In a past project I worked on, one of the tasks I was assigned to was to validate the simulation model of a Sour Water Stripping unit against the stream data provided in the Licensor Package. The simulation was conducted on Aspen Plus with ELECNRTL property package employed.
The constraints at the bottom of the Sour Water Stripping column were: H2S < 10 ppmw; NH3 < 35 ppmw, which could be achieved by allowing adjustment of the reboiler duty during the setup of the simulation model. Hence, the results were validated through the comparison against those provided in the Licensor Package.
At the overhead of the column, cooling of the vapour outlet relied on the cooled reflux which was essentially the side-draw liquid cooled by an overhead condenser. The flow rate of this stream and the number of theoretical stages were also adjusted such that both the overhead temperature and the flow rate were sufficiently close to those provided in the Licensor Package without exceeding the constraints at the bottom of the Sour Water Stripping column.
An online source stated that sour water stripping is a low-efficiency process and hence suggested the use of an overall column efficiency of 50%. The overall column efficiency arising from this validation task reveals that the figure falls within the range of 25% – 30% which is much more conservative than the suggestion.
The results of the converged simulation model were found to agree very well with the data provided by the Licensor. It was also found that stripping of NH3 is the determining factor in reboiler duty calculation as the pH of sour water feed was acidic which favoured the stripping of H2S.