Browsing by Author "Pirnia, Mehrdad"

Now showing 1 - 8 of 8

Constraint-Guided Machine Learning for Solving Optimal Power Flow Problem
(University of Waterloo, 2022-08-31) Lotfi, Amir; Pirnia, Mehrdad
Due to the nonlinear and non-convex attributes of the optimization problems in power systems such as Optimal Power Flow (OPF), traditional iterative optimization algorithms require significant amount of time to converge for large electric networks. Therefore, power system operators seek other methods such as DC Optimal Power Flow (DCOPF) to obtain faster results, to obtain the state of the system. However, DCOPF provides approximated results, neglecting important features of the system such as voltage and reactive power. Fortunately, recent developments in machine learning have led to new approaches for solving such problems faster, more flexible, and more accurate. In this research, a Deep Neural Network-based Optimal Power Flow (DNN-OPF) algorithm is implemented on small to large case studies to show the accuracy and efficiency of the ML-based algorithms. Since the ML methods such as NN are considered black-box approaches, the system operators are not satisfied with solving power system models using them, as such methods do not explain the reasoning behind the generated solutions. Moreover, there is no guarantee that the obtained solutions would be converging and close to optimality. To overcome such issues this research provides a novel approach to first classify the converging and non-converging ACOPF problems, and then suggests a constraint-guided method, based on normalizing outputs and using particular activation functions to satisfy the technical limits of the generators such as maximum and minimum generation. Furthermore, a post-processing approach is incorporated to check for the convergence of the power flow equations which are in form of equality constraints. The suggested method is applied on IEEE24-bus, IEEE 300 bus, and PEGASE 1354 bus systems and the results show significant improvement on execution time, comparing to traditional gradient-based methods, such as Newton-Raphson and Gauss–Seidel methods. Also, the approach has been benchmarked against DCOPF model and it is shown that the proposed DNN-OPF not only provides faster speed, but also ensures higher accuracy on the final results. Furthermore, since is a need to run ACOPF problem using different scenarios, to account for continuous changes in the demand, the suggested DNN-OPF is solved for various scenarios from 1 to 10,000 to appreciate the improved execution time obtained from the ML-based approaches. Our results show that DNN can improve execution time a factor of 400 to 800 for large to small networks.
Data-Driven Inverse Optimization with Applications in Electricity Markets
(University of Waterloo, 2023-01-19) Rafieepouralavialavijeh, Ali; Pirnia, Mehrdad; Pavlin, Michael
Due to the increasing penetration of renewable resources and demand response instruments in the electricity markets, generation planning models have become more complex and require detailed information on the inherent structure of the system, including generator and demand parameters. Demand should be met by cost-effective, adaptable, and efficient power plants to ensure that it is met even in the worst-case scenarios, such as an unanticipated peak or the failure of a critical generating unit. On the other hand, there is a need to consider short-term details in the Planning problems to address the needed system flexibility due to sudden changes in demand and renewables generation. Such short-term details increase the size of the models and their related computations. As a result, there is a trade-off between the complexity of the computation and the level of short-term operational details, which should be considered. Accessing electricity infrastructure data in North America is often difficult due to the lack of open data standards and the proprietary nature of much of the data. The regulations and policies surrounding the data also vary significantly from province to province, making it difficult to access the data uniformly. Additionally, privacy and security considerations can limit access even further. Despite these limitations, there are indirect methods such as inverse optimization(IO) to derive the market parameters using publicly available data; examples of these parameters include generator costs of generation, their capabilities, etc. The discovery of unobservable information via IO could aid energy models to account for operational details without increasing the complexity of their problem. Furthermore, this information can inform policymakers on potential interventions to improve the efficiency of the electricity market. In this research, a MIP model is developed to incorporate capital and operational costs associated with long-term planning problems. The operating costs of each technology are assumed to be approximated by a series of step-wise functions so that model outcomes, such as generation output, are as close as possible to real-world electricity market generation. The proposed method employs a two-stage algorithmic framework using data-driven inverse optimization and regression. In the first stage, constraints are generated based on relationships between cost and electricity prices. In the second stage, these constraints on costs are added to a problem that finds and reconciles the parameters of the cost functions. To evaluate the performances of the proposed IO-based method, it was applied to a DC-OPF model using the IEEE 24-bus system, which helped eliminate power flow constraints. This approach was then applied to a long-term planning model using Ontario's electricity market data. The results indicate that the proposed approach could find a close solution to the conventional models. In the long-term planning model, the IO-based approach showed more moderate investment policies, while the traditional methods tend to over or under-invest.
