Page 144 - Gas Purification 5E

134 Gas plirijkation

tures) by the use of the Reed and Wood correlation, Figure 2-88. Unfortunately, correlations
similar to those generated for MEA by Fitzgerald and Richdson have not been developed
for other amines, and lean solution loading estimates are usually based on the limited plant
operating data that are available.
DEA is a weaker base than MEA. As a result, the same stripping conditions will produce a
lower lean solution loading in DEA than in MEA solutions. Alternatively, a similar level of
stripping can be accomplished with less steam consumption by the DEA system. In fact,
improved steam economy is one reason many plants have been converted from MEA to
DEA. A survey of 24 DEA plants, conducted by Smith and Younger (1972), suggests that
1.2 lb steardgallon of solution is adequate to produce a lean solution containing less than 40
,gains HzS/gallon (about 0.007 mole HzS/mole DEA) under a variety of DEA plant operat-
ing conditions.
As with MEA, carbon dioxide is the principal acid gas remaining in DEA solution. Smith
and Younger found typical C02 loadings of lean DEA solution to be in the range of about
0.01 to 0.03 mole C02/mole DEA with high ratios of HZS:CO2 in the feed gas, and to be
less than about 0.09 with low H2S:C02 ratios. Other investigations report COz loading
much less than 0.09 mole C02/mole DEA, even at very low H2S:C02 ratios. Butwell and
Perry (1975), for example, describe two DEA plants treating natural gas with H2S:C02
ratios of only 0.04 to 0.07 that produced lean DEA loadings in the range of about 0.01-0.02
mole COz/mole DEA.
DGA is a primary amine similar to MEA with regard to basicity. As a result, the correla-
tions developed for MEA stripping can be used to provide a first approximation of lean solu-
tion loadings for DGA solutions. Additional approximate lean solution loading values for
DEA and DGA are provided in Table 2-19. Little data have been published on the stripping
of DIPA, and it is suggested that lean solution loadings estimated for DEA be used in the
absence of more specific data. For material balance purposes, lean MDEA loadings can be
assumed to be zero, as MDEA solutions are very readily stripped. For example, lean MDEA
solution loadings less than 0.004 mole acid gadmole amine have been reported by Dupart et
al. (1993A, B).
Determining the Rich Solution Equilibrium and Design Loadings

After the lean amine temperature and lean solution loading are determined, the designer
should use the following steps to determine the required circulation rate. Refer to Figure
2-93 for information on the amine contactor calculation envelope and to Table 2-20 for a
definition of the variables used in these calculations.

1. Establish the lean amine concentration and loading, LL (moles acid gadmole amine), and
break the loading down into LL.H?S (moles of H2S/mole amine) and LL,co2 (moles of
COz/mole amine).
2. Tabulate the feed gas conditions, including the following:
a. Feed gas flow rate, MF (moleshr) and WF (lbhr).
b. Feed gas inlet temperature, TF (“F).
c. Feed gas inlet pressure, PF @sia).
d. Feed gas composition (mol a).
e. Water content, MF,H~O (moleshr) and WF*,O (lbhr).
f. Latent heat of water in the feed gas, hF (Btunb).
3. Set the lean amine temperature, TL (e.g., TF + 10).