Mixing renewables and back-up gas power is going to be wasteful

The purpose of this draft paper is to assess what will happen if, as expected, many gigawatts of  intermittent renewables are added to the UK grid alongside large amounts of standby gas power. I use actual data from spring 2013 to model what will happen in 2030 if the expected portfolio of low carbon sources of electricity is constructed. In particular, I try to estimate how much the back-up gas plants will be used and how much surplus and unusable electricity will be generated when the wind is blowing strongly or demand is relatively low.

I suggest that the two principal issues facing the UK grid will be using the huge seasonal surpluses of electricity arising in late spring, summer and early autumn and, second, how to finance the construction of tens of gigawatts of standby power which may be used less than 10% of the time.

I conclude by tentatively putting forward a view that the right way to deal with these issues may possibly be to invest even more heavily in renewables (and less in gas standby) and use the increased surpluses to produce methane for use in the gas grid. This may be a cheaper and more energy-secure solution than current proposals. It will also make decarbonisation of heat easier than currently expected.

(As always, I’m very grateful for comments on this article, however critical.)

Electricity demand and supply in 2030

Electricity supply.

a)      Generation portfolio. With adjustments, I use the latest figures from the Committee on Climate Change (CCC). [1] The CCC says that ‘decarbonisation of the power sector is key to reducing emissions across the economy and would also enhance energy security’. It proposes four broadly similar scenarios that enable the UK to cut emissions from electricity generation to below 50 grams of CO2 per kilowatt hour. All these scenarios involve large amounts of nuclear power, wind generation, biomass burning, electricity production from fossil fuels using carbon capture and storage and unabated generation from conventional combined cycle gas plants (CCGT).  I have merged these four possible future plans into a single estimate but simplified it to only include nuclear, wind, CCS, biomass, unabated gas and solar PV.[2]

The figures in table 1 are based on the following key assumptions

a)      Nuclear. The government reaches agreement with the power and construction companies over the price that will be paid for electricity produced by nuclear power stations. As a result, two or three consortia built about 8 power stations by 2030.[3]

b)      CCS. The government must also reach agreement over the price paid for power from CCS-equipped fossil fuel plants. As important, experiments and pre-commercial deployments need to happen quite quickly, if large plants are to be ready by 2030.

c)       Wind. Offshore and onshore farms are projected to both contribute about 25 GW of capacity. The key issue is whether offshore wind costs decline at the rate expected by the CCC.

d)      Biomass. Electricity generation from burning biomass is included in the government’s plans for 2020 and financial incentives are provided. I have assumed no expansion of biomass beyond the CCC’s figure for 2020.

e)      CCGT. For the UK to have security of electricity supply, it needs to have enough ‘dispatchable’ generation capacity to meet peak winter needs. At the moment of likely maximum power requirements in early evenings in December and January, no PV is available and wind output may be negligible. So CCGT and nuclear, biomass and CCGT must cover the total requirement. In the last few years, the peak has been about 60 GW and I have suggested a figure of 40 GW of gas-fired power generation, leaving 20 GW to be provided by the other reliable power sources and a 5 GW safety margin. The CCC’s scenarios range between 40 GW and 46 GW.

Table 1

Central scenario for 2030 – generation capacity

Nuclear 12 gigawatts (GW)
CCS 11 GW
Wind 50 GW
Biomass 4 GW
CCGT 40 GW
Solar PV 10 GW
Total 127 GW

Electricity demand

In recent years electricity use has been gently falling. This change has been caused by improved energy efficiency and a weak economy. In addition, higher electricity prices have choked off demand.

The CCC assumes that electricity generation will rise from about 350 terawatt hours (TWh) in 2011 to between 403 and 465 TWh in 2030, an increase of between about 15 and 30 percent. The Committee isn’t clear in its most recent report why this increase will happen but previous publications from the CCC have pointed to the likelihood of increased electricity use from the use of heat pumps for domestic heating and from growth in the number of electric cars.

Other sources are less bullish about power demand. (Electricity demand is lower than electricity generation because of losses in the distribution system and use of electricity by the generators themselves). The most recent work by the National Grid suggests lower figures than those projected by the CCC.[4] National Grid’s three scenarios offer estimates of 2030 electricity demand that vary from a figure very similar to today’s level in its ‘Slow Progression’ scenario to estimates about 10% higher in its Gone Green forecast and 20% higher in the ‘Accelerated Growth’ view.

Projections for 2030

In the following section, I have estimated how much electricity is generated by the portfolio of generating plant in Table 1 during a portion of 2030. My approach was to assume that the daily pattern of spring demand in 2030 – rising from about 5 am to a plateau from 9am to 7pm and then falling sharply – is identical to 2013.

