Search

This engineering doctorate (EngD) thesis has investigated and improved the understanding of Demand Side Response (DSR) aggregators, DSR estimation methods, and developed a new load profile-based estimation method. The primary motivation for this research was to develop and improve the understanding of different DSR estimation methods and their effectiveness for assessing new sites as suitable for DSR. DSR aggregators play a key role in facilitating DSR uptake by providing over 80% of DSR capacity. Therefore, this research has focused on the estimation methods that a UK-based aggregator uses to determine the suitability of new end users. As an intermediary in the DSR assessment and programme enrolment process, aggregators need to ensure that each end user site is suitable for DSR. Otherwise, both the aggregator and the end user could be negatively impacted if financial returns from participation fail to cover DSR implementation costs. Therefore, this research was undertaken with the aim of better quantifying the uncertainty in DSR estimation methods for new sites, with a view to improving the assessment of their suitability to participate. The research was undertaken in conjunction with KiWi Power Limited, a UK-based DSR aggregator, by establishing and then addressing three interlinking objectives. The first objective mapped out the criteria used by KiWi Power to determine the suitability of an end user's site for DSR and found that the highest priority for KiWi Power during the assessment process is understanding the DSR potential of a site's assets. The second objective compared the outcome uncertainty and information input requirements of four existing DSR estimation methods using as the example asset HVAC Chillers and their sub-meter usage data from two UK hotel sites. The comparison results showed a range of uncertainty levels which produced mean average percentage error (MAPE) levels of between 39% to 159%, with the estimation methods costing between £10 to £180 to perform on new sites. The third objective developed and evaluated a new method that uses load profiles of assets to reduce the uncertainty of DSR potential estimation during an aggregator's assessment process. The new method compares favourably against the existing DSR estimation methods, as it generated the second lowest MAPE level of 46.5% with an estimated usage cost of £26. The new method demonstrated additional benefits of being usable earlier in the assessment process for a new site when compared to the existing methods, and offered the ability to use pre-calculated uncertainty levels enabling users to adjust the estimation outputs based on an organisation's risk appetite.