Assessing capacity credit of demand response in smart distribution grids with behavior-driven modeling framework

In smart grid, demand response (DR) provides the utilities with a new alternative to mitigate the operational uncertainties and achieve the power balance target However, unlike conventional generation units, the performance of DR is strongly dependent on behavioral pattern of customers. As such, to what extent DR programs could be utilized to provide capacity support and contribute to the adequacy of the supply turns out to be an important concern for the utilities. In order to resolve this issue, this paper presents a new methodological framework for assessing the reliability value of DR in a context of distribution grid. The proposed approach is developed on the generation-oriented concept of capacity credit (CC) and it extends the CC application to a DR setting. As the major contribution of this work, the proposed framework accounts for the impacts of both physical and human-related factors on the availability of DR; furthermore, the uncertainty issue that associated with demand-side performances is also explicitly considered in our study. To properly handle the ambiguities of customers' willingness for DR participation, a novel Z-number-based technique is introduced. Through such an approach, not only the inherent randomness accruing from the demand-side could be captured, but the impact of information creditability would also be accounted for, which could allow a more realistic characterization of DR as compared with existing studies. By jointly using fuzzy-expectation technique and the centroid method, the different types of uncertain variables (probabilistic and Z-numbers) involved in our analysis can be normalized into comparable quantities and then used for the CC evaluation of DR. The proposed framework is illustrated based on both a small test race and a real distribution system, and the obtained results verify the significant role of DR in enhancing the reliability of supply, as well as its sensitivity to different influencing factors.