Soft Machines

The UK’s industrial strategy landscape was overturned again in the March budget, with the previous strategy (as described in the 2017 White Paper from Greg Clark, Business Minister in the May government, superseded by a Treasury document: “Build back better: our plan for growth”. Is this merely a “rebranding”, or a more substantial repudiation of the very idea of industrial strategy?

From what I can deduce, it is neither of these extremes – instead it reflects some unresolved tension inside government between two views of how industrial policy should be framed. In one view – traditionally associated with HM Treasury – the government should restrict itself to general measures that it is thought will promote productivity growth across the whole economy, resisting any measures that selectively one sector of the economy over another. This is often called “horizontal” industrial policy, in contrast to so-called “vertical” industrial strategy, in which particular sectors of the economy that are thought to be of particular importance are singled out for special support. The 2017 White Paper did signal some return to “vertical” industrial strategy, though we can see recent precursors for this going back to Mandelson’s return to the Department of Business, Innovation and Skills (as BEIS – the department for Business, Energy and Industrial Strategy was called then) in 2008, and in the continuing support for sectors such as aerospace, automotive and life sciences since then. It seems that the Treasury “Plan for Growth” marks a swing of the pendulum back towards a focus on “horizontal” industrial policy, though the signals remain somewhat mixed.

The biggest signal of a change of direction following the March budget was the abolition of the Industrial Strategy Council. This was a non-statutory body set up by BEIS to monitor and provide advice about the implementation of the Industrial Strategy, chaired by the Bank of England’s Chief Economist, Andy Haldane, and featuring a stellar array of economists and business people. The Industrial Strategy Council’s final annual report gives a great outline of what an industrial strategy should be – “a programme of supply-side policies to drive prosperity in and across the economy”, whose key ingredients should be “scale, longevity and policy co-ordination”. The regional dimensions of industrial strategy, they say, should be co-created with businesses and regional actors (as we’ve seen in the development of local industrial strategies). The report calls for the use of clear metrics to judge success by, but to look “beyond these “traditional” drivers of productivity to measures of social, human, and natural capital, as well as broader welfare impacts”. Naturally, the Industrial Strategy Council thinks it’s a good idea to have an independent – and preferably statutory – body to provide independent monitoring and advice. There’s an good summary in this FT article by Andy Haldane – UK industrial strategy is dead, long may it live.

Leaving aside the signals that winding up the Industrial Strategy Council might be sending, what’s the substance in the new Treasury document “Build back better: our plan for growth”?

I don’t find a lot to argue with in the diagnosis of the problems. The UK’s poor productivity performance since the global financial crisis is placed front and centre. A telling graph highlights the growing gap in productivity between the UK and France, Germany and the USA, while another graph (shown below) makes it clear that this isn’t just an abstract issue of economics – the stagnation of wages and living standards the UK has seen since the financial crisis closely tracks the productivity slow-down. The UK’s persistent regional disparities in productivity, about which I’ve written at length in the past, are highlighted, too, with the problem identified (correctly, in my view) as arising from “cities outside London not fully capturing the benefits of their size”. The level of analysis of the causes of these issues is somewhat more sketchy, with the Treasury ascribing the problem primarily to be persistent low investment in physical capital and skills.

A lost decade. UK Labour productivity and real wages since 2000. From HM Treasury’s Build back better: our plan for growth. Open Government License.

The new framework for Treasury industrial strategy is built on three “pillars for growth” – infrastructure, skills and innovation. This is classical “horizontal” industrial strategy, without a focus on any particular sectors. But there are priorities – three goals, each of rather different character.

The first of these is “levelling up” – a (commendable) commitment to “ensure the benefits of growth are spread to all corners of the UK”, tackling regional disparities in health and education outcomes, supporting struggling towns, and ensuring that “every region and nation of the UK [has] at least one globally competitive city, acting as hotbeds of innovation and hubs of high value activity”. The second is the 2050 net zero greenhouse gas target, where the stress is laid on the number of “green jobs” this will produce. The third priority is the post-Brexit one of “taking advantage of the opportunities that come with our new status as a fully sovereign trading nation” – as “Global Britain”.

The plans for building on these three pillars and three priorities remain vague – some existing commitments are reasserted, such as the plan to increase public infrastructure spending, to meet an total R&D spending target of 2.4% of GDP, to deliver the FE White Paper, to introduce the new science funding agency “ARIA”, and to introduce “Freeports”. Further details are promised later, including an Innovation Strategy and the R&D Places Strategy.

