INCREASED PENETRATION OF DISTRIBUTED ROOF-TOP PHOTOVOLTAIC SYSTEMS IN SECONDARY LOW VOLTAGE NETWORKS: INTERCONNECTION IMPACT ANALYSIS AND MITIGATION
| dc.contributor.advisor | Liang, Xiaodong | |
| dc.contributor.advisor | Chung, Tony | |
| dc.contributor.committeeMember | Bui, Francis | |
| dc.contributor.committeeMember | Karki, Rajesh | |
| dc.contributor.committeeMember | Zhang, Chris | |
| dc.contributor.committeeMember | Wahid, Khan | |
| dc.contributor.committeeMember | Ventaktesh, Bala | |
| dc.creator | Asiri, Elisha C | |
| dc.creator.orcid | 0000-0001-6567-5680 | |
| dc.date.accessioned | 2023-09-25T19:58:22Z | |
| dc.date.available | 2023-09-25T19:58:22Z | |
| dc.date.copyright | 2023 | |
| dc.date.created | 2023-08 | |
| dc.date.issued | 2023-09-25 | |
| dc.date.submitted | August 2023 | |
| dc.date.updated | 2023-09-25T19:58:22Z | |
| dc.description.abstract | The worsening global climatic condition has necessitated increased investments in renewable energy resources and in turn increased penetration of these resources in electricity grids worldwide. Distributed photovoltaic (PV) energy is one of the rapidly growing and viable forms of renewable energy. Distributed PV systems currently exist in two modes, a few numbers of large or utility scale systems and a plethora of small or residential scale roof-top systems which are rapidly growing in terms of number. Residential systems are alternatively called Behind the Meter (BTM) systems because they are not directly monitored by utility operators and are therefore invisible vis-à-vis their performances. While an individual Behind-The-Meter (BTM) system's size holds little significance in comparison to the inertia of the utility grid, the collective presence of numerous interconnected BTM systems within a single feeder has the potential to jeopardize the stability and security of utility operations. Conventional protective devices within distribution networks are designed to accommodate a unidirectional downstream power flow. However, as the integration of PV generators into utility grids intensifies, the prospect of reverse or upstream power flow becomes more probable. This development raises various apprehensions, including the potential for voltage level breaches and a notable reduction in the operational longevity of these devices. BTM systems generally have a wide geographical coverage within a region and each system operates independently of others as well as the fact that their real-time performances are concealed in the net-load data relayed by electricity meters. Consequently, traditional forecasting methods have proved insufficient in predicting the outputs of PV systems on a regional level requiring the development of spatial aggregation approaches. Three basic sub-areas aimed at increasing the penetration of BTM PV generators in utility grids are the principal focus of this study. The sub-areas include performance analysis of BTM systems; day-ahead regional scale PV power forecasting model and a PV ramp events extraction model. The first sub-area tries to address the challenges with small scale solar power performance data access on a regional basis. The performance analysis was aimed at evaluating the credibility and reliability of BTM data from public webpages and their representativeness for high profile research. Consequently, this sub-area proposes and investigates the feasibility of the instrumentation of every invisible solar system for near real-time data monitoring. The investigation involved detailing the convergence between simulated and reported power outputs on a spectrum of orientation and tilt angles. Two simulation methods as well as two case studies public web repositories from which a subset of representative solar sites were adopted to provide a basis for the proposed approach. The results show that the proposed model is viable and feasible depending on the participation of certain key stakeholders in electricity market discourses. Day-ahead forecasts are required by electricity market investors to make informed decisions on the trading floor. Whereas it is relatively easier to predict the performance of a few large-scale PV systems, a large number of small-scale PV systems with a wide geographical spread poses more challenges because they are not metered for real-time monitoring. This sub-area proposes an artificial neural network (ANN)-based model to achieve regional-scale day-ahead PV power forecasts from numerical weather predictions of weather variables excluding solar irradiance as inputs. The model was first implemented by dividing a region into clusters and selecting a representative site for each cluster using data dimension reduction algorithms. Solar irradiance forecasts were then generated for each representative PV system and the corresponding PV power was simulated. The cluster power output was obtained using a linear upscaling model and summed to produce regional-scale power forecasts. The model’s accuracy is validated using power generation data of several distributed systems in California. Compared with available benchmark models with similar objectives, the proposed model performed significantly better. Insufficient information on solar power ramp events is counterproductive to the operational flexibility and economics of electricity grids. Accurate solar ramp extraction and characterization in terms of ramp magnitude, rate and duration are useful to power system operators for system planning especially with regards to ensuring supply security and sizing ancillary services. The characterization of ramp events in historical databases is also useful for testing forecast models’ accuracy in predicting significant solar ramp events that are of more concern to utility operators. A novel technique for solar power ramp events (SPREs) detection using the modified swinging door algorithm (MSDA) considering different time resolutions and weather profile is proposed in this sub-area. Firstly, the swinging door algorithm (SDA) is used to create ramp segments of the solar power data that are collected from different randomly selected systems. Afterwards, the power generation variability patterns of these segments are studied. The SDA is then modified to merge adjacent segments according to the observations made by comparing the variability patterns. The solar power data simulated from irradiances measured with different time resolutions is utilized for performance validation and testing. The proposed technique shows much improved performance than existing detection algorithms with respect to the number of detected ramps, detection accuracy and in some cases, computation time. | |
| dc.format.mimetype | application/pdf | |
| dc.identifier.uri | https://hdl.handle.net/10388/15054 | |
| dc.language.iso | en | |
| dc.subject | Roof-top photovoltaic systems, Behind-the-meter systems | |
| dc.subject | Residential PV systems, Day-ahead prediction | |
| dc.subject | Clusters and representative systems | |
| dc.subject | k-means clustering | |
| dc.subject | Principal Component Analysis | |
| dc.subject | Ramp events characterization | |
| dc.subject | Swinging door algorithm | |
| dc.subject | ramp definition | |
| dc.title | INCREASED PENETRATION OF DISTRIBUTED ROOF-TOP PHOTOVOLTAIC SYSTEMS IN SECONDARY LOW VOLTAGE NETWORKS: INTERCONNECTION IMPACT ANALYSIS AND MITIGATION | |
| dc.type | Thesis | |
| dc.type.material | text | |
| thesis.degree.department | Electrical and Computer Engineering | |
| thesis.degree.discipline | Electrical Engineering | |
| thesis.degree.grantor | University of Saskatchewan | |
| thesis.degree.level | Doctoral | |
| thesis.degree.name | Doctor of Philosophy (Ph.D.) |