Jump to content



Featured Articles

Check out the latest featured articles.

File Library

Check out the latest downloads available in the File Library.

New Article

Product Viscosity vs. Shear

Featured File

Vertical Tank Selection

New Blog Entry

Low Flow in Pipes- posted in Ankur's blog

Air Drying


This topic has been archived. This means that you cannot reply to this topic.
9 replies to this topic
Share this topic:
| More

#1 sskumar

sskumar

    Junior Member

  • Members
  • 17 posts

Posted 17 February 2005 - 08:24 AM

I have a doubt regarding air drying.

Normally people recommend Activated Alumina for drying. Why silica gel or molecular seives are not recommended. In our plant we are using activated alumina as adsorbent for dryers of compressed air plant and we are using TSA and dew point(DP) at the outlet is -40degC (DP at 1 atm.). In our nearby Captive Power Plant, they are using silica gel for their air dryers and they also use TSA. I heard people say that the crushing strength of silica gel is less and hence they are not recommended for high pressure operation. But in our plant and in Captive Power plant, the dryer operating pressure is the same , 9atm abs. Then where is the difference?

Also, in our Air separation plant, we have dryers for removing moisture and other impurities(like CO2,H2S, hydrocarbons etc ) from air. Dryer is divided into two portions. In the first portion, moisture will be removed by activated alumina and then in the second portion other impurties will be removed by Molecular Seives. People here say that inorder to reduce loading on Molecular Seives, activated alumina is used to remove the moisture. I think molecular seives can as well remove moisture. Then why activated alumina is provided? Here in this case both TSA and PSA are being used to get dew point of -100 degC(DP at 1 atm).

Please tell me the advantages with activated alumina and disadvantages with silica gel and molecular seives for air drying.

Thank you,
Regards
sskumar

#2 Art Montemayor

Art Montemayor

    Gold Member

  • Admin
  • 5,782 posts

Posted 17 February 2005 - 05:57 PM

sskumar:

On the issue of drying compressed air or other similar gases down to dewpoints of -50 oF or lower:

The main advantages of using Activated Alumina (there are several grades available) are that it has superior "ruggedness" and yields a lower dewpoint on the product gas when compared to silica gel under the same operating and regeneration conditions. By ruggedness is meant the ability to sustain and remain in its original physical shape (without breakage or "dusting") while retaining its dynamic sorptive capacity. It also has a definite sorptive attraction for lubricating oils and this serves to capture all of the oil discharged in the gas from an oil-lubricated compressor like a reciprocating or screw machine.

The main disadvantage with employing silica gel is that it "dusts" or undergoes attrition much easier than other adsorbents such as Activated Alumina. This tendency to attrition is pronounced when the adsorbent bed is subjected to mechanical shaking or movement caused by the instantaneous pressure difference across the bed when the same bed is changed from regeneration phase to the adsorption phase. Note that I have not said that so-called high pressure adsorption operation causes attrition or dusting. I have never seen this effect nor have I ever heard of it. What I have witnessed is the crushing and dusting of adsorbent - silica gel, Activated Alumina, or Molecular Sieve - when the same adsorbent(s) are subjected to bed movement caused by rapid or careless change-over of the bed from regeneration to adsorption service (or vice-versa, depending on the direction of the related gas streams). Operating an adsorbent at relatively high pressures doesn't "crush" it; what crushes the adsorbent particles is the force generated by a pressure difference - not a pressure prescense. Be careful in understanding how a physical force is applied on an object. The adsorbent is porous and any pressure imposed on it will be imposed on all its surfaces - interior and exterior - thereby causing a homogeneous pressure throughout and no pressure difference across the adsorbent. You have to think as a mechanical engineer - but as an engineer, nevertheless.

