Welcome to European Tribune. It's gone a bit quiet around here these days, but it's still going.

The real cost of electricity - some numbers (well, lots of them)

by Jerome a Paris Sun Mar 5th, 2006 at 07:08:21 PM EST

I've been accumulating a number of sources on the "real" cost of electricity generation, and I'd like to put them together in one place - here.

There are several items that influence the cost of electricity:

  • the initial investment amount, and the discount rate used to amortise it over the life of the power plant;

  • the fuel costs (for those power plants that need a fuel, like coal, natural gas or uranium);

  • the operating costs;

  • the externalities, i.e. the cost imposed on society by the power plant, if not internalised by regulation.

Any comparison between various power sources that does not explicitly states which assumptions are made with regards to fuel costs and discount rate should be considered as dubious.

Click on all small graphs to enlarge them.
All sources listed at the end of the diary

Now, the main alternatives for power generation are well known: coal-fired plants, gas-fired plants, nuclear, hydro and wind. There are other renewables sources, and various subcategories in the above, but they represent most of the installed capacity. Hydro, being a mature industry with few prospects in the West, will not be discussed here.

Coal-fired plants are one of the oldest ways to generate power, and they still represent most of the generation capacity in a number of countries, including the USA (more than 50%) and China (more than 80%).

  • initial investments are moderate, typically 1000-1200 $/kW, but plants can be of many different sizes, so the absolute level of investment is quite variable;

  • fuel costs have historically been low, and are expected to remain so to an extent, as coal is considered to be plentiful in a number of countries;

  • externalities are the big outstanding issue for coal, as it is a highly polluting source of energy (both during mining and during generation), and the worst producer of greenhouse gases;

Gas-fired plant has been the big story of the last decade, with massive investment in a number of countries, especially the USA (see graph below), and it has reached a significant share of generation (20% in the USA) and often acts as the price setter for wholesale electricity as it usually is the marginal cost producer (i.e. that with the highest short term cost, and thus the one setting the price to provide a given number of kWh at any given moment)

  • initial investments are low, typically 500 $/kW, and size is quite flexible, which makes it possible to amortise that investment fairly easily, making the technology a favorite of the financial markets and thus of private power producers;

  • fuel costs are linked to the price of natural gas (duh!), which, until recently, was quite low. In the past 2 years, that price has suddenly shot up enormously, making the sector much less competitive (see more about that further down);

  • pollution is much less than with coal as natural gas is mostly methane and burns pretty cleanly, but that combustion generates carbon dioxide and contributes to carbon emissions. The issue of security of supply of gas has come to the forefront recently, with fears of shortages in both the US and the UK, due to faster than expected field depletion (and, in the US, damage to Gulf of Mexico production facilities)

Nuclear power plants were massively developped in the 70s following the first oil crisis, and provide about 20% of power in the rich world (and up to 85% in France). There has been little investment in the sector in the past 20 years as fears of accidents à la Chernobyl or Three Mile Island and concern about waste storage prevailed; the recent energy price hikes are putting nuclear energy back on the agend a of a number of politicians.

  • initial investment is heavy, typically 1700-2000 $/kW, and plant size is necessarily big (1,000 MW or more), thus making the absolute investments extremely high (at least $2 bn per plant);

  • fuel costs are still fairly low for now;

  • externalities are hard to quantify. Direct carbon emissions are negligible (but those linked to construction subject to much more controversy). Pollution is low, but the risk of accident, terrorism, or misuse of fissile materials (low probability, but potentially high impact) is harder to assess, and decommissoning and waste storage push costs onto the future.

Wind power
Wind is the renewable technology that has developped fastest in recent years, as it is not very far from being cost competitive and its development has been supported by governments in a number of countries.

  • initial investments are high, around 1000 $/kW. This is high because a kW of wind power produdes fewer kWh (typically, a third or a quarter as many) than a kW of coal, nuclear of gas)-fired power, due to the intermittent nature of wind. Thus the cost per kWh over the life of the plant of that investment is much higher. On the other hand, it is possible to invest small amounts as  individual wind turbines are relatively cheap ($2-3M today for modern models);

  • fuel costs are nil, as is the case for most renewable sources;

  • externalities are extremely small. There is no pollution, very little impact on the environment (impact on birds and on landscapes can be a issue in some locations), and there are no carbon-emissions. On the other hand, the fact that the supply is intermittent and unpredictable can create a burden on the network that needs to be properly accounted for.

