Process for removal of acid gases from a fluid stream

Abstract

In a process for removal of acid gases from a fluid stream the fluid stream is contacted with an absorbent to obtain a treated fluid stream and a laden absorbent. The absorbent comprises a diluent and a compound of the general formula (I) wherein R.sup.1 is C.sub.1-C.sub.3-alkyl; R.sup.2 is C.sub.1-C.sub.3-alkyl; R.sup.3 is selected from hydrogen and C.sub.1-C.sub.3-alkyl; and R.sup.4 is selected from hydrogen and C.sub.1-C.sub.3-alkyl. ##STR00001##

Claims

1. A process for removing acid gases from a fluid stream, wherein the fluid stream is contacted with an absorbent obtain a treated fluid stream and a laden absorbent, the absorbent comprising at least one diluent and a compound of the general formula (I) ##STR00005## wherein R.sup.1 is C.sub.1-C.sub.3-alkyl; R.sup.2 is C.sub.1-C.sub.3-alkyl; R.sup.3 is selected from hydrogen and C.sub.1-C.sub.3-alkyl; and R.sup.4 is selected from hydrogen and C.sub.1-C.sub.3-alkyl.

2. The process according to claim 1, wherein R.sup.3 is C.sub.1-C.sub.3-alkyl.

3. The process according to claim 1, wherein the compound of the general formula (I) is selected from 3-(tert-butylamino)propane-1,2-diol, 1-(tert-butylamino)-3-methoxy-propane-2-ol, 3-(iso-propylamino)propane-1,2-diol, and 3-[(2-methylbutan-2-yl)amino]propane-1,2-diol.

4. The process according to claim 1, wherein the diluent comprises water.

5. The process according to claim 4, wherein the absorbent additionally comprises an acid.

6. The process according to claim 1, wherein the diluent comprises a non-aqueous organic solvent.

7. The process according to claim 6, wherein the organic solvent is selected from C.sub.4-10 alcohols, ketones, esters, lactones, amides, lactams, sulfones, sulfoxides, glycols, polyalkylene glycols, di- or mono(C.sub.1-4-alkyl ether)glycols, di- or mono(C.sub.1-4-alkyl ether) polyalkylene glycols, cyclic ureas, thioalkanols and mixtures thereof.

8. The process according to claim 1, wherein the absorbent comprises at least one activator selected from a sterically unhindered primary amine and/or a sterically unhindered secondary amine.

9. The process according to claim 8, wherein the activator is piperazine.

10. The process according to claim 1, for selective removal of hydrogen sulfide from a fluid stream comprising carbon dioxide and hydrogen sulfide.

11. The process according to claim 1, wherein the laden absorbent is regenerated by means of at least one of the measures of heating, decompressing and stripping with an inert fluid.

Description

(1) The invention is illustrated in detail by the appended drawings and the examples which follow.

(2) FIG. 1 is a schematic diagram of a plant suitable for performing the process of the invention.

(3) FIG. 2 is a plot of the selectivity of H.sub.2S over CO.sub.2 as a function of acid gas loading of an aqueous solution of methyldiethanolamine (MDEA, 26 weight-%) and an aqueous solution of 3-(tert-butylamino)propane-1,2-diol (TBAPD, 32 weight-%).

(4) FIG. 3 is a plot of the acid gas loading over time of an aqueous solution of methyldiethanolamine (MDEA, 26 weight-%) and an aqueous solution of 3-(tert-butylamino)propane-1,2-diol (TBAPD, 32 weight-%).

(5) FIG. 4 is a schematic diagram of a twin stirred cell arrangement used to determine the relative CO.sub.2 absorption rates of absorbents.

(6) According to FIG. 1, via the inlet Z, a suitably pre-treated gas comprising hydrogen sulfide and carbon dioxide is contacted in countercurrent, in an absorber A1, with regenerated absorbent which is fed in via the absorbent line 1.01. The absorbent removes hydrogen sulfide and carbon dioxide from the gas by absorption; this affords a hydrogen sulfide- and carbon dioxide-depleted clean gas via the offgas line 1.02.

(7) Via the absorbent line 1.03, the heat exchanger 1.04 in which the CO.sub.2- and H.sub.2S-laden absorbent is heated up with the heat from the regenerated absorbent conducted through the absorbent line 1.05, and the absorbent line 1.06, the CO.sub.2- and H.sub.2S-laden absorbent is fed to the desorption column D and regenerated.

(8) Between the absorber A1 and heat exchanger 1.04, one or more flash vessels may be provided (not shown in FIG. 1), in which the CO.sub.2- and H.sub.2S-laden absorbent is decompressed to, for example, 3 to 15 bar.

