The return of the academic
The New Zealand electricity market is similar to Tolkien's hobbits: it may be small in size, but it shows others the way. This post describes its most remarkable features. Describing what is missing is often as important as characterizing what is there when it comes to understanding the organization and performance of markets. We will therefore start out by presenting the basic structures in place, then the elements which New-Zealanders have thought best to leave out, before drawing some lessons.
1. What is there
The wholesale market is organized in accordance with best practices suggested by academic researchers specializing in energy (and plain common sense).
A strong, independent regulatory authority
The entire edifice rests on good governance. The Electricity Authority, (hereafter referred to as "the Authority") is responsible for the organization of electricity markets. It is entirely independent from the government, i.e. the Ministry of Energy. Its mandate is clear: to contribute to improving efficiency in the electricity sector for the long-term benefit of consumers. The budget for the Authority is approved by parliament every year, and financed by a tax on electricity sales.
A consensus seems to have emerged in New Zealand: the electricity industry is a technical undertaking which is best left to professionals. The role of politicians is limited to ensuring the governance of the sector is adapted to the objectives. Unlike many others, elected officials in New-Zealand do not appear to feel the need for regular and intempestive interventions in the organization and performance of the electricity market.
The Authority sets the rules of the game but, contrary to regulators in most other countries, it is also responsible for managing wholesale markets. However, to avoid potential conflicts of interest, it is required by its statute to delegate this management to other entities.
The system operator is distinct from the market operator. Transpower, the owner and operator of the transport network, is responsible for operating the market. They are responsible for security, reliability and dispatch. The New Zealand stock exchange is the operator for the wholesale market, making a market platform available to participants, calculating and publishing prices using dispatch information provided by the System Operator. The market operator is also responsible for clearing, billing and making up the gap between predicted and actual consumption. EMS, a subsidiary of Transpower, auctions off and allocates the financial transport rights (presented below). The market operator is responsible for clearing and billing for these rights.
This separation between system and market operators can also be found in other countries (in Europe, for example). The unique feature of the New Zealand model comes from bringing together these two activities under one roof - that of the regulator, which sets the rules of the market. The advantage is that, in the event of problems, responsibility can be clearly attributed to one institution.
The market structure
The market is structured around four vertically integrated producers / providers which emerged after the break-up of the former public monopoly, and now meet around 80% of demand. Three of these four companies are 51% owned by New Zealand. Eighteen other retailers, in a variety of sizes and configurations (some of them vertically integrated, others simply retailers) serve the remaining 20% of demand.
Transport activities, which were traditionally integrated within the state monopoly, have been vertically separated and assigned to a separate public entity, Transpower. The distribution networks are run by local companies, some of which are private while others are owned by local authorities.
The wholesale electricity market is a "real" spot market, unlike the European and North American markets, which are "day ahead" markets. To understand this feature of the electricity industry, it is important to remember that thermal production, in particular coal power plants, is relatively inflexible. These types of plant cannot operate at 100% of capacity one hour and fall to 0% the next. For this reason, the electricity industry has traditionally broken down the problem of the optimal use of the means of production available (optimal dispatch) into two stages: i) determining today which production units will be active tomorrow (unit commitment), then ii) optimal use at every hour of active production units. The underlying mathematical problem is fairly complex, and has led to some important scientific advances in the field of operational research.
When the electricity markets were set up in the 1990s, most countries adopted an architecture structured around these two stages. In other words, for a given period, megawatt hours are exchanged and valued on two spot markets: the day ahead market, where the production plan is determined for every hour of the following day (thus solving the unit commitment problem), followed by the adjustment market(s) where production plans from the previous day are adjusted to take into account more recent information relating to demand and available capacity. The market architects thus internalized the technical constraints for the production units, and produced a replica of the solution previously developed by engineers for integrated companies. But this dual structure ("two settlements market") is complex. What's more, it opens up the possibility of strategic arbitrage between the two spot markets.
New Zealand has the advantage of most of its electricity production being hydraulic. The "unit commitment" problem therefore only arises for a relatively small number of power plants. The architects of the New Zealand market chose a simple solution: a unique spot market.
The architects of the New Zealand market adopted nodal prices as the price setting mechanism for transport network use. This approach, which was developed in the US in the 1980s and 1990s, is based on a self-evident fact: the value of usage and the cost of electricity do not only vary over time, but also throughout space. At each point in the network (known as a "node"), the nodal price balances supply and demand; it also includes the impact of an injection or retraction at this node on congestion and losses over the lines. The transport cost between two nodes is simply the difference between nodal prices. This technical and economic approach provides the most efficient solution for setting prices for congestion and losses throughout the electricity transport networks. In New Zealand, every half hour, between 240 and 260 nodal prices are calculated and used. At night time (low demand, no congestion) they are practically all the same. During the day time (strong demand, variable congestion depending on lines), they differ.
