Introduction
My name is Clare MacGregor. I focus mostly on the renewables
work that we do at Numeritas. Following the release of the draft Energy Bill in
May of this year, I have written some articles around some of the debates that I
think it raises and how it might impact the assumptions in, and modelling of
projects. This one concerns the risks around contract volumes.
One of the key points in the proposed Electricity Market
Reform (EMR) is that intermittent generators, which will be predominantly wind,
will use the day-ahead market for the sale of electricity.
Under the current proposals, the generator will be
responsible for the contract volumes which they demand for electricity
produced. According to Poyry[i],
the considerations in deciding contract volumes will include:
1) “Expected levels of day ahead versus ex-post pricing
2) The predictability of their demand or generation
3) The flexibility of their demand or generation”
For wind generators this could, in the short term, be
problematic.
Current weather forecasting is relatively reliable. However,
the error in forecasting “increases dramatically as the time interval
increases”[ii].
A day-ahead forecast still leaves much room for error, especially where the
risk is being borne by the generator.
Over-Forecasting
Over-forecasting would leave generators having to buy
electricity on the market in order to fulfil their contracts. As wind
generation is often correlated across a region, a point of low wind is likely
to also be one where the spot price of electricity is higher. Purchasing energy
at peak prices to fulfil contracts introduces risk and effective price
uncertainty.
One of the key purposes of the EMR is stated as reduction in
price uncertainty to attract institutional investors to clean and renewable
energy. This re-introduces projects to market pricing exposure.
It could be argued that although projects are exposed to market
prices, it is still only for a portion of the energy they generate, rather than
all, as under the RO. In addition, due to the correlated effect of wind
electricity would be more likely to be sold at low prices under the RO.
However, I would argue that it is an increased risk. Rather
than potentially getting proportionally lower revenue, the project could be
exposed to negative prices. If the spot price of electricity is higher than the
contract value the generator could be paying to fulfil their contract.
It is also an increased risk because of the greater exposure to peak prices than to average prices.
Under-Forecasting
Under-forecasting is also not without its perils. There are
two main risks to the generator, which also affect the systems operator (SO).
Firstly, the generator would be producing energy not under
contract. This would remain the risk of the generator, who would be tasked with
selling the excess energy, or risk losing any associated revenue.
There would also be additional pressure on the SO to balance
the system, where more energy is being sold into the half-an-hour-ahead market
than anticipated.
This could lead to constrained plant. This has been
considered in the bill. It is proposed that the CfD should be paid against that
which is reported to the SO, rather than the constrained amount. This reduces
the risk to the generator for constrained MWh.
Modelling
From a modelling perspective, incorporating this cost could
be difficult.
Over-forecasting would need project developers and
generators to have a better grasp on the spread in price forecasting to use a sensible
peak number to understand the impact on having to buy on the market.
Under-forecasting would need generators to have a strategy
in place for the deployment of un-contracted energy generation and sufficient
resources to carry it out. It would also need a full understanding of the cost
associated with the strategy.
Both of these could cause problems particularly with smaller
generators. They are not as vertically integrated, requiring a potentially more
complex trading supply chain to market. This could be difficult to staff, or
expensive to outsource. Where the pricing has been set for the assumed fixed
costs of the energy, it might not take into account the differing effective
cost of energy between market players.
Both scenarios also require assumptions around the level of
failure to deliver or rate of over-generation. The volumes on a daily basis
could be hard to predict without several years of active site data to use. The
government argues, rightly, that will incentivise generators to invest in
better forecasting technologies. However, if developed in-house, it would be by
larger generators and utilities who can tolerate the R&D cost and the
technologies could become proprietary. There would be little incentive to share
this with competitors. This would again leave smaller generators with a larger
risk burden. The optimum solution is an industry investment in weather
forecasting systems or a third party, such as the Met Office. The solution
could then be sold in the market on a project by project basis, as their
virtual met mast data is currently. Although this would be an additional
project cost, it would at least be accessible to all.
Conclusion
This all has huge implications for players hoping to finance
projects at the start of the new regime. The risks of the RO are well
understood whilst the ‘unknowns’ of the CfD mechanism are much less so. Far
from attracting investors not currently engaged with renewables, it could put
off some investors who are unwilling to take on less well understood risks.
Broadly speaking though, the CfD appears to be a sturdy long
term mechanism. It addresses many of the problems of the RO and the
consultation process has helped address some concerns with the mechanism early.
Hopefully, with the legislated ‘room for manoeuvre’ and the development of
better demand side management and forecasting technologies, these issues
will soon be ironed out, leaving an industry which people will clamour to invest in.
Comments (0)