EP0007956A1

EP0007956A1 - Adsorbent regeneration and gas separation utilizing dielectric heating

Info

Publication number: EP0007956A1
Application number: EP19780900252
Authority: EP
Inventors: Eugene Julius Mezey; Salvatore Terrence Dinovo
Original assignee: Battelle Development Corp
Current assignee: Battelle Development Corp
Priority date: 1977-10-31
Filing date: 1979-05-21
Publication date: 1980-02-20
Also published as: GB2036587A; JPS54500045A; WO1979000251A1

Abstract

Le chauffage dielectrique (rayonnement micro ondes ou haute frequence) d'adsorbants satures afin de separer les materiaux adsorbes donne une regeneration plus rapide, plus efficace et plus sure que le chauffage conventionnel. La separation d'un melange gazeux en deux courants concentres de ses composants est possible par l'adsorption d'une fraction par un adsorbant selectif suivie de la separation de cette fraction par chauffage dielectrique avec peu ou sans gaz de purge. Des derives utiles de ce procede de separation sont ainsi recuperes economiquement.Dielectric heating (microwave or high frequency radiation) of saturated adsorbents in order to separate the adsorbed materials gives faster, more efficient and safer regeneration than conventional heating. The separation of a gas mixture into two concentrated streams of its components is possible by the adsorption of a fraction by a selective adsorbent followed by the separation of this fraction by dielectric heating with little or no purge gas. Useful derivatives of this separation process are thus economically recovered.

Description

DESC ;

ADSORBENT REGENERATION AND GAS SEPARATION UTILIZING DIELECTRIC HEATING

Technical Field

Adsorbents are solid materials used to selectively remove contaminants or components from fluid process streams. Activated charcoal, zinc oxide, activated alumina, and molecular sieves are typical examples of the variety of known adsorbents. The adsorbent may be self-supporting or may be fixed to a substrate. The process stream is caused to contact the adsorbent for the required time period for the desired removal of the contaminant or component. The adsorbent may also be used to effect a separation of a gas stream into two components if, after removal of one component, the flow of the process stream is stopped long enough to desorb and recover the adsorbed component. Even if the adsorbed material is not- recovered for further use (e.g., it may merely be a contaminant and not a useful by-product of the separation process) , it still must be removed period¬ ically after it saturates the adsorbent or the adsorbent will cease its function. One method of regeneration is by heating the adsorbent (and substrate) to a temperature sufficient to desorb the adsorbate. Typically, this heating is accomplished with a flow of hot gas since it not only heats the adsorbent, but also purges^' the adsorbate as it is desorbed.

Effective adsorbents create a problem in regeneration since, generally, the more effective the adsorbent, the more difficult it is to remove the adsorbate. Long regeneration times and large purge gas volumes are therefore required for a moderate temperature purge gas to desorb the adsorbate. This is due in part to the poor thermal transport inherent

IfUREΛ^

OMPI in this method and with these adsorbent and substrate materials. Moreover, heat requirements are high since not only the adsorbent (which can be very large in itself) but also the adsorbent support, the adsorbent column associated conduits and large quantities of purge gas often must be heated. The long regeneration times are poor on efficiency and large volumes of purge gas result in effluents which are very dilute in the adsorbed (desorbed) component. Usefulness of the adsorbed component as a by-product is thereby reduced since more processing would be required to recover the component from the dilute gas.

Higher temperature purge gas could be used for more rapid desorption but this may also degrade the adsorbent capacity and shorten the life of the support. Large volumes of purge gas and dilute off- gas would still result.

Background Art

The present invention provides dielectric heating which eliminates the problems and provides benefits unobtainable with hot purge gas heating and with conventional heating, and with an energy savings. In the prior art, dielectric heating (in particular, microwave heating) has been suggested for many heating processes including drying, vulcanization of rubber, detoxification of dangerous substances and polymer¬ ization of fiberglass laminates. For example, U.S. Patent 3,771,234 suggests the use of microwave radiation for removing volatile polar vehicles from non-polar materials (dielectric loss factors of about 0.0001 to 0.1), specifically drying of synthetic polymers.