The reduction of emphasis in this document on the sectors that were so prominent in the previous industrial strategy – such as aerospace, automotive, and life sciences – has clearly caused some anxiety in business circles. There’s been a response to this, in the shape of a joint letter from BEIS Secretary of State Kwarteng & Chancellor of the Exchequer Sunak. This emphasises continuity with the previous industrial strategy, asserting that the new plan “builds on the best of the Industrial Strategy from 2017 and makes the most of our strengths right across the economy”. They promise that “this government remains committed to its industrial sectors” and that the existing sector deals (e.g. for Aerospace, Automotive and Life Sciences) will be honoured. And there is the promise of more in the future – “we will follow up the plan for growth with an Innovation Strategy, as well as strategies for net zero, hydrogen and space” and “we will also develop a vision for high-growth sectors and technologies.”

There’s another indication that the words “industrial strategy” may not yet be completely unspeakable in the current government – shortly after the budget, the Ministry of Defense published their “Defense and Security Industrial Strategy”. I think this is positive – another way of creating some priorities in industrial strategy without entirely going down the sector route is for the government to focus on strategically important, long-term goals of the state, and systematically to evaluate what innovation is required and what industrial capacity needs to be built to deliver those goals.

What other goals should be pursued besides defense? The obvious two are net zero and healthcare. As a matter of urgency, the government should be developing a long-term Net Zero Industrial Strategy, to accompany a more detailed road-map for the huge job of transforming the UK’s energy economy. And as we recover from the pandemic, there needs to be a refocused Healthcare Industrial Strategy, building on the successes of the old “Life Sciences Strategy” but focusing more on population health, and learning both the positive and negative lessons from the way the UK’s health and life sciences sector responded to the pandemic. The lately departed Industrial Strategy Council produced a very helpful paper on the lessons that industrial strategy should learn from the state’s involvement in the development of the Oxford/AstraZeneca Covid-19 vaccine.

What would worry me most if I were in government is time. “Developing visions” is all well and good, but budgets are now set until 2022, and there are suggestions of funding “pauses” in some parts of the existing industrial strategy, such as the industry research and development supported by the Aerospace Technology Institute. If new programmes are to begun in 2022, they will take time to ramp up. Meanwhile other dates will be creeping up – 2027 is the date for the 2.4% R&D target, which needs the private sector to make decisions to commit substantial extra funds to business R&D in response to any increase in government R&D. And although 2050 seems far away now for the net zero greenhouse gas target, the scale of the transition and the lifetime of the assets, and the need for innovation to bring down the cost of the transition, means that the next ten years is crucial.

Not least, the latest date the government can hold an election is the end of 2024. Having repealed the Fixed Term Parliament Act, the government will probably want to use the regained flexibility to hold the election as much as a year early. There are some who say that this is a government that likes to mark its own homework. Ultimately, though, the homework will be marked by the voters. The government has raised high expectations about a return to economic growth and a levelling up of living standards, especially in the so-called “Red Wall” seats of the Midlands and the North. There’s not a lot of time to demonstrate that the country has even started on that journey, let alone made any substantial progress on it. So whatever the government has decided is the future of industrial strategy, it needs to get on with it.

A couple of weeks ago, on 9 October 2020, I took part in a seminar for the Tony Blair Institute for Global Change, called “UK Industrial Strategy’s three horsemen: COVID, Brexit and trade wars”. This featured as speakers me, the economist Dame Kate Barker, and Anand Menon (Director of “UK in a changing Europe” at Kings College London), and was chaired by Ian Mulheirn. There is a YouTube video of the event here. Here is a slightly tidied up version of what I said.

It’s a real pleasure to be speaking at this event – and especially to be sharing a virtual platform with Kate Barker, from whom I learnt so much as a colleague working on the Industrial Strategy Commission back in 2017. Our final report then was intended to inform the discussion around the 2017 White Paper on industrial strategy. Now industrial strategy is back on the agenda – we read that the the government is planning to “rip up” the 2017 strategy, producing a new document with a heavy focus on science and technology.

Despite everything that’s happened since 2017, I agree with Kate that the principles we laid down do stand the test of changing times. Since then, the focus of my own work has been on the link between R&D, innovation and productivity, and the way regional imbalances in economic performance reflect regional imbalances in state spending in R&D.

