Objective/Scope When operating an Electric Submersible Pump (ESP) in a location with high installation or worker costs, the installation can cost much more than the ESP itself. Having a reliable ESP with good average run life is critical to the operational economics. In sweet or benign reservoirs, ESP run life can easily exceed five (5) years through careful design, selection, manufacture, installation and operational practices. When taking an ESP system with a proven track record into sour environments, the average run life deteriorates significantly and correlates with the partial pressure of the H2S in the produced fluid. This proven track record is important to bear in mind because an ESP system that has inherent reliability issues will fail of its own accord long before the effects of H2S become obvious which makes failure analysis challenging. In a region with both sweet and sour fields coupled with high intervention costs, it is vital that ESP run life is reliable and predictable in any field. Years of dismantle inspection failure analysis (DIFA) on all ESPs has allowed identification of modes of failure attributed to H2S, prioritized those issues and led to the development of a high integrity ESP system to combat the modes of failure allowing ESPs in sour fields to improve run life to be on par with those in sweet fields. This paper will describe the specific problems H2S causes an ESP and what was done to resolve them. Methods, Procedures, Process H2S will cause significant damage to an ESP system that may eventually cause a failure. If the H2S partial pressure is low, the damage may not reduce the run life significantly or at least lower it enough to cause significant inconvenience and high operational cost. If an ESP can run five (5) to seven (7) years, on average, without failure, this would be more than adequate in most cases. While having an ESP that never fails might seem to be a worthy goal, it isn't really practical. Changing reservoir conditions or production strategies with time will make it necessary to pull and replace an ESP at some point. While a variable speed drive (VSD) does expand the useful window of operation of an ESP, it only helps up to a point. Fortunately H2S does not appear to significantly impact the ESP run life performance. To date, corrosion resistant coating (CRC) on carbon steel has been effective as long as the coating is intact. If CRC is to be applied to a housing, it is necessary that the coating be applied as soon after bead or shot blasting as possible. Even a few extra minutes of delay will allow an oxide layer to form on the newly cleaned housing surface making it impossible for the coating to adhere to it. H2S attacks the ESP cable, seal section and motor in specific ways. For this paper, the ESP completion discussed has a deep set, non-vented packer located approximately 200 feet above the pump discharge used to prolong the life of the production casing. Based on empirical data, the most significant reduction of run life is failure of the MLE system which includes the packer penetrator lower connector, the MLE pothead and the MLE lead jacketed cable itself in that order of importance. While some penetrator manufacturers perform better than others, nobody has yet designed an H2S proof system. Whether the penetrator is factory molded or field attachable, H2S will eventually cause it to fail. Without going into the specifics of one design over another, probably the most important factor impacting the reliability of the penetrator connector is the length of the leakage path from the sour production fluid to the copper of the cable and connectors. The longer the distance, the longer the useful life of the connector.