Researchers Implement Models for Large-Scale Dynamic Studies of Electric Power Systems

Achievement date: 
2016
Outcome/accomplishment: 

Researchers at the NSF-funded Future Renewable Electric Energy Delivery and Management (FREEDM) Systems Engineering Research Center (ERC), headquartered at North Carolina State University, have implemented and validated models for large-scale dynamic studies of electric power systems in a simulation known as OpenDSS (an electric power Distribution System Simulator).

Impact/benefits: 

This is an original contribution that extends the capability of OpenDSS to perform dynamic simulations on large distribution systems that can capture the power-system dynamics and interactions. With this advance, power-systems researchers will now have a method to analyze systems with large numbers of nodes and be able to expand the size and capability of the Large Scale System Simulation (LSSS) testbed to reflect a representative distribution system. Large-scale simulation is very important because, ultimately, acceptance by utilities of the FREEDM system's transformation of the electric grid will depend heavily on a verified simulation capability that can assure the system will perform as intended.

Explanation/Background: 

This research has three main segments: dynamic-algebraic-equation (DAE)-based modeling of the converter-interfaced generation (CIG), verification of the CIG model, and dynamic simulation of the CIG model in large, unbalanced distribution systems. Dynamic simulation of the CIG model in OpenDSS has been verified by comparing simulation results with the ones from PLECS® (a tool for high-speed simulations of power electronic systems). PLECS is used only to validate the OpenDSS results in small enough systems that PLECS can handle and, once the validation is done, OpenDSS is used solely for the dynamic analysis on large systems.

Detailed dynamic studies involving fault current analysis, microgrid operation with droop control, and anti-islanding tests on the large distribution system have been carried out. Both balanced and unbalanced cases have been studied. The results show the close matching between the PLECS and OpenDSS simulations (see figure).

The basic components of the DAE-based CIG model include filter, current controller, and phase-locked-loop (PLL). The state-space function of the CIG model is written into Pascal to build the DLLs (Dynamic Linked Libraries), which are then interfaced with OpenDSS during simulations. The approach benefits from the usage of both the power electronics tools (PLECS and Simulink) and the extending of OpenDSS from the mere snapshot and quasi- static analysis of the latter. The DLLs of the user-defined models for different CIGs, such as (photo-voltaic) PV inverters and solid-state transformers (SSTs), are created using Embarcadero Delphi.