Abstract
The objective of the research project was to investigate the use (incorporating the environmental impacts) of ammonia and amines to minimise acidic corrosion in the overhead system of a Crude Distillation Unit without the danger of these additives forming corrosive hydrochloride deposits. It is hoped that the information obtained will enable refiners to select amines or amine mixtures that reduce both acidic and hydrochloride deposit corrosion to a level acceptable in today's refinery. Additonally, a framework has been developed to allow the refiner to select which of these chemicals has the lowest environmental impact in a particular refinery. A Computer Model that accurately predicts the pH profile of the acidic condensate of the overhead condenser system was developed. Although many similar models exist in the literature and in the marketplace, this model incorporates two novel aspects - the raising of the water dew point temperature by hydrochloric acid and the partitioning of amines between the water and hydrocarbon phases. The Knudsen Effusion technique was used to determine previously unknown vapour pressure data for the hydrochlorides of several commercially used amines and to validate existing data for ammonium chloride. This data can be used to predict the temperatures in the Crude Distillation Unit at which hydrochloride deposition occurs. With this information, the refiner should now be able to minimize the danger of hydrochloride salt corrosion in the overhead system of the Crude Distillation Unit. Furthermore, by combining the amine pH profiles from the computer model with their hydrochloride vapour pressure data it was possible to identify two amines (DMIPA and MOP A) that had a superior ability to neutralise the acidic condensate without the danger of hydrochloride deposition. A laboratory scale chamber was designed that accurately reflects the deposition of ammonium chloride and amine hydrochlorides in the overhead system of the Crude Distillation Unit. This Deposition Chamber was used firstly to validate assumptions made with the Knudsen Effusion data and secondly to check whether pure vapour pressure data could accurately predict the actual hydrochloride deposition temperatures in the Crude Distillation Unit. It also showed that the use of pure vapour pressure data to predict amine hydrochloride deposition temperatures in the Crude Distillation Unit is valid for structurally dissimilar amines. However, the observed hydrochloride deposition temperature of a mixture of two structurally similar amines was found to be approximately 5°C higher than the theoretical deposition temperature predicted from the vapour pressure data, indicating that they should not be used together.