Plant-based foods

This article was published in Riverford’s excellent online magazine Wicked Leeks. Thanks to the editor Nina Pullman for suggesting this topic.

The large packaged food companies are working on a huge variety of new products that closely imitate meat, but actually contain no animal matter. 

Giants such as Nestlé and Unilever are developing foods as various as lamb and smoked salmon substitutes from raw materials that often include peas, soya beans and coconut oil. Many analysts see huge continued growth in plant-based alternatives, such as the vegetarian burgers of Beyond Meat and Impossible Foods, the American companies that helped kindle interest in meat-free products.

Fake meat burgers from the likes of Impossible Foods are made from processed plants

Should we welcome the growing interest in replacements for meat? The carbon footprint is certainly lower than the original products. And vegetarian meat substitutes also save water and use less land.

While the precise numbers are still argued about, the overall conclusion is not. From a carbon footprint viewpoint, plants are a lot better than plant-based meat, which in turn is better than meat (particularly beef).  

We cannot avoid the conclusion that all agricultural beef generates lots of greenhouse gases, even when cultivated under the very best conditions. In my experience, I think that most beef farmers now (reluctantly) accept this. Even the argument that small numbers of truly free-range cattle add to carbon storage in the soil doesn’t really stand up to scrutiny, however much sympathy we might have with it.

But in many other ways, the current generation of meat substitutes offer few advantages over traditional products. They tend to contain far more added salt and, surprisingly, can also have more saturated fat. 

The new foods sometimes also have more calories than the foods they are trying to replace; the Beyond Meat burger contains about 250, slightly more than some of its competitors. It is very likely that improvements in meat substitutes will eventually bring them into line with their competitors but there’s currently little dietary reason to eat the new products.

Perhaps most importantly, meat replacements are almost all highly processed. The manufacturers have invested huge sums in making these products as similar as they can in their taste to the meat equivalents. They are also trying to make them look exactly like the product they are imitating, using beetroot juice, for example, to imitate the blood in beef. 

This has meant considerable engineering of foodstuffs to mimic the taste and texture of animal products. The Impossible Burger, one of the most successful entrants to the market, has more than 20 ingredients including some added vitamins and minerals. 

After water, the largest constituent is ‘soy protein concentrate’ but this is followed by a wide variety of other heavily manufactured ingredients, such as preservatives. 

These meat substitutes can also be very expensive. When I looked at the Tesco website this week, I could buy standard burgers for about half the price of the Beyond Meat plant equivalent.

What about the taste? The evidence is ambiguous; many people like meat substitutes but others find the texture or ‘mouth feel’ very unappealing. The makers of the first generation of processed plant products have tried hard but haven’t yet accurately mimicked all the complex characteristics of meat.

And although they certainly offer substantial carbon savings over meat, we would do much better to consume the ingredients in their unprocessed form.

When eaten simply cooked, the carbon footprint of key ingredients, such as soya, potatoes or peas, is a fraction of when they are served as a highly manufactured meat substitute. 

For those seeking to move away from animal products, it may make sense not to switch to the foods that are trying the hardest to imitate meat. 

Hydrogen is the best way of stabilising the electricity system

The last few days have been windy across northern Europe and turbines have provided a large fraction of power needs from Ireland to Finland. But these storms follow months of unusually European quiet weather, causing wind’s share of electricity supply to fall well below typical levels. This swing between scarcity and abundance illustrates the central question of the energy transition: how can countries cope with the huge variations in electricity supply arising from the unreliability of wind and solar?

The answer is by using hydrogen. When electricity is in surplus, as it was in many countries as Storm Eunice swept over Europe on Friday 18th February, electrolysers can separate water into hydrogen and oxygen. This can soak up the electricity that would otherwise have no productive use. At times of shortage, the hydrogen can then either be burnt in a gas turbine or fed into a fuel cell to generate power. 

The standard criticism of this idea is that is wasteful of energy; we will get back less than half the energy initially fed into the electrolyser as the storm blows. But the value of the power used will be very low because of the surplus in the electricity markets. When the hydrogen is turned back into electricity, the reverse will be true because of the shortage of supply. Hydrogen power stations will be generating valuable power. Perhaps even more importantly, conversion to hydrogen is the only way that huge quantities of energy can be economically retained for months at a time. Batteries would be a far more expensive way of achieving very long duration storage.

 A national electricity system that continues to invest heavily in renewables, complemented by the development of a hydrogen infrastructure to deal with the variability of supply, is an idea that has not been properly explored. It needs serious consideration now, not least because the continued expansion in renewable electricity sources will make periods of electricity surplus much more frequent. 

To give one example, the UK is proposing to add almost 30 gigawatts of electricity from offshore wind by 2030. In the last few days national demand for power in the hours after midnight has been less than this figure. Without taking in account the existing wind capacity of almost 25 gigawatts, and all the other low carbon sources such as nuclear and biomass, wild winter storms will soon overwhelm the UK’s ability to use the electricity generated.

Using hydrogen as the means by which the electricity system provides guaranteed power faces three major obstacles. None are unsurmountable although all are challenging. The first is the requirement to build up sufficient electrolyser capacity. The second is to locate the cheapest and safest means of storing enough hydrogen to cover many weeks use and the third is providing the gas turbine and fuel cells required to balance electricity demand and supply when the wind isn’t blowing.

Let’s look at these issues in turn. Can hydrogen can be generated in sufficient quantities to provide the guaranteed back-up that we require? When the UK has added an extra 30 gigawatts of offshore wind, it will probably need at least 10 gigawatts of electrolysers to convert periods of surplus power into hydrogen. That’s at least twenty times the total global installed capacity today. Nevertheless industry researchers project well over 200 gigawatts by 2040 and every forecast that is issued increases the expected growth rates for electrolysers. New factories to build many gigawatts a year are being constructed around the world and the largest proposed installations are now as much as 800 megawatts, more than today’s global installed base.[1]Although acquiring enough electrolysers to handle surpluses from offshore wind will be challenging, the supply industry is capable of growing at the required rate.

 The next question is how the hydrogen will be stored - possibly for very long periods - after it has been generated. Some northern European countries, including the UK, are in the lucky position of having extensive underground salt deposits that can used to create impermeable storage caverns. This is almost certainly the cheapest and safest way of holding large reserves of hydrogen. Researchers recently suggested that Europe has at least 7,000 terawatt hours of potential hydrogen storage capacity, nearly twice the continent’s annual electricity use.[2]

The first salt cavern for storing hydrogen was built in the UK in 1972 and is still operating today.[3] It does not store a large quantity but plans are being developed at sites in Germany and the US for caverns that can hold more than a hundred and fifty gigawatt hours of energy, equivalent to several hours of UK electricity use. As with electrolysers, salt cavern use will need to be increased by many orders of magnitude to enable countries to store months of energy, but the technology exists and is economic.

 The last major obstacle to a plan for using hydrogen to balance electricity supply and demand is the need to convert existing gas power stations. The major gas turbine manufacturers now offer products that operate on high percentages of hydrogen in a natural gas mix. Within a few years, 100% hydrogen turbines will be widely available. The cost will not be significantly different from standard natural gas equipment and the operating efficiencies will be very similar.

 The UK and other countries can develop large hydrogen-fired power stations. One in the Netherlands is planned to begin operating in 2023.[4] At a much smaller scale, hydrogen can be used to operate fuel cells to deliver power into electricity grids. The most advanced plan is the French territory of Guiana, where a solar farm will produce power during the day, and convert some to hydrogen to be used in a fuel cell to deliver overnight electricity.[5]

The world understands it need to shift rapidly to an energy economy based around renewable electricity. This will be cheaper than gas and coal and emissions will be close to zero. But the variability of electricity supply requires huge storage capacity. Batteries cannot economically provide this. Hydrogen is the only realistic way of dealing with the peaks and troughs of renewable supply. And it would reduce dependence on currently expensive gas and the associated emissions. This idea needs urgent investigation in the UK and elsewhere. 

[1] NEL Hydrogen presentation 16th February. https://nelhydrogen.com/quarterly-presentation/

[2] https://www.sciencedirect.com/science/article/abs/pii/S0360319919347299

[3] https://innovation.engie.com/en/articles/detail/hydrogene-souterrain-stockage-sel-cavites-mines/25906/12/1

[4] https://www.nsenergybusiness.com/projects/nuon-magnum-power-plant/

[5] https://www.pv-magazine.com/2021/09/30/large-scale-solar-plus-hydrogen-project-secures-25-year-ppa-in-french-guiana/

How much electricity will the UK need as it switches to electric heating and cars?

Decarbonisation depends on moving countries off natural gas and fossil fuel cars, largely replacing them with electricity. This is widely understood. 

However I can find few surveys of how much power demand will change as the transitions happen. This article represents a very approximate estimate of the impact of complete electrification of heating and transport in the UK. The numbers are necessarily imprecise and I think a proper programme of research is worthwhile. 

We need to think about three things:

 ·      Total power needs over the course of the year so that we can size the renewable and storage fleets correctly 

·      Peak electricity requirements in order to ensure the distribution and storage system scan can handle demand

·      Requirements at the final substation level so that upgrades can be costed and planned. I suspect that this will need to start soon because of the explosive growth of electric cars.

My rough calculations can be summarised as follows: 

Section 1: Total annual power needs

·      Total UK electricity demand will rise from 279.2 terawatt hours today to about 394.7 TWh when 100% of vehicles are EVs.[1] (The calculation assumes no change in car or commercial vehicles numbers). Just under half of this rise arises from the electrification of private cars.[2]

·      Switching to heat pumps for 100% of UK homes would move the UK from 394.7 terawatt hours to 505.5 TWh. (Calculation assumes no growth in numbers of homes).

·      The impact of moving both sectors to full electrification would therefore be about 226.3 TWh, or 81% increase in annual UK power needs.

Electricity annual demand estimates, personal calculations from published numbers

Section 2: Peak electricity requirements

·      I assume that a large fraction of EV charging demand can be shifted away from peak times. But I think it is an error to assume, as some do, that all charging can be shifted to night-time periods of low demand. My rough approximations suggest that EV charging will add about 5 gigawatts to the level of electricity demand at evening peak.

·      Heating is different. Heat pumps will probably have a usage pattern not significantly different to conventional gas and oil boilers, with peaks in the morning and early evening. This is shown in academic research. It will be challenging to meet this demand using the existing electricity system.

·      My guess is that the impact of 100% heat pump penetration in domestic homes will add about 65-70 gigawatts, or around 135% to peak electricity demand. This is assuming much electricity demand for heat pumps can be shifted to low demand hours of the day. But even shifting all heat demand to the depths of the night - an impossibly challenging target – will still mean a more than doubling of peak demand.

·      The total impact, combining domestic heat and electric vehicles will be about 70-75 gigawatts or an extra 140-150% added to peak demand.

Peak electricity demand, personal calculations from published data

Section 3: Maximum requirements at the most local level

·      Domestic heating and car charging largely occur at the very edge of the network in people’s homes. The average home today typically uses about 0.4 kW over the course of the day and year, although this will obviously vary enormously from hour to hour, depending on whether the dryer, cooker, toaster or other power-hungry appliances are in use. Home car charging can be up to 7 kW and a big heat pump can use over 4 kW. In other words, the new applications can multiply peak usage in the home several times.

·      Heat pumps and electric vehicles are likely to grow fastest in more prosperous areas. I guess that at the edge of the network maximum electricity needs are likely to rise many fold from today’s numbers in some areas. This means that upgrades will be required in the local distribution systems and at the neighbourhood substations often covering about a hundred homes. 

Section 1: Total annual power needs

Domestic heating, including hot water and cooking. 

 Gas demand across the UK for domestic purposes was about 299.3 TWh in 2020. This figure was probably slightly inflated by the numbers of people staying at home to work. The 2020 figure was about 1.5% up on the previous year. In addition, the amount of oil used for domestic heating was equivalent to about 26.7 TWh in the 4% or so of UK homes using this fuel. Coal used for home heating adds another 4.1 TWh or so.[3] Total input of fuels was therefore about 330.1 TWh

I assume the following approximate percentages for the efficiency with which the fuel is turned into heat

       Gas boiler       - 85%

      Oil boiler         - 75%

      Coal                 - 40%

These figures give us an estimate of 276.1 TWh for the heat currently used by UK homes. We do know, by the way, that the UK’s poor insulation standards across its home network mean that many people do not heat their home sufficiently for their health and comfort.

 If we move to 100% heat pumps, we won’t need as much energy as this. Heat pumps provide more heat that the electricity input required. Estimates vary but approximately 2.7 units of heat are delivered by the typical UK heat pump for every unit of electric power used.[4] This measure is called the Coefficient of Performance (COP). 

 Dividing the 276.1 TWh total heat need by 2.7 means that the UK would need an extra 102 TWh or so of electricity to meet all domestic requirements for powering heat pumps in homes currently using gas, oil or coal. 2020 total electricity use was about 279.2 TWh, a figure pushed down by the pandemic.[5]

Of this, domestic electricity consumption in 2020 was about 107.8 TWh but this includes some electricity used today for heating in storage radiators or other electric appliances. About 8.5% of the UK’s 27.8m households use electricity for heating, principally in the form of storage radiators, usually charged with heat overnight. I estimate that these use 7,000 kWh of electricity per year on average, implying a total demand for electricity for heating of about 16.5 TWh.

If these homes are switched to heat pumps, they would need less electricity. At a COP of 2.7, the aggregate saving for the UK is 10.4 TWh. 

So the impact of switching all homes to heat pumps is as follows[6]

      UK electricity need before a 100% switch to heat pumps    = 279.2 TWh

     100% heat pumps for gas, oil and coal heated homes         = 102.2 TWh

Reduction in electricity need from converting currently electrically heated homes to heat pumps                                                           = minus 10.4 TWh

 Total demand after conversion to heat pumps of all UK homes = 371.0 TWh

(Percentage increase as a result of switch to heat pumps   =32.9%)

(NB, a small fraction of electricity is lost during transmission and distribution. More power needs to be generated than is consumed. This loss currently amounts to about 7%.)

Car charging

The typical car in the UK drives about 7,500 miles/12,000 km a year. On typical journeys – not just fast motorway drives or very slow urban trips – an EV probably consumes about 1 kilowatt hour of electricity per 6 kilometres. So each car needs approximately 2,000 kilowatt hours a year. There are approximately 32 million passenger cars in the Great Britain, and probably about 33 million including Northern Ireland, although I cannot find the data to confirm this number. Assuming a total of 33 million, the total electricity consumption of the passenger fleet would be 66 terawatt hours.[7]

What about the impact of electrifying all transport, including buses, light vans and larger lorries? The easiest way to provide an approximate estimate is to look at the consumption of fuels (diesel and petrol) for all types of vehicles. The RAC Foundation says this figure was about 46.9 billion litres in 2019. I have ignored the artificially low number for 2020.[8] The RAC number combines diesel fuel and petrol. Diesel has an energy value per litre about 10% greater than petrol and the average is probably about 10.4 kilowatt hours per litre. 

These numbers suggest a total energy consumption of all the vehicles on UK roads of around 487.8 TWh.

Petrol and diesel are less efficient at moving vehicles than electricity. Whereas the motor in an electric car will deliver perhaps 90% efficiency, the average internal combustion engine probably offers no better than 25% conversion of energy to motion. 

When the UK has replaced all internal combustion engine vehicles with batteries (and therefore assuming that fuel cell cars and lorries never gain a significant share of sales) the amount of energy required will therefore fall a long way from today’s levels. I estimate the number will be about 135.5 TWh, roughly double the electricity need for private cars alone.

Current electricity consumption is about 279.2 TWh. Full electrification of all vehicles would add about 48.5% to this figure

Current electricity consumption                                                    = 279.2 TWh

1)    Electricity needed for private cars alone                                = 66.0 TWh

(Percentage increase as a result of switch to electric cars         = 23.7%)

2)    Electricity needed for full conversion to battery vehicles    = 135.5 TWh

(Percentage increase as a result of the switch of all vehicles    = 48.5%)

Summary Section 1: electrification of domestic heating and all vehicles.

The total extra amount of electricity needed to handle the requirements for all domestic heating (not commercial buildings, offices or industry) and all road vehicles will be approximately as follows:

Current electricity demand                                                                 = 279.2 TWh

Extra demand created by shifting 100% to domestic heat pumps   = 91.8 TWh

Extra demand created by shifting all transport to electricity            = 135.5 TWh

Total increase                                                                                       = 227.3 TWh

 (Percentage increase as a result of the two switches                     = 81.4%)

Is this a manageable increase? Very approximately, it will equate to the power output of 50 GW of offshore wind. This is roughly the amount that is likely to be in place in UK waters by the early part of the next decade. In other words, it is a large increment to electricity demand but is clearly within our reach.

