natural gas made from renewable hydrogen and CO2 of around $66. The carbon prices of today would not cut substantially into this difference.

What should we conclude? First, synthetic methanol stands out as an obvious focus for a renewable fuel. Second, that hydrogen from electrolysis may be competitive in some circumstances. It can be used for local energy storage in particular, and then converted back to electricity in a fuel cell. By contrast, renewable methane looks expensive, particularly in places such as the US where natural gas is extremely cheap. Last, ammonia is particularly interesting because it can substitute for natural gas in CCGT power stations and can be made in relatively small quantities for local use as the key ingredient for fertilisers in remote places.[6] We can envisage microgrids that provide electricity but store surpluses as ammonia, either for food production or for combusting for electricity purposes at times of seasonal lows in renewable production.

Most important of all, we just need to do more work on the economics and practicality of synthetic fuel. Full decarbonisation demands it. And, in parts of the UK, we have the potential to produce very carbon fuels at prices lower than most of the world.

[1] I think this presentation is absolutely outstanding. LUT has produced much of the most insightful research, both practical and academic, into P2x. Fasihi’s work summarises and extends existing knowledge.


[3] Source: $65 Northern Gas Networks Page 260      $118, McKinsey Page 58

[4] Source: Methanex published prices.


[6] See the Proton Ventures web site for some details of small-scale Haber Bosch plants.