I always endeavor to design my adsorption beds with the high pressure stream (usually the process wet gas) flowing in the downward direction, and NEVER in the upward direction. I always see to it that the bed support grids and screens are designed to withstand the force generated by the presssure difference upon b ed change-over. Many manufacturers and users fail to recognize the attrition effect of the bed movement when the high pressure stream is flowing upwards and there is no upper restraining grid or screen to keep the bed intact and without relative movement of the individual adsorbent particles. The worst effect is when actual fluidization of the bed takes place. This is to be avoided at all costs. It causes bed destruction, excessive dust, clogs filters and screens, reduces the efficiency of the dryer and generally results in downstream chaos.

You asked for the advantages and disadvantages and I've given those above.

Additionally, I have achieved a consistent and constant dewpoint of -90 oF (@ atm pressure) with gases such as air, CO2, Oxygen, and Nitrogen. The cycle used was a TEMA 8 hr cycle with low pressure dry gas regeneration at a temperature of 650-700 oF. You will find that if you were smart and designed your adsorber vessels to withstand 800-900 oF, you can regenerate the Activated Alumina up to 700 oF and obtain super-dry gas product. ALCOA experts will confirm this operation to you. If I were you, I would eliminate the double inventory of two adsorbents and concentrate on using only one - Activated Alumina.

In my field experience I have found that the two adsorption processes you have in series in your Air Separation Unit is the best and the most flexible design to incorporate in such an operation. While I have used Mol Sieves to do simultaneous drying and purification of air prior to introduction into an Air Separation Unit, I would not do this today. The cost of Mol Sieves is high. To subject them to mundane moisture separation is a waste of money when compared to their important downstream purification role. Activated Alumina serves as a great adsorbent up front, doing the "dirty and mundane" work of removing the moisture, dust, and oil out of the initial feed air. Once the front-end cleanup of the air is done, the real important and specific job of purifying the feed air can be undertaken by the Mol Sieves with specific ease and simplicity.

I hope this experience helps you out by responding to your query.

Art Montemayor
Spring, TX

#3 sskumar

sskumar

    Junior Member

  • Members
  • 17 posts

Posted 19 February 2005 - 05:04 AM

Art Montemayor:

Thank you very much for your detailed reply. You are absolutely correct. Velocity depends on differential pressure. As I told you, in our Compressed Air Plant (CAP), we use only thermal swing adsorption. Regeneration of bed will be carried out only under pressure. Hence chances of bed seeing high diff. pressure does not arise. If bed does not see high differential pressures, chances of attrition of adsorbent won't be there and in those conditions, can we go for silica gel.
Here with activated alumina, for regeneration we heat up part of the incoming air to 175-180 degC and pass it through the bed under regeneration step. If we use silica gel or molecular seives, upto what temperature we have to heat up the air ( regneration temperature ) to do the regeneration for getting the same quality of air when the regenerated bed comes in line for adsorption step. Will the temperature be the same as with activated alumina?
In case of air separation unit dryers, we use both pressure swing and thermal swing adsorption. Here chances of bed seeing high differential pressures are there we cannot go for silica gel. I agree fully with your argument of unnecessarily loading molecular seives for removal of moisture when it can do efficiently do the job of removing other impurities.
Thanks once again
Regards
sskumar

#4 Guest_Guest_sskumar_*

Guest_Guest_sskumar_*
  • guestGuests

Posted 22 February 2005 - 11:06 PM

QUOTE (sskumar @ Feb 19 2005, 05:04 AM)
In our Compressed Air Plant (CAP), we use only thermal swing adsorption. Regeneration of bed will be carried out only under pressure. Hence chances of bed seeing high diff. pressure does not arise. If bed does not see high differential pressures, chances of attrition of adsorbent won't be there and in those conditions, can we go for silica gel?

Here with activated alumina, for regeneration we heat up part of the incoming air to 175-180 degC and pass it through the bed under regeneration step. If we use silica gel or molecular seives, upto what temperature we have to heat up the air ( regneration temperature ) to do the regeneration for getting the same quality of air when the regenerated bed comes in line for adsorption step. Will the temperature be the same as with activated alumina?

