Conventional systems for the absorption of acid gases employ a liquid solvent; typical solvents include amines such as methyldiethanolamine (MDEA), monoethanolamine (MEA) or diethanolamine (DEA) and mixtures of solvents. These solvents absorb CO2, NOx, H2S and other acid gases. The solvent is contacted with the sour gas mixture (gas mixture including acid gases) in a column which may be a packed column, a plate column or a bubble-cap column, or a column with some other form of contact medium. In these systems, the gas and liquid streams flow countercurrently.

The prior art absorption systems suffer the disadvantage that in order to achieve a significant degree of gas/liquid contact, the columns have to be large and their operation is hampered by excessive foaming. In addition, the subsequent stripping section which removes the acid gas from solution must also be large, to handle the large volume of solvent used.

Since the operation normally takes place under high pressure and the fluids involved are highly corrosive, the capital costs of the large columns and subsequent stripping section is high.

It is an object of the invention to provide a method of selectively absorbing a fluid component from a fluid mixture with a high degree of efficiency and more economically than in existing methods. In particular,

it is an object of the invention to provide a method of selectively removing a selected gas component from a gas stream with a high degree of efficiency.

According to one aspect of the invention, there is provided a method of absorbing a selected gas component from a gas stream which comprises: bringing the gas stream into contact with a liquid including a solvent or a reagent for the selected gas component and subjecting the gas stream and liquid to turbulent mixing conditions in an ejector, thereby causing the gas component to be absorbed by the solvent or reagent.

The invention also extends to the apparatus for carrying out this method.

The turbulent mixing is very intense and results in extremely efficient gas liquid contact. The mixing regime is preferably turbulent shear layer mixing. The liquid entrained in the gas may be in the form of droplets for gas continuous fluid phase distribution.

The efficient mixing means that absorption can take place very rapidly and in a relatively small amount of solvent compared to that required in conventional absorption columns. This in turn means that the liquid duty in the equipment is dramatically reduced resulting in a consequential reduction in the size of any downstream regeneration section. At the same time, the mixing system used is simple and inexpensive compared to prior art systems, leading to reduced costs. Finally, an efficiency of approaching 100k for the removal of acid gas can be achieved for certain applications.

In addition, conventional absorbtion methods involve the evolution of heat which must then be removed from the system. While the method of the invention is capable of operation with a relatively low pressure drop across the mixing means, when greater pressure drop is employed, a cooling effect is achieved and this may

render the need for additional cooling unnecessary.

The absorption may be achieved by simply dissolving the gas or by way of a chemical reaction with the solvent.

Preferably, the ejector is a jet pump. In this application, the term "jet pump" is to be interpreted as an ejector further comprising a venturi passage.

Preferably, the method further includes the step of separating a gas phase and a liquid phase after the turbulent mixing. Preferably, the liquid phase is subsequently treated to remove the absorbed gas component.

The method is carried out as a continuous process with the gas mixture and solvent flowing co-currently.

The co-current flow eliminates the problems associated with foaming, since separation can easily be effected downstream of the ejector.

The turbulent mixing may be achieved by any convenient means, preferably an ejector, more preferably a jet pump. In one regime, the gas stream is supplied to the nozzle in the ejector, whereby the liquid is drawn downstream by the gas stream and the two phases are mixed. In another regime, the liquid is supplied to the nozzle in the ejector, whereby the gas is drawn downstream by the liquid and the two phases are mixed.

It will be appreciated that the invention is applicable to any absorption application where the reaction kinetics are rapid, for example the absorption of acid gas. The invention is also applicable to chemical reactions with fast reaction kinetics, where good mixing of the reactants is a requirement.

According to a more specific aspect of the invention, there is provided a method for removing a single selected component from a mixture of gases.

Alternatively, the method extends to removing a

plurality of gas components from a gas stream, either using a common solvent or reagent, or by respective solvents or reagents. According to a further aspect of the invention, the gas stream is a single gas which is absorbed.

Preferably, the gas stream and the liquid are formed into a homogeneous mixture by the ejector, the homogeneous mixture being cooled prior to separation into a gas phase and a liquid phase. Optionally, this phase separation occurs in a hydrocyclone.

