[Paper Review] Towards a systematic characterization of the potential of demand side management

David Kleinhans|arXiv (Cornell University)|Jan 8, 2014

Smart Grid Energy Management18 references16 citations

TL;DR

This paper proposes a time-resolved, storage-equivalent framework for systematically characterizing demand side management (DSM) potential by modeling load shifting capabilities of individual devices and sectors as equivalent energy buffers with time-dependent size and capacity. The method enables direct integration into large-scale energy system models by abstracting operational complexities, offering a standardized, comparable metric for DSM that mirrors energy storage characteristics while accounting for real-world constraints like scheduling and shifting windows.

ABSTRACT

With an increasing share of electric energy produced from non-dispatchable renewable sources both energy storage and demand side management might gain tremendously in importance. While there has been significant progress in general properties and technologies of energy storage, the systematic characterization of features particular to demand side management such as its intermittent, time-dependent potential seems to be lagging behind. As a consequence, the development of efficient and sustainable strategies for demand side management and its integration into large-scale energy system models are impeded. This work introduces a novel framework for a systematic time-resolved characterization of the potential for demand side management. It is based on the specification of individual devices both with respect to their scheduled demand and their potential of load shifting. On larger scales sector-specific profiles can straightforwardly be taken into account. The potential for demand side management is then specified in terms of size and capacity of an equivalent storage device. This intermediate layer of abstraction isolates the gross effects and opportunities of demand side management from questions concerning the operation of individual devices, which eases the integration of the results in existing schemes for the commitment of production units and system models. The novel framework is applicable for conceptual or predictive approaches and for real-time optimization and operation of energy systems involving demand side management.

Motivation & Objective

• To address the lack of systematic, time-resolved characterization of demand side management (DSM) potential in large-scale energy system modeling.
• To develop a method that abstracts DSM from individual device control, focusing on aggregate buffer size and capacity for use in system-level simulations.
• To enable direct comparison between DSM and energy storage by expressing DSM potential in equivalent storage terms (size and capacity).
• To support conceptual, predictive, and real-time optimization in energy systems by providing a scalable, intermediate abstraction layer.
• To formalize DSM potential in a way that accounts for time-dependent constraints such as load scheduling and shifting windows.

Proposed method

• The framework models individual devices based on their scheduled load profiles and their potential for load shifting, defined by a time window Δtc.
• It introduces time-continuous load functions Lc(t) and load capacity functions Λc(t) to represent demand and flexibility.
• The maximum and minimum buffer sizes for DSM are derived using integrals over time, estimating the energy that can be stored or shifted forward/backward.
• For discrete data, the method uses piecewise-constant approximations of load and capacity, replacing integrals with direct summation over time intervals.
• The resulting buffer size envelopes Emax(t) and Emin(t) are computed using piecewise conditions based on time overlaps between shifting windows and load periods.
• The approach abstracts away device-level control and rebound effects, focusing on gross system-level potential for integration into unit commitment and system models.

Experimental results

Research questions

• RQ1 How can the time-resolved potential of demand side management be systematically characterized in a way that enables integration into large-scale energy system models?
• RQ2 What is the equivalent storage capacity and size of demand side management actions when considering load shifting constraints and scheduling?
• RQ3 How can DSM be formalized as a storage-equivalent buffer with time-dependent size and capacity to allow direct comparison with energy storage technologies?
• RQ4 What is the impact of discrete time data on the estimation of DSM buffer envelopes in practical applications?
• RQ5 How can class 2 DSM actions (which affect long-term energy balance) be generalized within the proposed framework?

Key findings

• The proposed framework enables a systematic, time-resolved characterization of DSM potential by modeling it as an equivalent energy buffer with time-dependent size and capacity.
• The method provides a standardized abstraction layer that isolates DSM potential from device-level operational complexities, facilitating integration into existing energy system models.
• Buffer size envelopes Emax(t) and Emin(t) are derived using piecewise functions that account for shifting windows and load scheduling, allowing accurate estimation without continuous integration.
• For discrete time data, the framework replaces integrals with direct summation over time intervals, improving computational efficiency and practical applicability.
• The approach supports both class 1 DSM (no long-term energy imbalance) and can be extended to class 2 DSM by relaxing the energy balance constraint.
• The resulting storage-equivalent metrics allow direct comparison between DSM and energy storage in terms of size and capacity, enhancing system-level planning and optimization.