Technical and Economic Models for Distributed Energy Resources Integration

Abstract

The increase in distributed energy resources (DERs) in power systems presents society with both opportunities and challenges. On the one hand, DERs tend to reduce carbon emissions, introduce competition, and increase the flexibility to power systems. On the other hand, DERs fundamentally disrupt the technical and economic nature of power systems. As a result, they raise challenges such as less predictable load, more complex control of resources, disruption of power flow patterns, and obsolescence of market designs. This dissertation addresses five of those challenges: i) the need for models of building load flexibility and ii) models of interaction between DERs and their aggregator, iii) the problem of coordination among DERs, iv) the risk of market power abuse, and v) the long-term DERs planning problem. Regarding i), we present a data-driven and statistically robust model that describes the flexibility of thermostatically controlled loads in buildings. The model's simplicity makes it suitable for a wide-range of power system applications such as economic dispatch, optimal power flow, and reserve allocation problems. Regarding ii), we present an interaction model for DERs and their aggregator that predicts their long-term equilibrium. Predicting the eventual equilibrium is valuable because it predicts the aggregator's behavior in the electricity market and profit allocation among players. Regarding iii), we present a mixed-integer linear program adaptation of the Datzing-Wolfe decomposition algorithm for decentralized coordination of a building and a fleet of electric vehicles (EVs). This algorithm is suitable for buildings and EVs whose operation is coupled (e.g., by common infrastructure) and cannot formulate a joint problem due to data privacy concerns or software impediments. Regarding iv), we present a pricing mechanism that mitigates the market power of a generic firm (e.g., a generator, a demand-side bidder, or an aggregator of DERs). The pricing mechanism is attractive because it incentivizes socially optimal bids by the firm, requires no private information to be formulated, and provides an instrument for the regulation of the firm's profit. Finally, regarding v), we present a DER planning problem for deferral of capacity expansion (i.e., a non-wire alternatives planning problem) and an algorithm to solve it. Our contributions in this domain are twofold. We explicitly model deferral of capacity expansion an additional value stream of DERs and provide a scalable and tractable algorithm to solve this non-convex, large-scale problem.