Preferably, the solvent or reagent in the liquid phase is subjected to a regeneration treatment to remove the absorbed selected gas component. Preferably the regenerated solvent-containing liquid phase is recycled to the ejector.

Preferably, the regeneration is carried out by heating and/or by flashing off the absorbed gas component in a flash tank. Preferably, the post mixing cooling and the regenerative heating are achieved, at least in part by mutual heat exchange. Preferably, in instances where the gas stream is at a low pressure, the liquid is pumped to the ejector and thereby draws the gas stream with it through the ejector. Preferably, when the gas stream is at high pressures, it is conveyed to the ejector at a high pressure and thereby draws the liquid with it through the ejector.

The invention also extends to apparatus for carrying out such a method, comprising: an ejector; a cooler for the fluid stream from the outlet of the ejector; a hydrocyclone arranged to separate the cooled fluid stream into a gas phase and a liquid stream; a regenerator arranged to treat the separated liquid stream; and a recycle line arranged to convey the regenerated liquid stream to the ejector.

The apparatus may include a pump arranged to supply

liquid to the ejector. Preferably, the regenerator is a heater and/or a flash tank. Preferably, the ejector is a jet pump.

The invention may be considered to extend to the use of an ejector for absorbing a selected gas component from a gas stream by bringing the gas stream into contact with a liquid including a solvent or a reagent for the selected gas component, thereby causing the gas component to be absorbed by the solvent or reagent.

Preferably, the ejector is a jet pump.

The separation/absorption/reaction systems described are single operations, however it will be appreciated that multi separation/absorption/reactions may be performed. These may be carried out simultaneously or sequentially and may also be carried out in series or in parallel.

The improved efficiency possible for the removal of, for example, acid gases makes the present invention particularly valuable as awareness is increased of the potential damage to the environment that can be caused by acid gases in effluents such as combustion gas.

Furthermore, the small size of the apparatus compared to conventional absorption columns render the invention especially applicable to use in marine applications, such as on board shuttle tankers.

The invention may be put into practice in various ways and two specific embodiments will be described by way of example to illustrate the invention with reference to the accompanying drawings, in which: Figure 1 is a flow diagram of the process for use when the gas is under low pressure; Figure 2 is a flow diagram of the process for use when the gas is under high pressure; Figure 3 is a view of an ejector which is preferably used in this invention to generate turbulent

mixing conditions; Figure 4 is a block diagram of the apparatus as used in the batch test procedure for a mixture of N2 and CO2 as the test gas; Figure 5 is a block diagram of the apparatus as used in the batch test procedure for exhaust gas as the test gas; and Figure 6 is a view of the contactor as used in the batch test procedure.

Although the embodiments and experiments refer to a contactor, it will be appreciated that the method is applicable to ejectors, particularly jet pumps, since they generate similar homogeneous gas/liquid mixtures and turbulent mixing conditions.

In one embodiment of the invention, a continuous process operation for the removal of carbon dioxide (and other acid gases) from exhaust gas is shown in figure 1.

A liquid solvent stream 1, for example MEA (monoethanolamine), is conducted by a pump 2 to an ejector 3 capable of inducing turbulent mixing. An exhaust gas stream 4, including the CO2 which is to be removed, is drawn into the ejector 3 by the low pressure generated in the venturi by the liquid stream after it has passed through the pump (stream la). This arrangement provides an automatic means of self- regulation as the gas mixture to solvent ratio can be maintained for varying flow rates. At the outlet of the ejector 3 the liquid solvent and the exhaust gas stream are in the form of a homogeneous mixture (stream 5) and the mass transfer of the CO2 from the gas phase to the liquid occurs very rapidly.

The mixed two phase stream 5 is then conveyed to a cooler 6 and on into a hydrocyclone 7. The gas stream 8 is taken off and the liquid stream 9 passes on to a regeneration system. At this point in the circuit all

the CO2 is in the liquid phase (stream 9) and the gas stream 8 is free of CO2.