Figure 1 shows how electricity production varied for each of 4,300 half hour periods from February 23rd to May 23rd.

Figure 1

Electricity production, including imports, for half hour periods, expressed in megawatts

 Total electricity demand

The chart shows the expected daily variation with production rising to a peak at the end of the working day/beginning of the evening and then falling to a much lower nighttime level. It also exhibits the weekly cycle of lower weekend and Bank Holiday power needs. Average demand levels were highest during March 2013 because of the unusually cold weather, peaking consistently at over 50,000 megawatts (50 GW).  By May, peak demands averaged around 40 GW, with nightime levels falling to around 25 GW, down from 30-35 GW in March.

The next step was to estimate the source of electricity by type of generator in 2030.  I have done this in the following way.

a)      I used the 3 months electricity production data from late February to late May 2013. The manager of UK electricity trading arrangements, Elexon, publishes data for each half hour period. This information includes data on how much power is produced by nuclear, wind and all other types of generator.[5]

b)      This 3 months of data provides estimates of the total amount of electricity generated by wind power in each half hour. For this 3 month period, the capacity of wind connected to the UK National Grid was 7.15 GW. The estimate in Table 1 is that this will rise to 50 GW by 2030. Therefore if the weather patterns in spring 2030 were to exactly the same as in spring 2013 we can plausibly assume that seven times as power will be generated. (7.15 GW multiplied by 7 equals about 50 GW, the assumption in Table 1 for the amount of wind capacity in 2030).

c)       Nuclear power stations will operate continuously although some portion of capacity may be unavailable because of maintenance. I assume that 10 GW work continuously during the 2030 spring period.

d)      CCS plants will also be working continuously. This is because they will be paid a standard and unchanging contract price for electricity. It will therefore make financial sense to operate them all the time.

e)      The same is true for the 4 GW of biomass capacity.

f)       I have assumed a standard daily profile for PV production, starting at daybreak and rising to midday and falling as the afternoon proceeds. This profile varies only by the month of operation with May output being much higher than production in late February. Average daily output figures for each month per kW of installed capacity are taken from records of a small rooftop PV installation.

g)      Imports of electricity: in spring 2013, the UK imported significant quantities of power from France and the Netherlands. The 2030 forecasts assume no net imports. (This assumption is relaxed later in this paper).

To summarise: I have taken actual 2013 electricity generation for each half hour in a three month spring period and then used the predicted portfolio of generating capacity in 2030 to show the makeup of electricity production in that year. Nuclear, CCS and biomass plants (totalling an available figure of 25 GW, slightly less than maximum capacity because some plants will be undergoing maintenance) work continuously in 2030. Production from wind turbines is seven times greater than the 2013 actual figures for wind generation. Solar PV generates electricity during the daytimes according to a set pattern. CCGT plants operate when the amount of wind and PV generation would be insufficient to bridge the gap between the dispatchable power sources (nuclear, CCS and biomass) and the total 2013 demand levels.

The results

Overall

25 GW of generating plant that is working continuously will generate almost enough power to cover the minimum needs of a night in late spring. By contrast, peak demands of 55 GW in the early evening of the last days of February will require either substantial wind-generated electricity or the use of some of the 40 GW of CCGT plants.

When the wind is blowing strongly, the UK is likely to have a substantial surplus of power. By contrast, quiet days will mean continuous use of gas-fired generation.

In the three months under study, UK generators attached to the main distribution network and adding in imports produced about 81.2 terawatt hours (TWh) of electricity (or 81,200 gigawatt hours). In 2030, the portfolio of predicted generating plant working under the same conditions would produce 94.5 terawatt hours. This would be made up as follows

Table 2

Makeup of spring 2030 electricity production if demand conditions and wind speeds the same as 2013

Nuclear 23.6 TWh
Wind 32.0 TWh
CCS 21.5 TWh
PV 2.4 TWh
Biomass 8.6 TWh
Total before CCGT 88.1 TWh
CCGT 6.4 TWh
Total including CCGT 94.5 TWh
Surplus over electricity generation need 13.4  TWh (about 16%)

 

This is the key result: with the portfolio identified by the CCC, and the same demand pattern in 2030 as in spring 2013, CCGT plants need to generate 6.4 TWh and, second, the variability of wind means that the UK nevertheless produces 13.4 TWh too much electricity. This surplus must be exported, stored or dumped because electricity demand must match electricity supply every minute of every year. Figure 2 shows when the UK is in surplus, and by how much, over the 4,300 half hour periods under study.