That’s the first question answered, albeit with considerable uncertainty. Over the course of a year we can switch the most polluting activities to electricity, as long as we invest enough in renewable energy. But of course industrial and commercial heating will add additional electricity needs when they switch to low carbon heat.

What about the second question? Can electricity match the needs on an hour by hour basis, as well as year to year? This is much more difficult question to respond to. 

Section 2: Peak electricity requirements

Domestic heating including hot water and cooking

 A 2019 paper in the journal Energy Policy gave us very useful data on the hourly patterns of gas use in domestic homes.[9] It looked carefully at dates of very high heating need, focusing on a very cold period in December 2010. 

The research showed that typical heat requirements in natural gas heated UK homes on these days peaked at around 170 GW at about 17.00 in the early evening. This is approximately the time of the maximum UK electricity demand as well. 

Since 2010, electricity demand from homes has fallen sharply. Gas demand has declined much less, and there has actually been a rise in recent years as the efforts to improve insulation have stalled. I have assumed that a really cold day still requires about 170 GW of peak gas availability.

Peak electricity demand so far in winter 2021/22 has been about 48 GW. This occurred at around 17.00 on 10 January 2022. Average temperatures were higher than the figures recorded in 2010, so the comparison between 48 GW of electricity demand and 170 GW of gas use is not entirely fair. A better estimate might be that the need for electricity on a really cold day is now approximately 50 GW, before taking demand management into account. But under any assumptions, peak heating demand with gas is still several times that of electricity use. 

The next phase in the calculation is to add in the use of oil and coal for heating to supplement the gas needs. If we assume that these alternative fuels provide the same percentage of heat requirements as they did in the analysis in section 1 above, the total demand for all domestic heating fuels would be about 187 GW on the coldest day in 2010. Boilers aren’t 100% efficient, as discussed above, and this figure delivered about 156.4 GW of heating into the home.

If we transfer all heat demand to heat pumps, we will not need as much input energy. In the section above, I use a COP estimate of 2.7 across the whole year. On a very cold day, this figure will be much lower. I suggest a number of 2.0 is more appropriate, implying that total electricity demand for home heating will be about 78.2 GW. (If we include all non-metered domestic buildings, that number would be well over 100 GW). 

 The conclusion is therefore as follows

Peak UK electricity demand on a very cold day in 2022                  = About 50 GW

Extra electricity demand from a 100% conversion to heat pumps  = About 78.2 GW

(without demand management) 

Total electricity demand peak with 100% domestic heat pumps    = About 128.2 GW

(Percentage increase as a result of the switch                                 = About 156%)

This is a very different set of numbers to the ones in Section 1 above. It can be argued that they are too pessimistic; heat pumps can be temporarily turned off at times of highest electricity demand and their output could be replaced by thermal stores of heated water for many hours a day. But these stores will add substantially to the installation cost of new heat pumps. And in cold winter period, electricity demand never falls to very low levels. So moving heating requirements to periods of the day in which demand is lower – principally between 23.00 and 07.00 does not shave much from peak demand.

 On 12th January 2022, minimum demand was about 28 GW, compared to a maximum demand of 48 GW. If all the needs for electricity for heating were shifted to the lowest use periods, maximum demand would therefore fall from about 128.2 GW to about 108.2 GW, or over twice 2022’s highest level. But this switch is implausible; the reality is that heating demand serviced by heat pumps cannot be entirely switched to the overnight hours. There won’t be enough thermal storage to make this possible. 

 These numbers are indicative but they strongly suggest that a UK with 100% heat pumps will need to accommodate peak electricity demands of well over twice today’s levels. Perhaps a figure of around 115-120 GW is plausible compared to the current 50 GW on the coldest days.

In the high summer, of course, heating demands will fall to near zero. The seasonal variability of electricity need will therefore be significantly magnified. The low peak demand levels of around 35 GW in high summer will stay the same, but midwinter peaks will rise to perhaps 120 GW or almost four times as much.

 Charging of all road vehicles

In section 1 I asserted that demand increases from moving to 100% electric cars and other vehicles would add about 135.5 TWh to annual UK electricity needs. Spread equally over all 8,760 hours of the year suggests a continuous need of about 15.5 GW to add to peak needs of around 50 GW on today’s coldest days. This is an addition of about 31%. 
We can improve this by prompting drivers to move to charging at periods of low overall demand, particularly by using varying prices according to the time of day. But the extent to which this will affect charging patterns is as yet unclear. 

A 50 kWh battery charging at a 7 kW home charger will move from 20% charged to 100% in about eight hours, and possibly more. (Charging rates drop as the battery comes close to being full). So if the regulations and incentives (such as lower prices) can be put in place, most domestic car charging could be carried out between 23.00 and 07.00. 

 But if personal car charging is being carried out at slower rates for example at the workplace, it will not be possible to only charge at the periods when overall electricity needs are at their lowest. And for those car owners taking long journeys during the day, and needing to charge away from home or work, only using electricity at the periods of lowest demand will be impossible.

On the other hand many commercial vehicles will be able to charge overnight. Buses, for example, can be filled up at the end of the working day. But some other vehicles, such as overnight delivery vehicles for supermarkets, will be operating at times of low overall electricity demand and therefore will not be available for battery charging.

I guess that the UK can probably hold peak vehicle charging demand in a 100% electrified system to around 5 GW by carefully incentivising drivers to fill batteries at times of low overall electricity demand.

 Summary Section 2: electrification of domestic heating and all vehicles.

Peak electricity demand on a very cold winter day is about 50 GW and occurs at around 17.00 on working days.

The 100% use of heat pumps is likely to raise that figure to about 115 – 120 GW. And vehicle charging might be expected to add another 5 GW.

Total electricity needs are therefore going to peak at close to 125 GW, or two and a half times current levels. This is a very substantial challenge for electricity supply.

Section 3: Maximum requirements at the most local level

It is inevitable that at the edge of the electricity distribution network, the impacts of electrification of heating and of transport will have even more dramatic effects. Merely by statistical chance, the pattern of hourly demand at local substations serving perhaps a hundred homes will be more varied than across the country as a whole. 

So, whereas electrification of heat and transport may raise electricity demand overall to about 250% of the current level, at individual substations or transformers the effect will be even more striking. In a small area all homeowners could decide to charge all their cars at once and turn up the thermostats to heat their houses to higher temperatures. 

The consequence of this possibility is that the local distribution companies will be obliged to increase the capacity of all substations by a percentage far greater than is going to be required for the central skeleton of the transmission system (‘the National Grid’). Whereas at the core of the network capacities will need to be raised to two and a half times present levels, individual end-of-branch transformers will have to be improved so that the scope for carrying power will be multiplied by perhaps five or ten times. Only experience will tell us what these new capacity levels will have to be.

Multiplying power flows by five or ten times at the very edge of the network will be costly and difficult. I have been unable to find estimates for the money that will have to be spent at individual transformers. However there are about 400,000 edge of network transformers in the UK and if the cost is £10,000 at each location the bill will be about £4bn or around £140 per home.[10]

Conclusions

I believe that handling peak demand for home heating using electricity will be a very difficult challenge to address. We are likely to see a two and a half times multiple of existing peak electricity demand. As a consequence, the UK and other countries with poorly insulated houses may need to use hydrogen for some fraction of domestic heat. 

An alternative, which I talked about in a recent post, is to use hydrogen as the key storage medium for a world in which peak electricity demand has risen several fold from today’s levels. The seasonal variability of peak needs will rise sharply as a result of heat electrification and I strongly suspect it makes real sense to store surplus power in the form of hydrogen to address these steep swings in demand. This can either mean the use of hydrogen power stations for reconversion to electricity or the distribution of hydrogen via the existing gas grid to fuel cells close to the point of consumption. This idea has not been explored sufficiently yet. 

[1] For the purpose of making these calculations I assume that these future EVs are pure battery vehicles, rather than plug-in hybrids. The market share of hybrids in the UK and elsewhere is tending to fall and most industry commentators suggest that the pure EVs will dominate within five years.

[2] I use numbers in this analysis rounded to the first decimal point. I am not intending to imply this degree of precision but wanted to ensure that the additions are clear.

[3] These numbers are derived from the DUKES report of the UK government. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1023276/DUKES_2021_Chapters_1_to_7.pdf

[4] This study, for example, gives an estimate of 2.65 for Air Source Heat Pumps and 2.81 for Ground Source: https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/606818/DECC_RHPP_161214_Final_Report_v1-13.pdf

[5] Electricity production was about 312 TWh, a number that includes losses in transmission and electricity consumed by the energy industry itself.

[6] I have ignored the very small number of UK homes with existing heat pumps.

[7] I have not included the existing battery car fleet of about 1.5% of the total  in these calculations, nor the 3% of plug-in hybrids (which often principally use petrol or diesel because of the small size of their batteries).

[8] https://www.racfoundation.org/data/volume-petrol-diesel-consumed-uk-over-time-by-year

[9] https://www.sciencedirect.com/science/article/pii/S0301421518307249#bib34

[10] https://www.emfs.info/sources/substations/final/

The Aquind interconnector

A month ago the UK energy regulator stressed the importance of adding more links between the electricity systems of Britain and the surrounding countries.[1] This will help ensure power is available when the wind isn’t blowing. When it is, excess electricity can be sold to our neighbours. 

Aquind’s schematic illustration of its proposal

Ofgem wrote

Currently Britain has seven operational electricity interconnectors, connecting it to Ireland, France, Belgium, the Netherlands, and Norway, providing almost 7% of the UK’s electricity last year.

The UK Government wants to more than double existing interconnector capacity by 2030 to support its target of quadrupling offshore wind capacity by the same date.

Give the importance almost everybody places on the development of interconnectors it was strange to see the government turn down an application for a 2 GW connection with France a week or so ago.[2] This article looks at why the interconnection was blocked and suggests that the government’s decision is vulnerable to court action in the form of what is called ‘judicial review’. Judicial review allows judges to tell government or its agencies that they have not acted in accordance with their responsibilities or that their decision is patently flawed. The rejection of this interconnector looks vulnerable to being overturned.

The proposed interconnector 

A new electricity interconnector (Aquind) between Normandy in northern France and the UK’s southern coastline would have created a 2 gigawatt two-way link. Potentially able to supply over 5% of Britain’s electricity needs if fully utilised, the link was strongly recommended by the government’s planning advisers. Although the planners saw some limited problems with the proposal, these were heavily outweighed by the advantages. Very few, if any, large scale projects like this that are accepted by the planning body are then rejected by government. 

So why was this application different?

The proposed 2 GW pipeline would have required cabling through the city of Portsmouth, disrupting traffic and other activities during the construction period. It might also have had detrimental impact on some local landscapes. 

But the landscape effects were minor. The required interconnector requires construction of an inland convertor station next to the connection point to the electricity network. This would affect people’s view at a Grade II* listed cottage. The Planning Inspectorate, who carried out the detailed assessment of the project, regarded this adverse impact as ‘less than substantial’. 

In addition, a small building of a height of 4 metres at the point at which the electricity cable and fibre optic links carrying internet data would have had some effect on a historic site. The applicants intended to mitigate the impact by surrounding it with trees. The planners also did not regard this as a serious problem. 

The plan for the interconnector was, however, deeply unpopular in Portsmouth and the immediately surrounding area. Local politicians seem to have been united in trying to block the new link. The concerns were principally over the congestion caused during the construction period.

Route from Portsmouth up to the proposed substation. (With apologies for lack of clarity of the map).

 Central government faced a dilemma. Its own planning inspectorate had summarised its recommendation to proceed as follows:

overall, the need case for the Proposed Development strongly outweighs the identified disbenefits[3]

and 

The ExA (the planning inspectorate) therefore considers that the final balance indicates strongly in favour of granting development consent.

So how could government block the plan and so avoid political problems in the Portsmouth area? It used a reason that its own rejection letter acknowledges is extremely unusual. It said that its adverse decision was based on the failure of the applicants to fully assess a possible alternative site for the connection of the 2 GW cabling to the national electricity network (‘the Grid’). 

Although the vast bulk of applications for permission to construct a building or to carry our other construction works do not require the applicant to show that its plans are better than all other possible plans, the government made an almost unprecedented exception for this proposal. In fact it admits that applicants need to prove their proposal is better than all others ‘in exceptional circumstances only’. 

In this case, although the applicants had fully assessed a wide range of different Grid substations to connect to, they had missed out one particular location. This substation, called Mannington, is about 60 km to the west of the original choice of connection point. 

 Why hadn’t they evaluated this particular site? The reason appears to be that when the interconnector was first being planned the Mannington substation was proposed as the point at which electricity from a large offshore wind farm (‘Navitus Bay’) was fed into the Grid. The interconnector applicants would have assumed that the substation was not available for a further very high powered interconnector. 

But then the wind farm proposal was rejected by the planning inspectors, and in this particular case the recommendation was followed by the government in September 2015. The substation therefore became an additional possible site for linking the proposed interconnector to the Grid. It seems as though the applicants knew this but failed to include a careful evaluation of the route in their application documents.

The government’s letter of a week ago to the company wanting to build the interconnector gives the failure to assess the advantages and disadvantages of using the substation at Mannington as the principal reason for blocking the application. It is not there is anything particularly wrong with the existing plan for the interconnector, they indicate, it is that using Mannington just might be better. 

Except that we already know that it would not be. Why? Because some of the principal reasons for turning down the offshore wind farm proposal in 2015 were to do with the use of the Mannington substation.[4] The substation sits within the south east Dorset Green Belt, an area supposed to be protected from any forms of new building, such as would be required for the structures and buildings to bring more electricity into the substation. So the planning inspectorate called the proposal to connect the offshore wind farm to Mannington ‘inappropriate development’. Exactly the same argument would apply to bringing the electricity from the interconnector.

And to get to the substation would require crossing the New Forest National Park. To give permission for this (and the habitat destruction that would, at least temporarily, result) would require what the planning inspectors called ‘exceptional circumstances’. They say that these exceptional circumstances do not apply.

 There is no reference in the recent rejection letter from the government to these earlier refusals. But, to summarise, the government has blocked a 2 gigawatt interconnector on the basis that the applicants hadn’t assessed a potential alternative connection to the Grid. But the government itself had rejected an earlier application to use that site for a similar purpose in 2015.

 This looks to me like an obvious case for judicial review on the basis of the manifest irrationality of this month’s decision. But by the time this is all sorted out we will have lost yet more valuable time in the race to get Britain better connected to the energy networks of continental Europe to help decarbonisation. 

[1] https://www.ofgem.gov.uk/publications/ofgem-gives-green-light-investment-new-interconnectors

[2] https://infrastructure.planninginspectorate.gov.uk/wp-content/ipc/uploads/projects/EN020022/EN020022-004431-EN020022%20-%20Secretary%20of%20State%20Decision%20Letter.pdf

[3] https://infrastructure.planninginspectorate.gov.uk/wp-content/ipc/uploads/projects/EN020022/EN020022-004425-EN020022%20-%20Final%20Recommendation%20Report.pdf

[4] https://infrastructure.planninginspectorate.gov.uk/wp-content/ipc/uploads/projects/EN010024/EN010024-000043-Examining%20Authority%20Recommendation%20Report%20-%20%20Main%20Report.pdf

Transporting energy as hydrogen not electricity

I want to put forward what may seem an utterly ridiculous idea. My scheme is a way of dealing with the need to electrify as much of our energy requirement as possible, while minimising the total cost of the power and providing 100% reliability. I suggest we transport a large fraction of our energy need in the form of hydrogen rather than electricity.

The idea stems from the high price of transporting electricity compared to natural gas. The large UK utilities publish annual statements that summarise their financial accounts. For the largest, Centrica, the cost of transporting electricity to domestic users was about 4.6 pence per kilowatt hour in 2020[1]. By contrast, it was less than 1.2 pence for natural gas, or just over a quarter as much. Typically, the cost of transporting electricity is not much less than the price of buying from the generator.

This sets off the idea: would it better to run our future energy system by converting some electricity to hydrogen and then shipping the hydrogen in today’s gas mains? The hydrogen would arrive at homes, or perhaps at local electricity substations, and then be converted back to electricity in a fuel cell for use in the home. 

