Accurate reliability prediction for tidal turbines is challenging due to scarce reliability data. To
achieve commercialization, it is widely acknowledged that reductions in maintenance costs are vital and robust
component reliability assessments can help drive this. For established technologies, reliability prediction either
...
Accurate reliability prediction for tidal turbines is challenging due to scarce reliability data. To
achieve commercialization, it is widely acknowledged that reductions in maintenance costs are vital and robust
component reliability assessments can help drive this. For established technologies, reliability prediction either
involves a statistical assessment of historical failure data, or a physics of failure approach based on dedicated
accelerated testing. However, for low/mid Technology Readiness Level tidal developers these common
approaches are difficult. Thus, developers require a method of making reliability predictions for components
in the absence of tidal turbine specific failure data and physical testing results. This paper presents a failure
rate model for a tidal turbine pitch system using empirical Physics of Failure equations, with associated
uncertainties. Critical component design parameters are determined and their effects on the failure rate
investigated via a sensitivity analysis. The modelled failure rate is then compared with wind turbine failure
data from a series of turbines. The tidal turbine failure rate is approximately 50% lower, however high
reliability requirements mean this is unlikely to be acceptable. The developed model can assist turbine
developers in estimating failure rates and determining reliability critical design parameters for the failure
critical pitch system.