Energy transition: renewables, networks and electric vehicles (session one)
Angela Simonovska
PhD candidate
The University of Melbourne
Presentation title: Electrical model-free optimal power flow for PV-Rich HV-LV distribution networks
Abstract: The operation of distribution networks with high penetration of residential photovoltaic (PV) systems, requires active management of setpoints of dozens to hundreds of controllable elements to manage customer voltage issues, while catering for network constraints. The conventional AC Optimal Power Flow (OPF) has shown great potential for this purpose, however OPF-techniques require detailed three-phase LV network models which in practice are not readily available for most distribution companies. The growing adoption of smart meters on the other hand, is creating many opportunities for data-driven approaches that could be used for the calculation of setpoints of controllable elements, becoming an alternative for the conventional AC OPF.
This work proposes the optimal calculation of PV setpoints of controllable elements to mitigate voltage problems in a low-voltage (LV) feeder using, instead of power flow equations, a neural network (NN) trained to capture the nonlinear relationships among recent (a few weeks) historical smart meter data (active power, reactive power, voltage) and the corresponding LV feeder. In other words, an electrical model-free OPF that can be used by distribution companies to manage PV-rich high-voltage (HV)-LV distribution networks, without the need to produce detailed network models. Depending on the engineering problem, different objective functions can be adopted, however, for this study, the adopted objective function is the maximization of PV generation, subject to the equations and parameters of the trained NN that are used to calculate voltages within the optimization problem, coupled with thermal limits for key assets.
To assess the performance of the proposed electrical model-free OPF, a realistic three-phase LV feeder with 31 residential (single-phase) customers and 70% of PV penetration, from Victoria, Australia was investigated, using synthetic 5-min data. To account for the effect of the upstream HV network on LV network voltages, an integrated HV-LV network with 3,400+ customers is considered with 20% of PV penetration. To demonstrate the performance of the proposed approach, one representative summer weekday (in January) was selected. The proposed electrical model-free OPF shows up to 4% mismatch in calculating the PV setpoints for the observed PV-rich HV-LV feeder, and 6 times quicker solution time (less than a power flow), with respect to the AC OPF. Therefore, these results indicate the future application and use of the proposed approach in distribution companies to perform active management of setpoints of controllable elements within the operational control cycle and without the need for network models.
Rosalind Archer
Head of School
Engineering & Built Environment, Griffith University
Presentation title: Electric vehicles: experience from across the Tasman
Abstract: New Zealand has had a strong base of renewable energy for decades - positioning it well for electric vehicle uptake. This presentation explores New Zealand's experience with electric vehicles. The growth in electric vehicle imports into New Zealand will be outlined, accompanied by a discussion of the tax and policy interventions that supported electric vehicle adoption. Electric vehicle adoption in New Zealand is growing rapidly with 23% of car sales in August 2022 being fully electric (and 44% of sales that month being at least hybrid models). NZ wants 30% of its entire vehicle fleet to be fully electric by the year 2035 if it is to meet the emissions budget it aims to stay within. To reach this point there is a target for 50% of all light vehicle registrations by 2029 to be electric, and 100% by 2035.
What could Australia learn from New Zealand's experience? Australia is working to decarbonise its electricity generation which could provide a natural boost toward electrification of transport. The number of kilometres driven annually by households in New Zealand and Australia is very similar – so perhaps electric vehicle range should not be deemed to be major factor in the difference between electric vehicle ownership in each country.
What is the experience of owning and driving an electric vehicle like in each country? This presentation will compare the differences in public charging infrastructure in both countries. Home charging is often influenced by the style of residential properties – so this presentation will also compare the mix of property styles (e.g. apartments, duplexes, stand-alone homes in each country). Finally the availability and cost of typical electric vehicle models will be compared.
The set of reflections in this presentation will offer a view on why electric vehicles are few and far between on Australia roads (relative to New Zealand), and what interventions may change that.
