Keeping the lights on

Humans are making a mess of energy and endangering their planet, says Walt Patterson of London’s Chatham House. In his new book (excerpted here), he outlines a different way to think about energy and power issues, particularly electricity.

(Excerpted with permission from Keeping the Lights On: Towards Sustainable Electricity, by Walt Patterson; Earthscan/Chatham House 2007. ISBN 978-1-84407-456-3)

The electricity story used to be a documentary, based on fact. Now it looks more like fantasy, wishful thinking, out of touch with reality. The reality today is that one-third of us don’t have electricity at all. Those who do have trouble keeping the lights on. The International Energy Agency estimates that electricity will require investment of $10 trillion by 2030 — more than one thousand billion dollars every three years. But the past decade has cost many in the electricity business their jobs, their shirts or their companies — hundreds of billions of dollars of losses already. Future electricity investment could now be so risky it might not happen. The main technologies of traditional electricity — large dams, coal-fired and nuclear power stations, and overhead transmission lines — are long since in major trouble, financial and environmental. Yet traditional electricity diehards are now trying to stampede us into more of the same, to make matters worse. The electricity story could become a horror story. We need to rewrite it, urgently.

To get the story right we have to get the premises right. To start with, we have to get electricity itself right. Reiterate the essentials; we should now take them for granted. Electricity is not a fuel. It’s not a commodity. It’s a process, occurring simultaneously and instantaneously throughout an entire interconnected circuit. A process cannot be stored. A fuel such as coal, oil or gas comes out of a hole in the ground at a particular place. If you want to use it anywhere else, you must carry it there. But you can start the electricity process anywhere, in an extraordinary range of ways, from vast to minute. Electricity exists only in the infrastructure of assets that generate, deliver and use it, and through which it flows. Electricity is a function of infrastructure. Understanding this is the key to the necessary changes. You can produce and use electricity without fuel, but not without infrastructure. The flow of electricity through the infrastructure is easy to measure; but the price of a unit of electricity is ultimately arbitrary. The so-called “electricity market” is illusory. The price of a unit depends not only on the price of any fuel involved, but on asset accounting, taxation, regulation, risk, subsidies, network and system effects and other factors usually unmentioned. The arbitrary price of an ephemeral kilowatt-hour is not an adequate basis for the requisite finances, transactions and business relations.

Electricity means infrastructure

Instead of a quasi-commodity market, we should deal explicitly with the physical assets of the system — generators, networks and end-use technologies. What matters is this infrastructure — who owns it, who has access to it, who uses it and on what basis. What we need is not batch transactions in a quasi-commodity, but contracts for services.

To deliver electricity services more reliably and sustainably we need not only to upgrade the electric infrastructure, all of it, especially the end-use technologies, but to transform it. This is where the new story starts. Traditional electricity is based on a technical model dating back more than a century, to when the best available generating technologies were based on water power and steam power. Economies of scale in generating with these technologies shaped the model. As a result, all over the world, a century later, we still generate electricity in large remote central stations as synchronised alternating current, and deliver it to users over a network including long high-voltage transmission lines. The network is essentially radial and one-way. It also has to divide up the electricity to distribute it to loads mostly thousands to millions of times smaller than the generators.

This centralised configuration used to make sense. It no longer does. Consider some of the obvious drawbacks. Most central-station generators operate either intermittently or at only partial load most of the time, misusing costly assets. Fuel-based central generators waste two-thirds of the fuel energy before it even leaves the power plant. On many systems, line losses cost another significant fraction. The configuration is inherently vulnerable to disruption, by mishap or malfeasance, over a wide area and almost instantaneously. Traditional electricity assumes that every load is essentially equivalent, requiring the same high quality of electricity. This is akin to our absurd water-management policy, in which we purify water centrally to drinking-water quality, and then use most of it for flushing toilets, washing cars and watering lawns. In the same way, we produce high-quality electricity as required by sensitive loads, then use much of it for undemanding services such as heating and cooling. Most electrical loads, moreover, are inherently intermittent or variable; but large fuel-based generators are inherently inflexible. Traditional arrangements are almost a total mismatch.

Anyone who looks dispassionately at traditional electricity has to think, “We must find a better way to do this”. Fortunately, now we can. […] [T]he catalogue of innovative generating technologies already or soon to be available extends far beyond steam and water power, with very different attributes. Wind turbines, microhydro, biomass generators, photovoltaics, gas engines, microturbines, fuel cells, Stirling engines and microcogeneration all exhibit economies not of unit scale but of series manufacture. The more we make, the cheaper they get. We can often locate these small-scale generators close to loads, even on site, dramatically altering network requirements and operation. Instead of a radial one-way network, a decentralised system would have a two-way meshed network, with loads and generators of broadly comparable sizes more or less uniformly distributed across the system. Monitoring and control technologies now indeed offer the possibility of completely self-stabilising systems, in which loads and generators talk to each other continuously and react accordingly.

