Guard layers for rapid cycle pressure swing adsorption devices

TECHNICAL FIELD [0002] The present invention relates to embodiments of a method for protecting beds in rapid cycle pressure adsorption (RCPSA) devices against contamination. In particular, it concerns the construction of protective layers that can be used on the supply side of deposits.

BACKGROUND ART [0003] Pressure swing adsorption (PSA) devices are used to separate at least one preferentially adsorbed component from at least one more difficult adsorbed component in a liquid feed mixture. Gas separation by means of PSA is achieved by synchronous pressure change and gas inversion by means of a set of adsorber beds which adsorb preferentially adsorbed component (s) in the feed gas mixture. During each cycle, the pressurized feed gas mixture is first introduced to the feed side of the adsorber bed. The more difficult adsorbed component (s) flows through the adsorber beds, while the preferential adsorbed component (s) is (are) absorbed. Thus, the gas taken up on the side opposite the supply side of the bed (i.e. on the product side) is concentrated in the harder adsorbed component (s). The adsorbent in the beds is regenerated later in the cycle by cutting off the pressure supply

06PL00

Feed gas mixtures, reducing the pressure in the bed, and thus desorbing preferentially adsorbed component (s) and expelling them from the bed or cleaning the bed.

[0004] A simple PSA cycle may therefore include a single pressure boosting step in which the concentrated gas in the harder adsorbed component (s) is obtained on the product side of the deposit, and a pressure reduction step in which the concentrated gas in easily adsorbed ( component (s) is (are) discharged from the supply side of the deposit. However, in order to improve purity, performance and efficiency, typically complex PSA cycles are used in this field. These more complex cycles use gas flows with pressure reduction and re-pressure between the supply side and the product side of the adsorbent bed at various stages in the cycle. Many more adsorption beds are needed in these more complex PSA cycles.

[0005] Conventional commercial PSA equipment currently uses solid bed adsorbents in the form of beads or pellets with dimensions from about 1 mm to 4 mm. In order to obtain higher cycle speeds, the gas velocities within the adsorbent beds need to be increased, especially for devices with multiple adsorbent beds. The maximum cycle speed for such conventional PSA devices with bead beds is, however, limited by factors such as fluidization, bead abrasion, and to some extent valve speed and valve life.

[0006] Recently, rapid-cycle PSA (RCPSA) devices have been developed that operate at cycle speeds greater than about 2 cycles per minute. The use of structured adsorbent beds containing

06PL00

The laminate sheets of the immobilized adsorbent makes it possible to eliminate problems associated with fluidization and abrasion of the beads, and also allows reducing pressure drops in the beds. The use of such an adsorbent in the form of laminate sheets, in combination with the use of compact high-speed rotary valves, allows obtaining high PSA cycle speeds with high efficiency.

[0007] US Patent Nos. 4,968,329 and 5,082,473 and Publication Number 2002-0170436 disclose preferred embodiments for an RCPSA bed containing helically wound adsorption sheets of 1 mm or less in thickness. The adsorption sheet is wound in a spiral together with the spacer sheet, e.g. a wire mesh spacer sheet such that the spacer sheet defines flow channels between adjacent adsorbent sheets. US Patent No. 5,082,473 suggests that the ratio of half the thickness of the adsorption sheet to the channel gap (b / t) is preferably near unity, but may be between 0.5 and 2.0, or in other words, the channel gap may be somewhere between 0.25 and 1 thick adsorption sheet. This means that the share of channels in the deposit (where the share of channels is defined as the ratio of the channel volume to the total volume of the bed) is less than 50%.

[0008] In many PSA applications, the feed streams may contain small amounts of impurities that are even more preferably adsorbed by the adsorbent than the one or the component (s) to be adsorbed (s) . Such impurities can be characterized by very high, and sometimes irreversible, adsorption and can deactivate or poison the adsorbent, thus impairing its efficiency and selectivity, and thus its ability to function properly. For example, in the separation of oxygen from air, strongly selective for is commonly used

06PL00

Nitrogen exchangeable for cations, characterized by a low silica-alumina ratio zeolites, but these zeolites are very sensitive to water contamination in the feed stream.

[0009] Various methods can be used in conventional PSA to remove contaminants from the feed gas stream and thereby protect the adsorbent bed from degradation. These include purification of the feed gas upstream of the inlet (e.g. cooling of the feed gas, followed by condensation upstream of the PSA unit) or adsorption on regenerative adapted protective beds (which are usually located on the feed side in the adsorbent housing of the PSA unit itself). Protective beds are used to adsorb virtually all contaminants from the feed stream before it enters the main adsorbent bed. And, the protective beds are regenerated at the same time as the main adsorbent bed in a typical PSA cycle. A desiccant is usually used as a protective layer on the supply side of the bed to remove water impurities from the feed stream.

[0010] Protective layers for controlling contaminants in the PSA bed do not contribute to the main adsorption process, and thus effectively add undesirable dead volume to the PSA bed. Preferably, the void space on the adsorbent bed sides should be minimized for better regeneration. It is therefore desirable to minimize the length and volume of the inner voids of such protective layers, while effectively removing impurities in the feed stream. In the prior art, this is generally done by maximizing the amount of adsorbent protective material present in the protective layer, while allowing permissible gas flow through the protective layer. In the PSA applications in which it is used

06PL00

Zeolite adsorbents, in which the main impurity is water, usually from 5 to 30% of the adsorber bed is occupied by a protective layer comprising alumina, silica gel, activated carbon, or a combination thereof. The feed gas is dried to 0.1 to 5 ppm water vapor before contacting the zeolite adsorbent layers.

[0011] Conventional PSA devices are less sensitive to the presence of contaminants in the feed stream than the recently developed RCPSA devices with faster cycle times. The previous ones have relatively longer adsorption beds, on which diffusion of impurities must occur and have a relatively larger amount of adsorbent, so that if a given amount is deactivated, it corresponds to a smaller share of the total. In addition, the degradation rate depends on the cumulative number of cycles that occur, which is lower in conventional PSA devices over a given period of time.

[0012] This unexpected sensitivity of RCPSA devices to feed stream contaminants has been experimentally recognized in US Patent No. 7,037,358. It also discloses various methods to protect RCPSA equipment against pollution, especially against water. For example, protective layers similar in construction to the main adsorption layers have been used on the supply side of the adsorber bed. These layers have been generally disclosed as thin and having a large surface area, with flow channels that have a narrow hydraulic radius to overcome mass transfer restrictions. It is necessary to reduce the length of the narrow flow channels to maintain the desired small pressure drop in the protective bed.

BRIEF DESCRIPTION OF THE INVENTION [0013] The present invention relates to embodiments

06PL00

The method relates to adsorbent beds in RCPSA devices and to the protection of these beds against contamination. In particular, it concerns the construction of protective layers used on the supply side of deposits. Such protective layers preferentially adsorb contaminant components (e.g., water) from the feed gas and expel them when the bed reverses flow during the RCPSA drainage or purge phase.

[0014] It has been found that the effectiveness of the protective layer can be improved by using less protective adsorbent, and instead using structures that provide greater access to contaminant adsorption.

[0015] According to a first aspect of the invention, there is provided a method of protecting beds in a RCPSA device against contamination, which RCPSA device comprises at least one adsorbent bed and a protective layer on the feed side of said adsorbent bed, which protective layer comprises a laminate sheet containing a protective adsorbent, what this method contains:

the use of the proportion of channels in the protective layer greater than 50%, which proportion of the channels is defined as 1 minus the proportion of the adsorption sheet, which proportion of the adsorption sheet is defined as the volume of the laminate sheet divided by the volume of the protective layer; and the introduction of a textured surface into the laminate sheet, and thus the introduction of surface voids into the laminate sheet, which surface voids are defined as the volume defined by the thickness TA from the top to the top of the laminate sheet minus the volume of the envelope of the laminate sheet, with the proportion of empty surfaces of the laminate sheet

06PL00

EP 2 125 162 B1 is between 20 and 50%, and the proportion of surface voids is defined as the volume of surface voids divided by the sum of the volumes of the laminate sheet plus the volumes of surface voids.

[0016] The adsorption bed according to the invention is intended for use in a PSA device, in particular an RCPSA device, and has a protective layer on the supply side of the bed. The protective layer comprises a laminate sheet and this sheet contains a protective adsorbent. The protective layer uses less protective adsorbent and there is a corresponding increase in the volume or proportion of channels in the protective layer. The proportion of channels in the protective layer is greater than 50%, and usually ranges from 50% to 75%. In some disclosed embodiments, the proportion of channels in the protective layer is greater than 50% to at least 75%.

[0017] Water is a common impurity in many RCPSA applications. In such applications, the protective adsorbent used may be a desiccant or may contain a desiccant such as activated alumina.

[0018] The laminate sheet may be characterized by a thickness between about 5 and 200 microns and a surface area / (volume of protective layer) ratio greater than about 5000 square meters per cubic meter.

[0019] In order to increase the effectiveness of the protective layer, the surface of the laminate sheet is textured, which introduces surface voids into the laminate sheet. These surface voids are between 20 and 50%. The laminate sheet can be made using a wire mesh. A method of texturing the surface of the sheet is by coating the appropriate wire mesh with a suspension of protective adsorption material containing a small amount of substance

06PL00

Solids (i.e. "pouring").

[0020] The protective layer can be made by spiral winding the laminate sheet and the spacer sheet together. The spacer sheet may contain a wire mesh.

[0021] The invention is preferred for RCPSA devices operating at a cycle speed greater than about 5 cycles per minute.

[0022] An experimental method was also developed to test the effectiveness of the protective layers and thus to assist in determining the appropriate operating parameters for the RCPSA device. This method includes providing a series of windows for gas sampling at various locations along the length of the test bed exposed to the RCPSA cyclic process and determining the concentration of contaminant components at these locations using appropriate sampling devices (e.g., moisture analyzer, gas chromatograph).

BRIEF DESCRIPTION OF THE DRAWINGS [0023]

Figure 1 is a schematic diagram of an RCPSA system comprising 2 spiral wound beds with protective layers for adsorbents in each bed. Sample sensors for testing the effectiveness of the protective layers were also presented.

Figure 2 is a schematic cross-section of a protective layer comprising a wound spirally textured laminate sheet and a wire mesh spacer.

Figure 3 is a graph of the distance from the supply inlet (in inches) as a function of water concentration (ppm) showing the concentration of water as a function of distance from the supply side for the protective layers according to the invention tested in the Examples.

06PL00

EP 2 125 162 B1

Figure 4 is a graph of the distance from the supply inlet (in inches) as a function of water concentration (ppm) showing the water concentration as a function of distance from the supply side as a function of the raw material / product ratio (F / P) for one of the protective layers according to the invention tested in the Examples.

Figure 5 is a graph of the distance from the supply inlet (in inches) as a function of water concentration (ppm) showing the water concentration as a function of distance from the supply side as a function of cycle speed for one of the protective layers according to the invention tested in the Examples.

DETAILED DESCRIPTION [0024] Unless expressly stated otherwise, all technical and scientific terms used herein have the same meanings as commonly understood by one of ordinary skill in the art.

[0025] The effectiveness of the protective layer in RCPSA devices can be improved by using laminate sheet-based structures that use less protective adsorbent than is generally used in laminate-based adsorption layers. Improved protective layers have channel proportions that are greater than 50%.

[0026] FIG. 1 is a diagram of a simple RCPSA 1 system containing two spirally wound beds 2 with protective layers for adsorbents in each bed. Each bed 2 includes a main adsorbent layer 3 for separating the main components in the feed gas mixture, as well as a protective layer 4 for protecting the main adsorbent layer 3 from impurities present in the feed. Mixture 5 of the feed gas mixture. The gas mixture 5 is alternately fed to the supply sides of each bed 2 via the supply side rotary valve 6. Gas 7, which is a product, is staggered

06PL00

Obtained on the product side of each bed 2 via the product side rotary valve 8. In this simple arrangement, the exhaust gas 9 is withdrawn from the supply sides of the bed 2 through the rotary valve 6 during the regeneration stage in the RCPSA cycle. In FIG. 1 also shows example sampling sensors 10 that can be used to test the effectiveness of protective layers in experimental RCPSA equipment.