[QUOTE]

could some body answer the above
rolleyes.gif

#5 Art Montemayor

Art Montemayor

    Gold Member

  • Admin
  • 5,782 posts

Posted 23 February 2005 - 12:21 PM

In my opinion, Silica Gel will not yield the same, low dew point as Activated Alumina when both are used in the same TEMA cycle, at the same process and regeneration temperatures and pressures.

I have found Silica Gel to be lacking in yielding the industrial low dew points (-70 to -100 oF). It will not tolerate regeneration at the higher temperatures in a TSA cycle.

Just because "regeneration of bed will be carried out only under pressure", does not mean that there will not be dusting or adsorbent attrition. An adsorbent bed that is not fixed and held down properly will always shift, settle and move (however slightly) as it undergoes cycle after cycle. And Silica Gel will dust while the bed shifts and settles.

I have already addressed the question of Mol Sieve regeneration: it requires much higher regen temperatures - approximately 500 to 800 oF. Silica Gel, in my opinion, cannot tolerate these temperatures. You haven't stated if your regeneration air is saturated with moisture or totally dry. I have to assume it is "wet", since you say it is "incoming air". If you are regenerating with wet gas (which is OK, except you have to take special design steps) you CANNOT flow the hot (& cool) regen gas in a COUNTER-FLOW direction to that of the main feed air. If you do, you will be saturating the adsorbent (with cool, wet gas) during the last step of the regen phase and upon subsequent bed change-over, the outlet product gas out of that bed will be "spiked" with moisture and fail the product dew point specs.

A strong reason for rejecting Silica Gel in ultra-dry, process adsorption applications is that it fails to achieve mechanical and physical stability under the higher regen temperatures required to obtain properly regenerated adsorbent and consequently yield very low dew points in the product gas. And, don't forget, the main reason for using Adsorption as the Unit Operation of choice in a process is because you desire very low dew points in the product gas. If you don't require the low dew points, then you shouldn't be using or applying Adsorption as your drying Unit Operation. Use another, cheaper method and save on the process energy.

I hope the above answers your questions.

#6 Guest_Guest_*

Guest_Guest_*
  • guestGuests

Posted 25 February 2005 - 10:14 AM

QUOTE (Art Montemayor @ Feb 23 2005, 12:21 PM)
I have to assume it is "wet", since you say it is "incoming air". If you are regenerating with wet gas (which is OK, except you have to take special design steps) you CANNOT flow the hot (& cool) regen gas in a COUNTER-FLOW direction to that of the main feed air. If you do, you will be saturating the adsorbent (with cool, wet gas) during the last step of the regen phase and upon subsequent bed change-over, the outlet product gas out of that bed will be "spiked" with moisture and fail the product dew point specs.

A strong reason for rejecting Silica Gel in ultra-dry, process adsorption applications is that it fails to achieve mechanical and physical stability under the higher regen temperatures required to obtain properly regenerated adsorbent and consequently yield very low dew points in the product gas. And, don't forget, the main reason for using Adsorption as the Unit Operation of choice in a process is because you desire very low dew points in the product gas. If you don't require the low dew points, then you shouldn't be using or applying Adsorption as your drying Unit Operation. Use another, cheaper method and save on the process energy.

I hope the above answers your questions.

Art Montemayor:

Now it is clear why we should prefer activated alumina over silica gel and molecular seives for air drying/moisture removal from any gas. The problem with silica gel is dusting and low quality air at the o/l as it can't be regenerated at high temperatures. For the same quality of air as with activated alumina, the regeneration temperatures are much higher for molecular seives under the same conditions. Am I right, sir?