By moving toward innovative decentralised electricity, we can tackle directly the most serious shortcomings of traditional centralised electricity. But first we have to explain and clarify this new story to policymakers. Ignore the reported “cost of generation” by different means. It usually claims that traditional large-scale remote fossil-fired generation is “cheaper” than smaller-scale renewable or cogeneration closer to loads. As noted earlier, such comparisons, in fractions of a penny per unit, with no qualification as to the accounting or financial framework, tax treatment, subsidies, risks, system and network effects or other essentials, including environmental effects, are meaningless. They should have no influence whatever on policy. The corollary bears repeating until it is second nature. Policy determines costs — not the other way round. That indeed should be the aim of electricity policy, sensibly and coherently developed.

The whole story, the whole system

Consider, for example, taxation. If we treat electricity as a commodity, taxation applies only to the unit price, and to batch transactions in measured amounts. Even for fuel-based electricity this is unsatisfactory. For what we might call “infrastructure electricity”, such as wind, hydro or solar, such tax treatment misses the point completely. If, however, electricity is treated correctly, as an infrastructure issue, tax policy should focus not on flows of electric current but on taxation of assets in electricity infrastructure — all assets, explicitly including end-use technologies and the buildings that contain them. To upgrade electric infrastructure, to improve performance and reliability of services and reduce unwanted side-effects, differential asset taxation is key.

Until recently, such tax policies have been fragmentary, tentative and ad hoc, hardly recognised as energy policy. Now, however, innovative electricity, including small-scale decentralised generation close to loads, offers more cogent reasons and more attractive opportunities to integrate and optimise entire local systems, including both generation and the technologies it drives. End-use technologies — lights, heaters, motors, freezers, electronics and so on — and even passive infrastructure such as buildings are part of the system that delivers comfort, illumination, motive power, refrigeration, information and all the manifold electricity services we take for granted. Upgrading end-use technologies is the most effective way to deliver better services more reliably at lower cost and with lower impact. But most electricity users don’t know or care enough to do anything about it. Worst still, companies whose business is selling or delivering units of electricity want us to use more, not less. Inefficient lamps and motors benefit their cash flow. More appropriate tax treatment of electricity assets, especially end-use technologies and buildings, can provide a potent incentive to invest in upgrading infrastructure. It can give the incentive specifically to those whose business it is to deliver better services.

Users do not want reliable “electricity”; they want reliable electricity services. With local generation, under local control, driving local technologies, the responsibility for keeping the lights on can be similarly local and coherent, and accordingly much more manageable. Moreover, this responsibility can be the focus of well-defined contracts between those enjoying illumination and comfort and those providing them, a more stable, less nerve-racking business than competing to sell ephemeral units of electricity.

This story feels much more coherent and convincing. Regard electricity not as a commodity but as a process delivering services. Improving the whole process benefits reliability and quality of service, while reducing vulnerability to disruption. It also offers the potential to shift progressively away from fuel-based to infrastructure generation, a key to sustainable electricity. If we can set this in train for electricity worldwide, we may eventually begin to recognise that all energy services, even including transport, are not commodities but processes. The challenge is always to optimise the entire process — an inherently positive undertaking for human society.

How do we get there from here? We have to start changing the way society thinks about electricity, and about energy. Everything else follows. That’s why we have to get the story right. That’s where those of us who are policy academics, analysts and commentators come in — right at the beginning, right here, right now.


Real energy policy must deal explicitly with buildings as well as all the other technologies — not just the technologies that produce and deliver energy carriers, but also the technologies that deliver the energy services we actually want. Real “energy policy” is not just about commodity fuels and electricity. It is about infrastructure. Moreover, it is not just about so-called “energy supply” infrastructure, such as refineries and power stations. It is about what we should explicitly and emphatically call the “energy service” infrastructure.

Process and infrastructure

We see that most clearly when we consider electricity. At the beginning of the 1990s, when the UK government liberalised electricity, it imported the language and concepts of the hydrocarbon industry. It made electricity into a quasi-commodity, bought and sold like barrels of oil, in batch transactions in a competitive “electricity market”. The objective of introducing competition was to make the price of a unit of electricity, a kilowatt-hour, as low as possible. It still is, as if this were what users wanted. It is not. What users want is a low electricity bill – not the same thing at all. What users also want is reliable services – keeping the lights on. We now know, as we might have anticipated, that a low unit price and reliable electricity may not be compatible.

