How much does wind power really cost?

2nd August 2010

Last week, Chris Huhne, the Secretary of State for Energy & Climate Change, gave the first annual energy statement to Parliament, putting reducing carbon emissions by 80% by 2050 at the heart of the UK’s energy strategy.

To achieve this is going to require an enormous expansion of renewable energy – and Huhne has been immediately leaped on by the usual suspects for daring to claim that wind power can be ‘incredibly competitive’.

How much does wind actually cost? Christopher Booker, in the Daily Mail article linked above, claims that wind is twice as expensive to build as gas, coal or nuclear, and that the electricity from offshore wind turbines is three times as expensive as that from conventional power plants. It’s not clear where these figures are from – but one can’t really blame an op-ed in a tabloid newspaper for not including references (especially a tabloid who specifically doesn’t require op-eds to be able to back up their factual claims).

Part of the problem in this debate is that there’s a wide variety of range in cost estimates for costs for energy generation. Let’s take three examples: the RenewableUK briefing sheet on costs, the PB Power Study for the Royal Academy of Engineering, and the estimates given by DECC in their ‘Pathways’ document which was released to coincide with Huhne’s energy statement.

RenewableUK claims that onshore wind (£1,600/kw) is cheaper to build than coal(£1,650/kw) and nuclear(£3,000/kw), although more expensive than gas (£650/kw). Offshore wind costs are £3,600/kw at the top end. DECC’s figures are similar, although they include the cost of Carbon Capture and Storage when considering coal, which comes out as £2,500/kw. The PB Power report is very different, which may be an artefact of its age. It says the cost of building new coal plant is £820/kw, gas is £300/kw for a combined-cycle plant, nuclear is £1,150/kw, onshore wind is £740/kw, and offshore is £920/kw.

It’s therefore not clear where Booker got his numbers from – none of our sources show that the cost of building onshore or offshore wind is twice the price of gas, coal and nuclear, although it is clear that gas is by far the cheapest to build. Offshore wind is comparable to nuclear in terms of construction costs.

Let’s look at the second challenge: how much electricity actually costs from each of these types of generation. The DECC report doesn’t include this cost, so we’ll concentrate on the PB Power Report and the RenewableUK Briefing Sheet.

RenewableUK claims that electricity from onshore wind costs on average £90/MWh, while gas is £49/MWh, coal is £69/MWh, and nuclear is between £57MW/h and £86MW/h. Offshore wind is significantly more expensive at £160MW/h. Of course, this is largely dependent on the cost of fuel – RenewableUK uses a price of £14/MWh for gas. The 2009 prices used in the DECC report gives a cost of about £20/MWh for gas, rising to £26MW/h by 2030 on a likely case scenario.

The PB Power Report claims that power from gas costs £22/MWh, coal £25, and nuclear £23. Onshore wind and offshore wind are treated differently; their costs are given so as to include additional back-up to account for the variability of wind. Onshore costs £37/MWh without backup, and £54/MWh including it. Offshore is £55/MWh and £72/MWh correspondingly.

The cost for nuclear-generated electricity seems out of place compared to every other report – it seems that PB Power dramatically underestimated the cost of nuclear by specifically excluding financing costs, which are considerable for long-term projects such as nuclear. The biggest single factor in the cost of nuclear power is the price of money. Nuclear plants are a long-term risk, and financing costs reflect that.

The PB Power report only applies the cost of standby generation to wind, despite the fact that the current standby generation capacity in the UK stands at about 14GW out of 80GW – compared to 4 GW of wind. This is because the grid requires the ability to quickly compensate for major reactors ‘tripping’ (i.e. unexpectedly going offline) simultaneously – for example, this report shows that Sizewell B, representing 0.85GW of the UK’s generation capacity, ‘tripped’ four times over the 18 months prior to the end of the first quarter of 2009. On average about 1GW of thermal plant in the UK ‘trips’ every week, and so requires back-up – but for some reason this wasn’t included in the PB Power report. Indeed, in the past quarter Sizewell B has been down entirely owing to water leakage: http://www.hse.gov.uk/nuclear/llc/2010/sizewellb2.htm. It’s therefore not correct to claim that only wind requires back-up – any analysis needs to apportion back-up costs relative to the amount of time they’ll be required by each type of generation.

Despite this, it does seem that wind generation is more expensive than fossil fuel generation, although it’s competitive with nuclear. However, the costs given above don’t tell the whole story – the cost of fossil fuel generation isn’t just in its market price, but in its long-term impact on the environment. This is why part of the Secretary of State’s statement involved the price of carbon – the future cost of pollution emitted by fossil fuel plants, hypothecated into a cost for producers of that pollution now. The PB Power report adds a potential price of carbon to each of the fossil fuels, and it’s immediately clear that even including the full price of back-up generation, onshore wind is very competitive against fossil fuel generation. It’s also clear that offshore wind is not three times more expensive than other forms of generation – unless you don’t think carbon emissions have any impact on the environment. Of course, Christopher Booker doesn’t think that – but somehow he neglects to mention it when talking about wind.