[0027] In commercial embodiments, the layers of the main adsorbent 3 may in fact contain more than one adsorbent material to separate more than one main component in the feed gas mixture. Different adsorbent materials can be configured sequentially along the adsorbent layer 3 (i.e., in a series of adsorbent layers). In a similar manner, the protective layers 4 may also contain more than one protective material to remove multiple contaminants from the feed gas mixture. Also in this case, various protective materials can be configured sequentially along the protective layer 4. For example, in applications where a range of operating temperatures can be found, multiple protective layers may be needed to remove one or more contaminants from the feed gas mixture.

[0028] In the RCPSA device of FIG. 1, the adsorbent layers 3 comprise a porous laminate adsorbent sheet comprising a suitable adsorbent that is spirally wound together with a suitable spacer sheet. A spacer sheet is positioned to determine the passage channels for gas flow and access to both sides of the wound laminate adsorption sheet. The thickness of typical laminate adsorption sheets is in the range from 5 to 200 microns. The height or gap of the channels formed by the spacer sheet is

06PL00

Usually less than or equal to the thickness of the laminate adsorption sheet.

[0029] In one embodiment of the present invention, the protective layers 4 also comprise a porous laminate sheet containing a suitable protective adsorbent that is spirally wound together with a suitable spacer sheet. However, in this case, the channels defined by the spacer sheet occupy more than 50% of the volume.

[0030] FIG. 2 is a close-up of a schematic cross-section of the protective layer 4. The view shown is parallel to the rolling axis. In FIG. 2, two adjacent turns of the laminate protective adsorption sheet 15 separated by the spacer sheet 16 are visible. (It should be noted that these turns are obviously curved, but this is not so apparent in the approximation shown in FIG.

2.) The adsorption sheet 15 is a porous sheet containing protective adsorbent particles and a suitable adhesive coating the woven wire mesh substrate. FIG. 2 shows the envelope defined by the porous adsorption sheet 15. (This envelope is the theoretical surface that surrounds the solid places in sheet 15 and which connects the holes of its pores. Sheet 15 shown in FIG. 2 has a textured or "wavy" surface 15a. The thickness TA is the thickness of the adsorption sheet 15, measured from apex to apex on a textured surface. The spacer sheet 16 is a woven wire mesh. FIG. 2 shows the envelope defined by the spacer sheet 16. The thickness TS is the thickness of the spacer sheet 16, measured from the top to the top on the envelope of the spacer sheet. The adsorption sheet 15 physically contacts the spacer sheet 16 at many places. However, significant texturing of the adsorbent sheet 15 creates relatively large voids between the two

06PL00

The boundaries shown in FIG. 2 as surface voids 17. In this case, the proportion of the channels is defined as 1 minus the proportion of the adsorption sheet, where the proportion of the adsorption sheet is equal to the volume within the envelope of the adsorption sheet divided by the total volume of the protective layer. With reference to FIG. 2, the proportion of the channels is therefore also the same as (volume of channels / total volume), where the volume of channels is equal to the volume inside the envelope of the spacer grid plus the volume of surface voids. (The relatively small volume occupied by the wires in the spacer mesh itself thus remains within the volume of the channels.) In the protective layers 4 according to the invention, the proportion of the channels is greater than 50%, and typically is between 50% and 75%.

[0031] TA and TS thicknesses are easily measurable with a micrometer. Estimation of hollow surfaces can be done by estimating the effective thickness of the sheets using optical microscopy of cross-section samples. However, the preferred method is based on liquid buoyancy testing. In this method, the pores in the test sample of the adsorption sheet are filled with a liquid such as water. Then the water-filled adsorption sheet is immersed in a second liquid that is immiscible with water (for example chlorofluorocarbon) and the volume of the displaced second liquid is determined. The volume displaced is therefore the volume inside the envelope. The surface void is then calculated as the difference between the volume containing surface voids (i.e. the surface area Ta * of the sheet sample) minus the volume inside the envelope of the adsorption sheet (i.e. the volume of the second displaced liquid).