The regeneration of the liquid solvent is achieved by boiling off the CO2 in a heater 10. The CO2 is taken off as a gas stream 11 and the liquid solvent is optionally passed through a flash tank (not shown) to remove any residual dissolved gas before being recycled into the feed stream 1. The liquid solvent in stream 1 is topped up from the reservoir 12 as necessary to maintain a regular flow rate around the system.

It will be clear to a person skilled in the art that the cooler 6 and the heater 10 may be combined to form a heat exchange unit.

An alternative system for the removal of CO2 from a high pressure gas stream is shown in figure 2. A high pressure gas stream 20 containing the CO2 which is to be removed is conveyed to an ejector 21. The high pressure of the gas draws a controlled amount of liquid solvent, for example MEA, from the recycle stream 22 and, if necessary, from a reservoir 23 into the ejector 21.

At the outlet of the ejector 21 the two phases are in the form of a homogeneous mixture (stream 24) and the mass transfer of the CO2 from the gas phase to the liquid solvent takes place. The residence time may be as little as 0.1 seconds since, for example, the reaction kinetics for the absorption of CO2 by MEA are very rapid, although this residence time will vary with the solvent used and the gas to be transferred from the gas to the liquid.

The two phase mixture (stream 24) passes through a cooler 25 to a hydrocyclone unit 26. The gas stream free of CO2 is taken off in stream 27 and the remaining liquid stream 28 including the CO2 is passed to a regeneration system.

The liquid stream 28 is fed into a heater 29 to

remove the CO2 as a gas stream 30. This regenerates the solvent for re-use in the system. This solvent (stream 22) is then drawn into the ejector 21 by the low pressure generated in the ejector. Any shortfall in the solvent liquid is made up by addition from the reservoir 23. As in the first embodiment, the heater 29 and the cooler 25 can be combined to form a heat exchange unit.

The ejectors referred to in the above embodiments are preferably jet pump arrangements which are capable of inducing turbulent mixing. Figure 3 shows an ejector 120 comprising a first fluid inlet 121 for the high pressure fluid and a second fluid inlet 122 for the low pressure fluid. The high pressure fluid draws the low pressure fluid along the length of the ejector 120 to the outlet 123. The fluids are well mixed into a homogenised mixture in the region 124 at the outlet of the high pressure inlet 121.

It is clear that in a first regime, the gas stream is supplied to the nozzle 121, whereby the liquid is drawn downstream by the gas stream. In a second regime, the liquid is supplied to the nozzle 121, whereby the gas is drawn downstream by the liquid.

The invention is further illustrated by reference to the following examples. These serve to verify the operating principles of the two embodiments described.

In the first series of batch experiments conducted, the gas stream was a mixture of nitrogen (N2) and CO2 and the liquid solvent was a mixture of MEA and water. The reservoir pipe was kept under pressure using nitrogen gas. The contactor used was a Framo contactor generally as described in EP 379319 and shown in figure 6. The contactor injection pipe was adjusted to yield gas/liquid ratios in the range of about 3 to 5, depending upon the total flow rate.

As stated previously, it will be appreciated that

the method employed by the contactor is equally applicable to ejectors, particularly jet pumps, since they generate similar homogeneous gas/liquid mixtures and turbulent mixing conditions.

A schematic diagram for the first series of experiments is shown in figure 4. The contactor 51 is charged with an amount of the liquid solvent mixture from the reservoir 54 which is controlled by a valve 55.

A gas source 50 of the experimental N2/CO2 gas mixture is conveyed to the contactor 51 via a pipe 52 controlled by a valve 53.

At the outlet of the contactor 51 there is a 1 metre section of pipe 56 in which the mass transfer occurs. This section provides the residence time for the contacting materials. A set of 2 simultaneously acting fast closing valves 57 and 58 form a 1.5 metre analysis section 59 where the gas/liquid mixture can be captured, separated and sampled. At the top end of the analysis section there is a sampling point where a sample of the gas can be drawn off (not shown). At the lower end of the section there is a further sampling point where a sample of the liquid can be drawn off (not shown). The lower section of the sampling section is provided with means for cooling the liquid sample prior to its removal (not shown for clarity).

