JPH0351449B2 - - Google Patents

Description

[Detailed description of the invention]

Desiccant dryers have been commercially available for many years and are widely used throughout the world. A common type consists of two desiccant beds, one of which is regenerated while the other is in the drying cycle. The gas to be dried is passed unidirectionally through one desiccant bed in the drying cycle and then after a predetermined period of time such that a hazardous condition is created such that the moisture level of the effluent gas cannot be maintained at the required low value. Note that when the desiccant becomes hygroscopic, the incoming gas is switched to the other bed, and the spent bed is heated and/or discharged, and/or the dry cleaning stream is normally applied to said bed in the opposite direction. It is regenerated by communicating. There are generally two types of desiccant dryers on the market today: heat regeneration, in which heat is applied at the end of the drying cycle to regenerate the spent desiccant; and unheated element dryers, in which the purified stream of drying gas is used without heating to regenerate the spent desiccant at the end of the drying cycle. The purge stream is typically the effluent gas from a bed in a drying cycle and is passed through the spent bed at low pressure and at rapid cycles to maintain the heat of adsorption to aid regeneration of the spent bed. It will be held in However, the use of low pressure for regeneration gas pressure due to the pipe pressure of the gas to be dried is not limited to non-heating element dryers; It was also used in regenerated desiccant dryers. Both types of dryers are operated in drying and regeneration cycles of constant, usually equal duration, the length of the cycle being determined by the available desiccant volume and the moisture content of the incoming air. The cycle time is fixed to be much less than allowed, so that the moisture content of the effluent gas always meets the requirements of the equipment. As the drying cycle progresses, the desiccant bed becomes increasingly saturated from the inlet end to the outlet end, and its capacity to adsorb moisture transported by the incoming gas becomes less and less. Moisture removal from the incoming gas depends on the flow rate of the gas, the rate of moisture adsorption, and the moisture content of the adsorbent, and the temperature and pressure of the gas in the bed. The rate of adsorption by the desiccant decreases as the desiccant becomes loaded. Since the moisture content of the incoming gas is hardly constant, the demand on the desiccant bed varies, sometimes very quickly.
And sometimes over a wide range. As a result, the constant drying cycle time must always be short to give a safe margin for moisture removal at the maximum moisture content of the incoming gas, so that the moisture adsorption capacity of the bed does not become too low. This means that a certain cycle time often ends up being too short in order to finish the cycle time. Of course, in the average cycle,
This means that the moisture holding capacity of the floor is not fully utilized. The lifetime of a desiccant heated for regeneration is highly dependent on the number of regenerations. Commercial experience has shown that desiccant beds are effective for a few regenerations, but not for more. It is clear then that if the moisture holding capacity is not effectively utilized in each drying cycle, the useful life of the bed will be unnecessarily shortened. Additionally, in both regenerative and unheated element types, the inability to fully utilize effective bed capacity during each drying cycle means that extreme but temporary moisture content in the incoming gas occurs during a given drying cycle. This means that the volume of the desiccant bed must be larger than necessary in order to possess the necessary capacity to adsorb the levels. The purge gas is wasted considerably in each cycle because the moisture absorption capacity is not fully utilized. Purge gas is typically removed from the effluent gas for regeneration of the spent bed, thus reducing the amount of effluent gas. Each time a bed is transferred from a drying cycle to a regeneration cycle, an amount of purge gas equal to the open volume of the bed vessel is given up and is lost. Short cycles result in more purge gas loss than longer cycles. Such losses are particularly severe in unheated element dryers where more frequent cycling is required.
In practice, the choice between a heated regeneration type and a non-heated element dryer is often determined by the frequency of cycles required. U.S. Patent No. 2944627 (Skarstrom,
1960, 7, 12) describes a non-heating element dryer, which is several years earlier than U.S. Pat. No. 2,800,197 (Wynkoop, 1957, 7, 23).
Specification and British Patent Nos. 633137 and 677150
This is an improved version of the device described in the patent specification. No. 2,944,627 discloses that by providing a very fast cycle between adsorption and desorption in each region, the desorption cycle can effectively utilize the heat of adsorption for the regeneration of spent desiccant. Are listed. Therefore, in the adsorption cycle a time of not more than 2-3 minutes, and preferably less than 1 minute, and very preferably
It has been shown to use a time of 20 seconds or less.
This cycle time is, of course,
2800197, which is of the order of 30 minutes or more, or shorter than the 5 to 30 minute cycle times described in GB 633137. GB 677150 indicates that the adsorption and desorption cycles do not necessarily have to be identical. A disadvantage of the device shown in U.S. Pat. No. 2,944,627 is that a significant amount of purge gas is lost in each cycle; No.
A cycle time of, for example, 10 seconds is quite large compared to 30 minutes or more in US Pat. No. 2,800,197. Of course, little of the capacity of the desiccant bed is utilized during this short cycle, but if no heat is applied to regenerate the desiccant, the moisture content of the adsorbent does not exceed a certain minimum during the adsorption cycle, or During the regeneration cycle, the adsorbent cannot be effectively regenerated. The dryer is equipped with a
A humidity detector is installed in the outflow pipe. However, due to their slow response and insensitivity to low dew points, such devices allow the moisture front to leave the floor by the time the detector detects moisture in the effluent gas. Therefore, if a low dew point effluent gas or relative humidity is desired, it cannot be utilized to determine the dryer cycle time. U.S. Patent No. 3,448,561 (Seibert, Pellend,
1969, 6, 10) The specification states that regeneration is performed only when the moisture load on the desiccant bed necessitates regeneration, thereby making effective use of the moisture absorption capacity of the desiccant bed. A gas drying method and apparatus is shown that achieves optimum efficiency at times. In each adsorption cycle, the adsorbent is utilized to its maximum wetting capacity under possible regeneration conditions, with or without applied heat, and with or without vacuum. This detects the advancement of a moisture front in the bed based on the moisture content of the gas to be dried, and when the front reaches a predetermined position in the bed,
This is possible by stopping the drying cycle before leaving the floor. This involves installing a device in the desiccant bed to detect the moisture content of the gas to be dried, and in response to the moisture content, stops the drying cycle when the moisture content of the gas to be dried reaches a predetermined value. This can be done automatically by providing a device. The advancement of a moisture front through a desiccant bed as it gradually adsorbs moisture is a well-known phenomenon in desiccant drying technology and is described in many patent specifications, such as U.S. Pat. No. 2,944,627. It is discussed in the book. During most of the drying cycle, the adsorbent effectively adsorbs moisture from the gas passing through it. However, as the adsorption capacity of the desiccant approaches zero, the moisture content of the gas passing through it increases, sometimes very rapidly. If the moisture content, dew point or relative humidity of a gas is measured and plotted against time, an increase in moisture content will be indicated, and if the increase is rapid, a change in the slope of the graph will indicate can be observed. In either case, as the moisture content of the outflow gas increases, it exceeds a predetermined minimum value and approaches the moisture content of the inflow gas. The portion of this curve above the selected minimum moisture content effectively represents a moisture front;
This may be an S-shape or other shape in the case of a slope change, and if this is observed within the length of the bed, from the inlet end of the bed to the outlet end as the adsorption cycle progresses. I understand that it will progress. The objective is to end the cycle before the moisture front or slope change of the curve reaches the end of the bed, which then rises rapidly and prevents the delivery of effluent gases with undesirable moisture content. Because it will disappear. In U.S. Pat. No. 3,448,561, this is done in a device sufficiently spaced from the outlet end in the bed so that the drying cycle can be completed before the moisture front reaches the outlet end. This is prevented by detecting the advance of the front. In this invention, a method is provided for reducing the concentration of water vapor in a mixture of water vapor and a second gas below a critical maximum concentration, which method includes:
The mixture is flowed in contact with a bed of adsorbent having a preferential affinity for water vapor from one end to the other to cause the bed to adsorb water vapor and to maintain a concentration below the maximum value. forming an effluent gas having and in said bed during the continuation of the adsorption,
As the adsorption capacity of the adsorbent decreases, the concentration of water vapor in the second gas increases, forming a water vapor concentration gradient that gradually decreases from one end to the other end of the bed. in defining an advancing concentration front and sensing the advancement of the moisture front within the bed by determining a change in water vapor content of the adsorbent as a function of capacitance of a condenser, the adsorbent comprising: dielectric, and the capacitor is located at a sufficient distance from the edge of the bed to prevent effluent gas having a critical maximum water vapor concentration from escaping the bed, such that the effluent gas is and before the water vapor concentration in the second gas exceeds the maximum concentration, causing the gas mixture to cease flowing in contact with the bed, and before exiting from the bed. The method comprises the step of controlling the desorption time of the desorption cycle for desorbing the water vapor based on the output signal of the condenser that detects the concentration front. Further, the present invention provides an apparatus for reducing the water vapor concentration of a mixture of water vapor and a second gas to a limit maximum concentration or less, the apparatus comprising: a container; the apparatus is provided in the container; a chamber for a bed of adsorbent having preferential affinity for the chamber, a conduit for delivering the inflow gas to the inlet end of said chamber, and a conduit for delivering the outflow gas from the outflow end of said bed; A capacitor comprising two conductors having a surface area, the conductors being spaced apart a sufficient distance from each other to define a space containing an adsorbent as a dielectric, and the conductors having a surface area of A capacitor in which a change is detected, for example as a function of the dielectric constant, whereby the capacitance is changed, a change in said capacitance when a selected capacitance is reached or exceeded, indicating a selected water vapor content. a device for generating a signal in response to a signal, a device for blocking the flow of the incoming gas in response to the signal, and a device for controlling the desorption time of the desorption cycle for desorbing the water vapor based on an output signal of a condenser that detects a concentration front. It consists of a device that Thus, in the method of the invention, the concentration of the first gas in the mixture with the second gas can be reduced below the critical maximum concentration in the second gas, the step for which A bed of adsorbent having a preferential affinity for a first gas is brought into contact with and flowed from one end to the other, adsorbing a first gas onto said bed, and as adsorption continues, a first gas flows into the bed from one end to the other. forming an effluent gas having a first gas concentration gradient to the other end; and a first gas concentration gradient in the second gas.
