On the potential of parallel powertrains to reduce the cost of energy from offshore wind turbines

  • Godwin Akpan JIMMY

Student thesis: Doctoral Thesis


Offshore wind turbine operating conditions are challenging with access for maintenance being limited by weather to a greater degree than for onshore turbines, resulting in prolonged downtime and reduced availability. This makes operational costs (helicopter, crew transfer or heavy lift vessels) more expensive, leads to loss of energy production and tends to increase the cost of energy of offshore wind farms. It is therefore important to investigate potential strategies that could improve availability, energy production and at the same time reduce operation and maintenance (O&M) cost and cost of energy in the long run. One possible option for availability improvement and cost of energy reduction is through the powertrain design. Most of the existing wind turbine types could be distinguished through their powertrain configurations. Conventional wind turbine powertrain exhibits single-input-single-output topology (one gearbox coupled to a generator with a power converter) while some exist with no gearbox (gearless drive). Although some of the geared and gearless powertrains have some good availability, yet they are still susceptible to prolonged downtime and consequently significant energy loss. This has alarmed the need to introduce the concept of parallelism into the design of offshore wind turbine powertrain. This research, therefore, focusses on a configuration with single-input-multiple-output (parallel powertrain) subsystems as a strategy for improvements in availability, energy production and cost of energy reduction. The novelty of this work comes from the availability improvement of small parallel subsystem with reduced failure rate, extra energy production at failure states, reduction in (O&M) cost due to high repair rate and the resulting cost of energy reduction of parallel powertrain. The highest-level research question amongst all of the research questions answered in this work is: "Can parallel powertrains reduce the cost of energy of offshore wind turbines?" In attempting to address this key question and other secondary research questions, in Chapter 3 the author carries out survey and analysis of failure and repair rate data from published sources to determine how they vary with powertrain configuration, power ratings, and sizes. In Chapter 4, a baseline powertrain availability and that of different parallel powertrains are evaluated using Markov state space model (MSSM). In Chapter 5, the annual energy production (AEP) of parallel powertrain is analysed using Raleigh probability distribution and the rated power in order to quantify any extra benefit at below rated wind speed. The ideal AEP is analysed at rated power, rated wind speed and at no-failure state. Also, the losses and efficiency of parallel powertrain at failure states are evaluated. Chapter 6 estimates the O&M costs of parallel powertrains using offshore accessibility tool. Chapter 7 calculates the cost of energy of parallel powertrain using AEP and O&M cost results from previous chapters in combination to initial capital cost (ICC). Finally, a general conclusion is made in Chapter 8. The novel results from each chapter provide some new insight into the potential of the parallel powertrain. The thesis concludes that an increase in the number of parallel systems, N, does not automatically lead to a higher availability for a wind turbine powertrain; however, when failure and repair rates scale with module power ratings then there is an improvement. It is possible to have extra AEP at below rated wind speed and at the various failure states of parallel powertrain. Potential reduction in the cost of energy is also observed with the parallel powertrain at below rated wind speed and failure states.The results shown in this thesis will be useful for offshore wind farm developers, operators
Date of Award7 Jun 2019
Original languageEnglish
Awarding Institution
  • University Of Strathclyde
SponsorsUniversity of Strathclyde
SupervisorAlasdair McDonald (Supervisor) & David Infield (Supervisor)

Cite this