3.2 Inputs

3.2 Inputs Bruno Prior Tue, 15/12/2020 - 10:15

The choice of 2030 places some constraints on our assumptions. The lead-time on new nuclear is such that we can only assume that Hinkley Point C will be operating (3 GW). Meanwhile, most of the existing nuclear fleet should be closed by then, leaving only Sizewell B (2.4 GW). This is consistent with the Prime Minister’s recent announcement that the UK will refocus on small, modular nuclear. That may pay off in the long run (although the UK’s track record of doing nuclear its own way is poor), but from a standing start for an unproven technology, it’s unlikely to be contributing much by 2030. For the nuclear evangelists, this is not about what’s desirable, just what’s probable. You can re-run the model assuming more nuclear in 2030 if you think it’s credible.
We can assume the coal-fired power stations will be gone. They are almost dead already, with levels of utilisation so low that it cannot be worth keeping them operational for much longer. And even if it were, their lifespan is limited by the Industrial Emissions Directive, which will survive Brexit broadly intact.
Carbon pricing has already killed most oil-fired generation. Apart from small, standby/peaking units, the UK is unlikely to add much more.
Instead, the UK will build offshore wind farms, and lots of them. The Prime Minister has re-affirmed the government’s intention to quadruple offshore wind capacity by 2030. That means approximately 40 GW. New capacity will have a materially higher availability factor.
Onshore wind plays a smaller part in the government’s plans, but has continued to expand in the face of political ambivalence (at best). We can assume a few more GW of onshore wind, but not the scale of the offshore expansion. It is already over 10 GW, so we can assume more than that. Re-powering may add materially more even if few new windfarms are developed. New capacity will have a higher average availability factor: 35% vs the current 28%.
Likewise, the boom days for solar are past, but it continues to expand at a more measured pace, helped unintentionally by the hikes in electricity retail prices that owe much to the costs of delivering previous governments’ green agenda. Like onshore wind it is already over 10 GW. We can allow for a few more GW of solar by 2030.

There is little reason to believe there will be a significant further expansion of solid-biomass generation. The low-cost opportunities (converted coal-fired plants) have by-and-large been implemented or lost. New plants are too expensive in an environment where support is less generous, wholesale values are under pressure, and intermittents hurt load factors.

Technology	2016-18 (MW)	2030 (MW)
Nuclear	9000	5400
Coal	15000	0
Oil	1000	1000
Gas	34000	34000
Hydro (exc storage)	1600	1600
Biomass	4000	5500
Biogas	1000	1000
Offshore wind	5000	40000
Onshore wind	8000	13000
Solar	13000	18000
Total	91600	119500