[QUOTE]

I think if we regenerate the bed with dry air, we need not have to subject the bed to high temperatures (175 - 180 deg C). But normally we see dryers with high regeneration temperatures though dry air is used for regeneration. In our air separation plant, we use bone dry air for regneration of the bed ( in fact it is waste N2 stream from air distillation column after passing through vaporiser). Even then it is heated upto 175 deg C and passed through the bed under regeneration during the heating step. I think when we use bone dry air for regeneration, what is the need for heating it and wasting energy when cold dry air can pick up the moisture itself. I have done small experiment in our air separation unit when it was under shut down. I deliberately switched off the heaters during the heating step of regneration and the regeneration was completed ( 8 hr cycle ) with cold dry air. After completing the regeneration with cold dry air, it was taken in line and we got the same dew point as we were getting during normal operation. With this we were sure that activated alumina was getting regenerated perfectly with cold dry waste N2, but what we were not sure of was whether molecular seives ( for removal of co2, H2s, HCs etc ) was getting regenerated perfectly or not. Hence still we are heating the waste N2 to 175 deg C during the heating step of regn. cycle.

In our Compressed Air Plant (CAP), we use wet air for regeneration. During drying, air flow is from top to bottom as it should be. Its 8 hr. cycle. 5hr for heating and 3 hr. for cooling. As you told, last step of regeneration i.e., cooling can be done only in cocurrent direction ( top to bottom ) in order to avoid moisture spike in the o/l air. But heating can be done in counter-current direction. Here we do in the same way. Heating for 5hr is done in counter current direction and cooling for 3hr is done in cocurrent direction.
Regeneration is done under pressure, as I told you earlier.
Here we plan to do the regeneration of beds under depressurised condition using bone dry waste N2 from air separation plant which we are throwing out now. As we plan to use dry gas, we need not heat it up during heating step of regeneration. In this way we can save on electrical energy during heating step. I think this will work. We have submitted the modification proposal. Will it work sir? It should work thoeretically and I saw it practically in air separation unit.
Thanks
sskumar

#7 Art Montemayor

Art Montemayor

    Gold Member

  • Admin
  • 5,782 posts

Posted 25 February 2005 - 02:13 PM

sskumar:

What you are proposing is a clever, unique and cost-saving process change. If you indeed have waste gaseous, LP Nitrogen to spare then what you propose is the smart thing to do if you can justify the process change capital costs with the operating cost reduction and product air quality improvements.

I would incorporate the regeneration changes as follows:

1) The LP N2 waste, hot and cooling regen gas should be piped in COUNTER-CURRENT flow to that of the drying air;
2) Although you don't require the really hot regen temperatures with the bone-dry, LP regen N2, I would not use a regen temperature any lower than 300 oF;
3) Depending on the water loading and the TEMA cycle you are using, you will probably require a lowered heat-up time; this should be calculated and factored in with a contingency factor.
4) Vent the waste, regen N2 out to the atmosphere after it has done its regen job.


Another cost saving point to consider is that if you are not relying on the Oxygen content of the compressed air for process use, you could also use the waste, LP, bone-dry N2 as the feed to the air compressor and simply use the compressed, dry, N2 as the process fluid. This is often done when you are using the Process compressed gas (air or N2) as a pressure medium to drive pneumatic valves and instruments. If you are supplying the compressed air because of its inherent Oxygen content, then this scheme won't work for you. This is just another thought.

Good Luck.

Art Montemayor
Spring, TX

#8 sskumar

sskumar

    Junior Member

  • Members
  • 17 posts

Posted 25 February 2005 - 11:14 PM

Art Montemayor:

Thank you for your complements and thank you for your suggestions.

