By Albert Cheung
Head of Analysis
Bloomberg New Energy Finance
As I write this, staff at the U.S. Department of Energy are busy working on a study which, in just 60 days, must provide answers to the question, “[Are] federal policy interventions and the changing nature of the electricity fuel mix… challenging the original policy assumptions that shaped the creation of [the wholesale electricity] markets?”
Or put more simply, is the rise of subsidized clean energy causing power markets to misfire?
The study, commissioned by Energy Secretary Rick Perry, has been widely dismissed as a politically motivated attempt to bring back a bygone era of ‘base-load’ coal generation – a key Trump campaign pledge. Indeed Perry himself, speaking at the Bloomberg New Energy Finance Global Summit in New York on April 25, called for an end to the so-called “war on coal” instigated by his predecessors.
However, whatever the motivation of Perry in ordering the DoE study, there is, in the impact of zero-carbon energy on power markets, a question that is troubling many policy-makers, utilities and large consumers both in the U.S. and internationally.
America’s Federal Energy Regulatory Commission recently completed a two-day conference examining how state-level policies that support certain generation technologies are distorting the competitive energy and capacity markets. The most recent and controversial examples are the decisions by New York and Illinois to provide out-of-market payments to support their struggling nuclear plants, which have been undercut by gas and renewables. FERC and the regional market operators are growing anxious: subsidies distort price formation in the power market, and this might beget more subsidies for favored technologies, and ultimately lead to “unintentional re-regulation”, in the words of acting FERC Chair Cheryl LaFleur. (See our Analyst Reaction on the FERC conference: BNEF clients can access this on the web or on the Bloomberg Terminal).
European power markets have been down this road already. In Germany for example, the rapid uptake of wind and solar has famously contributed to drastic reductions in power prices, more frequent negative pricing events, the introduction of out-of-market payment programs to keep fossil plants in reserve, and ultimately the restructuring of energy champions RWE and E.ON. It is also leading to the dissolution of the single pricing zone for Germany and Austria. (Our 2017 Germany Power Market Outlook discusses the main trends: BNEF website | Bloomberg Terminal.)
The problem
All of this is enough to make market purists wince: if we believe that wholesale power markets deliver the most efficient outcomes, they say, then why do we keep adding layers of additional incentives and mechanisms?
Plants that receive feed-in tariffs, tax credits, contracts for difference and any other non-market revenue streams have an advantage in a competitive wholesale market. They may for example continue to generate – and profit – even when prices are extremely low or negative (a clear market signal that the power they produce is not needed). They are also able to compete at lower prices in capacity auctions.
We are seeing such effects now in California, where utility-scale solar is approaching 10% of power generation. As my colleague Will Nelson noted in his Summit session (presentation: BNEF website | Bloomberg Terminal), PV plants used to realize wholesale power prices well above those enjoyed by the average gas plant, because they captured the daytime demand peak. In 2011, solar PV realized 25% more value than the average around-the-clock power price. But as PV deployment accelerated, daytime power prices fell, and today the abundance of solar means that PV plants are capturing only half the average around-the-clock power price (so far this year). On the basis of wholesale prices, it would be folly to build a new PV plant in California.
This has also transformed the economics of gas plants in California, which increasingly have to operate at a loss during the daytime and then ramp up to capture the lucrative evening hours, when the sun goes down and power demand rises as people head home. This situation – gas operating at a loss – raises critical questions about the long-term security of supply in the Golden State, or in any high-renewables market.
Supporters of subsidies for renewables retort with two arguments. First, it is hard to find an energy source that has not at some point been the beneficiary of government or regulatory support. What is really at stake is whether or not states and countries are right to pick winners at all. Is it wise for Massachusetts to run procurement programs for offshore wind? Should New York and Illinois have a nuclear policy at all? Should Germany and the U.K. continue to run auctions for renewable energy? None of these are compatible with a strict definition of the primacy of power markets, which would dictate that technologies should live or die in the fire of pure competition. Even capacity payments are subsidies, in the eyes of many – they just happen to subsidize for reliability, rather than decarbonization.
Second, the focus on subsidies misses a bigger issue. Subsidized or not, variable renewable energy – wind and solar – is unlike anything power markets have seen before. Their generation is cost-free and abundant for some hours, and completely unavailable in other hours – not to mention behaving differently at different times of the year. And they are quickly becoming the cheapest forms of power generation in numerous markets around the world.
