This evening I attended an interesting meeting hosted by the Energy Institute, and held at the Royal College of Nursing in Cavendish Square, London. The speaker for the event was Dr Scott Milne, of the Energy Technologies Institute (ETI), who introduced us in a “meet the public” way to the recent launch of two sample scenarios for the future of Britain’s energy : “Clockwork” and “Patchwork” from the ETI’s Energy System Modelling Environment (ESME).
What follows is me typing up my notes that I made this evening. It is not intended to be a literal or verbatim, word-for-word record of Dr Milne’s words, as I took the notes longhand and slowly. Where I have put things in square brackets ( [ ] ), they are my additions.
[ Before the talk, I chat with somebody whose name I didn’t catch, who in all honesty asked me whether I thought fusion nuclear energy would be a likely energy technology choice by 2050. ]
So, what is the ETI ? It’s a public-private partnership, aimed at de-risking various technologies and technology families. We receive funding from BP, Shell, EdF, Caterpillar, Rolls-Royce […] We have a large number of stakeholders who take the work we put out for tender to be done. We aim to build internally-consistent models – using “exogenous assumptions” [ externally-imposed ]. We have about 250 profiles in the model – costs are added in. The ESME modelling is policy-neutral – unless where we intervene to state otherwise – for example, to say no nuclear power, or Carbon Capture and Storage (CCS) to be applied later rather than sooner. Our starting point is existing stocks of energy installations as of 2010, which are gradually retired out, and we are subject to supply chain constraints in replacing them. How quickly can we deploy new solutions ? We have a “spatial disaggregation” in the model – with 12 separate regions of the UK. We have offshore nodes, and storage points, and carbon dioxide capture and storage is pushed offshore. Our modelling is not as finely detailed as the National Grid’s power dispatch model. We have seasons, and five parts of a day – a model suitable for load balancing purposes. We assume a 1-in-20 risk of a cold snap – a “peak day” of consumption. There is a probabilistic element for each technology on cost, and the modelling is done using the Monte Carlo method (repeated random model runs). This helps us to identify which technologies are optimal. Our partners DECC (Her Majesty’s Government Department of Energy and Climate Change) and CCC (Committee on Climate Change) are users of the model, and the model provides an evidence base for them. The low carbon energy research models (ESME) are used by some academic groups. We came public with these for the first time this year, and we launched on 4th March 2015.
In the “Clockwork” scenario, transport continues to be liquid fuel options as we have today, and using carbon offsets from elsewhere in the energy system. There are a few things we need to believe as part of this scenario. We need to accept the “negative emissions” possibilities of Carbon Capture and Storage combined with biomass (Biomass+CCS) – this is still certainly open to question. By 2050 there should be ultra-low carbon vehicles. These two scenarios “Clockwork” and “Patchwork” are not extremes as in some modelling done elsewhere – they are more balanced between the two. The “Clockwork” scenario is not about decisions made at the household level – whereas “Patchwork” is – it involves engagement from householders, and includes influences and constraints besides decarbonisation – for example, the cost of energy and air quality. In the “Patchwork” scenario there is a limited role for biomass in space heating, and you see a greater push for low carbon transport. Plus, space heating is decarbonised in parallel [ partly through demand reduction ].
In “Patchwork” there is less central governance. You see experimentation in different regions, and only at the end see which technologies have been picked. There is a stronger burden on households in “Patchwork”, and more emphasis on renewable energy. Coal is switched off in both scenarios by 2030, and it is not replaced by coal-with-Carbon-Capture-and-Storage (Coal+CCS) but with Natural-Gas-with-Carbon-Capture-and-Storage (Gas+CCS). In the “Clockwork” scenario there is still a role for renewable energy, but not so significant. Hydrogen gas turbine generation takes over the “peaker plant” (on-the-spot generation at peak demand) role from Gas+CCS. The hydrogen comes from Biomass+CCS. There is large scale geological storage of hydrogen. In the “Patchwork” scenario, offshore wind plays a major role – the model assumes that the land available for onshore wind is capped (that’s a choice). Solar power is also a big factor in “Patchwork”, but still making a fairly modest contribution by 2050. Also, there is an assumption that biomass contributes directly for power generation. In the “Patchwork” scenario, solar power makes a major contribution to capacity (gigawatts) but less to generation (terawatt hours).