Figure 2

UK net electricity surplus (-ve) or deficit (+ve) before taking in account the use of CCGT plants, expressed in megawatts

 Surplus

 

Several features of this chart stand out.

a)      As winter ends, surpluses of electricity become more common. From mid-April onwards, periods of deficit, and hence need to back up intermittent wind using CCGT, become much more rare.

b)      Surpluses and deficits can be very large in any one period. Deficits peaked during March at levels over 20 GW. But deficits of this size tend to be very short-lived. The peak deficit of around 25 GW at point 903 was followed 5 hours later by a need for only 8 GW of back-up gas generation.

c)       Surpluses in the coldest months (such as around points 700 and 1400) are driven by major storms. These events last for several days and can produce sustained and very large surpluses. The importance of these sustained surpluses will be discussed later.

d)      The April/May surpluses are only interrupted by very short periods of need for CCGT back-up. The implications of this will also be mentioned in the section on storage.

Demand for gas generation

One criticism of the approach in this paper is to say that the CCC assumed higher overall electricity demand in 2030. (In future work I will examine the implications of greater aggregate demand). The calculations here suggest that 40 GW of CCGT back-up will only  provide about 6.4 terawatt hours of power in the late February – late March 2030 season. This is equal to only 7.5% of potential output. The CCC itself admits that the gas back-up stations will be used for ‘less than 20%’ of their capacity.[6]

An important question is whether the late February/late May 2013 electricity generation figures are reasonably typical. Multiplied up to the entire year, the figures under study suggest a total electricity generation of about 330 terawatt hours, just less than a quarter of actual total annual need for this year. In other words, the period I used has average daily electricity needs of very slightly below the mean for the year. But the variance is only about 5% and would not significantly affect the results in this paper: the total yearly demand for gas back-up generation will probably mean that CCGT plants stand idle for a very large portion of the time.

The CCC estimates that CCGT stations cost about £600 per kilowatt of power. The public policy question is therefore whether 40 GW of back-up stations are worth the £24bn that they are likely to cost (and for which electricity users will have to pay). If the late February/late May 2013 period is typical, these power stations will only provide about 8% of the total electricity production of the UK in 2030.

Storage

Swings in wind power production and variations in daily demand mean that electricity storage will become increasingly important. However, the results of my analysis suggest that conventional energy storage technologies are not particularly helpful in assisting management of electricity deficit and surplus. The reason? Periods of surplus, usually created by windy periods of a couple of days, generate far more electricity than could conceivably be stored using conventional technologies.

To validate this last assertion, I modelled the addition of 50 gigawatt hours of storage to the electricity network. The current storage capacity for UK electricity, almost entirely in the form of hydro-electric power plants that pump water to a high reservoir when demand is low and let it flow downhill when electricity is scarce, is only about 10 gigawatt hours. 50 GWh is a five times expansion of this capacity, which will also be created largely from new ‘pumped storage’ reservoirs, principally in the Scottish Highlands.[7] But even 50 GWh makes very little difference, particularly when the UK in 2030 will often have days or weeks of consistent surplus in the warmer, lighter six months of the year.

My simple model suggests that the periods of electricity surplus in the three months under study created a total excess of around 13.4 terawatt hours. Less than 10% of that (1.1 TWh) could be stored and regenerated as electricity during times of deficit in the hours and  days after the surplus was created.

Even huge scale expansion of conventional storage, costing billions of pounds, doesn’t solve the fundamental problem that electricity surpluses and deficits will not principally be diurnal, balancing out during the course of day, but multi-day (in the case of typical Atlantic storms in winter) or, more importantly, seasonal. With the pattern of electricity generation capacity proposed by the CCC, the UK will be in deficit in winter and in sustained surplus during the summer. 50 GWh is about 6% of the average daily electricity demand of the UK and therefore incapable of being a significant contributor to electricity supply. To put this more vividly, mid-April 2013 saw consistent winds that would have resulted in a total surplus of 4.0 TWh over a five day period (more than 1% of total UK annual electricity need). 50 GWh (0.05 TWh) of storage reservoirs would have reused little more than 1% of this.

Exporting the surplus

Instead of storing the electricity abundance during windy weather, the UK could export the surplus to countries linked to the National Grid through interconnectors. (At the moment, the UK is typically a net importer of power from France and other places).  Interconnectors have a limited capacity to take current. Presently, the links to France, Netherlands and Ireland have a total size of about 3 GW.

I modelled creating export capacity of 5 GW, 8 GW and 10 GW for surplus UK electricity. I did this by looking at each half hour period in the three month study period and when there was a surplus calculating whether it could be carried abroad on interconnectors of the three sizes. The results are in the table below.