A further valve 60 separates the sampling section from a reservoir pipe 61 and is used to control the flow rate through the system. The reservoir pipe 61 is pressurised to a predetermined pressure by an independent nitrogen gas source 62 via a pipe 63 controlled by a valve 64. This pressure will be lower than that in the contactor to provide a pressure difference which will force the fluids through the system. The reservoir pipe 61 is inclined with respect to the horizontal to enable the liquid collected to be

drained off via a pipe 65 controlled by a valve 66 to a measurement drum 67 which is used to determine the amount of liquid passing through the system on each run.

The drum 67 has a drainage pipe 68 controlled by a valve 69.

In operation, the contactor 51, pipe section 56 and analysis section 59 are filled with the suitable strength solvent solution. The simultaneously acting valves 57 and 58 are closed and valve 60 is set to a position carefully adjusted to yield the required mass flow rate through the system for the predetermined pressure difference between the mixer and the reservoir pipe.

In the first series of experiments the contactor 51 is pressurised with the test gas of CO2-rich nitrogen to a pressure of 50 barg. The reservoir pipe 61 is pressurised with nitrogen to a predetermined value typically between 16 and 48 barg, providing a range of flow rates through the system.

Before the experiment starts, a sample of the test gas is taken to determine the level of CO2 in the gas.

The experiment commences with the activation of, the simultaneously operating valves 57 and 58. The liquid and the gaseous solution flow co-currently through the system to the reservoir pipe 61. The pressure in the mixer is maintained at 50 barg during the 10 second test run by manual supply of the test gas from a cylinder fitted with an accurate manometer. This makes it possible to record the amount of spent gas for each experiment.

After 10 seconds the 2 operating valves 57 and 58 are closed simultaneously. A sample of gas from the analysis section is extracted from the upper sampling point immediately after the valves have closed. This is then tested for content of CO2 by gas chromatography.

The machine used was a Chromopack Model CP-2002 gas chromatograph.

In order to verify the mass balance, a liquid sample of the amine solution in the analysis section is taken from the lower sampling point. Before the sample is taken the liquid in the analysis section is cooled using nitrogen gas surrounding the pipe section 59. The liquid sample is analyzed using a titration technique specially developed for CO2.

At the end of each run, the liquid from the reservoir pipe 61 is released into the measurement drum 67 to measure the amount of liquid expended in the course of the run. The results of the tests are shown in Table 1 below: MEA mol% gas liquid total gas volume wt% CO2 in flow flow flow fraction exit rate rate rate gas (m3/hr) (m3/hr) (m3/hr) 50 0.005 10.34 4.63 14.97 0.69 50 0.003 11.76 3.92 15.68 0.75 50 0.005 12.12 3.92 16.04 0.76 50 0.002 10.87 3.92 14.79 0.73 50 0.006 10.08 3.96 14.04 0.72 50 0.007 11.7 3.6 15.3 0.76 50 0.019 10.44 3.24 13.68 0.76 50 0.006 7.2 3.24 10.44 0.69 50 0.007 15.48 3.24 18.72 0.83 25 0.009 10.08 4.68 14.76 0.68 25 0.005 9 ~ 3.96 12.96 0.69 25 0.006 9 3.96 12.96 0.69 25 0.003 6.84 3.6 10.44 0.66 25 0.005 14.04 4.32 18.36 0.76 5 2.03 14.4 3.6 18 0.80 5 0.5 15.12 3.24 18.36 0.82 5 2.95 17.28 3.24 20.52 0.84 5 3.65 18.72 1.8 20.56 0.91 5 1.63 12.6 3.96 16.56 0.76 5 2 14.76 3.96 18.72 0.79 5 2.13 15.84 3.6 19.44 0.81 5 0.31 7.92 3.6 11.52 0.69 5 1.25 7.92 3.6 11.52 0.69 5 2.32 10.44 3.6 14.04 0.74 5 2.67 11.16 3.6 14.76 0.76 5 3.4 18 3.6 21.6 0.83 Table 1

In all cases the gas feed composition was 10.5 mol per cent CO2 in nitrogen.