The crux is that competition in most power markets is determined by short-run marginal cost: how cheaply you can run your power plant in the next hour, compared to your competitors. Over two decades of energy liberalization, this market construct has arguably been able to deliver a competitive and inexpensive power supply (though the details of the story differ between countries). But how relevant is such a construct in a new world where a large portion of plants will have zero variable cost, and cannot be turned on nor turned off?
The flexibility imperative
As variable renewables gain ground, the grid will require more sources of flexibility. This term captures a vast array of possible resources: flexible generation, demand response, interconnection and storage; both large-scale and distributed; both dirty and clean. It also captures three very different time horizons: long-term seasonal back-up, short-term balancing reserves and frequency regulation. A back-up generator, a rooftop PV-plus-storage system, a smart EV charger and a gas turbine – all are flexible resources that can be enlisted in different ways.
In order to unlock the full diversity of flexible resources, policymakers and regulators will need to make sure they can all participate in the market, and be appropriately paid according to their value. This is a complex challenge, not only because these resources all have different characteristics, but also because they will be located in all sorts of places – different locations on the transmission network, connected at the distribution level, or embedded ‘behind the meter’. Markets will need to be able to value energy and flexibility in each location of the network (for example by judicious use of zonal and nodal pricing), and also unlock the value of distributed resources.
Realizing a cheap, clean and reliable energy supply
At the heart of this complex picture is the need to deliver the ‘trilemma’ – energy that is cheap, clean and reliable for all. There are numerous trade-offs between these three (the cleanest energy system may not be the cheapest; the cheapest may not be the most reliable), but the overall policy approach is fairly clear:
Clean: in order to drive deep, rapid decarbonization, continued intervention will be needed, either in the form of carbon pricing, or support for clean sources, such as via contracts-for-difference, power purchase agreements awarded at auctions, or tax credits.
Reliable: there will always need to be a regulatory obligation on system operators and utilities to ensure reliable service. Capacity mechanisms are one approach that these bodies can employ, but not the only one.
Cheap: market competition and regulation must be employed to deliver the other two objectives at the least cost to customers.
In order to achieve these three objectives, there are also five problems that markets need to solve. Without solving these, it is unlikely that the trilemma will be realized:
- Problem 1: Maintaining credible incentive signals for long-term investments in energy and flexibility
- Problem 2: Optimizing short-term dispatch and balancing of supply, demand and storage
- Problem 3: Driving capacity build and energy provision in the geographical locations where it is most needed
- Problem 4: Unlocking the value of distributed resources, by exposing them to the markets and price incentives, and allowing distribution utilities to make optimal use of them
- Problem 5: Ensuring that markets are technology-agnostic, encourage innovation and are future-proofed for emerging technologies
There will be tremendous diversity in how different countries, states and power markets achieve these goals – and how effectively.
Taking back control
Faced with such complexity, many markets have turned to capacity mechanisms to ensure reliability and availability of flexible resources – often layering a capacity market onto an array of other policies.
In the U.K., the Great Britain region, fearful of a possible generation shortfall, has emerged from a multi-year Electricity Market Reform process and now operates a capacity market with annual auctions for delivery in four years, as well as a carbon price floor, a scaled-back auction system for renewables build, a long-term contract for new nuclear build and a plan to phase out coal by 2025. France has also added a capacity market, and executed its first auction for this in December 2016 – alongside a program of renewable tenders.
The Canadian province of Alberta has resolved to introduce a capacity market, with first procurement starting in 2019. (See my colleague Alex Morgan’s Summit presentation on Alberta: BNEF website | Bloomberg Terminal.) This mechanism will work in concert with an economy-wide carbon price, a commitment to phase out coal by 2030, and a renewable energy tender program that will procure 5GW by the same year – a similar cocktail of policies to those in the U.K.
In these markets, power prices will have a limited role in determining the energy mix. At a recent BNEF client roundtable in London, one energy company executive remarked, “The power sector is unique among commodities in that the market does not drive investment, only dispatch.” Indeed, capital investments will be guided less by markets and more by central procurement in the form of renewable energy tenders and capacity auctions.
Proponents of this approach argue that the power markets are not capable of providing the long-term investment signals required to deliver on Clean and Reliable – especially as variable renewables grow. Renewable energy auctions (or other forms of support) are the fastest way to add carbon-free energy to the power system (Clean), and capacity markets are a proven method for ensuring that the system is Reliable. The Cheap part comes from competitive tendering.
