Modelling and Optimization of Natural Gas Dehydration System Using Triethylene Glycol

The problem with water vapour in natural gas stream, threaten the process facilities if the dew point temperature is not properly managed. Hydrate formation is inevitable at temperatures below the dew point. It becomes very important to reduce the water content in the gas stream to below or within the acceptable limit of 6-7lb/MMSCFD. There are many methods that can be used to reduce the water content in the natural gas stream of which adsorption and absorption are one of them, in this work absorption was employed. Natural gas composition analysis was carried out and Industrial data of a natural gas dehydrating plant was obtained and simulated using Aspen Hysys software. Process conditions of 72.6 barg and 38oC and gas flow rate of 10 lb/mmscfd, were inputted into the software and simulated. Different Triethylene Glycol (TEG) flow rates were used for the simulation. Results obtained show that for a TEG of 3.5m3/h (0.1480 kgmole/h), the water content in the dry gas was 6.6333 lb/MMSCF which is within the acceptable water limit and when the TEG flow rate was increased up to 15.7m3/h (0.6640kgmole/h) the water content in the dry gas was further reduced to 1.0930 lb/MMSCF from an initial value of 19.84lb/MMSCF. For the first simulation, a reboiled still column was used and 88% of TEG was recovered and 12% was lost with water vapour at the top of the still column but after the second simulation where a recovery separator, still column with reboiler and reflux condenser were used, 99.98% of the TEG was recovered which was compared with other methods in the industry. This implies that for a higher purity of TEG to be achieved, a recovery separator and a reflux still column needs to be used because increasing the reflux reduces glycol loss. The sum of the total capital cost and glycol loss per year is 3,873,590 USD for simulation 1, while there is a negligible TEG loss in simulation 2 so the total capital cost is 3,863,380 USD and the two simulations have the same energy consumption of 114,300KJ/h.