Energy Management Systems for Multi-Microgrid Networks Under Uncertainties
(University of Waterloo, 2023-08-08) Ceja-Espinosa, Carlos; Canizares, Claudio; Pirnia, Mehrdad
Environmental concerns have motivated a gradual transformation of power systems in recent years, mainly focused on replacing fossil fuel-based energy sources with Renewable Energy Sources (RESs) such as solar and wind energy. However, due to their variable nature, the large-scale integration of RESs poses several technical challenges for the safe and efficient operation of evolving power systems. The adoption of microgrids (MGs) has increased as a viable option to effectively integrate RESs into existing grids and reduce the dependency on conventional, centralized power stations, as well as enhancing the electrical supply resiliency. Furthermore, MGs can provide sustainable energy to remote areas in which a connection to the main power grid is not possible. In this context, the Energy Management System (EMS) of the MG, which is responsible for determining its optimal operation, is an important part of MG control. However, the variability of electricity demand and RESs within an MG complicates the adequate dispatch of the MG resources to maintain supply-demand balance. Hence, uncertainties inherent to an MG must be taken into account, which is one of the main topics of this thesis. The coordinated operation of multiple MGs as a multi-microgrid (MMG) system has recently attracted attention due to the potential benefits that originate from a coordinated operation, as opposed to the individual and independent operation of each MG. The collective operation enables the possibility of power exchanges among MGs and the main grid, which can mitigate the unpredictability of RESs, as well as reduce the operational costs by taking advantage of the heterogeneity of load and generation profiles in each MG. Furthermore, differences in generation costs and grid buying/selling prices can incentivize power exchanges and ensure the maximum utilization of RESs. Therefore, it is important to design EMSs that adequately consider the collective operation of a set of MGs while taking uncertainties into account, which is the primary focus of this thesis. In the first part of this thesis, a centralized MMG EMS model is proposed, which is formulated as a cost minimization problem that considers the operation of all MGs and their interactions among each other and the main grid as a single system. The model includes detailed operational constraints of thermal generation units and Energy Storage Systems (ESSs), as well as power capacity limits at the Point of Common Coupling (PCC) of each MG. A decomposition procedure based on Lagrangian relaxation is then applied, with the goal of separating the complete problem into subproblems corresponding to each MG, which can be solved independently with minimal information exchange through a subgradient-based distributed optimization algorithm. Demand and solar irradiance data from a realistic Active Distribution Network (ADN) in São Paulo, Brazil, are then used to design a system to test and validate the proposed models. The simulation results show that the distributed algorithm converges to the optimal or a near-optimal solution of the centralized model, making the proposed approach a viable alternative for the implementation of a distributed MMG EMS. Furthermore, the advantages of an MMG system are demonstrated by showing that the operational costs of the system are significantly reduced when MGs are able to exchange power among each other and with the main grid, compared to their costs in individual operation. In the second part of this thesis, the proposed centralized MMG EMS model is reformulated using an Affine Arithmetic (AA) optimization framework to consider uncertainties associated with electricity demand and renewable generation. First, the uncertainties are characterized by their affine forms, which are then used to redefine the variables, objective function, and constraints of the original model in the AA domain. Then, the linearization procedure of the absolute values introduced by the AA operators is explained in detail. The proposed AA model is validated through comparisons with the deterministic and Monte Carlo Simulation (MCS) solutions. The test system used in the aforementioned MMG distributed dispatch approach is utilized to show that the AA model is robust under a range of possible realizations of the uncertain parameters, and can be solved with lower computational burden and in shorter execution times with respect to an MCS approach, while considering the same range of uncertainties, which is one the main advantages of the proposed AA model. Furthermore, it is demonstrated that the affine forms of the solution variables can be used to find a dispatch for different realizations of demand and renewable generation, with no need to repeatedly solve the optimization problem.
Geographic-information-based stochastic optimization model for multi-microgrid planning
(Elsevier, 2023-04-01) Vera, Enrique Gabriel; Cañizares, Claudio; Pirnia, Mehrdad
This paper presents a model for the realistic planning of multi-microgrids in the context of Active Distribution Networks with the assistance of Geographic Information Systems. The model considers the distribution system grid as well as the geographic features of the Region of Interest. It also includes long-term purchase decisions and short-term operational constraints, and considers uncertainties associated with electricity demand and Renewable Energy Resources using an existing Two-Stage Stochastic Programming approach. Geographic Information Systems along with Deep Learning are used to estimate the areas of rooftops within the Region of Interest and model the Low Voltage grid. The planning model is used to study the feasibility of implementing a multi-microgrid system consisting of 4 individual microgrids at an Active Distribution Network in a municipality in the state of São Paulo, Brazil. The results of the model presented in this paper are compared with the results obtained using Monte Carlo Simulations and an existing, less detailed, Two Stage Stochastic model. It is demonstrated that the stochastic solutions are close to those obtained with Monte Carlo at a lower computational cost, and that the use of Geographic Information allows to determine both the capacity and location of the PV panels, batteries, and distribution transformers on the microgrids grid, thus providing more precise and useful planning results.
