Graeme Hawker

Synthesis of wind time series for network adequacy assessment

When representing the stochastic characteristics of wind generators within power system simulations, the spatial and temporal correlations of the wind resource must be correctly modelled to ensure that reserve and network capacity requirements are not underestimated. A methodology for capturing these correlations within a vector auto-regressive (VAR) model is presented, and applied to a large-scale reanalysis dataset of historical wind speed data for the British Isles. This is combined with a wind speed-to-power conversion model trained against historically metered data from wind farms on the Great Britain (GB) electricity system in order to derive a lightweight model for simulating injections of wind power across a transmission network. The model is demonstrated to adequately represent ramp rates, both at a site and network level, as well as the individual correlations between sites, while being suitable for network adequacy studies which may require the simulation of many years of operation.

Adequacy assessment of future electricity networks

Liberalisation of electricity markets, changing patterns in the generation and use of electricity and new technologies are some of the factors that result in increased uncertainty about the future operating conditions of our power system. In this context, planning for future investments in power system requires careful selection and assessment of future operating conditions. This paper revisits the notion of power system adequacy and highlights the need for consideration of some factors that have hitherto tended not to be part of a transmission expansion planning process, in particular in respect of the credible range of possible values of system operating conditions and transitions between successive operating states. Firstly, we present some definitions of power system operational regions. Secondly, we present a stochastic optimisation model that measures the adequacy of a transmission network for given future operating conditions. Uncertainties in demand and generation are modelled using a large number of scenarios. The optimisation model identifies the critical future operating conditions needing the special attention of a power system planner. The proposed model is simulated on a 39-bus network, whereby it is shown that this model can identify critical operating conditions that need the attention of a system planner.

The impact of maintenance contract arrangements on the yield of offshore wind power plants

In the optimisation of maintenance and vessel strategies for the operation of offshore wind plant, it is normally assumed that the off-taker of the power produced may directly control the dispatch of maintenance resources. However, in practice, services such as maintenance technicians and vessels are usually contracted from companies with larger arenas of operation, and so the organisational interfaces between these parties, and the different objective functions involved, need to be considered. This article looks at different current and future models for contracted maintenance, identifies interfaces and conflicts of interest, and constructs a quantified model demonstrating the potential impact on headline energy yields for a set of wind farms with a common contracted maintenance resource. The modelling illustrates that the performance of a site with contracted maintenance operations is not only dependent on the contracts held by that site but also on the effective competition in place with other sites for a centralised resource, and the performance of a site may be highly sensitive to the alignment of contractual incentives, relative travel distances, and the relative size of the site in terms of energy yield.

A Review of probabilistic methods for defining reserve requirements

In this paper we examine potential improvements in how load and generation forecast uncertainty is captured when setting reserve levels in power systems with significant renewable generation penetration and discuss the merit of proposed new methods in this area. One important difference between methods is whether reserves are defined based on the marginal distribution of forecast errors, as calculated from historic data, or whether the conditional distribution, specific to the time at which reserves are being scheduled, is used. This paper is a review of published current practice in markets which are at the leading edge of this problem, summarizing their experiences, and aligning it with academic modeling work. We conclude that the ultimate goal for all markets expected to manage high levels of renewable generation should be a reserve setting mechanism which utilizes the best understanding of meteorological uncertainties combined with traditional models of uncertainty arising from forced outages.

Analysis of wide-area availability of wind generators during storm events

During high wind speed shutdown (HWSS) events, the power outputs of wind power plants may be subject to high ramp rates, causing issues for the System Operator (SO) in predicting total wind output, allocating adequate reserve levels and minimising balancing costs. As the timing of these events is difficult to predict, it is proposed that individual turbines may be used as probabilistic early warning indicators of HWSS events across sites, and by extension to a wide geographical area. The shut-down history of two separate wind farms across Scotland is analysed to determine the likelihood and impact of such events. It is shown that in most cases, HWSS does not result in the full loss of availability. Factors such as turbine elevation and mean wind exposure are key indicators of the order of shut-down across a site. The suggestion that sites could be used as early warning indicators for the pattern of HWSS across a transmission zone is difficult to characterise and for the two wind farms studied, prediction was not consistent.