We thought of connecting LP N2 gaseous waste to air compressor's suction pipe as the compressor discharge air is used only for the operation of instruments and control valves and shut down valves. (O2 is not important )
This will reduce the moisture loading on the dryers and we can keep the dryer in service for more time. We have limited supply of this waste LP N2. Hence we want to use it for regeneration of Compressed Air Plant (CAP ) dryers only. This will help us in cost savings in two ways:

In our CAP, from compressors' discharge, air goes to a wet air receiver. From here air goes to a distributor where 1/3rd air goes to dryer under regeneration and remaining air will be joined by the regenerated air (after passing through heater ,dryer under regeneration and cooler ) before going to the dryer under service. In this way, no air will be wasted. But what is done is we are dropping the pressure of air by 0.5 atm across the distributor deliberately in order to join the regenerated air with the air stream going to dryer under service. This 0.5 atm drop amounts to huge loss when air flow rates are huge. By using LP waste N2, we can remove this distributor and thereby we can avoid this 0.5 atm drop. And also it helps in energy savings during heating step.
Thanks once again
sskumar

#9 Art Montemayor

Art Montemayor

    Gold Member

  • Admin
  • 5,782 posts

Posted 27 February 2005 - 01:21 PM

sskumar:

I am now back at my residence, after being out for over 5 weeks. I have some further advisements for you (and other Forum members) and add them to this thread. These come as a result of the discussion on substituting Low Pressure, waste Nitrogen from an Air Separation Plant for atmospheric air feed to a compressor, and using this fluid as the utility means to activate pneumatic valves and pneumatic instrumentation.

The first advisement is a strong PRECAUTION. Frequently (and I hope, more often) some of us senior forum members contribute our field experiences as examples and proof of what can often occur out in the real, engineering environment. When it comes to safety issues, I wish this were done more often. As graduate Chemical Engineers, we all know the attractiveness, the simplicity, and the economics of substituting available, bone-dry Nitrogen for atmospheric air as a pneumatic fluid for activating valves and instrumentation. However, when thinking as a TOTAL engineer, bear in mind that although this may seem as a benign and totally, theoretically safe application (the N2 is inert and totally dry - negating any fire hazard or corrosion), you may have potentially serious and fatal scenarios developed if you do not take the precaution of thoroughly identifying the fluid throughout its application route and also thoroughly training your personnel (including outside contractors). It was my experience in Geel, Belgium (where I built a grass-roots Furfuryl Alcohol plant) that we had a late delivery on the planned instrument air compressor. To avert a start-up delay, our contractor employed Nitrogen from our on-site source as the pneumatic fluid. This worked very well, allowing me a record start-up and immediate cash flow. What everyone forgot was that a lot of our pneumatic instrumentation was routed to our control room - which was remote and pressurized with double-doors, due to the presence of a Hydrogen Reformer on the site. During the second week of operation, I noticed some operators complaining of headaches and having to step outside the control room for refreshment. Suspecting something, I had an analysis done of the atmosphere and found our Oxygen content to be reduced down to approx. 15%. At this point, I immediately halted all operations and started an investigation; the result was that I discovered pneumatic instruments within our control room had been exhausting their N2 fluid within the captive and sealed confines of the operations Control Room. This had been a disaster waiting only on time to take its effect. This happened in 1970. Since that time, many other similar incidents have happened throughout the world - some involving control rooms, others involving "breathing air" hosed into confined spaces during maintenance inspections and vessel entries - and most of these have resulted in human fatalities due to asphyxiation. Nitrogen is a deadly and silent killer and all engineers - especially Chem E's - should be well versed in its dangers and hazards due to its use or "misuse". By misuse, I mean that I have often experienced process plant personnel employing Nitrogen source as a "purge" hose line to a vessel and then have people enter the confined space. If the N2 is not clearly and efficiently identified with colors and signs (as well as special valves and fittings - such as left-hand threads), then a fatal mistake can occur out in the field. My advice is that you should fore-warn all personnel involved with regards to this "tradeoff" involving the use of inert, "benign" Nitrogen. The tradeoff is a hard and fatal one and should be not only understood thoroughly, but it should be heeded 24-hrs/day.

