DISTILLATIVE SEPARATION OF ACID GASES FROM LIGHT HYDROCARBONS

Technical Field This invention is in the field of distillation.

Background Art

It is often desirable to separate acid gas compo¬ nents from light hydrocarbons in the processing of gas streams. Certain of these separations are, however, made difficult because of the tendency of mixtures of light hydrocarbons and acid gases to form azeotropes.

One such example can be found in the cryogenic dis- tillative separation of methane from acid gas components described in our copending application. Serial No. 94,226, filed November 14, 1979. This process is particularly effective for separating methane from high Cθ2-content feed in one distillation column without solids formation. The bottoms product of this process contains carbon dioxide, ethane, and higher hydrocarbons, and it is often desirable to separate the carbon dioxide and ethane components in this bottoms product. Such a sep¬ aration would produce a useful carbon dioxide product as well as an enriched ethane product which could be used for its heating value or as raw material in many chemical syntheses.

Although highly desirable., the separation of carbon dioxide from ethane by distillation has proven to be a difficult problem in practice. This difficulty is

σaused by the fact that carbon dioxide and ethane form an azeotrope of approximately two thirds carbon dioxide and one third ethane on a mole basis. For a feed mixture containing ethane and carbon dioxide, this azeotrope tends to form in the upper portion of the column, usually making further separation beyond the

_* azeotrope composition impossible. The common practice of employing two distillation towers operating at dif¬ ferent pressures to work around the azeotrope does not . help with the carbon dioxide/ethane system because pres^ sure has only minimal effect on the composition of the azeotrope. Because of this, attempts to separate carbon dioxide from ethane by distillation have heretofore resulted in an'overhead carbon dioxide stream containing approximately azeotropic amounts of ethane, which are unacceptable in many applicatόns.

Ethylene also forms an azeotrope with carbon • dioxide. Additionally, it is known that the acid gas hydrogen sulfide forms azeotropes with both ethane and propane. These and other possible azeotropes between acid gases and light hydrocarbons present limitations similar to those described for the carbon dioxide/ethane system when efforts are made to perform distillative separations on such systems.

Disclosure of the Invention

This invention relates to the distillative separa¬ tion of acid gases from light hydrocarbons in mixtures wherein such separations are normally limited by the tendency to form an azeotrope. The feed mixtures can, of course, contain additional components such as higher hydrocarbons, nitrogen, hydrogen, etc.

In the method of this invention, a distillation

column can be used to separate an acid gas and light hydrocarbon _r both present in the feed, into an overhead and bottoms product which are not limited by the azeo¬ trope composition of the acid gas or light hydrocarbon. This is achieved by adding to the distillation column an a _. gent which prevents formation of the acid gas/light hydrocarbon azeotrope. For example, an agent which prevents azeotrope formation can be employed in the distillative separation of a binary of carbon dioxide and ethane to produce an overhead stream having signif¬ icantly more than two thirds carbon dioxide, which is the limitation when an azeotrope forms.

In addition to obtaining a purer overhead pro¬ duct, this method has other inherent advantages. In many instances, for example, the energy requirements are lower than in the corresponding separation without agent because the relative volatility is improved or because the separation can be made at higher tempera¬ tures. This, of course, can lower the cost for the separation.

Sometimes materials capable of eliminating azeo¬ tropes are contained in the feed mixtures. For example, natural gas typically contains butane, which can be an effective agent for preventing azeotrope formation between carbon dioxide and ethane. Although the mere presence of such materials in the feed is not sufficient to prevent azeotrope formation, such materials can be separated from bottoms product and added back to the column at an appropriate point to prevent azeotrope formation. This takes advantage of materials already present in the feed as a convenient source of agents.

Brief Description of the Drawings

Fig. 1 is a plot of vapor-liquid equilibria data

for a binary system of carbon dioxide and ethane at two different pressures, as well as data from a computer simulated separation illustrating the effect on such a system of the addition of an agent for preventing azeotrope formation;

Fig. 2 is a plot of the relative volatility of carbon dioxide to ethane at various temperatures in a computer simulated distillation having a multi- component hydrocarbon mixture added to serve as an agent for preventing azeotrope formation?