(9) From the lower part of the desorption column D, the absorbent is conducted into the boiler 1.07, where it is heated. The steam that arises is recycled into the desorption column D, while the regenerated absorbent is fed back to the absorber A1 via the absorbent line 1.05, the heat exchanger 1.04 in which the regenerated absorbent heats up the CO.sub.2- and H.sub.2S-laden absorbent and at the same time cools down itself, the absorbent line 1.08, the cooler 1.09 and the absorbent line 1.01. Instead of the boiler shown, it is also possible to use other heat exchanger types for energy introduction, such as a natural circulation evaporator, forced circulation evaporator or forced circulation flash evaporator. In the case of these evaporator types, a mixed-phase stream of regenerated absorbent and steam is returned to the bottom of desorption column D, where the phase separation between the vapour and the absorbent takes place. The regenerated absorbent to the heat exchanger 1.04 is either drawn off from the circulation stream from the bottom of the desorption column D to the evaporator or conducted via a separate line directly from the bottom of the desorption column D to the heat exchanger 1.04.

(10) The CO.sub.2- and H.sub.2S-containing gas released in desorption column D leaves the desorption column D via the offgas line 1.10. It is conducted into a condenser with integrated phase separation 1.11, where it is separated from entrained absorbent vapour. In this and all the other plants suitable for performance of the process of the invention, condensation and phase separation may also be present separately from one another. Subsequently, the condensate is conducted through the absorbent line 1.12 into the upper region of desorption column D, and a CO.sub.2- and H.sub.2S-containing gas is discharged via the gas line 1.13.

(11) In FIG. 4, the following reference symbols are used: A=CO.sub.2 storage vessel, B=twin stirred cell, C=temperature regulator, D=metering valve, E=manometer. According to FIG. 4, a liquid phase of the absorbent to be tested is present in the lower part of the twin stirred cell B, and is in contact with the gas phase above it via a phase boundary. The liquid and gas phase can each be mixed with a stirrer. The twin stirred cell B is connected to the CO.sub.2 storage vessel A via a metering valve D. The pressure that exists in the twin stirred cell B can be determined by means of the manometer E. In the measurement, the volume flow rate of carbon dioxide is recorded, the volume flow rate being adjusted such that a constant pressure exists in twin stirred cell B.

(12) In the description of the examples, the following abbreviations were used:

(13) DMAPD: 3-(dimethylamino)-1,2-propanediol

(14) M3ETB: (2-(2-(2-tert-butylaminoethoxy)ethoxy)ethyl)methyl ether

(15) MDEA: methyldiethanolamine

(16) TBA: tert-butylamine

(17) TBAEE: 2-(2-tert-butylaminoethoxy)ethanol

(18) TBAPD: 3-(tert-butylamino)propane-1,2-diol

EXAMPLE 1

Preparation of 3-(tert-butylamino)propane-1,2-diol (TBAPD)

(19) 1,000 mL of absolute ethanol, denatured with toluene, were filled into a 4 L round bottom flask, equipped with an overhead stirrer. The flask was placed in an ice bath. 438 g (6.0 mol) of tert-butylamine (TBA) was added and thoroughly mixed with the ethanol. After the solution had cooled to 1° C., 222 g of glycidol (3.0 mol) were added continuously over 60 min using a dropping funnel. The temperature inside the flask was kept at 1 to 3° C. After the addition was complete, the ice bath was removed, and the content was allowed to warm to room temperature under intense stirring. After three hours reaction time, a sample was taken and analyzed by gas chromatography. The reaction mixture was stirred for a further 20 h at room temperature and then for 6 h at 47° C. After this period, conversion was complete. Excess TBA was removed in vacuo and the crude product was purified by distillation over a 10 cm distillation column and Claisen bridge (0.7 mbar, 93 to 95° C. head temperature). The yield of TBAPD was 302 g (68.4%), and the purity by GC was 99.8%.

(20) Gas chromatography was performed with a column type DB1 by Agilent, length 30 m, diameter 0.25 mm, layer thickness 1 μm. The temperature program was: 8 min at 60° C., 10° C./min to 280° C., and 35 min at 280° C.

EXAMPLE 2

Comparison of Absorption Properties of MDEA and TBAPD

(21) The experiments were carried out in an absorption unit (semi-batch system), comprising a stirred autoclave to which gas could be fed in up-flow mode, and a condenser. The autoclave was equipped with a pressure gage and a type J thermocouple. A safety rupture disc was attached to the autoclave head. A high wattage ceramic fiber heater was used to supply heat to the autoclave. The gas flows were regulated by mass flow controllers (from Brooks Instrument) and the temperature of the condenser was maintained by a chiller. The maximum working pressure and temperature were 1000 psi (69 bar) and 350° C., respectively.