:: ::

To get back to the main hypotheses listed above the fold, we can see that they have the following impact:

Interest rates / financing costs

The sources most sensitive to the discount rate used are, in decreasing order, wind, nuclear, coal and gas. Thus, making the hypothesis of a high financing cost structurally favors gas and coal against nuclear and wind. Conversely, providing cheap financing is most helpful to wind and nuclear.

It is thus not neutral at all to campaign for private ownership of generation assets, as it will  always skew investments towards gas-fired and coal-fired plants, unless you have - gasp - specific regulations or -double gasp - subsidies that encourage investments in other sectors like renewables (or nuclear).

To show you how significant the interest rate is, here are some calculations made by the French Ministry of industry (click to enlarge):

Changing from a 5% rate (typically the rate at which governments or public bodies can use to borrow long term) to 8% (a more typical rate for the private sector) increases costs:

  • for gas-fired plants by less than 5%
  • for coal-fired plants by a bit more than 10%
  • for nuclear plants by more than 30%
  • for windpower by just under 20%

Thus equal interest rates are not enough to compare competing power sources, the absolute level chosen matters as well.

Fuel costs

This graph tells us where the mental world of most investors in the sector is:

(Note: you need to multiply by approx. 1.5 to have prices in $/mbtu. Sorry for the poor quality of this graph.)

For more than 10 years, the fuel costs for both coal and gas were fundamentally flat, and if there was any trend, it was slightly downwards. Price risk on the fuel supply thus became a non-question for investors and the bankers lending them money. Coal and gas are cheap - and are thus expected to remain so (bankers are human too: depsite writing everywhere disclaimers saying that the past is no indication of future trends, they tend to believe it themselves...)

And as fuel costs constitute a large share of the production cost of coal- and gas-fired plants, the overall cost of the electricity form these sources was thus seen as low and stable, a nice combination risk-wise:

(This is from the French study mentioned above; don't look at absolute levels, but look at the relative importance of the sub-components: fuel is most important for gas, somewhat important for coal and slightly less for nuclear - and that's with 2003 price estimates, i.e. before the recent increases).

The evolution of the past 3 years has thrown out of the window all hypotheses on fuel prices (via Freecharts.com)::

The situation seemed to go back to normal after the Enron-induced crisis in 2000, but since 2002 the price has been going up relentlessly, with additional spikes which never come back down to where they started.

If you consider that the fraction above of the cost of gas-fired power that came from fuel cost was based on gas costing around 3$/mbtu, you immediately understand how gas-fired power has suddenly become a hell of a lot more expensive. You need to more than double the fuel component, which basically doubles the total cost... (and the current lowish prices after the Katrina peaks last autumn are due to the unseasonably warm winter so far).

Coal and uranium prices have also doubled, and, while the impact is not as massive, it is not trivial either.

The fact is that fuel prices have changed so much in the past 3 years that it has become really, really hard to make long term estimates of what these costs will be over the next 10-20 years, the period necessary to assess actual kWh production costs. Most available studies (all the links are provided throughout or at the bottom of the diary) still use pre-spike prices for natural gas and coal, and thus underestimate fuel costs for them.

Again, a very important hypothesis to keep in mind.

Externalities: carbon costs

One of the biggest uncertainties today on the price of electricity in a number of countries, especially those that use a lot of coal or gas, is how to take into account the impact of global warming. As pretty much everybody now agrees that it is caused by greenhouse gas emissions, and that the main culprit is carbon dioxide, power generation is right in the middle of it as coal- and gas-fired plants are amongst the biggest emitters of the stuff.

The numbers below come form one study, which may not be the final say on the topic, but it gives an idea of the orders of magnitude we are talking about: 1 MWH of electricity causes the emission of 1 ton of CO2 in a coal plant and half a ton in a gas-fired plant:

If we want to limit carbon emissions, as everybody seems to agree is a good idea, the solution will be to tax or to trade carbon emissions. Europe, as part of the Kyoto protocol, has set up a market for carbon emissions which has begun to trade last year.

With these prices, we are already adding 20 $/MWh to the cost of coal-fired electricity, and 10$/MWh to that of gas-fired power. Again, not an unsignificant change...

The above graph, which comes from an English study by the Royal Academy of Engineering in 2004 (The Costs of Generating Electricity), shows that the carbon cost essentially negates the cost difference between coal-fired or gas-fired and wind (at pre-2003 fuel costs!). Do note however that this would be for a price of carbon double today's price, which is probably more than the actual estimated impact of carbon emissions on the environment, as estimated below.