Fig. 3 is a plot of vapor-liquid equilibria data for a binary system of hydrogen sulfide and propane - at 200 and 400 psia;

Fig. 4 is a plot of vapoir-liquid equilibria data at 200 psia for the hydrogen sulfide/prόpane system containing 0, 20 and 40% butane in the liquid phase,-

Fig. 5 is a plot of the relative volatility of hydrogen sulfide to propane in systems containing 0, 20 and 40% butane in the liquid phase, plotted against liquid phase composition;

Fig. 6 is a plot of vapor-liquid equilibria for a binary system of hydrogen sulfide and ethane as well as data illustrating the effect of the addition of butane to this system to prevent azeotrope formation; Fig. 7 is a schematic flow diagram illustrating apparatus suitable for carrying out the invention described herein.

Best Mode of Carrying Out the Invention

This invention will now be further described in more specific detail with regard to the figures.

US_

- OM

-5-

Much of the data presented in the following descrip¬ tion, as well as that shown in the Figures, was obtained using a plate-to-plate column calculation program to simulate conditions within a distillation column for certain given or desired operating conditions. Unless otherwise stated, the program employed was the Process ^' Simulation Program of Simulation Sciences, Inc., Fullerton, California, Oct.-Nov., 1979. Vapor-liquid equilibria and thermodynamic data were calculated based upon the Soave-Redlich-Kwong equation of state. While the total accuracy of the data obtained cannot be assured, and in fact will change somewhat depending upon the constants chosen, the data is believed to be representative of actual data and is certainly appro- priate for illustrating and substantiating the benefits gained by use of an agent to prevent azeotrope forma¬ tion in distillative separations of an acid gas from a light hydrocarbon according to this invention. For purposes of simplifying the plots, data from systems which were not binary were plotted on a pseudo-binary basis in which mole fractions are calculated as if the components beyond those in the binary were not present.

The practical difficulty of obtaining a complete separation of carbon dioxide from ethane in a gas ix- ture containing both can be seen by referring to Figure 1. Figure 1 contains vapor-liquid equilibria data for a binary mixture of carbon dioxide and ethane at both 578 psia and 335 psia. Data for these binary mixtures were obtained by Arthur D. Little, Inc. using an in- house computer program and employing the Soave-Redlich- Kwong equation of state. As can be seen from the data, the binary mixtures form a minimum boiling point azeo¬ trope at both pressures in the range of about 65-70%

-6-

carbon dioxide. Thus, a mixture of carbon dioxide and ethane tends toward the azeotropic amount as the gas rises up the distillation column. Once the azeo¬ trope composition is reached, further separation of the binary does not occur. ^* Thus, the overhead product typically contains the azeotropic amount of ethane, which in this case would amount to approximately one third of the overhead product, on a mole basis. The data also demonstrate that the effect of pressure on composition of an azeotrope in this system is minimal, meaning that the use of two distillation columns operating at different pressures is not a viable solution to the problem.

The beneficial effect of adding an agent to eliminate the azeotrope is demonstrated in the upper plot of vapor-liquid equilibria data for carbon dioxide and ethane, but additionally including a mixture of C ₃ ~C _fi alkanes, which together function as an additive to prevent azeotrope formation between carbon dioxide and ethane. The exact compositions and operating conditions are discussed in more detail below with reference, to Simulated Run 3."

It can be seen from the data plotted that the azeotrope has been eliminated in the simulated tun with additive. Thus, further separation of carbon dioxide and ethane, beyond the azeotrope amounts of the binary, can be achieved.

The beneficial effect of the additive is also dramatically demonstrated by the data regarding the relative volatility of carbon dioxide to ethane which is plotted in Figure 2. As seen in Figure 2, the

IfUR

relative volatility is about 1.65 at the point in the column where the additive is introduced. It remains at or above this value at all points below the point of addition, which comprise the middle and bottom sections of the column. In the top section of the column, above the additive addition point, the relative volatility slips to values below 1, indicating that ethane is actually more volatile than carbon dioxide in the top section of the column above where the additive was introduced. Thus, efforts to recover the additive actually reverse the natural"relative volatilities of these materials.

Figure 3 is a plot of vapor-liquid equilibria data for a binary system of hydrogen sulfide and propane. In a manner similar to the binary system of carbon dioxide and ethane, the hydrogen sulfide/ propane system exhibits azeotrope formation at both 200 and 400 psia. Although the azeotrope occurs at a somewhat higher mole fraction of hydrogen sulfide, e.g., 84-85 mole percent, nevertheless the azeotrope formation does interfere with a more complete distil¬ lative separation.

