Project Background

Load modeling has significant impacts on studies of planning, operation and control of both distribution and transmission networks. Although many efforts have been devoted to load modeling, it is still challenging work because of the uncertainty, complexity and time variability of loads, lack of data, the implementation of demand response (DR) and increasing integration of distributed generators (DGs) and power electronic loads. Studies have shown that inaccurate representation of system loads usually leads to discrepancies between the simulated and recorded power grid responses, which may send misleading signals to system operators. However, most of the load models currently used were developed a few decades ago and are not adequately updated after subsequent changes in load structure and characteristics.

Scientific Opportunities

We developed a hierarchical load modeling structure to build time-varying, stochastic, customer behavior-driven and DR-enabled load models by leveraging practical utility data and laboratory experiments. The proposed framework solves the challenges that have not been fully addressed in previous load modeling efforts: representation of time-variant, geographically-dependent, and stochastic load behaviors; integration of DGs and power electronic loads; consideration of uncertain customer behaviors, multiple DR programs, physics and economics; and application of PMU, SCADA and AMI data for real-time load model identification and calibration.

Research Goal

The major benefits of this project to industry are significant enhancements in understanding static and dynamic characteristics of individual and aggregated loads/DGs, thereby improving the situational awareness, planning and analysis of power systems. This project aims to provide utilities the capability to analyze customer behavior-driven and demand response-enabled load characteristics industry can use the developed load models with recommended parameter ranges as well as the updated commercial software tools with new models to facilitate the analysis of power system transients, power system static/dynamic stabilities, load forecasting, demand response and conservation voltage reduction, and long-term planning issues such as transmission planning. By leveraging the proposed models, grid operators can operate the system within its capability without imposing unnecessary restrictions.

Research Plan

Load modeling in this project has been performed at four levels: 1) component level, 2) customer level, 3) feeder level and 4) substation level. Compared to existing load models and load modeling techniques, the hierarchical load modeling framework in this project has the following salient features:

• It leverages PMU, SCADA, and AMI data, as well as laboratory experiments to develop load models at component, customer, feeder, and substation levels.
• It provides equivalent models of networks with a significant amount of DGs and power-electronic loads.
• It considers the stochasticity, time-variability, flexibility, and controllability of loads.
• It considers customers’ behaviors and the price sensitivities of loads.
• It provides reasonable load composition and aggregation through data-driven methods.
• It develops new model identification algorithms that are robust to measurement noises and bad data.
• It leverages WECC and NERC models and provides the analysis approach accordingly.

Deliverables and Impacts

List of Deliverables:

1.  A set of static and dynamic load models at component, customer, feeder and substation levels, which are generic and applicable to various practical systems.
2.  A set of customer behavior-driven and demand response-enabled load models at component, customer, feeder and substation levels, which are generic and applicable to various practical systems.
3.  A set of load model identification techniques, which are robust to measurement noises and bad data as well as suitable for on-line identification of model parameters using PMU, SCADA and AMI measurements.
4.  Recommendations on typical load model parameter values, ranges and probabilistic distributions.
5.  A set of commercially available software tools with developed load models, which include PSS/E at transmission level, CYME at distribution level, and RTDS/OPAL-RT at customer and component levels as well as practical testing results with collaborating utilities.
6. Technical reports and journal papers with detailed descriptions of load models, assumptions/limitations, laboratory/utility data tests, demonstrations with commercially-available software tools.

List of Publications:

1. Dongbo Zhao, Qian Ge, et al, ​“Short-Term Load Demand Forecasting through Rich Features Based on Recurrent Neural Networks,” IET Generation, Transmission & Distribution, vol. 15, no. 5, pp. 927 – 937, March 2021
2. Zhaoyuan Fang, Dongbo Zhao, et al, ​“Non-Intrusive Appliance Identification with Appliance-Specific Networks,” IEEE Transactions on Industry Applications, vol. 56, no. 4, pp. 3443-3452, July 2020
3. Zixiao Ma, Zhaoyu Wang, et al, ​“High-Fidelity Large-Signal Order Reduction Approach for Composite Load Model,” IET Generation, Transmission & Distribution, vol. 14, no. 21, pp. 4888 – 4897, Nov. 2020
4. Bo Wang, Dongbo Zhao, et al, ​“Aggregated Electric Vehicle Load Modeling in Large-Scale Electric Power Systems,” IEEE Transactions on Industry Applications, vol. 56, no. 5, pp. 5796-5810, Sept. 2020
5. Tianqi Hong, Dongbo Zhao, et al, ​“Optimal Voltage Reference Setting for Droop based DER in Distribution Systems,” IEEE Transactions on Smart Grid, vol. 11, no. 3, pp. 2357-2366, May 2020.
6. Bo Wang, Payman Dehghanian and Dongbo Zhao, ​“Chance-Constrained Energy Management System for Power Systems with Renewables and Electric Vehicles,” IEEE Transactions on Smart Grid, vol. 11, no. 3, pp. 2324-2336, May 2020.
7. Chong Wang, Zhaoyu Wang, et al, ​“Robust Time-Varying Parameter Identification for Composite Load Modeling,” IEEE Transactions on Smart Grid, vol. 10, no. 1, pp. 967-979, Jan. 2019
8. Chong Wang, Zhaoyu Wang, et al, ​“SVM-Based Parameter Identification for Composite ZIP and Electronic Load Modeling,” IEEE Transactions on Power Systems, vol. 34, no. 1, pp. 182-193, Jan. 2019.
9. Anmar Arif, Zhaoyu Wanget al, ​“Load Modeling – A Review,” IEEE Transactions on Smart Grid, vol. 9, no. 6, pp. 5986 – 5999, November 2018
10. Zhaoyuan Fang, Dongbo Zhao, et al, ​“Non-Intrusive Appliance Identification with Appliance-Specific Networks,” IEEE Industrial Application Society Annual Meeting, Baltimore, MD, 2019

The developed load models will bring about major benefits to power industry entities, including utility companies, consumers, ISOs/RTOs, regulatory agencies, investors, and vendors/consultants.

Team and contact

Argonne National Laboratory (Lead)
National Renewable Energy Laboratory
Iowa State University
Siemens

Lab Lead Team Members
Dr. Yuting Tian (Postdoc)
Dr. Tianqi Hong (Postdoc)
Mr. Bo Wang (Intern)
Dr. Qian Ge (Intern)
Mr. Rahul Jha (Intern)
Mr. Gonzalo Constante (Intern)
Ms. Dan Lu (Intern)
Mr. Yiyun Yao (Intern)
Mr. Andy Fang (Intern)

Project PI: Dr. Dongbo Zhao
Principal Energy Systems Scientist
Energy Systems and Infrastructure Analysis
dongbo.​zhao@​anl.​gov
630-252-4307