Coventional heating plus a purge gas flow is also used (see U.S. Patent 4,011,306), but as a result of poor heat transfer this can-require long regener¬ ation times or high temperatures. Microwave heating and a purge gas are apparently disclosed in Japanese Kokai 76/43,394, 76/43,395 and 76/145,491 (Chemical Abstracts; Nos. 8522166187w, 8522166188x and 8702008041f) .

As is well known in the art, heating with microwaves provides high frequency oscillatory movement of the molecules within the material by the combined interaction of the electric and magnetic fields associated with absorbed electromagnetic energy. The rapid temperature increase of the material is caused by this molecular friction.

Disclosure of the Invention It is an object of the invention to provide a process for the rapid thermal regeneration of adsorbents so that throughput of process streams can be increased or the amount of adsorbent can be reduced. It is also an object to provide rapid regeneration of adsorbents with little degradation of adsorbent or support.

It is further an object of the invention to provide a process of separating a fluid process stream into two concentrated component streams whereby both streams can be economically treated and used.

It is an object of the invention to provide the above benefits while using less energy than with conventional forms of heating.

In accordance with the objectives, the invention is a process for the rapid, thermal regeneration of adsorbents wherein the saturated adsorbent is heated dielectrically to desorb the adsorbate, after which a small quantity of purge gas stream may flush the adsorbate from the area of the adsorbent. Heating dielectrically, preferably with microwave radiation, effects internal heating of the adsorbent thereby minimizing the effects of poor heat transfer of the absorbent and substrate which delays the effectiveness of a conventional form of heating. Consequently, this is not a mere substitution of one heat source for another, but the provision for a different form of heating which may lower energy usage and, at the same time, benefit the process in allowing selective, rapid heating of the absorbent, rapid desorption of adsorbed components and gas separation into two concentrated streams.

Further, the invention can be used to desorb materials which are physically adsorbed on the adsorbent and also to desorb chemisorbed materials which may have further undergone chemical reaction with the adsorbent. An example of the latter "absorption" is the reaction of S0₂ n flue gases with a modified ZnO absorbent. The effects of 'dielectric heating are also utilized in separating a fluid mixture into two or more concentrated component streams. Each repetition of the inventive process may result in separation into two components. Successive treatments over different adsorbents may eventually reduce a fluid mixture to many of its individual components.

In the separation method, the fluid mixture is passed over an adsorbent which preferentially adsorbs at least one component thereof. The component is adsorbed as the mixture passes and the fluid stream remaining after adsorption is collected. This effluent stream is more concentrated than the input fluid mixture in components other than those removed by the adsorption. Thereafter, or after the adsorbent is saturated with the adsorbate, the fluid mixture flow over the absorbent is temporarily halted and the absorbent is heated dielectrically to a temperature at which the adsorbed component desorbs (or to a temperature at which the "adsorbent" chemical reaction reverses and the fluid mixture component reactant is desorbed) . A small quantity of purge gas may then be passed over the adsorbent to remove the desorbed component to a collection area. This purge gas stream is now concentrated with the desorbed component. After purging the adsorbent, the original fluid mixture may again be passed over the adsorbent and the cycle repeated.

If the adsorbed (desorbed) component is truly a waste contaminant of the original fluid mixture it can be collected for disposal. But if the component has some value, it can be more easily upgraded or used from this concentrated form than it could in prior processes.

Use of a purge gas is optional since the process is effective in either case. Use without a purge gas is preferred, however, since it generally increases throughput. Without a purge gas, heating causes the desorption and an increase in pressure in the adsorber chamber. This pressure increase causes most of the desorbed component to exit the chamber to be collected. The remaining desorbed component is merely readsorbed when the heating is eliminated and the fluid mixture is again allowed to flow through the adsorber. The readsorption lowers efficiency per cycle but many more cycles can be run without the purge gas since purging requires appropriate valving and time consuming sequencing with the fluid mixture flow. In total, the speed of the regeneration without purging outweighs the slight efficiency advantage enjoyed by purging. Of course, purge gas also dilutes the desorbed component stream.

Separations that are advantageously effected by the method include air separation (O2 production) , CO2 and H2S removal from natural gas, light hydrocarbon separation (replaces distillation) , and SO2 or other contaminants from flue gas. Many other separations are possible in practicing the invention.