Table 3

How much of the 13.4 TWh hour surpus in late February/late May could be exported?

Interconnector capacity 5 GW 8 GW 10 GW
 
Exported 5.2 TWh 7.5 TWh 8.9 TWh
Remaining surplus 8.2 TWh 5.8 TWh 4.5 TWh
 
Total surplus 13.4 TWh 13.4 TWh 13.4 TWh

 

This analysis shows that a 10 GW interconnector could accept about two thirds of the surplus generated in the three month period. Even a 5 GW interconnector would be able to export over 40% of the excess. The problem remains that when the UK is in surplus, the rest of Europe probably will be as well. High winds over the British Isles will mean excess in other countries close to the UK. As an illustration of this, we can look at the point of maximum wind power in the UK in the three month period under investigation. At around 3.30pm on 22nd March, turbines were delivering 5.3 GW to the electricity grid. In Germany, the peak was at almost the same time.[8]  On that occasion Spain saw a daily peak of 12 GW an hour earlier than the UK.[9]

At times when the UK has surplus electricity, the rest of Europe – also heavily and increasingly reliant on wind power – will generally also be wishing to export. Although it may be possible to transmit the excess, the price of wholesale electricity will fall to zero or below. (Electricity can trade for negative prices if producers are paid for their production through subsidies such as feed in tariffs). Building bigger and bigger interconnectors to other European countries is not the solution to the oversupply problem.

The conclusions of this paper

This paper uses actual data from spring 2013 and modified CCC forecasts for the portfolio of generating plants to project the pattern of electricity demand and supply in 2030. It shows that the requirement that electricity demand is always met implies that the UK will have to have up to 40 GW of standby CCGT plants. If the UK acquires 50 GW of wind power (a seven fold increase on today but the CCC regards the figure as achievable) then the average gas plant will be used about 10% of the time. Expensive assets will lie unused for months on end but will have to be paid for by electricity users.

As importantly, a strong portfolio of nuclear and CCS plants will mean that baseload needs are met at periods of very low demand, such as summer weekends. 50 GW of wind power will mean about 40% of electricity demand is met from wind but much of this electricity will be – in effect – wasted because it cannot be stored and exports will have no value.

If the weather conditions of 2013 are replicated in 2030, the three month period of late February to late May will result in a surplus of about 13.4 TWh in the period. Extrapolation to a full year is difficult but might be as much as 40-50 TWh, or perhaps 15% of total electricity demand.

The following question arises. The CCC says that by 2030 all low carbon technologies, including CCS (and unabated gas) will be (very roughly) at the same cost.[10] Is the rational national strategy to choose 50 GW and some PV and expect substantial amounts of dumped electricity? Or will it be better to invest in ‘Power to Gas’ the only conceivable way of storing energy seasonally?[11] [12]

Demand for heat for homes and other buildings is a larger part of the UK’s total energy requirements than is electricity. It is also varies far more seasonally. Is the right route forward to hugely over-invest in renewable electricity sources, such as PV, and then convert the surplus on a sunny July day into methane for use in December? It seems to me that if the UK wants to decarbonise the entire economy, and not just electricity production, that this might well be the right way forward.  The rise and rise of solar PV makes this more and more likely. Within two decades we are likely to see PV on a large portion of all roofs, domestic and other. This will mean, as already in Germany, that for four hours a day, six months a year net demand for grid electricity will fall substantially below current levels. If export is unavailable, then storage of power as methane may be economically attractive.

I will try to explore these topics in a further paper.



[1] Next steps on Electricity Market Reform – securing the benefits of low-carbon investment, Committee on Climate Change, May 2013

[2] For reasons that are very unclear, the CCC almost ignores PV in its 2030 projections. However, recent investment interest has resulted in rapid expansion of large PV farms using ground mounted panels. See, for example, http://www.larkenergy.co.uk/news/uk-s-largest-solar-farm-completed-and-grid-connected/. I project that falling costs of PV installations will take the UK’s PV capacity up from about 3 GW at the end of 2013 to 10 GW in 2030. For comparison, Germany already had about 34 GW in May 2013

[3] These assumptions are distilled from Next steps on Electricity Market Reform – securing the benefits of low-carbon investment, pages 30 and 31 and elsewhere in the CCC document.

[5] This information is published here: http://www.bmreports.com/bsp/bsp_home.htm . 3 months power production data is located by clicking on ‘Current/Historic’ underneath the ‘Generation by Fuel Type’ graph. These charts work best in Internet Explorer.