There are flaws in this argument. As BNEF Advisory Board Chairman Michael Liebreich pointed out in his remarks at our Global Summit, capacity auctions administered by a central system operator are likely to over-procure due to risk aversion, and the high cost of getting it wrong and causing a black-out. Because the capacity payments flow through to utility bills, it is the customer that pays over the odds if over-build occurs. Power plant investors can sleep soundly knowing their investments are protected. In other words, capacity markets may solve ‘problem 1’ above, but possibly at the expense of Cheap.
Capacity markets are also difficult to design well. Both the Great Britain and PJM markets have inadvertently encouraged the use of polluting diesel generators, and demand response and energy storage providers have to fight running battles to make sure participation rules give them equal footing against generators (problem 5).
Similar criticism can be directed at renewable energy auctions and other direct support mechanisms. Of course, auctions and tariffs play an invaluable role in accelerating renewables from zero to double-digit percentages in the energy mix – and long may they play this role. But at higher penetrations, they start to look like blunt instruments.
Market design for higher shares of green power
There are alternatives. The starting point is this: opponents of central planning argue that wholesale markets can indeed provide the right long-term investment signals, without the need for centralized auctions, as long as they are well designed and have credibility.
To achieve Clean requires a carbon price or emissions target. The ideal is to place a carbon standard on the energy supplier (or utility, in American parlance), incentivizing them to find the cheapest ways to reduce emissions. Even when renewable energy is demonstrably the cheapest energy source, there will come a time when simply adding more renewable energy is not the cheapest or smartest way to decarbonize. A simple example is when there is already enough solar to meet daytime demand – a renewable energy auction might lead to more solar being procured, when what is needed is to reduce emissions from night-time fossil generation, or to increase energy storage capacity. A carbon standard on energy suppliers would provide a clear signal to the market that it must innovate to find new ways to decarbonize – solving problem 5 above. They might respond by promoting energy efficiency, reducing reliance on coal, investing in distributed generation, storage, or promoting electric vehicles to create flexible demand for renewable kilowatt-hours. Importing clean power from adjacent markets could also be an answer, as long as certificates of origin are available.
To achieve Reliable, wholesale price caps would have to be removed and prices allowed to rise to thousands of dollars per MWh at times of scarcity, to reward owners of rarely-used back-up capacity (as currently happens in Texas and Australia). Unlike in a capacity market, the risk is placed firmly on the power plant investor, not the customer – but the probability of high prices should induce investment, solving problem 1. Of course, some of these investments will go sour in seasons when scarcity does not occur, but if this happens then it is a sign that the back-up was not needed after all.
For short-term flexibility, trading must be allowed to occur as near to real-time as possible, so that generators, storage providers and energy suppliers can respond to variable conditions even as weather forecasts are updated during the day. The system operator will of course procure balancing resources and reserves to ensure reliability, but market participants should be made to pay the costs of the imbalances that they cause – for example when a suppler underestimates its customers’ demand, or when a wind generator produces less than expected. This approach should address problem 2.
This leaves problems 3 and 4: making sure that investments are optimized at the transmission and distribution levels. Approaches that include zonal or nodal pricing are in the right direction – it is important to acknowledge that the transmission grid is a real physical constraint, and price resources accordingly. A more radical approach would be to charge energy suppliers explicitly for the specific transmission links that they use when procuring energy from one location and supplying it in another.
At the distribution level, network regulations need to be updated to make sure that distributed resources can realize their full value – the value of grid upgrades deferred, ancillary services provided and energy supplied. This is the goal of recent regulatory efforts such as the New York REV, the U.K.’s RIIO reform and California’s Integrated Distributed Energy Resources effort, all covered in my colleague James Sprinz’s Research Note on this topic (BNEF website | Bloomberg Terminal). To optimize the deployment and usage of distributed resources, retail rates will also need attention. With smart meters rapidly gaining penetration, there is an opportunity to align pricing incentives with the needs of the overall energy system.
Let us know your thoughts
These issues lie at the core of how power markets will function as the energy system travels down a path of deeper decarbonization. Today, we have published an opinion piece written by BNEF founder Michael Liebreich, containing his views on how the market design issue could be approached.
Generally at Bloomberg New Energy Finance, our thinking on these issues is evolving quickly, and our team will be focusing efforts on understanding how these questions will play out in different markets. We look forward to hearing your thoughts.