For purposes of this disclosure the following word meanings will be used. The term microwave frequency radiation is defined as electromagnetic energy in the region of the spectrum having wavelengths of about 1 meter to 1 millimeter and frequencies of about 300 MHz to 300 GHz. However, this energy is conventionally operated in the region of 915, 2450, 5800, or 22,125 MHz in the industrial, scientific and medical (ISM) band as assigned by the Federal Communications Commission. The term radio frequency radiation will mean the same type of energy as microwave but in the lower range of frequencies of about 300 kHz to 300 MHz. The conventional ISM frequencies are 13.56, 27.120, and 40.68 MHz.

Adsorbent shall mean a solid material which has the property of being able to physically or chemically immobilize or adsorb molecules whether or not the adsorbent further enters into a chemical reaction with the adsorbate, as in the "adsorption" of SO2 by ZnO. For purposes of this disclosure the adsorbents shall have dielectric loss factors in excess of about 0.1 at operating frequency and be capable of being heated dielectrically. Adsorbents which are selective in any particular separation are well known in the art. For example, see U.S. Patent 4,011,306 which discusses adsorption of oxygen from

OMPI air with the cobalt chelate, bis (3-fluorosalicylal) ethylenediimine cobalt (II) .

Best Mode for Carrying Out the Invention

The best mode for carrying out the invention is demonstrated by the following examples. However, the benefits to be derived on a commercial scale are much greater since the size of adsorbers is much greater. For example, the laboratory adsorber is 2.5 inches in diameter and 4 inches deep whereas commercial adsorbers can be, for example 3-12 feet in diameter and 3-20 feet deep. Thermal problems and delays with conductive and connective methods of regeneration increase substantially at this scale. Further, the time necessary to heat (and cool) the adsorber is critical to the adsorbent inventory (or the amount of adsorbent needed) . Slow heating and cooling neces¬ sitates the high capital costs of higher adsorber inventory.

Examples of the Preferred Embodiments

Example 1

To simulate removal of C0₂ and H-S from natural gas, a fluid mixture of C0₂ and H-S in nitrogen was investigated. A source of the fluid mixture analyzing 14.78% C0₂ and 6.79% H₂S (balance nitrogen) was connected to the inlet of a 2 1/2 inch (6.35 cm) diameter cylindrical adsorber reactor chamber and a gas flow bellows meter was connected to the outlet thereof. The adsorber was contained within a microwave oven capable of delivering 200- 2500 watts of power at 2450 MHz. A nitrogen purge gas source was also connected to the inlet of the adsorber chamber with appropriate valving to admit only one source at a time through the inlet. Thermocouples were removably installed in the adsorber chamber at spaced locations to monitor temperature when microwaves were temporarily turned off. About 800 grams of molecular sieve 4A was placed in the adsorber column to a depth of about 12 inches (30.5 cm) and the power was turned on to run the column through a heating/cooling cycle to outgas the system. After outgassing the weight of the molecular sieve 4A was determined by difference to be 702 grams.

The adsorbent was then loaded by flowing the gas mixture therethrough for 30 minutes at about 0.5 cubic feet per minute (0.24/sec). Total flow through the adsorber was determined from the bellows gas meter to be about 10.5 cubic feet (298 liters). During the loading, samples were taken at elapsed times of 5, 15, 25 and 30 minutes at the outlet of the gas meter for chemical analysis by gas chromatography. ^* After loading, the gas mixture source was turned off and the adsorber was heated at 1000 watts power (150 watts reflected) for 5 minutes. A sample of evolved gases was taken after 3 minutes of heating. The temperature reached 275°C within the adsorber column. After heating, the pure nitrogen gas source was started at about 0.5 ft³/min (0.24j_/sec) for about 10 minutes to purge the desorbed C0₂ and H₂S. Samples of this purge gas mixture were also taken for analysis after 0.5 minutes and 10 minutes of purge flow. Measurements of sample composition indicated that during loading the effluent was substantially pure nitrogen. During heating (desorption) , the concentration of C0₂ and H₂S in the nitrogen purge gas effluent totaled at least 95%. Example 2 To quantify results found in Example 1, a second experiment was conducted to measure the degree of concentration which can be achieved in the removal of C0₂ and H₂S from a nitrogen carrier gas.

