Price, Cooling and Heating Demand Prediction Tools (Consultancy)
My research aims at developing cost-effective solutions and algorithms to pave the way to 100% renewable generation in the future grid. In that direction, our research is focused on demand flexibility exploitation and storage systems.
As the energy industry decided to exploit clean and sustainable resources, concerns grew over the inherent unpredictable variability in renewable resources. To address these concerns, my research over the last eight years primarily focused on achieving one goal: to efficiently and securely utilize high penetration of renewable resources in the power system while avoiding to jeopardize of the integrity of the system as a whole.
I started my research in power systems by developing models and control algorithms for renewable energy resources and their interconnection to the power system. It was the beginning of a long journey that has continued until today in my position as a Lecturer at the University of Adelaide, Australia.
While working on my M.Sc. thesis, I developed multiple intelligent techniques for very short-term wind speed prediction to start changing the blade’s pitch angle ahead of time in order to maximize extracted energy. An ultracapacitor with four-quadrant DC choppers was also utilized on the DC side to suppress a portion of variations in the output power. I also developed a dynamic model of the power system and multiple components, such as wind turbine and generator, and power electronic interfaces to complete the study. In collaboration with my colleagues in the lab, I worked on optimal sizing and placement of renewable resources in distribution feeders.
Over the course of my Ph.D., I worked on frequency and voltage regulation through optimal power management and real-time demand response. I proposed multiple controllers for responsive loads (specifically Electric Water Heaters) for islanded microgrid as well as bulk power system at the transmission level to regulate frequency in presence of large renewable generation. I also integrated storage in the optimal power management system for islanded microgrid.
At NEC Laboratories America (NECLA), my research revolved around Li-Ion batteries. I started with characterization and degradation modeling of Li-Ion batteries for battery energy management system. Then, I integrated the model in a post-processing module to avoid unnecessary battery degradation. Later, I developed a resilient battery charging algorithm within the energy management system. As my last research efforts, I worked on developing behind-the-meter (BTM) aggregation model for participation in the wholesale energy and ancillary services markets based on Transactive Energy (TE) concept.
I worked at DTU Compute as a postdoc fellow for about three months. I completed writing a grant proposal in collaboration with a European consortium led by NTNU. The proposal was submitted to the Horizon 2020 LCE-05-2017 call. I led a workpackage development in collaboration with the University of Newcastle, Aalborg University, TU-Berlin, and DIW-Berlin. In addition, I was working on two workpackages from SmartNet project.
I moved to Australia in 2017 to work as a research fellow at the Global Change Institute (GCI) and then the school of Information Technology and Electrical Engineering (ITEE) at the University of Queensland (UQ). I worked on one of the largest solar PV research facility in the world with 3.325 MWp generation in Gatton campus. The plant was hosting a 600 kW/760 kWh Li-polymer battery storage for peak shaving and voltage control. My research mostly will focus on the storage and its application for the integrated system.
In June 2019, I joined the University of Adelaide (UofA) as a Lecturer in the School of Electrical and Electronic Engineering (EEE). My research are focused on battery storage systems and market mechanisms, which are explained in more details under "Current Research Projects."
Price, Cooling and Heating Demand Prediction Tools (Consultancy)
Assessment, design and operation of battery-supported electric mining vehicles and machinery
3 PhD positions supported by the University of Adelaide in partnership with Watts, Denmark
Distributed Energy Resources sizing validation for ARENA grant (Consultancy)
School of Electrical and Electronic Engineering, the University of Adelaide
School of Electrical and Electronic Engineering, the University of Adelaide
I believe that impactful research in power systems engineering would be impossible without borrowing knowledge and expertise from other disciplines. I also happened to believe that one researcher cannot (and will not) be expert in more than one or two areas of research due to the amount of knowledge produced these days. Moreover, collaborative research provides a chance to disseminate the results faster on a global scale. Therefore, I advocate and practice multi-disciplinary and collaborative research in full capacity.
Over the years, I had the privilege to collaborate with many researchers and engineers in industry and academia alike. These collaborations resulted in publications with co-authors from 10 countries and 20 institutions until early 2019. My collaborators are mainly expert in the disciplines in which I lack knowledge to make useful research. They fill the gaps in my research by their contributions. Since the landscape of my research changes over time, new tools and methods are demanded, which needs expertise from new collaborators. So, I am always keeping an open eye for new collaborations that expand and deepen my research as well as my collaborators.
