Lateral analysis of gas reinjection compressors with uncertainties on bearings coefficients

Abstract

The tilting pad bearings add stiffness and are the main source of damping to a rotor-impeller-bearing system. The bearing stiffness and damping coefficients have uncertanties which impacts on these machines mechanical integrity design. This thesis studies two compressors used in gas reinjection at oil reservoirs. An stochastic model is constructed to simulate the compressors rotor dynamic behavior and a Thermo-Elasto-Hydrodynamic model is applied to the tilting pad bearings with a level of uncertanties related to the coefficients calculation. These stochastic models provide the industry a useful tool to antecipate whether the machine is reliable to be manufactured and therefore to avoid the costs of unsucessfull factory acceptance tests and the machine life cycle will not be compromised. Hence, some scenarios of unbalance magnitudes were built according to a variety of severity levels feasible in oil and gas industry. The API (American Petroluem institute) standards are used as rotordynamic acceptance criteria. The reliability methods, First-Order Reliability Method and Importance Sampling, are used to evaluate these machines different scenarios in the calculation of the probability of not meeting API 617 [4] criteria and to save computational effort.