The apparatus was similar to Example 1 using the cylindrical reactor, 1/16 inch pellets of molecular sieve 4A as adsorbent and a source of the fluid mixture analyzing 17.2% C0₂ and 5.5% H₂S, balance nitrogen. A quantity of the 4A sieve was placed in the adsorber column and outgassed. The weight thereafter was about 240 grams. The adsorbent was loaded by flowing the fluid mixture therethrough for 27 minutes at 2000 cc/min. Referring to Figure 1 it can be seen that break-out of C0₂ took place after about 5 minutes and break-out of H₂S after about 15 minutes. The bar graph shows the percentage of the components in the effluent stream at various times.

Microwave heating was commenced at t = 27 minutes for 8 minutes at 300 watts power. No purge gas was used during this period and for the next 3 minutes thereafter, but it can be seen from the cumulative graph that almost 80% of the adsorbed C0 and almost 60% of the adsorbed H₂S were desorbed by this time. The C0₂ and H₂S during this period made up more than 90% of the effluent gas. The C0₂ made up 36-64% of the effluent (versus 17.2% in the original fluid mixture) and H₂S made up 27-56% of the effluent (versus 5.5% in the original fluid mixture). Maximum temperature was about 230°C.

Figure 1 shows the percentages of the components in the effluent and the cumulative desorption over the period of desorption. Microwave heating was periodically started at 300 watts for 2 minutes at t = 45 minutes and 55 minutes and for 2.5 minutes at t = 65 minutes. A final heating at 125 watts was initiated at t = 76 minutes for 17 minutes. A nitrogen purge of 80 cc/min. was initiated at t = 38 minutes to eliminate the remaining desorbed components. This experiment shows that a substantial amount of the adsorbed component can be removed from the adsorber column without purge gas. Operating commercially, it might be advantageous to begin desorption immediately after break-out (after 5 to 10 minutes in the example) , and to end desorption and begin loading again after 80%, for example, of the components have been desorbed.

Example 3 An experiment to separate C0₂ and H_S from methane was conducted using the equipment of Example 1. In the experiment, the 4A molecular sieve was loaded until breakthrough of H₂S occurred as measured by gas chromatigraphic analysis. After breakthrough, feed gas was interrupted and microwave heating commenced. As the flow of desorbed gases diminished, nitrogen purge gas flow was initiated and followed by inter¬ mittent heating cycles to maintain the bed temperature at between about 200 and 260°C. Three load and desorption cycles were run to determine cyclic behavior.

In the first cycle, with a feed gas of 15% C0₂, 10% H₂S, 0.83% ₂, balance CH., it took 75 minutes at 2000 cc/min to obtain breakthrough of H₂S. Flow was stopped and heating at 300 watts was commenced for 8 minutes, followed by a 3 minute pause in heating and another 3 minutes of heating. At this time, a gas analysis showed the composition to be greater than 90% CO.,, H„S and COS. This accounted for almost 100% desorption of C0₂ and 30% desorption of the H₂S. On the second load cycle using a methane gas containing 16.2% C0₂, 7.2% H₂S and 0.9% N_?, it took only 40 minutes to get breakthrough of H-S. Heating at 300 watts for 11 minutes from this point resulted in removal of 68% of the C0₂ and 29% of the H₂S from the adsorbent. While heating periodically for a total of 46 minutes during the next 85 minutes at 150 watts to maintain the temperature at about 250°C, a total of 95% of the C0₂ and 50% of the H₂S were desorbed. In the third cycle the feed gas was methane containing 18.5% C0₂, 8.6% H₂S and 1.2% N,. Break¬ through of H₂S occurred in 25 minutes and microwave heating thereafter for 11 minutes at 300 watts resulted in 74% desorption of C0₂ and 20% desorption of the H₂S. Heating for 39 minutes of the next 80 minutes to maintain the temperature resulted in 99% of the CO- and 36% of the H₂S being removed from the adsorbent.