As regards space heating (the heating of the insides of buildings) : in the “Clockwork” scenario, heat pumps make a major contribution – and there are big step changes in the final decades compared to “Patchwork”. Gas boilers are being built for the 1-in-20 year cold snaps – but not for the home [ – for district heating ]. There is a high demand for heat in the “Clockwork” scenario – where householders are “comfort takers” and homes may be heated to 21 degrees Celsius. In the “Patchwork” scenario, people have more engagement with the management of energy, better at managing their use of energy at home, and so less heat is used. There is a strong role for retrofits [ for insulation for energy demand reduction ] behind the scenes. Population continues to grow and the number of individual households continues to grow.
As regards transport : Heavy Goods Vehicles (HGV) and Light Duty Vehicles (LDV) are important (although the graph only shows cars). In “Patchwork” there is a move towards urban living – and so people will be thinking more about how transport can be done – car pooling and car sharing. In “Clockwork”, we are seeing aspirations – people flash the cash – and pay more to do more. The Biomass+CCS carbon dioxide emissions offsets create more headroom for transport emissions in “Clockwork”. The model could explore lowering demand for transport – through a shift to gas from liquid fuels – fuel/gas hybrids actually [dual fuel]. There are implications for liquid fuel – significant in both cases. There are therefore implications for fuell stations – for example, if cars are coming to the forecourt less often for fuel because of vehicle fuel use efficiency. We need to maintain the liquid fuelling infrastructure – but we need electric vehicle charging and give hydrogen refuelling infrastructure as well. There is quite an overlap in investment. Even if we stop selling liquid fuel vehicles, they will stay on the road for some time – we assume 13 years.
In terms of what it means – in terms of cost compared to its fossil fuel “dark cousin” [ business as usual trajectory ] : “Patchwork” works out to be more expensive – these graphs show capex only [ capital expenditure on investment in assets and infrastructure ]. For “Patchwork” [ although capex is higher ], the resource cost is less [ owing to more renewable energy being sourced. ] These graphs give an idea of when money needs to be spent and how much – it’s not insignificant [ between 1.4 and 1.6 % of GDP ? ] To make the investments, buildings and space heating could be considered infrastructure [ and need central spending ? ] The costs of transport are heavier in “Patchwork”. Both have “negative emissions” (from Biomass+CCS). By having “negative emissions”, you are allowed to have some of these fossil fuel options. This is important as air travel and shipping will need fossil fuels. You cannot fly aeroplanes on hydrogen, for example. The outlook for industry takes a bit more explaining.
Taking action over the next decade is a no-regrets option. We need to replace energy installations – replacing them with low carbon options gives only a marginal extra cost. We lose very little by hedging – even if carbon action doesn’t take place. Developing the technologies enhances export capability – at least we will not be an importer. If we wait to implement low carbon technologies, we have less time for the transition. This model operates over a timescale of 35 years. Development of the technologies will involve some degree of redundancy [ not all developments will be useful going forward ], but we need to prove them up, cost them out. If we wait until it is clear we must act, we will have to jump to things that are not yet costed up. If there are no technological solutions worked out, we might have to slash energy demand – which would politically be very challenging – you can imagine how people would react to having a cap on the energy they are permitted to use at home. If we attempt to make an 80% reduction in carbon dioxide emissions later on, we will have higher cumulative [ overall ] emissions – and as a result we would need tougher carbon emissions cuts.
Things we have concluded from this modelling : we are not yet at a stage where we need to say definitively what needs to be used, for example, decide for nuclear power, CCS etc. Biomass+CCS is challenging – there are questions around the lifecycle carbon dioxide emissions. But if we don’t have it, it doubles the abatement cost. We have shown that a high level of intermittent renewable energy in the power sector is quite manageable – we can use the excess in renewable electricity generation for building up renewable heat – for example hydrogen electrolysis for hydrogen production [ “Power to Gas” or “WindGas” ] – which is not modelled. We hope these two scenarios can be a starting point.
[ Questions and Answers ]
[ Question from the floor ]
[ Answer from Dr Milne ] …For solar power we assumed the lowest cost profile. There are various studies for LCOE – Levelised Cost of Energy [ Levelised Cost of Electricity ] – they are not showing wider system integration costs – for example, the extra storage needed [ for excess generation that needs to be stored somehow for later use, when the sun has set ]. “Counterfactuals” – is this useful in this case or that case or … ? Model a whole range of scenarios around that.