[6] Page 31 of the CCC’s Next steps on Electricity Market Reform – securing the benefits of low-carbon investment

[7] The utility SSE has planning permission to build a 30 GWh pumped storage plant at Coire Glas at a cost of £800m.

[8] The German figure is taken from the data produced by Dr Bruno Berger at the Fraunhofer Institute. (www.ise.fraunhofer.de)

[9] Spanish electricity production data can be found at https://demanda.ree.es/demandaGeneracionAreasEng.html

[10] See Figure 1.3 in the CCC’s recent Next steps on Electricity Market Reform – securing the benefits of low-carbon investment

[11] Power to Gas refers to any process that takes electricity and uses it to generate hydrogen through electrolysis and then adds the hydrogen to CO2 to make methane, the principal ingredient of conventional natural gas.

  1. Michael Lloyd’s avatar

    Interesting article and a good start on this subject.

    I am a bit wary of the figures for the German paper on conversion of surplus electricity to methane. It looks like the theoretical maximum energy conversion has been quoted and does not take into account the energy requirement for running the whole industrial process. Nevertheless, it looks like a better bet than expecting wholesale conversion of main gas heating to heat pumps.

    One of the many future imponderables is going to be the impact of smart metering. This may bring in much greater differential pricing to modulate demand.

    One thing that did surprise was that DECC do not have energy consumption statistics down to local authority level on anything less than an annual basis. No figures for how demand varies monthly let alone daily. I would have expected some figures to be available from both the mains electricity and gas grids. We can anticipate a very highly distributed generation network and encouraging local consumption of local generation would seem sensible.

    I look forward to your next paper I this series.

    Mike

  2. Samuel’s avatar

    I’m wondering why you did not take in account the obvious conclusion that this study should have brought: instead of adding more wind capacity we should add more base load using nuclear.

    But still a refreshing writing showing the too often ignored externality of an energy source like wind.

  3. Oliver Tickell’s avatar

    Chris, your idea to produce methane from surplus wind electricity is very sound. Also of course it produce hydrogen and in turn ammonia as precursor to nitrogen fertiliser. If the process can be miniaturised, this would be very useful for remote wind assets. And of course ammonia can also be used as automotive fuel – much less problematic / dangerous than hydrogen.

    Another thing your analysis reveals is the considerable unhelpfulness of “baseload” – ie, always on – power in a high-renewables scenario. Get rid of the very expensive nuclear and coal/CCS baseload and you can have far higher renewables penetration, and save huge amounts of money.

    One thing you failed to mention at all is demand management – the need to modulate demand so as to soak up peaks of production, and to reduce when supplies are scarce. This can be as simple as having all the UK’s cold stores, immersion tanks, etc, going into overdrive to use up surplus wind/solar electricity, and coasting along when power generation is low.

  4. Oliver Tickell’s avatar

    @Samuel – actually you have got this the wrong way round. Baseload power now a very old-fashioned concept, as it will never rise to meet peak demand, or drop in line with troughs. It’s a bit like the stopped clock, that is right twice a day. Not much help. It is also fabulously expensive. EDF holding out for £100/MWh for 35 years, plus construction price guarantees, plus fixed price decomm and waste “management”, plus public liability waiver. And even then it looks like investors are wary, looking for more like 12% return than the 10% UKG has agreed on. And while renewables costs are constantly falling – solar PV now cheaper than nuclear as Chris showed last week – nuclear costs just rise and rise. A few years ago EDF was saying it needed no public subsidy for Hinckley C. It is now demanding a subsidy package that would cost UK consumers around £150 billion.

  5. Robert Wilson’s avatar

    Chris

    One of the rather obvious problems here is where the CO2 comes from in this power-to-gas scenario. The trials in Germany (http://www.rwe.com/web/cms/en/1846074/rwe/innovation/projects-technologies/energy-storage/project-power-to-gas/) are taking the CO2 direct from coal plants.

    Now, we could take the CO2 from the CCS plants and run them through this power-to-gas process, but I’m baffled by the point of this. Why not just sequester the CO2 from CCS and just burn plain old fashioned gas that we have taken out of the ground instead? It’s hard to see how the emissions will be lower using power-to-gas. I’m not a chemist, so I may be wrong on this, but consideration of the equation for methane combustion doesn’t bode well for this scheme:

    CH4 + 2O2 -> CO2 + 2H2O

    So, say we want to heat a million houses with gas. We can use plain, taken out of the ground, gas, or we convert CO2 (from a power plant) into CH4. The problem here is that the conversion process is not 100% efficient, so we will lose some CO2. How much is probably hard to find out until it is trialled extensively. However, it seems clear that storing the CO2 would likely result in lower emissions overall than converting it to methane.