Example 4 The results of the percentage desorptions in Example 3 were compared with the energy exposure of the adsorbent (the energy exposure being the output power of the microwave generator times the exposure time) to determine the effectiveness of this heating. An experiment was then run with the same loading but with conventional resistance heating to effect the desorption. The energy exposure of the adsorbent during this type of heating was calculated and compared to the percent desorption of C0₂, H₂S and COS to determine its effectiveness. The results showed that the first cycle with microwave heating produced the most effective desorption over the range of 0 to 5000 watt-minutes. The total desorption increased slowly to 25% over 2000 watt-minutes, then accelerated to over 70% within the next 1000 watt-minutes, and finally slowed to a steady desorption rate, reaching almost 100% desorption at 5000 watt-minutes.

The second and third cycles using microwave heating resulted in steady desorption up to 60% desorption after 5000 watt-minutes.

The resistance heating was provided by a 288 watt heat tape and produced desorption which increased steadily to only 20% after 5000 watt-minutes. It might be expected that desorption would begin to increase at a much faster rate if longer times were employed, however, time is a significant factor in throughput and extended times are not desirable on a commercial scale. Further, the temperature of the adsorbent near the wall was much hotter during the resistance heating, raising the problem of damage to the adsorbent if excessive power is employed in resistance heating. Moreover, the hotter the adsorber, the longer the cooling time which would be necessary to begin another cycle, again adversely affecting throughput.

Example 5 An experiment was conducted to show the ability to recover SO- sorbate from a sorbent. An effective 0 sorbent was prepared by reacting ZnO with S0₂ and H₂0 to form ZnS0₃ ^• 2 1/2 H-0 and then driving off the S0₂ and H₂0 by heating. The ZnO is an effective sorbent for S0₂ when H₂0 is present, and after the above pretreatment, is much more effective in sorbing S0₂ than is ZnO without the treatment. The increase in efficiency is probably due to a surface area increase.

82 grams of the pretreated zinc oxide was slurried with 360 ml of water and placed in an adsorber chamber such as in Example 1. Pure SO- gas was then bubbled through the slurry for about 30 minutes at about 2 grams/minute. The adsorbent was then dried overnight at about 90-98°C and weighed 5138.8 grams for a weight gain of 56.8 grams S0₂. This saturated sorbent was then heated dielectrically as in Example 1 for 4 1/2 minutes at 1000 watts power (350 reducing to 100 watts reflected) . The offgas was analyzed and found to be essentially 0 pure S0₂. This shows that a sorbent can be rapidly regenerated and the offgas stream can be highly concentrated in the desorbed gas.

Example 6 The apparatus of Example 1 was modified by 5 placing the bellows meter before the adsorption column and a wet test meter at the output side of the column. This provided more accurate loading and desorption flows from the separation of this example. A 13X molecular sieve was employed in the column to adsorb 0 C-, C- and C . hydrocarbons from the impure methane feed. The feed gas analyzed 81.6% CH., 10.5% C₂Hg, 5.3% C₃H- and 2.6% C^O.

The feed gas was started at t = 0 and continued at 2450 cc/min. until t = 86 minutes when 5 about 5% propane breakthrough occurred. During the loading, ethane was not totally adsorbed. It made up about 15% of the output with the balance being methane. Selection of a better adsorbent or a second adsorption would separate the ethane and leave essentially pure 0 methane.

At t = 86 minutes, microwave heating was commenced at 300 watts for 7.5 minutes, followed by a 1 minute pause, 2 minutes of heating, a 2 minute pause and 2 more minutes of heating. During this time 5075 cc of the adsorbed ethane and propane gases were desorbed. The composition of the offgas was between 85 and 97% ethane plus propane with the balance methane. Continued periodic heating with microwaves using a nitrogen purge produced an offgas mixture comprising 30-40% propane and butane with the balance nitrogen.

A second loading at 2450 cc/min. of the adsorbent resulted in a 5% propane breakthrough after 63 minutes. Eleven minutes of microwave heating at 300 watts followed by a 1 minute pause and 3 minutes additional heating desorbed 5800cc of the C--C. gases and resulted in a gas which was 95% C--C.. Inter¬ mittent heating over an additional 90 minutes removed a cumulative total of about ll,400cc of the adsorbed gases.