Conventional route

Electricity generated  -> Transmitted -> Arrives at home

Suggested alternative

Electricity generated -> Converted to hydrogen -> Piped as H2 -> Converted back to electricity at the destination.

Is this an insane idea? No, I don’t think so. 

Let’s assume that the wholesale price of electricity is 5 pence per kilowatt hour, or £50 per megawatt hour. This is about the typical level of recent years although at the moment – January 2022 – it would be very much higher. Add the price of transport of 4.6 pence, and the cost rises to 9.6 pence per kWh.[2]

 For the alternative route, the wholesale price would be the same – 5 pence per kilowatt hour. Assume 80% conversion efficiency at the electrolyser, and that cost rises to 6.25 pence. Then add the cost of hydrogen transport in the gas network, assuming it is the same as shifting natural gas, of 1.2 pence per kilowatt hour, meaning that is delivered to the home or local substation for 7.45 pence. If the fuel cell is 62% efficient, then the total cost to deliver the electricity to the home is 12.0 pence. 

At first sight, this is a relatively large difference. The kilowatt hours of ‘electricity’ delivered via the gas network cost 12 pence, compared to 9.6 pence across the conventional electricity grid. But the customer also has the value of the heat provided by the fuel cell. It would probably be rational to use this heat either for hot water or, in a world dominated by heat pumps, to increase the temperature of the water in the central heating circuit. This would approximately equalise the two costs for getting energy to the home. (And if the hydrogen is made by the electricity generator at times when power is cheap, as is very likely, the difference disappears).

If the two routes are similar in effective cost, why might we decide to go for the extra complexity of transporting part (or even all) of our energy needs in the form of hydrogen?

The reason is that a hydrogen distribution route running alongside the electricity network provides many other substantial advantages.

a)    Electricity is expensive to store, particularly in large quantities. Although batteries are getting cheaper, they are currently vastly more expensive that storing gas in underground salt caverns, which is the way hydrogen would be kept between seasons

b)    A hydrogen network would also provide storage in its pipes. The UK gas system has over 280,000 km of piping at various pressures. So when electricity was in short supply, perhaps because of a lack of wind, the gas in storage and in the network of pipes could be used to make extra electricity. We can also store months worth of hydrogen gas in underground salt caverns

c)     Keeping the gas network open - rather than eventually closing it when the UK stops using natural gas, which is the current intention - could remove much of the need to increase the transmission capacity of the electricity network to meet increased power demand. This benefit shouldn’t be underestimated. As the country switches to 100% renewable power and electrifies everything in the home, the electricity infrastructure will need upgrading, at a cost of many billions of pounds. The extra requirements will range from new high voltage transmission lines (‘pylons’) to local reinforcement of distribution links to major upgrades for most of the tens of thousands of small transformers dotted around residential areas. Shifting a large part of energy distribution to hydrogen can potentially save almost all of this cost. 

d)    Some offshore wind farms will be able to entirely avoid being connected to the electricity network and could make hydrogen instead. This is already being considered for several North Sea developments in non-UK waters. The turbines will be cheaper to construct and transmission costs will be far lower.

The largest single benefit would be that the inevitable intermittency of wind and solar can be safely and inexpensively accommodated while guaranteeing electricity supplies. If the wind isn’t blowing strongly, as has been the case for much of this winter so far, households would be drawing on hydrogen for electricity supply. Stored in huge salt caverns near to the points of electricity generation, hydrogen would provide a complementary energy source across the country.  This is an absolute requirement for a 100% renewable energy system. 

This week’s announcement of nearly 25 GW of new offshore wind in Scottish waters makes the point very effectively. This additional capacity will mean that renewables supply will now exceed electricity demand for many hours each year. Wind and solar already have to be curtailed on some occasions as electricity supply is greater than what is needed. To be useful, this excess power will have to be converted to hydrogen, perhaps for use in three months’ time. It then makes obvious sense to ship this energy still as hydrogen. Small fuel cells are as efficient as large power stations at converting it to electricity. 

Others have advocated using gas turbines with CCS to provide the necessary backup for when supplies are short. This is probably costlier and, even with good CCS, would still result in significant emissions from natural gas fugitive methane and lost CO2. And this solution doesn’t help at all with the increasing numbers of hours when electricity supply exceeds demand whereas conversion to hydrogen works both ways.

Household demands for power for heat pumps and electric vehicles will double within a decade or so if decarbonisation proceeds as hoped. Converting the existing natural gas distribution networks to hydrogen in order to make electricity close to the end user is worth detailed investigation as to technical feasibility and commercial viability.  

Some quick and tentative answers to some of the obvious questions

 Can we easily convert natural gas pipelines to carry hydrogen? Yes, most of the UK’s high pressure network is already fit to carry hydrogen and a large proportion of the local links. Most of the entire UK network of pipes will be fit for hydrogen by 2030. Compressor size may need to be increased but because less energy will need to be transmitted than currently, the pipelines will not be overloaded.

What will the electrolysers cost? The cost of the electrolysers is not included in the rough numbers offered in the article. Electrolyser prices are falling sharply and will continue to do for decades to come. Recent estimates for large scale installations are around €300 per kilowatt by 2025. At this level the cost of electrolysers will not add significantly to the price of hydrogen, as long as the electrolysers are operating a large fraction of the hours in the year.

Why do we need to convert back to electricity at the home or the local substation? Although gas might be useful for home heating, electric heat pumps are significantly more efficient, offering up to 4 times as much heat for each unit of electricity used. Similarly, homeowners will need electric power to charge the growing number of electric vehicles. So although fuel cells are only about 62% efficient, it remains better to convert hydrogen to power than combust the hydrogen for energy.

What will the fuel cells cost? Ballard, one of the world’s leading fuel cell manufacturers, has published estimates of $100 a kilowatt at large production volumes. The average home might typically need a one or two kilowatt system to complement electricity supply. The actual cost of installing the fuel cell, ensuring safety and including electronics that allow the cell to be controlled by the power company in order to respond to local shortages and surpluses, will be much greater than this figure. But this cost needs to be weighed against the many advantages of a having millions of fuel cells acting as a centrally controlled virtual power plant.

Would this proposal also possibly work in other countries? Yes, particularly in place which will need to transport electricity long distances to match supply and demand. Germany is the obvious example. The country is opposed to the construction of the necessary North/South pylon lines. The existing natural gas network could be used to shift energy instead. Political resistance to new high voltage electricity distribution links increases the attractiveness of using below-ground hydrogen pipes instead.

Would the economics be improved if the electricity used to make hydrogen was purchased at lower than average prices? Yes, the idea becomes more feasible the lower the price of electricity assumed for the manufacture of hydrogen. And it is entirely reasonable to suggest that hydrogen will only be made at times when power is cheap, because that is when electricity is abundant. 

[1] https://www.centrica.com/media/4745/centrica-2020-ofgem-css.pdf. The ‘transport’ charge includes all fees incurred shipping the power or gas from point of origination to point of use.

[2] In addition, of course, electricity prices include the costs of running the company that retails the electricity and the charges imposed to cover subsidies such as Feed-In Tariffs.

The first large-scale attempt to build a 'dispatchable' renewables project

Debate rages as to which applications hydrogen is best suited for. Making ammonia and decarbonising steel-making look obvious choices. Most people say that hydrogen isn’t right for home heating or powering vehicles, with the possible exception of long distance heavy trucks. 

There’s not enough discussion about another vital role of hydrogen: storing excess renewable electricity to avoid the problems of intermittent renewables. But this might be the single most important use by 2050. Trial projects are beginning around the world and one in Guiana, the overseas department of France on the northern coast of South America, is the largest.(1) Construction of this €170m scheme started late last year and opening is planned for 2024.(2)

An impression of what the CEOG project will look like.

The CEOG project will take electricity from a new 55 MW solar farm and distribute at least 10 MW into the local grid during daylight hours (8am to 8pm). Excess power will be stored in 38 MWh of batteries or converted to hydrogen via 16 MW of alkalne electrolysers provided by the French manufacturer McPhy. The hydrogen storage capacity appears to be almost 100 MWh. The electricity in the batteries plus conversion of hydrogen back to electricity via an adapted Ballard fuel cell will feed at least 3 MW into the distribution system for the hours of darkness (8pm to 8am), meaning the hydrogen storage is enough for at least four nights.

The structure of the CEOG scheme (batteries omitted).

Electricité de France (EdF) is the local electricity supplier and has signed a 20 year power purchase agreement guaranteeing the offtake of 10 MW in the day and 3 MW at night for the next 25 years. I think this makes the project by far the largest ‘dispatchable’ renewable electricity scheme in the world. It will provide guaranteed power for about 10,000 homes, allowing 300 watts a household at night, enough for a TV, fridge and some lighting. The scheme will be disconnectable from the wider grid, allowing it to continue to operate even during an electricity failure in the rest of this small and highly forested country. 

This is an expensive project and required substantial help from the French state and support from the European Investment Bank. The price to be paid by EdF for the guaranteed output has not been specified although it is said to be less than the cost of generating electricity by diesel, which would be the only realistic alternative. 

A quick estimate of the underlying economics shows how far from financially viable this scheme might be in mainland Europe. A 55 MW solar farm will produce very approximately an average 20% of its rated output in Guiana. The scheme will therefore generate about 100,000 MWh a year. If the capital cost is €170m or $200m and the project lasts 25 years, then the depreciation is at least $8m a year. The cost of the project per MWh produced is $80, considerable more than the typical (not today) wholesale price of electricity. This is without thinking about the running costs or the re-equipping needed for the electrolysers and fuel cells. 

But this a ‘first of a kind’ project and costs will fall, possibly dramatically, as solar farms, batteries, electrolysers and fuel cells all get cheaper to make.

In my view, this development will be the first of many similar schemes. The main developer behind this innovative project – Hydrogène de France (HDF) -  claims a list of prospects in at least 20 different countries with a total potential value of €3bn. I imagine most will be planned for remote areas currently using oil for power production.  

When completed, the Guiana scheme will be a vital demonstration that renewable electricity supply need not be ‘intermittent’. Adding hydrogen storage allows the operator to offer fully reliable and dispatchable power. We urgently need many copies of this scheme around the world to demonstrate whether or not ‘renewables plus hydrogen’ can be economic.

(1) https://en.ceog.fr (2) https://www.bloomberg.com/press-releases/2021-09-29/hdf-energy-breaks-ground-on-world-s-largest-green-hydrogen-power-project

The ECB doesn't think that the higher cost of renewables is raising energy prices

Professor Isabel Schnabel of the ECB spoke about the inflationary effects of the global push to cease using fossil fuels.[1] She told a US audience that the energy transition might continue to add to prices, and persistent inflation could result in a reduction in ECB bond buying and, partly as a consequence, higher interest rates.

Press coverage of this important speech has sometimes suggested that Schnabel was saying that the cost of renewables was tending to push energy prices up. This mirrors an assertion made elsewhere by the UK government adviser, Sir Dieter Helm.[2] The Financial Times headline was ‘ECB executive warns green energy push will drive inflation higher’.

It seems to be me that this is not what Professor Schnabel is saying. Nowhere does she assert that an energy system based on renewables is more costly. 

 Her argument is very different. Instead, she says that

·      Renewables growth is far too slow to meet Europe’s ambitions for emissions reductions.

·      At the same time, investors - and capital markets more generally - are demanding that fossil fuel exploration and production bear an increased cost of capital. This raises the costs of producing oil, gas and coal. It also reduces the incentives to explore for, and then develop, new sources of fossil energy. This is a necessary part of the transition.

·      And carbon taxes are becoming more prevalent around the world, while the levels of penalty imposed are rising sharply. In addition, the taxes are being applied across a wider range of emissions-producing activities. These taxes increase the cost of buying fossil fuels.

The impact of not having enough renewable electricity at the same time as crunching down on the cheap availability of fossil fuel is to inflate the price of energy. She comments that the effects are particularly acute in the case of natural gas, which is increasingly used in Asia and elsewhere to avoid burning coal in power stations. It is not that Europe has pushed too fast towards an energy system dominated by renewables, it is that it has not been able to move rapidly enough. ‘At present, renewable energy has not yet proven sufficiently scalable to meet rapidly rising demand’.

 In the past, sharp rises in energy costs rapidly corrected themselves. Exploration activity increased and more oil and gas duly appeared in response to the incentive of higher prices. This time is different. Schnabel writes ‘last year’s strong economic expansion, for example, was characterised by an atypically slow response of US shale oil production to rising oil prices, as such investments may no longer prove profitable to investors over the medium term − at least not to the same extent as they have done in the past..’

As a result, it may be that price inflation continues at a high level. Energy price rises were responsible for over a third of overall European inflation in 2021 and this pattern might continue during this year. 

She summarises the position as follows: 

‘As the shift in the energy mix towards cheaper and less carbon-intensive fuels will take time, a rising carbon price, higher tax rates across a range of fossil fuels, and relatively inelastic energy demand may lead to continuous upward pressure on consumer prices in the transition period’

Schnabel suggests that increased rates of inflation in other parts of the economy do not appear likely. Wages pressures remain subdued (though some would strongly argue with this conclusion) and this energy price shock, unlike those in the past, was not driven by an overheating world economy that might produce high levels of general inflation. Unless oil and gas prices continue to rise, inflation rates will therefore fall back before the end of 2022. Using future (‘forward’) prices for oil and gas, she shows that the markets expect fossil fuel prices to fall considerable from the end of this year (though they will still be far higher than in 2020).

The overall tone of the piece is strikingly optimistic in many respects. Although the world is not decarbonising fast enough, Professor Schnabel of the ECB stresses that ‘In our baseline scenario, the current energy shock is expected to fade over the projection horizon’. There are risks of continued inflation of prices, but they do not dominate the Bank’s thinking.

She asserts that the policies causing the price inflation are helping the decarbonisation drive and need to be continued.

‘In other words, even in the absence of a global carbon price, which remains essential, there are growing signs that the green transition is accelerating around the globe’.

And that carbon taxes are probably effective at increasing economic activity.

‘An emerging strand of empirical evidence finds no robust negative effects of carbon taxes on GDP growth and employment. If anything, the evidence is consistent with a modest positive impact’.

There is no reason, she says, to change energy policies. Although the pain of higher prices is substantial, particularly for the less well off, slowing the pace of the energy transition is not a good response to the current crisis. 

So what should governments do to ease the position of those people spending a large fraction of their income on energy? Schnabel supports either direct lump-sum payments to households or a reduction in taxes on employment. A lump-sum transfer in European countries, funded by the higher carbon tax revenue, could ‘largely cushion’ the cost of more expensive fuels.

There is no reasonable case for reducing VAT on energy, a policy currently being advocated in several European countries. This would simply increase the demand for oil and gas, prolonging the period of an inflationary mismatch between demand and supply. And we won’t solve the global warming problem by making fossil fuels cheaper. 

To conclude, nothing in Professor Schnabel’s remarks suggests any ambiguity about the need to continue pushing forward with the energy transition.

In fact, she argues the reverse. It is only by increasing the speed of decarbonisation, partly by the use of a global carbon tax, that long-run price stability will be encouraged. And the eventual result, in her words, wil be cheaper energy. However this goal must be accompanied by a commitment that temporary rises in oil and gas prices will not affect the living standards of the less well-off. 

[1] https://www.ecb.europa.eu/press/key/date/2022/html/ecb.sp220108~0425a24eb7.en.html

[2] http://www.dieterhelm.co.uk/energy/energy/luck-is-not-an-energy-policy-the-cost-of-energy-the-price-cap-and-what-to-do-about-it-2/

Use ammonia for shipping, synthetic hydrocarbons for aviation

Long distance transport over sea and air cannot be electrified in the foreseeable future. The energy density of batteries is too low to sustain movements of more than a few tens of kilometres. Fuel cell airplanes are possible, but are unlikely to operate successfully on journeys of more than an hour. Similarly some short distance sea travel can use batteries. Ferries are good examples. But most marine transport will need an alternative fuel.

This article argues that ammonia will be energy provider for much of global shipping but that aviation needs synthetic fuels made from hydrogen and captured CO2. 

Why? Ammonia is not energy dense enough for aviation, and raw hydrogen would use too much space. But for shipping, where space and weight used by the fuel are not important constraints, ammonia will have a more important role. It may well be cheaper than liquid hydrogen and certainly less costly than synthetic fuels. Nevertheless, synthetic methanol may challenge ammonia, partly because the transition away from fossil oil will be easier to manage.

Aviation.

