While a little off topic for our recent posts, one of the projects on our ‘someday’ list is to write a book on the economics of electricity markets. We’ve struggled to get this into our ‘in progress’ file, so what better way to progress it than a public commitment of posting the first real chapter[1]Chapter 1 will be an introduction. on our blog.
The aim of our book isn’t to cover every last detail of electricity markets, but rather the essential economic concepts that underlie the competitive electricity markets around the world today.
Our focus will start with thermal fossil fuel generators, but we will move to renewable generators as we go along. At this stage we’re not sure whether this will be in the same book or a second one.
We haven’t included references here as it was simpler to import text from Word without them. Please wait for the final book before judging whether we’ve plagiarised anything.
For the record our ‘go-to’ text on electricity markets is Steven Stoft’s Power System Economics, which is a great book but is now almost 20 years old, so doesn’t have some of the more recent developments in electricity markets.
We’ll aim to blog the book by posting chapters as we write them. This is a first draft out for comment that has not yet been edited. Please read our disclaimers and don’t rely on this text.
Please comment or contact us if you disagree with anything we’ve said or you feel we haven’t covered a topic sufficiently. Our aim is to produce something useful, so we’d appreciate your feedback on whether we’re achieving this.
Here we go:
2 What do Economists Need to Know about Electricity?
2.1 Background
Markets are everywhere. Almost everything we produce and consume comes through some form of market process. Today, electricity is no different.
However, electricity is not a ‘natural’ market, which is reflected in its historically heavily regulated nature (US) or public provision (Australia). Considerable government intervention and regulation is required to, we hope, gain the benefits of markets over public production.
The reasons[2]Stoft, p15. why regulation is needed are:
- Lack of economic storage. Electricity cannot be stored (e.g. like unsold cars), or production and/or consumption delayed until the other is ready (e.g. a restaurant meal). Electricity consumption ebbs and flows throughout each day and supply must exactly match demand at every moment.
- Lack of real-time metering. Electricity supply to individual customers cannot be measured in real time.
- Related to this, consumers receive no immediate feedback on how much electricity they use or the cost of that electricity. This leads to a lack of any meaningful demand-side response. Consequently, electricity demand is usually taken to be completely inelastic.
- Even if electricity to individual customers could be measured, power flow to individual customers cannot be controlled in real time. Supply to one house in a street cannot be curtailed without curtailing supply to all houses in the street.
Technologies are emerging that challenge some assumptions behind these flaws (e.g. ‘smart’ meters and batteries). We will cover these new technologies in Chapter _, but for now we will deal with the industry as it has existed to this point.
Markets are prone to price spikes they must supply as much electricity as needed to meet (inelastic) demand. There is therefore a tendency for price spikes that require regulation (including price limits and perhaps restrictions on the prices of individual offers).
Another major problem with electricity markets in the absence of centralised regulation is the missing money problem (see Chapter _ below), where generators dispatch (mostly) at short-run marginal cost which is reflected in the market price. Recovery of fixed costs requires price spikes (energy-only market) or a capacity payment (capacity market). More on this in Chapter _ below.
Specifically, electricity markets require:
- A centralised wholesale market including a real-time or spot market and perhaps a day-ahead market, with set of rules under which participants must operate. These markets are auctions where generator’s offers are dispatched according to their price competitiveness;
- A market or centralised purchaser to procure ‘ancillary services’ (see below);
- A set of market rules and a system operator to run the market according to the rules. This is the Australian Energy Market Operator (AEMO) in Australia;
- A rule making process to update the rules as required; and
- A regulator to ensure participants comply with the rules.
Long-term contracts or derivative markets such as futures can form outside of the centralised market but are not essential to the operation of the core market. They may however be essential in practice because the of large capital cost and long return period of electricity generation.
The operation of several common types of contract are discussed in Chapter _.
The markets most frequently cited in this book are Australia’s east-coast National Electricity Market (NEM) and Western Australia’s Wholesale Electricity Market (WEM). However, the book will consider United States markets where required.
2.2 Market Objectives and Economic Efficiency
Most markets specify set of objectives for the market to achieve, which act as a touchstone for their rules to be interpreted should the rules contain some ambiguity.
Market objectives can include factors such as lowering costs to consumers, incorporating low-carbon electricity and so on. However, this is a book for economists and so it will concentrate on the common objective of economic efficiency.