[0032] The choice of absorbent protective material depends on the use of separation. Typical materials suitable for use as protective adsorbents include activated alumina, aluminosilicate gels, gels

06PL00

Silica, zeolites (e.g. zeolite Y or zeolites with high silica content), activated carbon, carbon molecular sieves and combinations of these materials. Combinations of materials, for example activated alumina plus zeolite such as 5A and / or 13X, may also be commonly used. Water is a common contaminant in PSA applications and suitable water protective adsorbents include desiccants such as activated alumina.

[0033] In RCPSA devices, it has been found that having more access to protective adsorbent material is more important than having more material. The design with a larger proportion of channels provides greater accessibility. The measure of availability is the ratio of surface area to volume (SA / V) of the protective layer (where the surface area in question is the area of ​​the envelope envelope, not the adsorbent particles themselves, and the volume is the volume of the protective layer, and thus includes both the protective sheet adsorption and channel). As the SA / V ratio increases, the thickness of the protective adsorption sheet must decrease. It is difficult to measure the surface area of ​​the bounding box of a textured sheet as shown in FIG. 2. Thus, in the Examples below, the SA / V ratio was determined assuming that the protective adsorption sheet was a smooth sheet with a thickness of TA. In this case, the SA / V protective layer was greater than about 5000 square meters per cubic meter.

[0034] In addition, texturing the surface of the protective adsorption sheet increases the availability and effectiveness of the protective layer. In addition, the increased volume of channels introduced through the surface voids results in a reduced pressure drop over the protective layer and hence improved performance.

06PL00

[0035] The surface of the sheet can be easily textured by ribbon coating the corresponding wire mesh with a low solids (i.e. "pourable" or low viscosity) suspension. Preferably, the wire mesh is woven and therefore has a repeating pattern. After coating, the suspension settles in the holes in the woven mesh before drying, thereby introducing the texture into the finished dried surface. Texture (e.g., amplitude and period) are mainly functions of suspension properties (e.g., viscosity) and weave pattern (e.g., wire spacing in the weave). In this way, surface voids of between about 20 and 50% can be easily incorporated into the protective layer.

[0036] It should be noted that the protective layers 4 are exposed to high speed feed gas streams during the feed step in the RCPSA cycle, which can cause erosion over a very large number of cycles. To protect them, flow dividers / interfering elements of the multilayer perforated plate or grid type streams (and not shown in FIG. 1) to distribute the gas flow over the inlet surface of the protective bed. For example, a combination of a large metal mesh (e.g., number 10) and fine metal mesh (e.g., number 325) may be used. The former provides structural support and coarse interference / distribution of the streams, while the latter provides interference / distribution of the streams into fine.

[0037] RCPSA devices operate at cycle speeds greater than about 5 cycles per minute, significantly higher than conventional PSA devices. Therefore, gas velocities through the protective layer and pressure drop per unit length are also higher. When working at a high level of product recovery, another issue is

06PL00

The ratio of the volume of raw material to the gas being the product required for the proper effectiveness of the protective layer.

[0038] The volume ratio of raw material to product gas is a measure that is close to the volume of gases flowing through the protective layer in opposite directions during the feed and regeneration stages of the cycle. Operation with a relatively high ratio of raw material to product means that the device works at low product performance, which is undesirable from this point of view. However, this also means that a relatively large amount of gas is withdrawn back through the protective layer during the regeneration phase of the cycle to desorb the impurity from the protective layer. A lower ratio of raw material to product is better from the product performance perspective, but means less exhaust gas available for desorption of contamination, and thus may not be suitable for the effectiveness of the protective layer.

[0039] Regarding the effectiveness of the protective layer, when the feed gas is flowing, the target equilibrium equilibrium concentration of the impurities at the end of the protective layer must be low enough to minimize degradation of adsorption on the main absorbent behind it, looking in the direction of flow. In practice, this generally means that impurity levels should be reduced to <10 ppmv and preferably <1 ppmv. For example, in the case of steam on zeolites, the water level is preferably <10 ppmv, and more preferably <1 ppmv for long-term cyclic stability.

[0040] Testing the effectiveness of the protective layer on an RCPSA device may be more difficult than on conventional PSA devices. It was found that

06PL00

Conventional methods based on measuring the concentration of impurities in the product stream or monitoring PSA performance as a function of time have proved to be erroneous and unpredictable. Instead, a new method was developed based on measuring when contamination through the protective layer occurs during the feed or production phase of the PSA cycle.