Decarbonisation of long-distance air travel is not yet widely discussed, perhaps because it seems too difficult. Airbus has published some details of ‘concept’ aircraft designed to run on hydrogen.[1] Reaction Engines, a start-up from the space technology cluster around Oxford, recently announced a joint venture to build reactors that turn ammonia back into hydrogen for use on board airplanes and for other applications, including rockets.[2]

Neither manufacturers nor airline have detailed plans for decarbonisation although ‘Sustainable Aviation Fuel’ (SAF) is often mentioned. SAF is assumed to be made from biomass or waste but, as stated in the previous article on this site, nobody contends that more than a small fraction of total aviation fuel can be made in this way. 

Broadly speaking, the real options for full decarbonisation of long-distance aviation are liquid hydrogen, ammonia or synthetic fuels. My belief is that although synthetic aviation fuels are likely to be more expensive than fossil energy for decades they are nevertheless the logical way to decarbonise aviation. 

Ammonia is disadvantaged by its ratio of energy to weight. 

Ammonia                                  5.2 MWh per tonne

Conventional aviation fuel     11.9 MWh per tonne

So to provide the same amount of power to take off, cruise and then land would take over twice as much weight as today’s aviation fuel (and therefore also more than twice as much weight as an identical synthetic kerosene made from CO2 and H2). 

Does this matter? Yes, a lot. Please take a look at the numbers below.

                                                Empty weight             Max. fuel weigh

Airbus A380 – 900                  277 tonnes                  254 tonnes

Boeing 737 -900                      45 tonnes                  24 tonnes

For the Airbus 380 long distance aircraft, the amount of fuel that can be taken on board is almost the same as the empty weight of the aircraft. If it were ammonia, the energy value of this fuel would be less than half that of aviation kerosene. The maximum distance travelled would therefore be considerably less than half what it is with today’s fuel. (Why ‘considerably less’? Because takeoff consumes much of the fuel for a heavy aircraft, and this is the same whether the journey is 1 hour or 10 hours).

Before the pandemic, the longest scheduled Airbus A380 flight was just over 14,000 km (Dubai – Auckland).[3] With ammonia as the fuel, this aircraft would probably have only been able to travel between New York and London (5,600 km).

Would this disadvantage be outweighed by ripping out a few seats and reducing the number of passengers? No – when the Airbus 380 is full of fuel the weight per passenger of the aviation kerosene is around 400 kg. It doesn’t use that each flight but even a single London-New York flight might use around 150 kg per passenger. The industry assumption is that the typical passenger weighs just under 80 kg. So cutting the number of passengers will not solve the weight problem. 

Liquid hydrogen faces a different issue. Although it is more energy dense than aviation fuel in terms of kwh per kg, it requires about four times as much space to store a megawatt hour of energy. The airplanes that used it would have to have a very different shape and be completely re-engineered in other ways. Airbus has shown us photographs of how the planes might look, but the process of designing and producing these aircraft in large numbers will take decades. We also shouldn’t underestimate how expensive it will be to create liquid hydrogen. Not only is the energy cost of liquefaction approximately one third of the energy value of the H2, but the cost of the processing plant is also very high. One recent estimate from the US government is $800m for a 70,000 tonnes a year facility.[4]

We need to start decarbonisation now and synthetic aviation fuels are by far the best option.  Be wary of those who push either ammonia or liquid hydrogen.

Shipping

Ammonia will need far more storage space than fuel oil, and it will be difficult to handle and safely store. But it is also likely to be cheaper than synthetic fuel, or liquid hydrogen. This should be enough to ensure its eventual dominance as the core fuel for long distance shipping. 

I assume that the cost of the hydrogen necessary to make the ammonia is $1.50 a kg, a level likely to be reached within a decade or so in sun-rich countries. Simple projections of the cost of turning H2 and nitrogen into NH3 (ammonia) suggest an eventual price of around $60 per megawatt hour. At a price for heavy fuel oil or around $730 a tonne, conventional fuel has a cost of around $56 a megawatt hour. So ammonia is eventually likely to be about the same price as fuel oil.  But today it is likely to be at least twice the price at most ports around the world.

Synthetic direct substitutes for conventional oils are always to be more expensive per megawatt hour than ammonia, whatever the price of hydrogen. This is because green ammonia doesn’t require the capture of carbon dioxide and its processing into carbon monoxide in the way that zero-carbon synthetic fuels do.

Many shipowners ordering vessels now, and looking for a low-carbon fuel, see ammonia as the logical energy source (although relatively few orders have been made). But is synthetic methanol a viable alternative? Methanol (CH3OH) has a higher ratio of hydrogen to carbon than kerosene and therefore may offer a cheaper route for manufacturing because less CO2 has to be captured to make an equivalent quantity of fuel.  However, at whatever price of hydrogen we assume, methanol will probably be more expensive to make than ammonia.

That may not stop shipping businesses purchasing methanol vessels. After the announcement in the late summer that Maersk would buy 8 large dual-fuel methanol container ships, the Singapore shipping company Xpress Feeders announced this week that it would also commission 8 new smaller ships in 2023/24. 

The big advantages of methanol include fewer safety concerns and much easier storage and bunkering. It also has a greater energy density per cubic metre than ammonia, although the difference is not huge and methanol will require more space in the fuel tank than oil. Perhaps most importantly, a dual-fuel methanol/fuel oil vessel would be able to switch back to heavy fuel oil if methanol was not available.

Conclusion.

 Transport is often described as the core market for hydrogen. But we need to be somewhat nuanced about this conclusion. In the case of the most important market, batteries may be more suitable for heavy long distance road freight than we current assume. Shipping is likely to shift to ammonia and aviation will move to synthetic equivalents to aviation kerosene. 

[1] https://www.airbus.com/en/innovation/zero-emission/hydrogen/zeroe

[2] https://www.reactionengines.co.uk/news/news/press-release-joining-forces-deliver-world-leading-decarbonisation-technology

[3] https://simpleflying.com/longest-airbus-a380-flight/

[4] https://www.hydrogen.energy.gov/pdfs/19001_hydrogen_liquefaction_costs.pdf

How much will electrofuels cost?

The last few months have seen advances in the production of ‘electrofuels’, or liquid fuels made from green hydrogen and captured CO2 that can replace products made from fossil oil. 

·      Infinium, a Californian start-up raised $69m from Amazon and NextEra Energy to build manufacturing facilities that will make about 160 million litres a year of near-zero carbon liquid fuels.

·      Prometheus Fuels, also in California, raised money from Maersk. 

·      Norway’s Nordic Electrofuels recently signed an agreement to take waste CO2 from a metals processor before building a 10 million litres a year plant. 

·      In the last few days Zero Petroleum, a UK start-up, provided the fuel for the first ever flight using 100% renewable aviation gasoline anywhere in the world.  

I look briefly below at how low the prices of hydrogen and carbon dioxide will have to be for electrofuels to be competitive with oil. 

The level of interest today in electrofuels (‘efuels’, ‘synthetic fuels’) remains low. There is, for example, absolutely no commentary in UK media or assessment in public policy documents. Even though these fuels can decarbonise difficult sectors, particularly aviation, understanding of the potential is very limited.

Many confuse efuels with conventional biofuels, particularly in the context of aviation. This is unfortunate; electrofuels can be made in unlimited quantities without needing agricultural land. They do not result in major greenhouse gas emissions, which biofuels generally do. 

The reason for the lack of interest seems to be that electrofuels are assumed to be extremely expensive and unlikely ever to be competitive with liquid fossil fuels. This is a reasonable concern; at today’s costs electrofuels will be far more expensive to make than using fossil below. This will not always be the case.

In the table below, I write down the cost of the hydrogen and the CO2 that is necessary to make a tonne of C11H24, which I have used the representative hydrocarbon in aviation fuel.[1] I have not included any calculations of the cost to process the source chemicals into this hydrocarbon but this number will probably not exceed $200 per tonne of C11H24 and will not change the basic conclusion. (The details of the equations I have used are in the appendix to this note).

How the costs of H2 and CO2 affect the prospective price of aviation electrofuel

The calculations are described in the Appendix below.

What do these numbers show? In the left hand column, I suggest a price of $1.00 of hydrogen and $100 a tonne of CO2. These input price result in a total cost for a tonne of C11H24 of $746. On the right, costs of $2.50 and $250 result in an input cost per tonne of product of $1,865. 

Are these numbers high or low? The current price of a tonne of aviation fuel is about $750 in the US. So in order for electrofuels to be competitive, the price of hydrogen needs to be about $1.00 per kilogramme and CO2 about $100 a tonne. 

These are very demanding targets but - for example - the US government’s Earthshot programme is pushing for $1.00/kg hydrogen and $100/tonne CO2 by 2030. Direct capture of CO2 will eventually cost, according to leading industry player Carbon Engineering less than $100 a tonne at scale. H2 cost may well fall to $1.00 a kilogramme within five years in the sunniest locations where PV electricity is already very cheap.

Would a carbon tax on fossil fuels make much difference? Combusting a tonne of C11H24 will result in production of about 3.1 tonnes of CO2. A $100 a tonne carbon tax would therefore add about $310 to the cost of aviation fuel, taking it to approximately $1060, or about 5% below the electrofuels route at an H2 price of $1.50 and a CO2 price of £150.

In summary, at today’s costs for hydrogen and carbon dioxide, electrofuels would probably be at least five times the price of oil products, even when made in large quantities. But prices will fall as green hydrogen comes down in price and direct air capture of CO2 increases in scale. Even if electrofuels look expensive today, the logic of developing them is strong since they will work in existing applications such as aviation engines. Alternatives, such as pure hydrogen or ammonia require huge investment of time and money in designing, building and testing new aircraft and redesigned engines. Electrofuels may become the dominant source of energy for flying, even with today’s costs.

Appendix

A liquid fuel, such as aviation kerosene, consists of a mixture of hydrocarbons. A hydrocarbon is a molecule composed entirely of hydrogen and carbon atoms. 

It is possible to manufacture hydrocarbons in chemical processes that result in no net emissions to the atmosphere after combustion. Put simply, all we need do is generate some hydrogen in an electrolyser, collect some CO2 from the atmosphere, turn it into carbon monoxide and then react the carbon and the H2 to make a hydrocarbon(s). The technology is well understood. The Fischer Tropsch process using this approach and is over 100 years old.

Making synthetic aviation fuel from zero carbon sources

Aviation fuel is a complex mixture of chemical compounds. Kerosene is the most important group of hydrocarbons in the mix. To make kerosene from hydrogen and captured CO2 requires two reactions.

1, The CO2 needs to be turned into carbon monoxide (CO). This is done using what is called the reverse water gas shift reaction. 

CO2 + H2 -> CO + H20

This reaction requires hydrogen to be added. Not all the hydrogen that is added is captured in each run through the reactor but eventually it will be reacted with incoming carbon dioxide. The molecular weight of CO2 is 44 and that of H2 is 2. So for every tonne of CO2 converted to CO, 2/46ths of H2 will be needed. 

CO has a molecular weight of 28. So a tonne of CO will require 44/28ths tonnes of CO2 as input and 44/28 * 2/46 tonnes of hydrogen.

2, The CO needs to be reacted with hydrogen to make a kerosene-like molecules. I have chosen C11H24 as the representative hydrocarbon. (The Fischer Tropsch reaction that turns CO and H2 into hydrocarbons produces many different molecules). The reaction is 

23H2 + 11CO -> C11H24 + 11H2O

 The total molecular weight of inputs is 354, including 46 for the hydrogen required. The weight of the useful output (C11H24) is 156. So to make a tonne of aviation fuel requires 46/156 tonnes of hydrogen and 110/156 tonnes of carbon monoxide.

3, Total inputs. I calculate that to make a tonne of C11H24, the refinery will require 3.10 tonnes of CO2 and 0.44 tonnes of hydrogen. (The extra 2.54 tonnes is lost as water, H20).

[1] Aviation kerosene includes a wide variety of hydrocarbons and other molecules. I have used C11H24 as a typical ingredient.

The IEA finally starts to believe in the energy transition. A note from Kingsmill Bond

Kingsmill Bond, Energy Strategist at Carbon Tracker, is one of the very best analysts of the energy transition. (Listen to a podcast here with David Roberts of Volts). With his permission, I have attached below the text of an email he sent out this morning (October 13th 2021) after the IEA press conference.

The IEA press conference this morning was a tour de force, and well worth watching even the first 20 minutes. In the same way as it is much more profitable to buy an unloved stock where management turns it around, the impact of the IEA’s analysis is all the more powerful because they have moved over the course of the last couple of years from supporting continuity of the fossil fuel system to embracing renewables. And why not - when the facts change, you change your mind after all; I look forward to a similar change of view from those who followed the old IEA approach.

https://www.iea.org/events/world-energy-outlook-2021

So a few points from the press conference by Fatih Birol, Laura Cozzi and Tim Gould. The new heroes of the energy transition...

* We are moving to a new energy economy.

* It will be better than the fossil fuel economy. Cheaper, cleaner, fairer, more resilient, safer. With higher GDP growth.

* This is a key turning point in human history. Rising energy demand but falling fossil fuel demand is possible.

* Even if we don’t hit all the climate goals, the fossil fuel sector will be radically disrupted.

* This is the decade of disruptions. Coal demand has already peaked and oil and gas demand will likely peak by 2025. 2019 thus was the peak in fossil fuel demand, and we are bouncing along the plateau.

* There is a lot that we can do at no economic cost to reduce emissions this decade. By 6Gt extra in total. And we can deploy 800 GW of new solar and wind at no economic cost as they are cheaper than the fossil fuel alternative.

* A new geopolitics of energy will emerge as trade shifts from fossil fuels to minerals and hydrogen.

* New energy is a huge market opportunity, with demand increasing tenfold to over $1tn and a new opportunity bigger than oil today.

* Since Paris, we have reduced expected global warming in 2100 from 3.6 degrees to 2.6. We now need to use Glasgow to get it down to 2.1 degrees and ratchet lower. At Glasgow policymakers need to send the world a message: we will do this. If they do that, solutions will happen faster.

* We have to triple clean energy investments and get them out into emerging markets.

* If you continue to invest in dirty energy you may well lose money. You will make profits from the clean energy economy.

* There will be volatility at the top. But we need to plan for it and to build renewables quicker. Governments should have ‘people’ strategies as well as hydrogen strategies, to manage change.

* Technology costs of new energy will keep falling and this will help to solve the hard to solve sectors.

* The high fossil fuel prices in the last few months are caused by the demand bounceback in the middle of continuing COVID disruptions. Not by renewables. We need to build more renewables to solve this issue.

__

It is also worth highlighting some of the many old fossil fuel arguments that they have just bankrupted. With the ‘new’ IEA answer in bold.

__

The old fossil fuel arguments

*Energy transition means degrowth. No – it means higher growth

*Decline rates mean you have to keep investing in oil and gas. No - you need some maintenance capex but no new fields

*Peak demand for fossil fuels is highly unlikely. No – peak demand is now the central scenario.

* It will take ages for fossil fuels to be disrupted because they are so big. No – this is the disruption decade.

* Energy transition is not just. No – it is fairer, cheaper, local, and cleaner.

* Inertia will stop change. No – we have already bent down the curve of fossil fuel demand from rapid growth to no growth, and decline is in our sights

* Renewables are tiny and irrelevant. No – they are big enough to disrupt incumbents

* Renewables can’t grow rapidly. Yes they can and this is the central scenario

* Energy transition is expensive. No – it is much less expensive than inaction. And 40% of the necessary actions to speed up change have no economic cost. Household energy costs will be cheaper if we transition than if we do not.

* Fossil fuel demand always rises by some special law of history. Not any more. This is a historic turning point.

* We don’t have enough minerals to do this. Yes we do. They even worked out how big trade will be in 2050

* There is not enough land. No - land availability is no constraint. It just means trade in hydrogen.

* Intermittency stops change. No - this is soluble if you increase flexibility. They even worked out flexibility solutions for 2050 in detail.

* Hard to solve sectors stop change. No – change happens anyway. Hard to solve sectors are simply the last area to change. And anyway, falling renewables costs will make all this easier

* There will be job losses from the transition. No there will be net job gains of from 13 to 24 million. And 75% of them will be local.

* CCS will enable us to continue with business as usual. No – CCS is for the last 10% of emissions or so. You have to curb fossil fuel demand and supply.

* EVs are tiny. No – they are big enough to drive a peak in oil demand by 2025.

* Electricity is too small to matter. No – electricity is the key driver of change as it increases from 20% of final consumption to 50%. In any event, final consumption is a misleading metric because electricity is a much more efficient technology than fossil fuels.

