Page 166 - Gas Purification 5E

154 Gas PzmiJication

uct, and the rate is high enough to make MEA unsuitable as an absorbent for gases with high
concentrations of COS. DEA also undergoes both hydrolysis and direct reaction with COS.
As with MEA, the direct reaction results in the formation of a nonregenerable compound
however, the rate of formation is extremely slow, so DEA can be used effectively to treat gas
streams containing COS. According to Butwell et al. (1982), only about 28 of the COS in
the feed gas stream reacts irreversibly with DEA; while overall COS removal efficiencies of
7040% are attainable, primarily by hydrolysis. They claim that even higher levels of COS
removal can be attained with DEA solutions by eliminating the competition of HIS and C02
for reaction sites; Le., contacting the gas with lean DEA solution in a second stage after H,S
and C02 have been removed in the first stage. Additional data on the reactions of COS with
MEA and DEA are provided by Peace et al. (1961).
DIPA is reported to react with COS even more slowly than DEA (Danckwerts and Sharma,
1966). Aqueous DIPA solutions can thmfore be used for treating gases containing COS with-
out serious deterioration. Ac-g to Klein (1970), DIPA, as used in the ADP process, is
suitable for partial removal of COS from gas streams. The Sulfinol process, which uses a mix-
ture of DPA and Sulfolane, reportedly provides very complete COS removal (see Chapter 14).
DGA is a primary amine which is chemically related to MEA and also reacts quite rapidly
with COS. Fortunately, the main product of the reaction decomposes under conditions
encountered in a DGA solution reclaimer allowing DGA solutions to be used for gases con-
taining COS. Furthermore, the process can be operated to enhance COS hydrolysis and
thereby provide effective COS removal. As previously mentioned, two key factors in COS
hydrolysis are temperature and contact time. Two key features of the Fluor Daniel Improved
Econamine Process, which is claimed to provide efficient COS removal, are operating at ele-
vated temperatures and providing adequate contact time. Huval and van de Venne (1981)
report on experience with 5 large DGA plants operating in Saudi Arabia. These plants oper-
ated with lean amine temperatures as high as 150°F. One typical unit averaged about 90%
COS removal (see Table 2-26).
Singh and Bullin (1988) examined the kinetics of the reaction between COS and DGA
over a temperature range of 307 to 322°K (93 to 120°F). They concluded that the reaction is
controlled by hydrolysis and that the DGA has a catalytic effect on the hydrolysis reaction.
The reaction was found to follow a second order rate equation; first order in COS and first
order in DGA. A competing reaction occurs to form N,N’bis (hydroxyethoxyethyl) thiourea
(BHEEU) which, fortunately, reacts with water at high temperatures to regenerate DGA.
The absorption of COS in aqueous solutions of MDEA was studied by Al-Ghawas et al.
(1988). They concluded that the overall reaction between COS and MDEA is given by the
following equation:

R$J + HzO + COS = R3NH+ + HCO$ (2-46)

They performed laboratory tests to measure the rate of COS absorption in MDEA and
used a model based on the penetration theory to calculate the kinetic rate constant for equa-
tion 2-46. The results should be of value in rigorous design calculations to predict the frac-
tion of COS in a feed stream that will be absorbed in a commercial MDEA contactor.
Rahman et al. (1989) conducted an experimental study to determine the products of reac-
tion between COS and several different amines. The protonated amine thiocarbamate salt
was detected in the reaction products of COS with MEA, DEA, and DGA. The DIPA thio-
carbamate salt could not be detected, possibly because the DIPA-COS reaction was too slow
to provide enough reaction product for detection by the method employed. The tertiary