This fascinating table from a comprehensive study conducted for the European Commission External Costs - Research results on socio-environmental damages due to electricity and transport, pdf) shows the estimated cost of externalities caused by various power sources:

(1 cEUR/kWh = 10 EUR/MWh and 10$/MWh, approximately)

Global warming has a large impact, but the health impact of hydrocarbon-fired plants is, again, by no means negligible. who pays for it? The power consumer, in cash, or the citizen, in degraded health?

Externalities: network costs

Wind has one thing going against it: its intermittence does not make it a fiable power source, which is a problem in so far as electricity cannot be stored and production has to match demand all the time. When production changes in random fashion, like windpower tends to do, the network must be able to absorb these variations, and this has a cost.

Various studies have been made, and they have been contradictory. some, as that by the Royal Academy of Engineers have extremely high values for that cost, while others (see box below) have concluded that costs would be quite low, i.e. no more than 2-4 $/MWh altogether for wind power below 20% of total production.

reports on the cost of connecting windfarms to the network
  • GE Power systems energy consulting, the effects of integrating wind power on transmission system planning, reliability and operations, draft report phase 1, january 2004 ;
  • California wind energy collaborative, California RPS integration cost analysis phase 1, december 2003 ;
  • ESB national grid, impact of wind power generation in Ireland on the operation of conventional plant and the economic implications, february 2004 ;
  • LEPII-EPE, l'intégration de la production intermittente dans les marchés électriques libéralisés, mars 2003 ;
  • D. Milborrow, penalties for intermittent sources of energy, 2001 ;
  • D. Millborow, the real costs and problems of intergrating wind, 2001 ; Pacificorp, modeling wind energy integration costs, june 2003 ;
  • The Royal Academy of Engineering, the costs of generating electricity, march 2004
  • Ilex consulting, quantifying the system costs of additional renewables in 2020, octobre 2002

That question will most likely be studied further in coming years as the share of wind power goes up; it should be noted that the UK electricity market is one of the few that imposes balancing costs (i.e. producers are penalised if their actual production deviates from their announced numbers at any moment of the day), and some players make money out of this balancing mechanism (by providing easily switched on or off capacity) and have an interest to see prices be high...

:: ::

In any case, the fundamental questions are:

  • what's the financing cost hypothesis?
  • what's the fuel cost hypothesis?
  • what externalities are priced in?

which coincide, explicitly or nor with a number of eminently political questions:

  • is the sector State-owned, or strongly regulated? Are any sub-sectors encouraged by easy financing or similar schemes?
  • what long term price hypotheses are you making? Who do you rely upon to provide security of supply?
  • what regulations are applied with respect to pollution and carbon emissions?

And that's where we come to play. Our political campaigns to impose regulations or relax them will have real impacts on the cost of various sources, and on the investments made.

Speaking of which, here's a final, fun graph to feed your thoughts...

And with all this in mind, I'll post the various graphs that I have comparing prices in a future diary...

:: ::

Electricity Production costs - centralised plants (DGEMP - French Ministry of industry, pdf, full study in French)
Electricity Production costs - decentralised plants (French Ministry of industry, pdf, English summary)
Electricity Production costs - decentralised plants (DGEMP - French Ministry of industry, pdf, full study in French)

Ten Steps to a Sustainable Energy Future (by Rudolf Rechsteiner)

The Costs of Generating Electricity (The Royal Academy of Engineering)

Energy subsidies in the EU: a brief overview (European Environmental Agency)

External Costs - Research results on socio-environmental damages due to electricity and transport (European Commission)

Learning is fun :)

One externality you did not include was the cost of decommisioning different plants and restoring the land. I probably would not have thought about it either, except I talked with a nice man at a swedish environmental NGO. He told me how they had been pushing for all-steel structures (or at least a lot of metal) for wind-power, when swedish concrete giants had wanted to build wind-farm on concrete platforms. The reason for this was that as long as it was fairly pure metals, it would in all probability be worth it to harvest the scrap metal if and when it was decommisioned. Otherwise they might leave concrete plates all over.

Now there was a organisation that was thinking ahead.