But how is this all changed by the three horseman of the apocalypse – COVID, Brexit and trade wars – that we’re asked to discuss?

Brexit

We can talk about the link between Brexit and industrial strategy both in terms of cause and effect. Failures of industrial strategy contributed to the political conditions that led to Brexit, and the changes that Brexit will force on the UK’s economy will demand a different industrial strategy for the new economic model that the country will have to adopt.

Much written on the connection between “left behind communities” and the Brexit vote, and the relationship is possibly more complicated than simple accounts might suggest. But, as the economic geographer Philip McCann has demonstrated in his analysis of “geographies of discontent” , the UK is an outlier amongst developed countries in the scale of its economic imbalances. The greater Southeast looks like a prosperous Northern European country. The rest of the country looks like East Germany, Southern Italy or Portugal. In fact, East Germany has recovered from 40 years of communism faster than the North of England has recovered from the deindustrialisation of the 1980’s.

These regional imbalances are reflected in living standards and other measures of prosperity, including life expectancy and health outcomes. But at the root of the issue is a huge imbalance in productivity. In fact, the imbalances show up more strongly in productivity than in living standards, because the UK runs an effective transfer union – money is moved up from the greater Southeast – the only parts of the country to make a surplus on the government current account – to the rest of the country.

But the paradox is that while we transfer money to cover current spending, we concentrate the investments that build productivity growth in the already prosperous greater southeast. My own focus has been on Research and Development: here nearly half the public spending on R&D is concentrated in London and the two two subregions containing Oxford and Cambridge. R&D isn’t the only thing that matters in driving productivity growth, but it is associated with high value companies operating at the technological frontier and innovative start-ups, which anchor strong regional innovation ecosystems and produce highly prosperous knowledge intensive economies like those around Cambridge and Oxford.

Even more paradoxical is the fact that public spending on R&D is even more geographically concentrated than private sector spending. This means that potential spillovers from private sector R&D spending are being left uncaptured. We have regions like the North West, with highly productive, R&D intensive industries such as chemicals and pharmaceuticals, and the East Midlands, with its strong private sector innovation in automotive and aerospace, where the innovation potential of the regional economies isn’t being exploited to the full because the public money doesn’t follow the private.

On the other hand, there are some places that don’t have enough R&D of any kind, public or private. In Wales and the Northeast, for example, low investment in R&D leads to weak innovation economies, with poor productivity performance and low demand for skills

It is failures of industrial strategy that have led to a divided country, and those divisions have brought us sour politics.

Trade Wars

Moving onto the effects of Brexit – and the wider sense of a retreat from globalisation – we’re likely to find that the economic model that the UK has chosen is likely to be particularly threatened. Some of our highest productivity and most export oriented sectors have succeeded by becoming highly integrated in transnational value chains, and it is these sectors that are most at risk from the trade dislocations that Brexit threatens.

The UK’s automobile industry is a prime example. This has made a remarkable comeback from a low-point in the mid-2000’s – perhaps an unsung success of a modern industrial strategy which began with Mandelson’s time in the Business department at the end of the new Labour period, and persisted into the Coalition, with considerable policy continuity. But a finished car that rolls off a production line in Sunderland or Solihull combines components and parts that have been shuffled back and forth from a network of suppliers all across the world. This leads to efficient car production, but it’s going to be very difficult to adapt to a post-Brexit world where there are likely to detailed rules on local origin for the export of vehicles to the EU.

Brexit isn’t the only event that’s likely to give us hard lessons about technological self-sufficiency. We can see the effects of a much colder attitude to China in the USA in the pressure on the UK to lessen the involvement of Huawei in the 5G network, and the exclusion of the UK from the EU’s Galileo project for a satellite positioning has resulted in a scramble for a UK alternative. I suspect that politicians and policy makers severely underestimate the degree to which the UK has lost technological self-sufficiency in a whole range of sectors. I also wonder whether a wider sense of loss of what one might call technological sovereignty has itself contributed to the anxiety that culminated in Brexit.