The government will introduce another attempt to stimulate biogas (the Green Gas Levy), but we set out in the Introduction some of the reasons why this is likely to have limited success. Meanwhile, landfill gas is on its descent to oblivion, having lost most of its feedstock several years ago, thanks to the Landfill Directive and other measures to divert putrescibles. Most of the low-hanging fruit in sewage gas has been implemented. Thermally-produced biogas (e.g. gasification) is likely to prove as much of a mirage this time as it has proved previously. In any case, any new biogas will be pushed strongly towards grid injection rather than electricity generation.
New, big hydro is nearly inconceivable in the UK.
Not much new gas-fired generation is likely to be built under current conditions. But that may change when it becomes apparent that the UK needs to reinforce its dispatchable generation to compliment the government’s low-carbon plans. We will start by modelling the current position (i.e. assuming little extra gas-fired generation), but will modify that assumption when it is clear how problematic it is.
Instead of dispatchable capacity, we will model initially a significant amount of storage capacity as the favoured balancing solution of many techno-evangelists. For economic reasons, most storage is currently designed for short (e.g. 12-hour) charge/discharge cycles. We will start by modelling that sort of capacity (20 GW, 120 GWh of battery storage, on top of the existing pumped storage), before modifying our assumptions to include some longer storage, when we have seen that short-cycle storage does not address all the challenges of intermittency.
We also assume that interconnector capacity has been increased materially, by approximately 50% to 9.55 GW.
We will continue to improve the efficiency of lighting and equipment, although much of the low-hanging fruit has been picked. We’ll put electricity consumption for this use at 175 TWh (down from 184 TWh). Increasing population would otherwise mean higher demand, so the reduction is more significant than it appears.
Cooling demand (i.e. air-conditioning) should receive a double boost, partly to counteract rising temperatures, and partly because electrification will mean massive deployment of heat pumps, many of which will be capable of providing cooling as well. An increase from 5 to 10 TWh is a relatively conservative projection under these conditions.
It is widely recognised that improving the efficiency of our leaky buildings is an important part of decarbonising heat.
1. Not only does it reduce the amount of energy required for heating, but it also makes it more feasible to install heat pumps, one of the two core technologies for decarbonising through electrification (the other is electric vehicles).
2. Government data show how modest are the improvements that can be achieved through efficiency retrofits, especially when most of the easiest retrofits have already been installed under a series of incentives (CERT/CESP/EEC/ECO etc) over decades. But the UK government has decided once again to ignore its own data, and promote retrofit rather than rebuild to improve efficiency. We reflect this in our assumptions for this scenario.
3. Government statistics also indicate the numbers of buildings that have cavity or solid walls and lofts suitable for insulation, and categorise them according to whether they already have a high level of insulation, or whether they are easy or difficult to bring up to that standard. Noticeably, the majority of cavity walls and lofts have already been improved to a good standard, whereas most solid walls (which are much more difficult and expensive to insulate well) are poorly insulated. We assume for this scenario that by 2030 we will have improved all the remaining cavity walls and lofts that are deemed easy to upgrade, and will have done half the solid walls that need improvement. Most of the outstanding cavity walls and lofts are considered easy to upgrade, so this leaves less than 10% of these surfaces below standard.
4. Likewise, most windows are already double-glazed, but we assume for this scenario that most of the remainder will have been double-glazed by 2030. It is unlikely to be 100%, because existing windows in listed buildings can be difficult to upgrade.
We assume that there will be little change in the amount of heat used for hot water and cooking in buildings. Remember that, unlike national statistics, we define heat as the amount of energy used after conversion losses, so improvements in efficiency (e.g. through heat pumps) are incorporated at a different point in the model.

The government’s focus on retrofit rather than rebuild, the opposition of their back-benches to planning reform to expand housing supply in their constituencies, and the likely reduction in the rate of population growth post-Brexit, means that we assume a relatively conservative amount of new-build.

TWh	2016/18	2030
Existing homes
Space heating	271.6	262.8
Hot water	74.8	73.3
Cooking	12.0	11.5
New homes
Space heating	-	5.5
Hot water	-	2.9
Cooking	-	0.4
Services
Space heating	97.9	95.0
Hot water	14.5	14.5
Cooking	23.0	23.0
Industry
Space heating	19.6	18.6
Process (high)	36.3	37.0
Process (low)	58.4	59.6
Drying	18.6	19.0
Other	22.5	22.5

We have allowed for an extra 800,000 houses and 400,000 flats by 2030, built to a Fabric Energy Efficiency Standard quality suitable to achieve at least Level 3 under the Code for Sustainable Homes.
We assume that the UK’s recent track-record of building some of the pokiest little homes in the world will continue, as a reflection of our housing costs and planning incentives: we take 92 m2 on average for a house and 57 m2 for flats. Building energy consumption is broadly (in theory, ignoring practical behavioural factors) the product of the efficiency standard and the floor area.