* We don’t have all the technologies we need for a transition by 2050. Of course not, we are in 2021. But we do have everything we need for this decade.

* The gilets jaunes will stop the change. No – the benefits of change for the people are overwhelming. Governments need to figure out people centred strategies to find solutions for those at risk from change.

__

Not all will be rosy. There are some import and accurate caveats they make to their newfound optimism:

* This is not going to be easy. Because of inertia and incumbency and so on.

* There will be volatility at the top. There will be some losers from change. We need to help them out or they will seek to block it.

* We need to increase the capital flowing into renewables. (I see this as a policy issue not a capital issue)

It is of course possible to criticise the analysis; falling casts mean that change is likely to be cheaper and easier and so on. But this is still a big step forward.

‘Gradually, and then suddenly’; the unsurprising rise in EV sales

‘Is this an inflection point?’ asks Andrew Bergbaum, a managing director at AlixPartners. ‘I think the answer has to be yes.’

That was one of the comments in a recent article in the Financial Times (paywall) on the growth of EV sales, particularly in the UK and Europe. The tone of the article was one of astonishment; the author breathlessly noted the ‘extraordinary surge in demand’.

The really extraordinary thing is that the ‘inflection point’, at least in the UK, was actually over two years ago. It was then that the share of EVs (both pure battery and plug-in hybrid) began to accelerate sharply. But at the time none of us noticed. 

Yes, we knew that EV sales were growing and would probably eventually command a majority share of sales. However, as usual, we failed to understand that exponential change happens ‘gradually, and then suddenly’.[1]

Here’s a chart that shows the percentage of all car sales held by electric vehicles in the UK.  The graph doesn’t plot the numbers month by month but uses a twelve month rolling average. The data series that I compiled starts in October 2015, so the first twelve month period begins in September 2016.[2]

Source: SMMT

Between the beginning of the data series - when sales were running at around 1% share of all vehicles -to July 2019 sales increased at around 2% a month, or 27% a year. As a result, in July 2019 the share was around 2.6%. 

The EV market share the jumped sharply in August 2019 and the acceleration continued. From then on, market share has typically increased by over 7% a month.[3] Over a year, this means that the share rises over 130% a year. We’re noticing that effect now, but the trend has actually been there to see for over two years.

Of course the monthly figures have swung around, reflecting new product introductions, shortages and changes in tax incentives. But the underlying rate of rolling twelve month increases has been surprisingly stable. Nevertheless, the Financial Times article characterised the developments as follows. ‘A slow burning shift in the way the world works suddenly starts to gather pace at a rapid rate’.

Hmm, I say, this shift has actually been burning like wildfire for 25 months. And if we’d plotted the data this time last year and extrapolated the growth to September 2021, we’d have come very close to the actual sales share held by EVs last month.

Humans don’t notice fast changes when the absolute amount of variation is small. A rise of from 1% to 2% market share is far less noticeable to us than an increase from 5% to 10%, although the value of spotting the signal is far greater for the smaller numbers.

This failure in our understanding is tending to impede many of our responses to changes that the energy transition will produce. So, for example, it is only in the last few months that most of the world’s car industry has finally realised that the market demand for the components of internal combustion engines will fall very sharply in the next few years. It would have been better for all of us, but particularly employees in related manufacturing businesses, if planning had started two years ago, not today.

Exponential growth really shouldn’t surprise us any more after what we have seen in solar PV, offshore wind, electrolyser and battery costs.

[1] A quote from Ernest Heminway’s The Sun Also Rises’.

[2] I have adjusted the data from two months at the very beginning of the pandemic. In these months the share of EVs was exceptionally high – at 34% in the month of April 2020 - and I interpolated the shares between March and June. 

[3] This means that if the EV share of all personal vehicle sales is 10% one month, it is expected to be 10.7% in the following month. 

More on 'renewables plus hydrogen'

In the last piece on this website I argued that the most effective means of completely decarbonising electricity supply in the UK and elsewhere is to hugely expand the capacity of wind and solar power. This will mean frequent substantial surpluses of electricity when the unneeded power will be converted to hydrogen via electrolysis. This hydrogen can then be used an energy source for use in combined cycle gas turbine plants during periods of electricity deficit. 

I suggested that the UK needed approximately 4.5 times its existing resources of wind and solar power to carry out this strategy. These resources would provide enough electricity to cover all electricity needs, including in periods of deficit. The assumptions behind this number are overly simply but gave a good indication of how much new wind and solar would be needed. I also said that the country would have needed about 75 GW of electrolysers to ensure that all electricity generated would be used. 

This brief article now looks in a little more detail at the requirements for electrolysers. In particular, I ask the question ‘how much more electricity generating capacity would be needed if we restricted the availability of electrolysers to much lower levels?’. The logic is this: 75 GW would only be needed for a few half hour periods a year so if we had less capacity, we would lose relatively small amounts of hydrogen. We would have to make up this loss by installing more wind and solar but this might be less costly than installing huge amounts of electrolyser plant.

I also made one important change to the spreadsheet. I previously assumed that wind and solar would produce all the the electricity needed in each of the 17,520 half periods of the year to 30th June 2021. This ignored other low carbon sources. So for the exercise covered in this article I included nuclear power, biomass and hydro as low carbon sources. In each period I therefore took the total demand figure and deducted the amounts of power provided by these three types of generator. This reduces the electricity required from wind and solar. 

The change substantially cuts the additional capacity required from these sources of power. Instead of needed 4.5 times as much solar and wind as today, the UK will only need a 3.4 times multiple. As a consequence, the maximum demand for electrolyser capacity falls from 75 GW to around 55 GW. To be clear, a 55 GW need arises because in at least one half hour period in the year under analysis, a 3.4 times multiple of wind and solar capacity would have resulted in 55 GW of unneeded electricity that would have been available for conversion into hydrogen.

But is it worth investing in as much electrolyser plant? If much of the capacity is only used a few hours a year, would it be better to install a smaller capacity for making hydrogen? This would mean that some electricity would be wasted at times of high production. However this could be compensated for by increases in solar and wind capacity that would produce more power in periods of lower surpluses. 

The chart below shows how what multiple of current wind and solar power would be needed to compensate for reduced electrolyser capacity. This exercise shows that partially restricting electrolyser availability has little impact because the number of hours of very substantial electricity surpluses were few in number across the year under study The addition requirements for renewables are very small indeed until electrolyser installations fall to less than half the maximum 55 GW total.

I then compared the costs of installing 55 GW of electrolyser and a 3.4 times multiple of current wind and solar with an alternative of 20 GW electrolyser capacity and 3.59 times these renewables. 

I think the approximate number suggest that the reduction in required electrolyser capacity reduces costs by about twice as much as the increased need for investment in renewables. 

The conclusion seems clear. It would be much better for the UK to go for 20 GW of electrolysers (or perhaps less) and a 3.59 times multiple of wind and solar. In addition, there is further smaller advantage. A 3.59 multiple, rather than 3.4, means that in periods of deficit the requirement for combined cycle gas turbine capacity to turn the hydrogen back into power would be slightly less.

In addition, it is worth pointing out that the total amount of electricity generation necessary to cover the UK’s needs in the year to June 2021 would have been about 1.25 times eventual demand, after taking into account the energy losses from the conversion into hydrogen and back again. Let’s mention one of the implications of this. If new renewables can now be commissioned by commercial developers at prices of £40 or less, the full electricity cost of 1.25 times means an underlying wholesale price of electricity of £50 (before the costs of electrolysers and CCGT). ‘Renewables plus hydrogen’ can work effectively as the key decarbonisation weapon, and the cost is less than any alternative.

‘Renewables plus hydrogen’: a way to push fossil fuels out of electricity supply

What is best route to reduce the UK’s reliance on gas and fossil fuels for electricity generation? 

This note provides an outline of one approach, using real data from the last year.  The analysis shows that all the UK’s current electricity demand could be met by an expansion of onshore wind, offshore wind and solar PV to about 4.5 times current levels. 

Crucially, half hourly matching of supply and demand across the year is carried out by hydrogen. When electricity is in surplus, electrolysers are used to make hydrogen. When in deficit, the hydrogen is burnt in gas turbines to generate electricity. The chart below illustrates this for 25th June this year when wind and sun would have been converted to hydrogen during the day but hydrogen would then have been needed in the night hours.

In this simple model fossil fuels are never needed to supplement supply. The inevitable energy losses in the conversion processes mean that more electricity is needed to cover total needs over the year but the excess production is less than 25% of the total. Complete electricity independence would thus come from having about 50-60 gigawatts of each of solar PV, onshore wind and offshore wind, up from about 11-14 gigawatts each today.

The note only looks at replacing fossil fuels in the current electricity supply. As decarbonisation proceeds, more activities – particularly transport and building heating – will be switched to electricity. This will increase the demand for power but this can also be handled by an expansion in renewables, as long as a storage medium such as hydrogen is used to balance supply and demand.

The analysis

1, The data used in the article was provided by Drax Electric Insights for the 365 days from 1st July 2020 to 30th June 2021. (Thank you particularly to Iain Staffell of Imperial College for giving me access to the database). Electricity demand and supply is logged for each half hour period. Electricity supply is split by type of generator, including fossil fuels, nuclear, renewables, storage and international connections.

2, I have added extra columns to the spreadsheet that allow me to multiply wind and solar output in each line by a multiple that can be varied. Thus if solar output is 5 GW in one half hour period and I use a multiple of 2, the column changes the output to 10 GW. I can use different multipliers for onshore, offshore and solar.

 3, Sometimes the multiplied estimated output exceeds the electricity demand for that half hour. Sometimes, even after applying the multiple on all three sources of electricity, demand still exceeds supply.

 4, In those periods when there is a surplus from renewables, the spreadsheet diverts that surplus into making hydrogen. The conversion efficiency is assumed to be 70%. That is, the energy value of the hydrogen that is generated by the electrolyser is 70% of the electricity input. 10 MWh in to the electrolyser creates 7 MWh of hydrogen (lower heating value). This is slightly higher than standard PEM or alkaline electrolysers today but will almost certainly be achieved within the next two years. The hydrogen is assumed to be stored.

5, In periods of deficit when renewables output doesn’t cover power demand, hydrogen is assumed to be taken from storage and burnt in a combined cycle gas turbine (CCGT), much as natural gas is today. Many of today’s large gas turbines can be converted to burning hydrogen, say the market leaders, GE, Siemens and Mitsubishi. The conversion efficiency is assumed to be 60%, roughly the performance of modern turbines. For every 10 MWh of hydrogen burnt by the gas turbine, 6 MWh of electricity is generated.

6, The process of generating hydrogen is 70% efficient and then turning back into electricity preserves of 60% of the energy value. This means that any period of deficit will require electricity generation that is only 42% efficient (70% times 60%). 10 MWh of electricity turned into hydrogen in a half hour of surplus will at some point be converted back into only 4.2 MWh of power at a time of deficit. 

7, Simple calculations allowed me to work out how much electricity output from renewables needs to be expanded so that all electricity needs over the course of the July 2020 to June 2021 would be met by renewables with hydrogen storage. The UK’s electricity requirements could be met by many different combinations of offshore wind, onshore wind and solar. I found that the most ‘efficient’ mix of sources was to use approximately the same multiples of each of the three existing sources. By ‘efficient’ I mean the combination that requires the least amount of renewable electricity to be converted in hydrogen for storage. This mixture of sources therefore most closely matches the electricity required over the 17,520 half hour periods each year. 

The results

8, A combination of 4.54 times expansion of onshore, offshore and solar produces a total of 678,913,290 MW, for half hour periods, or about 339.5 TWh. The actual demand in the year under study was 276.8 TWh, a lower figure than usual because of the impact of Covid. So the electricity system would have produced about 63 TWh more power than would have been generated in a fossil fuel network. This extra amount is what it takes to make the hydrogen in periods of surplus and use it in deficit half hours.

9, With this mixture of resources, the electricity system would have been in surplus for about 10,300 half hour periods and in deficit for about 7,220. The system would have required electrolyser capacity of about 75 gigawatts if the target were to ensure all electricity generated was used. In reality, if it is only to be used a few hours a year this would never be a sensible electrolyser capacity to install. In this case, just 30 GW of electrolyser capacity would have captured all but about 15 TWh of the surplus, or about 25% of the total surplus generated. It probably would be cheaper to overbuild the renewables rather than increase electrolyser capacity.

10, Other combinations of renewable sources usually require more electricity to be generated to cover the periods of deficit. In other cases, when one renewable is expanded more and another less, the match between demand and supply is less good, requiring more storage in the form of hydrogen to meet periods of deficit. However the differences are not enormous.  For example, a system which multiplied onshore wind by 4, onshore wind by 2 and solar by 13.5 would have covered overall demand at the expense of generating about 352 TWh, rather than 339.5 mentioned in paragraph 8. 

11, The only route of a reliance on fossil fuels is to base the energy system on renewables. The transition will be painful and expensive but will eventually result in lower costs and far greater less exposure to geopolitics. The current crisis of gas supply should oblige us to begin to take the difficult steps towards complete decarbonisation.

Appendix

What does a 4.54 times expansion mean?

 1, Solar would move from 13.1 GW to 59.5 GW capacity

2, Onshore wind from 13.6 GW to 62.1 GW capacity

3, Offshore wind from 10.7 GW to 48.7 GW capacity.

The UK land and sea space is comfortably capable of accommodating this increase.

The issues with the analysis

1, The spreadsheet manipulation is overly simple. It assumes that all electricity supply comes from the three renewable sources. Other non-fossil sources (principally imports, nuclear and biomass) are excluded from the analysis. If they were included, the required multiple for renewables expansion would be lower. 

2, It assumes constant capacity of solar and wind during the period. In fact, renewables capacity rose by just over 1% during the year. I use the figures for the end of the year. This will have marginally increased the multiple of new capacity needed.

3, I don’t take imports and exports into account, not least because shortages and surpluses are likely to be continent wide. When it is windy in the UK, it is very likely indeed to be windy in northern Germany and Denmark, for example.

4, My assumptions are also conservative in assuming that future renewables capacity is no more productive than at present. In fact offshore in wind in particular is improving in its capacity factors as the size and height of turbines increases.

5, I haven’t taken into account any storage costs for hydrogen. Nor have I modelled whether it would be more efficient, for example, to install multiple gigawatt hours of batteries to act as the first reserve in periods of excess or deficit. Hydrogen makes sense when storage is over longer periods but batteries work well as the storage medium for intra-day periods. (Storage losses with batteries are likely to be less than 10%, compared to almost 60% with hydrogen).

Follow on work

1, This programme needs to be costed. Because of the conversion losses from electricity to hydrogen and then back, more electricity needs to be generated than is apparently needed. Does this make the proposed route to self-sufficiency costlier than alternatives? I don’t think so but this topic needs more research. This note suggests that from an energy supply viewpoint it makes good sense to expand solar, offshore and onshore by roughly the same multiple. But does this make sense financially? Might it be better, for example, to focus on adding solar PV because the Levelised Cost of Energy (LCOE) is probably lower than offshore wind at the moment?

2, It would be best to do much more sophisticated modelling of the impact of existing and future nuclear power, hydro and pumped storage, biomass and imports. In addition, it would worth assessing the potential effect of demand shifting during the course of the day. But this could only reduce the total need for new wind and solar. 

3, Critically we should also be assessing the impact of the future expansion of electricity demand due to heat pumps and transport electrification. This is complex; we would need to overlay temperatures to assess the need for electric heating. Much demand for electricity in transport can be deferred for hours, if not days. Modelling this would be very difficult indeed.

4, We also need to accurately model exactly what the capacity factor of new installations is likely to be, and exactly when the output will occur. Much solar PV is on household roofs not facing due south. But a very large percentage of new solar is likely to be in south facing solar fields so the total amount of power generated by 1 gigawatt of PV will be greater, but it will be concentrated in the hours around midday. We should also factor in the growth of large scale battery capacity. 

Chris Goodall

chris@carboncommentary.com

September 23rd 2021

Maersk's methanol fuelled container ships

Why does Maersk’s decision to buy eight new ships matter? After all, the company owns around 700 already. 

The reason is that the world’s largest shipping company is committing substantially more than a billion dollars to a set of new ships that may dramatically change the prospects for green methanol production around the world. It has ordered the vessels with engines that can either burn methanol or heavy fuel oil.