Yushan Hou
PhD candidate
The University of Melbourne
Presentation title: EV hosting capacity and voltage unbalance: an Australian case study
Abstract: Electric vehicles (EVs), particularly, passenger light-duty EVs, are becoming increasingly common worldwide. At the same time, as more and more EVs charge from home [1], mostly through single-phase chargers, voltage unbalance can increase in the charging period. This, in turn, may exacerbate voltage drop issues on a certain phase (i.e., for certain customers) and, eventually, limit the ability of the corresponding distribution network to host EVs (also known as hosting capacity). Once a distribution network reaches its EV hosting capacity, no more EVs can be accommodated, unless a solution (such as network reinforcements or EV charging management) is adopted.
This presentation will present the work of investigating how voltage unbalance affects EV hosting capacity and the extent to which a phase balancing solution can be beneficial. Crucially, the effectiveness of adopting statutory limits for Voltage Unbalance Factor (e.g., 2% in the IEEE Standard 1159-2019 [7]) from the perspective of EV hosting capacity will also be discussed. Unbalance is quantified by Voltage Unbalance Factor (VUF), calculated using three-phase voltage magnitudes at the head of the LV feeder [2]. To fully account for the uncertainties of EV location (i.e., which customer has an EV), charger size, and charging behaviour, a Monte Carlo-based approach is proposed. A realistically modelled Australian MV-LV network with over 1,300 customers, is used to demonstrate the proposed approach, and a sensitivity diagram is used to illustrate the effects of different EV penetrations (percentages of customers with EV) and unbalance on customer voltages which in turn defines EV hosting capacity.
Results show that even though, for most of the scenarios, the VUF is smaller than statutory limits (2%), for the same transformer and the same EV penetration, a high VUF greatly limits EV hosting capacity. Moreover, reducing the VUF will not only benefit the worst customer, but it would also increase the voltage of all the customers in the network and thus increase the EV hosting capacity. However, different transformers with the same VUF may have different EV hosting capacities. Thus, from the perspective of EV hosting capacity, it is ineffective to adopt a limit for VUF.
References
[1] “Global EV Outlook 2021,” IEA, 2021. https://www.iea.org/reports/global-ev-outlook-2021
[2] “IEEE Recommended Practice for Monitoring Electric Power Quality,” IEEE Standard 1159-2019 (Revision of IEEE Std 1159-2009), vol. 2019, 2019.
Soobok Yoon
PhD candidate
Monash University
Presentation title: Gird-to-vehicle (G2V) algorithm with voltage control for Monash microgrid
Abstract: Globally, the transport industry is one of the most responsible sectors for carbon dioxide emissions (approximately 23%) along with electricity generation. In this regard, the electrification of the transport sector serves as one of the most prioritised options when planning to build a carbon-neutral society.
As of 2019, the sales of Electric Vehicles (EVs) in Australia accounted for 0.9%, lagging behind the global average of 2.6%. Nonetheless, it is predicted to reach up to 70-100% of new vehicle sales by 2040 with strong initiatives towards decarbonisation along the global trend. In this context, EV sales in Australia increased by 200% in 2019 compared to the previous year, leading to increased EV charging by more than two times.
However, if done in an uncontrolled/unsophisticated way, the rapidly increasing need for charging EVs could easily negatively impact the electricity grid operations. The research on smart charging was born in this context. This concept makes it possible to utilise the EV fleet as a prosumer that participates in multiple grid services rather than as a simple consumer.
This study explores a centralised Grid-to-Vehicle (G2V) charging algorithm to safely accommodate rapidly increasing EV charging in Monash microgrid (Clayton campus). The G2V algorithm is designed for the central aggregator (Monash university), and the proposed framework focuses on the safe operations of the microgrid and the operating profits via arrangements of EV charging schedules.
The optimisation process consists of two steps. In the first step, the main aggregator arranges the EV charging schedule to maximise the trading profits of electricity without sacrificing the EV owner’s needs (driving range anxiety). In the second step, the algorithm performs power flow analysis via the Newton-Raphson method to investigate grid congestion and voltage violations. The performance of the proposed method is tested by IEEE LV European test feeder system and MATLAB Simulink, along with several reference studies.
The simulation result shows that the proposed G2V scheme effectively suppresses grid violations in the long term with better Net Present Value (NVP) and Internal Rate of return (IRR). In addition, this study will provide deeper insights into the optimal investment strategies for Electric Vehicle Supply Equipment (EVSE) and the business model of smart charging services in Monash microgrid from the tech-economics point of view.