Sweden's finest (and perhaps only) collaborative, leftist e-newspaper Synapze.se

by A swedish kind of death on Sun Mar 5th, 2006 at 08:43:03 PM EST
A number of countries include an obligation to decommission the wind farm as part of the permit. In at least Denmark and Spain, the developers are obliged to guarantee funding for hte decommissioning (either set aside a portion of the revenues, or provide a financail guarantee).

The cost of taking them down is not so high on a per turbine basis; you should also expect that in a few years, most useful sites will have been taken, and turbiens will be taken down - to put bigger ones in the same, already approved, spot...

In the long run, we're all dead. John Maynard Keynes

by Jerome a Paris (etg@eurotrib.com) on Mon Mar 6th, 2006 at 06:41:02 AM EST
[ Parent ]
Unless climate change changes the winds...

Anyway that is good news, but I was actually more thinking of possible lingering environmental effects from nuclear and coal plants sites. I do not know what they might be, but I guess the dismantling of Barsebäck here in Sweden might give a hint.

If there are such effects it would of course push the calculus in favor windpower.

Sweden's finest (and perhaps only) collaborative, leftist e-newspaper Synapze.se

by A swedish kind of death on Mon Mar 6th, 2006 at 09:49:48 PM EST
[ Parent ]
I have a silly question. Could have looked it up elsewhere, but why be ashamed, right?

Why can you not store electricity?
Do the plants always produce 1-to-1 with the demand?

A patriot must always be ready to defend his country against his government -- Edward Abbey

by serik berik (serik[dot]berik on Gmail) on Mon Mar 6th, 2006 at 04:33:33 AM EST
You can, but only in a very limited fashion: things like pumping water into reservoirs to feed hydroelectric turbines. Expensive, dependant on geography, environmentally awful and not always very worthwhile. What form would you store it in?
by Colman (colman at eurotrib.com) on Mon Mar 6th, 2006 at 04:39:32 AM EST
[ Parent ]
There's no such thing as an efficient storage technology. The best batteries are hugely inefficient and also far more expensive than generation. So the only possible storage systems are mechanical, like the water scheme that Colman mentioned.

If someone invented a room-temperature superconductor, it would eliminate distribution losses and could potentially also be used for loss-free storage. But current superconductors need super cooling, so they use more energy than they could possibly save.

by ThatBritGuy (thatbritguy (at) googlemail.com) on Mon Mar 6th, 2006 at 04:58:37 AM EST
[ Parent ]
One of the things that never seems to figure in nuke calculations is that there is no reliable long term storage of high level waste. Currently it's pulled out of reactors and kept in ponds, which is hardly a sustainable long-term solution.

This storage has to stay safe for 100,000 years or so, so it won't be cheap. There are very few places on the planet that are dry enough and stable enough to make reliable storage possible. And the costs of creating inert containers - the current favourite is a copper/steel mix - aren't trivial.

So when decommissioning costs are quoted, it's worth remembering that they don't yet include this long term storage.

by ThatBritGuy (thatbritguy (at) googlemail.com) on Mon Mar 6th, 2006 at 05:04:48 AM EST
There used to be a nuclear/chemical storage facility on an Aral Sea island in the times of the Soviet Union, when the Sea was a sea-like lake.

Now, considering that is looks more like a dirty, salty puddle, the base had to be relocated, due to safety reasons.

Funny, but the island was a state-protected national park back then. Almost no wildlife there now, is there?

A patriot must always be ready to defend his country against his government -- Edward Abbey

by serik berik (serik[dot]berik on Gmail) on Mon Mar 6th, 2006 at 05:33:47 AM EST
[ Parent ]
Last month I did read this report about that.

Don't read it, you gonna be sick for at least a day.

The struggle of man against tyranny is the struggle of memory against forgetting.(Kundera)

by Elco B (elcob at scarlet dot be) on Mon Mar 6th, 2006 at 06:15:26 AM EST
[ Parent ]
Nuclear/chemical? Aren't you rather talking about the big Soviet bioweapons test field? That was certainly on an island in the Aral Sea. Or was there all three?

*Lunatic*, n.
One whose delusions are out of fashion.
by DoDo on Mon Mar 6th, 2006 at 06:22:23 AM EST
[ Parent ]
Sorry, I should have put it in another way, since I was not sure. Instead of "nuclear/chemical" read "nuclear or chemical." :)

A couple of years ago I watched a Russian documentary about the island, in which they showed the bunkers with the stuff that were some sort of storage facility.