defining a concentration gradient that progressively progresses from one end of the bed to the other as the adsorption capacity of the adsorbent decreases due to an increase in the concentration of gas; sensing as a function of the capacitance of the capacitor with the adsorbent as dielectric, and that an effluent gas having a limiting maximum first gas content is in contact with said bed before it exits from said bed. stopping the flow of the mixed gas and controlling the desorption time of the desorption cycle for desorbing the first gas from the bed based on the output signal of a condenser sensing the concentration front. This invention applies to gas adsorption fractionators where part or all of the adsorbent bed is heated for regeneration, to equipment where no heat is applied for regeneration, to equipment where regeneration takes place under reduced pressure, and to equipment where purified gas It can be applied to devices that utilize flow, and to devices that combine one or more of these. Another feature of the invention is that the regeneration time need not be, and in most cases is not, the same as the duration of the drying cycle, so the bed being regenerated can be closed when it is complete. , heating, purification, evacuation, or any regeneration equipment may be deactivated. The remaining cycle time can be used, for example, to cool the regeneration bed, and when inlet gas flow to said bed is resumed, said cooled temperature is effective and convenient for adsorption. The drying apparatus of the invention comprises as an essential element an adsorbent bed which is regenerated periodically and preferably by counterflow, and one or more condensers are provided in the bed to Moisture content is detected at a sufficient distance from the outlet end so that the cycle can be terminated before the effluent gas in which the adsorbed gas exceeds a critical maximum concentration is discharged from the bed. I have to. Additionally, and/or if one wishes to observe bed regeneration as a function of capacitance, one or more condensers may be connected within the bed when desorption of the adsorbed gas reaches a selected low level or is complete. , is placed in a position where playback can be ended. Optionally, the device can be provided with a device for applying heat during said regeneration. This type of device can be arranged to extend throughout the bed, or only in areas of the adsorbent bed that have a high moisture content, on the order of 20% or more of the moisture-bearing capacity, at the end of the drying cycle. That is, during the drying or adsorption cycle, it may be limited to only those areas that are first contacted by the incoming fluid. In this case, the remainder of the adsorbent bed is not heated during regeneration, so that no heating device is provided at that location. The unheated portion of the bed volume can be as large as desired. Usually 1/ of floor volume
4 to 3/4, and preferably 1/3 to 2/3, portions are heated. Effectively, the unheated part of said bed constitutes a preliminary bed, which in normal drying cycles is seldom needed, and in each case the adsorbent is present in a relatively small part, e.g. Not only does it adsorb less than 20% of its moisture capacity, it also prevents moisture from being adsorbed insufficiently in floor areas with heating devices, preventing the delivery of undesired high moisture content effluent gases. established for the purpose of Since the moisture absorption capacity of the reserve section of the bed is hardly utilized, the preadsorbent is regenerated by the purge stream and the moisture carried forward from this section by the purge gas is After passing through the heating section of the bed, it is effectively removed from the bed. Although the apparatus of this invention can be comprised of a single desiccant bed, the preferred apparatus utilizes a pair of desiccant beds disposed within a suitable container and receiving the incoming gas to be dried. and connected to a conduit for the delivery of dry effluent gas. The drying apparatus may further include check or throttle valves for the purpose of reducing pressure during regeneration, and multichannel valves for circulating an incoming gas stream to the interbed and receiving an outgoing gas stream therefrom. Additionally, a volume or throttle valve may be included to divert a portion of the dry effluent gas as a purge stream in counterflow conditions to the bed to be regenerated. The capacitor can sense moisture content at any location within the bed as a function of capacitance. In some types of dryers, such as unheated element dryers that operate at very low total moisture content levels in the desiccant bed per drying cycle, the condenser is placed close to the inlet of the bed, i.e. It is desirable to place it between 1/5 of the length of the floor and half of the floor from the inlet. Typically, flow rates up to 45.8 m/min (150 ft/min) are recommended for optimal results.
The condenser is located between 2/3 and 1/5 of the length of the bed from the outflow end of the bed to prevent undesirable high moisture content of effluent gas from escaping. be done. However, if the condenser reaction is sufficiently fast or the possible moisture content of the effluent gas is of high value, the condenser can be placed adjacent to the effluent end of the bed. The location of the condenser in the bed is determined in part by the flow rate of the effluent gas through the bed, where the condenser is able to react to the moisture content of the effluent gas that exceeds the critical maximum moisture content before it exits the bed. I have to. Generally, at higher flow rates, the condenser is placed further away from the outlet end of the bed to prevent effluent gas above the critical maximum moisture content from leaving the bed and entering the operating equipment. must be detected fast enough to prevent it. The type of condenser and its location in the bed are selected to detect the moisture content level of the gas to be dried so that a signal can be given before the effluent gas exceeds a critical maximum moisture content and leaves the bed. be done. The safety margin (coefficient) required for a specific device can be determined by obtaining moisture content data for the specific device being used and plotting it.
Easily determined empirically. This can be determined without reference to this invention. Any capacitor of conventional design can be utilized. In its simplest form, it is a pair of mutually parallel flat metal sheets spaced apart by a variable or constant distance. Three or more flat parallel metal sheets can be utilized by connecting alternating sheets to form a stack of interleaved sheets. Several pairs of sheets can be attached to one level of the adsorbent bed, and the pairs are electrically connected in parallel. The cylindrical type has a tube-shaped conductor, and has a wire body,
A rod or tube-like conductor is mounted concentrically or eccentrically within it. The entire shell of the vessel containing the adsorbent bed has the function of one conductor, the other conductor may be a heating element or other device embedded in the desiccant. The conductors utilized for the plates can be of any shape as long as they are electrically conductive, such as wires, rods, sheets, strips, mesh, sintered metal, etc. The sheet conductors or cylindrical conductors need not be parallel or concentric, respectively, provided they are electrically insulated from each other. The conductor can be fixed or adjustable to allow adjustment of the capacitor. The distance between the two conductors of the capacitor is determined by the volume required to uniformly fill said space with desiccant. The sensitivity of the capacitors decreases as the spacing increases, so they should be as close together as possible. The conductor can be bare metal or, if the desiccant exhibits an undesirably high conductivity, coated with an electrically insulating layer to insulate it from the desiccant. Insulating layers of this type include nonconductive plastics such as polyolefins, polyethylene, and polyisobutylene, synthetic and natural rubbers, epoxies, ceramics, and glasses. Since the adsorbent between the two conductors of the condenser is utilized as a typical sample of the condition of the adsorbent bed, it is important that that portion receive all or a uniform portion of the gas flow through the bed. To this end, the conductor should be placed in-line with the fluid and not in a blind alley on one side of the flow. The entire shell of the vessel containing the adsorbent bed has the function of one plate of a condenser, with coaxial rods, wires, tubes or cylinders extending over a portion of the internal length of said shell and conducting the other. If configured, the condenser senses the total amount of water adsorbed by the desiccant bed rather than sampling the moisture load at a particular location on the bed. The transmission of a change in capacitance of a capacitor into an electrical, mechanical, acoustic or visual signal is accomplished in several ways. Manufactured by National Semiconductor
The LM2917N integrated circuit is utilized as a frequency-to-voltage converter, but with appropriate external components and connections it can also function as a capacitance-to-voltage converter. A bridge circuit excited by a sine wave can also be used to generate a voltage proportional to capacitance. Optional amplifier, e.g. National Semiconductor Part No. LM741N
A sine wave can be emitted by a circuit incorporating the .
The probe (needle) can be used as a capacitor in a series or parallel capacitor-conductor circuit excited by a constant frequency, constant amplitude sine wave. By appropriate selection of the excitation frequency and inductor value, this type of circuit can be designed to resonate at any capacitance value over the range sensed by the capacitor. The output of this type of circuit changes significantly when the capacitance reaches the value required to produce resonance. Again, the excitation sine wave is
Generated using an LM741N or similar differential amplifier. Capacitance is used as a feedback element in operational amplifiers, either reversing or non-reversing amplifiers, and for amplifying constant frequency, constant amplitude sine waves. This type of circuit has a sinusoidal output, the amplitude of which is adjusted by changing the capacitance. One or more condensers can be attached to at least one, and preferably all, adsorbent beds of the gas fractionator. Two or more condensers located at selected locations in the adsorbent bed can sense the development of a moisture gradient or an adsorbed component gradient in the bed.