I believe that the UK will need to rebuild some of its technological capacity if it is to remain a prosperous country, and that this will need to be a central ingredient of a more activist industrial policy. But that leaves some big questions. How much, in what sectors? The UK is a relatively small country in a big world economy. We will need to think very deeply about our place in the evolving trading systems of a world that might look very different to the post-cold-war, globalising world that policy makers have grown up in. An industrial strategy does need to be founded on a clear view about what kind of economy the UK wants – and can realistically hope – to become.

COVID-19

We’ve known for some time that increasing globalisation puts the world at greater risk of a pandemic, and with COVID-19 those fears have been realised. The closer entanglement of natural ecosystems with human society leads to more pathogens crossing from the animal world into people; then the worldwide traffic of business people and tourists spread the disease across the world before health systems have a chance to respond, as we have seen with such tragic consequences over the last year.

It’s too soon to unpick all the effects COVID-19 will have on the economy, and the implications that those effects have for the UK’s industrial strategy, but we can already start to see some themes emerging. Different sectors have been affected in different ways, with an obvious severe (but hopefully time-limited) blow to hospitality and tourism, and perhaps more far-reaching effects on commercial real estate as some pandemic induced changes in working practises are permanently adopted.

One very important question for the UK concerns the shape of the future civil aerospace industry. It’s difficult to know how future patterns of international mobility will change, but any permanent reduction will have a serious impact on one of the UK’s highest productivity industries. As a specific example, Rolls-Royce is one of the UK’s few world class innovative engineering companies of any size, but its dependence on long-haul air traffic makes the company – and the cities like Derby that depend on it – very vulnerable.

Rolls-Royce has been bailed out by the government once before, following its bankruptcy in 1971. I believe that letting Rolls-Royce fail now should be unthinkable, because of the dissipation of concentrations of high level skilled people, and the loss of innovation capacity in areas like the East Midlands that would follow. But what form should any bail-out take – and how should it take into account bigger imperatives such as the net zero greenhouse gas target?

What can we learn about our industrial strengths and weaknesses from the experience of our response to the pandemic? We went into the pandemic thinking we had the advantage of a world-class life sciences sector, but after the event we can’t say the UK has excelled in its response.

It’s certainly true that parts of our life sciences sector are excellent, and if the Oxford group or the Imperial group produce an effective vaccine against covid-19, and if the pharmaceutical industry is successful in rapidly scaling up its manufacture, that will be a huge achievement and an invaluable contribution to world health.

But in other important areas – in public health, diagnostics, the care sector – the UK’s weaknesses have been savagely exposed. We have learnt about weaknesses in supply chains for basic supplies like PPE and generic pharmaceuticals.

I think we made a category error in thinking that the “life sciences sector” is a sector at all. We have a strong pharmaceutical sector which historically been enormously productive (though not without recent difficulties), exploiting the UK’s excellent biomedical science base to produce drugs for the most lucrative world markets (particularly those of the USA). But we have done much worse in driving and implementing the kinds of innovation that serve the health needs of the UK’s own population.

The fourth horseman

There is, of course, a fourth horseman of the industrial strategy apocalypse, that is more important than any of the three we have been asked to discuss. That is climate change and the huge economic transition that the need to decarbonise our energy economy requires. It is an entirely positive development that the government has committed to a target of net zero greenhouse gas emissions by 2050, and that there is a wide political consensus in support of this target, or indeed a more ambitious goal. But I’m not convinced that policy makers and the public fully understand the magnitude of the task.

We need not just to decarbonise the electricity sector as it stands now; as we electrify other forms of energy use, we will need to at least double generating capacity. We need to decarbonise transport and domestic heating, probably using hydrogen as an energy store and vector, especially for hard to decarbonise industries like steel. We will need as much offshore wind as we can get (probably including new technologies like floating wind), we will need new nuclear build, possibly including new high temperature designs to make hydrogen from process heat.

This energy transition will be a huge dislocation. We have to do it, but we shouldn’t expect it to be without cost. People rightly talk about the new “green” jobs this transition will produce – but that’s not an unmixed blessing. If the new energy systems need to employ more people than our current fossil fuel based system, that implies a drop in productivity. We will have to apply more resources to achieve the same energy benefits, and those resources won’t be available to satisfy other wants and needs. Innovation, to create new zero carbon technologies and improve existing ones, will be urgently needed to drive down those costs, and that innovation – carried out in parallel with the deployment of existing technologies – should be a priority of industrial strategy.