Figure 4 presents vapor-liquid equilibria data for a binary hydrogen sulfide/propane system as well as this system with n-butane added to prevent azeo¬ trope formation between hydrogen sulfide and propane. It can be seen that the azeotrope illustrated in Figures 3 and 4, which occurs at 200 psia for the pure binary, is eliminated by the addition of n-butane to a level of 20%. The pinched portion of the vapor-

liquid equilibrium plot is opened even more when n- butane is added to a level of 40%. Thus, in addition to eliminating the azeotrope, the n-butane additive also increases the relative volatility of hydrogen sulfide to propane near the azeotrope composition. Further, the additon of n-butane also increases the column temperatures, which lowers the energy require¬ ments for refrigeration. Because of these factors, it might be desirable to add n-butane to hydrogen sulfide/ propane systems even though the separation will not be extended beyond the azeotrope composition.

, ^; Figure 5 is a plot of the relative volatility of hydrogen sulfide to propane, with and without n-butane as an additive. As illustrated, with no additive, the relative volatility falls below 1 for mole percentages of hydrogen sulfide above about 84%. Addition of n- butane raises the relative volatility well above 1, even for the higher percentages of hydrogen sulfide. Thus, as noted above, it can be desirable to add agent to enhance relative volatility even below the azeotrope composition, as illustrated.

Figure 6 is a plot of vapor-liquid equilibria data for the binary hydrogen sulfide/ethane system and for this system with n-butane at a level of 20% added to the liquid phase. In this system, the light hydrocarbon ethane is more volatile than the acid gas hydrogen sulfide below the azeotrope. As can be seen, the azeotrope present in the binary is eliminated by the addition of n-butane to a level of 20%.

In order to further describe this invention, the results of three distillations simulated on a computer will now be presented.

''

SIMULATED RUN 1

Summary

Feed, Mols^/Hr. 1 646 Feed CO2, Hoi % 65 Feed Ethane, Hoi t 9.9

Stages^, Total 20 Additive Rate, Hols/Hr 986.3

Ethane Recovery in Bottoms Product, % 64.7 Ethane Composition in Overhead, Hoi % 5.3 CO? Recovei in Overhead, % 96.8 Additive Loss Into Overhead, Hols/Hr 2.5 „ _< - . (CO2 Ohd)(Ethane Btm) 52.9 - ^{*p ~} * (CO2 Btm)(Ethane Ohd)

Col umn Operation - 500 psia

Heating

(Cool ing) Location

Duty HMBTU/Hr Temperature Stage

(4.8) (2.9) 30°F 1 Overhead Partial Condenser 2 thru 5 Top Section of Column

98°F 6 Additive Fed In 7 thru 11 Hiddle. "Section of Column

68°F 12 Feed 13 thru 19 Bottom Section of Column

18.5 11 .2 ^* 251°F 20 Reboiler

Material Balance

' Feed Additive Overhead Bottoms

Kols/Hr CO2 1072.3 1 .0 - ^• 1038.0 35.3 96.85 Recovery in Overhead

Kols/Hr Ethane 163. S 1 .0 58.7 105.8 64.75 Recovery in Bottoms

Kols/Hr Propane 142.2 .1 .0 .2 143.0

Kols/Hr ^• n-Butane 268.5 983.3 2.3 1249.5

TOTAL 1646.5 986.3 1099.2 1533.6

Hoi X CO2 65.1 0.1 94.4 2.3

Kol 5 Ethane 9.93 0.1 5.3 6.9

Hoi 5 Propane-Plus ; 25.0 99.8 0.23 90.8

"F Temperature 62.9 88.0 30.3 251 .4

Ho 5 Liquid 100 00 0 100

Hoi s CO, (Pseudo- bin (d). 86.8 94.6 25.0

(a) Stages are theoretical stages (trays , partial condensers, reboil ers) , 1005 efficient.

(b) For full amount of feed, as shown.

