Optimization of Electrical System for Offshore Wind Farms via a Genetic Algorithm Approach

Research output: Book/ReportPh.D. thesis

Abstract

Offshore wind farms seem to be more attractive than onshore farms. However, offshore wind farms cost more money than onshore wind farms in both installation and maintenance. Due to the fast development of power electronics, more kinds of configurations of offshore wind farm are possible, which lead to very different costs, system reliability, power quality, and power losses etc. Therefore, the optimization of electrical system design for offshore wind farms becomes more and more necessary.

There are two tasks in this project: 1) the first one is to construct an algorithm for finding the capacity of a grid-connected wind farm; 2) the second one is the optimization of electrical system for offshore wind farms (OES-OWF).

The capacity of a grid connected wind farm is limited by the transfer capability of the grid system, where the thermal limit of the transmission lines, the voltage stability, and the LTC limitation of transformers, the power generation limits and the voltage operation range are considered as the constraints. The optimization method combined with probabilistic analysis is used to obtain the capacity of a given wind farm site.

The OES-OWF is approached by Genetic Algorithm (GA). This platform is based on a knowledge database, and composed of several functional modules such as cost calculation, reliability evaluation, losses calculation, AC-DC integrated load flow algorithm etc. All these modules are based on a spreadsheet database which provides an interface for users to input corresponding parameters. The objective of the optimization is to minimize the mixture of Levelized Production Cost (LPC) and the reliability index of this work. LPC is the discounted life-cycle average cost per unit of electricity produced. It considers the capital costs, the maintenance costs, the power losses and power generation.

This work proposes a serial AC-DC integrated load flow algorithm for variable speed offshore wind farms (VSOWF). The model of DC/DC converters is proposed and integrated into the basic DC load flow algorithm by modifying the Jacobian matrix. Two iterative methods are proposed to respectively take into account the control strategy and power losses of PWM converters.

A reliability index, Loss of Generation Ratio Probability (LOGRP), is proposed to evaluate the electrical system of OWF. The LOGRP doesn't depend on the load demand and has weaker correlation with wind speed. Based on the particular characteristics of OWF, the algorithm to compute LOGRP is proposed in this report.

There are lots of different GA techniques which are used to improve the performance of the algorithm. In this work, the GA is implemented in OES-OWF and several different GA techniques are investigated.

Then, this optimization platform was applied to two real offshore wind farms. One is a small OWF and another one is relative large. Both applications showed satisfactory results. This platform provides a useful tool both for the future wind farm design and the design evaluation of existing wind farms.

Close

Details

Offshore wind farms seem to be more attractive than onshore farms. However, offshore wind farms cost more money than onshore wind farms in both installation and maintenance. Due to the fast development of power electronics, more kinds of configurations of offshore wind farm are possible, which lead to very different costs, system reliability, power quality, and power losses etc. Therefore, the optimization of electrical system design for offshore wind farms becomes more and more necessary.

There are two tasks in this project: 1) the first one is to construct an algorithm for finding the capacity of a grid-connected wind farm; 2) the second one is the optimization of electrical system for offshore wind farms (OES-OWF).

The capacity of a grid connected wind farm is limited by the transfer capability of the grid system, where the thermal limit of the transmission lines, the voltage stability, and the LTC limitation of transformers, the power generation limits and the voltage operation range are considered as the constraints. The optimization method combined with probabilistic analysis is used to obtain the capacity of a given wind farm site.

The OES-OWF is approached by Genetic Algorithm (GA). This platform is based on a knowledge database, and composed of several functional modules such as cost calculation, reliability evaluation, losses calculation, AC-DC integrated load flow algorithm etc. All these modules are based on a spreadsheet database which provides an interface for users to input corresponding parameters. The objective of the optimization is to minimize the mixture of Levelized Production Cost (LPC) and the reliability index of this work. LPC is the discounted life-cycle average cost per unit of electricity produced. It considers the capital costs, the maintenance costs, the power losses and power generation.

This work proposes a serial AC-DC integrated load flow algorithm for variable speed offshore wind farms (VSOWF). The model of DC/DC converters is proposed and integrated into the basic DC load flow algorithm by modifying the Jacobian matrix. Two iterative methods are proposed to respectively take into account the control strategy and power losses of PWM converters.

A reliability index, Loss of Generation Ratio Probability (LOGRP), is proposed to evaluate the electrical system of OWF. The LOGRP doesn't depend on the load demand and has weaker correlation with wind speed. Based on the particular characteristics of OWF, the algorithm to compute LOGRP is proposed in this report.

There are lots of different GA techniques which are used to improve the performance of the algorithm. In this work, the GA is implemented in OES-OWF and several different GA techniques are investigated.

Then, this optimization platform was applied to two real offshore wind farms. One is a small OWF and another one is relative large. Both applications showed satisfactory results. This platform provides a useful tool both for the future wind farm design and the design evaluation of existing wind farms.

Original languageEnglish
Place of PublicationAalborg
PublisherInstitut for Energiteknik, Aalborg Universitet
ISBN (Print)87-89179-67-6
StatePublished - 2006
Publication categoryResearch
ID: 17909237