An Innovative Approach to Optimizing Design of HP/HT Well Tubular Strings

Abstract

Abstract In traditional well completion design, the size and geometric configuration of various structural tubular components are determined by a set of stress limits based on the material properties and type of loading under the worst conditions expected during the projected service life. Uncertainties associated with design variables are presumably compensated by introducing safety factors derived primarily from past experience. It is intuitively obvious that a lower safety factor leads to a design with higher risk. However, those safety factors give no quantitative indication of risk assessment of an intended design, and no insight into the degree of risk. Moreover, this approach often leads to over-conservative design and therefore excessive cost, or over-risk design, which can cause failure in tubular components. This paper introduces an innovative approach to fill this gap. This approach integrates design of experiments (DOE) and stochastic study into the design process. It explicitly considers the variability and uncertainty of each design variable in probabilistic terms, and allows the assessment of probability of failure of the design; thus it is able to help achieve better designs and higher-reliability completion tools. Design of experiments provides an efficient methodology to study the effects of design changes and identify the most influential design variables, and, therefore, provides clear design performance improvement direction for complicated designs. Stochastic study addresses sensitivity of stress or strain limits, which are tied to the life cycle of the tools, under the influence of variability of design parameters, and provides quantitative estimation of probability of success of the optimized design. A deepwater high-pressure/high temperature (HP/HT) tubular string design example will illustrate this new approach. The analysis results indicate this approach leads to a more robust design.