The results show that virtually all the CO2 is absorbed from the gas to the liquid solvent for the 50% and 25% mixture for all the flow rates tested. Only on reduction of the MEA concentration to a mere 5% by weight does the amount of CO2 remaining in the gas reach appreciable levels.

From the measurements at the 5% level, it can be seen that the absorption efficiency decreases with an increasing gas flow rate and gas volume fraction. This result is expected since the already lean solvent mixture (only 5 MEA) has a diminishing capacity to absorb all of the CO2.

The gas chromatograph measurements of the CO2 were verified using the data obtained from the titration of the liquid sample. A mass balance calculation on the CO2 through the system showed that the CO2 which was in the test gas had been transferred to the liquid.

In a second set of experiments, the contactor 51 was only pressurised to a low pressure (in the range 0.5 to 2 barg) and the reservoir pipe 61 is left open to atmospheric pressure. This gave a driving force of between 0.5 and 2 bar. The only change to the apparatus from the first set of experiments is the addition of a small hydrocyclone at the top of the gas pipe to separate the gas and liquid after reaction. This means that there are no entrained droplets in the gas sample.

In these experiments, the liquid solvent mixture is a 50% solution of MEA and the gas feed composition was 9.4 mol per cent CO2 in nitrogen. As for the first set of experiments, the test run lasted for 10 seconds and the pressure in the contactor was maintained by manual supply of the test gas. The results are shown in table 2 below.

Contactor mol% gas liquid total gas P (barg) CO2 flow flow flow volume in rate rate rate fraction exit (m3/hr) (m3/hr) (m3/hr) gas I 0.5 0.59 2.16 4.68 6.84 0.316 0.5 0.87 1.80 4.32 6.12 0.294 0.5 0.80 2.16 3.96 6.12 0.353 1 0.80 3.24 4.68 7.92 0.409 1 0.95 3.24 4.32 7.56 0.429 1 1.20 3.42 4.32 7.74 0.442 1.5 1.10 4.68 4.32 9.00 0.520 1.5 0.76 4.68 4.14 8.82 0.531 1.5 1.27 5.04 4.32 9.36 0.538 2 0.73 6.12 5.22 11.34 0.540 2 1.10 6.48 5.76 12.24 0.529 2 0.82 6.12 5.40 11.52 0.531 0.5 0.13 2.52 3.96 6.48 0.389 0.5 0.61 3.60 3.96 7.56 0.476 0.5 (1) 0.45 2.16 3.69 5.85 0.369

Table 2 (1) -this experiment had a run time of 20 seconds.

The small pressure difference driving the fluids through the system results in there being more liquid relative to the gas than in the previous experiments.

Even at these lower gas volume fractions, most of the carbon dioxide is removed from the gas phase. It will be noted that there is no real trend from a pressure difference of 0.5 to 2.0 bar so it will be apparent that this method is applicable down to lower pressure differences than 0.5 bar. Such pressure differences may be present, for example, in exhaust gas systems.

In a third set of experiments, exhaust gas was used

in place of the experimental N2/CO2 mixture. A schematic diagram of the apparatus for these experiments is shown in figure 5. In general, the system is operated in a similar way to the system shown in figure 4. As for the previous experiments, the contactor 51, pipe section 56 and analysis section 59 are charged with an amount of the liquid solvent mixture from the reservoir 54. The exhaust gas comes from a diesel engine 75 and passes through the contactor with a minimum loss of temperature. In contrast to the earlier experiments, the contactor 51 is not pressurised.

In these experiments, the gas mixture is exhaust gas from a Yannmar 4TN84E 15 KVA water cooled diesel engine 75. A 30% load was placed on the diesel engine to increase the exhaust gas temperature and to obtain a higher level of CO2 in the exhaust gas. An orifice plate 74 is provided in pipe 71 for continuous flow measurement of the exhaust gas.