(22) During runs at atmospheric pressure, the pH of the solution was monitored in situ by using a pH probe (from Cole-Parmer), which was installed in the bottom of the autoclave. This pH probe was limited by a maximum temperature and pressure of 135° C. and 100 psi, respectively. Therefore, before carrying out experiments at a pressure above atmospheric pressure (“higher pressure”), the pH probe was removed and the autoclave was capped. In both cases (atmospheric pressure and higher pressure), liquid samples were collected by directly attaching a vial (atmospheric pressure) or a stainless steel cylinder filled with caustic (higher pressure) to the sampling system. A specifically designed LabVIEW program was used to control the absorption unit operation and to acquire experimental data like temperature, pressure, stirrer speed, pH (at atmospheric pressure), gas flow rate and off-gas concentration.

(23) The gas mixture used in the examples had the following properties:

(24) Gas feed composition: 10 mol-% CO.sub.2, 1 mol-% H.sub.2S, 89 mol-% N.sub.2

(25) Gas flow rate: 154 SCCM

(26) Temperature: 40.8° C.

(27) Pressure: 1 bar

(28) Volume: 15 mL (T=0.1 min)

(29) Stirring rate: 200 rpm

(30) The experiments of example 2 were performed by flowing gas mixtures as specified above through the autoclave.

(31) The autoclave was previously loaded with the respective aqueous amine solution, as specified below:

(32) TABLE-US-00001 molar weight molarity amount of amine molar ratio amine [g/mol] [mol.sub.amine/L] [weight-%] H.sub.2O:amine MDEA 119.2 2.17 26 18.8 TBAPD 147.2 2.17 32 17.4

(33) The acid gas mixture was fed to the bottom of the reactor. The gases leaving the autoclave were passed through the condenser, which was kept at 10° C., in order to remove any entrained liquids. A slip-stream of the off-gas leaving the condenser was fed to a micro-gas-phase chromatograph (micro-GC, from Inficon) for analysis while the main gas flow passed through a scrubber. After reaching breakthrough, nitrogen was used to purge the system.

(34) The slip-stream of the off-gas was analyzed using a custom-built micro-GC. The micro-GC was configured as a refinery gas analyzer and included 4 columns (Mole Sieve, PLOT U, OV-1, PLOT Q) by Aglient and 4 thermal conductivity detectors. A portion of the off-gas was injected into the micro-GC approximately every 2 min. A small internal vacuum pump was used to transfer the sample into the micro-GC. The nominal pump rate was approximately 20 mL/min in order to achieve 10× the volume of line flushes between the sample tee and the micro-GC. The actual amount of gas injected into the GC was approximately 1 μL. The PLOT U column was used to separate and identify H.sub.2S and CO.sub.2, and the micro-TCD was used to quantify them.

(35) The results are shown in FIGS. 2 and 3. The term “acid gas loading” as used herein stands for the concentration of the H.sub.2S and CO.sub.2 gases physically dissolved and chemically combined in the absorbent solution as expressed in moles of gas per moles of the amine.

(36) Aqueous MDEA has a maximum selectivity of about 5.8 at a loading of about 0.15 moles. The selectivity declines at higher H.sub.2S and CO.sub.2 loadings. In comparison, aqueous TBAPD has a maximum selectivity of about 6.5, with the maximum shifted to higher loadings of about 0.30 moles.

(37) The acid gas loading of aqueous MDEA over time shows a maximum CO.sub.2 loading of about 0.38 moles of CO.sub.2 per mole of amine after about 800 minutes, while the H.sub.2S loading rises to a maximum of about 0.05 moles of H.sub.2S per mole of amine after about 50 minutes, afterwards falling slightly to about 0.04 moles of H.sub.2S per mole of amine at about 200 minutes and remaining essentially steady afterwards. Probably, bound H.sub.2S is displaced by CO.sub.2 at higher loadings/longer residence times. Aqueous TBAPD shows a maximum CO.sub.2 loading of about 0.55 moles of CO.sub.2 per mole of amine after about 800 minutes, while the final H.sub.2S loading is about 0.08 moles of H.sub.2S per mole of amine at about 400 minutes.

EXAMPLE 3

Cyclic Capacity of MDEA, TBAEE and TBAPD

(38) A loading experiment and then a stripping experiment were conducted.