Example 7 The experiment of Example 6 was continued with a third loading/desorption cycle but using resistance heating in place of dielectric heating to desorb the adsorbed gases. A 288 watt resistance heating tape was wrapped around the adsorber column and heating equivalent to the microwave energy input to the adsorber was used. A feed gas similar to the gas of Example 6 but analyzing 80.5% CH₄, 10.1% C₂Hg, 3.0% C₃H_g, 3.8%

C4.H1, 0 and 0.6% N2-. was used. A loading rate of 2450 cc/min. was used for 71 minutes. Resistance heating was initiated for 38 minutes during which the gas composition approached 95% C₂C₄- However, during the first 12 minutes, only about 3300 cc of gases were desorbed. This is due at least in part to the poor thermal transport and conductivity inherent in this type of heating. Example 8 Apparatus as used in Example 6 was used to separate 0₂ and N₂ from air. A 5A molecular sieve was used to adsorb the nitrogen. The adsorbent column was first saturated with 98% 0₂ and then air was passed therethrough and the oxygen rich offgas collected until the exit stream contained only 30% 0₂ (70% nitrogen) . At this time the adsorbent was heated with microwaves at 500 watts power. During the first 2.5 minutes of heating a large percentage of the nitrogen was desorbed and the offgas increased from 70% ₂ to about 82% N-. Thereafter, less gas was desorbed but the nitrogen content increased to 94% after 10 minutes of heating. The final temperature of the column was about 125°C.

The data show that nitrogen and oxygen can be separated using the invention. The results are believed to be conservative in that a lag in measuring the composition of the offgas may be present because of the physical separation of the flow meter and the downstream gas chromatograph. Therefore the large volume of gas desorbed in the first 2.5 minutes may in fact be much higher in nitrogen than recorded.

Claims

CLAIMS ;

1. A process for the rapid, efficient regeneration of a saturated solid adsorbent without thermal degradation of the adsorbent which comprises heating the adsorbent dielectrically to a temperature sufficient to desorb the adsorbate.

2. The process of claim 1 wherein the adsorption includes a chemical reaction of the adsorbate with the adsorbent which comprises heating the adsorbent to a temperature sufficient to drive the chemical reaction in reverse and desorb the adsorbate.

3. The process of claim 1 wherein the adsorbent is heated by means of microwave radiation.

4. The process of claim 3 wherein a concentrated stream of the desorbed adsorbate is recovered which comprises, after heating the adsorbent, passing a volume of purge gas sufficient to carry away the desorbed adsorbate across the adsorbent and collecting the effluent purge gas and desorbed adsorbate.

5. The process of claim 3 wherein the adsorbent comprises zinc oxide and the adsorbate comprises sulfur dioxide.

6. The process of claim 3 wherein the adsorbent comprises molecular sieve 4A and the adsorbate comprises hydrogen sulfide and carbon dioxide.

7. A process for rapidly separating a fluid mixture into two concentrated streams by an adsorption- regeneration process which comprises:

(a) passing said fluid mixture in contact with an adsorbent material which preferentially adsorbs at least one chemical component of said mixture;

OMPI

: WIPCT (b) collecting a concentrated first fluid stream which remains from said fluid mixture after adsorption of said chemical component by said adsorbent; (c) temporarily halting the flow of said fluid mixture across said adsorbent;

(d) dielectrically heating said adsorbent material to a temperature at which said adsorbed chemical component is desorbed from said adsorbent material; and

(e) collecting a concentrated second fluid stream comprising said desorbed chemical component.

8. The process of claim 7 wherein the adsorption includes a chemical reaction of the adsorbate with the adsorbent which comprises heating said adsorbent to a temperature sufficient to drive said adsorption reaction in reverse and desorb said adsorbate.

9. The process of claim 7 wherein said adsorbent is heated by means of microwave radiation.

10. The process of claim 7 wherein said desorbed chemical component is purged from said adsorbent after desorption which comprises passing a quantity of purge gas across said adsorbent sufficient to remove said desired chemical component from contact with said adsorbent, and collecting said concentrated second fluid stream comprising said desorbed chemical component and said purge gas.