The condenser is mounted in the bed at a location where it usually adsorbs little or no water vapor or other adsorbed components, so as to provide an alarm when moisture or other adsorbed components appear in said portion of the bed. used for. Two capacitors placed at different levels in the bed can be used to compare capacitances, with a "reference" capacitor placed in a part of the bed that is either permanently moist or constantly dry; The "sensing" condenser is located in a section of the bed that cycles from a dry state to a wet state and back during normal operation. The capacitor monitors the loss of water adsorption capacity of the desiccant, reflected by a gradual decrease in the maximum capacitance, and is installed in a bed section that becomes saturated with adsorbed water at some stage of the operating cycle. , can be used to display the condition of the desiccant in the dryer (i.e. water adsorption capacity). A capacitor that determines the moisture content of the desiccant bed can, but need not be, placed in the adsorbent bed and has sufficient volume to effect a change in capacitance related to the moisture content of the main desiccant bed. It can be provided within a representative sample of the adsorbent bed and at separate locations remote from the bed. In the latter case, a portion of the treatment or purge gas is circulated through said section so as to produce a proportional change in the moisture content of the sample adsorbent bed, for example in terms of the dielectric constant. This type of condenser senses the state of the adsorbent in the sample bed relative to the state of the main adsorbent bed in a fluidized state;
Proper calibration provides a response signal before the moisture front of the effluent gas leaves the floor. The following provides examples of preferred locations in the adsorbent bed of the condenser to control adsorption and/or regeneration cycle times in heated and unheated dryers. In heated dryers, each bed has one condenser that is reached by the moisture front when the bed is exhausted. Can be installed in locations throughout the floor. When the moisture front passes through the condenser, the relatively large fluctuation in capacitance caused by the change in the amount of water adsorbed by the desiccant between the conductors of the condenser causes the bed to change from a flowing state to a non-flowing state, i.e. for regeneration. converted into a signal that starts the cycle. If two beds are provided, the regenerated bed is placed in a fluid state at the same time. The criterion for a bed being exhausted is the arrival of a moisture front at a location in the bed selected to maintain a predetermined effluent gas dew point. A separate condenser in the heated dryer bed is utilized to detect the completion of regeneration when the bed is in a non-flowing state for regeneration. Since how to optimize the cycling of a heatless element dryer is dictated by the constant cycle time required for this type of dryer, the condenser is used to control the duration of delivery of purge gas to the regeneration bed. used. The most direct and preferred method is to mount a condenser near the inlet end of each bed to prevent the moisture front from reaching that end of the bed while being chased towards that end by the purge gas during regeneration. It is to detect. When the bed is returned to a fluid state,
If the moisture front reaches a point where the dew point of the effluent gas can be maintained under maximum moisture load conditions, the purge gas is shut off. In another more complex and less desirable arrangement, the condenser is placed at a location in the floor through which the moisture front passes during typical operating conditions. In a drying cycle, the time that a moisture front passes through a condenser in the bed is accumulated and related to the rate at which moisture is adsorbed to the desiccant in the bed. During this time, for example, the microprocessor
It is used to calculate the purification flow rate required to regenerate the bed when it enters the regeneration section of the NEMA cycle. The fundamental advantage of the condenser is that it provides a direct indication of the moisture content of the bed versus other sensing and/or calculation methods for monitoring dryer performance and therefore This is related to the depth of the floor through which the moisture front passes and thus to the degree of floor usage. Even if the condenser is utilized to determine the position of the moisture front during drying or regeneration times, this information and the parameters of actual interest,
For example, there is a direct correlation with the dew point of the effluent gas. By utilizing quantitative changes in moisture content, sensed as dielectric constant, sensing and control circuits are designed to quantitatively measure desiccant loading, dew point, and other parameters. The response of the condenser to changes in the amount of water adsorbed on the desiccant between the condenser plates is instantaneous. The device that responds to changes in capacitance of a capacitor due to moisture content can be electrical, mechanical, chemical, or any combination thereof;
In response to a predetermined change in capacitance or a predetermined capacitance, an appropriate valve is signaled or controlled to terminate a drying cycle or a regeneration cycle and the moisture level of the effluent gas reaches a predetermined maximum value. When reached, it is preferable to allow inflow and outflow gas to be transferred from one tank to the other. The time required for the moisture content of the effluent gas to reach a predetermined level is directly correlated to the moisture absorption capacity and moisture content of the adsorbent. As the gas travels along the length of the desiccant bed, its moisture content gradually decreases according to the rate of moisture adsorption by the desiccant. Since the rate of moisture adsorption by the desiccant depends on the moisture absorption capacity, gas pressure, temperature and flow rate of the gas stream, for a given temperature and pressure of incoming gas, when the moisture load of the adsorbent reaches a given level, It will be clear that only then will the effluent gas reach a predetermined moisture content level. Since the condenser responds directly to the moisture loading of the adsorbent, in this invention the length of the drying cycle can be adjusted precisely according to the moisture content or loading of the adsorbent, thus without risk of breakthrough. , the moisture absorption capacity of the adsorbent can be effectively utilized in each drying cycle. As a result, the desiccant dryer of the present invention can be configured to operate at a predetermined moisture load of desiccant during each drying cycle. This means that if the moisture level of the incoming gas changes, the length of the drying cycle is automatically adjusted accordingly. As a result, the drying cycle does not end until it is needed and unnecessary regeneration of the desiccant is also eliminated. Therefore,
Since there is no need to incorporate any reserve capacity into the desiccant, and the drying cycle is dependent on the moisture holding capacity of the desiccant utilized, a smaller amount of desiccant than previously required may be sufficient. At the same time, the amount of purge gas lost during each cycle is also kept at an absolutely small amount. Effectively,
The desiccant dryer of this invention can automatically time its drying cycle according to the demands given by the moisture content of the incoming gas, so that the dryer of this invention can It is called a vessel. On the other hand, the regeneration cycle need not be, and preferably not automatically, the same time as the drying cycle. The drying cycle can be very long if desired, so even if the drying cycle continues, it can be timed to end when regeneration is complete. This also ensures that the energy available for heating the purge stream and the bed is not wasted unnecessarily. The regeneration cycle is connected to another condenser to terminate when the adsorbent has finished drying, or if this is done before the scheduled regeneration cycle is completed, or the condenser can be connected to the regeneration cycle. If so, the condenser or timer will not end the drying cycle until regeneration is complete. The same capacitor can also be used to control the regeneration cycle as described above. Prevents unnecessary abandonment of process gas in the adsorbent bed in unheated element dryers, prevents concomitant process gas losses when process gas flow rates are very low, and provides significant flow purification Without gas guarantee (the amount of gas lost during abandonment is very large compared to the amount of purge gas required for regeneration at very low process gas flow rates, this mode of operation is practically Hysteresis is introduced into the control mechanism to save money. A capacitive humidifier will terminate regeneration with a sensing capacitance slightly lower than that required to begin regeneration. Therefore, when a fully regenerated bed is switched to a flowing state with no process flow and no downward movement of the mass transfer region and no increase in the capacitance of the capacitive probe, the regeneration When entering the half cycle, no purge flow is required and there is no pressure drop. In contrast, in a bed that does not introduce hysteresis, the capacitance level at which regeneration ends is also interpreted as the (minimum) capacitance level required to start regeneration, so that when entering the regeneration half-cycle, The pressure will drop and it will clear for a very short time. The degree of hysteresis is determined by optimizing the overall efficiency. High hysteresis allows a substantial amount of moisture to accumulate in the bed before the purge flow is required, creating a pressure drop, thereby conserving the maximum amount of process gas that would normally be lost during pressure drop. . This means that for low flow rates, a large number of
No purifying flow is required for the NEMA cycle;
As a result, this means that the heat of adsorption may be lost. On the other hand, a relatively small hysteresis allows the heat of adsorption to be utilized by allowing a small amount of water to accumulate in the bed before a purge stream is required. This allows the dryer, at a given flow rate, a little time for the heat of adsorption to dissipate.
This means that more frequent purge flow is required. Its disadvantage is that a large amount of process gas is lost due to frequent pressure drops. These two factors are an optimal efficiency trade off.
This trade-off of different hysteresis for each dryer size is part of the control circuit design. The necessary hysteresis can be introduced in the electronic signal-conditioning circuit shown in FIG. Resistor H is the only element needed to introduce hysteresis. Resistance H
The value of determines the amount of hysteresis, ie, the difference between two capacitance levels at which the output of the circuit goes high or low. Hysteresis places two condensers at slightly different levels of the bed, using the upper condenser to signal the end of regeneration and the lower condenser to signal the downward flow dryer that it needs to start regeneration. It can be used by In this case, the amount of hysteresis can be adjusted by changing the vertical distance between the two probes. With this invention, the cycle is terminated before the moisture front, or change in the slope of the moisture content curve, reaches the end of the bed, after which it rises rapidly and prevents the delivery of undesired humid effluent gas. This is because it is almost impossible to prevent. This is prevented by the condenser by sensing the advancement of the moisture front in the bed at a location far enough from the outlet edge to allow the drying cycle to end before the moisture front reaches the outlet edge. Ru. The following comparative example shows the effect of two levels of capacitance controller hysteresis on process gas utilization efficiency:

[Table] Average purification time
The inflow conditions for the 150DHA utilized are almost identical: for the high test, the inflow rate is 125.8 scfm at a pressure of 4.7 Kg/cm ² (67.5 psig), and for the low test, the inflow rate is 4.76 Kg/cm 2 (67.5 psig). At a pressure of cm ² (68 psig) the inflow is
It is 124.5 scfm. Low hysteresis testing required purge flow on each cycle. High hysteresis testing saved a total of 66 abandonments. The amount of gas lost per abandonment is equal to approximately 11 seconds of purge flow utilized, resulting in high hysteresis operation saving process gas equal to 726 seconds of purge flow. Averaged over 201 half-cycles, this results in an average savings equal to about 3.6 seconds of purge flow per half-cycle. This reduces gas consumption equivalent to 186.4 seconds of purge flow per half cycle. Therefore, high hysteresis tests save on waste, but consume more gas to maintain outlet conditions than low hysteresis tests. This implementation is shown in FIG. Figure 1 shows the condenser starting from the outflow end of the floor.
Adsorbent bed determined as a function of bed capacitance by sensing at a series of X points 76.2 cm (30 in) and a series of Y points 61 cm (24 in) from the outlet end of the bed. The moisture content of 37.8°C (100〓) to 21.1°C (70〓) is plotted against the time the gas is dried.
shows a series of curves for drying a humid gas with a relative humidity of 80% at a temperature of . These curves indicate that the air line pressure is 6.44Kg/cm ²
(92psig), surface flow velocity 19.9mpm (65.1fpm),
A bed 60 with a length of 1.52 m (60 in) and a diameter of 31.4 cm (123/8 in) is coated with silica gel as a desiccant agent.
For a device similar to that in Example 1 utilizing 80.4 Kg (177 lb). But the data
This is typically obtained using any desiccant under any adsorption conditions. The principle of the method of the invention is to detect a change in the slope S of the moisture front before it reaches the outlet and to stop the cycle, i.e. in FIG.
Approximately 2.4 hours before the cycle time for the curve,
About 3.3 hours before the cycle time for the curve,
Approximately 4.5 hours before the cycle time for the curve,
This is approximately 6.3 hours before the cycle time for the curve. The curve in Figure 1 shows that this is the point X on the adsorbent bed.
and Y by terminating the cycle while the moisture content does not exceed 2%. This moisture content is detectable by a conventional capacitor. When the condenser is placed at points X and Y in each drying cycle as shown in FIG. 1, the curve shows the moisture content of the adsorbent bed when the condenser is activated. In all cases, the curve,,
and, the capacitor X is at points A, C,
E and G are activated, capacitor Y is activated at points B, D, F and H, all tilted.
This is before the changes in S ₂ , S ₃ , and S ₄ , and the moisture front is prevented from leaving the floor. The dryer shown in FIG. 3 consists of a pair of desiccant tanks 1 and 2. These tanks are arranged vertically. Each tank contains a desiccant bed 10, such as silica gel. The tanks 1, 2 are further provided with desiccant filling and discharging vents 9, 9 for filling and discharging the tanks with desiccant. Lines 3 and 4 connect the two tanks and introduce from line 6 the incoming gas with the moisture to be removed, and lines 7 and 8 introduce the dry incoming gas with the moisture to be removed after passing through the dryer. Check valves G and H for outflow gas.