(39) A glass condenser, which was operated at 5° C., was attached to a glass cylinder with a temperature-regulated jacket. This prevented distortion of the test results by partial evaporation of the absorbent. The glass cylinder was initially charged with about 100 mL of unladen absorbent (30% by weight of amine in water). To determine the absorption capacity, at ambient pressure and 40° C., 8 L (standard temperature and pressure, STP)/h of CO.sub.2 or H.sub.2S were passed through the absorption liquid via a frit over a period of about 4 h. Subsequently, the loading of CO.sub.2 or H.sub.2S was determined as follows.

(40) The determination of H.sub.2S was effected by titration with silver nitrate solution. For this purpose, the sample to be analysed was weighed into an aqueous solution together with about 2% by weight of sodium acetate and about 3% by weight of ammonia. Subsequently, the H.sub.2S content was determined by a potentiometric turning point titration by means of silver nitrate. At the turning point, the H.sub.2S is fully bound as Ag.sub.2S. The CO.sub.2 content was determined as total inorganic carbon (TOC-V Series Shimadzu).

(41) The laden solution was stripped by heating an identical apparatus setup to 80° C., introducing the laden absorbent and stripping it by means of an N.sub.2 stream (8 L (STP)/h). After 60 min, a sample was taken and the CO.sub.2 or H.sub.2S loading of the absorbent was determined as described above.

(42) The difference in the loading at the end of the loading experiment and the loading at the end of the stripping experiment gives the respective cyclic capacity.

(43) The results are shown in the following table:

(44) TABLE-US-00002 Absorbent CO.sub.2 loading H.sub.2S loading [mol.sub.CO2/mol.sub.amine] Cyclic [mol.sub.CO2/mol.sub.amine] Cyclic after after CO.sub.2 capacity after after H.sub.2S capacity Composition loading stripping [mol.sub.CO2/mol.sub.amine) loading stripping [mol.sub.H2S/mol.sub.amine] 30 wt.-% MDEA + 70% wt.-% H.sub.2O* 0.77 0.05 0.72 0.68 0.11 0.57 30 wt.-% TBAEE + 70% wt.-% H.sub.2O* 0.97 0.24 0.73 n.d.** n.d.** n.d.** 30% wt.-% TBAPD + 70% wt.-% H.sub.2O 0.94 0.11 0.83 0.91 0.12 0.79 *comparative example **n.d. = not determined

(45) It is clear that TBAPD shows a higher cyclic CO.sub.2 capacity than MDEA and TBAEE and a higher H.sub.2S capacity than MDEA.

EXAMPLE 4

Volatility

(46) The volatility of M3ETB, TBAPD, MDEA and TBAEE in 30% by weight aqueous solutions was examined.

(47) The same apparatus as in example 3 was used, except that the condensate obtained in the glass condenser was not returned to the glass condenser but was separated and analyzed for its composition after the experiment had ended via gas chromatography and Karl Fischer titration. The glass cylinder was regulated to 50° C., and 200 mL of the absorbent were introduced in each case. Over an experimental duration of 8 h, 30 L (STP)/h of N.sub.2 were passed through the absorbent at ambient pressure.

(48) The experiments were repeated three times. The obtained average values are shown in the following table:

(49) TABLE-US-00003 Conden- sate Water Amine Solution [mL] [wt.-%] [wt.-%] 30 wt.-% M3ETB + 70% wt.-% H.sub.2O* 30.1 99.2 0.7 30% wt.-% TBAEE + 70% wt.-% H.sub.2O* 30.0 99.3 0.7 30 wt.-% MDEA + 70% wt.-% H.sub.2O* 27.1 99.4 0.7 30% wt.-% TBAPD + 70% wt.-% H.sub.2O 34.1 99.4 0.4 *comparative example

(50) It is clear that TBAPD has a lower volatility compared to the M3ETB, TBAEE and MDEA.

EXAMPLE 5

pK.SUB.A .Values

(51) The pKa value of the amino groups of TBAPD and MDEA were determined by titration with at 20° C. An aqueous amine solution (0.005 mol/L) was titrated with hydrochloric acid (0.1 mol/L). The results are shown in the following table:

(52) TABLE-US-00004 TBAPD MDEA* pK.sub.A 10.0 8.7 *comparative example

(53) It is clear that the pK.sub.A value of TBAPD is greater than that of MDEA. It is believed that a high pK.sub.A value at relatively low temperatures (such as 20° C.) as exist in the absorption step promotes efficient acid gas absorption.

EXAMPLE 6

Comparison of Absorption Properties of Aqueous Solutions of MDEA, TBAPD and DMAPD Comprising Piperazine

(54) In a twin stirred cell (TSC) according to FIG. 4, the relative CO.sub.2 absorption rates of aqueous absorbents were measured.

