Turning dilute acid gas streams into high-quality sulphur plant feeds
In many instances it is not possible to remove the hydrogen sulphide from a gas stream and simultaneously produce a satisfactory sulphur plant feed using conventional gas treating technology.
Tofik K Khanmamedov, TKK Company
Ralph H Weiland, Optimized Gas Treating, Inc
Viewed : 2515
In these cases the H2S, once removed from the gas, can present a serious disposal problem because its concentration in the product acid gas stream is too low to be fed to a sulphur plant but is far too high to be incinerated. One option is to treat selectively the acid gas stream itself and raise its H2S content to a level suitable for a sulphur recovery unit. However, conventional enriching of the acid gas stream still may fail to yield satisfactory results. This paper analyses how HighSulf, a family of new and exciting processing concepts, can produce a Claus plant feed stock of excellent quality even from sour gas streams that would otherwise present disposal problems. HighSulf is also examined as a means of producing a high quality Claus feed directly from a raw gas such as sulphur plant tail gas. The analysis is done using the mass transfer rate approach afforded by the ProTreat amine simulator.
Sulphur plants operate best with a feed containing 55% or more H2S. The balance of the SRU feed stream is CO2 and water, with small amounts of hydrocarbons, inerts, and other components. Lower concentrations of H2S can be processed only with increasing levels of sulphur plant complexity, larger equipment, and higher cost. Streams having less than 32% or so H2S are near the lower limit for a straight-through Claus process.
There are numerous sources of acid gas streams that are too dilute for sulphur recovery in Claus plants but too concentrated to vent. In plants treating high CO2:H2S ratio gas, adequate CO2 slip even using the most selective solvent often cannot produce an SRU feed of sufficient quality. For example, conventional treatment of a high pressure gas containing one or two percent CO2 and a few 100s of ppmv H2S using even the most selective amine available cannot produce an acid gas stream with more than a few mol% H2S, completely inadequate for a conventional Claus SRU. Even in plants with a moderate CO2:H2S ratio of say 4:1, if complete acid gas removal is necessary (LNG for example), using a single contactor will necessarily produce an acid gas stream containing only 20% H2S.
HighSulf is a patented process strategy (Khanmamedov, 1996, 1998, 2003) that can be applied incrementally in amine treating plants to increase the H2S concentration in the off-gas from the regenerator and produce an increasingly high quality Claus sulphur plant feed. A conventional amine treating plant (in this instance, a tail gas treater, or TGTU) is shown in simplified form in Figure 1. Tail gas is sulphur plant effluent, essentially a stream of H2S and CO2 diluted with the nitrogen remaining after the combustion and catalytic processes that convert H2S into elemental sulphur in a Claus sulphur plant. Frequently, the off-gas is of only very marginal quality as a sulphur plant feed. There are two approaches to making it more suitable. One is to apply various levels of HighSulf directly to the TGTU. The other is to reprocess the off-gas in another, smaller amine plant and produce by selective absorption a second off-gas of higher quality. This secondary treating unit (also called an acid gas enrichment, or AGE, unit) might apply HighSulf strategy, too. HighSulf recognises that the higher the H2S content of the gas being treated in an amine unit, the greater will be the H2S concentration in the off-gas from the regenerator. The family of HighSulf processes produce a more concentrated product stream, as discussed by Khanmamedov (1996, 1998, 2000, 2003).
The role of selectivity
Achieving the maximum selectivity for H2S over CO2 is done by using the right solvent under the right process conditions in the right equipment—it is paramount to successful tail gas treatment and to the operation of AGE units. The perfect process in these applications removes all the H2S and none of the CO2 from the raw gas because an acid gas consisting entirely of wet H2S is produced upon solvent regeneration. This would be perfect selectivity. Detailed discussions of selectivity have been presented in many places including Anderson et al. (1992), Khanmamedov (1998, 2000), and Weiland et al. (2003), so only a brief review is given here.
The equilibrium solubilities of H2S and CO2 in selective solvents such as methyldiethanolamine (MDEA) do not differ radically. In other words, chemical solvents do not have great thermodynamic selectivity so vapour-liquid equilibrium plays as critical a role as one might suppose. The differences in absorption rates are really determined partly by reaction kinetics, and partly by the hydraulic and mass transfer characteristics of the contacting equipment vis à vis the relative magnitudes of gas- and liquid-side mass transfer coefficients. The mass transfer characteristics of the contacting equipment is a factor that has been largely overlooked by many practitioners, possibly because of poor understanding of separation equipment from the standpoint of its inherent mass transfer rates and what affects them.
Tertiary amines, of which MDEA is the most commonly used in selective treating, react with H2S and CO2 at chemical rates that are at opposite ends of the spectrum. H2S absorption is accompanied by an instantaneous proton transfer reaction associated with H2S dissociation and amine protonation. CO2, on the other hand, reacts very slowly indeed, forming bicarbonate ion by reaction with water, but no amine carbamate. Thus, from a reaction kinetic standpoint, MDEA is highly selective for H2S.
As devices for carrying out mass transfer, trays and packing (both random and structured) behave quite differently hydraulically. The most obvious reason for this difference is that a properly operating tray (froth regime, not spray regime) usually has the liquid phase continuous and the gas phase dispersed, while in packing the opposite is generally true. The liquid flows over packing in films that are relatively quiescent compared to the highly agitated state of the liquid flowing across trays. Vapour flows are quite turbulent for both trays and packing. Therefore, these types of equipment should be expected to have different mass transfer characteristics, most pronounced with respect to their relative liquid-phase resistances, but also with respect to their vapour-phase resistances (to an extent that depends greatly on the exact internals involved). These differences in characteristics are important to establishing selectivity because the mass-transfer resistance to H2S absorption is primarily in the gas phase, while for CO2 it is in the liquid phase. The result is that to some extent selectivity can be controlled by controlling the relative resistances to mass transfer offered by the two phases through the judicious selection of tower internals. Phase resistances are functions of the type (trays, random packing, structured packing) and mechanical details (tray passes, weir heights, packing brand, size, crimp angle, etc.) of the contacting equipment itself as well as the way it is operated hydraulically (flow rates and phase properties that depend on temperature and pressure). Tail gas treating and acid gas enrichment are processes whose performance is completely dependent on relative rates of mass transfer. Only a true heat- and mass-transfer-rate based model such as ProTreat stands any realistic chance of reliably predicting performance in a specific piece of equipment. Reliable simulations cannot be done unless the simulation tool itself is cognizant of the mass transfer behaviour of the internals, and the engineer doing the calculations also keeps in mind the hydraulic regime in which the column is operating (e.g., spray versus froth regimes for trays).
Add your rating:
Current Rating: 2