Comments

The rub, of course, is how much wind actually reduces fossil (or nuclear) fuel use.

According to DECC: http://www.defra.gov.uk/environment/business/reporting/pdf/20090928-guid..., the grid factor for carbon is 0.543 kg/CO2 per KWh, which given the low penetration of renewable energy is a useful rule of thumb for CO2/KWh. By NETA's figures, right now wind is producing about 400MW of electricity. Over the course of this hour, wind will displace about 217200kg of CO2 from the system, or about 235 barrels of oil.

That is a theoretical displacement. Where is the actual record of how much less fossil fuel is burned per kWh consumed?

NETA is the record. If you're getting 400MWh from wind, that's 400MWh you don't need to get from other sources. The grid factor is the carbon intensity of electricity production in the UK - i.e. the average amount of carbon it takes to generate one KWh. This is averaged out across all forms of electricity production. If you wanted the precise figure, you'd need to calculate the carbon intensity of electricity production right now that doesn't involve wind or other renewables - currently 46% gas, 32% coal, and 12% nuclear, according to this half-hour's NETA figures. Wind would be displacing that average figure per KWh, rather than any other particular type of generation.

But that does not account for the behavior of other plants in ramping their production, e.g., spinning reserve (i.e., fuel continuing to burn to keep the plant "warm"), ramping/restarting costs (i.e., more fuel burned than otherwise), lower capacity (at which the plant burns its fuel less efficiently).

Nor does it even consider the preference of which plants are ramped in response to wind. Notably, it doesn't assume that hydro is most likely to respond, followed by gas.

Actually, the carbon factor for a given fuel type does include those factors - it's the average amount of carbon it takes for a given amount of electricity to be produced, so it's the sum of all power generated by a given fuel type against the amount of carbon released by that fuel type. That obviously includes the added carbon costs of ramping up production. As I said in the other thread, you really can't take anything from REF at face value.

What's your source for a preference for hydro over gas?

I don't think it does, since it simply describes the average over the entire system over, say, a year.

I don't have a ready source, but it is well known that where there is hydro, that is the source most likely to be ramped to balance wind production. Likewise, gas is preferred over other thermal plants because it too can respond quickly enough.

So really, your NETA average should include only the proportion provided by gas and hydro to more accurately estimate wind's effect.

Yes - and that average includes the inefficiencies caused by ramping up and down - which the grid does daily, in any case. Given that daily ramping actions involve spreads of 15-20GW and wind capacity in the UK is 4GW, any additional inefficiencies caused by wind would be overwhelmed anyway.

I'm sorry, but unless you can provide a source for your claims about hydro, I'm forced to conclude that you're wrong - National Grid's 'Gone Green' consultation shows that of the balancing actions they took in Nov 2008, 45% involved coal, 33% gas, 16% pumped storage - less than 1% involved hydro. They took that month as indicative. http://www.nationalgrid.com/NR/rdonlyres/32879A26-D6F2-4D82-9441-40FB2B0...

This is actually quite different to the fuel used by baseload generation, which of course involves significantly more gas and nuclear. It's therefore likely to be the case that using the grid factor for carbon displaced by wind generation, if anything, underestimates the carbon displaced.

The National Grid paper you cite describes 4-hour-ahead balancing, which indeed is feasible while wind is remains a very small percentage of the total. As wind's "penetration" grows, that model will no longer be adequate, because balancing will require more immediate responses, which can be provided only be gas and hydro.

By the way, pumped storage is hydro.

You're going to need to give me a source for the volatility you're presuming there - National Grid make the reasonable assumption that 4-hour forecasting will be sufficiently advanced to permit only a modest increase in e.g. STOR capacity - they give a figure of only 4GW of additional reserve with 32GW of wind on the system.

Pumped storage is not hydro, except in the sense that it involves water. It requires input to work, meaning that while it's a useful way of storing energy (and therefore is a complement to windpower), it's not displaced generation, as it requires generation in the first instance.

No-one is talking about building new hydro or pumped storage to meet the 4GW requirement for the expansion of wind - but they are talking about gas. The only place in which hydro is used to balance wind is in Denmark, which is heavily interconnected with Sweden & Norway, allowing the Danes to buy the excess. We only have 1.5GW of hydro on the system, so we wouldn't be able to use it for balancing in any case.