1: \begin{abstract}

2: 

3: Environmental contours offer descriptions of extreme environmental conditions commonly applied for reliability analysis in the early design phase of marine structures. The definition of these contours base themselves on restricting the failure time, specifically the first hitting time of the underlying environmental process into a usually unknown failure set. \PURP{(There is too much information crammed into each sentence)}

4: 

5: Certain technical issues arise when computing such contours under the assumption that the failure sets are convex. In this setting, the contours are usually constructed as the boundary of an intersection of half-spaces, and with the requirement that all the corresponding hyper-planes must overlap with the contour. However, in some cases the hyper-planes become separated from the intersection, and as a result fails to properly restrict the failure time. This can be caused by e.g.\ estimation errors of the environmental distributions or by the dependence structure between the different environmental factors. These issues have been previously studied, but no method for the computation of minimal contours circumventing this problem exists in the literature. \PURP{(This is impossible to follow)}

6: 

7: This paper presents a novel algorithm for computing convex environmental contours in order to solve this issue. We will consider a minimisation problem where the smallest contour in the sense of mean width is computed while ensuring the properties needed to provide the desired restrictions on the failure time is preserved. Additionally, numerical examples are provided to illustrate the presented methods. Lastly, rigorous convergence analysis is performed, both in terms of convergence of mean widths and in the sense of the Hausdorff metric. \PURP{(We should mention the use of linear programming)}

8: \end{abstract}

9: