A utility-centric and fairness-based design for smart demand response tariffs

Abstract

Demand response tariffs (DRT) have long been commended for the economic benefits they deliver to electricity utilities and to end users. Conventionally, DRTs apply varying prices at different times of the day, and include time of use, critical peak pricing, and real time pricing. However, three challenges with DRTs remain to be resolved, namely, adverse user selection, low-cost efficiency in deployment, and fairness. To address these challenges, this thesis seeks to design new smart demand response (SDR) tariffs that take advantage of recent innovations in smart meter data and advanced analytics in relation to tariff pricing and user enrolment. Developing SDR tariffs involves balancing multiple conflicting principles, including sufficiency, efficiency, and fairness. We propose two SDR tariff designs that represent two different approaches to balancing these principles. The first is a fairness-based tariff design. It takes into account the impact of the tariff on users of different income levels, in addition to ensuring revenue sufficiency for utilities. The second is a utility-centric tariff design. It focuses on cost reduction and revenue sufficiency for utilities, while leaving the aspiration for fair cost distribution to other government programmes. The first tariff design combines smart user identification with fairness-based principles. It aims to create a progressive cost distribution between users of different income levels, while simultaneously maintain the profitability of utilities. Our SDR tariff design features a differential tariff deployment approach based on price responsiveness and income, and a filtering process that ensures our SDR tariff design meets the fairness-based principles. A novel user-identification method is put forward to identify users who are responsive to price changes. Using data from Irish Time of Use (TOU) trial, the results show that our SDR tariff design can achieve both a significant reduction in total user cost and a progressive cost distribution across the low-income and the high-income user groups. The proposed tariff is also more cost-efficient than the conventional TOU tariff and can reduce the cost of deployment. The second SDR tariff design combines advanced data analytics of user segmentation with utility-centric principles. The tariff design aims to increase the profits of utilities under the constraint of bill neutrality. It segments users into different groups according to their typical load profiles and adopts an algorithm to determine the TOU tariff for different user groups. Using data from Irish TOU trial, our results show that the utility-centric tariff design can effectively address adverse selection and prevent revenue erosion. Our study is therefore among the first to demonstrate how SDR tariffs can be designed to overcome the challenges faced by their conventional counterparts. In particular, it fills the missing link between data analytics and tariff design. Furthermore, it demonstrates methodological innovation via developing algorithms to determine the price levels of smart TOU tariffs, and to identify which users are responsive to price changes. These algorithms will inspire the design of next-generation SDR tariffs, as they are intended for wider adaptability and transferability across jurisdictions