Can Current Electricity Markets Cope with High Shares of Renewables? A Comparison of Approaches in Germany, the UK and the State of New York.

• Author: Pollitt, Michael G.

1. INTRODUCTION

   One assumption that lies behind this special theme is that the electricity markets we currently have are not well set up for a world where renewables are a significant share of electricity generation. If this is the case they will need to adapt to cope with high shares of renewables as these become a reality within many markets. Through this study, we aim to unpack this idea in the following ways.

   First we look at the empirical and theoretical background. We do this in steps: first we examine what is meant by "high shares" of renewables; next we consider what we mean by electricity "markets"; then we discuss what the term "cope with" implies; before returning to the suitability of "current" electricity markets.

   Second, we explore three examples of jurisdictions with specific aspirations for decarbonisation and the role of renewables. Each exhibit very different approaches to the way they are adjusting their electricity market design to cope with high shares of renewables, namely Germany, the UK and the State of New York in the US.

   Germany is one of the EU leaders on renewables with an electricity share of around 27% in 2014. However, specific surcharges have been applied to household electricity prices to subsidise this expansion (e.g. Erneuerbare-Energien-Gesetz--EEG surcharge). (1) Germany is the country with the second highest residential electricity price in the European Union, after Denmark. The most recent Renewable Energy Law (EEG 2014) is looking at stabilising the rapid growth of renewables by the application of a market based approach by 2017 (market premium), lower degression rates, technology-specific annual growth targets for newly added capacity and the introduction of a surcharge on solar self-consumption.

   On the other hand, the UK is amongst the most reluctant in terms of binding renewable commitments. According to the European Commission (EC, 2015) the UK and France are among the EU nations that may not meet their 2020 renewable energy targets. In the UK, this underachievement is due in part to regulatory issues and administrative barriers (Keep on Track, 2015 (2)). However, the UK is one of biggest supporters of carbon reduction with specific power sector decarbonisation objectives, (3) in agreement with the EU 2030 targets. (4)

   By contrast, New York is among the US states with the most aggressive Renewable Portfolio Standard (RPS) (5) requirement, with a target of nearly 30% by 2015, which reflects its desire to decarbonise the electricity sector.

   The three cases present different visions of electricity market arrangements that might support the integration of high quantities of renewables.

   Our contention is that we seem set to experience a new wave of global electricity "experiments" (Pollitt, 2008), should technological developments and subsidy regimes continue to favour the connection of more renewable capacity to existing electricity grids. This wave of experiments will mirror the attempts at electricity market liberalisation around the world, which took place from 1982 (in Chile) onwards (Pollitt, 2004). As with liberalisation experiments, this new wave of experiments will share broadly similar objectives, there will be some template designs (as with the EU single electricity market (Jamasb and Pollitt, 2005) and the US standard market design (Joskow, 2006)), but lots of local variation and, most likely, a wide variation in outcomes. We discuss our three case studies in ways that bring out the contrasts between them and the choices facing policy makers, regulators and other industry stakeholders seeking to promote and facilitate greater renewable penetration.

2. EMPIRICAL AND THEORETICAL BACKGROUND

   2.1 What is meant by "high" shares of renewables?

   This has significance when compared to the basics of fossil fuel electricity generation (Stoft, 2002). A typical fossil fuel based generation system has the following characteristics. The system operates with the peak capacity margin of 10%. This means that if peak demand is 100 GW, then capacity needs to be 110 GW. If peak demand is 100 GW average demand is typically 2/3 of this. Thus average demand might be 67 GW. This implies that the average load (capacity) factor is around 60% (67/110).