It is adapted to be delivered to the outflow line 5 via. Inflow switch valves A, B, only one open at a time, direct the inflow gas flow into one of the inflow lines 3, 4 which introduce the inflow gas into the top of each tank. A desiccant screen support 11 of sintered stainless steel mesh is provided at the bottom of each tank to maintain the desiccant bed 10 within the tanks 1,2. The outflow lines 7, 8 from the bottoms of the tanks 1, 2 each lead to an outflow line 5. The pipe line 12 connects the pipe lines 7 and 8, and is provided with check valves E and F for controlling the purification flow for regeneration in one direction. Line 12 is connected to outlet line 5 by line 13 via a valve and orifice 14 . The check valves G and H allow flow in only one direction toward the outlet line 5 in the lines 7 and 8, and the check valves E and F allow flow from the outlet line 5 through the orifice 14 and the valve. A unidirectional flow is possible in the opposite direction to line 12 and further flowed in the opposite direction via the other of lines 7, 8 to the bottom of the tanks 1, 2 so that the desiccant 10
It is designed to purify and regenerate. The valve is utilized to control the delivery purge flow rate to the unflowed bed during regeneration. The conduit 13 is provided with a purified flow pressure gauge P. At a position above the support 11 inside each tank,
A pair of moisture sensitive capacitors 20, 21 are each arranged to measure (determine) the moisture content of the desiccant as a function of the capacitance of the capacitor, where the adsorbent is a dielectric and It's summery. A conduit 15 is pre-arranged in the conduits 3, 4 and discharge valves C, D are provided to control the flow rate from the tanks 1, 2 to the discharge conduit 16. During regeneration of the adsorbent beds 10 in tanks 1, 2, only one of the valves C, D is opened at a time. The capacitors 20, 21 are of plate type and are connected to a detector (not shown) set to respond to a capacitance corresponding to the moisture level to ensure that the moisture front does not leave the floor. . For example, if the maximum atmospheric exposure allowed for the effluent gas is -62°C (-80〓), the detector will
(−40〓) to −17.8°C (0〓) to respond to a capacitance corresponding to a dew point. The illustrated detector is of the LM2907N frequency-to-voltage conversion type, but any other type of detector may be utilized. Disposed at the inlet end of each bed 10 and extending substantially the entire length of the bed is a row of long heating elements 22, in this case ten heating elements. These heating elements are evenly spaced throughout the floor. However, depending on its heat capacity, more or fewer heating elements can be provided. The inlet end of the heating element is provided with an electrical connection 23, which connects the tanks 1, 2.
extending through the wall and connected to an electrical device such that the heating element is switched on when the floor is in a regeneration cycle and switched off at the end of a predetermined period of time sufficient for effective regeneration of the desiccant. , the predetermined time may be less than or equal to the drying cycle time, depending on the time required to activate the capacitors 20,21. Here, when tank 1 is in the drying cycle, tank 2 is in the regeneration cycle, and valves A and D are open and valves B and C are closed, the dryer operates as follows. Humid gas entering via line 6 at a line pressure of 1.75 to 24.5 Kg/cm ² (25 to 350 psig) is connected to valve A,
B is delivered from pipe 3 to tank 1, and the floor 10
It flows downward through the condenser 20 and is sent to the outlet through the conduit 7 to the valve G.
and is sent to the outflow pipe 5. Valve E and H
prevent flow in conduits 12 and 8, respectively. A portion of the effluent flow controlled by the purge valve passes through line 13 and through orifice 14 where the pressure is reduced to approximately atmospheric pressure by open discharge valve D and then by check valve F (valve E is connected to line 14). 13) through conduit 1
2 and in a regeneration cycle state.
, flows upwardly through the bed 10 towards the inlet, through valve D, through line 15 and is discharged via purified discharge line 16 to the atmosphere. The row of heating elements 22 in the regenerating tank 2 is energized and the drying bed is baked while receiving a purge stream for the time necessary for the desiccant to be completely regenerated. This time is much less than the drying cycle time and is of course determined not by a fixed cycle time, but by the moisture level in the floor. As a result, the heating element 22
The heating element and timer are timed to be energized for a period of time necessary for complete regeneration of the desiccant, at which point the heating element and timer are automatically shut off. The flow of purge gas is continued for a sufficient period of time to cool the desiccant bed to room temperature, the temperature at which adsorption is more effective, and the flow is adapted to continue for a sufficient period of time to cool the desiccant bed to room temperature, the temperature at which adsorption is more effective, and the flow is adapted to continue at a temperature at which adsorption is more effective. Closing valve D provides an automatic shut-off, allowing the spent bed to be repressurized and ready for the next cycle. Typically, 30 minutes to 2 hours is sufficient for complete regeneration if the bed is heated by heating elements to a temperature of 100-250°C. However, it will be apparent that other temperatures and times may be used depending on the desiccant utilized. When the condenser 20 detects a predetermined moisture content in the floor 10 of the tank 1, a timer is activated and valves A, B, and C are switched to close valve A and open valve B to prevent inflow. Gas is pipe 4
The second tank 2 is diverted to the drying cycle state.
At the same time, the discharge valve C is opened. Therefore, the purified flow is pipe 13, orifice 1
4 and via line 12 and valve E to the bottom of tank 1, which is put into a regeneration cycle. When valve A is switched, heating element 22 of floor 10
The switch is turned on and the bed is heated to activate the desiccant. This cycle continues until condenser 21 senses a predetermined adsorbent moisture level in tank 2, then valve A,
B, C, D are switched again and the cycle repeats. In the dryer of FIG. 4, no heat is used to regenerate the spent desiccant, and condensers 50 and 51 are used to detect the progress of the desorption steam front during desorption. The dryer consists of two tanks 31, 32, with suitable line connections 33, 34 for delivering the humid inlet gas from line 36 via valves A, B, and lines 37, 38 from each tank. Connections for delivery to the outflow line 35 via check valves G, H and desiccant fill and discharge ports 47, 48 are provided. Desiccant 40 is supported on a screen support 49 within each tank. A cross conduit 42 is bridged to the outflow conduits 37, 38 and a conduit 4 extending to the outflow conduit 35.
3 are provided with two check valves E and F on each side. A pressure reducing orifice 44 is provided in the pipe line 43, and the pressure exceeding this is removed by a purification discharge valve C.
Alternatively, when D is open, the pressure is approximately atmospheric, and a purification adjustment valve I for measuring the flow rate passing through the pipe line 43 is disposed. Said valve I controls the purification flow rate taken from the effluent gas for regeneration of the spent tank and is read from the purification flow rate indicator P. Another conduit 45 extends between conduits 33 and 34 and is provided with purification discharge valves C and D, respectively, through which, when opened, the purification fluid is discharged into the atmosphere via a discharge conduit 46. . approximately 1/3 down the length of the floor of each tank.
A pair of capacitors 50, 51 are provided which sense the moisture level of the adsorbent bed by capacitance, as in FIG. Once completed, the operation of valves A, B, C, D and both pressure valves I is controlled in response to the moisture level. If tank 31 is in the drying cycle and tank 32 is in the regeneration cycle, with valves A and D open and valves B and C closed, the operation of the dryer is as follows: e.g. 7Kg / ^cm2 (100psig) pressure, 305s.cf
At a flow rate of m., humid inlet gas saturated at 26.7°C (80°) enters line 33 via inlet line 36, passes through valve A, flows into the top of the first tank, and then It flows through a desiccant bed 40, e.g. The gas flows into the gas outlet line 35. The effluent gas is delivered at a pressure of 6.35 Kg/cm ² (95 psig) and a flow rate of 265 s.cfm, with a dew point of -100°C. Check valve H prevents drying gas from entering line 38. the remainder of the dry effluent gas,
In this example, 40s.cfm is at the outlet and pipe 3
5 through line 43, passed through valve I and orifice 44 to reduce the pressure to approximately atmospheric pressure, and then passed through line 42 through check valve F for regeneration. It is sent to the bottom of the second tank 32, which is in a cycling state. The clarified stream passes through desiccant bed 40 and exits into line 34 and then passes through clarified discharge valve D, from where it is discharged to the atmosphere via line 46. This cycle continues as the moisture front of wet vapor desorbed from desiccant 40 travels upwardly through bed 32 and past condenser 51 . Once the moisture front has passed, indicating that moisture has been desorbed from the floor below the condenser, a signal is given, valve D is closed, and tank 32
is repressurized and the purge flow is stopped. Each bed has a
Since the time in the drying cycle is usually longer than the time required to regenerate the spent bed, inlet valves A and B are activated after a period of time sufficient for complete regeneration and repressurization of the regenerated desiccant. It is timed as follows. Once this time has elapsed, valves A and B are automatically switched and the incoming gas flow is diverted to the regenerated bed. When condenser 51 senses a predetermined moisture level in tank 32, indicating completion of desorption, valve D is closed and bed 32 is repressurized. When a certain period of time for adsorption has elapsed in the tank 31, a timer closes the valve A and opens the valve B, so that the humid inflow gas flowing through the inlet 36 is transferred to the tank 32 via the pipe 34. and dry effluent gas is sent from tank 32 via line 38 to dry gas distribution line 35. The purge gas flow in cross conduit 42 is then reversed, and the purge flow flows in conduit 43 through valve I, orifice 44, to conduit 42, and through valve E.
through line 37 to tank 31 in the regeneration cycle and via bed 40 to line 33.
It flows through pipe 45, passes through valve C, is sent to pipe 46, and is discharged into the atmosphere. This cycle is determined by the condenser 50 and
The regeneration of tank 3 continues until the regeneration of tank 3 is completed, then the condenser closes the purification discharge valve C and
1 is repressurized. Tank 3 in drying cycle
Operation at 2 continues until the timer reverses valves A and B, then the cycle begins again. Usually the drying cycle is superatmospheric, i.e. 1.05~
It is carried out for gases at pressures on the order of 24.5 Kg/cm ² (15-350 psig). The combination of orifice 44 in cross conduit 43 and purge exhaust valves C, D causes the regeneration cycle to occur at a pressure significantly lower than that at which the adsorption cycle occurs. This equipment contains safety devices and
A deactivation timer is provided to count the duration of the NEMA half cycle to prevent damage from occurring when the floor disassembles and fails to regenerate. If the moisture front for the desorbed wet vapor does not reach the condenser within the maximum possible time for a purification NEMA half cycle,
Valves C and D are closed anyway, repressurization takes place, and an alarm is issued. In any case
At the end of the NEMA half cycle, valve A is closed, valves B and C are opened, and valves E and
F, G, and H are operated to fluidize the unfluidized repressurized regeneration bed, causing incoming gas to flow downward, and depressurizing the fluidized bed to begin regeneration. This operation ensures that the maximum purification flow is only required for the required amount of time, rather than in a fixed and usually excessive number of cycles, resulting in purification at start-up of the heatless element dryer. A particular advantage is that gas consumption can be minimized. Other controls for non-heated element dryers include complex control circuits. A condenser located in the area of each bed through which a moisture front is allowed to pass during the adsorption half-cycle of the bed is utilized to determine when during the adsorption half-cycle the front passes through the condenser area. By appropriately choosing the location of the condenser, the time at which the moisture front passes through the condenser location will be a function of the rate of movement of the moisture front during the adsorption half cycle and the rate at which moisture is adsorbed. The moisture front then passes through the condenser location and accumulates until the floor enters the regeneration half cycle. This period is stored as voltage or converted into digital information. When the bed enters the regeneration half cycle, the accumulated time is utilized to calculate the time of purification flow required to effect the regeneration of the bed. When a dryer equipped with the aforementioned control configuration enters a cycle change phase, the controller uses the flow time of the clarified stream during the previous half-cycle to determine the flow time of the purge stream that should continue for complete regeneration of said bed. The accumulated elapsed time required for the front to pass is utilized.
