The Oil and Gas (O&G) industry demands high investment costs and operational expenditures. Its profitability depends on systems reliability, availability and maintainability (RAM). The earlier is the value identification in an asset design the better is the value realization in following phases, operation included. RAM analyses estimate the operational efficiency of non-existing assets through Monte-Carlo Simulations (MCS) of reliability block diagrams with the equipment information from similar existing units. A comprehensive analysis of the total cost of ownership can be obtained using the costs structures and the RAM analysis results by means of Life cycle costing (LCC), but they are time consumers activities and sensitivity analyses simulations are needed later. The development of a technique to anticipate the synthesis of optimized options is necessary and this paper proposes a methodology to improve the value identification in early design that also can contribute to a design schedule reduction. To optimize the system arrangement an association of MCS, a tool to estimate stochastic variables, with the Non-dominated Sorting Genetic Algorithm (NSGA-II) method, which deals with combinatorial problems, is made. This combination produces Pareto fronts with optimal solutions. Expert opinion can provide pseudo-spatial data to complement RAM. It contains subjective information of anomalies/deviations in similar units, usually not recorded, but that are part of the expert knowledge. Using pseudo-spatial data, Bayesian inference can estimate anomalies in non-existing units.