   Now consider the difference renewables makes to such a system. Typical wind capacity factors are 30%; while solar capacity factors are 11% (DECC, 2013). Thus for a 50:50 wind: solar capacity mix, the capacity factor is approximately 20%. 100 GW of such a mix of renewables added to the existing 110 GW of fossil, on average generates 20 GW out of an average demand of 67 GW, i.e. 30% of total electrical energy. However the system will delivery coincident peak renewable output of 50% joint capacity, thus 50 GW. (6)

   Thus, with 30% renewable energy output, we have a situation where 45% (100/210) of the capacity is renewable; but at times renewable output could be 75% (50/67) of system demand. As renewables become more significant in their energy share the impact of their low capacity factors in creating excess generation becomes more extreme. 60% renewable energy requires renewables to be 65% (200/310); but renewable output could peak at 150% (100/67) of system demand.

   At the other extreme is the problem of non-availability of wind and solar power. Solar is not available at night, but in northerly countries peak demand occurs on a winter evening. While it is windier in the winter and the wind does blow in the evening, this is not always the case. There can be shortages of wind output for prolonged periods, with wind availability down at 5% of capacity. (7)

   The peakiness of renewable output and the fact that it cannot be turned up when the weather is unfavourable to generation, means that some significant combination of reserve fossil generation (especially in colder countries where demand peaks at night in the winter), matching demand response (up and down), electrical storage (e.g. batteries) and interconnection (to other territories) will be required as renewable shares increase within individual jurisdictions. Renewable generation can be turned down when demand is too low but often politicians are unwilling to see this happen.

   For our three cases, the current renewable target shares for electricity generation are 80% in Germany (by 2050); 35% in the UK (by 2030); and 29% in the State of New York (by 2015). The latest, 2014, renewable energy shares are 27% for Germany; 19% for the UK; and 22% for New York. See Appendix 1 for additional figures.

   2.2 What do we mean by electricity markets?

   There are lots of markets for electricity related products (Stoft, 2002).The most well-known are energy only wholesale markets for MWh. These can be operated by the system operator or by power exchanges. Usually these involve bids for electricity to be supplied/demand within a given half hour (or one hour) period. The markets around this product include day ahead and futures markets. Then there are ancillary services markets for non-energy products associated with the supply of electricity. (8) These include markets for balancing services markets that involve near real time matching of supply and demand conducted by the system operator. These include black start and ramping services, which are about providing system support following an outage of another plant. Most organised power markets have these markets.

   Organised wholesale energy markets in the US often also have auctions for financial transmission rights (FTRs) every six months which allow congested transmission corridors to be properly priced in presence of nodal (locational marginal) pricing (LMPs) which reflects short term congestion constraints within the transmission system (Hogan, 1998). LMPs are computer-generated price signals (often changing every 5 minutes), which lead to local deviations around the wholesale energy price faced by generators and loads at given locations.

   However in addition there may be capacity markets which involve auctions for the supply of peak capacity (in MW) (Cramton and Stoft, 2005). These can be for one month ahead or up to 4 years ahead. Capacity markets cover the all of the expected capacity requirement, or can exempt certain generation or can simply be for strategic reserves (extra capacity at the peak).

   A key characteristic of all of the above markets is that they are organised at the transmission system operator / national / international level. They are not organised at the local distribution level in general. Thus the price signals within the distribution system do not reflect local network constraints, even when there is locational pricing at the transmission level within the US. Within the EU pricing at the transmission level is not particularly cost reflective, as the EU has favoured a policy of market coupling which joins large zones up to create periods when the wholesale price is the same across the whole organised market or periods when prices reflect constraints on major international interconnectors (not within the zones). (9)

   It is important to position renewables within these markets. Fossil fuel generators are often, rightly, exposed to prices, which are time and location varying and which pay them separately for capacity, energy and ancillary services. This is not usually the case for renewables (Anaya and Pollitt, 2015c). Renewable generators often face fixed feed in tariffs (FiTs) for their energy regardless of the system condition or the network constraints. Thus a major issue is the extent to which renewable generation currently exists outside organised electricity product markets. A key reason for exempting them from a requirement to be exposed to market forces has been the desire to remove barriers to entry to renewable investment. However this is a form of subsidy that may not be appropriate as renewable shares increase.

   It is important to highlight the fact that electricity markets have been built around incentivising flexible...