Summary: The upstream sector of the Oil and Gas (O&G) industry is recognized by its capital-intensive projects and complex and hazardous associated recovery and production processes, thus susceptible for large and financially damaging accidents. In this context, to avoid the risk and impact of high expenses, O&G companies usually acquire insurance contracts. In practice, although the contract format is typically pre-specified, its parameter magnitudes can be adjusted aiming at maximizing the company total wealth. Therefore, this work proposes a holistic methodology to assess the optimal parameter specification of an insurance contract in the upstream sector of the O&G industry. A non-convex stochastic optimization problem is constructed aiming at maximizing a risk-adjusted measure of the policyholder total wealth. The modeling takes into account the uncertainty on the financial loss of an accident by making use of the safety barriers and precursor information framework. The non-convex optimization problem is cast as an equivalent mixed-integer linear programming problem by combining the scenario-based representation approach with a set of binary reformulation procedures. We illustrate the applicability of the proposed methodology with a set of numerical experiments. In a nutshell, we found that the proposed parameter specification methodology resulted in greater predictability when compared to two quantile-based specification policies and an uninsured company. In fact, the second best policy presented a standard deviation 103% higher than the proposed methodology. Furthermore, the model also provided greater protection against extreme events, since the second best policy presented a Conditional Value-at-Risk 41% higher than the proposed methodology.