Ambrose Evans-Pritchard (AEP) has written a series of well-informed and persuasive articles on energy in the UK’s Telegraph newspaper over the summer holidays. His topics included wind power and batteries. He also wrote with enthusiasm about carbon capture and storage, a technology that many people think will be needed at enormous scale if the world is to reduce emissions quickly.

I’d like to believe him. If we could find a way of adding inexpensive CO2 capture units onto existing power stations we might be able to continue to burn coal and gas into the long-term future. The world would have plentiful wind and solar, ready to be supplemented by fossil fuel power when necessary.

Unfortunately, I don’t think AEP is right. CCS will probably always add more cost to electricity than can be financially justified. I work out some numbers below for a power station in Canada with CCS to try to support my assertion. I'm sorry it takes a large number of paragraphs to do this.

Rather than seeing CCS as a way of complementing intermittent renewables, we are better advised to invest in energy storage to provide the buffers we need. When the sun is shining or the wind blowing, we will siphon off power and put it into batteries or transmute it into storable gases and liquid fuels. This is cheaper, and will become cheaper still every passing year.

The AEP vision

· Add CCS to all fossil power stations

· Collect and sequester all the CO2

· Run these power stations all the time, minimising the huge capital cost of CCS per unit of output.

What I say in The Switch

· Overbuild wind and, particularly, solar PV

· Take the surplus electricity and use to provide the energy to make renewable fuels (see the previous post on this web site on Daniel Nocera, for example)

· Store these fuels for times when the sun isn’t shining nor the wind blowing

The CCS process

At a power plant with CCS - of which there is really only one in the world, at Boundary Dam in Saskatchewan, Canada - a fossil fuel is burnt and the flue gas is passed through a solution containing chemicals that bond the CO2 into bicarbonate. The solution is then heated, the bicarbonate breaks up into CO2 and other molecules and concentrated CO2 is collected. This is a relatively simple, well understood process that has been in use for eighty years. Most – perhaps 90% - of the CO2 is collected, and almost all is then regained and can be stored.

In the UK, we envisage storing the CO2 in old oil and gas reservoirs. Storage of the CO2 in this way will add some cost. In other places, the CO2 actually has value because it can be injected into oilfields that are still producing. It enhances the production of fuels. However, it should be said that some of that carbon dioxide returns to the surface dissolved in the extra oil. Only about 75% of the CO2 sent down into a depleting oilfield stays below ground for ever.

CCS costs

Boundary Dam is an old power station that burns lignite on the border between the US and Canada. It is composed of several separate units. One of these boilers – number 3, usually called BD3 – was coming to the end of its life. Its owners, SaskPower, a public utility, decided to replace this unit with a new generating plant capable of producing about 139 MW of electricity. This is enough to meet about 2% of Saskatchewan’s power needs.

The CCS process uses large amounts of energy. About 29 MW of power is devoted to extracting the CO2 and then regaining it. Very roughly, a power station gets about 20% less usable power from its plant with CCS. There are two separate costs arising from the parasitic effect of carbon capture. First, CCS means less electricity output for each dollar of capital expenditure building the power station. Second, the plant has to spend money on fuel to provide the heat and power to run the CCS process.

The third, and much the largest, cost is the carbon capture plant itself. At Boundary Dam, this equipment cost around CAN $900m, or about US $700m.

Lastly, the plant needs people and materials to run the CCS process. The figures for this are the least visible to the outside world, although SaskPower has provided some estimates. They include the cost of manning the CCS plant and purifying and replacing the solution that absorbs the CO2.

How much do these four elements add to the cost of producing electricity?

First of all, I need to specify some assumptions. I guess that Boundary Dam and other CCS plants will last about 30 years. This is a figure you often see as the length of life of today’s coal fired power stations although many of today’s plants in the industrial world will last longer. I assume that the power station works 8,000 hours a year. I use a figure of 5% for the cost of capital, and assume zero inflation.

The price of lignite, the fuel that Boundary Dam uses, is about US $20 a tonne on the US/Canada border. It has an energy value of about 4,500 kWh per tonne. Boundary Dam delivers about 40% efficiency, meaning that one tonne of lignite provides about 1,800 kWh of electricity.

Very roughly, one megawatt hour (1,000 kWh) produced at Boundary Dam results in one tonne of CO2 being emitted. About 90% of all CO2 produced at BD3 is currently being captured.

We’re now in a position to estimate how much CCS costs per unit of electricity produced. And how much per tonne of CO2 captured.

Cost 1. The extra capital needed to build the electricity generating plant because 20% of its output is needed to power the CCS.

The power station part of the 139 MW Boundary Dam unit cost CAN $562m, or about US $450m. 20% of this is US $90m. At a 5% cost of capital over 30 years, the implied yearly cost is about US $6m. The power station produces about 880,000 MWh a year, and the cost is therefore about US $7 per MWh. This figure appears to be omitted from other estimates of the cost of CCS.

Cost 2. The extra lignite burnt to create the power and heat that is used by the CCS apparatus.

The 29 MW of the electricity initially produced at Boundary Dam is devoted to the CCS process. To make this much electricity at a conversion efficiency of 40% requires 72.5 MW of coal energy. This means that each hour about 16 tonnes of coal are needed to meet the electricity (and heat) requirements for CCS. Over the course of the year, the cost is just over US $2.5m dollars and just under US $3 per MWh. For simplicity, I round this number to $3.

Cost 3. The capital equipment needed for carbon capture.

The kit needed to carry out carbon capture cost over CAN $900m, or about US $700m. Over 30 years, and at implied cost of capital of 5%, this adds about US $52 per MWh. (If the cost of capital was 0%, this figure would still be over US $26).

However this figure is the one that may come down sharply when more CCS plants are constructed. SaskPower says the next unit might be 30% cheaper and 50% reductions are possible in time.
Let’s be generous to CCS and use a figure of US $25 per megawatt when the technology is mature.

By the way, the next retrofitted CCS plant, at Petra Nova in Texas, will cost about the same per MW as Boundary Dam and will probably come on stream in about six months. And don’t even mention the extraordinary new build at Kemper in Mississippi. This power station looks as though it will come in at over US 7bn for a coal gasification plant, combined with CCS, totalling less than 600 MW. That makes it more expensive than Hinkley Point per megawatt of output. As importantly, only 65% of the CO2 will be captured. So the optimistic figure of an extra cost $25 per megawatt hour of electricity produced is a really generous assumption.

Cost 4. The annual cost of operating the CCS plant.

A SaskPower presentation seems to suggest a figure of about CAN $9 a MWh, or US $7. . It may go down a bit in future plants but I have not included any improvement because it is likely to be quite small.

(I have had to make the critical assumption that the y axis marks are each CAN $10 on the relevant chart towards the middle of the presentation. This fits with the rest of the SaskPower presentation).

The total

Add these figures together and we get to US $72 per megawatt hour for the implied extra cost of power at BD3. This may go down to US $35 when the CCS technology is completely mature. This will take several decades.

US $72 is substantially more than the current wholesale price of Canadian electricity, which lies in the high US $30s. The implied cost of electricity at Boundary Dam has therefore been nearly tripled by the addition of CCS. Even after future cost reductions, CCS will add almost 100% to the cost of power.

The position is actually even worse for CCS. Boundary Dam has been so expensive that it has added substantially to the power bills of provincial residents. One think tank said

With the cost of electricity at 12-14 cents per kilowatt-hour and rising, the province’s economic competitive position will be weaker. Saskatchewan no longer has affordable electricity and it is likely to get more expensive in future, especially if Boundary Dam 4 CCS is built.

This means that the relative attractiveness of wind and solar are inevitably going to grow. Saskatchewan has been blocking wind power for decades, even though conditions on the northern Great Plains are highly favourable for turbines. A cynical observer might suggest that the presence of lignite and a commitment to using it in power generation has warped the decision-taking of the Province. Other accusation, such as undue influence of the company transporting the CO2 for oil recovery, fly about. But at some point the far lower cost of wind than coal electricity is bound to mean a larger number of wind farms across the Province.

Of course Canada is not the best place for sun. But the average PV panel on a house near Boundary Dam will produce at least 20% more than the best UK locations. At some point, PV electricity will replace the need for coal. When that happens, the implied cost of CCS per megawatt hour will rise as the plant is used less and less and costs need to be spread over a smaller amount of electricity.

Neither Canada, nor any other place in the world, should be investing now in generating capacity that needs to work every hour of the year in order to use its capital productively. What we need are sources of energy that can be available for the relatively small number of hours each year that neither the wind nor the sun are present.

The cost of the CO2 savings

After very severe teething problems, including over 6,000 maintenance calls, Boundary Dam is now producing almost as much sequestered CO2 as planned. 2017 will probably see about 1,000,000 tonnes pipelined to the oil field for increasing output. Of this, about 700,000 tonnes will stay in the ground for ever.

This has cost almost US $70m, or $100 a tonne, assuming constant operation apart from maintenance intervals. After further development, we might be able to get this to about $60, if future plants are fully used for 8,000 hours or so a year.

The alternatives

The last chapter of The Switch looks briefly at some of the alternatives to CCS that provide a renewables-based energy system with its need for month-long buffers and stores. (Short term storage will be offered by batteries). In summary, I write in the book that conversion of surplus electricity at times of high wind or solar output into gases and liquid fuels looks far cheaper than conventional CCS. Direct capture of CO2 from air will probably become cost competitive to the hugely capital intensive process of putting CCS plants beside coal-fired power stations.

Wind on the Great Plains is now producing power at less than 4 cents a kilowatt hour, or sub $40 a megawatt hour, and solar will be at similar level within five years at the Canadian border. Even if 50% of the energy value is lost in a conversion process to natural gas or gasoline, cheap renewable electricity for storage use will cost far less than today’s US $72 per megawatt hour at Boundary Dam. And we won’t have the 10% of fugitive CO2 emissions being added to the atmosphere all the time.

(NB The arguments about CCS on steel, cement and plastics plants are more complex and I have failed to address them here).