Abstract
The increase in natural gas production has allowed more opportunities for equipment to be exposed to corrosive contaminants. Field monitoring tools provide a measurement of the current corrosion rates as they occur, but cannot predict the impact of future operational changes. If a plant wants to change operations or gas feed, it would be safer to be prepared for corrosion risks in advance with a prediction model rather than experience a crisis if unacceptable corrosion rates occur.
A process simulation modeling tool was recently used to help a customer predict general acid corrosion risks in a liquefied natural gas (LNG) plant that was in the design phase. The customer provided flows, temperatures, estimated gas feed compositions, and other design specifications, which were applied to the model. The only acidic components in the gas were carbon dioxide and hydrogen sulfide, which are soluble in water. There were no basic species present, so any stream with free water was acidic. The model determined which streams have free water and the quantity of corrosion chemical needed to minimize general acid corrosion. Based on model results, corrosion chemical injections, corrosion monitoring equipment, and sampling recommendations were made. The customer utilized these recommendations and added these features to their design. This paper presents how a process simulation model can predict where corrosion will occur and how the model was applied to an LNG plant design.
