Page 125 - Gas Purification 5E

Alkanolamines for Hj>drogen Sulfide and Carbon Dioxide Removal 11 5

achieve adequate recovery of amine vapors. Typical DEA and MDEA stripping columns use
two to four trays, while MEA systems use four to six trays above the feed point Equilibrium
conditions alone would indicate that the above numbers are overly conservative; however,
the trays above the feed point serve to remove droplets of amine solution, which may be
entrained by foaming or jetting action, as well as amine vapor.
Overall Gas Absorption Coefficients

Early studies on the absorption of C02 and H2S were reported in terms of the overall gas
absorption coefficient (&a) and the Murphree vapor phase tray efficiency (Em) for packed
and tray columns, respectively. Both approaches were based on the assumption that the
absorption rate is controlled by the vaporfiquid equilibrium, which is an oversimplification
when a chemical reaction occurs in the liquid. However, the approaches are very simple to
use and can be helpful when comparing data for the same amine and acid gas at similar ow-
ating conditions.
Early experimental studies of the rate of absorption of C02 and H2S in alkanolamines in
packed towers were reported by Cryder and Maloney (1941), Gregory and Scharmann
(1937), Wainwright et al. (1952), Benson et al. (1954, 1956), Teller and Ford (1958),
Leibush and Shneerson (1950), Shneerson and Leibush (1946), and Eckart et al. (1967).
Packed column performance data on the absorption of HzS in MDEA solutions in the pres-
ence of C02 were presented by Frazier and Kohl (1950) and Kohl (1951). Much of the early
work on acid gas absorption in alkalis and amines was reviewed by Danckwerts and ShamLa
(1966), who proposed design procedures based on fundamental concepts.
Most of the recent studies on the development of design techniques for packed towers have
been based on the use of an enhancement factor to account for the effect of chemical reaction
on the liquid phase mass transfer. The empirical approach using &a based on overall gasfliq-
uid equilibria is losing favor, but still occasionally used. Strigle (1994), for example, describes
the use of &a values for amine plant design. He provides Koa values for the absorption of
Cor! by NaOH solution and notes that a 3 N MEA solution will produce a Koa for C02
absorption about twice that of 1 N NaOH under the same conditions; the &a for C02 absorp-
tion is about 40% of that for H$ absorption under the same conditions; and the &a for
absorption into DEA solution is only about 50 to 60% of that for MEA solution of the same
normality. The reported + values must be used with extreme caution because of the strong
effects of variables such as acid gas partial pressure on the mass transfer coefficient.

Theoretical Stages and Stage Efficiencies

The theoretical stage approach is still employed in several design procedures. It is particu-
larly useful for the design of plants for C02 removal where a high removal efficiency is
desired, and for the design of MEA stripping columns, where high tray efficiencies are
encountered. Gagliardi et al. (1989), for example, suggest that contactor and strippers be
sized by graphically determining de required number of theoretical stages, and then apply-
ing appropriate tray efficiency or HETP values. They base their design procedure on Air
Products & Chemicals, hc.'s (APCI) extensive experience with MEA plants that remove
CO, from high-pressure hydrogen and synthesis gas.
APCI plants operate with C02 partial pressures from 5 to 75 psia, C02 removals to less
than 100 ppmv, and MEA concentrations up to 32%. The systems are designed with reflux
ratios of 1.0 to 1.2 moles H20 per mole of C02 in the stripper overhead; lean solution load-