Effective Amine Technology: Controlling Selectivity, Increasing Slip, and Reducing Sulfur 1

Twenty years ago, achieving the gas treating goals regularly reached today was virtually unimaginable. The norm was complete removal of CO2 and H2S using tried and true amines like MEA, DEA, DGA, and DIPA. Then MDEA began to be used and it was found possible to leave a certain amount of CO2 in the treated gas (and get paid for it). MDEA found quite extensive use in operations such as tail gas treating where efficient H2S removal and high CO2 rejection were the goals. In the mid-80s mixtures of amines began to be used, formed mostly by adding varying amounts of reactive amines to MDEA. Thus it became possible to control selectivity and to produce a gas having a specified composition with respect to both H2S and CO2. A rather interesting additive was a mineral acid or other substance which was capable of neutralizing some of the MDEA. In addition, a class of sterically-hindered amines was developed to target selective removal applications—we will not deal with these sterically-hindered amines here. The focus of the paper is the development of a fundamental understanding of how modern amine solvent technologies work and validation of this understanding through computer simulation and, where possible, comparison with actual plant performance data. This paper first discusses how MDEA (and certain other amines) achieve good CO2 slip, while other amines do not. The development embraces chemical reactivity and equilibrium considerations, as well as the influence of detailed equipment design parameters. The engineering basis for our understanding is validated against actual plant performance data, including detailed tower temperature profiles. The use of reactive amines as activators or promoters for MDEA is considered next. We discuss in detail the basic mechanism by which these promoters function and the various ways in which they can influence selectivity. The basic understanding developed in the paper is illustrated with detailed tower simulation. Finally, we examine how partial neutralization (inactivation) of the amine can be used to achieve the seemingly astonishing result of much lower H2S residual levels in the treated gas. Again, the approach taken is fundamental, this time at the level of shifting reaction equilibria and its influence on vapor-liquid equilibrium. Although no plant performance data using a partially neutralized solvent are currently available to us, we examine fundamental modeling predictions in terms of projected treating plant performance.