Around 70% of residential consumption sites are equipped with connected meters which measure consumption at 30 minute intervals. What is unusual here is not so much the figure itself (Italy and some American states have similar levels of penetration), but the method. The providers, not the distributors, installed these meters voluntarily, because they believed the advantages which could be derived from them would outweigh their cost. Several companies competed for the contracts to install these meters.
Competition on the retail market
Entering the retail market is made easier by several factors. First, payments from providers to network companies and the wholesale market are synchronized with payments from customers to their providers, reducing the need for working capital.
Second, changing provider is a very simple and fast process, which takes 24 hours on average. This is possible for two reasons: a single database centralizes all the information for each point of delivery. Changing providers simply means changing an entry in this database. Second, with consumption measured every half hour for 70% of consumers, final billing issues are greatly reduced.
The part consumers play in the supply-demand balance
Like for all markets, efficiency requires users adapting their consumption to prices. They must therefore adapt to the wholesale price of electricity, which means transferring their peak consumption to off-peak times. Industrial customers in New Zealand are able to submit a demand curve in the dispatch algorithm, i.e. indicating that their consumption will be lower if prices are higher. Similarly, two providers now provide an offer for residential consumers where customers are billed based on wholesale prices.
2. What isn't there
A capacity mechanism
There is no capacity mechanism in New Zealand, and no plan to set up such a mechanism.
This is not an oversight, but a deliberate choice. In an electricity system based around thermal and nuclear power, the main risk is the combination of a dip in production and a very high peak in demand. Rationing is implemented in the shape of selective power cuts: power to some consumers is cut for a few hours, to "pass the peak". In a hydraulic system, the main risk is a dry year, i.e. a lack of water every five or ten years, which leads to a lack of energy. In this case, rationing takes the shape of a reduction in consumption for a week (or more).
The problems related to investing in peak production are slightly different: in a thermal system, peak production sources (which are also thermal), are operational a few hours per year on average (a little more during a harsh winter, less in a mild one). In a hydraulic system, peak production sources are active in dry years, and less or not at all during wet years.
Even though the details are specific to each energy mix, the underlying problem is the same: how to ensure that the system has the capacity to provide the levels of energy needed? In New Zealand, the answer comes in three parts. First, New Zealand has abandoned the technical-administrative criterion of "generation adequacy". By going from an integrated monopoly to a market system, New Zealanders went through a shift in paradigm: the technical-administrative criterion which structured investment decisions in the world of the integrated monopoly with no price mechanism is made obsolete, and has been replaced by the Value of Lost Load (VoLL), which indicates the value of scarcity.
Secondly, the Authority and other actors carry out studies into the electrical supply/demand balance for the next ten years. Market participants therefore have access to all the information required to make investment decisions.
Third, the Authority has developed a futures market where participants can trade energy blocks for the next five years, which contributes to financing future capacity. It also requires participants to report their exposure and coverage strategy for the coming years to their boards. In the event of a dry year, suppliers cannot ask for support from public authorities, on the grounds that they were unprepared. The public authorities would simply reply they had the time and means to prepare.
Direct subsidies for renewable energies
There are no direct subsidies for renewable energies in New Zealand. Wind farms sell their megawatt hours on the market. It is true that wind power has a high availability factor in New Zealand - around 40% for land-based wind farms, compared with 27% in the United Kingdom.
Some solar panels are starting to appear on roofs. They do not receive direct subsidies either. Each producer sells their energy to their supplier in a free market.
However, solar power enjoys indirect subsidies through network use rates. This is essentially variable and is expressed in $/MWh. Installing a solar panel not only introduces savings in terms of energy but also in terms of the price of electricity on the network. However, network costs are not reduced when a panel is installed. As the number of MWh is reduced and costs are constant, the rate in $/MWh increases. The other users pay a little more following the installation of a panel. Solar panels therefore receive an indirect subsidy.
3. What lessons can be drawn from these observations?
The challenge for all market architects is to reconcile management of the traded product's physical and technical characteristics with a need for efficiency, which requires the rules in place to be simple and transparent. In the case of electricity, market architects in North America and Europe have lent a great deal of importance to the first requirement, to the detriment of simplicity. Thus, we have devised very complex market rules, which try to follow historic practices as closely as possible, but which limit opportunities for new actors to enter the market (or make it more difficult for them to do so).
Broadly speaking, three models have emerged. In the US, the market architecture attempts to reflect technical realities as closely as possible: the spot market is centralized, i.e. all producers, suppliers and clients trade in a single market organized by the system manager (TSO), which operates the day before to resolve the "unit commitment" problem. Electricity prices are calculated for each node, to take into account congestion and loss in transport networks, markets are organized for different reserve and auxiliary systems. Increasingly sophisticated algorithms are used to define products which reflect engineering practices increasingly accurately. Very sophisticated capacity mechanisms make up the final piece in this architecture. In the end, the main difference with the old, regulated organization is that prices have replaced the dual values calculated using optimization algorithms at the old integrated companies.