You can see a list of my previous and current collaborators on the left and some basic information about them. This also gives more insights into my research interests.
In the last nine years, my research seamlessly was focused on promoting and development of smart grid technologies and concepts. While working on my M.Sc. thesis, I developed multiple intelligent techniques (based on artificial intelligence, Markov chain modeling, and wavelet networks) for very short-term wind speed prediction for real-time wind turbine control applications. The proposed techniques could provide a time margin for the wind turbine controller to start changing the blade’s pitch angle aiming to maximize extracted energy and to partially prevent abrupt swings in the generated power. The outcome of my researches on this very topic have been published in two journal papers and three proceedings of international conferences.
Over the course of my Ph.D. research, I primarily focused on the application of demand response (DR) in the distribution systems as well as transmission level to provide voltage/frequency support and balancing reserve. I started my research by developing multiple simulation frameworks for grid-tied and islanded microgrids to simulate the frequency/voltage behavior of the system under different level of renewables generation. Then, I proposed control methodologies to provide frequency regulation by compensating variability of renewable generation and load demand. At the transmission level, my contribution was twofold: The first part was introducing dynamic DR control loop in the traditional load-frequency control (LFC) model which was accompanied by extensive analytical evaluation and modeling for single-area and then multi-area power systems. In the next step, I designed and assessed application of aggregated electric water heaters (EWH) for providing balancing reserve at the transmission level with high wind generation. I created individual and aggregated EWH model based on actual hot water demand data. Then, I proposed an innovative thermostat setpoint control for EWHs to participate in balancing reserve services and assessed the economic impact of participation on the consumers.
Power system modeling and simulation was the core of many courses in my undergrad education which developed a great deal of interests in me. For that reason, I selected to work on implementation of dynamic model of DC motor in MATLAB/Simulink with three speed control techniques as my final project in undergrad. Since then, modeling of power system and its component became a part of my research at different levels. Over the course of my M.S. education, I developed dynamic model of wind turbine with pitch control, self-excited induction generator with capacitor banks, super-capacitor with four-quadrant DC chopper as a controller for smoothing the output power, and back-to-back converter connected to the grid. I also implemented a thermo-electrical model of PV with maximum power point tracker (MPPT) by effective control of DC-DC booster and 3-phase inverter controlled by PWM signals in connection to the grid.
During my Ph.D., I extended my knowledge of power system modeling by developing dynamic model of individual and aggregated EWH model, thermo-electrical model of valve-regulated lead-acid (VRLA) batteries and electrical model of Li-Ion batteries, IEEE 13-bus standard distribution system model for dynamic and stability analysis, a model of the upper grid to resemble weak and strong power system in grid-tied microgrid studies, proton-exchange membrane fuel cell (PEMFC), and solid-oxide fuel cell (SOFC). I developed these models in MATLAB®/Simulink and m-files, and PSCAD which have been used in numerous research studies.
Designing grid-tied and islanded microgrids by technology selection and sizing, and developing power/energy management was a part of my Ph.D. dissertation which is continued until today. I carried out MG design studies whenever a MG was required for my research. To do that, I gained experience by working with sizing packages such as HOMER and WebOpt in different occasions. My research on power/energy management algorithm at MSU resulted in publishing two journal papers. These management algorithms were specifically designed for islanded operation. In the latest research at MSU, I proposed a multi-timescale power management framework considering battery storage and dynamic DR using EWHs with detailed model of each component.
At NEC Laboratories America (NECLA), where I am currently appointed as a researcher, I continued working on this area by developing energy management system for grid-tied and islanded microgrids equipped with renewables, diesel generator, and battery storage for time-of-use (ToU) and demand charge (DC) management for a day-ahead. We developed very sophisticated and industry-standard EMS which is currently going through prototyping and final evaluations before getting to product line. We are closely working with one of the largest Li-Ion battery provider in the world (NEC Energy Solution). My research in this area has resulted in one granted U.S. patent application, which has been implemented in a telecommunication tower in remote areas of Malaysia and as a part of a larger EMS and DR in a 7-eleven store in Indonesia for prototyping.