If the required purge time exceeds the maximum purge time possible, the maximum purge flow is delivered and an alarm is sounded. Under normal operating conditions, after regeneration is complete, the exhaust valve is closed and the regenerated chamber is allowed to repressurize. The time that the moisture front travels through the condenser in a new bed is stored for subsequent use to calculate the purge flow time required to regenerate the bed when entering the regeneration half cycle. At the end of the illustrated NEMA half-cycle, valves are actuated to flow incoming gas through the regeneration bed, flow purge gas through the spent bed, depressurize the regeneration bed, and begin purification. This arrangement minimizes the maximum number of purge cycles required during dryer startup. The dryer shown in FIG. 5 utilizes this device. The dryer consists of two tanks 61, 62, with suitable piping connections for delivering moist inlet gas to each tank and dry effluent gas from each tank, and desiccant fill and discharge ports 63, 64. It is provided. Desiccant 60 is supported on a screen support 65 in each tank. The flow of humid inlet gas from inlet line 66 is controlled by valves A, B, and the inlet gas is directed into line 67 or line 68 and into tanks 61,62. Dry effluent gas is routed from the tank to line 69 or 70.
Both pipes 69 and 70 are connected to a drying gas outlet 74. Check valves G and H are installed in each pipe.
is provided. The crossing pipe 71 is the outflow pipe 6
Two check valves E, F are provided on each side of the parallel conduit 72 which is bridged at 9, 70 and extends to the outflow conduit 74. A pressure reducing orifice 73 is provided in the pipe line 72, and the pressure at a position beyond this is reduced to approximately atmospheric pressure or below when the purification discharge valve C or D is open. A purification regulating valve I is provided to meter the flow rate through line 72. This controls the purification flow rate taken from the effluent gas for regeneration of the used tank and is read by the purification flow rate indicator P. The conduit 75 extends between the conduits 67 and 68, and is provided with purification discharge valves C and D for controlling the purification flow rate discharged into the atmosphere via the conduit 76, respectively. A pair of condensers 80, 81 are provided approximately in the center of each tank, which provide a signal when the floor 60 reaches a predetermined moisture level and control a timer, the timer being adapted to control a valve. A, B, purification discharge valves C, D, and repressurization valve I are controlled. If tank 61 is on dry cycle and tank 62
is in the regeneration cycle, valves A and D are open,
When B and C are closed, the operation of the dryer is as follows: e.g. 1.75 Kg/cm ² (25 psig)
At a pressure of 305 s.cfm, humid inlet gas at a flow rate of 305 s.cfm is passed through inlet 66 into line 67, passes through valve A and flows into the first tank 61, where it absorbs a desiccant, e.g., a silica gel bed. 60, through a condenser 80 to an outlet, through line 69, through check valve G, and into dry gas delivery line 74. Outflow gas is 1.4Kg/cm ²
(20psig) pressure and 267s.cfm flow rate. Check valve H prevents drying gas from entering line 70. the remainder of the dry effluent gas,
That is, in this example, gas at a flow rate of 38 s.cfm is removed from the outlet via line 72 and passed through valve I and orifice 73 such that the pressure is equal to It is depressurized to approximately atmospheric pressure and then passed through valve F via lines 70 and 71 to second tank 62, which is in a regeneration cycle. The purge stream flows through desiccant bed 60 into lines 68, 75 and then flows through purge discharge valve D into line 76 for exhaust to the atmosphere. After a predetermined time controlled by a timer, purification valve D is closed. Then, once repressurization is complete, a timer switches valves A and B to place tank 62 in a flowing state and tank 61 in a regenerating state. In the adsorption cycle, the condenser 80 is connected to the tank 6.
The process continues until a predetermined moisture level is detected in one adsorbent. Adsorbent bed 60 by condenser 80
When a predetermined moisture level is detected, a signal is generated and the time required to reach this level is stored in electronic storage. While the tank 62 is in the regeneration state, the purge flow continues until the time stored in the storage device has elapsed, and then draining is stopped by closing valve D and the tank 62
is repressurized. When the timer is reenergized and the floor of tank 62 is repressurized, valve D is closed and purge flow is shut off. Then tank 6
1 continues adsorption until a certain period of time has elapsed, then the timer closes valve A,
switching the flow from tank 61 to tank 62 via line 68 and then opening valve C;
Dry effluent gas is routed from dry gas delivery line 66 to tank 62 . The purge gas flow in cross conduit 72 is then reversed, and the purge flow passes through conduit 72 to valve I.
, via line 71 to valve E, to tank 61 in the regeneration cycle, through the floor via line 67 , via line 75 to open purge discharge valve C and via line 76 . and is emitted into the atmosphere. The cycle continues as before, then the valve is reversed and the cycle begins again. Typically, the drying cycle is carried out with gas at atmospheric pressure or higher, on the order of 1.05-24.5 Kg/cm ² (15-350 p.sig). Owing to the orifice 73 in the cross conduit 72 and the discharge valves C and D, the regeneration cycle is carried out at substantially atmospheric pressure, which is considerably less than the pressure at which the adsorption cycle is carried out. The time allotted for the regeneration of a used bed to take place within the time for the regeneration cycle is:
Depending on the volume of desiccant and the pressure at which adsorption takes place,
Adjusted together with the flow rate of the purge stream. Non-heating element dryers operate at equilibrium, and equilibrium must be maintained under all conditions to which the dryer is subjected during its use. The use of capacitors is a particular application in gas fractionators that utilize electronic sequence timers with digital integrated circuits including oscillators and binary calculators. The condenser is operatively connected to a drive for a plurality of solenoid valves, said solenoid valves being movable between permissible positions and closing gas flow through the bed during adsorption and regeneration; Gas flow through the adsorber gas fractionator is controlled according to a cycle time defined in part by a timer and in part by the moisture content of the adsorbent. The electronic sequence timer consists of conventional commercially available electronic components, which do not form part of this invention, but which, in combination and in the circuit described below, determines the period of time required to operate the gas fractionator. can be stipulated. The heart of an electronic device is an oscillator that emits electronic impulses at selected times. A timer is effectively a self-exciting electronic circuit whose output voltage is a periodic function of time. The oscillator must be able to introduce a time delay between pulses so that the desired time can be selected; the short pulses provided by the timer or oscillator are the basic blocks, and the long times are collected in a binary calculator. . In theory, a timer emits a pulse at a selected time. These pulses are sent to a binary digital computer which counts these pulses and which consists of multiple stages or bits and stores information in the number of pulses at multiple time intervals. A plurality of logic gates located in the logic module are utilized to determine the output state of the computer, responsive to a selected output combination corresponding to a desired time interval, and actuate the solenoid drive accordingly. , thereby achieving the selected time interval for each stage. One type of timer oscillator utilizes a circuit that can trigger itself and free-run as a multivibrator. External capacitors are charged through one set of resistors and discharged through another set. Therefore, the timing interval can be varied within a desired range by changing the values of these two sets of resistors, and this can be easily done simply by selecting resistors with the required resistance values. An example of this type of oscillator is the 555. Other types available are flip-flop multivibrators, capacitive delay operational amplifiers (op-amps) with positive feedback, and capacitive connections.
There is a NOR gate. A binary digital computer receives the pulses from the timing oscillator and counts them. The calculator can contain the desired number of units required for the timing interval that needs to be determined. In the device shown, 14
Stage or bit type binary calculators are utilized because they are readily available and of a satisfactory type. In the computer shown, each counting stage is a stationary master-slave
It is a flip-flop, and the calculator is advanced by one count for each negative transition of the input pulse. but,
Other models are also available. This binary calculator has a series of stages, each having one input and one output. The output of each stage (o _o ) is connected to the input of the next stage. The logic output of each stage is reversed when its input changes from a logic 1 to a logic 0. Therefore, a complete cycle of any stage requires two cycles of the previous stage. This results in a frequency reduction of 2 ¹⁴ (or 16,384:1) in this 14 stage binary calculator. This reduction causes a 10 minute cycle to be driven by the 27.3Hz oscillator. Generally, oscillators are more accurate at higher frequencies. The stages are Q ₁ −Q ₁₄ . of this calculator
Q ₁₄ , the final (or slowest) stage, divides the entire cycle into two halves. In the first half it is at logic 0 and in the second half it is at logic 1. Similarly, Q ₁₃ takes 4 cycles
Q ₁₂ is divided into 8 parts, and Q ₁₁ is divided into 16 parts. By monitoring the outputs of these last four stages, the timer can
It can be determined whether it is in a uniform division or in a sequence. By selective placement of AND, NAND, OR, and NOR gates, these four outputs are interpreted and the appropriate output transistors are activated to operate the solenoid valves. The first ten stages of the binary calculator (not connected externally) have only the function of frequency reduction. However, they are utilized for more advanced cycle position analysis if required for more precise applications. The logic module includes a number of logic gates that are selectively arranged to combine to provide an output that operates a solenoid drive during a defined time interval for each valve function. AND,
The functions of NAND, OR, and NOR gates are well known, and the particular arrangement of these gates will of course depend on the time chosen and the timed oscillator and binary computing device utilized, so it is appropriate for a given circuit. Special arrangements that may be utilized will be apparent to those skilled in the art. The illustrated arrangement is exemplary of possible combinations. Minor modifications to the circuit provide greater accuracy and repeatability when required. This includes removing the oscillator and driving the binary calculator with an unfiltered connection to the secondary winding of the power supply transformer. This essentially uses wire frequency instead of an oscillator. Although the wire frequency is very accurate over long periods of time, it has the disadvantage that this frequency cannot be adjusted. This problem can be alleviated to some extent by providing "division by n computers". This is an integrated circuit adapted to provide one output pulse for each n input pulses, where n can be any integer from 3 to 9. By choosing the right n and the right number of binary calculator stages, different combinations are obtained and a wide range of cycle durations can be selected. The dryer of FIGS. 6-8 utilizes an electronic timer as described above and is comprised of a pair of desiccant tanks. These tanks are arranged vertically. Each tank contains a desiccant bed 71, such as silica gel or activated alumina. The tank is provided with desiccant filling and discharging ports 78 and 79 for filling and discharging desiccant. Only two lines are required to connect the two tanks at the top and bottom, respectively, to introduce the incoming gas containing the moisture to be removed and to pass through the dryer to remove the moisture. The removed dry effluent gas is to be delivered and valves A, B, C, and D are required to switch the flow to allow inlet and effluent gas to flow into and out of each tank. It is. Four valves A, B, C, D are pilot valves AD, BD, CD and operated by solenoids.
