Manish Mohanpurkar

Dynamic modeling of adjustable-speed pumped storage hydropower plant

Hydropower is the largest producer of renewable energy in the U.S. More than 60% of the total renewable generation comes from hydropower. There is also approximately 22 GW of pumped storage hydropower (PSH). Conventional PSH uses a synchronous generator, and thus the rotational speed is constant at synchronous speed. This work details a hydrodynamic model and generator/power converter dynamic model. The optimization of the hydrodynamic model is executed by the hydro-turbine controller, and the electrical output real/reactive power is controlled by the power converter. All essential controllers to perform grid-interface functions and provide ancillary services are included in the model.

Real time optimal control of supercapacitor operation for frequency response

Supercapacitors are finding wider applications in modern power systems due to a controllable fast dynamic response. Use of power electronically interfaced supercapacitors for frequency support is a proven technique. In practical applications the heat generated from the Equivalent Series Resistance (ESR) can significantly reduce the life of supercapacitor. Hence thermal issues must be addressed for optimal operation. It is infeasible to use traditional optimization control methods to mitigate the impacts of frequent cycling due to the lack of active thermal management. This paper proposes a Front End Controller (FEC) using Generalized Predictive Control based on real time receding optimization. The constraints in the optimization are based on thermal management to enhance the efficiency of utilization life of the supercapacitors. A rigorous mathematical derivation is provided and supporting simulation results are obtained using Real Time Digital Simulator to demonstrate the effectiveness of the proposed technique.

Potential for Plug-In Electric Vehicles to provide grid support services

Since the introduction of Plug-in Electric Vehicles (PEVs), scientists have proposed leveraging PEV battery packs as distributed energy resources for the electric grid. PEV charging can be controlled not only to provide energy for transportation but also to provide grid services and to facilitate the integration of renewable energy generation. With renewable generation increasing at an unprecedented rate, most of which is non-dispatchable and intermittent, the concept of using PEVs as controllable loads is appealing to electric utilities. If incentivized suitably, this could serve as an additional driver for PEV adoption. It has been widely proposed that PEVs can provide valuable grid services, such as load shifting to provide voltage and frequency regulation. The objective this work is to address the degree to which PEVs can provide grid services and mutually benefit the electric utilities, PEV owners, and auto manufacturers.

Empirical study of simulation fidelity in geographically distributed real-time simulations

Interconnection of digital real-time simulators over wide-area communication networks is an innovative approach to extend local real-time simulation capabilities to enable large-scale simulations. Furthermore, it allows the integration of geographically distributed assets as Power Hardware-in-the-Loop and Controller Hardware-in-the-Loop, thus providing a flexible framework for performing unique research experiments. In most cases, it is not possible to perform large scale real-time simulations and comprehensive experiments locally due to lack of simulation capacity and unavailability of unique assets. Main challenge associated with geographically distributed real-time simulation is to ensure simulation fidelity of the same degree as in the case when the entire simulation is performed at the same location. Simulation fidelity in geographically distributed real-time simulation is investigated and an empirical characterization is provided in this paper. Fidelity degradation caused by different values of time delay and sending rate of data exchange between two digital real-time simulators is presented. Two methods for representation of interface quantities in co-simulation interface algorithms are considered for performing simulations. The first method is based on representation of interface quantities as root mean square of magnitude, frequency, and phase angle of the current and voltage waveforms. The second method utilizes representation of interface quantities in form of time-varying Fourier coefficients, known as dynamic phasors. The empirical study is performed for transmission-distribution co-simulation using two racks of Real-Time Digital Simulator (RTDS®).