Summary Pilot tests have been conducted at a Levelland (TX) Pilot tests have been conducted at a Levelland (TX) gasoline plant to evaluate triethanolamine (TEA), a tertiary amine, for bulk CO2 separation from a hydrocarbon-rich gas, which simulates that produced during CO2 miscible flooding. Many operating parameters were evaluated, including contactor pressure, amine temperature, amine circulation rate and concentration, CO2 content in the feed gas, and various methods of amine regeneration. Bulk CO2 removal from streams containing from 37 to 67% CO2 was demonstrated successfully, with minimum input of heat and good hydrocarbon selectivity. Introduction The current emphasis in EOR is toward CO2 miscible flooding. The gas produced from the floods contains large quantities of CO2 with significant amounts of hydrocarbons. It is desirable to recover the CO2 for reinjection and the hydrocarbons for further processing. An economic comparison indicated the TEA CO2-removal process to be a viable alternative for achieving bulk process to be a viable alternative for achieving bulk removal of CO2 from the produced gas. To evaluate feasibility of the TEA process on a gas similar to that produced from miscible flooding, a pilot test was conducted. produced from miscible flooding, a pilot test was conducted. Process Description Process Description A standard skid-mounted amine unit was adapted for pilot testing at Levelland. The basic unit was designed for an inlet gas flow of approximately 600 std m3/d] with an amine circulation rate of 60 gal/min [3.78 dm3/s]. Fig. 1 illustrates the basic process. The hydrocarbon feed for the pilot plant is a sweetened casinghead gas. This is combined with CO2 from storage to meet the required test conditions. The CO2-rich feed gas is contacted with TEA solution under pressure in the contactor. The concentration of CO2 in the contactor overhead depends on contacting pressure, amine temperature, amine rate, and feed gas pressure, amine temperature, amine rate, and feed gas composition. The rich amine from the base of the contactor is flashed in the stripper column to slightly above atmospheric pressure. The reduction in pressure releases the bulk of the pressure. The reduction in pressure releases the bulk of the dissolved CO2. Since the stripper is operated below the boiling point of the amine solution, the reboiler is used to make up point of the amine solution, the reboiler is used to make up only the latent heat of water vapor lost in the stripper overhead and sensible heat losses in the system. The lean amine is drawn from the reboiler, filtered, boosted to contacting pressure, and recycled to the contactor. pressure, and recycled to the contactor. Performance Prediction Performance Prediction The prediction of performance for the TEA process is based on available equilibrium data for CO2 solubility in aqueous TEA solution. The data are from two published sources:Mason and Dodge for CO2 partial pressures below atmospheric in 0.5, 1.0, 2.0, 3.5, and 5.0 normal TEA solutions at temperatures of 32 to 167F [0 to 75C] andLyudkovskaya and Leibush for CO2 partial pressures ranging from 40 to 600 psia [275.8 to 4137 kPa] partial pressures ranging from 40 to 600 psia [275.8 to 4137 kPa] for 0.5, 2.0, 3.5, and 5.0 normal solutions at temperatures of 77, 122, and 167F [25, 50, and 75C]. Mathematical models based on the available data for acid gas and hydrocarbon equilibrium solubility in monoethanolamine (MEA), diethanolamine (DEA), and TEA, as well as for solution physical properties, are incorporated in an in-house computer physical properties, are incorporated in an in-house computer program. The vapor/liquid equilibrium correlation of each amine/ program. The vapor/liquid equilibrium correlation of each amine/ CO2/water system is based on a chemical reaction equilibrium model. The program uses an iterative Newton-Raphson technique for heat and material balance calculations and converges all trial-and-error calculations to within 0.05% or less. The computer program has been tested with operating data from conventional MEA and DEA sweetening units, and agreement has been satisfactory. Therefore, this program was employed to simulate the TEA pilot plant operations over a range of important operating variables. Fig. 2 is a typical plot of net CO2 loadings (expressed as standard cubic feet of CO2 plot of net CO2 loadings (expressed as standard cubic feet of CO2 removed per gallon of solution) predicted for operation with 50 wt% TEA. Each curve applies for a given lean amine temperature and a partial pressure of CO2 in the feed gas. The loadings were computed for four theoretical trays in the absorber and stripper, representing 20% average tray efficiency for the 20 valve trays installed in each of the pilot plant columns. JPT p. 1603
Read full abstract
Apr 1, 2002
Juan P Gorrino + 2
Open Access