We assume that the ambitions of all parties (including the current government) to sponsor a “green industrial revolution” and to bring some offshored manufacturing back onshore will yield (allowing for some process-efficiency improvements) modest increases in industrial activity reflected in modest (c.2%) increases in industrial energy consumption.
The combined effects of these assumptions (including the limited effect of building-efficiency retrofits) on the assumed energy requirements can be seen in the table above.
In total, that makes 645.6 TWh of heat needing to be generated, of which 507.5 is “building heat” (i.e. space heating, hot water and cooking), of which 363.3 TWh is space heating. (Again, remember, that is post-conversion, which is why it is a smaller figure than in government statistics).
For this scenario, a lot more of that heat needs to be supplied by heat pumps than is currently the case. But the government is hedging its bets on heat decarbonisation, with large demonstration projects planned for hydrogen as well, so they are unlikely to have pushed heat pumps to a dominant position by 2030.
1. Experience has shown that air-source heat pumps (ASHPs) are likely to dominate other types of heat pump (ground/water-source), because of their cost and simplicity. This model assumes they will be supplying 25% of our heat by then.
2. Ground/water-source heat pumps (GSHPs) will also have increased significantly from their current, modest levels. We assume they will maintain a similar ratio to ASHPs as now, and allocate them 2.5% of the heat market in 2030.
Most electric heating is currently direct (i.e. resistive) heating. This is split between industrial uses, for which it will often have special characteristics that mean it is unlikely to be changed, and building heat, much of which is likely to be replaced by heat pumps under this model. We assume direct electric heating’s share of the market will have halved to 6%. (Electric heating appears to contribute less than 12% in national statistics, but [again] that is on the basis of energy input not heat output.)
Solar thermal has proved surprisingly uneconomic under the RHI. We assume its contribution barely nudges upwards, to 0.1%.
The national statistics in some cases lump biomass boilers, wood stoves and fires and other bioenergy together, and in other cases separate them into arbitrary categories (“plant biomass”) that have little connection to any real-world distinction. We split solid biomass heat into two categories, representing the fundamentally different technology, efficiencies and uses employed: (1) biomass boilers, and (2) wood fires and stoves.
1. The RHI turbocharged and then throttled the biomass-heat industry. Policy is now highly adverse. But it remains the only practical technology for decarbonising many buildings. We assume there will be some modest growth, rounding up from 2.73% to 3% of heat.
2. Much of the domestic wood-burning in the national figures is a statistical fiction. But its contribution looks less significant anyway expressed in terms of heat output rather than fuel input, because of the low conversion-efficiencies assumed for those statistics. Until the fiction is acknowledged, we have retained it as the basis for our default figures for this technology. However, we can reasonably assume a material reduction in this technology’s share by 2030, as government policy is exceedingly hostile because of concerns about urban air-quality. We have assumed that its share of heat output will have shrunk from 1.8% to 1% by 2030.
Biomass CHP is an excellent technology that can maximise the energy-value of the fuel in suitable applications, whilst functioning at a scale that allows remote location and substantial back-end cleanup to address the air-quality concerns around smaller biomass heat. However, the best applications for CHP are relatively stable heat loads, unlike the majority of heating uses. The best way to marry the two is large-scale district heating, but governments remain all talk and no trousers in that regard. Retrofitting district heating where people have existing gas boilers is an engineering and commercial challenge, and governments have been reluctant even to do much to ensure that district heating is incorporated into substantial new developments. Perhaps they are deterred by the distribution losses, though the real reasons are probably more political than technical. The opportunities for biomass CHP will therefore probably remain limited, especially as it seems to have been overlooked in post-RHI proposals. We have assumed its share, like solar, only nudges up to 0.1%.
The government has consulted on a Green Gas Levy to try once again to stimulate biogas grid-injection. But we explain in the Introduction why this is likely to disappoint.
1. Anaerobic digestion for grid injection will remain constrained by the resource of suitable feedstock, unless governments go for massive energy-crop planting like Germany. The politics of that makes it unlikely to happen. This scenario assumes the GGL will push its share of heat up to 1%.
2. SynGas from gasification is treated as biogas or green gas nowadays. The problems with the technology remain as plentiful as they were when I helped write a report on it 30 years ago for Southern Electric, or as they were when Ernst & Young and National Grid were predicting 11 years ago that it would be making a massive contribution by now. Someone may crack it one day, but we won’t count the chickens on this technology. We have assumed there will be some response to the GGL stimulus, which looks like a manyfold increase given its negligible penetration to date. But that still only takes it to 0.2% of heat in this model.