The energy transition does have potential benefits for regional economic inequality, though. Much of the innovation and deployment of low carbon technologies should happen outside the prosperous Southeast – for example in Teeside and the Humber, in Cumbria and the Wirral. This should be an important part of the “levelling up” agenda.

An industrial strategy for our times

To sum up, the ravages of these four horseman mean that our economy will need to be transformed. That transformation needs to be driven by innovation, and it needs to be informed by a clear view of the enduring challenges the UK faces and a realistic assessment of the UK’s place in the world. As Kate stressed, the challenges are obvious: climate change, weak wage growth, the cost and effectiveness of health and social care, failing places. The need for a new start does give us a chance to spread the benefits of innovation more widely across the country, and we should seize that opportunity.

I did a webinar a couple of weeks ago, for the Society of Chemical Industry, about the role of the chemicals industry in addressing the UK’s problems of stagnant productivity and regional economic disparities. The recording of the talk should be on their website soon, but in the meantime here (5 MB PDF) are the slides I used. Here’s a summary of what I said.

I started by setting out the economic context the UK finds itself in. The very slow productivity growth since the 2007/8 global financial crisis has had the result that real wages have stagnated, while economic performance across the regions of the UK remains very uneven.

The most important contributor to productivity growth – and thus to rising living standards – is what economists call “total factor productivity” – the measure of how effectively an economy converts inputs, in the form of labour and capital, into valuable outputs. This includes, but is not limited to, the technological advances that allow us to produce existing products more efficiently and to create entirely new products and services.

We can thus map the different sectors of the UK’s economy on 2 dimensions – how big a share of the economy they take, and how much their total factor productivity increases. I argue that industrial strategy should focus on those areas that are both significant in scale relative to the economy as a whole, and that are dynamic in terms of showing long-term increases in total factor productivity. The three crucial sectors in the UK economy by these measures are knowledge intensive business services, information and communication technologies, and manufacturing. Within manufacturing, transport equipment – automotive and aerospace – stand out, but chemicals and pharmaceuticals are also highly significant.

Cumulative growth in total factor productivity in selected UK sectors and sub-sectors, indexed to 1995. Data from EU KLEMS Growth and Productivity Accounts database.”

Looking at the changes in total factor productivity over the last couple of decades offers an instructive window on the way the UK’s economy has changed.

Because normally manufacturing to grows productivity faster than services, we’d usually expect total factor productivity in the manufacturing sector to grow faster than the whole market economy. In the UK, that wasn’t so in the mid-1990’s – manufacturing lagged behind the economy as a whole. But from 1998 to the global financial crisis, manufacturing TFP grew faster than the economy as a whole; since the crisis both have stagnated.

Part of the explanation for this comes from the figures for the financial services industry. This showed very fast growth in the late 1990’s, booming right up to the financial crisis – since when it has fallen precipitately. It’s at least possible that some of the apparent boom was due to the way value is measured – or mismeasured – in financial services, but it’s clear that this sector, so influential politically, has been a drag on the whole economy over the last decade.

Focusing on manufacturing subsectors, transport equipment – including automotive and aerospace – stagnated in the late 90’s, began a recovery in the 00’s, which took off dramatically after the global financial crisis. It’s intriguing that the timing of this recovery almost exactly coincides with the UK government’s rediscovery of industrial policy – with an initial focus on automotive and aerospace industries. Pharmaceutical total factor productivity boomed from the late 90’s to the end of the 00’s, then collapsing, for reasons I’ve discussed extensively elsewhere.

But the surprise – to many, I suspect – is the performance of the chemicals sector. Written off in the late 90’s as the “old economy”, the chemicals industry has delivered the steadiest gains in total factor productivity, its cumulative performance exceeding both financial services and pharmaceuticals.

What’s more, if we look at where the chemicals industry takes place, in the context of regional economic inequality and the “levelling up” agenda of the government, we find that it is located outside the prosperous southeast, in Northwest England, the Humber and Teeside.

What sectors should industrial strategy focus on? My criteria would look at relative scale, the potential to produce significant and sustained gains in total factor productivity, and to contribute to economic growth in economically lagging parts of the UK. The chemicals industry qualifies on all counts.

What, though, of the future? Economic statistics don’t capture some of the costs of the chemicals industry, but these costs are borne by society more widely. The feedstocks it uses may be unsustainable and deplete the planet’s natural capital, pollution may damage local environments and ecosystems. Improper disposal of products – like plastic packaging – at their end of life causes yet more environmental damage.