The economics editor of the Guardian described the Maersk decision as ‘a drop in the ocean’.[1] He was wrong: this is the clearest sign yet that a swing to decarbonised shipping can conceivably happen within twenty years. Methanol isn’t new as a fuel – and few doubt its feasibility for shipping – but low carbon manufacture of this simple chemical is not well-established. By providing an outlet for zero-carbon methanol production, Maersk will transform its production. And if the methanol doesn’t arrive on time, the company will be able to persist with using fuel oil in the engines. 

A Maersk feeder container ship similar to the one ordered in February of this year with dual fuel (methanol/oil) capacity.

In March of this year a shipping magazine carried a quotation from a shipping analyst on why the major carriers were continuing to buy container ships powered by heavy fuel oil even though decarbonisation targets were inevitable.[2]‘There is no alternative’, he said. ‘Methanol ships are in development. There are trials with ammonia. There are studies of ships with hydrogen. None of these ships are orderable yet’.

 Five months later, Maersk has shown that this conclusion was too pessimistic. It has chosen Hyundai in South Korea to build the dual fuel ships, probably powered by MAN engines. Delivery will start in 2024. This follows a commitment earlier in the month to source renewable methanol from a Danish supplier for the first - and much smaller - dual fuel vessel it ordered earlier this year.

Maersk’s principal gambles are as follows:

·      About 20 dual methanol/oil ships are on the waters today. The earliest – a Stena Line ferry – entered service in 2015. But the existing vessels are far smaller than Maersk’s proposed new ships. The largest today are equivalent to a vessel carrying 2,000 standard containers. Maersk’s order is for ships able to transport 16,000 over long distances. The engine technology will need to be different. MAN says this is possible.

·      Methanol is available as a bunker fuel at many of the world’s largest ports. But if Maersk is to only use green methanol it will have to ensure that this fuel is widely distributed around the world’s major termini.

·      Will green methanol actually be available in sufficient quantities? About 200,000 tonnes a year are produced today, but very little would meet the company’s requirements for ‘zero-carbon’ synthetic fuels or for bio-methanol made from forest or field wastes. 

·      Green methanol is likely to be much more expensive than conventional bunker fuel, which is cheap partly because only ships can burn the least valuable output of oil refineries. Will customers pay the premium?

Dual fuel ships in the context of the Maersk portfolio 

Maersk sails about 700 ships, mostly larger vessels that travel long distances. Its share of container shipping worldwide is about 17%. The new methanol-powered ships will carry a total of 128,000 containers or around 3% of Maersk’s carrying capacity. (The vessels on order are much larger than the typical ship operated by this shipping line). Very roughly, the dual fuel container ships will move around 0.5% of the world’s container freight.

Maersk’s ships use about 10 million tonnes of bunker fuel a year at the moment. That results in almost 30 million tonnes of CO2 emissions, about the same as Denmark, Maersk’s home country. Methanol carries far less energy per unit weight than oil but will probably be more efficient at generating motion in a ship’s engine. The company says it will need about 360,000 tonnes of low carbon methanol and this fuel will save about 1million tonnes of carbon emissions compared to using heavy fuel oil.

Methanol made from fossil fuels is not in short supply. The world makes about 80 million tonnes a year at present, mostly as a precursor to chemicals production. It is a simple alcohol, containing carbon, hydrogen and one atom of oxygen. (CH3OH). About 200,000 tonnes today is manufactured using captured carbon dioxide and hydrogen. This is often called ‘green’ methanol, although the CO2 has usually been derived from a source that uses fossil fuels, usual a flue gas from an industrial process. This is not really ‘green’; the carbon dioxide will eventually end up in the atmosphere when burnt in the ship’s engine. 

Green methanol

Two types of methanol production may meet Maersk’s requirements for truly ‘green’ CH3OH. The first is when the CO2 is captured at a facility that is processing agricultural or biomass wastes. In this case, the gas has been extracted from the atmosphere by photosynthesis and so the methanol made using it will simply return the carbon dioxide to the air when burnt. 

The second type of plant does not exist yet. It will capture CO2 directly from the atmosphere and chemically merge it with hydrogen from electrolysis using renewable electricity. This is technically feasible. The only question is cost. Maersk paid about $300 a tonne of fuel in the last financial quarter and this represented a little over 10% of its revenues. So the price of fuel matters, although not overwhelmingly so, particularly in times of astronomical freight rates, such as at present.

The plans for green methanol plants around the world show how much of a boost to the market Maersk might be giving. The first site where methanol was made without fossil fuels was probably at Carbon Recycling in Iceland. There, CO2 from a geothermal power plant that would otherwise be vented is used with hydrogen made from electrolysis. This technology has worked well for more than a decade. Carbon Recycling is now expanding internationally and is planning a 100,000 tonne factory in Norway and a 110,000 tonne equivalent in China. (Both are reliant on fossil fuel exhaust gases and so are not properly ‘green’).

Liquid Wind in Sweden is a start-up that wants to construct factories making 50,000 tonnes a year. Seven of these would be needed to cover the demand from Maersk alone. REintegrate, the Danish company making the green methanol for the smaller ship Maersk ordered in February, is offering just 10,000 tonnes a year.

The effects of Maersk’s decision

The biggest impact of the Maersk initiative will be to transform the prospects for genuinely green methanol. One beneficiary might be the HIF project in southern Chile which intends to make a petrol/gasoline substitute using a process which has methanol produced in an intermediate step. Its plans are not advanced but would be sufficient to roughly cover Maersk’s entire needs. I expect the project executives have already booked their tickets to visit Maersk’s HQ in Copenhagen. 

What about the final gamble listed above? Will the cost be bearable? At the moment grey methanol trades for about $500 a tonne. And the energy value is little more than half that of fuel oil at $300 a tonne although combustion efficiency might be slightly better. This means that fuel costs would almost triple just by using fossil fuel derived methanol (and twice as much fuel will need to be carried for the same journey, reducing container capacity). 

The average container cost about $3,000 to ship last quarter, according to Maersk’s quarterly report. Current fuel costs might have been around $400, which would rise to over $1,000 if grey methanol were used, adding about 20% to the overall shipping cost. 

Green methanol is going to start by being more expensive, although possibly not significantly so. The two key costs are going to be the cost of hydrogen and captured CO2. CH3OH is 1/8 hydrogen by weight so a tonne of methanol requires 125 kg of green hydrogen. At the target price of $1.50 per kilogramme by 2025[3], the cost of H2 will be around $190. The capture of CO2 from air has a target price of around $100 per tonne and a tonne of methanol will need about 1.6 tonnes, implying a cost of at least $160. It might be less if the CO2 came from biological sources. Thus it is at least conceivable that green methanol might eventually cost no more than today’s fossil variety. But there is still a substantial price premium over the cheap heavy fuel oil used today.

 What will be the effect of the world’s (eventual) carbon tax? Buying 360,000 tonnes of green methanol will save about 1 million tonnes of emissions. At a carbon tax of $100 a tonne, Maersk will therefore save $100 million, or about $280 a tonne of methanol. There’s still a big gap between the fuel cost of methanol and that of heavy fuel oil, even of the desulphurised variety.

 Will Maersk’s clients pay the 20% increment that will be needed to cover the extra costs of green methanol when costs are reduced to today’s grey methanol prices? The company appears confident that it has obtained a sufficiently large base of customers prepared to pay the price. A long list of major global enterprises, ranging from H&M to Amazon and Novo Nordisk, seems to be committed to backing Maersk’s initiative. 

This appears to me to be a sensible gamble from the world’s largest shipping company. And its commanding stature in its industry will encourage others to follow. In pushing for methanol, Maersk is indicating that it is not yet ready to take the bigger risks on low carbon ammonia as a fuel.

[1] https://www.theguardian.com/business/nils-pratley-on-finance/2021/aug/24/sunaks-stamp-duty-holiday-hard-to-square-with-levelling-up-rhetoric (Last part of the article).

[2] https://www.freightwaves.com/news/inside-container-shippings-sudden-newbuild-ordering-spree

[3] This is the figure target by electrolyser company Nel in 2025 in low cost electricity locations.

The delusions of the UK hydrogen strategy

Long in gestation, the UK’s hydrogen strategy[1] gives the strongest possible sense of a country heading rapidly in the wrong direction. Where most countries and regions are pushing the development of hydrogen from green electricity, the UK has committed to a future dominated by the conversion of natural gas with carbon capture. Most strikingly, the UK is projecting costs for hydrogen vastly greater than anywhere else in the world.

Always sceptical about hydrogen made from electrolysis, the government has exceeded its normal standards of myopia. Below, I compare the estimates just provided by the UK government and by the EU in June 2020, 14 months ago.[2]

These figures show jaw-dropping differences between the EU’s and the UK’s estimates of the cost of hydrogen made from renewables in 2020 and 2025 respectively.

Estimates for the cost of hydrogen from electrolysis of renewables  (2025 or 2020)

2025 (UK)        £112/MWh

2020 (EU)        £64-£114/MWh[3]

These figures say that the EU saw the cost of hydrogen in 2020 as lower than the UK forecasts it will be in 2025. In the lowest cost locations, it was little more than half the price. The mid-point of the EU’s range is over 20% cheaper then than the UK forecasts for four years’ time. (It may be worth saying that nobody, but nobody, sees anything but the price of green hydrogen falling sharply in the next few months and years).

The differences widen. The EU estimates for the cost of renewable hydrogen in 2030 are up to 60% below the UK’s costs for 2050, twenty years later. 

Estimates for the cost of hydrogen from electrolysis of renewables (2030 or 2050)

2050 (UK) £71/MWh

2030 (EU) £28-£62/MWH

And it’s is also worth noting that the US is targeting a hydrogen price of $1/kilogramme by the end of this decade. That’s equivalent to a price of just under £22/MWh, or 70% below the UK’s estimate of costs twenty years later. The world’s largest electrolyser manufacturer, Norway’s Nel, aims for $1.50/kilogramme by 2025 (£33/MWh), less than half the UK’s 2050 estimates.

 What is going on? Why is the UK so jaw-droppingly less optimistic about renewable hydrogen than other entities? The suspicion must lie with the country’s devotion to the future of hydrogen from natural gas. It does very much look as though the policy-makers have determined that green hydrogen must remain more expensive than traditional routes of hydrogen manufacture using natural gas. 

Estimates for the cost of hydrogen from different technologies (UK 2050)

Renewable electrolysis          £71/MWh

Steam methane reforming

of natural gas                         £67/MWh

Autothermal reforming           £65/MWh

As an aside, it’s notable that the the UK strategy paper talks of ‘small projects expected to be ready to build in the early 2020s’ using renewable electricity but ‘large scale projects expected from mid 2020s’ for those employing natural gas and carbon capture. In other words, renewable electrolysis is still a toy. Even by 2050, the typical project seems to be expected to use a 10 MW electrolyser, when everybody else is talking of schemes today of one hundred times this scale.

Nowhere else in the world expects hydrogen to be cheaper to make using natural gas with carbon capture than electrolysis by mid-century. The UK government numbers are truly staggering.

 But, of course, there’s no reference that I could find in the UK strategy paper to any data or opinions from abroad.  That’s despite many major economies publishing their own policies over the last year. 

 All one can say is this: if green hydrogen made in the UK does cost £71/MWh in 2050, there’s absolutely no point in trying to build an industry here. It will be vastly cheaper to import the gas from Spain or Portugal by pipeline or Chile by liquid hydrogen carrier. The whole UK strategy will come to nothing, using a lot of taxpayers’ cash in the next four decades.

[1] https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/1011283/UK-Hydrogen-Strategy_web.pdf

[2] https://ec.europa.eu/energy/sites/ener/files/hydrogen_strategy.pdf

[3] €2.5-5.5/kg, using the Lower Heating Value and an exchange rate of €1.17/£.

The Hydrogen Revolution by Marco Alverà

The Hydrogen Revolution; a Blueprint for the Future of Clean Energy by Marco Alverà is published by Hodder Studio on 26th August at £20.00

The history of hydrogen’s use as a man-made source of energy is older than we might think. Late in the nineteenth century, the Danish inventor Poul la Cour improved the design of wind turbines. Realising that his machines were often generating excess electricity that needed to be stored for later use when wind speeds were low, he used electrolysis to make hydrogen. Production of up to 1,000 litres an hour was kept in a tank and then burnt to make light at the high school where la Tour taught. From 1895 to 1902 the machinery worked successfully, keeping the lights on at the school every single day. (Although the building’s windows seem to have been very occasionally blown out by minor explosions of hydrogen)

Similar anecdotes fill Marco Alverà’s new book about the vital role of hydrogen in the world’s energy transition. A senior businessperson now running Italy’s Snam, Europe’s largest natural gas distribution grid, Alverà is the perfect individual to give us sense of how hydrogen is going to be used in the zero-carbon future. The book bubbles with his optimistic fervour and mixes technical detail with personal opinions. It doesn’t remotely read like an earnest manifesto from a corporate leader protected by his public relations team from saying anything too specific. Instead it offers an impassioned argument for a rapid transition to a global economy whose entire energy needs are met by renewable electricity and the hydrogen generated from it.

Alvera - and the Snam strategy team that helped with his research - side with the proponents who foresee a world that gets 25% of its energy needs from hydrogen, more than the share held by natural gas today. Other analysts, such as the International Energy Agency and the Energy Transitions Commission, have recently offered projections that suggest figures of about two thirds of this level. 

The book argues that almost every activity that cannot easily be electrified should be converted to hydrogen. So, for example, Alverà suggests that many domestic heating boilers should be switched away from natural gas to hydrogen. He is unusual in thinking this; most analysts recommend the use of electric heat pumps in the large majority of houses. His judgment, perhaps too strongly influenced by the constraints on the Italian electricity grid, is that the infrastructure investment needed to deliver reliable electric power for heating at the coldest times of the year is too great to be envisioned. And, as with the hydrogen at Poul la Cour’s school, hydrogen can be used to provide energy at times when renewable electricity isn’t available.

Marco Alverà sees a much larger role for hydrogen in transportation than most commentators, believing that the longer range of fuel cell cars will help beat off the competition from battery vehicles. The book also enthusiastically promotes hydrogen airplanes, even though they will require a fundamental redesign and may find it difficult to compete with existing aircraft configured to use synthetic zero-carbon fuels. 

Alverà gives us analysis, delivered as usual with flair and enthusiasm, that shows how a programme of investment in electrolysers and renewable electricity could push the cost of hydrogen down to below $2.00 a kilogramme within five years. His models show that this is attainable in large parts of the world after 25 gigawatts of electrolyser capacity have been installed, making about 5 million tonnes of hydrogen, or less than 10% of today’s production of the gas from fossil fuels. In the locations with the cheapest renewable electricity this target will be achieved earlier.

But should we believe the chief executive of a natural gas distribution company about the stellar future of alternative energy sources? Without a large role for hydrogen, which can be transported in refurbished natural gas pipelines, Alverà’s company has very poor prospects in the post-carbon world. An increasing number of voices argue that hydrogen is simply a front for ensuring that the main fossil fuel companies, such as Snam, continue in business.

But the author’s case is strengthened by his analysis showing how cheap it would be to build and operate hydrogen pipelines around Europe and beyond, bringing hydrogen from sunny places to the industrial centres where it is most needed. Gas pipelines are very much cheaper to construct and operate than long-distance electricity networks. Alverà suggests new connections both to North Africa and to parts of the Middle East where solar energy will be particularly cheap. Pipelines also offer large amounts of energy storage, with one kilometre containing a maximum of 12 tonnes of hydrogen, with an energy value equivalent to the daily electricity use of 40,000 homes. Even larger volumes can be stored in underground caverns easily created in the salt strata that exist underneath many countries around the world, including the UK and other countries in northern Europe. 

Marco Alverà is keen to show us how a world of solar, wind and hydrogen can allow the world can continue to live much as it does now. In his view our lifestyles, including our car and air travel and purchasing of clothes, electronics and other goods, will not need to be curtailed. I think this where he moves on to dangerous ground. The climate crisis is just one part of the unfolding ecological disaster that includes the loss of biodiversity, rising pollution from other sources than energy production and the destruction of many natural environments. Although hydrogen has an absolutely central role in averting the worst of our climate problems, it is not a universal panacea.

Hydrogen will be cheaper than today's natural gas prices by 2025

Natural gas prices are high around the world. In the UK, recent prices of above 114 pence per therm (about 3.9 pence/5.4 US cents per kilowatt hour) exceed any figures since the wholesale market became established in the 1990s. Similar records are seen across Europe and around the world. Forward contracts predict that these high prices will remain for many months. 