My second point is that related to your statement that you can successfully regenerate adsorbers by simply flowing bone-dry gas through the spent adsorbent bed. This is theoretically true and can also work empirically - however, the time required for total bed regeneration is going to be very extensive and will (or may not) suit your TEMA cycle time(s).

You may feel that you successfully did this type of regeneration on your shut-down Air Separator dryers, but I would add that you did this on adsorbers that have a generous contingent adsorber charge. An Air Separator cannot tolerate a "spike" or a moisture content "break-through" in the feed air without causing you an immediate clogging and plug-up within the cold box piping and valves - leading to a shutdown and subsequent plant defrosting. You normally cannot tolerate an un-planned Air Separator shutdown; it is usually far too critical as a source of required utilities: Oxygen and Nitrogen. What I believe leads you to believe that you "succeeded" in regenerating the dryer with dry gas is that you have a down stream "safety net" or "parachute" in the form of the Mol Sieve adsorber - which also has sorptivity towards the mosisture. If your TEMA cycle is 8-hrs, then I suspect your Air Separator is not a large, tonnage one and your adsorbers have a generous contingent adsorbent charge - both in Activated Alumina and Mol Sieves.

Adsorption depends on van der Waals forces that act on the surface of the adsorbent and hold the contaminants (moisture, in this case) on the surface. Regeneration is required with de-sorption energy which also includes the means that any liquid on the adsorbents surface has to be converted into a vapor in order to evacuate it out during regeneration. If you have a bed temperature indicator and recorder, you will note that once you heat the bed up to approx. 220 oF, the temperature remains constant - indicating latent heat addition. After an appropriate time goes by, the bed temperature rapidly climbs according to the regen gas temperature input value and rate. This is why, as I recommended, you should regenerate no lower than 300 oF (when using LP regen gas). For bed regeneration to be thorough and efficient, you must thoroughly drive out all the bed contaminants (with the exception of some compressor oils that get into the bed and remain as permanent de-activators) with energy (heat) and subsequently cool down for the next, sequential drying period.

I hope this experience is of value to you.

Art Montemayor
Spring, TX

#10 Guest_Guest_*

Guest_Guest_*
  • guestGuests

Posted 01 March 2005 - 09:20 AM

Art Montemayor:

Regarding your first advisement : Yes, that danger is there in case of pneumatic instrumentation and in case of electronic instrumentation, that danger is not there if we use N2 in the place of instrument air. What I was referring to was LP waste N2 which consists of 30% O2 and 70% N2. Hence danger of asphyxiation is altogether not there. Yes, what you told is absolutely correct, N2 and air lines should be clearly distinguished. In our plant also there was near miss fatal accident happened because of confusion with air and N2 lines. For vessel entry permits, we keep air purging on (this is wet air, not instrument air which is dry ). We call this air as plant air. This plant air header is colored white whereas N2 header is white with yellow bands. The given plant air hose was removed by somebody for some reason. Before entering into the vessel, maintenance staff gave the hose connection, by mistake to N2 header thining it as Plant air header. One contract person entered the vessel and fell unconscious. To rescue him, another contract person entered without PPE, he was also overcome. Departmental supervisor became panic and he also entered without PPE and he also became unconscious. Last contract person there before entering the vessel, informed the control room over PA system about the incident and then he also entered the vessel and he was also overcome by N2. Because of timely action by rescue team, all were rescued in time and all survived. Then onwards we started using metallic red tags marked "danger" on N2 hose station points.

Regarding second advisement:

Yes, you may be correct. Molecular Seives may be taking care of the leaked out moisture from activated alumina and thatswhy we were getting the same dew point with cold bone dry air regeneration. We have to repeat the experiment with Compressed Air Plant(CAP) dryer which consists of activated alumina only and see. After some time moisture may spike in the outlet air. We will carry out the experiment in Compressed Air Plant with dry LP waste N2 and see

Thanks once again for your invaluable advice.
Regards
sskumar




Similar Topics