By Dave Elliott
Imperial College London and the NERA consultancy have produced studies of energy system integration costs and grid balancing options for the government’s advisory Committee on Climate Change. They focus on flexible generation and backup systems and conclude that ‘flexibility can significantly reduce the integration cost of intermittent renewables, to the point where their whole-system cost makes them a more attractive expansion option than CCS and/or nuclear’.
The Imperial/NERA research team argues that adding system integration costs, for extra transmission/distribution and backup plant/grid balancing services, to the cost of building and operating low-carbon generation capacity, ‘represents a critical input into planning for a cost-effective transition towards a low-carbon electricity system’ and the researchers look in detail at the implications under a variety of low carbon scenarios. A key finding is that the system integration cost of low-carbon generation technologies will significantly depend on the level of system flexibility and that ‘very significant cost savings can be made by increasing flexibility’. Flexible options considered include ‘application of more efficient and more flexible ration technologies, energy storage, demand side response, interconnection’ and also reducing the need for various balancing services through ‘improved system management and forecasting techniques’.
They conclude that ‘enhancing system flexibility reduces system integration cost of renewables by an order of magnitude’. For example, on the basis of their 2030 modeling, ‘the whole-system cost disadvantage of wind generation against nuclear reduces from circa £14/MWh in a low flexibility system to £1.3/MWh in a fully flexible system achieving 100 g/kWh emission intensity’ while the ‘whole system cost of solar PV reduces from being £2.3/MWh higher than nuclear to being £10.7/MWh lower than nuclear as the result of improved flexibility’.
That means more variable renewables can be added to the system while still maintaining system viability, and cost effectiveness, although (impartially!) the team adds that it also makes room for more inflexible nuclear, should that be wanted. They also say that the alternative, for meeting the emission targets, would be to add more CCS, which they see as a flexible option. That’s debatable: the economics of plants with CCS, already uncertain, would surely be worsened if they had to vary their output.
However, running all this through their WeSIM modeling system gives a range of possible mixes of supply/balancing capacity, depending on the 2030 emission targets and the degree of flexibility. Sticking with a low flexibility system, in both low (50g) and high (100g) emission runs, while gas use stays high (respectively 7.7-18.6GW of OCGT (open cycle gas turbine) and 44.5-47GW of conventional gas plant), nuclear has a large contribution (32-37GW respectively). But with high flexibility, then renewables dominate in both low and high emission cases (solar PV 43.3-43.9GW, wind 47.2-51.2GW, respectively), with storage making a small contribution (12.7GW in both cases) but nuclear down to 6 and 18GW, respectively.
These optimals are based on total system costs, not just individual component cost of energy – a holistic approach that the team sees as vital: at present CfDs only focus on the latter with the Capacity Market then trying to back it all up. To that extent they agree with the conclusion of the Energy Research Partnership’s study of grid balancing issues (see my earlier post): the use of standard levelised cost of energy figures (LCOEs) is not enough. But unlike the ERP study, which saw the various grid balancing options as being quite limited, they see flexibility as able to counter system costs with renewables at up to 50GW or possibly more, although at higher levels, system costs rise – doubling between the 100 and 50g targets. For example, at 100g/kWh, the system costs for wind are put at £6.2-7.6/MWh , but rise to £12.5-15.6/kWh at 50g/kWh in a wind-led system.
Even so, that is no reason to avoid expansion, since ‘as the flexibility increases, progressively higher volumes of wind and PV are able to replace nuclear generation, given that their cost disadvantage due to higher integration cost diminishes with enhanced flexibility. This replacement results in a net benefit of between £4.14bn to £4.93bn per annum for 100g scenario and between £5.23bn and £5.99bn per year for 50 g/kWh scenario, representing the net effect of removing nuclear (with higher LCOE) and adding lower cost wind and PV generation’.
Their WeSIM modeling assumes that UK electricity exports are balanced by imports on an annual basis, and that demand-side response (peak load time-shifting) does not extend beyond 24 hours, so it is not based on a very radical shift in practice. There is a lot of flexible gas capacity with reserve capacity and standby plants playing a role for longer term support, while wind and solar curtailment is used to deal with any residual surpluses not fed to the relatively limited storage. The team notes that the system cost savings benefits driven by flexibility include the avoided cost of building additional grid interconnector capacity between Great Britain and continental Europe. So while in the low flexibility scenarios 4.3GW more interconnector capacity is needed in the 100g case and 11.7-12.5GW in the 50g wind and solar-led scenarios, in the high flexibility scenario, no extra grid links are needed at all in either case.
The bottom line condition for all the scenarios was that ‘the system can cope at times of stress (e.g. lots of wind, very low wind over several days, unexpected nuclear outages, low fuel prices, high demand) and achieve the carbon target’. And it seems, given the necessary reserve capacity and flexibility, they all can do that, including the high renewables/low emissions 2030 scenarios. Indeed, the team claims that ‘additional flexibility not only helps to decarbonise the system more cost-effectively, but also allows the system to cope more efficiently at times of stress, for instance by using storage and DSR resources to manage periods of high wind output and low demand or vice versa.’ Also, the approach adopted by their model ‘will ensure that sufficient backup capacity is in place to supply the demand even in the event of prolonged periods of low renewable output’.
The team also looked at a scenario in which the UK power sector achieves emissions intensity of 10g/kWh by 2050. This scenario is characterised by medium nuclear capacity (21 GW), and high wind (90 GW) and PV capacity (100 GW). They let the model add CCS, (some) storage and conventional capacity if needed to ensure security of supply and meet the emissions intensity target at minimum cost, but kept DSR utilisation the same. The integration cost of wind and PV were found to be broadly twice as high as in the wind-dominated 50g/kWh core scenario. And they concluded, ‘even more intermittent generation can be accommodated in the long-term with relatively moderate level of renewable output curtailment (5-6%)’, although ‘this would require significant increases in system flexibility such as significant deployment of new storage capacity, a very advanced utilisation of DSR potential and also a significant expansion of interconnection capacity with continental Europe’.
A bit of Power to Gas conversion of surpluses might also help to limit curtailment and aid balancing, but they don’t go into that. Even so, it’s an impressive bit of modeling work, making it clear how important flexibility is as a way to limit costs and balance grids with high renewable inputs. https://www.theccc.org.uk/publication/system-integration-costs-for-alternative-low-carbon-generation-technologies-policy-implications/
The Committee on Climate Change subsequently produced a report which included a call for more attention to be given to flexibility. It estimated that intermittency for the currently planned levels of renewable supply would impose a grid balancing cost of ~ £10/MWh and said that would rise at higher levels of renewable penetration – to £15 for wind at 50GW and £25 for PV solar at 40GW. But as the Imperial/NERA reports argue, enhanced flexibility of generation and grid balancing services could contain that: https://www.theccc.org.uk/2015/10/22/new-low-carbon-electricity-generation-is-cost-effective-option-for-uk-power-sector-investment-in-2020s-and-beyond/