Pneumatically actuated by DD, said pilot valve is connected to and controlled by an electronic sequence timer T having the circuit shown in FIG. solenoid valve
Duration for AD, BD, CD and DD is 8th
As shown in the figure. As shown in Figure 7, 24V DC power is
36V center tap transformer T1 and rectifier D
1, D2 and is filtered by a 2200 Mfd electrolytic capacitor C1. Low voltage logic potentials are maintained by powering a 6.2V, 0.4W Zener diode D10 from a filtered 24V.DC supply through a power dissipating 470Ω resistor R1. This zener diode regulation adds the benefit of supply noise isolation for the reduced supply voltage function, but the initial filtered supply voltage and solenoid actuation voltage are within the logic IC's operating range (15V or less). In some cases, it is unnecessary. This low voltage is maintained by diodes D3 and D4.
The 250 Mfd capacitor C2 charge is available to maintain a small leakage current to the integrated circuit IC2 and maintain cycle position memory during short-term power failures. These two diodes are used to maintain the same supply voltage (Vcc: approximately 6V) for all integrated circuits. Timer IC1 of integrated circuit 555 is set to oscillate at 27.962 KHz for a 10 minute cycle by accurately selecting resistors R2, R3 and capacitor C3 utilized in the oscillator circuit. The 555 timer is a highly stable device that provides precise time delays or oscillations for normally on and normally off output timings with adjustable cycles from microseconds to hours, and is capable of providing non-stable and single oscillations. Can operate in stable mode. In the illustrated device, operated in an unsteady mode, the timer triggers itself and free-runs as a multivibrator. The external capacitor is charged via R2+R3, and R
Discharge through 3. Therefore, the duty cycle is set precisely by the ratio of these two resistors, and the resistors are
It can be changed as necessary. The capacitor charges and discharges between 1/3Vcc and 2/3Vcc. Charge and discharge times and frequencies are independent of supply voltage. The charge time is given by the following formula: t ₁ = 0.693 (R2 + R3) C3 And the discharge time is given by the following formula: t ₂ = 0.693 (R3) C3 Therefore, the total time is T = t ₁ + t ₂ = 0.693 (R2+2R3)C3. You can choose any time cycle,
IC1 is set accordingly. The output of the oscillator is a 14-bit type IC binary computer IC.
drive the first stage of 2. IC2 has the function of reducing the frequency of the output of IC1. From this comes two frequencies,
One is from pin 7 and drives pin 1 of LN2907, the other is from pin 1 and drives pin 1 of IC3.
Drive 0. IC3 is a CMOS 12-stage ripple-carry binary computer/
Divider, pulse input shaping circuit, reset/
Line drive circuit and 12 ripple carry 2
It consists of a hexadecimal computer stage. Buffer output is obtained externally by stages 1-12. The calculator is reset to an "all zero" state by bringing the reset inverter input line high. This reset is not utilized in this application. Each computer stage is a static master-slave flip-flop. The calculator is advanced one count for each negative transition of the input pulse. In this integrated circuit, each bit changes between a logic 0 and a logic 1 in a stage when triggered by a negative going pulse (from a logic 1 in the previous stage to a logic 0). Each stage therefore reverses the logic state at half the frequency of the previous stage, maintaining a 12-bit binary record of the unit's cycle position, as shown in FIG. During the final portion of the timing cycle, all 12 bits are in a logic 1 state. Due to the next negative oscillation of the oscillator,
All bits 2 are driven to a logic 0 state and the next cycle begins. The last four bits of this calculator, Q ₉ , Q ₁₀ , Q ₁₁ and Q ₁₂ contain the information necessary to divide the cycle into 16 uniform parts and to know which part the unit is in at any given time. are doing. These four bits are sent to a series of logic gates to determine the appropriate logic state combination that satisfies each of the four outputs being driven. This circuit consists of two NAND gates N1 and N2,
It includes two AND gates A4 and A5. One input of AND gate A4 is NAND gate N2
while the other is connected directly to Q ₁₂ .
The third input is connected to the condenser output for the chambers controlled by valves A and B. The output of A4 is transmitted through a drive transistor (not shown).
Connected to solenoid valve AD. Solenoid A is energized only if all three inputs are 1. One input of AND gate A5 is NAND gate N2
one input is connected to a NAND gate N1. The third input is connected to the condenser output for the chambers controlled by valves C and D.
The output of A5 is a drive transistor (not shown)
Connected to solenoid valve DD via. Q ₁₂ is connected to the solenoid valve BD via a drive transistor (not shown) without any intervening gate. It is activated when the output of _Q12 is 1.
NAND gate N1 is connected to solenoid valve CD via a drive transistor (not shown), and AND gate A5 is likewise connected to solenoid valve DD. Both NAND gates are three-input type, and when all three inputs are 1, the output is 0. However, since all three inputs of N1 are connected to the same stage _Q12 , this NAND gate is simply an inverter, and only when the output from _Q12 is 0 is it connected to the solenoid CD through its output transistor. and nothing else. NAND gate N2 has stages Q ₉ ,
It has three inputs connected to Q ₁₀ , Q ₁₁ and therefore produces an output of 1 unless all Q ₉ , Q ₁₀ , Q ₁₁ are 1. The solenoid valve DD is energized by an AND gate A5 through its drive transistor, said AND gate A5 having three inputs, one from N1, one from N2, and one from N2. One is from the capacitive probe circuit. Therefore, only when both N1 and N2 produce an output 1, and the probe detects moisture, will an output 1 be sent from A5 to the solenoid valve DD drive transistor. NAND gate N1 is at 1 during W and X times and 0 during Y and Z times. NAND gate N2 is W and Y
Hour is 1, X and Z hours are 0. AND gate A
4 is 1 at Y time and 0 at W, X and Z times,
AND gate A5 is at 1 during W and 0 during X, Y and Z, where the appropriate capacitor probe has detected sufficient moisture to require a purge flow. Therefore, the time indicated by the timer and capacitor is shown in FIG. Solenoid valve CD is energized during W and X times, and solenoid valve DD is energized during W times. Solenoid valve BD is energized during Y and Z times and solenoid valve AD is energized during Y time. The power output of these gates is provided by solenoid drives or drive transistors Q ₁ , Q ₂ , Q ₃ and Q ₄
(not shown), current limiting resistors R4, R5, R
6 and R7 (not shown) to switch on and off. These transistors are connected to solenoid valves AD, BD, and
In addition to driving CD and DD, diodes D5 and D
6, D7 and D8 (not shown) provide protection from induced fly-back. The W+X and Y+Z times correspond to the tank and drying cycle times, respectively. The W and Y times correspond to the tank and regeneration stages, respectively, and the X and Z times correspond to the repressurization stage when regeneration is complete. The condenser operates valves A and D to control the regeneration flow and stop the chamber outflow flow when regeneration is complete;
Repressurization occurs while the timer activates valves B and C, diverting incoming fluid from one chamber to the other. Conduit 72 directs humid incoming gas to valves A, B,
The four components including C and D are sent to the inflow switch valve 74. One of the valves C, B directs the incoming gas flow to one of the two inlet lines 75, 76;
One of the pipes 75 and 76 always introduces the inflow gas to the top of each tank, and the other pipe 75 and 76 sends the purified flow of the regenerated outflow gas to the pipe 81 through valves A and D.
It is then sent to the exhaust section via the muffler 82 and discharged into the atmosphere. Within each desiccant bed 71, capacitors 90, 91 are provided at the bottom and connected to AND gates A4 and A5 as described above. A desiccant support consisting of a perforated metal cylinder is provided at the bottom of each tank to maintain the desiccant bed 71 in and within the tank. Outlet lines 83 and 84 from the bottom of the tank lead to a pair of ball check valves 85 and 86, respectively. Valve 74 is actuated by an electronic sequence timer via a solenoid actuated pilot valve, while valves 85 and 86 are pressure actuated. The ball in the flowing tank and the outflow line from the
The fluidization switch in the pipes 83, 84 is moved at the time of start-up, but the other of the balls 85', 86' is moved to the valve seat at the time of said switch, and the pipe 83, which leads to the regeneration chamber under reduced pressure, 84 is closed, thus delivering the main effluent stream via outflow line 87. A filter screen is disposed in each outlet line 83, 84, which is movable and made of sintered stainless steel mesh. This is floor 7
The outlet valve 8 retains desiccant particles that may be transferred from the desiccant support 77 through the desiccant support 77.
5, 86 and the rest of the device from becoming contaminated with said particles. Dry effluent gas distribution line 8 from valves 85 and 86
7 extends and is adapted to deliver the drying effluent gas from the dryer to the equipment to be supplied. The conduit 87 can be provided with an outflow pressure gauge P5 and a humidity gauge H, but these are optional and can be omitted. A transverse conduit 89 having a narrow passage is connected to the outflow conduit 8
3 and 84, the valve 8
When 5 and 86 are closed, these valves are bypassed and the purge flow is routed to lines 83 and 84 for delivery to the non-flowing tank. The pipe line 89 has a pressure reducing function due to its small diameter, and the pressure on the downstream side thereof is reduced to approximately atmospheric pressure when one of the purification valves A and D is open, and is also used for the regeneration of the used tank. Valves 85 and 86 meter the purification flow rate removed from the effluent gas. Purification discharge valves A and D are connected to pipe line 7 by a signal from an electronic sequence device.
5, 76, said sequencing device is adapted to open and close said valves at appropriate times via appropriate solenoid operated pilot valves. Another restricted line solenoid valve E is actuated during repressurization to cause the dryer to perform this process at a faster cycle time. This is optional depending on the size and speed of the dryer. If the left tank is in the dry cycle and the right tank is in the regeneration cycle, then valve 74B
and D are open, 74C and A are closed, and the operation of the dryer is as follows: e.g. 7Kg/cm ²
(100psig) pressure, 305s.cfm flow rate, and 26.7
The humid inflow gas saturated at ℃ (80〓) flows into the inflow pipe 7.