Before the experiment starts, a sample of the exhaust gas is taken at point 72 to measure the CO2 content in the exhaust gas. In operation, the valve 70 is closed, allowing exhaust gas to enter the contactor 51. When a pressure of approximately 0.4 barg has built up in the contactor, the two valves 57 and 58 are opened simultaneously. As in the previous experiments, the liquid and the gaseous solution flow co-currently through the system for 10 seconds into the reservoir pipe 61 before the valves 57 and 58 are closed simultaneously.

A sample of gas from the analysis section 59 is extracted from the upper sampling point immediately after the valves are closed. As before, the sample is tested for content of CO2 by gas chromatography using a Chromopack Model CP-2002. At the end of each run the expended liquid is released from the reservoir pipe 61 to the measurement drum 67 and weighed. In these experiments, the liquid solvent mixture is a 50% solution of MEA. The results for these tests are shown in Table 3 below: C1 T S C2 QG QL QT G/L 1.4 30 15 0.03 45 5.40 50.40 8.33 1.4 30 15 0.04 45 5.40 50.40 8.33 1.4 30 14 0.06 45 5.04 50.04 8.93 4 50 14 0.19 45 5.04 50.04 8.93 4 50 16 0.15 45 5.76 50.76 7.81 4 50 14 0.09 45 5.04 50.04 8.93 4 1 50 14 0.08 45 5.04 50.04 8.93 4 50 13 0.10 45 4.68 49.68 9.62 15.5 65 12 0.10 45 4.32 49.32 10.42 15.5 65 15 0.10 45 5.40 50.40 8.33 15.5 65 16 1.40 45 5.76 50.76 7.81 15.5 65 15 1.00 45 5.40 50.40 8.33 15.5 65 14 0.20 45 5.04 50.04 8.93 2.8 122 15 0.22 59 5.40 64.40 10.93 2.8 133 15 0.07 59 5.40 64.40 10.93 2.8 128 15 0.06 59 5.40 64.40 10.93 2.8 132 14 0.06 59 5.04 64.04 11.71 2.2 136 15 0.10 59 5.40 64.40 10.93 2.2 133 14 0.30 59 5.04 64.04 11.71 3.4 123 5.5 0.37 59 1.98 60.98 29.80 3.4 123 6.5 0.25 59 2.34 61.34 25.21 3.4 123 6.5 0.10 59 2.34 61.34 25.21 3.4 123 6.5 0.27 59 2.34 61.34 25.21 3.4 123 6 0.27 59 2.16 61.16 27.31 9.98 118 7 0.22 59 2.52 61.52 23.41 9.98 118 7 0.01 59 2.52 61.52 23.41 9.98 118 - 6.5 0.01 59 2.34 61.34 25.21 Table 3

Key to Table 3: C1 - mol % CO2 in exhaust gas T - Temperature of the exhaust gas (OC) S - Expended solvent (1) C2 - mol % CO2 in exit gas QG - gas flow rate (m3/hr) Qt - liquid flow rate (m3/hr) QT - total flow rate (m3/hr) G/L - gas/liquid ratio As can be seen from the above results, virtually all the CO2 is removed from the gas and absorbed into the liquid solvent. It is also clear that the removal efficiency is higher for higher concentrations of CO2 in the feed gas which is significant for gas turbine applications. However, the efficiency of the system is still high for low concentrations of CO2 in the feed gas.

It is noted that there is no significant trend when the temperature of the exhaust gas is varied. This is probably because there is a "quenching effect" when the cool solvent solution contacts the exhaust gas. Reducing the amine flow rate does not significantly change the removal efficiency indicating that the system can be operated with higher gas/liquid ratios, for example higher than 30.

It will be apparent to a person skilled in the art that the results from the three sets of experiments above are not dependent upon the gas to be absorbed or on the solvent used to absorb that gas. Therefore it is clear that the above method of selective transfer of a gas from a mixture of gases to a liquid solvent for that gas is applicable to any gas and any respective solvent.

As stated above, although the examples have been conducted using a turbulent contactor such as that manufactured by Framo Engineering A/S (as shown in figure 6), the method is equally applicable to ejectors, especially jet pumps.