Some renewables are variable, so how do we account for the cost of balancing them? There are a variety of approaches, some looking at the cost to consumers, some at the costs to the energy system overall. The costs may not be huge with well-designed smart grid systems, but someone has to pay them. Dave Elliott explains
A paper from Imperial College proposes a new conceptual framework for understanding competition in electricity markets that includes variable input but marginal cost renewables. Noting that one of the primary drivers for consumers to switch from grid-supplied electricity to self-generated electricity (e.g. home rooftop PV with batteries) is cost savings, the researchers constructed a model that forecasts when going-it-alone “grid defection” by consumers may become widespread. In reality, few domestic consumers in the UK are likely to want to go entirely off grid. At least not for some time. Grid links are needed and useful for backup, e.g. for when there has not been enough sunshine for a while and consumers’ batteries are discharged, and also to sell any surplus power they can generate, beyond what they can store. However, the grid defection analysis is still a useful conceptual exercise, not least since it gives us some idea of the cost of balancing/backing up variable renewables. And, in time, some users may want to try the off grid option.
Based on detailed modelling of technology cost curves, the Imperial researchers estimate the year in which three types of consumers switch to on-site power generation that offers similar reliability and lower cost to grid-supplied electricity. But in so doing, instead of just using the normal concept of “grid parity”, the point of economic indifference between the cost of on-site renewable energy (e.g. roof top solar plus backup batteries) and the cost of conventional supply, they use a new concept, firm power parity. Building on the notion of cost equivalence, firm power parity is reached when “on-site renewables deliver the same service at the same cost as conventional electricity supplies”. Firm power is available when “the wind doesn’t blow, or the sun isn’t shining”.
Their results suggest that, “While it will become increasingly profitable for consumers to generate and store their own electricity, profitably disconnecting from the grid is more than a decade away in most markets. For consumers who already enjoy reliable transmission and distribution infrastructure, the cost of replicating grid reliability (even on a single-day basis) will remain significant.” But by 2030 the researchers say that may change so that these technologies will become disruptive to conventional sources, and it’s already the case in some off-grid areas, for example in developing countries, with local mini grids being an option.
There are some limits to the approach: the researchers note that “Our analysis does not yet account for the value obtained from a variable or time of use (TOU) tariff, which would likely act to accelerate the date of firm power parity. Furthermore, it does not incorporate the social costs of greenhouse gas emissions from grid-supplied electricity. To keep things simple, we fixed the price of electricity in each market, keeping it constant in 2017 real money terms throughout the forecast period.”
There are some parallels with the “equivalent firm power” concept proposed by Dieter Helm (see my earlier post), although that is part of a wider set of policy suggestions and it concerns the system level costs, as faced by supply companies and grid managers. For another, much more general, system choice metric, with some social and eco-costs also added, see “Co-production in distributed generation: renewable energy and creating space for fitting infrastructure within landscapes“.
There various new metrics are quite complex and to some extent an engineering approach to identifying the extra system cost is perhaps easier, e.g. the cost of grid balancing can be estimated directly and seems likely to be in the range of 10-15% extra on generation costs. So then you can judge if its worth it to make the change, e.g. at the system level. But that doesn’t give you an insight into the social and eco-costs and benefits. Neither does the approach that Trump tried to get adopted in the US – it seemed mainly concerned with protecting the conventional energy market and supply system, and coal and nuclear plants particularly.
To that end, energy secretary Rick Perry proposed that the Federal Energy Regulatory Commission (FERC) develop and implement rules that accurately price generation resources necessary to maintain the reliability and resiliency of the US bulk power system. As proposed, the final market rules would allow for the recovery of costs for what the Department of Energy calls “fuel-secure” resources that provide “reliable capacity, resilient generation, frequency and voltage support and on-site fuel inventory”. Eligible units would be required to have a 90-day fuel supply on site in the event of supply disruption. The result would clearly benefit fossil and nuclear.
Energy security and grid balancing are obviously important, but this approach, with renewables seen as introducing extra risks and costs, might be seen as a bit negative, ignoring the climate-change issues and the role that renewables can play in avoiding them:
Fortunately, is was ruled out of order, at least for now. However, what seems a similar view had also emerged in Australia, where wind and solar farms may be forced to meet tougher standards to guarantee reliable energy. The idea emerged in a review of system reliability by chief scientist Alan Finkel, which, among other things, recommended that individual wind and solar farms be responsible for providing “dispatchable” generation via a “generator reliability obligation”, or contracting with other suppliers to meet this requirement.
The idea is being fought by green energy backers, with, according to Renew Economy, the industry “struggling to understand why each new plant would need to add battery storage or strike a deal for ‘firm capacity’ with a neighbouring gas plant”. They argue that “reliability isn’t a function of each individual power station but all of the system”. That’s clear: no one expects each gas, coal, nuclear plant to have its own backup. Grid balancing is best provided at the system level.
Nevertheless, it is reasonable that each generator should pay its share of the cost of this. Fossil plants can also have variable outputs, due to unplanned outages, as can nuclear plants, but the variations with solar and wind are larger. The proposed new Australian system would certainly make this visible, as would that proposed in the USA. In the UK, some of these extra cost are already covered by “use of system” charges and grid development/management costs charged by grid companies, who, typically, are responsible for balancing the system and keeping the lights on. But there are pressures to make renewable generators pay more and similar issues are emerging in the EU in relation to the priority dispatch provisions that renewables enjoy – their output is given priority (see Clean Energy News and Energy Transition.) Clearly, if we want low carbon green power, the priory dispatch approach promotes it, but rivals like nuclear don’t like it, and renewables do require balancing measures, which have to be paid for.
Political battles like this aside, in terms of methodology and easy assessment, by contrast to the defensive approaches being proposed in the USA and Australia, and even now the EU, stressing the costs and risks of renewables, and supporting their rivals, the firm power parity approach is more forward looking, focusing on the process of change, taking balancing costs into account, from the consumers point of view. And certainly, despite its limitations as currently configured, it can provide a rough view of when changes may occur: maybe quite soon. Charles Donovan, director of Imperial’s Centre for Climate Finance, says “The results of our research are exciting as they show we will soon be entering a period where reliable and profitable solar power production by residential energy consumers becomes a reality in relatively cloudy places like London.”
Meanwhile, in terms of the system level requirements, the capacity market provides balancing capacity, with the UK’s latest auctions (T1 for next year, T4 for four years ahead) contracting mostly gas-fired capacity, to be available to meet shortfalls. Coal plants and diesel plants mostly got shunned, but storage and demand management are beginning to make an impact, though it’s still small.
Oddly, however, nuclear plants were included in T4, despite not being able to load follow. They do offer some synchronous frequency support – an issue I explore in my next post.