Elsevier

Energy Policy

Volume 61, October 2013, Pages 423-429
Energy Policy

Demand response from the non-domestic sector: Early UK experiences and future opportunities

https://doi.org/10.1016/j.enpol.2013.06.051Get rights and content

Highlights

  • Empirical demand response data from non-domestic sector evaluated.

  • Load profiles suggest strong sector dependence on availability response at system peak.

  • Majority of aggregated demand response still stems from stand-by generation, not from demand turn down.

  • Scope for substantial increase in demand response capacity if response times were extended.

Abstract

Demand response is believed by some to become a major contributor towards system balancing in future electricity networks. Shifting or reducing demand at critical moments can reduce the need for generation capacity, help with the integration of renewables, support more efficient system operation and thereby potentially lead to cost and carbon reductions for the entire energy system.

In this paper we review the nature of the response resource of consumers from different non-domestic sectors in the UK, based on extensive half hourly demand profiles and observed demand responses. We further explore the potential to increase the demand response capacity through changes in the regulatory and market environment.

The analysis suggests that present demand response measures tend to stimulate stand-by generation capacity in preference to load shifting and we propose that extended response times may favour load based demand response, especially in sectors with significant thermal loads.

Introduction

The transition towards low carbon electricity systems brings about new challenges of system integration. Variable sources of generation, such as wind and photovoltaics, are set to displace output from thermal plants, which presently contribute towards load following and peak demand provision.

Demand response has been suggested as a potentially valuable resource in future electricity systems and could constitute an alternative to potentially more costly means of integrating low carbon technologies, such as backup generation, network expansion and physical electricity storage (Barton et al., 2013).

Furthermore, demand response has the potential to reduce peak demand and ease impending peak generation capacity shortages, resulting from expected plant closures over the coming decade in the UK (Ofgem, 2012), and potential increases in peak demand from electrified heat and transport in the medium to long term (Strbac et al., 2010).

However, SEDC (2011) argues that regulatory barriers are holding back demand response provision across Europe. To date, demand response still plays a relatively minor role in electricity system operation, which is dominated by electricity generation. Macleod (2012) estimates that even in the UK, the EU country where demand response provision is most prominent, only around 3% (84 MW) of responses stem from aggregated load reduction.

UK policy makers recognise the need to give an ‘equal role for the demand-side’ (ECCC, 2010). The Draft Energy Bill DECC, 2012a sees ‘significant opportunities’ for changes in demand and the proposed capacity mechanism is intended to enable the demand side to bid into future markets. However, fundamental questions remain over how such mechanisms should be configured and administered (House of Commons, 2012).

From a whole systems perspective, supply and demand side measures can indeed perform very similar functions. For stakeholders engaging in and relying on the services of demand response, some critical differences arise, relating to reliability, response time and response duration (Grünewald and Torriti, 2012).

One means by which the demand side can already participate in the UK is the short term operating reserve (STOR) market. STOR is defined by National Grid, the UK system operator, as ‘a service for the provision of additional active power from generation and/or demand reduction’. The conditions for participation include the provision of 3 MW or more of generation or demand reduction within 4 h from instruction for at least 2 h. Participants must have the ability to provide STOR at least three times a week and are given 20 h ‘recovery’ between service provision events.

Typical loads of most electricity consumers are well below this 3 MW threshold and demand aggregators offer a service of contracting several smaller sites and to pass on their collective response capacity on to the system operator. In this study we have collaborated with one such UK demand aggregator, Kiwi Power, and draw on their client data and experience in this field.

Demand side response is defined by DECC (2012a) as an ‘active, short term, reduction in consumption’. If, as is the case in this study, consumption is seen as the metered load of a site, then local stand-by generation can contribute towards a reduction in this metered-load, leading to an apparent reduction in consumption, while actual consumption need not have changed. Thus, demand response can be the result of reduction in demand through demand turn-down or an increase in local generation.

This paper seeks to understand the balance between these two modes of response in current demand response provision and asks how policy and market instruments may lead to more effective use of genuine load shifting.

The methodology and sources of demand response data are introduced in Section 2, before we assess the data in three stages. First, the demand profiles for selected sectors are translated into probability distributions in Section 3 to provide a sense of their availability at different times and how these differ between sectors. The response capacity for two of these sectors is presented in detail and discussed in Section 4, before Section 5 suggests potential trends for the scale of these capacities in response to changes in response time and duration. Section 6 concludes and draws policy implications resulting from the findings of this analysis.

Section snippets

Methodology

Regardless of whether demand response is provided by turn down or stand-by generation, the load at the time of request constitutes an upper limit for the capacity that can be provided. Turn down can at best reduce the load by as much as would have been consumed in the counterfactual case (the baseline consumption). A generator, similarly, can only displace the local demand, unless an export licence was in place. We have not considered sites with an export licence in this study, since these

Demand profiles in selected sectors

To what extent a site or sector can be relied upon to deliver a demand response depends on their load at the time of response need. A site that does not usually consume electricity at a certain time is unlikely to make a significant contribution towards demand reduction. From the load profiles in this study it is in some cases possible to establish probability distribution of the level of load under particular conditions. Fig. 3 shows some of these probability distributions.

Several conditions

Analysis of empirical demand response data

As mentioned in the Introduction, we encountered two classes of demand response among the sites reviewed here: demand turn down and standby generation. The theoretical response capacity available from these two groups differs significantly, with stand-by generation dominating. Only around 16% of the sites in this set of data provided demand turn down. Here we present response trial data from each class for sectors that provide predominantly one or the other type of response.

Stand-by generation

Capacity under changing response conditions

One of the possible limiting factors preventing higher penetration of turn-down demand response may be found in the regulatory conditions under which the service is presently provided.

The primary mechanism in the UK for demand response is the STOR market, which is designed to include load reduction (Hamidi et al., 2009). Participants can bid to the system operator to provide response capacity for given periods. Small loads are excluded from direct participation due to the 3 MW minimum provision

Conclusion

We have presented empirical load profile and demand response data from four different sectors in the UK. These data support the argument that present regulatory and market frameworks, while claiming to seek to create a ‘level playing field’ between demand and supply side measures, may inadvertently favour distributed generation assets in the provision of demand response. One of the reasons for this bias towards stand-by generation could be related to the fast response times expected from

Acknowledgements

This study was funded by the Technology Strategy Board under Project number 130821. We would like to express our gratitude to Marina Hod and Darryl Pope from Kiwi Power Ltd. for their support, their insights and their generosity in making these data available for research.

References (19)

There are more references available in the full text version of this article.

Cited by (61)

View all citing articles on Scopus
View full text