[ Question from William Orchard ] Results all depend on assumptions in the models. How doees it treat waste fossil heat [ heat from burning fossil fuels for power generation at centralised power plants ] ? The European Union treats renewable heat dumped in the sea as renewable [ ? ] but considers waste heat in […] as non-renewable – the difference is significant. It also depends on your COPs [ coefficient of performance ] in district heating networks. Did you model nuclear reactor CHP [ combined heat and power ] ? What COPs did you use for the heat networks ? How did you treat biomass emissions ?
[ Answer from Dr Milne ] We don’t have to consider what the EU thinks. We do have an option to meet the RED targets [ Renewable Energy Directive ]. Waste heat from large scale power plants plays a huge role in our model – free heat. We build pipelines to link waste heat sources to networks. Question – how to build the heat network ? We need to justify building big pipelines to transport heat. [ Why not transport the heat in the form of gas ? That is, use the waste power plant heat to manufacture gas to distribute to local CHP schemes via a much smaller pipeline than a heat pipeline would need ? ] For Biomass CHP, we considered a range of scales. We gave it a 92% carbon credit. We also have biomass imports in the scenario – a 67% carbon credit. It’s a “pump”. Do we think we can ? We take an off-model view first of all and then apply it to the model.
[ Question from the floor ] This work is well overdue. Thank you for doing it. You say you will change from coal to gas. Why are you not considering more offshore wind – you can expect to bring on nuclear power more slowly ? I’m worried when you put in 60 more years of gas when you put Gas+CCS in. Have you considered fracking [ for shale gas ] ?
[ Answer from Dr Milne ] In the “Clockwork” scenario, it relies on [ strong early development in ] nuclear and CCS mostly – there is a stronger role for renewable energy in “Patchwork”. “Patchwork” is the more moderate speed [ of development of nuclear power and CCS ] as old capacity retires – this is why there is a role and space for other technologies. What the model wants is gas – but it’s not saying where that gas is coming from.
[ Question from the floor ] Have you put any cap on gas ?
[ Answer from Dr Milne ] The only new gas built is CCGT+CCS (Natural Gas-fired Combined Cycle Gas Turbine plus CCS). As you get more [ stringent carbon controls ] will need hydrogen turbines.
[ Question from the floor ] What are the key parameters that break the model ? That you can’t do without ?
[ Answer from Dr Milne ] Biomass+CCS for sure. If you make a lot of assumptions – such as no extra energy demand – then yeah, we’ll be fine. Otherwise, we need Biomass+CCS.
[ Question from the floor ] Where do you get your metrics from ? Isn’t District Heating less efficient than people say ? Isn’t there an anti-competition issue – as District Heating is a single source of supply ? And what about the parasitic loads ? And what happens if there’s not such a big demand for heat [ for example, due to high levels of building insulation ] ?
[ Answer from Dr Milne ] We used central projections from government – we test the cost of energy. Our members used to build some of this stuff. We replace data sets with studies – more independent sources. We have diversified out data set over time. The District Heating networks – it will need a different way of doing markets. It may not be policies that stop you… We assume that 90% of the housing stock remains – we see “difficult households” – not “low-hanging fruits” [ ripe for change ]. We envisage these will need complex packages – if you think it’s going to be received. We need to work this up more.
[ Question from the floor ] Have you calculated the carbon emissions ?
[ Answer from Dr Milne ] Zero or negative. The power sector is 100% de-carbonised by 2030. I can get the figures from our database – gCO2/kWh
[ Question from William Orchard ] MARKAL (previously favourite energy modelling tool) was not fit for purpose for modelling heat networks… MacKay…
[ Answer from Dr Milne ] MARKAL has been shelved, replaced by UK-TIMES…
[ Question from William Orchard ] …fundamentally has the same problem as MARKAL – uses the same algorithms. It wasn’t able to generate appropriate answers to the question of whether it was cost-effective to build heat networks…
[ Answer from Dr Milne ] We use the Biomass Value Chain Model (BVCM). This is new and includes hydrogen and CCS. We include the “tortuosity factor” (kinkiness) of pipeline layout. We model 9 types of buildings. With a hydrogen network – would you want to start small, for example with distributing cannisters… ?
[ Wrap up ]