  6. Michael Knowles CEng MIMechE’s avatar

    Chris – re your statement that Wind. Offshore and onshore farms are projected to both contribute about 25 GW of capacity. The key issue is whether offshore wind costs decline at the rate expected by the CCC.

    TABLE1 says 50 GW for 2030, and assumes offshore wind cost will reduce to presumably the hoped-for£100/MWh wher currently the subsidised RO cost is 2ROCs X 44 (2011/12OFGEM RO Accounts) + 44 wholesale cost/MWh =£122/MWh and all from offshore nearshore R1 & 2 wind farms that are underperforming, on average, by 18%.

    To get 25 or 50GW capacity inevitably means going further offshore for R3 concessions. Here the DECC thinks that the costs will be much higher 2011 RO banding report said ““Analysis suggests a ROC range of 2.0-3.0 ROCs for R2 offshore wind (with an operation start in 2014) and an ROC range of 2.6-3.9 ROCs for R3 offshore wind (with an operation start in 2017)”. ref (2) IMechE response to DECC consultation on RO banding levels for 2013 to 2017http://www.imeche.org/docs/default-source/public-affairs/3236-consultation-ro-banding-response-form.doc?sfvrsn=0

    We must be realistic about future costs for offshore wind! Onshore wind developments will struggle to meet the future predictions for growth, won’t they?

    Storage whether by pumped hydro or other will be very expensive. I share
    Michael Lloyds wariness over conversion of electricity to methane and other simialr schemes such as the Slough Cryogenic storage. What overall efficiency will the operate at? National Power (RWE) tried the 12MW Regenysis electrolyte flow storage at Didcot back in 2003 and it was found to be too expensive. NPower were aiming for £1000/KW at the time!! Where has it gone to?

    Clearly it is worth trying to get the storage cracked but —–????

  7. Robert Beith’s avatar

    The impression is given that wind variation is seasonal but in fact there is continuing cycling between full load to minimal load frequently over each month. With 50 GW of wind this becomes a significant challenge. With Nuclear, CCS and biomass necessarily as base load there will be big swings needed to be balanced. Exporting is a sensible concept but we are limited to about 4GW by cable so far. Currently Germany who had 29GW wind and 25GW solar by end 2011 were exporting and importing about 10 times as much as we are currently for balancing purposes. It is worth noting their solar only produced 3% of total output. Renewables performance varies and can disappoint and again we need enough balancing for the worst not the average scenario. So I would go along with 40GW gas back up. The 24 Billion Pounds stated as the build cost is of the order one third the cost of building the same wind turbine capacity. Perhaps the economic solution would be for wind turbine farm owners also to own gas back up plant, then whichever was operating they would get paid for!
    Finally I consider a higher nuclear content essential–It is only build costs of new designs which are extremely expensive but they have a design life of 60 years typically (20 years for wind)and operating costs reduce and are stable.

  8. Martin Normanton’s avatar

    Surely the first use for surplus electricity is for heat (process heat, water heating or space heating) as a substitute for gas. This is what is already done on Fair Isle where excess wind electricity is sent to domestic immersion heaters and storage heaters to reduce the use of oil for heating. This would start with large users of gas etc. for process heat being persuaded by very cheap elec at times of surplus to make their plant dual fuel. This would also be greatly helped by a carbon tax (or a cap and trade system which actually worked) to make the gas alternative more expensive.
    Remaining surplus electricity could then be used to electrolyse water as you suggest. This is done most efficiently at high temperatures. (source Wikipedia – High temperature electrolysis) So the electrolysis plant and gas turbine should be sited near baseload plant (fossil fuel, biomass or nuclear) to use some of the waste heat.
    But why make methane rather than just hydrogen? Burning the methane releases the CO2 previously captured somewhere, unless it is captured a second time. Burning the gas, hydrogen or methane, in a gas turbine on site (when demand is high) has the same effect as putting it into the gas main, but gets around the difficulty that the gas mains are insufficiently impervious to hold the tiny hydrogen molecule well.
    Be jealous of countries with large resources of high head hydro, such as Norway which allows Denmark its high wind penetration via their grid links. A link from Iceland to Britain (and on to the continent) has been proposed to allow us to import cheap hydro electricity, but if we also exported surplus wind electricity down it that might improve its economics.

    Finally you show up the folly of heat pumps in the UK where they only exacerbate the seasonality of electricity use. By contrast electric cars are a year round use, and much of it can be overnight when electricity is in surplus supply.

