Program Topic: Renewables Integration & Optimization – Solar

W3-1 – Wednesday 13:20-15:00

13:20 The Transient Behavior of the Volt/Var Control of Photovoltaic Systems for Solar Irradiation Variations

Insu Kim (Alabama A&M University, USA); Ronald Harley (Georgia Institute of Technology, USA)

Nowadays, photovoltaic (PV) systems supply reactive power in limited situations with mutual agreements so that they can minimize an increase in over- and under-voltages. However, since PV systems depend on weather conditions, including solar irradiation, they have, in nature, intermittency in their output. Therefore, the objective of this study is to analyze the effect of ramp events of solar irradiation on the generation output of PV systems in the transient state. For this purpose, this study models a distribution system that includes a PV system (e.g., a PV array, a DC bus and capacitor model, a controller, and a static generator) in DIgSILENT of Power Factory, generates a ramp event of solar irradiation as input data to the PV system, and examines the transient behavior of the PV system in the electromagnetic transient mode of DIgSILENT. Finally, the effect of reactive power control for the ramp event on over- and under-voltages will be addressed.

13:40 An Adaptive Neuro-Fuzzy Inference System-based MPPT Controller for Photovoltaic Arrays

Farhad Khosrojerdi (Université du Québec en Outaouais (UQO) & Université du Québec en Outaouais (UQO), Canada); Shamsodin Taheri (Université du Québec en Outaouais, Canada); Ana-Maria Cretu (Université du Québec en Outaouais, Canada)

This paper presents a Maximum Power Point Tracking (MPPT) method applying an Adaptive Neuro-Fuzzy Inference System (ANFIS) for a stand-alone photovoltaic (PV) system. The proposed ANFIS-based MPPT technique determines the optimal operating point of a PV system which is designed in conjunction with a Z-source DC-DC converter as an interface between the PV array and the load. In the ANFIS-based MPPT controller, real meteorological data are used to define the two input membership function plots assuming that the PV array is located in Ottawa, Canada. The performance of the proposed MPPT technique in tracking the maximum power point (MPP) is assessed numerically in the MATLAB/Simulink environment. The simulation results highlight the benefits of determining reference voltage and duty cycle as output membership functions of the control system without using the current and voltage sensors. Moreover, in comparison with conventional control systems, the proposed solution can reduce the complexity and cost of the control system by eliminating the PID controller.

14:00 Tilt Angle Optimization for Maximum Solar Power Generation of a Solar Power Plant with Mirrors

Isuru Vidanalage (Ryerson University, Canada); Kaamran Raahemifar (Ryerson University, Canada)

Solar power generation is mainly based on direct, diffused and reflected solar radiation. This paper will give an insight of the strategy of the implementation of optimization of the tilt angle of the solar panel to maximize the electricity generation, at presence of mirrors. Mirrors will improve the reflected solar radiation, leading to increase the radiation on solar panel. For the purpose of analysis, as the site Toronto in Canada was selected. Renewable energy data was gathered through National Renewable Energy Laboratory (NREL). Energy increment due to the addition of mirrors are discussed and it will be a really useful input for designing solar equipment.

14:20 Sensitivity Analysis of a Current-Source Inverter-Based Three-phase Grid-connected Photovoltaic System

Prajna Dash (Energent, Canada); Mehrdad Kazerani (University of Waterloo, Canada)

In this paper, first, an accurate small-signal model of a three-phase current source inverter (CSI)-based grid-connected photovoltaic (PV) system is derived. The model takes into account the nonlinear characteristics of PV arrays, as well as the controller and grid-side dynamics. Then, the developed small-signal model is employed to investigate the impact of different parameters on the stability of the CSI-based PV system, through sensitivity analysis. The sensitivity to variations of different parameters is assessed by tracking the positions of real parts of the eigenvalues corresponding to the dominant state variables in the s-plane. Insolation level, controller gains, and grid inductance are the parameters considered to study the sensitivity of state variables.

14:40 Impact of Systems Cost Incentives on the Economics of Distributed Solar Power in Canada

Hillary MacDougall (University of Ottawa, Canada); Steve Tomosk (DAI Inc, Canada); David Wright (University of Ottawa, Canada)

Systems cost incentives have been used in USA and Canada to stimulate the deployment of solar power. This paper analyzes the impact of such incentives on the internal rate of return (IRR) from residential, offsetting projects in Canada, focusing on Regina, a high irradiance city. For a given percentage systems cost reduction, the IRR increases by a higher percentage. The impact on the IRR for photovoltaic (PV) projects is higher than for concentrating PV (CPV) projects. During the period 2016 to 2020 systems prices decline and electricity costs increase resulting in an increase in IRR from deferring the project start date. That increase is higher for CPV than for PV. Orienting the PV modules so as to maximize revenues has a minimal impact on IRR.