Impacts of COVID-19 on Ontario's Electricity Market
(University of Waterloo, 2022-01-05) Elsarague, Menna; Pirnia, Mehrdad
The COVID-19 outbreak has not only threatened global health but has also significantly affected the energy sector. Most countries around the world have faced sudden changes in the electricity load as a result of the strict measures that have been taken by mid-March 2020 to limit the spread of the disease. In order to investigate the patterns of changes in the electricity sector and to predict future load, machine learning (ML) techniques, such as descriptive data analytics, clustering, and forecasting methods, have been used widely in practice. This research, in particular, studies the impacts of the pandemic on Ontario’s electricity market by investigating changes in the electricity demand and prices. It further provides insights into incorporating ML methods for electricity load forecast and prescribes enhanced solutions for the pricing of electricity by assessing Ontario’s Market Renewal pricing system during COVID-19. The analysis of demand and price changes due to the pandemic is presented through a comprehensive study of Ontario’s hourly electricity demand and hourly electricity prices (HOEP) considering annual, monthly, and daily granularity. Furthermore, the impact of the pandemic on load forecasting is investigated using a short-term Feed Forward Neural Network (FFNN) model, as in such rare events, load forecasting becomes more challenging and less accurate, causing high risks in the electricity system operation. Finally, the potential efficiency of Ontario’s Market Renewal during COVID-19 is assessed through a comparative analysis between Ontario’s current electricity market and New York’s electricity market, which has a comparable electricity system with respect to load and supply of electricity. In order to conduct this study, Ontario’s hourly electricity demand and price data, as well as the hourly weather data are used. Our data-driven analysis shows that although the electricity demand dropped by 12% during the beginning of the pandemic in March, it started unexpectedly rising by the end of May 2020 to levels that exceeded the electricity demand in 2019. A similar pattern is observed for Ontario’s HOEP. The load forecast model performance is evaluated using the mean absolute percentage error (MAPE) during three distinct periods: pre-pandemic, beginning of the pandemic, and during the pandemic to illustrate how the sudden changes in the early stage of COVID-19 have affected the load forecast compared to other periods. The results of the forecast model show an overall MAPE of: 3.21%, 13.86%, and 4.23%, respective to the periods identified. Expectedly, the performance of the model during the pandemic is significantly affected. However, the model is still considered plausible, as a MAPE index between 10% and 20% is classified as good forecast accuracy. Finally, through the comparative analysis between the current Ontario’s uniformed price market and New York’s locational marginal price (LMP) based market, it is observed that Ontario’s current pricing system is less efficient and that consumers’ welfare could increase with an LMP pricing system, which will be part of the proposed Ontario’s Market Renewal.
Planning of Multi-Microgrids Considering Uncertainties and Spatial Characteristics
(University of Waterloo, 2023-03-13) Vera, Enrique; Canizares, Claudio; Pirnia, Mehrdad
Global warming is a serious issue that is being tackled from various fronts, one of them is the decarbonization of electrical energy systems, which may be addressed by introducing clean Distributed Energy Sources (DERs) such as Renewable Energy Resources (RESs) and Energy Storage Systems (ESSs). These types of technologies can be clustered to form Microgrids (MGs), which have proven to be technically and financially feasible solutions to supply electricity demand while reducing emissions and increasing resiliency. MGs can operate isolated or connected to the grid, both in rural and urban settings, which allows them to interact with the existing electricity grid to enhance its capabilities and functionalities, while improving power quality, reducing network congestion, increasing efficiency, reliability, and flexibility, and delaying investments in transmission and distribution systems. Hence, this thesis focuses on various relevant and timely aspects of MG planning, in particular for isolated Remote Communities (RCs) and for the interconnection of MGs and their integration with Active Distribution Networks (ADNs) to form Multi-Microgrid (MMG) systems. The deployment of clean MGs to satisfy RC electricity needs, considering their inherent geographic characteristics, imposes a series of challenges that must be taken into account when planning them. Thus, delivering electricity to RCs is economically and environmentally expensive, as the main source of electricity is diesel generators, which present significant Greenhouse Gas (GHG) emissions, and Operations and Maintenance (O\&M), transportation, and fuel costs. Therefore, an optimization model for the long-term planning of RC MGs to introduce RESs and ESSs is proposed in this thesis, with the objective of reducing costs and emissions. The presented model considers lithium-ion batteries and hydrogen systems as part of ESSs