Europe has chosen an intermediary model. Markets are decentralized. They are not organized by centralized national TSOs. These are only responsible for the adjustment between production plans -and transfer plans proposed by the markets (or demand-supply balance managers) and the physical constraints of the network. Like in the US, the day ahead market is the market of reference, which is supplemented by a sequence of adjustment markets. Congestion on transport networks between member states is taken into account by combining national markets but, unlike in the US, congestion within each state is ignored. Finally, like in the US, capacity mechanisms were progressively introduced to complete the system. These mechanisms are national which, in the long term, raises the issue of their compatibility with the principle of a single European electricity market.
In New Zealand, market architects have made different, and probably more judicious choices. Like in the US, they have adopted nodal prices, which take into account the reality of transporting electricity. They have made the choice to not take into account "unit commitment", which enabled them to have a single spot market. They also chose an "energy only market", that is to say remunerating producers solely based on actual production rather than capacity in place.
These choices were the result of technical characteristics which are particularly favourable to the New Zealand Electricity system, but also deeply-held convictions about the efficiency of competition as a means of allocating resources.
The Europeans and Americans (following the California crisis) have limited confidence in the application of the principles of competition in the electricity industry. They therefore found themselves in a vicious circle. Public authorities set the rules whereby markets operate. These develop over time, following changes in context, objectives or government. Industry actors therefore adopt a strategic behaviour faced with a changing a set of rules, to which public authorities respond with new, even more sophisticated rules, which are inapplicable in practice. This situation is profitable to consultants who advise public authorities and participants, and for researchers in the field of the economics of energy, who are kept busy, but it has led many observers to question the effectiveness of opening the electricity industry to competition.
New Zealand, on the other hand, has entered a virtuous circle by structuring the market in a way which favours competition and innovation. Public authorities – the Ministry and the Authority - are committed to intervening as little as possible, and to doing so predictably. Participants adapt their behaviour to this environment and try to make a living in other ways than by demanding subsidies.
* * *
It is too early to tell whether this virtuous circle is sustainable. How will society and elected officials act in the face of power cuts during a particularly dry year? And how long will the government resist the calls for subsidies which, like the call of the ring, are extremely powerful? For the time being, the architects of the New Zealand electricity market, like the four famous hobbits, continue undaunted.
Title: In reference of course to the “Return of the King”, although we do not imply that academics are as important as kings.
 After all, in The Hound of the Baskervilles, Sherlock Holmes solves the mystery by analysing why the dog didn't bark.
 One of the authors of this blog was invited to visit the Authority's premises in Wellington.
 "The Electricity Authority promotes competition in, reliable supply by, and the efficient operation of, the New Zealand electricity industry for the long-term benefit of consumers". Extract from the Authority web page.
 http://www.emi.ea.govt.nz/Reports/VisualChart?reportName=R_MSS_C&categoryName=Retail&reportGroupIndex=0&eventMode=Async&reportDisplayContext=Gallery -
 This problem has led to a rich strand of scientific research, for example: “Testing for Market Efficiency with Transaction Costs: An Application to Financial Trading in Wholesale Electricity Markets “, by Akshaya Jha and Frank A. Wolak, September 2015, “Efficiency Impact of Convergence Bidding in the California Electricity Market” par R. Li, A.J. Svoboda, and S. S. Oren, Journal of Regulatory Economics, December 2015, 48(3) pp 245-284, and more recently, “Virtual Bidding and Electricity Market Design”, by Bill Hogan.
 Fred Schweppe and his co-authors at MIT formalized the theory of nodal prices in their work Spot Pricing of Electricity, published in 1988. In his article Contract Networks for Electric Power Transmission (1992) Bill Hogan, a professor at the Kennedy School (Harvard University), showed how nodal prices can and should be used in spot electricity markets.
 http://www.emi.ea.govt.nz/Reports/VisualChart?reportName=W_P_C&categoryName=Wholesale&reportGroupIndex=0&eventMode=Async&reportDisplayContext=Gallery -
 Like New Zealand, most of Columbia's electricity is hydraulic. The country is currently facing a dry year, following El Nino. To manage this risk, the Columbian government has implemented a capacity mechanism, which shows that the choice made by New Zealand is not simply a product of the production technology available to them. Furthermore, it would seem that the capacity mechanism in Columbia is not without its issues: http://market-analysis.co.uk/PDF/Academic/Britain'selectricitycapacitymarketfinal10April2014.pdf
 Energy is one of the seven industrial sectors which is the responsibility of the Ministry of Business, Innovation and Employment. One of the six other sectors is film (Screen sector), which brings us back to Lord of the Rings.