W4-1 – Wednesday 15:20-17:00

15:20 Power Integration for Solar Applications

Tanya Gachovska (Solantro Semiconductor Corp., Canada); Tudor Lipan (Solantro Semiconductor Corp, Canada); Gabriel Scarlatescu (Solantro Semiconductor Corp, Canada); Chris Gerolami (Solantro Semiconductor Corp, Canada); Nikolay Radimov (Solantro Semiconductor Corp, Canada); Mihai Varlan (Solantro Semiconductor Corp, Canada)

— Power integration in power electronic circuits is important for new circuit designs since it tends to decrease the bill of materials, cost, and size of an application, with the added benefit of an increase in reliability. A functional description of a novel architecture called a Micro-Inverter Controller (IXC64) integrated circuit is discussed. The IXC64 has a 32-bit RISC microcontroller core with 64 kB of RAM, as well as advanced switching timers and sensing features for control of different topologies for solar applications. These topologies may include stages such as an Active Clamp Forward converter, Buck converter, Unfolding Bridge, Dual Buck converter, Double Antiphase Buck converter, Chopper, and Current-fed Push-pull converter. Characterized test results and results of applications using the IXC64 is presented

15:40 Modeled Estimates of Solar Direct Normal Irradiance in Al-Hanakiyah,Saudi Arabia and Boulder, USA

Emad Abyad (University of Ottawa & Taibah University, Canada); Christopher Valdivia (University of Ottawa, Canada); Joan Haysom (Leidos Canada & University of Ottawa, Canada); Ahmad Atieh (University of Jordan, Jordan); Karin Hinzer (University of Ottawa, Canada)

Accurate knowledge of the solar resource is required for solar energy generation projects, including preferably data for both the direct normal (DNI) and diffuse horizontal (DHI) components of irradiance, but in many solar datasets, only a single global horizontal (GHI) may be available. This paper undertakes the evaluation of three models-Reindl*, BRL and DISC – which estimate DNI and DHI in such cases. The models are tested against actual data from Boulder, USA, and Al-Hanakiyah, Saudi Arabia. Generally, a good correlation between hourly measured or calculated DNI and hourly modeled DNI is observed for the three models in Boulder and Al-Hanakiyah. The three models overestimate DNI in Al_Hanakiyah while Reindl* and BRL model exhibit less bias for estimating DNI in Boulder. The results show that the BRL model outperforms other models in Boulder with RMSE of 121 W/m² and standard deviation (𝜎) of 121 W/m². Overestimation of DNI has been noticed in BRL and Reindl* models with values that exceed 1100 W/m². On the other hand, Reindl* model’s performance is better with MBE of -28 W/m², MAE of 63 W/m², and RMSE of 86 W/m² in Al-Hanakiyah. Other locations such as Ottawa, Canada and Ma’an, Jordan as well as DIRINT model will be included to evaluate the performance of each model under different weather conditions in the future.

16:00 Design and Application of a Photovoltaic Powered Domestic Solar Water Heating System in Regina, SK., Canada

Somkene Mbakwe (Memorial University of Newfoundland, Canada)

Canadians use an average of 75 Liters of hot water, per individual, at home every day for washing dishes and clothing, cleaning and personal hygiene. With the increasing cost of electricity, as well as the current and growing spate of environmental and climate change issues, the cost of heating domestic water that will meet this demand, through retail electricity, can add up quickly for the average family. This study presents a sustainable solution for domestic water heating – an indirect, non-pressurized, closed-loop with drain back tank domestic solar water heating system, driven by a photovoltaic-powered dc pump – which provides optimal techno-economic performance, reliability, availability, and maximized environmental performance. The analysis results of the proposed system show that the proposed system meets the performance criteria set for the project with annual savings of $744. The proposed system also demonstrates to be financially viable based on all key indicators assessed – NPV, IRR and yielding a simple payback period of approximately 12.8 years. A net annual GHG emission reduction of 3.1 tCO2/yr and cumulative net GHG reduction of 72 tCO2 over the duration of the project life is achieved, with a potential to even further improve both energy savings and emission reduction by up to 18% on an annual basis.

16:20 Power Variability Analysis of Megawatt-Scale Solar Photovoltaic Installations

Mohammad Sedighy (Hatch, Canada); Mark Mitchell (Hatch Ltd., Canada); Michael Campbell (Hatch, Canada); Kathryn Klement (Hatch, Canada)

Power output variability of solar PV installations is an increasingly important topic as more renewable power generation is connected to electrical grids. This paper presents a summary of solar power fluctuations for two solar installations in Eastern Ontario as measured over a two-month period. The installations use the same cell technology, and are in the same relative geographical area. In this analysis the 99.5th percentile value has been used to represent the maximum rate of change due to irradiance variability whereas the absolute maximum (100th percentile) rate of change would have included variability resulting from events such as curtailment step changes or breaker trips. Further analysis results are presented on the maximum dip from daily peak power output caused by weather conditions. The measurement setup allowed for the two 10 MW installations to be segmented and combined, giving a total of five different configurations subject to similar meteorological conditions. The results are also compared to other published studies on this topic.

16:40 A Predictive Direct Power Control Technique for Transformerless Grid Connected PV Systems Application

Nima Safari (University of Saskatchewan, Canada); Osama Aslam Ansari (University of Saskatchewan, Canada); Chi Yung Chung (University of Saskatchewan, Canada)

Transformerless grid-connected PVs are more economical and efficient compared to transformer connected PVs. However, in this type of connection high leakage current flows between the PV arrays and the grid due to the parasitic capacitances of the system. This leakage current is one of the main culprits in causing electromagnetic interference in power system, saturation of distribution transformers and poses safety issues. Common mode voltage (CMV) is primarily responsible for the existence of this leakage current. To cope with issues which arise from omitting galvanic isolation (transformer), in this paper a predictive direct power control (PDPC) with reduced common voltage (CMV) called active zero predictive direct power control (AZ-PDPC) is proposed. Both theory and simulation show that the proposed method reduces the level of CMV along with general good performance. The proposed method’s characteristics are compared with the PDPC technique.