Renewable energies require new approaches to the operation of power systems. In this work, we present a novel Demand Response (DR) program to mitigate grid congestion. The program stands out by adjusting electricity prices online and solely based on observed aggregate electricity consumption. It thereby avoids privacy concerns and expensive information system infrastructure investments. Additionally, it can cope with non-elastic and time-interdependent demand—which is typical for storage devices and industrial processes. Beyond the DR program, this paper contributes to research on flexible load modeling. Specifically, we model four flexible load types and provide a unified framework of load modeling under uncertainty and with time interdependencies. Numerical experiments show that our DR program achieves considerable and stable cost savings of 40% to 60% in comparison to conventional DR programs. Moreover, the solution approach based on Deep Reinforcement Learning reaches these savings after a short learning period corresponding to only 25 simulation days. Regarding load flexibility, we find that the responses of the four load types strongly differ depending on prices and the lag between the announcement and the application of price changes (`notification interval’) which is an important design parameter. Our results imply that system operators should flexibilize their DR programs with the help of learning algorithms and tailor their program to the local load composition, congestion frequency, and forecasting quality. For researchers, our solution approach and our unified framework can help to solve pricing problems with unknown and heterogeneous demand, which occurs on online platforms and in other environments.
Local electricity markets (LEMs) are a promising approach to integrate flexible appliances into electricity systems and manage system constraints on the distribution level. In this article, we suggest an approach to derive bidding functions for time-interdependent electricity-based services and use our framework to analyze the welfare implications of an LEM. Previous work have left such bidding functions undefined which are, however, an important bridge between customer preferences and technology in a smart grid. Furthermore, while welfare analyses exist for the transmission level, we are not aware of equivalent studies for residential distribution systems. We specify a bidding function for Heating, Ventilation, and Air Conditioning (HVAC) systems — a major load in residential distribution systems — and pursue a case study of 437 houses. We find that the introduction of an LEM can realize welfare gains of more than 17,000 USD over a year. These benefits are largely driven by savings in energy procurement costs during a few weeks. Moreover, all houses contribute to this gain and houses which contribute more benefit over-proportionally. Furthermore, LEMs can contribute to the management of constrained systems. We derive the marginal value of investment and show that optimal grid expansion is less than under a fixed retail tariff.
Our results demonstrate that LEMs can provide system benefits, however, important design questions are still open, for instance who should incorporate the role of an LEM operator.
Load aggregation can be an effective approach to bundle residential loads and market their flexibility and central markets. In this work, we propose a business model for an aggregator with uncertain generation. By deploying flexible demand resources, the aggregator is able to balance short-term fluctuations of generation and avoid balancing costs. We design a contract menu for customers with HVAC systems and private temperature sensitivities and show that an incentive-compatible contract menu is feasible. This extends existing literature to time-interdependent operations and specific electricity-based services. Furthermore, we use a numerical example to demonstrate that customers are better off under a curtailment contract than under a fixed retail tariff, that low-comfort customers are better off than under real-time pricing, and that the bundling of generation and flexible load realizes synergies for the aggregator as opposed to the separate marketing of generation and flexible load.
Rausch, B.; Mayer, K.; Arlt, M.-L.; Gust, G.; Staudt, P.; Weinhardt, C.; Neumann, D.; Rajagopal, R. (2020): An Enriched Automated PV Registry: Combining Image Recognition and 3D Building Data. Accepted for NeurIPS 2020, Workshop Tackling Climate Change with Machine Learning.
Mayer, K.; Wang, Z.; Arlt, M.-L.; Neumann, D.; Rajagopal, R. (2020): DeepSolar for Germany: A deep learning framework for PV system mapping from aerial imagery. 2020 IEEE Smart Energy Systems and Technologies (2020 IEEE SEST), Istanbul, Turkey, September 07 – 09, 2020.
Arlt, M.-L.; Neumann, D.; Rajagopal, R. (2019): Curtailment Contract Design for HVAC Systems and Benefits for Management of Uncertain Generation. Accepted for IEEE PES General Meeting, Atlanta, USA, August 4-8, 2019.
Arlt, M.-L.; Rajagopal, R. (2019): Curtailment Contract Design for HVAC Systems. Accepted for 2019 American Control Conference (ACC), Philadelphia, USA, July 10-12, 2019.
Arlt, M.-L.; Cardoso, G. and Weng, D. (2017): Hydrogen Storage Applications in Industrial Microgrids. 2017 IEEE Green Energy and Smart Systems Conference (IGESSC 2017), Long Beach, USA, November 6-7, 2017 (Presentation and publication in conference proceedings).
Arlt, M.-L., and Astier, N. (2020): Where can Drivers Charge Their Electric Vehicle for Free in the U.S.?, IAEE Energy Forum
Arlt, M.-L. (2018): Preparing electricity markets for renewable energies. Newsletter of the European Centre for Energy and Resource Security, King’s College London, Issue 76, June 2018.
Ecofys and Fraunhofer IWES (2017): Smart Market Design in German Distribution Grids (German Smart-Market-Design in deutschen Verteilnetzen). Study commissioned by Agora Energiewende. Contributors: Ropenus, S.; Nabe, C.; Arlt, M.-L.; Döring, M.; Holzhammer, U.; Gerhardt, N.