technologies. The model is used to investigate the feasibility of integrating these DERs in an MG in Sanikiluaq, an RC in the Nunavut territory in Northern Canada, where several planning scenarios with various combinations of resources are considered in order to assess the impact of different technologies. The results show that wind resources along with solar and storage technologies can play a key role in satisfying RC electricity demand, while significantly reducing costs and GHG emissions. Independent MGs can be interconnected to form MMG systems in the context of ADNs, bringing valuable benefits such as energy use, power quality and stability improvements, as well as flexibility and thus economic enhancements for both costumers and utilities. Therefore, a Two Stage Stochastic Programming (TSSP) model is proposed for the planning of MMGs within ADNs at Medium Voltage (MV) levels to minimize the total costs, while benefiting from interconnections of MGs and considering uncertainties associated with electricity demand and RESs. Furthermore, the model includes long-term purchase decisions and short-term operational constraints, using Geographical Information Systems (GIS) to realistically estimate rooftop solar limits. The planning model is used to study the feasibility of implementing an MMG system consisting of 4 individual MGs at an ADN in a municipality in the state of São Paulo, Brazil. The results show that the TSSP model tends to be less conservative than the deterministic model, which is based on simple and pessimistic reserve constraints, while being computationally more efficient than the usual, Stochastic Linear Programming (SLP) and Monte Carlo Simulations (MCS) approaches, with adequate accuracy. Finally, the MMG planning model at MV is further extended to include the Low Voltage (LV) grid. Thus, a model is proposed for the realistic planning of MMGs in the context of ADNs, with the assistance of GIS. The model considers the distribution system grid with an adequate level of detail for multi-year planning as well as the geographic features of the studied region. Similar to the MV model, it also includes long-term purchase decisions and short-term operational constraints, and considers uncertainties associated with electricity demand and RESs using a TSSP approach. GIS along with Deep Learning (DL) are used to more accurately estimate the rooftop areas within the studied region for solar PV deployment, as well as for modelling the LV grid. The planning model is then used to study in more detail the feasibility of implementing the MMG system previously considered in São Paulo, Brazil. The results of the extended TSSP LV grid model are compared with the results obtained using MCS and the less detailed TSSP MV grid model, demonstrating that both TSSP solutions are close to those obtained with MCS at a lower computational cost, while providing accurate and practical planning results.
Renewable Energy Integration in Canadian Remote Community Microgrids: The Feasibility of Hydrogen and Gas Generation
(Institute of Electrical and Electronics Engineers (IEEE), 2020-12-02) Vera, Enrique Gabriel; Canizares, Claudio; Pirnia, Mehrdad
Approximately 1.1 Billion, or 14%, of the global population do not have access to electricity due to the challenges associated with energy supply. Around 84% of those without electricity access reside in rural areas, with more than 95% being in sub-Saharan Africa and the developing parts of Asia. In Canada, about 72% of off-grid aboriginal and nonaboriginal communities use fossil fuel (oil: 71%, natural gas: 0.8%) as their main source of electricity generation, and only 4.7% of these communities rely on renewable energy sources (RES). In addition, 17.9% fulfill their energy demand through interconnections with other communities as they don?t have enough resources to support their own needs. The remaining 5.6% are reported to rely on unknown sources of electricity (see Arriaga et.al. 2014).
Two-Stage Stochastic Optimization Model for Multi-Microgrid Planning
(Institute of Electrical and Electronics Engineers (IEEE), 2022-10-03) Vera, Enrique Gabriel; Cañizares, Claudio A.; Pirnia, Mehrdad; Guedes, Tatiana Pontual; Trujillo, Joel David Melo
This paper presents a Two Stage stochastic Programming (TSSP) model for the planning of Multi-Microgrids (MMGs) in Active Distribution Networks (ADNs). The model aims to minimize the total costs while benefiting from interconnections of Microgrids (MGs), considering uncertainties associated with electricity demand and Renewable Energy Sources (RESs). The associated uncertainties are analyzed using Geometric Brownian Motion (GBM) and probability distribution functions (pdfs). The model includes long-term purchase decisions and short-term operational constraints, using Geographical information Systems (GIS) to realistically estimate rooftop solar limits. The planning model is used to study the feasibility of implementing an MMG system consisting of 4 individual Microgrids (MGs) at an ADN in a municipality in the state of São Paulo, Brazil. The results show that the TSSP model tends to be less conservative than the deterministic planning model, which is based on simple and pessimistic reserve constraints, while performing faster than a simple Stochastic Linear Programming (SLP) algorithm, with higher accuracy.