  9. Chris Clews’s avatar

    Chris, have you fully considered using hydrogen instead of methane, not for combustion but in situ generation of electricity for transport and industry via fuel cells? Though greater efficiency that’s a lot less to carry around so less “dangerous” – but same could be said for any energy concentration? Surplus electricity could be used to produce hydrogen by electrolysis of water. Very simple cycle then?

  10. Michael Knowles CEng MIMechE’s avatar

    Re Martin’s comment on using the surplus electricity for heat or electrolysis, the electricity is already very expensive from renewables viz., 8.5/kWh for onshore and 14p/kWh offshore wind and that before it is transmitted, ditributed and delivered to your home and Gov. obligations and taxes (taxes e.g., 5% VAT environmental and Renewables Obligation 14%),operating cost paid for & profit added for the supplier. So add another 10p/kWh to make it 18.5 to 24p/kWh or 4 to 5 times the cost of gas. Why should the surplus electricity be subsidised? Only to satisfy unrealistic and uneconomic ideals.

    Heat pumps are used mainly in the winter when the demand for elctricity is highest too.

  11. Martin Normanton’s avatar

    In reply to Michael and to summarise my post above: using (surplus) electricity to make hydrogen (or methane) which inter alia can then be burnt to generate more electricity is a pretty inefficient process, while using it for heating is also wasteful. So it is a matter of choose your poison. If you have sufficient pumped storage capacity (not realistic in the UK) using that is a no brainer as it is about 85% efficient. But we should not refuse to sell renewable elec at a loss, since just dumping it or switching the WT off is an even bigger loss. Remember that the cost saving from switching a WT off is minimal (a bit of maintenance saved), while switching off a fossil plant saves real money for the fuel, not to mention the carbon saving.
    “Heat pumps are used mainly in the winter when the demand for electricity is highest too”. That is what I was trying to say, that heat pumps make the seasonality problem worse. They make sense in countries like Japan or California where peak elec demand is in the summer.

  12. John Goldsbrough’s avatar

    Martin
    ‘Be jealous of countries with large resources of high head hydro, such as Norway which allows Denmark its high wind penetration via their grid links.’

    Instead of being jealous we could develop our own storage.
    http://www.isentropic.co.uk/
    http://www.highview-power.com/

    However, my view is the problem will be solved with tens of millions of micro electrolyzers, fuel cells and batteries, linked together in a smart grid.

    Michael
    Is renewable energy very expensive?
    If I peddle hard on a bicycle generator I can produce about 200W.
    It would take me 5 hours of hard labour to get a kWh. 24p now seems unbelievably cheap. As a bonus it is relatively clean and sustainable.
    The real problem we face is that coal and gas are dirt cheap.

  13. Chris Goodall’s avatar

    Chris,

    Absolutely agree for small quantities. But large-scale storage of hydrogen is extremely problematic. (By ‘large scale’, I mean hundreds of gigawatt hours). Whereas methane (3 times the energy density of hydrogen) can be stored in the pipes, tanks and reservoirs of the natural gas grid.

    Chris

  14. Chris Goodall’s avatar

    Robert, avoiding the need to store the CO2 in aquifers and depleted hydrocarbon fields saves money. Perhaps £20=30 a MWh. That’s why it seems to me, perhaps wrongly, to be better to methanate the CO2 and then reburn it. Or put it in the gas grid where it can be burnt by my nice new condensing boiler as ‘low carbon’ gas at 90% efficiency.

    Chris

  15. Chris Goodall’s avatar

    Sam,

    I don’t disagree that it *might* be better just to build nuclear. But we will still face a 2.5 times winter maximum to summer minimum electricity demand ratio. So we might well always have a significant need for storage, particularly as grid parity solar PV will sharply reduce net summmer electricity demand.

  16. Paul Hughes’s avatar

    Chris,

    Excellent article and approach to the ‘waffle’ surrounding intermittancy and renewables. Some thoughts

    1. The CCC model for 2030 sees a significant electrical load for plug in electric / hybrid cars. Clearly, the connection of large battery capacity to the grid at peak demand (early evening), with the capacity to re-charge either overnight or during the day, would go some way towards addressing the storage question.

    2.

  17. Joseph’s avatar

    Very interesting article. A few points, which don’t necessarily affect the validity, but just observations.

    Wind should not scale exactly with capacity, but should be a slightly smoother version of 2013, due to more geographically distributed generation, and also more offshore wind, which has higher load factors and more consistent wind generation.