However, in more important than economic efficiency is reliability. Economic efficiency of production sits on top of a base of reliability.
The job of the system operator is to reliably supply electricity in the most efficient manner possible. But there can be no efficiency without reliability. This includes maintaining the network’s frequency at all times.
‘Reliability’ of an electricity system is often confused with ‘security’ of that system. The Australian Energy Market Operator defines the difference as:
‘security’ means that the power system operates within defined technical limits, and ‘reliable’ means that the supply of electricity is sufficient to meet the demand for electricity.
‘Adequacy’ of available capacity is the goal of long-term system planning. A system is reliable if it has adequate capacity, including an allowance for unexpectedly high demand or generator outages, and sufficiently flexible generation to meet the variability of demand.
The reliability requirements for the two major Australian markets are:
- NEM[3] (at least 99.998 per cent of forecast customer demand to be met each year);
- WEM[3] (a 7.6 per cent reserve margin over forecast peak demand should be calculated to a probability level that the forecast would not be expected to be exceeded in more than one year out of ten, with limit expected energy shortfalls to 0.002% of annual energy consumption).
In an efficient market, the goods and services that are produced are the ones that are most valued by society, produced at least cost, and allocated to those who value them most highly, thereby maximising community well-being.
There are several dimensions to economic efficiency. These include:
- Allocating resources to their most productive use (“allocative efficiency”), which can be achieved by setting the prices of goods and services to reflect the cost of providing an additional unit of the good or service.
- Providing goods and services at least cost (“productive efficiency” or “technical efficiency”), which can be achieved, for example, through using the most efficient, least-cost production technologies or management methods that reduce costs, without compromising service standards.
- Ensuring that investments are optimal over the long-term, in their timing and location (“dynamic efficiency”; that is, considering change over time). An example of this is timing capital investments so that costs are minimised over the long-term, and that they reflect any changes in consumer preferences and available technology over time.
Chart 1: Stylised Firm Economic Costs
[As an aside, you can see how to do this chart here.]
Economic efficiency, particularly allocative efficiency, is often presented in stylised fashion. The ubiquitous textbook example of producing at a profit-maximising point where Marginal Cost (MC) equals Marginal Revenue (MR) under an upward sloping supply curve (MC above the shutdown point) is shown in Chart 1 above. The firm’s MR is equal to Price (P) in a perfectly competitive market.
Under such a cost structure the firm can recover its costs (variable and fixed) at an allocatively efficient single price if that price is high enough. In Chart 1 the price level P* allows for normal profits to the firm, which it obtains by maximising its profits at quantity Q*.
However, reality is more complex than this. Specifically, regarding short-run (allocative) efficiency:
- Thermal electricity generators have high fixed costs and low marginal costs, so marginal cost may not be able to rise high enough to recover fixed costs; and
- Generators MC curves do not rise as dramatically as in the standard diagram, or may even fall (Chapter _), meaning that MC may never rise above Average Variable Cost (AVC). That is offering electricity at MC might lead to a sort-run loss, or the supply curve above the shutdown point is never reached.
Designers of electricity markets have offered ingenious solutions to these problems, which we will examine below.
Dynamic efficiency is achieved because the long-run cost of thermal generators also aligns with the short-run dispatch priority of these generators. This is examined in Chapter _.
There is, however, no proof that any design is superior to another design, or any proof that any design achieves short or long-run efficiency. A ‘leap of faith’ is required in all circumstances, which is examined in Chapter _.
2.3 Wholesale Electricity Market Basics
To start our tour of electricity markets, let us examine the basics of electricity relevant to the operation of a market.
Firstly, we need to set the context of wholesale electricity markets. These markets bridge the gap between generators and final customers. Generators sell into a wholesale market which retailers purchase . The task of retailers is to purchase electricity, managing the associated price volatility, and selling to retail customers at more stable prices. We refer to retailers as ‘customers’ in wholesale markets.
Microeconomics’ exposition of firms and markets starts with a price and a quantity of a product, in this case electricity. Electricity is fungible, like the products in our microeconomics textbooks, at least most of the time . However, it does come with a few additional complications.
In wholesale electricity markets generators make offers into the market with an associated price and quantity. Electricity generators compete on price to be dispatched against (usually) completely inelastic demand for each unit of time.