(c) Per 1000 ols/hr. of feed CO- ^•

(d) Pseudo-binary C02 composition = {-^ — + ethane^ ^{1 0} °

(e) Hols cited are b- ols -

Conditions at Sel ected Stages

Stage Number 1 6 12 20

Temperature, °F 30.3 98.0 67.6 251

Liquid, C0 ₂ , Hoi ϊ 94.7 35.6 46.7 2.30

Liquid, Ethane, Hoi t 4.45 4.23 10.1 6.89

Liquid, Propane-Plus, Hoi 1 0.85 60.2 43.2 90.8 ^'

Liquid, Reduced Temp. 0.892 0.822 0.828 0.966

Liquid Flow, Hol s/Hr 1100 2093 3634 1534

Vapor, C0 ₂ , Ho 1 94.4 82.1 81 .5 5.86

Vapor, Ethane, Kol 5 5.34 5.34 10.4 12.8

Vapor, Propane-Plus, Hoi 2 0.23 12.6 8.1 81 .3

Vapor Flow, Hol s/Hr 1099 1803 2355 3510

Relative Volatil ity, 0 ₂ : Ethane 0.831 1 .83 1 .69 1 .37

_. Q

SIMULATED RUN 2

Summary

Column Operation - 400 psia

Heating (Cooling) Location Duty HH3TU/Hr Temperature Stage b c

(10.9) (1.1) 15°F 1 Overhead Partial Condenser

2 thru 5 Top Section of Column (13.7) (1.4) 62°F 6 Intermediate.Partial Condenser

Additive Added to Stage 6

7 thru 13 Kiddle Section of Column

80°F •14 Feed Vapor Addition 80 ^β F 15 Feed Liquid Addition

16 thru 24 Bottom Section of Column

94.0 9.4 210°F 25 Reboiler

-12-

Haterial Balance

Ethane-Plus O2 Product Product

Feed Additive Overhead Bottoms _β

Kols/Hr Hethane 32.1 0.0 0.0 32.1

Hol s/Hr C02 3189.8 0.0 - ^' 3058.1 131 .7 95. ,95 Recovery in Overheβ'

Hols/Hr Ethane 1851 .7 0.0 167.0 1684.7 - 91 . ,05 Recovery in Bottoms

Hols/Hr H ₂ S .235 0.0 .0040 .231 9£ 1.35 ; El imination from Overhead

Hols/Hr Propane 1026.0 60.9 2.8 1084.1

Hols/Hr Butane-Plus 3844.8 6025.7 5.0 9865.5

TOTAL 9944.6 6086.6 3232.9 12798.3

Hoi X C02 32.1 0.0 93.7 1 .03

Hoi X C2 18.6 0.0 5.11 13.2

Hoi X Propane-Plus 49.0 100.0 .24 85.8 ppra H S 23.6 0.0 1 .22 18.1

•F Temp 75.8 57.0 14.8 209.5

Hoi X Liquid 83.6 100.0 0.0 100.0

Hoi X CO2 (pseydp- 63.3 - 94.8 7.3 bin J)

(a) Stages are theoretical stages (trays, partial condensers, reboilers), 1005 efficient.

For full amount of feed, as shown.

13 Per 1000 mols/hr. of feed C0 ₂

. ,

(d) Pseudo-binary CO. composition x 100

CO _j + ethane

(e) Kols cited are l i -εsols

Conditions at Selected Stages

Stage Number 1 6 15 25 Temperature, °F 14.8 61 .9 79.5 210 Liquid, C0 ₂ , Hoi X 95.1 39.3 25.1 1. Liquid, Ethane, Hoi f 4.01 4.46 18.3 13. Liquid, Propane-Plus, Hoi X 0.76 56.1 56.5 85. Liquid, Reduced Temperature 0.865 0.780 0.811 0. Liquid Flow, Hols/Hr 2250 12173 21238 1276 Vapor, C0 ₂ , Hol/Hr 93.7 86.4 63.4 3. Vapor, Ethane, Hoi X δ.n 5.10 24.2 30. Vapor, Propane-Plus, Hoi X 0.24 7.8 12.1 66. Vapor Flow, Hol s/Hr 3265 4930 8395 1323 Relative Volatil ity C0 ₂ : Ethane 0.773 1 .92 1 .91 1 .