[0041] The new test method consists in installing gas sampling windows along and inside the protective layer, as well as in an intermediate gap between the protective layer and the main adsorbent layer. The appropriate arrangement of the sampling windows 10 for the disclosed embodiments is shown in the complete RCPSA device 1 in FIG. 1. However, it is expected that research for development purposes will typically be performed on a test stand using a single test bed. The bench uses computer-controlled solenoid valves to control the gas sequence and timing to and from the test bed, and manual control valves to control the flow rate. Accumulation tanks are also used to collect the evacuated gases and recycle them to the same test bed to simulate cyclic flows in a multi-PSA. Such a test stand enables the experimental assessment of the effectiveness of a multi-PSA cycle using a single test bed over a wide range of cycle speeds.

[0042] In the test device, a small proportion of internal gas is extracted for analyzing the content of contaminants (for example with a moisture analyzer) at a desired point in the cycle. If sampling is only needed during the supply or production stage, each sampling window can be equipped with a distributor that only opens during the stage

06PL00

EP 2 125 162 B1 high pressure supply. It has been found that an overpressure valve or non-return valve with a predetermined response pressure is an effective means of sampling gas only during the feed stage with the highest PSA cycle pressure. A computer-controlled solenoid valve is a suitable alternative. For continuous measurement there is an effective low-flow control valve connected to each valve of the sample window and a section of capillary tube with a small opening is a suitable alternative. It was found that this sampling procedure did not affect the performance of the test bed (confirmed experimentally by establishing a cyclic steady state with and without sampling and observing the same results within experimental error).

[0043] The materials, methods and examples described herein are illustrative only and are not intended to limit the invention to the specific materials, methods and examples disclosed.

Examples of laminate adsorbent [0044] Three experimental spiral wound laminate test beds were made using zeolite as the main adsorbent and activated alumina as the protective layer drying agent. Zeolite laminate sheets have been reinforced with an inert binder and stainless steel woven mesh. These laminate sheets were wound helically together with a stainless steel woven mesh spacer sheet on a central mandrel to form the main layers. The protective layers contained protective sheets of activated alumina with variable surface voids. They were also reinforced with an inert binder and woven steel mesh

06PL00

They were spirally wound together with a stainless steel mesh spacer sheet to form protective layers. Each test bed consisted of one experimental protective layer and a main adsorbent layer enclosed in a stainless steel protective tube.

[0045] These experimental beds were tested in a single bed test device, based on an electromagnetic valve and controlled by a computer, which was used to simulate operation in a real RCPSA working device. The test feed gas was humidified compressed air. For regeneration, the beds were simply allowed to reduce pressure on the supply side.

<td>alive (these</td><td colspan="2">would be</td><td>in the same</td>

<td> 4,4</td><td>cm</td><td>from</td><td>page</td>

<td> 0,3</td><td>cm</td><td>in</td><td>place</td>

<td colspan="2">protection</td><td>and</td><td>layer</td>

[0046] As previously described with reference to FIG. 1, gas samples were taken to determine the water content at various locations during the cycle of adsorption (production). (Water content was not measured during regeneration). The sampling windows were located at a distance of 1.3 to 5.1 cm from the supply side of the protective layer), at the supply distance (it would be in the AA transition gap between the protective layer of the zeolite) and at a distance of 7.6 cm from the supply side ( in the zeolite layer alone). The Amatek Model 5800 Moisture Analyzer capable of measuring from 0.02 ppmv to 1000 ppmv was used for continuous monitoring of water concentration during the measurement of low concentrations. For intermediate water concentrations, Cosa XDT-PM / LPDT humidity and temperature transducers were used. To measure the highest concentrations, Vasiala HPM238 Moisture and Temperature transducers or Elextronics EE29 were used.