Let us start with the Q in our model. Electricity is more complex than standard commodities because it is provided in terms of energy, which the physicists define as power over time. Power in most wholesale markets is measured in Megawatts (MW), or some multiple of Watts.
The capacity of electricity generators, or the maximum stable amount of power they can produce, is measured in MW. If a generator produces 100 MW for one hour, then it produces 100 Megawatt hours (MWh) of energy or electricity.
This is important in electricity because, unlike other industries, demand must equal supply at very moment. If a car manufacturer falls behind in production it can, within limits, run extra shifts to catch up. If it knows a period of high demand is approaching, it can build up stocks of cars to run them down later.
However, in electricity, if demand is 105 MW for a few minutes, but only 100 MW is available, supply must be rationed amongst consumers (load must be shed). That is, lights will go out, factories will shut down and/or life support devices will be threatened (i.e. blackouts).
Different types of generators have different capabilities. A 100 MW thermal generator without fuel constraints can produce 100 MW indefinitely, but a 100MW battery with 200 MWh of storage can produce 100 MW for only 2 hours or 50 MW for four hours. A 100 MW solar system can produce 100 MW when the sun is shining, but less than this (even zero) when the sun’s strength is reduced.
Generator outages occur when there is a schedules or unscheduled interruption to electricity assets (generation units, electricity transmission or distribution assets). Generator or network assets are said to trip, causing an unplanned outage, if suddenly lose connection to the network. This may occur perhaps due to a failure of equipment in a generator.
Electricity markets are based around a spot (real-time) market and perhaps a day-ahead market. A day-ahead market is required if markets separate commitment, or whether a generator will be required, and dispatch, or the exact electricity generated.
Centralised electricity markets (see Chapter _ below) use the day ahead market to specify whether a generator needs to be started and running in 24 hours’ time. These generators are compensated separately for start-up costs, running costs and generation costs, should the market price not cover these costs.
In contrast, decentralised markets require generators to assess whether their single price offer will see them dispatched and must plan accordingly. Consequently, these markets do not need a day-ahead market to operate, but some (such as the WEM) contain a day-ahead market as a means for managing risk.
Electricity generators are dispatched according to their price competitiveness for periods of time called trading intervals. These are 5 minutes for the NEM and 30 minutes for the WEM.
Generators are paid according to the market price over pricing intervals. In the WEM pricing intervals equal 30 minutes, the same as trading intervals in this market. However, in the NEM, pricing intervals are also 30 minutes, or the average price over six trading intervals.
Gate closure is the time before the trading interval that a generator that a generator may submit no new offers. This is two hours for most participants in the WEM , but generators in the NEM may submit re-bids may be submitted up until the start of the relevant five-minute dispatch interval by re-allocating offered volumes within the nominated price bands.
2.4 Essential/Ancillary Services
The other major complication in electricity markets is that there is a need for another class of markets within the electricity market. These are known as essential or ancillary services markets.
This section covers the most important of these services.
We cannot, at least not yet, run a competitive price-based market in real time. There will be a lag between gate closure and the precise moment in which supply and demand must balance.
Additionally, disturbances in generation (contingencies) such as large generators suffering sudden outages or transmission constraints means that electricity markets must have sufficient capacity available to quickly ramp-up production from other sources.
Variation between submitted offers and actual demand, both due to forecast error and that it is impractical to run an auction for every second of dispatch, is covered by load following, which can be centrally planned or procured in its own market. Load following requires a generator to run at a level above its minimum generation but below its full capacity, so that the system operator can turn it up or down in real time.
The largest Allowance for contingencies that might befall the market is spinning reserve. This is unused but instantly available capacity in case, for example, a major generator suffers an outage. Providing spinning reserve involves a generator running above minimum generation but below maximum generation.
Different markets procure ancillary services differently. The WEM currently procures of mandates these services directly , while the NEM tries to minimise the total cost to the entire system of procuring energy and ancillary services. This is known as co-optimisation.
The market’s system operator can override the market should stability of the system be threatened. Generators may be constrained on or off no matter how competitive their offers are.
Finally, the market operator will procure black start or system restart services. This is needed in case the system collapses and needs to be restarted from scratch.
For the market to incentivize generators to offer spinning reserve, the generators must be able to profit as much from offering load following as they could in the balancing market.
These factors complicate electricity markets enormously, which is beyond the scope of this book. From this point the book will concentrate on spot market to assess how it might achieve economic efficiency.