In the last two years, my primary focus at NECLA was on Li-Ion batteries for grid-scale energy storage and most recently for behind-the-meter applications, by working with our industrial partner (i.e., NEC Energy Solutions). I started my research by developing a comprehensive degradation algorithm for Li-Ion battery by conducting statistical analysis to find the most significant input parameters in the degradation process, and then using artificial neural network (ANN) to develop the model. I developed individual models for cyclic and calendar aging of the battery, which later has been combined in a universal framework for integration of the both processes based on daily operation profile. I have filed one patent application with U.S. Patent office for the individual degradation models, and another patent application under preparation by the legal team at NECLA on the integration framework. This effort has been recognized by a Spot Recognition Award at NECLA in Nov. 2015. I also recently presented my paper on the integration algorithm at the IEEE PES ISGT-North America conference in Minneapolis, MN. In parallel with my research on Li-Ion battery, I participated in battery optimal sizing studies for different applications such as an islanded microgrid in Australia belonging to a mining company (to reduce diesel fuel consumption), a gigantic banana plant in Indonesia (to be independent from upper grid), and a utility company in Lithuania (for frequency regulation in 15- second interval). I was partially responsible for problem formulation, implementation in GLPK package in C++, problem size reduction, and integrating daily operation and battery degradation in sizing study which was very critical for our customers. Later, I worked with Gurobi to change the problem formulation from a LP to MILP since our customer had several diesel engine which unit commitment became inevitable in the formulation. Since Li-Ion battery degradation is a nonlinear process, I tried SNOPT as a nonlinear optimization solver to integrate the battery degradation directly in the optimization formulation. Through these real-world research studies, I gained an in-depth knowledge of classical optimization techniques which can conveniently be utilized in any planning study with optimization nature.
As another attempt to improve the battery operation by avoiding unnecessary degradation and selfdischarging,I proposed and implemented a resilient battery charging algorithm. I have filed a patent application on this innovation which is under preparation for lodging with U.S. patent office by the legal team at NECLA. In collaboration with another researcher and a summer intern, I participated in an a research study to develop a battery sizing algorithm for BTM application (typically for Commercial/Industrial consumers) considering two different DR offered by the PG&E utility in Northern California. The proposed algorithm also considered ToU and DC management while accounting for the battery degradation impact. This work has been presented in the IEEE PES ISGT-North America recently.
Most recently, I hired an intern to work with me on developing a market structure at the distribution level to create extra revenue for BTM storage owners by participating in wholesale energy market. A two-layer controller based on incentive signal from aggregator to the individual customers is designed and uncertainty analyses are carried out. The proposed market structure and two-layer controller is developed based on Transactive Energy (TE) concept. An intellectual property application is filed on this research and the first draft of a paper is completed for submission to the next IEEE PES General Meeting. On the same topic, I started a collaboration with a Postdoc fellow at the University of Manchester to develop a market structure at the distribution level for multiple home-microgrid. The proposed market mechanism fulfills TE requirements and proposes a game theory formulation for market participants. The paper is almost ready for submission to the IEEE Transaction on Smart Grid. I have so many ideas on this new topic which I believe a position like this one at the Monash University can help me to pursue.
My research efforts reflected on the first topic is closely related to this topic as well. On top of that, my research experience have some components which can be directly related to this topic. Over the course of M.Sc. education, I collaborated with lab-mates on developing optimal siting and sizing algorithms at the distribution level. The algorithm was intended to find optimal size and location of distributed generation resources to minimize losses and increase short-circuit level. Later during my Ph.D. research, I developed a comprehensive model of a distribution system with distributed PV generation to assess the impact of renewable variations on voltage and frequency profiles. I also included a moving cloud functionality in the simulation to take the model one step closer to reality.
Cost-based operation of distributed generation, including their ownership costs, is an integral part of the smart grid concept. So, accurate operational cost calculation plays an important role for profitable operation of DERs for their owners. I worked on an idea in this regard which is almost completed and a paper is almost written on it. In this research, I developed multiple new cost ownership calculation models by considering uncertainty and risk analyses.