Perhaps most importantly, the energy the industry uses produces carbon dioxide and thus accelerates climate change. 3% of the UK’s greenhouse gas emissions are directly associated with the chemicals industry, which accounts for about 20% of all emissions associated with manufacturing.

There is another side of the ledger, too. The products of the chemicals industry – like batteries and fuel cells – will be crucial in decarbonising the economy. In the future we might see the widespread use of hydrogen as an energy vector, direct capture of carbon dioxide from the air, and the synthesis of hydrocarbons from green hydrogen and captured carbon dioxide for zero-carbon aviation. Much of the net zero agenda is in fact a chemical industry agenda.

We need an industrial strategy for the UK chemicals industry, justified by its scale and its record of steady total factor productivity improvement. It’s a pity that the government hasn’t responded to the Chemistry Council’s proposed Sector Deal, which would provide a good start. In addition to a focus on productivity growth, that strategy should have a regional element, building on the existing chemical industry clusters in the North West and North East with further interventions to promote innovation and skills and all levels. Above all, it should emphasise the important role and responsibility of the chemicals industry as part of the wider economic transformation that needs to take place to achieve the government’s 2050 Net Zero emissions target.

The good news from the latest release of the UK government’s energy statistics is that the fraction of electrical power generated from renewable sources in 2019 reached a record high of 37.1%, driven largely by an increase in offshore wind of 20%, to a new high of 32 TWh a year. The bad news is how little difference this makes to the UK’s overall energy consumption – of the 2300 TWh used, 78.3% was obtained from burning fossil fuels. This is a decrease from last year’s fraction – 79.4% – but progress remains much too slow.

It’s tempting to focus on the progress we are making in decarbonising the electricity supply, and this isn’t insignificant. But while the UK used 346 TWh of electricity in 2019, the country directly burnt gas to provide 512 TWh heat for domestic and industrial purposes (not counting here the gas converted to electricity in power stations), and 152 TWh of petrol and 301 TWh of diesel to power vehicles. We’ve no chance of reaching net zero greenhouse gas emissions by 2050 without displacing this directly burnt fossil fuel contribution. And given the longevity of energy infrastructures, we haven’t got long to start building out the technologies to do this at scale.

Can hydrogen help? This technology – or more accurately, group of potential technologies – is having a moment of attention, not for the first time. I think it could well make a significant contribution, but there are some awkward choices to make. Implementing any use of hydrogen in our energy system at scale will involve massive, long-term investments, and making the right choices involve difficult economic judgements, not just about the technologies as they currently exist, but as they may evolve under the pressures of energy markets across the world. Of course that evolution can be steered by incentives, regulation, and targeted support for research and development.

To begin with the basics, because there aren’t any reserves of molecular hydrogen lying around, it isn’t a source of energy, but a way of storing, transmitting and using energy. When burnt, or combined with oxygen in a fuel cell, it produces nothing but water. So the issue is how you make it without producing carbon dioxide in its manufacture. There are three broad options:

Currently, most hydrogen is made from natural gas through a process called steam methane reformation. By adding heat to water and methane, with suitable catalysts, one can obtain hydrogen and carbon dioxide. The carbon dioxide produced in the reaction, and any that results from generating the heat needed to make the reaction run, would need to be captured and stored underground in old gas fields. This process, including separating the carbon dioxide, are mature technologies, used for example at scale to produce ammonia for fertiliser.

If zero-carbon energy is available cheaply, from wind, solar or nuclear, intrinsically zero carbon hydrogen can be produced by electrolysis of water. The most effective current technology uses a proton exchange membrane to separate anode and cathode.

If zero-carbon process heat is available cheaply, from high temperature nuclear reactors or solar concentrators, hydrogen can be made by the thermochemical splitting of water. (As a combination of the last two ideas, given both process heat and electricity, high temperature electrolysis is another option).

How then might the hydrogen be used to attack the carbon dioxide currently produced by the nearly 1000 TWh of energy we derive from burning gas, petrol and diesel for heating and transport?

Right now we could add some hydrogen to natural gas – perhaps up to 20% – making a significant lowering of its carbon intensity without substantial changes in our existing systems.

