Abstract
In the traditional power system demand response, customers respond to electricity price or incentive and change their original power consumption pattern accordingly to gain additional benefits. With the development of multi-energy systems (MES) in which electricity, heat, natural gas and other forms of energy are coupled with each other, all types of energy customers are able to participate in demand response, leading to the concept of integrated demand response (IDR). In IDR, energy consumers can response not only by reducing energy consumption or opting for off-peak energy consumption but also by changing the type of the consumed energy. Taking the traditional demand response in power system as a starting point, the studies of the fundamental theory, framework design and potential estimation of demand response in power system are reviewed, and the practical cases and software development of demand response are introduced. Finally, the current theoretical research and application of IDR are assessed.
Introduction
The landscape of global energy management is undergoing a fundamental shift. Traditionally, Demand Response (DR) has been a cornerstone of power system stability, encouraging consumers to modify their electricity usage in response to price signals or incentives. However, as we move toward Multi-Energy Systems (MES)—where electricity, heating, cooling, and natural gas networks are deeply interconnected—the "single-commodity" focus of traditional DR is becoming insufficient. This paper reviews the evolution from electricity-centric DR to Integrated Demand Response (IDR), a paradigm that leverages the coupling of diverse energy forms to enhance grid flexibility and economic efficiency.
The Evolution of Demand Response
Traditional demand response research has matured over the last decade, focusing on three primary pillars: fundamental theory and framework design, potential estimation, and practical implementation. Potential estimation often utilizes data mining and clustering algorithms (such as K-means or Support Vector Clustering) to identify typical consumption patterns and quantify shiftable loads.
As illustrated in the framework above, the research trajectory has moved from isolated power system optimizations toward a holistic "Integrated Demand Response" model. This transition is driven by the need to coordinate systems with vastly different physical properties, such as the instantaneous nature of electricity versus the slow-moving thermal dynamics of heating networks.
Core Concepts of Integrated Demand Response (IDR)
The defining characteristic of IDR is energy substitution. While traditional DR focuses on when energy is consumed (load shifting), IDR introduces the dimension of which energy is consumed. For instance, if electricity prices spike, an IDR-enabled facility might switch its cooling load from an electric chiller to a gas-fired absorption chiller.
A critical technical enabler for this is the Energy Hub (EH) concept. The EH treats a building or district as a single unit with multiple energy inputs and outputs, using conversion and storage technologies to optimize flows. Furthermore, IDR leverages the "thermal inertia" of buildings. Because heat loads change slowly, heating systems can act as "virtual batteries," buffering volatility in the power grid without impacting occupant comfort.
Methodologies and Optimization
Coordinating these complex systems requires advanced mathematical tools. The paper highlights several key approaches:
- Optimization Techniques: Mixed Integer Linear Programming (MILP) and Model Predictive Control (MPC) are used to manage the timing and type of energy flows.
- Game Theory: Nash Equilibrium models are applied to balance the competing interests of energy suppliers and consumers in a multi-market environment.
- Automation: The shift from manual response to "plug-and-play" IDR terminals allows for sub-second decision-making, essential for maintaining transient balance in grids with high renewable penetration.
Practical Applications and Results
The theoretical benefits of IDR are supported by significant real-world data:
- Economic Impact: In the PJM Interconnection (USA), DR capacity accounts for 6% of peak load and saved customers approximately $11 billion in a single year.
- Efficiency Gains: Demonstrations like the Shanghai Mansion project showed energy efficiency improvements of 22%, significantly reducing annual operating costs.
- Response Speed: Automated systems in industrial parks, such as those in Jiangsu, China, can reduce demand by over 500,000 kW in just one second.
- Regional Markets: Analysis of European markets indicates that DR has reduced peak loads by an average of 2.9% across eight major regions.
Challenges and Future Prospects
Despite its potential, several hurdles remain for widespread IDR adoption. Current models often use simplified constraints that do not fully capture the physical nuances of gas and heat networks. There is also a pressing need for unified pricing mechanisms that span multiple energy sectors to break down existing market silos.
Future research must focus on high-resolution modeling for non-electric energy carriers, hardware standardization for "plug-and-play" terminals, and robust uncertainty management to handle the stochastic volatility of weather-dependent renewable energy sources. By breaking the boundaries between energy sectors, IDR offers a much larger optimization space, leading to superior economic and environmental outcomes.
Summary updated