    I’d question the feasibility of a lot of the generation capacity mix. Beyond the scope of this I know, but regarding the baseload generation:

    First nuclear plant is scheduled for 2023, but negotiations on a seemingly exorbitant CfD price are ongoing, cost overruns and delays on the two being built in Flamanville and Finland (6 years late, 5.5bn over budget), make that questionable. Of course, should be better on the next ones, but assuming 8 plants by 2030 means 1 a year. I doubt if the industry can support that pipeline, or that utilities can financially support it, when EDF is heavily debt burdened and can just about get the financing for one, and is close to walking away from that, even with government guaranteed revenue.

    As you say CCS should run baseload. If it is technologically and economically viable, (as of yet there have only been small trials), I believe it has to, the CCS doesn’t work well with varying output. Point above about utility finance also holds. The cost/MWh is comparable to nuclear, but with higher fuel costs, I don’t see the attraction, except for converting existing plants, but by 2030, Drax might be the only coal plant left. All the other will be 60-70 years old by then. Maybe for new or recently built CCGTs?

    It’s dubious about the carbon savings of large scale biomass. See for example. When Drax have announced that they’ll burn wood from an area 4 times the size of Rhode Island to convert half the plant (1.8GW?), currently shipped in from Georgia, I don’t see how this will be acceptable or possible.

    50GW of wind by 2030, when there’s about 8GW on the system now means we need to build 6-7MW/day, so 1-2 turbines, depending on size. Ambitious to say the least.

    Interconnector will give only limited ways to alleviate excess wind generation. Interconnectors are expensive. The recent UK-Irish one, EWIC, cost about €1000/kW, and it’s relatively short. Extrapolating, building 10GW to export excess wind, would be over €10bn then. It also would go unused much of the time, given the variability of wind, and so would have to recoup it’s costs over a much smaller amount of generation. And as you say, when it’s windy enough across the UK for wind to be that high, it’s probably windy across most neighbouring areas. Last Sunday due to German wind and solar, prices in France went to -€200/MWh, So you would have had to pay €200 to export a MWh of power, or be paid €200 to import one. That’s higher than UK subsidies, so it would make sense to import and constrain off even more UK wind. As you say, interconnectors are not the solution, and may even exacerbate the issue depending on the structure and subsidies of the connected market.

    Constraints such as this happen occasionally in Scotland already, where windfarms are paid not to generate as there is limited connections between Scotland and England. With that much more wind, it will happen much more frequently, but constrained within GB, not Scotland. Obviously expensive, as windfarms will expect to get their subsidy anyway. See the Telegraph for complaints about sporadic times where windfarms get £100k for shutting down overnight. That will happen much more frequently, and in much higher £ figures.

    I also, as other commentators have said, don’t see the economics of creating methane as a storage option. Even if the conversion was 100% efficient, and I assume it’s far from that, 1 MWh of power is the energy equivalent of 34 therms. At 4 ROCs/MWh for offshore wind and £40/ROC, that’s £4.70 a therm. Onshore is 1 ROC, so about £1.20. Current market prices are somewhere around 60p.The high DECC gas price projections for 2030 are about £1, so it’s an uncompetitive way of storing gas. Of course, if you’re committed to pay the ROCs no matter what, the power is effectively free, and the £1 you get in 2030 just has to cover the cost and running of the conversion infrastructure. Whether or not it does, I’ve no idea.

    Also though, it does also have to cover the cost of storage. There’s limited amount of long term storage in the UK, basically Rough, which is reasonably expensive, and also fully in use. The economic case for new storage in the UK is limited and several projects have been cancelled in the past while. They basically depend on a consistent Summer-Winter spread in gas cost. This has dropped significantly in recent years as UK production declined. Now the marginal gas supplier is LNG which has no seasonality. Also, if all those heat pumps and other efficiency measures come in, that should reduce the excess Winter gas demand, further reducing spreads and the need for storage. So effectively if there is a low or no Summer-Winter spread, storing it for winter doesn’t make economic sense.

    It might be more viable to store it short term, and burn it in a CCGT when wind drops off again. Even then, the cost of conversion and storage infrastructure and running have to be covered, so it depends on the economics of those. It could also be competing with gas that is much lower than the high DECC case, e.g. if shale has an impact. There will be some shale, but how significant it will be, I don’t know.

    More generally, without some subsidy scheme, it would be an investment with significant risk, and risk free investments (wind, nuclear, networks, etc) are sucking up all the limited available investment capital that utilities have at the moment. Given the complete dysfunction of the European power markets now, and the fear the same could happen in the UK, I think it’s unlikely that even if assumptions show it’s viable that it will progress without a guaranteed return. See for example the desire of UK utilities for a capacity mechanism to eliminate investment risk in the new CCGTs needed, and the refusal to invest otherwise.

Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>