Any prospect of electricity price recovery in the wholesale market? A detailed look at the fundamentals of bidding



Author: ABB European Advisors

We all know from our economics classes that in a perfectly competitive market, price will equal marginal cost – the cost of producing the next (or incremental) unit. A unit is one MWh of generation in the case of the electricity market. However, over the longer run, firms at the margin need to recover their operating costs, so will attempt to set prices at least at average variable cost, and hopefully recover some or all of their fixed costs in order to achieve profitability.

As the electricity market has to continually match demand and supply in order to “keep the lights on”, price setting tends to be more transparent than in most markets. Actual market designs differ significantly, but in general, electricity market prices are generally set by “bids” into some kind of a market; generally the cheapest viable “bids” are accepted to yield sufficient volume to meet total electricity load demanded by consumers. When generators bid in an attempt to set prices higher than their pure marginal incremental cost, we call this “bidding up”.

A look at the current electricity market in key European countries such as Germany, shows that wholesale prices have been running at or below cost for key marginal plant such as gas plant. For example, clean spark spread prices have been persistently negative for some time, indicating that these plant are making little contribution to fixed costs unless they are able to run only during the on peak hours.

EPR_Germany
Source: ABB, inc.

ABB has analysed recent market fundamentals – and projected them forward – to assess where electricity wholesale prices are relative to short run marginal costs, and to analyse whether incumbent players are able to “bid up” to be able to recover their production costs, either now or in the future. A key component of this analysis is a deconstruction of the supply curve, showing how different bidding and strategies impact prices, generation and revenues.

Take a look at the German supply curve in a decade’s time, and an analysis of residual demand. As is indicated on the below graph, 90% of residual demand is met by generation units with an incremental cost of between €67/MWh and €76/MWh – a range of just under €10/MWh. A plant’s operating costs need to change by less than 15% for it to move from serving 90% to just 10% of residual load. This indicates a highly competitive merit order, where lots of similar plant are jostling for position, and puts a lot of pressure on portfolio optimisation and fuel purchasing strategies in particular as fuel is the chief variable cost component for fossil plant.

EPR_Curve
Source: ABB, inc.

Germany is not alone – a quick look at the British supply curve is even worse as near-identical gas fired CCGT units are all competing together making the supply curve even flatter than in Germany. In Britain, just a 9% change incremental costs would cause an individual unit to slide right through the merit order from one end to the other of the 90% band.

EPR_Curve_GB
Source: ABB, inc.

Is an individual unit able to improve its position in the supply stack, and recover more of its fixed costs? There are two ways of doing this: either by lowering its costs and running more, or by “bidding up” and attempting to increase prices – so running less overall, but operating more in higher priced periods. Cost reduction is challenging – and will probably only be achievable in practice with an unusually favourable gas contract. Given how flat the supply curve is, if an individual unit attempts to “bid up”, it will move a long way through the merit order, and will have very limited (if any) impact on prices with a very significant loss in running hours – a high risk strategy.

If a new unit comes into the supply curve, while it is likely to be at the “bottom” (i.e. cheaper) end, it will likely also only gain very slight competitive advantage – as most of the units already there are also relatively new, efficient units too.
If a portfolio player – or indeed everyone – “bids up” all its units together, then some impact on prices may be achieved – but is this likely to happen, and would it be enough to make generation profitable again for either incumbents or new entrants?
In order to examine this further, as part of its Reference Case forecast for Northwest Europe, ABB developed three scenarios:

  • Scenario A: units bid at their short-run, incremental cost of generation
  • Scenario B: units bid at their short-run, average cost of generation and cycling units also bid-up to recover their start costs.
  • Scenario C: as scenario B, but units also try to recover their fixed operating (but not financing) costs at times of higher residual demand. The ability to recover these additional costs is phased in from 2018 to 2021 as reserve margins tighten across NW Europe.

In reality, no market is perfectly competitive, as in Scenario A, and prices and bidding today in power markets in northwest Europe look most like Scenario B. If the outcomes of Scenarios B and C are compared to Scenario A, pure incremental cost, then plant do start to recover some of their fixed costs, and prices increase by as much as €15/MWh in Scenario C in France (see below graph).

Germany – change in market prices compared to pure incremental cost (Scenario A)
EPR_ScenarioA
Source: ABB, inc.

So there is some future prospect of market price recovery – but this is dependent on all plant “bidding up” in a Scenario C type world. If the market out-turns at Scenario B levels, they will remain at around €5/MWh on average above incremental cost levels, and €5/MWh-€10/MWh above cost at peak times.

Is this enough to keep the market whole going forward? It may provide a reasonable contribution to old, incumbent plant with much of their fixed costs already written off, especially if they are flexible enough to “cherry pick” the highest priced hours. However, to achieve full cost recovery of all fixed and financing costs for a new entrant mid-merit (i.e. running around half the time) CCGT plant requires more than double this amount. A German CCGT with a 45% load factor, entering the market in the early 2020s, would need an average premium of €29/MWh above incremental cost to cover all of its costs from the energy market only – substantially above Scenario C levels.

It may be hoped that as reserve margins tighten, generators will be able to move their bidding behaviour closer to Scenario C, and prices will recover sufficiently to enable both (partial) fixed and variable cost recovery in the energy market. However, there is a risk that the reverse may happen. The EU’s proposed single electricity market (the Electricity Target Model), and/or expanded interconnection capacity across the region, could promote more free flow of electricity across Europe, meaning even more generating units (including cheap renewables) are bidding to operate in each country’s market, potentially leading prices to collapse further towards the marginal cost levels of Scenario A.