[0047] Table 1 below summarizes some significant

06PL00

Physical properties of the protective layers tested. Each desiccant sheet in protective layers was prepared as described above by coating the wire support mesh with a low viscosity suspension containing activated alumina. In all cases, a 70 mm stainless steel woven mesh (made of 0.08 mm wires) was used. The length of each protective layer in the bed was

6.4 cm. The thicknesses from the top to the top of the laminate drying sheet and spacer were measured with a micrometer. The total desiccant content and surface area per unit volume of desiccant was determined. The surface voids of the drying sheets were measured using the method described above

Archimedean.

activated

In this case, the alumina-containing sheets were first saturated with water to completely fill the voids contained therein. The samples were then immersed in a Vertrel chlorofluorocarbon liquid that was not mixed with water to determine the displaced volume. Thus, this displaced volume did not contain surface voids of the drying laminate. The surface voids were then calculated as the difference between the calculated volume of the drying laminate (based on its measured length, width and thickness TA, as determined in FIG. 2) and this displaced volume. Table 1 also shows the effective thickness of the drying sheet and the effective height of the channel. The effective thickness of the drying sheet represents the equivalent thickness of a flat, non-textured drying sheet that would occupy the same volume as the actual textured drying sheet. And, the effective channel height represents the channel equivalent height that would exist if the actual textured desiccant sheet in the protective layer was replaced by an equivalent, flat, non-textured desiccant sheet. At the end

76P35206PL00

The proportion of channels (equal to the effective height of the channel divided by the effective height of the channel plus the effective thickness of the laminate) is given.

Table 1

<td>Protective layer sample number</td><td> 1</td><td> 2</td><td> 3</td>

<td>The amount of desiccant in the laminate sheet GAA / m²)</td><td> 103,3</td><td> 56</td><td> 39</td>

<td>Drying sheet thickness, T_AND (Mm)</td><td> 0,16</td><td> 0,14</td><td> 0,14</td>

<td>Thickness of spacer sheet, T_S (Mm)</td><td> 0,14</td><td> 0,12</td><td> 0,12</td>

<td>Activated alumina in a protective layer (total, g)</td><td> 30,7</td><td> 18,4</td><td> 12,8</td>

<td>SA / V ratio (m²/ m³)</td><td> 6766</td><td> 7452</td><td> 7569</td>

<td>Share of surface voids in the drying sheet, (% vol.)</td><td> 20%</td><td> 39%</td><td> 50%</td>

<td>Effective thickness of the drying sheet (mm)</td><td> 0,13</td><td> 0,089</td><td> 0,074</td>

<td>Effective height of the channel (mm)</td><td> 0,17</td><td> 0,18</td><td> 0,19</td>

<td>Channel share</td><td> 57%</td><td> 67%</td><td> 72%</td>

[0048] Each experimental bed operated under simulated RCPSA conditions at 20 cpm, 40 ° C, a humidity level of 40,000 ppm on feed, and a raw material to product gas ratio of about 7. Figure 3 shows the water concentration measured as a function of distance measured from the side power supply for the protective layers according to the invention tested in the Examples. These experimental protective layers have channel proportions generally greater than 50% and have surprisingly good performance).

[0049] Further research was conducted on sample 1 protective layer to illustrate the effect of different raw material / product ratio (F / P from 3.4 to 6.8) on performance. Figure 4 shows the water concentration as a function of distance from the supply side as a function of the ratio of raw material to product under otherwise similar operating conditions. As seen in FIG. 4, with a raw material to product ratio of 3.4, to the zeolite adsorption layer

An unacceptable level of water vapor (31 ppmv) was getting in.

[0050] Further testing was carried out on a sample of protective layer 2 to illustrate the effect of varying RCPSA cycle speeds (from 10 to 30 cpm) on performance. Figure 5 shows the water concentration as a function of the feed side distance as a function of cycle speed under otherwise similar operating conditions (F / P ratio was always around 7.1). On the data of FIG. 5 there is a slight influence on the effectiveness of the operation_for the tested cycle speeds. A shift of the water front site by about 0.64 cm (or about 10% of the length of the protective layer) was observed when the cycle speed was changed from 10 to cpm.

[0051] The present invention has been described with reference to examples of preferred embodiments. It will be apparent to a person skilled in the art that changes and modifications can be made without departing from the present invention as defined in claims 1 to 14. For example, the invention can be used in protective layers for RCPSA equipment applications other than water.

Air Products and Chemicals, Inc.

Proxy:

76P35206PL00

EP 2 125 162 B1