On how to understand and present the uncertainties in production assurance analyses, with a case study related to a subsea production system

Production assurance analyses of production systems are in practice typically carried out using flow network modelling and Monte Carlo simulations. Based on the network and probability distribution assumptions for equipment lifetime and restoration time, the simulation tool produces predictions/estimates and uncertainty distributions of the production availability, which is defined as the ratio of production to planned production, or any other reference level, over a specified period of time. To adequately communicate the results from the analyses, it is essential that there is in place a framework which clarifies how to understand the concepts introduced, including the uncertainty distributions produced. Some key elements of such a conceptual framework are well established in the industry, for example the use of probability models to represent the stochastic variation related to lifetimes and restoration times. However an overall framework linking this variation, as well as "model uncertainties", to the epistemic uncertainty distribution for the output production availability, has been lacking. The purpose of the present paper is to present such a framework, and in this way provide new insights to and guidelines on how to understand and present the uncertainties in practical production assurance analyses. An example related to a subsea production system is used to illustrate the framework and the guidelines