• Energy Research

Electricity imbalance pricing provides the ultimate incentive for generators and suppliers to contract with one another ahead of time and deliver against their obligations. As delivery time approaches, traders must judge whether to trade-out a position or settle it in the balancing market at the as-yet-unknown imbalance price. Forecasting the imbalance price (and related volumes) is therefore a necessity in short-term markets. However, this topic has received surprisingly little attention in the academic literature despite clear need by practitioners. Furthermore, the emergence of algorithmic trading demands automated forecasting and decision-making, with those best able to extract predictive information from available data gaining a competitive advantage. Here we present the case for developing imbalance price forecasting methods and provide motivating examples from the Great Britain’s balancing market, demonstrating forecast skill and value.

There is a reduction in the percentage penetration of synchronous machines within the Great Britain (GB) power system leading to a decrease in inertia, and an increase in system rate of change of frequency (RoCoF) resulting from power imbalances. This raises the challenge of containing frequency deviations to within the relevant operational limits. As a result, steps need to be taken by the system operator to manage the risk to system security. In order to better understand this risk, this paper presents the challenge in light of the changing energy landscape and the current and future frequency response services available to contain frequency deviations. Although the current GB frequency response services may be capable of containing most events within frequency limits, in low inertia scenarios these responses alone are not capable of containing excursions within practical RoCoF limits. Consequently, further action must be taken to ensure system security. The system operator currently employs an interim solution of limiting the largest loss risk, depending on system inertia and the RoCoF limit. While this is suitable in the short-term, it is unlikely that this option will be cost-effective in the future.

Power systems require a wide range of ancillary services in order to function and renewables will be expected to provide such services in line with their increasing penetration. This paper focuses on the participation of wind energy in response and reserve markets. We compare the present situation in Great Britain (GB) and Spain, and make recommendations to support future development. Wind is already participating in a limited range of ancillary services in both countries: frequency response in GB and reserve services in Spain. We analyse the effects of market design, subsidy arrangements, and systemspecific needs on participation of wind in these markets, and then make policy recommendations designed to enable increased participation from wind. Our recommendations include the use of short-term markets to enable the use of accurate wind power forecasts, capacity-based subsidy schemes to avoid distorting ancillary service markets, and facilitating the participation of aggregated (single and mixed technology) resources. Country-specific recommendations include revising the current settlement process in GB to remove the incentive to over-estimate short-term generation forecasts, and establishing a competitive frequency containment reserve market in Spain. These recommendations are supported by analysis of publicly available market data.

This paper describes two methods for creating improved probabilistic wind power forecasts through the use of turbine-level data. The first is a feature engineering approach whereby deterministic power forecasts from the turbine level are used as explanatory variables in a wind farm level forecasting model. The second is a novel bottom-up hierarchical approach where the wind farm forecast is inferred from the joint predictive distribution of the power output from individual turbines. Notably, the latter produces probabilistic forecasts that are coherent across both turbine and farm levels, which the former does not. The methods are tested at two utility scale wind farms and are shown to provide consistent improvements of up to 5%, in terms of continuous ranked probability score compared to the best performing state-of-the-art benchmark model. The bottom-up hierarchical approach provides greater improvement at the site characterized by a complex layout and terrain, while both approaches perform similarly at the second location. We show that there is a clear benefit in leveraging readily available turbine-level information for wind power forecasting.

Electricity load forecasting is a necessary capability for power system operators and electricity market participants. The proliferation of local generation, demand response, and electrification of heat and transport are changing the fundamental drivers of electricity load and increasing the complexity of load modelling and forecasting. We address this challenge in two ways. First, our setting is adaptive; our models take into account the most recent observations available, yielding a forecasting strategy able to automatically respond to changes in the underlying process. Second, we consider probabilistic rather than point forecasting; indeed, uncertainty quantification is required to operate electricity systems efficiently and reliably. Our methodology relies on the Kalman filter, previously used successfully for adaptive point load forecasting. The probabilistic forecasts are obtained by quantile regressions on the residuals of the point forecasting model. We achieve adaptive quantile regressions using the online gradient descent; we avoid the choice of the gradient step size considering multiple learning rates and aggregation of experts. We apply the method to two data sets: the regional net-load in Great Britain and the demand of seven large cities in the United States. Adaptive procedures improve forecast performance substantially in both use cases for both point and probabilistic forecasting.

Abstract Installed capacities of wind and solar power have grown rapidly over recent years, and the pool of literature on very short-term (minutes- to hours-ahead) wind and solar forecasting has grown in line with this. This paper reviews established and emerging approaches to provide an up-to-date view of the field. Knowledge transfer between wind and solar forecasting has benefited the field and is discussed, and new opportunities are identified, particularly regarding use of remote sensing technology. Forecasting methodologies and study design are compared and recommendations for high quality, reproducible results are presented. In particular, the choice of suitable benchmarks and use of sufficiently long datasets is highlighted. A case study of three distinct approaches to probabilistic wind power forecasting is presented using an open dataset. The case study provides an example of exemplary forecast evaluation, and open source code allows for its reproduction and use in future work.

Outputs from the Network Innovation Allowance project "Control REACT" (workstream 2), sponsored by National Grid Electricity System Operator (NGESO). This deposit contains underlying data used in this project. The R code (Rmarkdown) and html renders of these workbooks are available in a separate deposit linked below. See description there for further details. In order to run the R scripts, data and code must be arranged in the directory structure given in "Directory Structure.pdf". Wind, solar and net-demand data are derived from raw data made available by Elexon and Solar Sheffield via public APIs. See respective websites for details, our processed (aggregated and cleaned) versions of this data are shared here under a CC-BY license. Weather forecast data are derived from historic operational forecasts from the ECMWF HRES model and are shared under a CC-BY licence. For details on how these were processed please see references. {"references": ["J. Browell and M. Fasiolo, \"Probabilistic Forecasting of regional net-load with conditional extremes and gridded NWP\", IEEE Transactions on Smart Grid, vol. 12, no, 6, pp. 5011-5019, 2021", "C. Gilbert \"Topics in high dimensional energy forecasting\", J. Browell & D. McMillan, degree supervisors; Centre for Doctoral Training in Wind and Marine Energy Systems; Department of Electronic and Electrical Engineering Thesis [PhD] 2021"]}