Multi-State Load Models for Distribution System Analysis

Abstract

Recent work in the field of distribution system analysis has shown that the traditional method of peak load analysis is not adequate for the evaluation of emerging distribution system technologies. Voltage optimization, demand response, electric vehicle charging, and energy storage are examples of technologies with characteristics having daily, seasonal, and/or annual variations. In addition to the seasonal variations, emerging technologies such as demand response and plug-in electric vehicle charging have the potential to receive control signals that affects their energy consumption. To support time-series analysis over different time frames and to incorporate potential control signal inputs, detailed end-use load models that accurately represent loads under various conditions, and not just during the peak load period, are necessary. This paper will build on previous end-use load modeling work and outline the methods of general multi-state load models for distribution system analysis.

Modeling of GE Appliances in GridLAB-D: Peak Demand Reduction

“…This allows users to develop and simulate models which more accurately predict the behavior of loads, and how they interact with the power system, including voltage and temperature dependencies, power system and load control functions, and the complex interactions that occur between devices in such an interconnected system. Previous work has highlighted the advantages of developing and simulating load models as multi-state load models and applying these models in a single simulation environment that can represent loads, electrical infrastructure, and energy markets in a single solution [2]- [7]. This work will build upon those efforts by developing load models which represent the GE DR-enabled appliances as multi-state models with voltage-dependent behavior.…”

“…Because a cooling system operates to maintain internal air temperature within a band, parameters such as near term history of operation, time of year, outside air temperature, building construction, and terminal voltage will impact the instantaneous power consumption, as well as the energy consumption. To examine these issues, a physical model of the cooling system and the structure of the building is constructed using an equivalent thermal parameter (ETP) model [2]. Because the ETP model has been shown to be an accurate representation of residential and small commercial building instantaneous power draw, as well as energy consumption, it will be used for the formulation of the physical model.…”

“…V T is the terminal voltage of the system compressor motor. Additionally, it should be noted that Q hvac is expressed in terms of British thermal units (Btu) consistent with the conventions of the heating/cooling industry in the United States and the derivation of the ETP model of [2], while S losses is expressed in SI units. As a result, the two terms of (C-5) must be converted using the conversion of 1.0 Btu/h = 0.2931 W.…”

“…The cooling system's heat removal rate, Q hvac , can be solved using heat transfer equations based on the available mechanical torque of the compressor [2]. When (C-5) is implemented in a time series simulation, the result is a model that determines the energy consumption, both real and reactive, of the cooling system as a function of the outside air temperature, the inside air temperature, equipment parameters, terminal voltage, and occupant-controlled setpoint.…”

“…Equation (C-4) is the second order differential equation that describes the heat flows shown in Figure C-2 [2]. Its solution determines the time-varying temperature of the house, both air and mass, given the thermal inputs.…”