Simulation of TEG dehydration plants
MetadataShow full item record
The purpose of natural gas dehydration is to prevent condensation of water in production units and transport pipelines. Liquid water can cause severe problems such as corrosion and hydrate formation. The aim of this thesis is to evaluate various models used for simulation of triethylene glycol (TEG) dehydration plants by comparing simulation calculations to experimental and field data. Three different simulation tools are included in the evaluation: Pro/II (V9.1), Hysys (V8.3) and ProMax (V3.2). In Pro/II, the glycol package is applied for the simulations. Four different fluid packages are included in Hysys: Glycol Package (GP), Peng-Robinson (PR) and two versions of the Cubic-Plus-Association-model (CPA) developed at Technical University of Denmark (DTU, V3.8) and Statoil (NeqSim). Soave-Redlich-Kwong (SRK) and PR are included in ProMax. Statoil provided field data from Gullfaks A (March 2012) as a comparison for the simulation calculations. The dehydration unit on Gullfaks A utilizes an absorption process which consists of a contactor, a regenerator with condenser and reboiler, and a separate stripping column. TEG is used as absorbent. Relevant experimental data is also gathered and compared with the calculations done by the simulation tools. The thesis focuses on parameters relevant for gas dehydration: wet gas water content, dry gas water dew point, rich TEG composition, reboiler duty, lean TEG purity and TEG loss. Hysys CPA NeqSim proved to be the best suited fluid package for dehydration of natural gas using TEG as an absorbent. This fluid package provided results in good agreement with both experimental and field data for all relevant parameters. ProMax SRK and ProMax PR also provided accurate results, and only minor deviations were found such as a slightly high lean TEG purity. Both packages are well suited for dehydration of natural gas, and the simulations showed no significant difference between the two packages. Heating of TEG was shown to count for roughly 65 % of the calculated reboiler duty. Other heat consumptions were primarily related to evaporation of water (25 %). Pro/II and Hysys CPA DTU provided low reboiler duties. This was related to low calculated heat capacities of TEG. This was the only serious flaw discovered in Pro/II, making it better suited than both Hysys GP and Hysys PR, which provided inaccurate results for respectively wet gas water content and dry gas water dew point. The wet gas water content calculated by Hysys GP was 9 – 14 % lower than the field data. This was expected due to low calculated water content in methane. The low water content in wet gas in Hysys GP had a slight influence on the dry gas water dew point, reducing it by roughly 4 %. For the composition and flow rate of lean TEG given in the field data, the calculated dry gas water dew point was shown to be primarily dependent on the calculated water content in the vapor phase in equilibrium with TEG. Hysys CPA DTU provided consistently very low water dew points, while Hysys PR calculated values higher than both experimental and field data. Gases are soluble in TEG. This thesis studied the solubility of methane, ethane and CO2 in rich TEG. Hysys GP calculated the lowest solubility of these components in rich TEG and the largest solubility in binary simulations with TEG. These opposing results were shown to be related to the water content in rich TEG which decreased the solubility of especially methane and ethane significantly. No data were available as a comparison. However, considering the relatively low water content in rich TEG, the decrease was too large. The lean TEG purity from the reboiler was shown to be primarily dependent on the state of equilibrium between TEG and water. Hysys CPA DTU calculated a higher amount of water in the liquid phase, leading to lower purity from both reboiler and stripping column compared to the other fluid packages. The purity from the stripping column was lower than the measured value. In addition, Hysys CPA DTU calculated low circulation rate of TEG due to low TEG density. Consequently, Hysys CPA DTU was evaluated as the least suited fluid package for dehydration of natural gas of those included in this thesis.