A patriot must always be ready to defend his country against his government -- Edward Abbey

by serik berik (serik[dot]berik on Gmail) on Mon Mar 6th, 2006 at 07:27:19 PM EST
[ Parent ]
Don't forget on the other hand that highly radioactive waste is also the one with the shortage half-life (in tens of years at most). The less "hot" stuff is not really dangerous, even if it remains that way for a very long time.

So I am not convinced that there is no solution to the waste problem. I personally think that it plays on our fears of an invisible killer more than anything else. There are lots of things that are really more dangerous to us and which we should worry about before (like most pollution from coal...)

In the long run, we're all dead. John Maynard Keynes

by Jerome a Paris (etg@eurotrib.com) on Mon Mar 6th, 2006 at 06:44:07 AM EST
[ Parent ]
One factor which hasn't been considered is the opportunity cost of using natural resources now instead of later. For example, the oil companies have just had a  huge increase in the worth of their reserves because the price of oil has gone up. They did nothing to get this gain other than sitting on their reserves. If, let's say, the price of oil is expected to increase 10% a year they are better off not selling now and waiting for the price to go up. The fact that this is partly politically impossible doesn't mean that it isn't being factored into the oil company's long range plans.

Another way of looking at opportunity cost is that natural hydrocarbons are hard to make and are much more useful as chemical feedstock than as fuel. So by burning them now we can expect higher costs for synthetic materials later.

As non-renewable resources it is also hard to calculate the cost to society after they run out. Currently the way this is handled is exactly backwards. Extraction industries get a tax break for using up natural resources (depletion allowances) instead of being charged for permanently removing them from the commons. Standard economic models always assume substitutes will exist for any commodity. Which implies that the fuel costs of Uranium will rise so that the effective cost per KWh will tend to be the same as with other fuels. The differences in the cost of power generation by technology used are mostly because the marketplace hasn't had a chance to catch up with the new raw material situation.

While no profit making company is going to factor in the opportunity costs or the long range ecological costs when making business decisions, governments can do this via tax and other public policies. Unfortunately there is hardly any real long range planning being done outside of academia. I've coined a new acronym. We have NIMBY (not in my back yard) to express the unwillingness of people to allow new developments. My new one is NIMLT (not in my life time). People are unwilling to make any changes in their lifestyles and push the sacrifices needed into the future.

Policies not Politics
---- Daily Landscape

by rdf (robert.feinman@gmail.com) on Mon Mar 6th, 2006 at 08:42:36 AM EST
One of Jerome's recent graphs showed that almost half of the electricity generated was lost in transmission.  We could double the supply of electricity if we could transmit it efficiently, or like Edison, produce it locally.  One of the Edison/Westinghouse DC/AC arguments was that if there was a powerplant in each neighborhood the high losses of transmitting Direct Current electricity would be reduced.

If we had a power plant on every block, and the fuel could be transported more efficiently than electricity, we could double our supply for the same fuel/carbon cost.

Now, who wants this down the block? (Don't all raise your hands at once)

by dmun on Mon Mar 6th, 2006 at 10:17:38 AM EST
But we can install solar panels, solar heating and wind turbines in every block, can't we?

A society committed to the notion that government is always bad will have bad government. And it doesn't have to be that way. — Paul Krugman
by Migeru (migeru at eurotrib dot com) on Mon Mar 6th, 2006 at 10:23:28 AM EST
[ Parent ]
Prolly not turbines...
by asdf on Mon Mar 6th, 2006 at 08:08:47 PM EST
[ Parent ]
You could put wind turbines at the centre of roundabouts...

Whether or not it makes sense to install small turbines on rooftops seems to be controversial, but I see no reason why you couldn't install a 10m tower on a 3-storey building spanning a city blosk. It wouldn't look disporportionate either.

A society committed to the notion that government is always bad will have bad government. And it doesn't have to be that way. — Paul Krugman

by Migeru (migeru at eurotrib dot com) on Tue Mar 7th, 2006 at 03:35:13 AM EST
[ Parent ]
Well I don't know how it is over there on your side of the pond, but here they put them in places where the wind really blows. I've never lived in a city where the wind blew nearly as much as it does on the Great American Desert.
by asdf on Tue Mar 7th, 2006 at 11:09:35 PM EST
[ Parent ]
Notice that the coal was transported to the Pearl St. station by horse and wagon!
by dmun on Mon Mar 6th, 2006 at 10:47:24 AM EST
[ Parent ]

Go to: [ European Tribune Homepage : Top of page : Top of comments ]