Part of the reason is geopolitics. Russia is ‘encouraging’ Europe to move ahead with the Nord Stream 2 pipeline by restricting exports on existing routes, forcing prices higher. Another reason is climate change. Gas demand is higher than usual because soaring summer temperatures are prompting high electricity use for air-conditioning. And in places like Brazil long-term drought has sharply reduced hydro-electric generation, forcing the country to ship in more gas from the US.

The price rise ought to make us even more eager to move away from fossil fuels such as natural gas. The effects of the cost inflation will be felt predominantly by the poorest groups. In the UK, the least well off decile of households devote 8% of their spending to paying for gas and electricity, almost three times as much as the richest decile.

At current gas prices in many parts of the world it will soon be cheaper to use hydrogen as a fuel for heating or for making electricity. (Many existing types of gas turbines can use some portion of hydrogen instead and the main suppliers say that 100% hydrogen will be possible soon).

Earlier this year, the large electrolyser supplier Nel published a formal target for the cost of green hydrogen in 2025, just four years away.[1] It looks for a figure of $1.50 per kilogramme, based on expectations for its own electrolyser costs and, most importantly, a figure of $20 per megawatt hour for electricity supply. Translate $1.50 into a cost per kilowatt hour and we arrive at a figure of 4.5 cents. Compare that to the price of wholesale gas in the UK today, which costs around 5.4 cents per kilowatt hour, making hydrogen prospectively almost 17% cheaper. 

The continuing belief that green hydrogen is always ‘too expensive’ to be useful needs to be challenged. We need a robust international supply chain for low carbon hydrogen, probably coming from the lowest cost locations such as Australia and Chile, to help keep heating and electricity prices down. And as reports this week have made clear, it has to be ‘green’ hydrogen because the alternative ‘blue’ product, made from natural gas with CCS almost certainly involves emissions that rule out its widespread use.[2]

[1] https://nelhydrogen.com/wp-content/uploads/2021/05/Nel-ASA-Q1-2021-presentation.pdf

[2] https://www.theguardian.com/environment/2021/aug/12/uk-replace-fossil-gas-blue-hydrogen-backfire-emissions

Hydrogen made at the wind turbine

Both offshore wind and hydrogen generation are increasingly seen as central to global decarbonisation. And over the last year we’ve seen a striking increase in the number of linked proposals that seek to develop offshore wind farms that are at least partly devoted to making hydrogen. Most of these schemes envisage generating electricity at the turbine which is then taken onshore via a high voltage undersea cable. This electric power will then be fed into a large electrolyser complex onshore to make hydrogen for various uses.

An illustration of how a central electrolyser complex on a platform in the middle of the Aquaventus project might look

But the story is changing rapidly; we are seeing a striking increase in the number of schemes that propose to make hydrogen in the wind turbine itself or at a platform in the middle of the wind farm. Instead of electricity, the farm will feed hydrogen onshore via a pipeline and then directly to users or to a repurposed gas grid for wider distribution. Although these new schemes are still at an early stage, it looks probable that a substantial fraction of all offshore wind turbines will eventually make hydrogen rather than transmit electricity.

What is driving this largely unnoticed change in the plans of large offshore developers? 

1, It is usually (much) cheaper to transport hydrogen than it is to move electricity. 

2, Having the electrolyser in the turbine or on a nearby structure enables the electronics in the turbine to be simpler. 

3, If making hydrogen is the ultimate purpose of the electricity made at the wind farm, then it may make sense never to attach the turbine to the grid. The advantages of this include cost – no need for a substation onshore, for example – and flexibility. There is no risk of enforced disconnection if the grid temporarily cannot handle the electricity that the turbine produces.

4, In addition, the hydrogen pipeline to the shore can act as a very efficient storage medium. The pressure can be increased and more hydrogen ‘packed’ into the pipeline.

1, Cheaper to transport hydrogen than electricity

Hydrogen is surprisingly easy to transport via pipeline. It can be compressed to approximately the same level as natural gas in conventional pipelines and flows more easily because of its lower viscosity. (But it does have a lower energy value per cubic metre at comparable pressures). 

The German utility RWE wrote as follows about its proposed AquaSector project in the North Sea that will eventually make hydrogen offshore (see below for more details about this proposal).

Compared to the transport of electricity generated offshore, the hydrogen production at sea and the transport via pipeline could offer clear economic advantages. The pipeline could replace five High Voltage Direct Current (HVDC) transmission systems, which would otherwise have to be built. It is by far the most cost-effective option for transporting large volumes of energy over long distances.[1] 

One of the largest offshore hydrogen schemes is likely to be NortH2 in the Netherlands sector. The developers say this[2]:

As wind turbines are placed further out to sea, hydrogen production close to the source becomes more attractive. After all, the generated energy must also be transported to land. This can be done via heavy electricity cables, but it is cheaper and more efficient to transport hydrogen gas molecules. That is why NortH2 is also looking at possibilities to convert the generated wind power into hydrogen directly at the wind turbines: electrolysis at sea.

Siemens, which makes many of the offshore wind turbines now being installed around the world, seems to be convinced of the operational advantages of putting the electrolysis process either into the turbine itself or close by. One of its recent presentations suggests three reasons for favouring transporting hydrogen rather than electricity including the following comment. [3]

* Capex reduction by replacing high cost HV infrastructure with pipes network

The precise cost of transport to shore will depend on the distance and the complexity of linking to the hydrogen users or to the gas grid. The EU offered a recent estimate of around 5 Euro cents per kilometre per megawatt hour of energy.[4]This implies a cost of about €6/MWh for a 100 km link. The number was based on earlier research by Bloomberg NEF. 

Other experts, including the CEO of the Italian gas distributor Snam, offer much lower figures suggesting that hydrogen transport by pipeline can cost as little as one eighth as much as electricity transfer.[5] (However this largely assumes using existing natural gas pipelines). 

Unlike the transmission of electricity, which will always see losses of power in the transmission process and in the conversion from one voltage to another, hydrogen can be transported in gaseous form almost without loss.

2, Simpler electronics in the turbine itself.

A wind turbine connected to the grid, even via a long undersea cable, requires substantial electronic equipment to condition the power that is generated. Second by second, the varying amounts of electricity being produced have to be converted to a standard voltage and aligned to the frequency of the grid. This means that the power coming from the turbine requires substantial manipulation before it can be connected to the grid. The equipment required is costly.

3, Less equipment onshore and complete independence from the requirements of the electricity grid.

As renewables grow around the world, the percentage of time that wind turbines produce power which cannot be accepted by the national or regional grid is tending to rise. At these times the electricity is wasted. One of the attractions of not connecting turbines to the electricity system and directly making hydrogen instead is that all the power produced will be productively used.

In addition, there will be no need for a substation or substations onshore that connects the power coming in from the wind farm to the main grid. The hydrogen can either be connected to a gas network or used directly by customers such as an oil refinery, a steel plant or a fertiliser manufacturer.

4, The hydrogen pipeline can act as storage.

Natural gas pipeline networks can reduce or increase the pressure in their pipes to provide a buffer between supply and the demand for energy. The same will be true of hydrogen pipelines. When the wind is blowing and the turbine is making hydrogen but customers only need their standard amounts of the gas, it will make good sense to store the temporary surplus in the pipeline. Marco Alverà of Snam says that I kilometre of pipeline can store 12 tonnes of hydrogen, with an energy value of around 400 MWh.[6] Batteries can in theory provide the same service for a turbine that produces electricity but the cost for an storage capacity equivalent to a kilometre of hydrogen pipeline would be fifty million dollars or more. 

The routes to a full coupling of offshore wind and hydrogen.

Let’s now look in a little more detail at some of the main pilot projects to produce hydrogen directly at the turbine or at central node in a wind farm. We will start with early experiments that are precursors to full hydrogen production and then move through to some of the outline plans for gigawatt-scale projects. All the schemes I could find were in Europe.

Denmark: a trial to show that a wind turbine can directly power an electrolyser while unconnected to the grid.

 At Brande in Denmark, close to the its Danish headquarters, Siemens is trialling joint operation of a 3 MW onshore wind turbine and an electrolyser.[7] One of the most important features of the experiment is the attempt to show that the turbine can be ‘islanded’, meaning it can power the electrolyser without any grid connection, thus replicating how an offshore turbine without the means to connect to the electricity network can make hydrogen. The hydrogen from Brande will be used to power Copenhagen’s fleet of fuel cell taxis.

The site for hydrogen production from an islanded onshore turbine at Brande, Denmark

Perhaps more importantly, Siemens’ wind business is also investing heavily in the development of a variant of its largest offshore turbine that that will incorporate an electrolyser in the base of the tower. This work is being supported by the Siemens division that makes electrolysers.

The Netherlands: installation of an electrolyser on an existing natural gas platform.

The PosHYdon project in the Netherlands North Sea will use an existing natural gas platform to host a 1.25 MW NEL electrolyser, making a maximum of about 500 kilogrammes of hydrogen a day from water that has been demineralised.[8]The hydrogen will be added to the natural gas produced at the platform which will then be piped onshore. (Domestic appliances can burn natural gas that contains some hydrogen. The percentages that are allowed vary from country to country). 

This trial will not use electricity from a nearby wind turbine but instead will use power that has been delivered to the platform from onshore. The installation is one of the few fully electrified offshore oil and gas platforms in the world.

An image of how an electrolyser might look on the Dutch offshore gas platform

The experiment will test the durability of a NEL electrolyser in a rugged marine environment and also examine how well the electrolyser copes with sharp variations in the availability of power.

The project is backed by a large number of energy companies and science research funds of the Netherlands government. 

Norway: making hydrogen in times of surplus electricity and storing it on the sea-floor.

A pilot project will use seawater to make hydrogen close to an offshore wind farm in periods when electricity is in excess supply.[9] The hydrogen will be made and stored on the sea floor. Fuel cells will convert the gas back into electricity at the wind farm at times of shortage. In this case hydrogen does not avoid the need for electricity transmission. Instead it uses the gas to help make electricity supply more reliable. 

This scheme is still at an early stage but involves a large number of partners with strong interests in hydrogen, including Finnish utility Vattenfall and Spanish oil company Repsol. It is being led by Technip FMC, a leading supplier of equipment for undersea oil exploration. The €9m project is being part-funded by Innovation Norway.

Underwater electrolyses and fuel cells in the projected Technip FMC scheme

UK: A trial project to make hydrogen on a floating offshore wind turbine base.

One of the potential advantages of using floating turbines for hydrogen production is that the turbine base provides a horizontal surface on which to place the electrolyser and other equipment. It may be easier and more economical to use floaters than fixed foundation turbines. Consulting firm ERM has received funding to develop a trial site in the middle part of this decade that will install a large 10 MW turbine, probably off the Scottish coast, to make hydrogen.

The design of the ERM floating turbine with hydrogen electrolyser

ERM’s technology may be used at a much larger Scottish wind farm.[10] The company recently signed a memorandum of understanding with the developers of a 200 MW project that will be constructed before 2030. The Scottish gas grid operator SGN is also part of the possible scheme to use floating turbines to make hydrogen to be piped onshore.

Italy: making electricity and hydrogen from an offshore wind and solar site.

The engineering company Saipem and its partners intends to make hydrogen at a new 450 MW offshore wind and offshore solar site in the Adriatic Sea off Ravenna.[11] The offshore farms will have an electrical connection to the mainland but also make hydrogen on converted offshore oil and gas platforms that will be decommissioned. The hydrogen will be piped both to the shore, and thus to the local natural gas network, and to marine refuelling platform that will accommodate ships that will transport the hydrogen to other locations.

A illustration of some aspects of the Saipem project, including offshore hydrogen production and hydrogen transport ship.

Netherlands: investigation of the opportunity to directly produce hydrogen at an offshore wind farm.

NortH2, a much larger Dutch project, is examining the options for directly making hydrogen from wind turbines.[12] It envisages as much as 4 GW offshore wind capacity in the Netherlands North Sea by 2030. In the early design, the power will be transported to a port on the Dutch coast but the consortium is backing research that looks at making hydrogen either at the turbines, on a platform shared between many turbines or at a man-made island close to the wind farm. This scheme is backed by the Norwegian oil company Equinor, Shell and several other entities including the innovative Netherlands gas grid operator Gasunie. 

Germany

Aquaventus is the most ambitious of all European projects for making hydrogen offshore.[13] The scheme proposes eventually to use 10 GW of offshore wind to make hydrogen to be transported by pipeline to the German island of Heligoland. The first phase of this enormous scheme involves the installation of about 300 MW of offshore wind, producing about 20,000 tonnes of hydrogen a year by 2028. (Current global demand for hydrogen is about 70 million tonnes, and this is likely to rise sharply). 

The partners behind the Aquaventus project, which include the utility RWE and Shell, regard it as a ‘proof of concept’ for the larger set of wind farms to be built by 2035, all of which will be connected to Heligoland by hydrogen pipeline. The gas will then be piped to mainland Germany. 

An illustration of the 10 GW wind farm feeding hydrogen into north west Germany via Heligoland

In almost all respects this scheme seems similar to existing plans for North Sea wind farms producing electricity with the only difference being the generation of hydrogen. If it comes to fruition it will show that direct production of hydrogen at wind farms may eventually be as financially attractive as the conventional model of producing electricity.

What can we conclude from these diverse experiments and pilot projects?

Although some participants in these pilots are still checking that electrolysis at sea is possible and makes financial sense, there seems to be an increasingly strong view that a large fraction of total hydrogen supply will come from offshore wind turbines. Many of the largest European utilities are heavily involved in the project proposals.

Underlying this view is the sense that the demand for hydrogen will be sufficiently broad to make investment in direct manufacture at renewables sites an attractive proposition. The usual objection to making hydrogen from renewables is the loss in energy value resulting from the electrolysis process. But if the market needs huge quantities it is not a question of whether hydrogen should be manufactured but how best to do this. The number of large companies crowding into the hydrogen from offshore wind business suggests a high confidence that many millions of tonnes of hydrogen will be needed. 

[1] https://www.rwe.com/en/press/rwe-renewables/2021-07-23-aquasector-partnership-on-first-large-scale-offshore-park-for-green-hydrogen-in-germany

[2] https://www.north2.eu/en/blog-en/offshore-electrolysis/

[3] https://www.siemensgamesa.com/en-int/-/media/siemensgamesa/downloads/en/products-and-services/hybrid-power-and-storage/green-hydrogen/210318-siemens-energy-hydrogen-day.pdf page 10. 

[4] https://op.europa.eu/en/publication-detail/-/publication/7e4afa7d-d077-11ea-adf7-01aa75ed71a1/language-en?WT.mc_id=Searchresult&WT.ria_c=37085&WT.ria_f=3608&WT.ria_ev=search

 [5] This information is taken from Marco Alverà’s forthcoming book, The Hydrogen Revolution.

[6] Marco Alverà in The Hydrogen Revolution.

[7] https://www.siemensgamesa.com/en-int/products-and-services/hybrid-and-storage/green-hydrogen

[8] https://poshydon.com/en/home-en/

[9] https://www.energyvoice.com/renewables-energy-transition/hydrogen/289798/technip-fmc-offshore-green-hydrogen/

[10] https://www.offshorewind.biz/2021/08/04/scottish-floating-wind-project-forms-green-hydrogen-tie-up/

[11] https://www.saipem.com/en/media/news/2020-08-25/saipem-protagonist-offshore-wind-will-develop-wind-farm-italy

https://www.agnespower.com/en/progetto-adriatico/

[12] https://www.north2.eu/en/blog-en/offshore-electrolysis/

[13] https://www.rwe.com/en/press/rwe-renewables/2021-07-23-aquasector-partnership-on-first-large-scale-offshore-park-for-green-hydrogen-in-germany

The struggles to make CCS work

The continuing difficulties facing the huge Gorgon carbon capture project in Western Australia must make us concerned about the viability of CCS elsewhere in the world.[1] As an informed Australian commentator said after recent announcements from the gas field, the Gorgon experience implies that CO2 storage will be more ‘expensive, slow and difficult’ than was hoped.[2] Each project will need to be carefully tailored to the precise geologic circumstances of the reinjection site. In his words, the difficulties at Gorgon show that CCS will be only a ‘vital and important, but niche, component’ of the energy transition. 

Part of the offshore infrastructure for the Gorgon project. Source: Chevron

This would also be the conclusion of many of those associated with an earlier large CCS project to reinject carbon dioxide at the In Salah gas field in central Algeria. This experiment ran into similar geological problems and was abandoned after several years because of concerns that the CO2 might escape. 