No matter what free-market ideologues may say, electricity is not and never can be a commodity like oil or gas. Electricity as we use it is a process, taking place simultaneously throughout a complete system of physical assets – generation, network and loads. You can have electricity without fuel, but not without infrastructure. Electricity exists only in infrastructure. We must shake off the misguided belief that electricity policy can be treated as commodity policy, and electricity business as commodity business. Electricity policy should be first and foremost infrastructure policy – that is, policy to guide and shape the evolution of the physical assets that use electricity to deliver the services we want.

We already have an array of policy levers available to guide, shape and upgrade our energy service infrastructure. We should begin by recognising explicitly that they are energy policy levers, more important than any policy levers that focus purely on fuel or electricity or their price per unit. We keep hearing that only “higher unit prices” will make us improve our use of energy. That is simply nonsense. Minimum performance standards, planning requirements and government procurement are all potent policy levers to improve infrastructure. Perhaps most potent of all is asset taxation, the tax treatment of buildings and other energy technologies. But tax policy worldwide still stubbornly persists in a fundamental flaw […]. If you invest in an asset to sell its measured output, say a refinery or a power station, government considers it a business asset, and taxes it accordingly. If, however, you invest in an asset to deliver an energy service, say a well-insulated, comfortable home, it is not a business asset. You don’t measure or sell by the unit its output of “comfort”. In general, governments treat such investments much less generously for tax purposes than business investments whose outputs are to be sold. That one single assumption skews all investments in energy infrastructure, in favour of investments to supply saleable energy carriers and away from investments to deliver better energy services. Of course we do sometimes get grants, tax breaks, rapid write-offs, or other financial and fiscal benefits for improved end-use technologies. Unfortunately, however, such measures tend to be ad hoc, short-term, inconsistent and incoherent. They have never been subsumed into coherent policy for energy infrastructure. They should be, as a matter of urgency, starting immediately.

Energy security means reliable services, not just fuel and electricity. The best way to minimise vulnerability to price rises or power cuts is to minimise dependence on fuel and electricity, by upgrading the infrastructure. The UK government, for example, has tens of thousands of buildings all over the country that are its own responsibility — everything from offices to prisons. Suppose it decided to upgrade all these buildings — to ensure they were solidly built and adequately insulated, with high-performance lighting, heating, ventilation, and electronics, and indeed with on-site generation and cogeneration of their own heat and power. Such an undertaking would entail major pump-priming contracts to energy-service companies. It would boost skilled employment all over the country. It would cut the unit costs of the necessary technologies, by tooling up for expanded use. It would reduce the vulnerability and improve the reliability of all the energy services provided. It would demonstrate dramatically the vast potential for improvement. It would be a spectacular international public-relations coup. What’s more, it would save taxpayers money. Imagine what such an approach could accomplish over the entire European Union, in OECD countries generally, or indeed anywhere in the world. Sustainable government procurement, not just of fuel and power but rather of energy services, could bring about a historic change of direction, towards real energy policy.

Real energy policy also points the way to real climate policy. You can put it like this:

•Climate is an energy issue.

•Energy is an infrastructure issue — not a commodity issue, an infrastructure issue.

•Therefore, climate is an infrastructure issue.

To get climate right, we have to get energy right. To get energy right, we have to start by changing the way we think about it. We have to explain this change, with its profound and exciting implications, to the politicians, financiers and journalists who still fail to understand it. If we are serious about tackling energy security and climate change, if we are serious about keeping the lights on, we have to start by getting energy right.

(Excerpted with permission from Keeping the Lights On: Towards Sustainable Electricity, by Walt Patterson; Earthscan/Chatham House 2007. ISBN 978-1-84407-456-3)

Walt Patterson is an associate fellow of the energy, environment and development programme at Chatham House (the Royal Institute of International Affairs) in London. His personal website, Walt Patterson on Energy, can be viewed here.

Copyright Walt Patterson/Royal Institute of International Affairs, 2007

•chinadialogue readers who would like to purchase a copy of "Keeping the Lights On" at a special discount may do so via Earthscan website here.

•While the website already offers a 10% discount on the price of the book, chinadialogue readers can receive a further 10% off.

•Here’s how to get it: After entering your credit-card payment details on the "Review and Confirm Your Order" page of Earthscan’s site, simply type the code R72 into the "Additional Notes" box beneath your card details. The additional discount will not appear in the "Amount" on your screen, but will be applied nonetheless and will be noted on the final invoice accompanying your book.