Dr. Genaro Longoria successfully defended his PhD thesis at his Viva Voce on 6th April 2020, which took place remotly with Dr. Stepan Ivanov as Internal Examiner and Dr. Steven D. Prestwich as External Examiner. Dr. Longoria’s thesis is titled “Open-market Energy Procurement Strategies to Integrate Wind Energy”. His research focused on the sustainability of renewable power sources by handling the uncertainty to increase the share in the aggregated power supply. Dr. Longoria’s work has been published in top tier journals.
To contextualize his work, the speed at which technological breakthrough have been accomplished in recent years has no precedent, for example in the ICT domain and electric vehicles. Moreover, the constantly decreasing prices have increased the popularity and affordability to these technologies. From the hard to the fog, basically all infrastructures have been influenced or have had an impact from the strepitous progress. To make this possible at the core lies a key factor: Energy.
Blessed by the geographic location, Ireland has one of the best wind resources in the world. Integrating wind energy has many advantages. The running cost of a wind farm incurs in nearly zero marginal costs, the variable cost is basically negligible. It also decreases the dependence on scarce indigenous and imported fossil fuels. In addition, deployment at different scales can be done with ease and decentralizes power generation which helps to alleviate transmission congestion.
The hard infrastructure of wind power (for example blade and nacelle aerodynamics) is at a mature state which is reflected in the falling Levelized cost of Energy. Nevertheless, generating electricity from wind power is still dependent on support schemes. Incorporating wind energy into portfolios represents a significant challenge for Load Serving Entities. Therefore the sought after characteristics of wind energy hinder potential hefty balancing costs or penalty fees leading to higher operational costs, percolating to the end-customer, and thus meager or negative profit. As with other new technologies in the past, governments around the globe have incentivized renewable generation with tax-based safety nets to provide significant risk mitigation, foster investment and access to low cost finance.
Dr. Longoria’s research is on reducing the need of exogenous support to integrate wind energy. As a first milestone the research identified the operational context of a typical price-taker wind power producer and the gaps in the literature. This gives place to a splitting of the portfolio concocting in two strategical horizons. The first horizon is the long-term energy procurement. This is based on non-cooperative behavioral techniques and evolutionary programming to determine long-term bilateral contracts between power wholesalers and electricity retailers. This is followed by a risk constrained continuous stochastic model for spot electricity prices and a wind power for a given market. A challenge in modeling energy prices, unlike other commodities, is taking into account the lack of storage of scale for long periods. The effect of this are sudden short-lived price escalations and a mean reversion process. The former is modeled with a Compound Poisson term compensated for trends and the latter with an Ornstein–Uhlenbeck process. For decision-making the framework introduces Conditional Value-at-Risk and Excess Cost to assess the power mix uncertainty. Overall this provides a methodology to determine the expected cost of energy procurement for different levels of risk attitudes.
The second part addresses the short term horizon. In particular, the internal trading of a hybrid power producer and with a power exchange. Real data from Nordpool (one of the leading power exchanges in Europe) is used and random power outages to test performance. The result is a multi-policy computer agent aimed at running a renewable power plant without external financial supports. This is accomplished by reducing exposure to volatile markets (for instance the post dispatch regulatory markets), optimal allocation of energy flows and utilization of the hybrid power plant. The agent comprises a wisdom extracting algorithm from bulky data, a sequential forecast engine of short-term price events and the management of energy flows within the power plant and externally to the market.
As a summary the research comprises Mixed-strategies Game theory to determine the equilibrium share of renewable and conventional power. To find the Nash portfolio a tailored made solver is developed. Then the work proposes a theoretical continuous framework whose output is risk assessed thus providing an answer for different risk policies. Lastly an autonomous energy trader. The backbone of the agent implements a wait-and-see receding horizon with a conjunctive reward function of the multiple objectives. The agent can make optimal decision on behalf of a subsidies-free price-taker wind farm with a hydro pumped plant.
Dr. Longoria will continue extending his research to uncertainty constrained cost-benefit analysis of the transmission network expansion for different technologies like district heating and hydrogen extraction.