Bioliquids are a convenient way for people with oil-fired boilers to decarbonise. But there is a lot of competition for alternative uses for a limited supply of fuel, and we assume that this fuel will continue to be focused mainly on road transport. We have allowed for its share multiplying, but that still only takes it to 0.1% of heat.

% / TWh	2016-18	2030
ASHPs	2.46% / 16.0	25% / 161.4
GSHPs	0.18% / 1.2	2.5% / 16.1
Direct electric	12.03% / 78.4	6% / 38.7
Solar	0.09% / 0.6	0.1% / 0.6
Biomass boilers	2.73% / 17.8	3% / 19.4
Wood fires	1.8% / 11.7	1% / 6.5
Biomass CHP	0.09% / 0.6	0.1% / 0.6
Biogas (AD)	0.73% / 4.8	1% / 6.5
Biogas (SynGas)	0.01% / 0.07	0.2% / 1.9
Bioliquids	0.01% / 0.07	0.1% / 0.6
Coal	2.39% / 15.6	1% / 6.5
Oil	8.96% / 58.4	4% / 25.8
Gas	68.52% / 446.3	56% / 361.5

One of the inexplicable idiosyncracies of British energy policy is that very little has been done to tackle some of the lowest-hanging fruit in the sector: oil- and coal-fired heating. Indeed, oil-fired heating benefits as much as any other domestic heating fuel from the special low VAT rate. So little has been done that oil’s share of the heat market has actually increased over recent years, despite governments’ claims to be trying to decarbonise. We assume there will finally be some action on this by 2030:
1. Coal’s current share is a surprisingly large 2.39% (more, if measured in terms of fuel input). We assume that will have been knocked down to 1% by 2030.
2. Oil’s current share is nearly 9%. We assume that will be down to 4% by 2030.
Our model assumes gas continues to supply the balance. Following the government’s recent announcement, some of that may be hydrogen, but our model does not yet cover that, and it’s premature to make assumptions about such an immature technology. Excluding hydrogen, the balance of the heat market supplied by gas (lumping natural gas and LPG together) is 56%.

Much of the focus of the recent policy announcements was on electrifying cars (not transport, nor even road transport in general, but cars specifically). Cars are a very large part of total transport energy-use. The plan to ban the sale of petrol and diesel cars and vans by 2030 will not eliminate those cars from the road in that year. But it will skew the market heavily in favour of electric vehicles as 2030 approaches, because of the impact on resale values. Assuming the policy holds, we can assume a roughly linear increase in electric cars’ proportion of total sales towards 100% in 2030. That should mean that electric cars constitute a large proportion of the total by 2030.The fact that the government has not announced similar plans for other road vehicles indicates the greater difficulty and lower priority of electrifying them, and we may assume electric vehicles constitute a smaller proportion. As this is largely a question of sticking a finger in the air, let’s go with a round number and assume half of road-vehicle fuel is electric by 2030. Given the differences in efficiency, that translates into around 71 TWh of electricity and a residual 241 TWh of petrol/diesel, out of the current 482 TWh.

TWh	2016-18		2030
TWh	Fossil	Electric	Fossil	Electric
Road	482.0	0.2	241.1	71.0
Rail	12.4	4.6	3.4	8.0
Air	151.0	0	151.0	0
Water	10.8	0	10.8	0

The recent announcements included nothing abour rail electrification, but there is a longstanding intention to electrify those parts of the network for which it is feasible. Again, there is little of substance on which to base a 2030 estimate. We assume there has been material progress in electrification, from 4.6 to 8 TWh, but a material amount (3.43 TWh) remains diesel.
There is little substantial policy on decarbonising air and water transport. We assume no progress.
We use the default costs described in section 2.2.
We start with seed data from 2017, assuming normal (i.e. not mild or severe) weather. All three years of seed data (2016-18) were somewhat milder than the historic average, so the “normal” scenarios are probably slightly on the mild side of a “normal” year.