Multicriteria optimization of design of a grid interfaced PV-wind power system

Abstract

Integrated renewable energy systems are considered to be the most promising sustainable power generation solution in the near future. Due to the technological developments in the last two decades, development of renewable energy system has greatly increased to face the negative environmental impact caused by conventional energy sources and continuous rising of prices of conventional fuels (coal, gas and oil etc). Among several renewable energy source alternatives, hybrid configuration of photovoltaic (PV) and wind turbine (WT) system steadily connected to the grid is a potential resource. Development of an optimal design approach for hybrid configuration of PV-WT system interfaced to a grid requires to take into consideration of different system variables such as potential solar irradiance and wind speed profiles, daily load profile, technical specifications of devices used in PV-WT system, grid electricity cost, initial investment cost, operation and maintenance cost of PV-WT system, land availability and cost of land etc. Optimal balancing among PV, WT and grid electricity requires particular attention to achieve a good engineering solution. In previous literature, different design approaches represented by several optimization algorithms such as particle swarm optimization (PSO), genetic algorithm, modified PSO, discrete harmony search algorithm etc are used to make hybrid configuration of renewable energy sources. These optimization algorithms are used to minimize single objective function, this objective function is represented mainly by the total system cost. Sometimes technical and/ or economical and/ or environmental requirements are converted into a single mathematical expression to make single objective function. These methods are not always practical since some of the system variables might not be easily converted into a single unified unit. Multi-criteria optimization without the need for conversion of several design criteria to a single function is a very significant design solution in this respect. Few researches used multi-criteria optimization approach for making hybrid configuration of renewable energy sources considering two or three criteria. This research deals with problem of optimal electricity sharing among PV, WT and grid considering four criteria such as technical, economical, social and environmental criterion. Annual average electricity share from grid is considered, as there are some locations where solar and wind resources are not sufficient enough to generate electricity to meet the annual electricity demand of those locations. Technical criterion is represented as energy index of reliability (EIR). EIR indicates the capability of energy sources to meet the load for different instances throughout the day. Economical criterion is represented as cost per unit electricity (kWh) generation. Environmental criterion is represented as equivalent CO2 emission for usage of electricity from PV and WT system. Social criterion is represented as social acceptance by the inhabitants. Although PV and WT system emits low equivalent CO2 but usage of land is larger than that of conventional power plant. Considering land usage and emission of equivalent CO2 for generation of 1MWh electricity per day social criterion is modeled using fuzzy logic. In this thesis PV panels are considered to be annually fixed to a particular tilt and azimuth angle. Considering variations of solar irradiances in different months of a year for a particular location, optimal tilt and azimuth angle of PV surface is determined using genetic algorithm. As the value of power conversion co-efficient (Cp) of wind turbine varies for different wind speeds, electricity generated from wind turbine in different wind speed conditions are directly calculated from power curve of the wind turbine provided by the manufacturers. Due to aging of PV panels and WT generators, capacity of electricity conversion is reduced. To take into account of de-rating of PV and WT systems due to aging effect, linearly de-rating curve is considered according to the technical information provided in technical data specification sheet supplied by the manufactures. Sixty six numbers of alternatives are generated for different sharing of electricity among PV, WT and grid sources to meet the local demand of electricity. Each alternative consists of individually different integration of percentile sharing of electricity sources. Performances of multiple criteria are evaluated for every alternative. The Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) is the one of the powerful multi-criteria decision analysis which is used for ranking the alternatives. As cost is the most important consideration for planning of a large scale power plant, economical criterion should be given more preference. This research proposed the double weighing of economical criterion than any other criteria. A case study is performed for the location of Kutubdia Island, Bangladesh. A comprehensive electrical design of grid tie PV-WT system is adopted for the best alternative considering solar wind resources and daily load profile of the location of Kutubdia Island, Bangladesh.