Candidate: Al-Baraa El-Hag
Title: Enhancing System Reliability Utilizing Private Electric Vehicle Parking Lots Accounting for the Uncertainties of Renewables
Date: September 11, 2019
Time: 3:00PM
Place: EIT 3151-3153
Supervisor(s): El-Shatshat, Ramadan A.
Abstract:
Integration of renewable energy sources into electric grids comes with significant challenges. The produced energy from renewable sources such as wind and solar is intermittent, non-dispatchable and uncertain. The uncertainty in the forecasted renewable energy will consequently impact the accuracy of the forecasted generation. That, in turn, will increase the difficulties for the grid operators to meet the demand-supply balance in the grids. Moreover, the shift to electric vehicles (EV) also adds complications to grid operators planning since their demand profiles are unlike anything that is currently connected to the grid.
With the advent of the smart grid, many new interesting and practical technologies will become a reality. Unfortunately, most of these elements will not be physically realized in systems for the immediate future. This thesis will maintain an overarching constraint of only using aspects of the smart grid that can be implemented, given today’s infrastructure, within the next three to five years. For that reason, all loads connected to the system are considered to be uncontrollable except EVs when they are connected to a “commercial parking lot”. The goal of this body of work is to investigate the benefits of the utility providing incentives, in terms of reducing the price of electricity for charging EVs through a commercial parking lot, for the sole goal of enhancing reliability through optimized scheduling of charging time periods. Moreover, since the penetration of renewable energy is only predicted to increase over time, their impact will also be investigated.
Since both EVs and renewables add uncertainty and randomness whenever connected, these elements need to be accurately and adequately modelled. There will be five electrical components that will need stochastic models built. The first two, base electric load and dispatchable/distributed generators, will be modelled based on the IEEE - Reliability Test System (IEEE-RTS), with the later using a Monte-Carlo (MC) simulation. The next two, solar and wind energy, will be extrapolated from historical weather patterns through a Markov-Chain Monte-Carlo (MCMC) simulation. Finally, the EV will be virtually generated from historical commercial parking lot data also using a MCMC.
Three scheduling algorithms were implemented in this work. The first is a base case, in which the EVs charged in a first-come first-serve basis, this situation that would arise if no information at all is shared with the parking lot owner. The results of the simulation had the value of Expected Energy Not Served (EENS) came out to be 38.05 MWh/year (the lower the better). With the second algorithm, a basic Demand Side Management (DSM) algorithm was implemented, with the result being that the EENS decreased by 8.35 %. The information shared was the demand shape of all consumers for a given day. Lastly, an algorithm that will be called Grid and Demand Side Management (GDSM) by this thesis proved to be even more successful, having the EENS reach a value of 29.85 MWh/year, a decrease of 21.55 %. The GDSM scheduling algorithm needs the grid to share not only the consumer behavior but also the expected generator behavior. Based on these results, recommendations are made to the electric utility in terms of the benefits it may reap if it expands and develops a communication infrastructure that includes information of generator availability.