Optimization and control of Distributed Energy Resources in electric power distribution grids

Abstract

As renewable energy integration grows, the power grid is undergoing a rapid transition from a conventional, unidirectional system to a bidirectional and distributed network. In this new paradigm, end-users are becoming prosumer, individuals who both consume and produce energy. To enable prosumer participation in the power grid, new market mechanisms must be designed to align with both participant interests and network-level requirements.

We develop the concept of transactive multi-agent systems (MAS) as a tool to model transactions between prosumers in a power grid. Each prosumer is represented as an autonomous agent capable of making decisions about electricity usage and trading based on personal preferences, with pricing signals guiding their decisions. These agents have local energy generation, e.g., from rooftop solar panels, which they can either consume or trade within the network, generating income. Beyond electricity markets, carbon permit trading systems offer another real-world application of transactive MAS. In such systems, agents represent entities (e.g., nations or corporations) that hold carbon emission permits, engaging in peer-to-peer carbon trading. Transactive MAS facilitate decentralized decision-making and efficient resource allocation by dynamically pricing resources according to market mechanisms.

Efficient resource allocation is particularly crucial in systems where total demand must match total supply to ensure safe operation. For example, in power grids, maintaining this balance between electricity supply and demand is vital for ensuring frequency regulation. Microeconomic theory suggests that appropriate pricing of resource flows can ensure this balance. We adopt the concept of competitive equilibrium from microeconomics as a market-clearing mechanism, defining it as a pair of resource allocation and prices that maximizes individual agent payoffs while ensuring market clearance.

The social acceptance of the resource price is an important issue for the adoption of transactive MAS. For example, if the electricity price is considered to be too high for some agents in a transactive energy system, they cannot compete in the market and all resources will be allocated to those who have dominated the market. Specifically, in emergency situations, when some generators stop working or the demand is too high to be met by the available generators, the energy market fails. Acceptable resource pricing is also crucial in climate-economic systems, where international agreements like the Kyoto Protocol (1997) and the Paris Agreement (2016) have aimed to reduce emissions. Different locational prices is another social acceptance issue that arises in transactive MAS, which can cause social inequalities by restricting access to goods and services based on geographic location. For example, in electricity markets, locational price variations arise due to binding transmission and distribution constraints, placing greater limitations on users at the end of the feeder.

Designing social shaping mechanisms to ensure that resource prices, such as those for electricity and carbon emissions, are socially acceptable is crucial for the implementation of transactive MAS. This thesis introduces social shaping as a method to address such social acceptance challenges. We propose social shaping methods that maintain resource prices under a competitive equilibrium below a socially accepted threshold, mitigating high price spikes. Our design includes both static and dynamic formulations, forming Research Questions 1 and 2. In the static formulation, decision variables and prices remain fixed over time, whereas in the dynamic formulation, they evolve. Additionally, to address disparities in locational pricing caused by binding network-level constraints, we develop social shaping methods that promote uniform prices across the system, enhancing the social acceptance of transactive MAS. This constitutes Research Question 3. Show more