This paper describes the methodology behind the partial safety factors and design fatigue factors of the DNV GL standard for certification of tidal turbines. The standard follows a risk based approach that allows the adjustment of the design requirements to the identified risk level of the system. The work undertaken in this study has involved the identification of the uncertainties during the design process. Tidal turbines are located in highly energetic sites which are very difficult to characterise, hence site conditions are one of the largest sources of uncertainty. Key parameters like turbine inflow conditions or predictions of extreme values are still grey areas. Load simulation tools are still quite uncertain and are often dependent on the experience of the people running them. Both partial safety factors and design fatigue factors have been calibrated to different target safety levels with due account for the uncertainties introduced by load models as well as site characterisation. The different target safety levels have been selected in accordance to the risk associated to structural failures of tidal turbine support structures and aim for a more streamlined design from the safety requirements point of view.

Tidal energy is nowadays one of the fastest growing types of marine renewable energy. In particular, horizontal axis tidal turbines (HATT) are the most advanced designs and the most appropriate for standardisation. It is however, in the interest of the industry to provide a set of standard practises in order to help in the process of designing this type of marine converters. DNV GL is producing this standard with the support of Alstom and as part of the ReDAPT project commissioned and co-funded by the ETI (Energy Technologies Institute). The work undertaken by DNV GL to produce this standard involves the identification of the uncertainties that designers need to address during the design process. Unlike other marine structures, HATTs are usually located in very energetic areas where no other marine structure has been before. Site characterisation is one of the largest sources of uncertainty e.g. turbulence. Key inputs like turbine inflow conditions or predictions of extreme values are still grey areas due to the limited site measurements and the uncertainty of the metocean models. Numerical models of HATTs are still quite uncertain often dependent on experience of the people running them. As part of ReDAPT project there is an ongoing effort in validation and evaluation of the accuracy of these numerical models and some of the results are used in this calibration study. The new standard for HATTs deals with the loading uncertainty in a whole new way by introducing a new parameter that is added to the traditional partial safety factors for loads. This new standard uses the traditional safety factors from the offshore structures standards and allows changing them based on the level of uncertainty that was introduced during the design process. This paper describes the process of calibration of the partial safety factors in ULS for loads of HATTs that was part of the work in the creation of the new standard. The reliability based calibration involved the formulation of failure criteria, the identification of stochastic variables in the failure criteria, calculation of reliability against failure and ultimately the new set of partial safety factors for loads and a methodology for adjustment of the factors in a case by case basis.

Tidal energy is nowadays one of the fastest growing types of marine renewable energy. In particular, Horizontal Axis Tidal Turbines (HATTs) are the most advanced designs and the most appropriate for standardization. This paper presents a review of actual design criteria focusing on the identification of the uncertainties that technology developers need to address during the design process. Key environmental parameters like turbine inflow conditions or predictions of extreme values are still grey areas due to the lack of site measurements and the uncertainty in metocean model predictions. A comparison of turbulence intensity characterization using different tools and at different points in time shows the uncertainty in the prediction of this parameter. Numerical models of HATTs are still quite uncertain, often dependent on experience of the people running them. In the reliability-based calibration of partial safety factors, the uncertainties need to be reflected on the limit state formulation. This paper analyses the different types of uncertainties present in the limit state equation. These uncertainties are assessed in terms of stochastic variables in the limit state equation. In some cases, advantage can be taken from the experience from offshore wind and oil and gas industries. Tidal turbines have a mixture of the uncertainties present in both industries with regard to partial safety factor calibration.