There are many rapidly emerging technologies becoming available to electricity consumers around the world in economically viable form. These technologies allow for self-generation of electricity using green and renewable means (such as PV panels), storage (household batteries such as the Tesla Power Wall, or electric vehicles), demand response (smart meters and smart appliances) and/or combined heat and power units. The common theme of these technologies is that unlike large hydro, wind, or thermal power plants, they do not benefit from economies of scale and they are economically viable for individual households. As electricity consumers move towards self-generation and storage, utilization of the electricity grid will decrease. This project is aimed at constructing a suite of analytics tools that can be easily utilized to assess the impact of various scenarios of uptake of technology and demand response on the grid, transmission pricing and any natural consequences of under utilization of the grid.

Investment in electricity generation and transmission is subject to considerable uncertainty. In the short term, wind intermittency means that the investment mix must include enough fast ramping and peaking plant to cover shortfalls in capacity when the wind does not blow. In the medium-term security of energy supply in dry winters needs to be included in provision and use of energy on an annual basis. Over the long term, variations in the rate and location of demand growth will necessitate robust investment plans to avoid poor choices of generator location and technology. The future of prices on carbon emissions also affects this choice. Finally, there is considerably regulatory uncertainty that affects electricity prices, and hence the income that generators earn from their investments. This PhD project will entail the development of a suite of stochastic integer programming models that integrate these sources of uncertainty into a system that can be used to study the long-term evolution of the New Zealand electricity market. These will build on and complement the GEMSTONE model developed by Giradeau and Philpott. This research is also linked to strategic models of investment.

Most electricity pool markets have a two-settlement structure. Agents offer generation to a day-ahead market that arranges dispatch for most generating plant in advance. In real time a secondary market manages variations about this as load varies. In New Zealand, we have a single real-time dispatch mechanism that is operated over a rolling horizon with a 2-hour "gate closure". The original day-ahead structure was proposed in 1996, but not implemented for various reasons, including the additional complexity involved in a system with 244 grid injection/exit points. Advances in computer speed and solver capabilities mean that the day-ahead model can be revisited. This PhD project will develop and investigate day-ahead pool markets in the context of the NZEM and evaluate their effectiveness when compared with a single-settlement market. Part of this project will also consider single-settlement day-ahead markets that are based on stochastic programming (see e.g. this paper by Pritchard et al.).

As a result of a big research investment drive in the United States, there has been a lot written recently about the engineering of the so-called smart grid. This project will focus on potential market arrangements that might arise from the implementation of new technologies that will emerge from this research. The smartness of the grid will emerge from prices incentivizing optimal behaviour from grid users. For example, wind power is intermittent, and so it is lost unless it can be used immediately or stored. Energy storage can be built by merchants who pay for it by arbitraging energy prices over time. Agents who store energy on behalf (without taking a position on energy price) might also participate in such a market, and price their services accordingly. Agents might also offer demand-side response in such an environment that combines with storage to produce an optimal outcome. In this setting risk averse agents will seek policies and hedging instruments to decrease risk. This PhD project will focus on developing and investigating pricing models that can be devised to lead to close to optimal generation and consumption outcomes in the smart grid.

Security of supply is a frequently addressed concern in electricity markets such as the NZEM. Sufficient investment in generation is needed to ensure the security of electricity supply in such markets yet, electricity generators are risk averse with respect to such large investments and governments mindful of gold-plating the system and/or interfering in deregulated markets. We will develop a model that employs coherent risk measures to reflect a generating firm's risk aversion level and recommend under an assumed stochastic process for the price of electricity and perhaps another stochastic process for fuel cost, if that firm should invest in building a generation plant.

This project aims to extend the existing EPOC models and build new models that address the question of optimizing revenue for a hydro-electric generator whose actions influence the price of electricity. Such generators are often major generators. As their offer policy influences the outcome of the dispatch, hence the price of electricity, they cannot observe price as an exogenous process.

The reviews released on the NZEM in 2009 has resulted in an electricity industry bill currently under review in the Parliament, where 21 changes are proposed to the structure of the NZEM. One of these changes is to explicitly incorporate demand response functions in the objective of the SPD software. SPD is the side constrained network optimization problem that runs every period to determine the nodal prices of electricity and quantities dispatched. This project will devise methods to enable a major consumer of electricity to submit an optimal demand response bid to the wholesale electricity market in one period. We will then go on to model a decision process that would drive the major consumer's decisions over a time horizon beyond the single period.

The reviews released on the NZEM in 2009 has resulted in an electricity industry bill currently under review in the Parliament, where 21 changes are proposed to the structure of the NZEM. One of these changes advocates limiting the the number of tariffs available to distributors of electricity. The reason for this change is that (household) consumers of electricity are overwhelmed by the large number of options hence refrain from changing distribution companies or tariffs. This project will look at the optimal set of tariffs that would maximize total social welfare in an electricity market and what may be important attributes of such tariffs.

The reviews released on the NZEM in 2009 has resulted in an electricity industry bill currently under review in the Parliament, where 21 changes are proposed to the structure of the NZEM. One of these changes proposes divestiture of Tekapo A and B generation stations from Meridian to Genesis. The bill then outlines that a water management agreement must be reached between Meridian and Genesis to ensure that consent conditions, flood management requirements and Meridian's obligations under the Tiwai point contracts must be met. We will look at the agreement design from from the perspective of the two generators and attempt to design a steady state water management agreement. We will then assess the impact of this agreement on offer strategies of the generators.

Electricity distributors such as Transpower and Vector are among the regulated industries in NZ and have a cap on their annual revenue. Hence electricity distributors seek a policy of setting their fixed, variable and peak charges so as to meet a target revenue and discourage peak consumption. Peak consumption leads to (future) investment cost that is to be avoided if possible. In order to discourage peak consumption, they need a model for consumer behaviour. We will consider peak shaving and peak shifting models for consumers in a distribution network and analyze the distributor's revenue optimization problem as a 2 stage leader-follower game. This project will be extended to consider coincident peak charges (where more than a single major consumer is involved,) stochastic spot prices of electricity, and the network effects (i.e. the location of the consumers).