A methodology for forecasting impact of demand response on capacity expansion planning

Abstract

In parallel to tighter energy regulations and increasing demand for emissions reduction, the Department of Energy (DOE) has set a goal to reduce energy consumption in the building sector to 50% of 2010 levels by 2030. This encourages the use of advanced operational strategies and demand side management concepts to improve energy efficiency and reduce peak energy loads. In response to this societal need, many communities such as cities and university campuses are trying to transform their energy systems into smart and sustainable ones. At the same time, capital planners in these organizations are interested in deferring the need to expand energy supply capacity, e.g., new chillers for district cooling systems, to avoid incurring those costs until farther in the future. A campus level district energy system is in the mixed position of an energy producer and consumer. It consumes the electrical energy to produce the thermal energy for multiple buildings within the community. A district energy plant tends to have the excessive capacity with a redundancy to ensure the system reliability. Advancement in building energy technologies and communication systems enables the concept of Demand Side Management (DSM) to become a reality. Among the DSM concepts, Demand Response (DR) is now the most established practice for the power system management. The established DR can be utilized by the owner of the district energy system to alleviate the oversizing problem and defer the capital investment on new chillers if it can replace a redundant chiller. This thesis proposes a methodology to quantitatively evaluate the impact of DR on a district energy system’s operation and planning. The proposed methodology utilizes measured data, converts them into actionable information by developing models to capture the interaction between demand and supply sides, and provides an insight into the planning design space by connecting the planning optimization and the reliability analysis modules. The methodology consists of seven steps including preparatory steps. Each of steps is developed from research questions seeks for efficient modeling and analysis methods. Data driven cooling demand model discerns the amount of load shed by a specific DR method, and the physics based chiller plant model is approximated as a performance curve for fast assessment of DR. A capacity expansion planning is formulated as a Mixed Integer Linear Programming problem and following simulation based reliability analysis. The proposed methodology is applied to an example system, which is based on the real system data, to demonstrate the capability. Results show that the proposed method can yield the trade-off space between the capital cost and the reliability of the optimal system expansion plans and discover a hidden trend in the capacity planning design space. Comparing to the baseline N+1 system, it shows how long the plant expansion can be deferred and how much cost can be saved from the preferred plan.