The complete replacement of natural gas by zero-carbon hydrogen for domestic heating and many industrial processes is probably technically feasible, but quite a lot more expensive. Some changes will need to be made to the gas distribution system (e.g. replacement of iron/steel pipes with thermoplastic pipes), and boilers and appliances would probably have to be replaced too.

Hydrogen can be used for transport, as fuel for internal combustion engines, or more likely, converted to electricity via fuel cells to power cars and trucks.

Finally, hydrogen might make possible the very large scale seasonal storage of energy (potentially on the scale of 10’s or even 100’s or TWh) generated by intermittent renewables, by storing it underground in rock salt formations.

All of these ways of making hydrogen and using it are technically possible. They’re also all potentially enormously expensive, with the potential for locking the country into solutions which turn out to be inappropriate or made redundant by rival technologies. Some experimentation is necessary, and some blind alleys are probably inevitable, but what needs to be taken into account as we make our choices?

To start with the basic physics and chemistry, hydrogen is a light gas which burns completely and cleanly to yield only water vapour. Perceptions of hydrogen are inevitably shaped by the Hindenburg disaster – but all flammable gases are potentially dangerous, and these are risks of the kind that industrial societies have got used to managing. Hydrogen is more easily set aflame than methane and it burns hotter, but on the other hand at atmospheric pressure burning a given volume of hydrogen produces less energy than the equivalent volume of methane, and much less than petrol vapour. In fact it’s this low volumetric energy density of hydrogen that poses the biggest problem. Even compressed to 70 MPa (as it would be in typical compressed gas tanks) its energy density is only 1.3 MWh per cubic meter, compared to petrol or aviation spirit eat about 10 MWh per cubic meter. Even liquified its energy density is still only 2 MWh per cubic meter, and this needs a temperature of -250 °C, considerably colder than liquid nitrogen.

Moving on to economics, how can we find the most cost-effective solutions? The problem is that technologies don’t stand still – indeed, it’s essential that costs come down, and substantial research efforts are needed to make sure that happens. Where can we hope to see the biggest cost reductions? Existing technologies – like steam reforming of natural gas with carbon capture – are probably the cheapest options with current technology, but being mature further improvements are likely to be more difficult to find than with newer technologies like proton exchange membrane or high temperature electrolysis.

It’s important to remember that the UK accounted for just 1.4% of the world’s energy consumption in 2018, and this fraction will inevitably (and desirably) fall over the next few decades. The choices we make must take into account what the rest of the world is likely to do; while the UK might hope to influence that path, perhaps by helping develop new technologies cheap enough for wide adoption, the UK isn’t a big enough market to be able to make unilateral decisions about technology directions. If battery electric vehicles win the race for zero-carbon personal transport, it would be pointless for the UK to develop a hydrogen network for fuel cell cars. Likewise, if the UK is the only country to back hydrogen boilers for domestic heating while the rest of the world chooses electric heat pumps, it won’t be a big enough market to justify the development of hydrogen domestic boilers by itself, so its plans would be left high and dry.

We have well developed existing energy distribution systems, so the question for any new energy vector is whether these systems can be incrementally adapted, or do new ones need to be built out entirely from scratch? We currently have a well developed electricity distribution system. Distributed PEM electrolysis plants could take zero-carbon from the grid, and produce hydrogen locally. We also have systems for distributing natural gas: it’s likely that the core high pressure network would have to be entirely rebuilt for hydrogen, but the low pressure local distribution system could be adapted. We don’t have a cryogenic liquid distribution system at scale, and this is likely to limit global trade in hydrogen.

Finally, we have to consider our plans for low carbon electricity. Whatever we do, we need to replace the 512 TWh of gas we use for heating, and the 453 TWh of petrol and diesel we use for transport, with zero carbon alternatives. If this involves electrification – either directly or through the production of hydrogen from zero-carbon electricity – this will need a huge expansion of power generation capacity from the current 346 TWh/yr. I find it difficult to see how this can happen without both a massive increase in offshore wind – possibly including floating offshore wind – and new nuclear build, possibly next generation nuclear able to produce high temperature process heat for production of additional hydrogen.

These are difficult choices, but we haven’t got much time. Let’s get on with it!

Some references:

Current UK energy statistics from DUKES 2020.
Hydrogen supply chain evidence base.
On hydrogen storage (US Dept of Energy PDF) –
Royal Society Policy Brief Options for producing low-carbon hydrogen at scale.