(55) The twin stirred cell had an internal diameter of 85 mm and a volume of 509 mL. The temperature of the cell was kept at 50° C. during the measurements. In order to mix the gas and liquid phases, the cell according to FIG. 4 comprised two stirrers. Before commencement of the measurement, the twin stirred cell was evacuated. A defined volume of degassed absorbent was added to the twin stirred cell and the temperature was regulated at 50° C. The stirrers were already switched on during the heating of the unladen absorbent. The stirrer speed was selected such that a planar phase boundary formed between the liquid phase and gas phase. Development of waves at the phase interface has to be avoided since there would otherwise be no defined phase interface. After the desired experimental temperature had been attained, carbon dioxide was introduced into the reactor by means of a metering valve. The volume flow rate was controlled such that the CO.sub.2 partial pressure was constant at 50 mbar over the entire experiment. With increasing experimental duration, the volume flow rate decreased since the absorbent became saturated over time and the absorption rate decreased. The volume flow rate was recorded over the entire period. The experiment was ended as soon as no further carbon dioxide flowed into the twin stirred cell. The absorbent was in an equilibrium state at the end of the experiment.

(56) For the evaluation of the experiments, the absorption rate in (mol of CO.sub.2)/[(m.sup.3 of absorption medium)×min] was determined as a function of the loading of the absorption medium. The absorption rate was calculated from the recorded volumetric flow rate of carbon dioxide and the volume of absorption medium in the twin stirred cell. The loading was determined from the cumulative amount of carbon dioxide which had been fed into the twin stirred cell and the mass of absorption medium in the twin stirred cell. Further, the maximum loading at the end of the experiment was determined.

(57) The median absorption rates were determined as follows: starting from the maximum loading of the absorption medium (virtually equilibrium state at a CO.sub.2 partial pressure of 50 mbar and a temperature of 50° C.) the absorption rates were determined at 75%, 50% and 20% loading of the maximum loading and the mean was determined. Absorption rates at less than 20% loading were not taken into account in the taking of the mean, since the absorption medium in the industrial process passes into the absorption apparatus with a residual loading of CO.sub.2.

(58) The following aqueous solutions were examined: 5 wt.-% piperazine and 30 wt.-% MDEA, 5 wt.-% piperazine and 30 wt.-% DMAPD, and 5 wt.-% piperazine and 30 wt.-% TBAPD. The results are shown in the following table:

(59) TABLE-US-00005 Relative median Maximum absorption loading Aqueous solution rate** [Nm.sup.3/t solvent] 5 wt.-% piperazine + 30 wt.-% MDEA* 100% 17 5 wt.-% piperazine + 30 wt.-% DMAPD* 119% 22 5 wt.-% piperazine + 30 wt.-% TBAPD 142% 31 *comparative example **based on a solution of 5 wt.-% piperazine and 30 wt.-% MDEA

(60) It is clear that the absorption rate of the aqueous solution comprising TBAPD is significantly higher than that of the solutions comprising MDEA and DMAPD, respectively. Further, the maximum loading of the aqueous solution comprising TBAPD is higher than that of the solutions comprising MDEA and DMAPD, respectively.

EXAMPLE 7

Synthesis of 1-tert-butylamino-3-methoxy-propane-2-ol

(61) A 250 mL 4-necked flask equipped with a cooler, magnetic stirrer and thermometer, was charged with tert-butylamine (tBA, 34.9 g) and water (9 mL) and the mixture was heated to 55° C. Then, glycidol methyl ether (45.0 g) was added drop-wise. The reaction mixture was heated to 65° C. The reaction was exothermic and the mixture boiled at 89° C. When no further heat evolved, the mixture was stirred for another 3 hrs at 65° C.

(62) Workup: Excess tBA was removed in vacuo. Remaining mass 71.7 g, the product was isolated with a purity of 97% by temperature ramp gas chromatography. The yield was 99.6% of theory. The column used for gas chromatography was a DB1 with a length of 30 m, a diameter of 0.25 mm and a film thickness of 0.25 um by Agilent. The temperature ramp profile was as follows: Injection at 60° C., increasing the temperature at a rate of 15° C./min until a temperature of 280° C. was reached. This temperature was held for a further 1 minute.

(63) 1H NMR (400 MHz, Deuterium Oxide) δ 3.91-3.74 (m, 1H), 3.58-3.42 (m, 2H), 3.39 (s, 3H), 2.73-2.49 (m, 2H), 1.11 (s, 9H).