2, through valve 74B (valve C is closed) into the top of the first tank, and downwardly through the interior desiccant bed 1, for example made of silica gel or activated alumina. and flows to the bottom of the tank, then through support 77 and line 83, valve 85, and dry gas outlet line 87.
sent to. The effluent gas is delivered at a pressure of 6.65 kg/cm ² (95 psig), a flow rate of 265 s.cfm, and a dew point of -73.3°C (-100〓). Ball 86' prevents drying gas from entering line 84 other than through line 89. The remainder of this dry effluent gas, i.e. 40 s.c.
The fm is removed via line 89, where the pressure is reduced to approximately atmospheric pressure, and then sent via line 84 to the bottom of the second tank in the regeneration cycle. The purge stream passes upwardly through desiccant bed 71 and enters line 76 from the top and into valve 4D.
, passes through a pipe 81 and a muffler 82, and is discharged from there into the atmosphere. This cycle continues until regeneration is complete as sensed by capacitor 91, at which time capacitor 91 emits a signal and pilot valve DD
By stopping the operation of the purification discharge valve D
will be closed. Line 89 therefore repressurizes the slack tank. Operation of the system with the tank in the dry cycle continues until a fixed cycle time W+X has elapsed, at which time the electronic sequence timer reverses valves 74C,B and the cycle is restarted for the opposite chamber. Ru. The time each bed is in the drying cycle W+X (and Y
+Z is the time W required to regenerate the used floor
(and Y) by time X (and Z). When the regeneration time has elapsed, as detected by condenser 91, valve D (or valve A, if detected by condenser 90) is shut off and the regeneration tank is automatically and slowly refilled via line 89. be pressured. This repressurization is accelerated by opening optional valve E. After a certain cycle time W+X has elapsed, an electronic sequence timer switches valves 74C and B to cause the humid inlet gas entering via inlet line 72 to be delivered via line 76 to the top of the tank. , and the check valve 86 moves,
Line 84 is opened and check valve 85 is moved to close line 83 so that dry effluent gas is routed from the bottom of the tank to dry gas delivery line 87 and line 83 is closed and traversed. The flow is reversed, with the exception of the purge gas flow which bypasses valve 85 via line 89. The clarified stream is sent via line 83 to the bottom of the tank in the regeneration cycle, then flows up the bed into line 75, through valve 74A, line 81, and muffler 82.
From there it is emitted into the atmosphere. Normally, the drying cycle is 1.05~24.5Kg/ ^cm2 (15~24.5Kg/cm2
It is carried out at superatmospheric pressures on the order of 350 psig). Due to the combination of the orifice in cross conduit 89 and the purge exhaust valves A and D, the regeneration cycle is conducted at a significantly reduced pressure than the pressure at which the adsorption cycle is conducted. The drying apparatus of this invention can utilize any type of adsorbent that adsorbs moisture from gas. Activated carbon, alumina, silica gel, magnesia, various metal oxides, clays, clay, bone char, and Mobilbeads, and similar hygroscopic compositions can be utilized as desiccants. Molecular sieves can also be used as they often have dehumidifying properties. This type of material contains natural and synthetic zeolite, the pore size of which varies from a few angstroms to 12-15 Å or even more. Chabasite and analcite are typical natural zeolites that can be used. Available synthetic zeolites are described in U.S. Patent No.
Including those disclosed in 2442191 and 2306610. All these materials are well known as desiccant agents and the reader is referred to the literature for a detailed description. All of the illustrated and described dryers are applicable to clarified stream regeneration where the clarified stream flows in a counter-direction to the humid inlet gas. As is well known, this
This is the most effective method when using a desiccant bed. Since the humid gas passes through the desiccant bed in one direction, the moisture content of the desiccant bed gradually decreases, and typically the least amount of moisture will be adsorbed at the outlet end of the bed. Therefore, it is practical to introduce the regeneration purge gas at the outlet end in order to avoid moving moisture from the wet to the dry parts of the bed, thus increasing the required time of the regeneration cycle. It is. When the purge stream is introduced into the outlet end, a small amount of moisture present therein is removed by the purge stream and carried towards the moist end of the bed. The bed is thus regenerated sequentially from the outflow end and all moisture is transported the minimum possible distance through the bed before being released from the inflow end. However, for some purposes it may be desirable to have the purge stream flow in the same direction as the inlet stream. In this invention, it is possible to maintain the moisture content of the desiccant at a much higher level than normally possible, which allows flow to be interrupted at more precisely metered moisture levels than previously possible. This is due to the protective effect of the humidity detection element. As a result, in many cases, when the bed is nearly saturated throughout, there is little difference between when the purge stream enters at the inlet end or the outlet end, and the present invention Both types of operation (forward and reverse flow) are contemplated, although of course reverse flow regeneration is preferred in most cases. Each of the illustrated dryers utilizes one condenser per tank. However, it is also possible to utilize two, three or more such capacitors per tank. In this case, operation of the device is guaranteed even if one or more capacitors become defective in a group. Condensers are placed at different levels within the desiccant bed,
It may be possible to follow the successive passage of moisture fronts onto the floor. As previously mentioned, during the continuation of the drying cycle, the moisture front moves from the inlet end toward the outlet end of the bed. Therefore, due to the passage of the moisture front, condensers farther from the outlet end are activated sooner than condensers closer to the outlet end. Two mutually spaced capacitors are operated at different times and this fact can be utilized to operate different stages of the cycle, such as regeneration and repressurization, at different times. Therefore, one condenser can be placed at a point a considerable distance from the outflow end of the bed, e.g. at a mid-point in the bed, to detect the moisture front at time A of the cycle and turn on the heating elements of the regenerating bed. shut off, e.g., before the bed is placed in a drying cycle.
The heating element can be utilized to shut off early to allow sufficient cooling during the regeneration cycle. A second intermediate condenser is available to close the purge exhaust valve and repressurize the regeneration bed. A third condenser off the edge of the bed is available to switch on the cycle and end the drying cycle. In this case, a timer is unnecessary and the playback time is determined by the capacitor rather than the timer. The incoming gas stream may sometimes be contaminated with gases and liquids other than the substance to be adsorbed, such as water, in which case when adsorbed on the adsorbent,
The dielectric constant changes in a state different from that of the intended adsorbent,
Desiccant is adversely affected. Such contaminants reduce the adsorption capacity to adsorb water, or coadsorb with water, thereby altering the response of the capacitor by displacing water with a higher dielectric constant and lower or It is replaced by a contaminant with a higher dielectric constant. These responses will result in a spurious moisture content indication at a given level of the adsorbent bed. False responses are corrected by placing two condensers in each bed, one at a selected control point and another away from the inlet end of the bed. The control capacitor is affected by the aforementioned contaminants before the second capacitor. By considering the effect of bed position changes on the capacitor output in the absence of contaminants, the effect of the aforementioned contaminants on the capacitor response can be determined by comparing the outputs of the two capacitors, either continuously or intermittently, A contamination condition is detected when it produces a higher output than the control capacitor. The above comparison reveals whether the adsorption capacity of the control capacitor location has been impaired by contaminants or the dielectric constant has changed due to contaminants, and whether the control capacitor is in a faulty state. The desiccant bed condenser can be placed at any depth within the diameter of the bed, but the distance from the outlet depends on the gas velocity and temperature, which affects the moisture front and flow rate in the bed. Other factors previously discussed are the moisture content of the incoming gas and the moisture content and level at which the condenser is operated. Depending on the moisture content of the adsorbent, condensers can be positioned to detect desired moisture levels in the effluent gas even at very low dew points or relative humidity, but typically at dew points of approximately −34.4°C (−30°C) ) can detect a higher range of moisture content. The moisture level of the drying effluent gas from the desiccant bed is typically around -90°C (-130°C) at the outlet end of the bed.
and no less. The exact location of the condenser within the bed depends on two factors: the length of time to regenerate the bed;
and prevention of breakthrough of the dew point of the effluent gas. Clearly, the condenser must be positioned and adjusted to detect high moisture levels in the adsorbent bed under the most unfavorable conditions of inlet gas velocity, humidity and temperature, before the outlet gas dew point becomes excessive. Must be. This is done as shown in FIG. However, the condenser must be positioned such that the amount of water required to saturate the bed sufficiently to operate the condenser can be desorbed during a given regeneration cycle time. Therefore, in dryers where the regeneration time increases disproportionately to the increasing moisture content, for example in beds with unheated elements, the condenser is moved closer to the inlet and the bed is lower than the heated bed. It is considered to be used at total moisture content. As mentioned above, considering the aforementioned factors, for any given adsorption condition, the appropriate location of the condenser for sensing the moisture front at the appropriate time to determine the drying cycle can be determined empirically; As shown in FIG. 1 of US Pat. No. 3,448,561, hourly dew point or hourly relative humidity data and graphs are obtained for the dryer. A preliminary test on the change in the dielectric constant of the desiccant with respect to the change in the amount of water adsorbed on the desiccant was conducted using 704.7.
cm ² (109.25in ² ) and 1.9cm
(0.75 in.) using a test device consisting of two metal plates spaced apart. The space between these plates is filled with a dried or saturated sample of alumina or silica gel desiccant. The measured capacitance values are as follows. Dry Moisture Dry/Moisture Ratio Alumina 94.3pF 591pF 6.3 Silica Gel 97.0pF 331pF 3.4 A test to determine the effect of oil on the capacitance of a desiccant also measures the capacitance of a sample of dried alumina in the apparatus and This was done by removing the sample, mixing it with excess oil, returning the alumina-oil mixture to the test equipment, and measuring the capacitance again. Dry alumina provides a capacitance of 97.5 pF and the alumina-oil mixture provides a capacitance of 102.0 pF, so the effect of normal levels of oil contaminants on the aforementioned capacitors is considered negligible. The following example, in the view of the inventors, shows a preferred method of operation of the drying device according to the invention. Example 1 A two-bed heat regenerable dryer of the type shown in Figure 3, with two desiccant beds of 1.52 m
(60in.) length, 31.4cm (12 3/8in.) diameter, and
A dryer with 80.4Kg (177lb) of silica gel is
80% relative humidity, 6.44Kg/ ^cm2 (92psig) inlet pressure, and 350scfm flow rate, and 37.8℃ (100〓)~
It was used to dry air at 21.1℃ (70〓). The surface velocity of air is 37.8℃ (100〓) per minute
It was 19.9m (65.1ft). Two condensers, X, Y, operable at approximately 3% relative humidity corresponding to a 2% sorbent moisture content are placed in the bed, with X located 76.2 cm (30 in.) from the outlet end of the bed. and Y was located at 61 cm (24 in.). The table below shows the results obtained using this device under the conditions described above.