In both cases, the projects have been run by some of the world’s largest fossil fuel companies, all with huge experience in understanding geology and deep drilling. If these businesses cannot manage to achieve successful CO2 storage in nearly ideal conditions, there must be real doubts about whether carbon dioxide can be effectively stored in oil and gas formations.

Nevertheless, some governments around the world, and many fossil fuel companies, see CCS as a saviour technology that will allow continued large scale use of fossil fuels. The experience at Gorgon, and at almost all other CCS projects, suggests that this unthinking reliance on carbon capture is mistaken. The world will need to store CO2, but it cannot be a central plank of our decarbonisation strategies.  Australia’s community-funded Climate Change Council summarises the history of global carbon storage in a unequivocal fashion - ‘no CCS project has yet been delivered on time, on budget or to agreed performance’. [3]

Gorgon CCS

Gorgon is a series of large offshore gas fields, operated by Chevron with shareholdings also held by Exxon Mobil and Shell as well minor stakes taken by Japanese gas supply companies.  The project is one of the world’s largest sources of natural gas. Most of the production is liquefied to LNG and then transported to Asia. 

The Gorgon gas field off Western Australia, with pipelines going onshore via Barrow Island, where the CO2 separation occurs. Source: Chevron

As with many other gas fields, the Gorgon output naturally contains some carbon dioxide. Percentages range from 1% up to about 15% depending on which of the several separate fields the gas comes from. Even small percentages of CO2 cause particular problems for the liquefaction process. CO2 freezes to a solid at higher temperature than those at which gaseous hydrocarbons turn to liquid. This causes damage to equipment at gas liquefaction plants, such as those that process the Gorgon output. 

So the CO2 has to be separated from the natural gas. This is relatively simple. Some chemicals naturally absorb carbon dioxide and passing extracted natural gas over these chemicals will result in the CO2 being captured. The carbon dioxide can then be released again by simple heating, completely separating it from the hydrocarbons in natural gas. 

In most places around the world where gas liquefaction takes place, the CO2 is released to the atmosphere. Gorgon was meant to be different. The CO2 was intended to be injected back into the sandstone formation from which the natural gas originally came. The developers promised to put back at least 80% of the CO2 that had been separated out. It hasn’t turned out as well as hoped. 

An outline of how the Gorgon CCS scheme operates. Source: Chevron

The Gorgon project was started in 2009 and CO2 capture was intended to begin in 2016. The difficulties faced by the project meant that no carbon dioxide was actually injected until 2019. Since then, the sequestration process appears to never to have been fully operational and the amount stored is a fraction of what was expected. As a result, Chevron and its partners may have to pay fines of up to AUS$100m/$74m. (In the context of the project, this is an insignificant penalty).

What has gone wrong? The first problem was that when mixed with water CO2 forms carbonic acid, a weakly corrosive molecule. After the CO2 is injected into the sandstone formation, which is filled with water, the carbonic acid starts to dissolve the metal equipment in the injection well. 

The injection of carbon dioxide into the sandstone increases the pressure in the formation. Unchecked, this would eventually result in underground rock fracturing and the possibility of the return of the CO2 to the surface. This eventuality has previously been vehemently denied by the CCS industry.  In order to avoid leakage, Chevron created another set of wells to extract water from the formation to reduce the pressure. The wells did not work properly because both sand and water rose to the surface, eventually clogging the pipes. The difficulties resolving this eventually forced the Australian regulator to ask Chevron to reduce the rate at which CO2 was being injected into the formation so that the pressure did not rise too fast.

This problem seems to be persisting, reducing the rate at which the carbon dioxide is stored. Industry estimates suggest that only 2.5m tonnes a year are being sequestered rather than the 4m tonnes which was promised at the beginning of the project. Thus far, the CCS portion of the Gorgon project is said to have cost about AUS$3bn ($2.2bn) and has injected a total of about 5 million tonnes. If the current collection rates continue, the total amount sequestered is likely to be around 50 million tonnes during the lifetime of the field, about half of what was initially promised.

 The CO2 capture and storage will be much more expensive than first forecast. Assuming the $2.2bn figure applies to the full 50 million tonnes collected, the capital alone will imply a cost of around $45 a tonne of CO2. The full price, including operating costs, will be much higher.

The experience at Gorgon mirrors the most signifcant earlier attempt by the oil and gas industry to sequester the CO2 originally mixed into natural gas.

The In Salah experience 

BP and Equinor (formerly Statoil) are shareholders in the In Salah field in central Algeria. The operator is state-owned Sonatrach, the largest African oil and gas company.

 The In Salah field first began producing gas in 2004. It is expected to continue in operation until 2027. As in the Gorgon fields, the Algerian gas contains too much CO2 and the excess has to be removed. The target was to capture about 1 million tonnes a year and reinject it back in to the sandstone formation from which the gas has been extracted.

The In Salah gas field. Source: Sonatrach

The project was never fully successful. By 2011, when the CCS project was abandoned, about 4 million tonnes had actually been injected back into the gas-bearing sandstone formation.

What went wrong? In this case, there appears to have been no attempt to reduce the pressure in the CO2 storage areas by extracting water. CO2 was injected directly into the sandstone formation and caused the pressure to rise to levels sufficiently high to fracture the rocks above, raising the possibility of a leak.

The following paragraph is taken from an academic paper written by engineers from BP, Equinor (then Statoil) and Sonatrach after the project was abandoned.[4]

 ‘Following  the  2010  QRA (Quantified Risk Assessment),  the  decision  was  made  to  reduce  CO2  injection  pressures  in  June  2010.  Subsequent analysis of the reservoir, seismic and geomechanical data led to the decision to suspend CO2 injection in June 2011. The future injection strategy is currently under review and the comprehensive site monitoring  programme  continues.  Concerns  about  possible  vertical  leakage  into  the  caprock  led  to  an  intensified  R&D  programme  to  understand  the  geomechanical  response  to  CO2  injection  at  this  site’.

The diagram below shows where the engineers suggest fracturing may have already occurred by the time the project was abandoned. (See, for example, the near vertical line in the centre of the graphic). 

Visualisation of some of the problems at one of the CO2 injection wells at In Salah. Source: https://www.sciencedirect.com/science/article/pii/S1876610213007947

Although the risk of excess CO2 pressure producing or enhancing rock fractures was considered before the project began, it was not initially regarded as likely. The engineers had carefully selected the reservoir for injection, saying that it had ‘big storage capacity with a good insulation’ of rock over the top.[5] This turned out not to be the case.

BP engineers on the project had earlier described the storage geology at In Salah as ‘very similar to that of the North Sea’, where the company also hopes to develop large CCS projects.[6] We have long been told by specialists in CCS that injection of CO2 into depleted fossil fuel formations held no risks because the geology had already proved itself by retaining the gas or oil for hundreds of millions of years. The experience at In Salah and at Gorgon suggests that this does not provide sufficient security, perhaps because the volumes of CO2 stored result in pressures that are higher than projected by the geologists.

It is possibly a trivial finding but one other feature of In Salah needs mentioning, if only because the oil company engineers themselves discuss it in some detail. Parts of the land above the CO2 injection wells have risen very slightly (by up to 20mm) in response to the carbon dioxide stored at pressure over two kilometres below the ground. The direct significance of this is small, but it does indicate that large volumes of injection even into very deep formations can have unexpected effects on geology.

What does this mean for the future of CCS?

The world needs carbon capture and storage if it is to get to net zero. There may always be activities, such as the making of cement, that cannot be carried out without CO2 emissions and these must be safely stored. However the evidence from Gorgon and In Salah is that successful storage in oil and gas formations is almost certainly;

a)    More difficult and expensive than expected.

b)    Very dependent on geology. An approach to CCS that might work in one location might fail in another. 

c)     So rolling out CCS rapidly and at gigatonne scale in many hundreds of places around the world is not easy to envisage. We are still in the stage of CCS experimentation, and are well before a standardisable and inexpensive approach can be widely used.

d)    Areas, such as the North Sea, which are touted as perfectly suited to geologic storage, may well be more difficult to use than currently expected by government and by the oil and gas industry. 

e)    Of particular concern is the development of a ‘blue hydrogen’ industry around NW Europe, which will probably rely entirely on finding CO2 storage sites in the North Sea. However, as our knowledge stands today, the injection of carbon dioxide is likely to be more costly and much more limited in tonnage stored than is being currently modelled.

[1] The two cases discussed in this note both involve injecting CO2 into the geologic formation that contains gas but at a location away from the gas field itself. We cannot conclude that all types of CCS, including injection into working oil fields, will experience similar problems. However very large scale storage (hundreds of millions of tonnes) does now look more difficult than we believed.

[2] https://www.abc.net.au/radionational/programs/sundayextra/chevron-gorgon-ccs/13467950 Interview with Peter Milne. Absolutely fascinating and highly recommended.

[3] https://www.climatecouncil.org.au/resources/what-is-carbon-capture-and-storage/

[4] https://reader.elsevier.com/reader/sd/pii/S1876610213007947?token=CA1B347BA1CD3EFB86A7F2B30B81BE638206DB66453686410CD6A56CC773892ED08E3CD4D2C7EC601DC59A5BE0C679A6&originRegion=eu-west-1&originCreation=20210730104253

[5] https://www.opec.org/opec_web/static_files_project/media/downloads/press_room/HaddadjiSonatrach_Algeria.pdf page 27

[6] https://ec.europa.eu/clima/sites/default/files/lowcarbon/ccs/docs/colloqueco2-2007_session2_3-wright_en.pdf page 10

ArcelorMittal says it will be producing zero carbon steel in 2025

ArcelorMittal’s Gijón steel plant in north west Spain

ArcelorMittal is the second largest steel-maker in the world, trailing only Baowu, the huge Chinese producer. It produces about 80 million tonnes of the metal a year, or about 4% of the global total. Making steel is a process that uses coal and generates large amounts of CO2, meaning the company is alone responsible for about 0.3% of world emissions. So its actions matter. Recent company news suggests a new willingness to invest in the full transformation of its business away from coal and towards hydrogen.[1] The significance of the move appears to have been missed by the world’s media. 

Earlier in July, ArcelorMittal announced a plan to build what will be its first zero-carbon steel-making facility. If the target opening date of 2025 is achieved, the 2.6 million tonne plant at Gijón in north west Spain promises to be the first full scale low carbon steel works in the world. It will beat the current leader, Sweden’s SSAB, by a year. SSAB is already producing trial quantities of metal without using coal but only promises commercial quantities in 2026. 

At Gijón, green hydrogen, made from solar electricity, will be used to reduce iron ores (oxides of the metal) to sponge iron, from which steel can be made. Up until this point ArcelorMittal had begun several experiments of varying size and financial cost that attempt to reduce the greenhouse gas intensity of steel-making. None promised full carbon neutrality. Some involved the reuse of waste gases or their conversion to ethanol.

This month’s announcement is important because it seems to commit ArcelorMittal for the first time to a large scale investment at an existing steel plant that will produce zero-carbon metal. Until now, the company sometimes appeared to be toying with the carbon problem, making unclear promises to ‘eventually’ move to green hydrogen use in some of its German plants or to recycle waste gases containing CO2 in other European steel works. 

The proposed process at Gijón – ‘direct reduction’ or DRI – is already extensively used around the world although it conventionally employs natural gas to create synthesis gas (carbon monoxide and hydrogen, usually called ‘syngas’), rather than using hydrogen directly. DRI plants are less expensive to build than conventional blast furnaces and can be economically operated at a smaller scale. ArcelorMittal, perhaps aided by the US company Midrex that dominates DRI manufacturing technology, appears to be committing to using pure hydrogen at Gijón at a much greater scale than ever before planned in the world steel industry.[2]

Why now?

After dragging its feet during recent years, and making very few specific promises on decarbonisation, the company seems to finally made a full scale plan for greening part of its production. Why now? Probably the most important reasons will have been - 

·      The willingness of the Spanish government to help ArcelorMittal with the capital costs of the new plant, and probably its operating expenditures as well. The ArcelorMittal announcement of the Gijón plan came after the signing of a memorandum of understanding with the Spanish government which indicated that Spain will provide financial help but without being specific as to the amount.

·      As the weeks pass, the chance of the EU imposing carbon taxes on steel are rising. Not only is it increasingly likely that steel makers will have to pay for their ETS allowance but the probability of a carbon price at the borders of the EU within ten years has grown. This will make steel made from coal substantially more expensive. A tonne of steel typically requires about 0.75 tonnes of coal, and is therefore responsible for about 1.9 tonnes of CO2emissions. At an ETS price of around $60 a tonne, carbon taxation might add over $110 to the price of steel made with coal. (Steel usually trades for around $600 a tonne, so the carbon price could make a real difference). 

·      The cost of green hydrogen is falling fast, largely because of the fall in price of renewable electricity. Spanish solar parks could probably now produce electricity for less than $25 per MWH (around €20). I have estimated elsewhere that a tonne of low carbon steel will probably require about 4.25 MWh of electricity, costing therefore about $107. At today’s metallurgical coal prices of around $135 a tonne, steelmaking costs around the same whether using electricity or coal. And this is before taking carbon taxation into account.

·      After a long period of lukewarm interest in solar PV - Spain has less photovoltaic capacity than the UK -  the Spanish government has allowed substantial expansion of production capacity in the past year. The Gijón plant will need very large amounts of electricity to make hydrogen; I calculate it will probably require about 6 gigawatts, or about 50% of current national installed PV capacity. The national administration is making clear that it will encourage the development of the new solar fields in the local area that will be needed to deliver the 4% extra national electricity production that Gijón alone will require. 

 What are the wider implications?

·      It seems to me that the Spanish support for ArcelorMittal must inevitably produce similar offers from governments in the other main steel-producing countries in Europe. If ArcelorMittal goes ahead at Gijón I guess it is likely that no new steel furnaces will be built on mainland Europe that don’t use hydrogen. (Germany’s finance minister has already made a commitment to the local steel industry that promised whatever support is needed for a transition to hydrogen). Many steel furnaces inside the EU are reaching the end of their lives and it makes increasing sense to convert to hydrogen DRI instead of the costly rebuilding of ageing steel furnaces. 

·      We are beginning to get a sense of what the transition to hydrogen in the steel industry will cost. ArcelorMittal has previously said that it thought its transition to zero carbon using hydrogen would require investment of around $40bn for its own plants. The limited financial figures released for the Gijón project are consistent with this estimate. Grossed up to the global industry, we can expect an investment of around $1trn, or slightly more than 1% of global GDP. The benefit will be a reduction of about 8% in world CO2 emissions.  The will be spread over perhaps 25 years, implying annual investment requirements of less than 2% of world steel industry turnover. Even in an industry that goes through frequent financial crises, this is manageable.

ArcelorMittal invested about $3.5bn in new fixed assets in 2019. (The unusual 2020 figure was much lower than this.) The new Spanish DRI plant is therefore a large fraction of the company’s typical annual capital expenditure. But the output of the new Gijón plant represents over 3% of ArcelorMittal’s total steel production, meaning that a full conversion to DRI over the thirty years to 2050 should be fully financeable within the company’s existing capital budget

·      The investment world may come to recognise that steel making will almost inevitably shift to areas of the lowest electricity prices. Spanish PV can compete but it is less certain that German offshore wind can provide the cost-competitive electricity prices that the local industry needs. Australia, with good supplies of accessible iron ore and ultra-low potential renewable energy prices is very strongly positioned to regenerate its steel-making industry, possibly in the Pilbara region in the north west of the country. 

·      It is far too early to be certain but other steel makers that have been experimenting with partial use of hydrogen in existing coal furnaces, such as VoestAlpine in Austria, may soon conclude that it is better to shift to 100% low carbon rather than take intermediate steps that might result in perhaps half the CO2 saving. This is analogous to the car industry; why continue to invest in designing and making hybrids when the world is swinging so fast to fully electric autos?

After five years of small experimentation and promises to spend tens millions of dollars on carbon reduction, ArcelorMittal now says it will invest in a 2.6m tonne DRI plant costing a billion Euros, with the help of the Spanish government. This has worldwide significance.

[1] https://corporate.arcelormittal.com/media/press-releases/arcelormittal-signs-mou-with-the-spanish-government-supporting-1-billion-investment-in-decarbonisation-technologies

[2] I say ‘appears to be committed’ because Arcelor Mittal’s corporate material is usually particularly cleverly worded to avoid any absolute promise to take any particular route.