^C? 1£

SIMULATED RUN 3

Column Operating Conditions - 400 psia

Heating (Cooling) Location

Duty MMBTU/Hr Temperature Stage b c (3^5) (73) 16 ^β F 1 Overhead Partical Condenser

Top Section of Column

(5.44) (3.1) 30 ^β F 2 thru 5 Intermediate Partial Condenser

Additive Added to 6th Stage Hiddle Section of Column

34 ^β F 7 thru 16.. Feed Addition

Bottom Section of Column

23.0 13.0 205 ^β F IB thru 25 Reboiler

( __0?PI

Haterial Balance

Ethane-Plus

COg Product Product

Feed Additive Overhead Bottoms

Kols/Hr C0 ₂ 1069.0 0.0 1058.1 10.9 99.05 Recovery in Ohd

Hol s/Hr Ethane 310.0 0.0 - 20.0 290.0 93.55 Recovery in Btm

Hols/Hr H ₂ S 5.0 .0.0 0.0083 4.992 99.85 EHmin.from Ohd

Hols/Hr Propane 253.0 59.7 3.2 309.5

Hols/Hr n-Butane 92.0 463.6 . 0.7 554.9

Hol s/Hr n=Pentane 36.0 362.T 0.0 398.1

Hols/Hr n-hexane 11 .0 109.4 0.0 120.4

TOTAL 1776.0 994.8 1082.0 1688.8 *

Kol 5 C0 _? 60.2 0.0 97.8 0.65 Hoi X C ₂ 17.5 0.0 1 .8 17.2

Hoi X Propane-Pl us 22.1 100 0.36 81 .9 ppra H2S 2800 0 7.7 3000 - ^β F Temperature 30.7 31 .0 16.3 204.6 .

Hoi 5 Liquid 79.1 100.0 » 0.0 100.0

Hoi 5 CO? (pseudo- .« - bin ^(d} ) 77.5 98.1 3.6

(a) Stages are theoretical stages (trays, partial condensers , reboil ers) , 1005 efficient

(b) For full amount of feed , as shown, (cj of feed Q. i d) r cnommpπnocsfiftlifloni __: f & ) x 100

COg + ethane. (e) Hols cited are I b-mols

Conditions at Sel ected Stages

Stage Number 1 6 17 30

Temperature, ^β F 16.3 30.2 34.4 205.0

Liquid, C0 ₂ , Hoi X 97.9 65.5 51 .9 0.64

Liquid, Ethane, Hoi X 1.37 1 .89 16.2 17.2

Liquid, Propane-Plus, Hoi X 0.76 32.6 31 .7 81 .9

Liquid, Reduced Temperature 0.867 0.777 0.797 0.90

Liquid Flow, Hols/Hr -790 3313 4758 1689

Vapor, C0 ₂ , Hoi X 97.8 95.7 80.1 2.60

Vapor, Ethane, Hoi X 1 .85 1 .66 14.9 41 .2

Vapor, Propane-Plus, Hoi X 0.37 2.67 - 4.15 ^• 55.6

Vapor Flow, Kols/Hr 1082 1829 3429 2874

Relative Volatil ity C0 ₂ : Ethane 0.740 1 .66 1 .68 1 .68

-15-

Since the azeotrope composition is between 65-70% carbon dioxide, the feeds presented in the three simula¬ tions of 87%, 63% and 78% (pseudo-binary basis) are on both the high and low sides of the azeotrope co - 5 position. The respective bottom products lean in carbon dioxide of 25%, 7% and 4% are significantly- lower than the feed composition. The overhead products, rich in carbon dioxide at 95%, 95% and 98%, are sig¬ nificantly richer in carbon dioxide content than the 0 azeotrope composition. Thus, the additive has made possible a more complete separation of carbon dioxide from ethane in all three simulations. A distilla¬ tion without additive could not have produced such complete separations. Simulation 3 is particularly 5 noteworthy in that the carbon dioxide overhead stream was substantially free of ethane and the ethane bottoms product was substantially free of carbon dioxide. Even more complete separations could be achieved.

The heating and cooling duties are tabulated per 0 1,000 moles/hour of feed. In this regard. Simulated Run ^* 2 has the least consumption of heating and cooling whereas Simulated Run 3 has the greatest consumption of each.

In regard to the split function. Simulation Run 1 has the least thorough split whereas Simulated Run 3 has 5 the most thorough split.

In general, any material or mixture of materials which causes the relative volatility of an acid gas to a light hydrocarbon from which it is to be separated by distillation to be significantly different than 1 0 over the range of interest is satisfactory as an agent for this invention. Liquids which are miscible with the • acid gas and light hydrocarbon, such as C ₃ -C ₈ alkanes, are

^' BU EΛTT o pi

preferred agents because they are typically present in feed mixtures, are easy to separate and recycle, and often have a very beneficial effect in causing the acid gas to be more volatile relative to the light hydrocarbon. Natural gas liquids (NGL) contain such alkanes and can . often be separated from bottoms product in conventional separation equipment. Thus, NGL or components thereof can be conveniently recycled to provide agent. It is also clear that materials satisfactory for the agent need not be pure materials. In general, the agent should be liquid at the overhead conditions in the distillation column. It is desirable, of course, to have agents which have volatilities lower than the components to be separated. The agent should- also have a .freezing point sufficiently ^* low to avoid solids formation in the column.