Data from a drying cycle, in each case the drying cycle ended when condenser Y was signaled:

[Table] As is clear from these data, capacitors X and Y each provide an alarm at the end of the drying cycle when the effluent gas is at a safe moisture level. Also, as evidenced by the different cycle times, the condenser allows the length of the cycle time to be adapted to variations in the moisture content level of the incoming air, thus substantially reducing the number of regenerations and extending the life of the desiccant. keeping. 37.8°C to deliver the effluent gas at the appropriate moisture content level.
(100〓), the cycle time must be set to less than 2.4 hours, e.g. 2.2 hours, and when operated above that condition, the moisture content level of the effluent gas This is because there is a possibility that the value exceeds the required value. The cycle time can be extended to 6.2 hours if the air is at 21.1°C (70°) and therefore less moisture content air is introduced. Obviously, if the moisture content of the incoming gas is further reduced, the cycle time will be lengthened accordingly. Longer times are possible if the flow rate or inlet temperature is reduced. Example 2 A two-bed unheated element dryer as shown in Figure 4 with a desiccant bed 1.35 m (53 in.) long and 20.3 cm long.
(8in.) diameter, containing 38.6Kg (85lb) of alumina per bed, at 50% relative humidity, 21.1
℃ (70〓) and inlet pressure 4.76Kg/cm ² (68psig)
was used to dry the air. The flow rate is
125 scfm equals a surface velocity of 17.7 m (58 ft) per minute. As shown in Table 2, the eight sensors are spaced apart at a distance of 15.2 cm (6 in.) to detect 21/2 to 16% adsorbed water.
Two condensers A-H were placed in the bed. These condensers are used to detect the concentration gradient from the inlet to the outlet side of the bed by measuring the moisture content of the adsorbent alumina at each location.

TABLE The moisture content of the adsorbent bed was plotted against time as shown in FIG. The upper line corresponds to the adsorbed moisture profile of the bed at the end of the half cycle when the bed is in the flow state. The lower line corresponds to the adsorbed moisture profile of the bed at the end of the regeneration half cycle. Floor outflow end 1.36m
The amount of adsorbed water at (53.5 in.) is calculated assuming that it is in equilibrium with the outflow air. Therefore,
The cycle time can be terminated at any point as required to deliver the effluent gas at the appropriate moisture content level. Although the invention has been described primarily in terms of a desiccant dryer and the drying of gases, the device can also be used to separate one or more gaseous components from a gas mixture by appropriate selection of adsorbents. . In that case, the adsorbed components are removed from the adsorbent during regeneration by application of heat and optionally reduced pressure. Accordingly, this process can be used to separate hydrogen from petroleum hydrocarbon streams and other gas mixtures containing the same components, to separate oxygen from nitrogen, to separate olefins from saturated hydrocarbons, etc. Adsorbents that can be used for these purposes will be obvious. Adsorbents often useful for removing moisture from the air, such as activated carbon, glass wool, adsorbent cotton, metal oxides, and clays such as attapulgite and bentonite, white clay,
Bone char and natural and synthetic zeolite can also be used to adsorb one or more gaseous components from the mixture. Zeolites are particularly effective in removing nitrogen, hydrogen and olefins such as ethylene or propylene from propane and high paraffin hydrocarbons, or butenes or high olefin mixtures. The selectivity of zeolite depends on the pore size of the material. Since the selective adsorption properties of the available zeolites have been shown in the literature, material selection for a particular material is straightforward and does not form part of this invention. In some cases, adsorbents can be utilized to separate multiple materials in a single pass. For example, activated alumina can adsorb water vapor and carbon dioxide, whereas Mobilbeads only adsorb water vapor. The equipment used for this purpose is
It is the same as shown and described in the figures, with appropriate modifications according to the proportions of the components to be separated, the operating pressure and temperature, and the volume of adsorbent. However, it will be clear that this method has particular application in the drying of gases and is the preferred embodiment of the invention.

[Brief explanation of drawings]

1 is a graph showing the moisture content of the adsorbent bed as a function of gas drying time; FIG. 2 is a graph of the moisture content of the adsorbent bed determined in Example 2 as a function of capacitance versus time; FIG. The figure is a schematic diagram of a dryer using a two-bed type heated regenerated desiccant according to the present invention, FIG. 4 is a schematic diagram of a dryer using a two-bed type non-heating element desiccant according to the present invention, and FIG. FIG. 6 is a schematic diagram of a two-bed dryer using a heated regenerated desiccant controlled by an electronic timer according to the present invention; FIG. 8 is a timing diagram showing the time sequence of the timer circuit of FIG. 7. 1, 2... Container, 3, 4, 6... Inflow gas distribution pipe, 5, 7, 8... Outflow gas delivery pipe, 20,
21... Capacitor.

Claims (1)

[Scope of Claims] 1. A method for reducing the concentration of a first gas in a mixture of a first gas and a second gas to below a maximum concentration limit, the method comprising: flowing in contact with an adsorbent bed from one end of the adsorbent bed to the other, causing said bed to adsorb a first gas and forming an effluent gas having a concentration lower than said maximum value, and continuing the adsorption. forming a first gas concentration gradient in the bed from the one end to the other end, and increasing the concentration of the first gas in the second gas as the adsorption capacity of the adsorbent decreases; defining a progressively advancing concentration front in the bed from one end thereof to the other; and determining the progression of the first gas front in the bed as a function of the capacitance of the condenser to the first gas content of the adsorbent. in which the adsorbent is dielectric and the capacitor is located sufficiently far from the edge of the bed that an effluent gas having a critical maximum first gas concentration is detected by determining a change in the gas concentration. The gas mixture is prevented from escaping from the bed, before the effluent gas exits the bed, and before the concentration of the first gas in the second gas exceeds the limit maximum concentration. A method in which the flow is stopped in contact with a bed, and the desorption time of a desorption cycle for desorbing the first gas from the bed is controlled based on an output signal of a condenser that detects a concentration front. . 2. The method of claim 1, wherein the first gas is water vapor. 3. The first gas is desorbed from the bed by flowing a purified gas stream having a low concentration of the first gas in contact with the bed, and the adsorption and desorption cycle is sequentially repeated. 1st
The method described in section. 4. The first gas is desorbed from the bed in a high temperature state sufficient to desorb the first gas.
A method according to claim 1. 5 At low pressure due to the pressure at which adsorption takes place,
A method as claimed in claim 1, characterized in that a first gas is desorbed from the bed. 6. The method of claim 1, wherein the first gas is desorbed from the bed at a pressure lower than atmospheric pressure. 7. The method of claim 1, wherein flowing the gas mixture in contact with the bed is automatically stopped when the concentration front has advanced a predetermined distance through the bed. 8 Utilizing two adsorbent beds, one in the first gas adsorption cycle and the other in the cycle of desorbing the first gas with a purge stream consisting of the effluent from the first bed. A method according to claim 1. 9. The method of claim 8, wherein said bed in a desorption cycle receives a room temperature purge stream. 10. The method of claim 8, wherein said bed in a desorption cycle receives a purge stream at an elevated temperature sufficient to assist in desorption of said first gas. 11. The method of claim 9 or 10, wherein the bed in the desorption cycle receives a lower pressure purge stream in the adsorption cycle. 12 A method for reducing a first gas concentration of a mixture of a first gas and a second gas below a critical maximum concentration, the method comprising: placing the mixture on an adsorbent bed having a preferential affinity for the first gas; the first gas is flowed in contact from one end to the other to adsorb a first gas in the bed, and the concentration of the first gas in the bed gradually decreases from one end to the other during the continuation of the adsorption. forming a gradient effluent gas, whereby an increase in the concentration of the first gas in the second gas defines a concentration gradient in the bed that progressively progresses from one end to the other with a decrease in the adsorption capacity of the adsorbent; , detecting the advancement of the gradient in the bed as a function of a change in the first gas content of the bed and thus a change in the capacitance of a capacitor with the adsorbent as the dielectric; a desorption time of a desorption cycle in which the flow of the gas mixture in contact with the bed is stopped and the first gas is desorbed from the bed before a quantity of effluent gas exits the bed; A method in which the concentration gradient is controlled based on the output signal of a capacitor that detects the concentration gradient. 13. The method of claim 12, wherein the first gas is water vapor. 14 A device for reducing a first gas concentration of a mixture of a first gas and a second gas to a critical maximum concentration or less, the device comprising: a container; provided in the container;
a chamber for a bed of adsorbent having a preferential affinity for the gas, a channel for delivering the inflow gas to the inflow end of said bed, a channel for delivering the outflow gas from the outflow end of said bed; A capacitor comprising two conductors having selected surface areas, said conductors as a dielectric to define a space for containing an adsorbent;
a selected first capacitor spaced a sufficient distance from each other such that the capacitance is varied by sensing a change in the first gas content of the adsorbent as a function of dielectric constant; a device for generating a signal in response to a change in capacitance when a selected capacitance indicative of gas content is reached or exceeded; a device for interrupting the flow of the incoming gas in response to the signal; A device for controlling a desorption time of a desorption cycle for desorbing the first gas based on an output signal of a condenser that detects a concentration front. 15 the condenser is located in the center of the floor;
From the outflow side end of the bed to 1/50 of the length of the bed
15. The device according to claim 14, arranged in the 2/3 position. 16. Claim 14, wherein a pair of containers are arranged, each container having a chamber therein for a bed of adsorbent and a passageway for delivering the incoming gas and for delivering the outgoing gas. The equipment described in section. 17. Claim 16 comprising a device for diverting a portion of the effluent gas from said one vessel to the other vessel for desorbing adsorbed first gas from said bed.
The equipment described in section. 18. Claim 1 comprising a device for heating a bed of adsorbent within said vessel to a sufficiently high temperature condition to assist in desorption of a first gas adsorbed on said bed.
The device according to item 4. 19. The apparatus according to claim 14, comprising a device that reduces the pressure during desorption to a level lower than the pressure during adsorption. 20. The apparatus of claim 14, wherein the container is not equipped with a heating element. 21. The apparatus of claim 14, comprising a device activated by the condenser during a regeneration cycle to shut off the adsorbent bed and stop the regeneration cycle after completion of regeneration of the adsorbent. 22. The first stage of the bed is located at a sufficient distance from the outlet end of the bed so that the effluent gas exceeding the critical maximum concentration of gas to be adsorbed can terminate the cycle before exiting the bed. 15. The device according to claim 14, wherein a plurality of condensers are arranged to detect the content of one gas.