In addition to the preferred.materials mentioned above, there are other classes of materials which meet these requirements. For example, other hydrocarbons such as higher alkanes and naphtheήes, halogenated hydrocarbons such as fluoro-chloromethane and fluoro- chloroethane compounds; sulphur dioxide, etc.. are believed to be suitable. Those skilled in the art will know, or be able to ascertain using no more than routine experimentation, other suitable agents for use with the invention described herein.

The amount of agent added will be dependent upon factors such as the composition of the feed, operating pressure, throughput of the column, recovery of over- head and bottoms product desired, etc. Such factors can be taken into account by those skilled in the art by determining the operative amounts for. any given sepa-

BϋR

O ^'

-17-

ration using no more than routine experimentation.

Agent is added to the tower at a point above where feed is introduced since this is where azeotrope forma¬ tion normally occurs. Although, some materials which, are suitable agents are contained in the feed in some cases, this alone is not sufficient to prevent azeotrope format tion. This is because the agent is usually not suf¬ ficiently volatile to rise up the column to the problem area.. Thus, even if present in the feed, the agent should be separated and added to a point above the feed _* Although it is possible to add feed at the top of the column, including into the condenser, this is usually not desirable because agent cannot then be separated efficiently from overhead product. Thus, it is preferable to add agent in most cases at a point below the column top to thereby allow separation of additive from the desired overhead product.

In some cases, it is desirable to add agent at more than one column location on a simultaneous basis. An apparatus for carrying out a separation of carbon dioxide from ethane according to this invention is schematically illustrated in Figure 7. Therein, feed mixture 10, containing a mixture of carbon dioxide and ethane, and usually other components such as heavier hydrocarbons, nitrogen, etc., enters through feed line 12 into distillation column 14. Column 14 contains a number of vapor-liquid contact devices such as trays or packing, with the exact number of contact stages depend¬ ing upon the required operating conditions. Overhead stream 20 is rich in carbon dioxide and . passes to partial condenser 22 from which the remaining vapor stream 24 exits as carbon dioxide product. This

product stream also contains, of course, components present in the feed which are more volatile than carbon dioxide, such as any nitrogen present in the feed. Liquid from the partial condenser returns to column 5 14 in line 26 where it serves as reflux for tower 14. - Condenser 22 is cooled by external cooling source 28.

The bottoms stream exits from the lower portion of column 14 in bottoms line 30 and contains ethane and other less volatile hydrocarbons or other components, 1.0 and any agent added to prevent azeotrope formation. A portion of the bottoms product is passed through reboile 32 and back to column 14 in line 34. Reboiler 32 is ' heated by an external heat source 36.

The bottoms product passes in line 38 to further 5 separation equipment 40, such as another distillation column. Separation equipment 40 is _. employed to separate out the agent which is recycled in line 42 back to the : column. The amount of recycled agent can be controlled by flow control valve 46. An ethane fraction is also 0 ^" separated in equipment 40 and is directed in line 44 to suitable ethane product facilities ^' .

Agent for preventing azeotrope formation may also be added to the system through line 50 and flow control valve 52. Such externally added agent may be used in 5 lieu of recycled agent or in conjunction with recycled agent. In either case, the agent is cooled in heat exchanger 54, cooled by cooling source .56, and directed through flow lines 58 back towards the column 14.

Agent can be added at a number of different loca- 0 tions, either individually or at ^" several locations

simultaneously. As illustrated, agent can be directed in line 64 _. to flow control valve 66 and flow line 68 and introduced directly onto a tray in the upper section of column 14. Similarly, agent can be added to a " higher column tray, such as by passing it in line 70 through control valve 72 and line 74. Agent can also be introduced into condenser 22 by directing agent through line 60, flow control valve 62 and line 63. Other suit¬ able points of addition can be determined, of course, for each particular separation being performed.

Industrial Applicability

This invention is useful in the distillative separation of acid gas components from light hydrocarbon components. .

Equivalents

Those skilled in the art will recognize, or be able to determine using no more than routine experimentation, other equivalents to the specific embodiments described herein.