10 replies to this topic

[dodong]

Dear All,

I would like to seek for help in calculating the amount of silica gel that we shall place in a humidifier.

The air entering the dehumidifier is chilled, and saturated at -6 C and 1 bar. It passes through a dehumidifier, with Silica gel of specific weight .7012 kg/L. The expected dew point of air outside the bed is -60C and a dry bulb temperature of 20 C. To calculate the amount of silica gel, what are the iformations that I need to obtain and how do I go about it?

I intend to use one bed in 12 hours while the second bed will undergo regeneration in an 8 hour cycle. The brochure from our supplier indicates that the Silica can absorb 22 kg of water per 100 kg silica, but I have to indicate the temperature of bed. Which temperature do I consider? the outlet temperature of the dry air which is the dry bulb?

Thanks,
Dodong

[Art Montemayor]

Dodong:

I recommend that you contact the supplier of your Silica Gel and get some factual and real information from him/her.

You will be extraordinarily lucky to realize a sorptive capacity of 22/100 = 22% water adsorbed per mass of dry adsorbent. I will bet you all my Social Security benefits that you will never achieve this high of water adsorption after you have regenerated the Silica Gel in situ and are running normally. You will be fortunate if you achieve 8-10 kg of water adsorbed on 100 kg of regenerated Silica Gel. And this is assuming that you regenerate your bed to at least 350 ^oF. That has been my experience with a lot of Silica Gels.

What you need to find out is the practical, dynamic sorptive capacity of the Silica Gel after it has been regenerated at least 100 times. That is the real-life design capacity that you need to know. And even that figure should have a contingency factor added on to it.

It is very important that you also obtain the recommended bed temperature recommended for what your supplier defines as optimum regeneration status. You should reach this bed temperature (at the exit of the regeneration gas) and hold it for at least 2 hours before starting your cooling cycle.

I assume you have used the “Search” function on this Forum and have research and read all the threads where we discuss adsorption and adsorption dryers. If you have, you will find out that some very important factors in adsorption drying are: Type and quality of regeneration gas, pressure and temperature of regeneration cycle, superficial velocities of adsorption and regeneration, proper bed supports and tie-downs, etc., etc.

I assume that you are well-trained in unit operations and are knowledgeable about Adsorption. Note that you have used the erroneous spelling “absorb”. I assume you made a typo error and know the difference between these two words. If you don’t, you shouldn’t be involved in designing an adsorption unit.

I understand you to state that you intend to feed your adsorption dryer with chilled air at -6 ^oC. This will be of great help in reducing the amount of adsorbent required to produce what you say will be air with a dew point of -60 ^oC (I don’t know what you mean by a dry bulb temperature of 20 ^oC). The dew point is normally referred to atmospheric pressure.

You state you have to indicate the temperature of bed. What temperature do you mean – the adsorption average temperature, or the maximum regeneration temperature? Please refer to the thread “Replacing Desiccant Air Dryers” posted some days ago and my response and attached adsorption sketch.

[dodong]

Dear Art,

Thank you for your response.

Our objective is to dry our process air. Our target is to attain an air after the dryer of -60 C dew point. What we are getting is -48 C. So we are suspecting leaks prior to the Adsorber.

This is our current system: Our process air coming in (from blower) is 100C. We cooled it down up to 44C through a heat exchanger, then further cool it through 2 heat exchangers, using cooling water and glycol. At glycol heat exchanger, we can bring the process air at 6C. Im sorry it is not -6C as mentioned in my previous thread. At this point we already condensed water. This air (we call it chilled air) will directly go to adsorber (not absorber).

Our problem is this: we are not attaining the dew point of -60. Possible reasons are:
1.) Leaks in heat exchangers
2.) Silica Gel is poisoned
3.) Silica Gel reached its life
4.) Leaks in valve (during regeneration atmospheric air comes in during cooling process)
5.) Rgeneration temperature is not attained

we have checked items 1,4,5. No leaks whatsoever.

We are looking at the silica gel we are using. I am attaching the picture of Silica that I have obtained. It is very odd to have this silica color. Do you have any experience with this discoloration? Please see attached.

Then we want to replace SILICA Gel, that is why I want to know how do I obtain the right amount of Silica I should be placing.

Thank you Art, this is a very great help for us.

Dodong

Attached Files

•  Silica_Gel.jpg   805.34KB   65 downloads

[Art Montemayor]

Dodang:

Thank you for the timely and corrected data and information – as well as the photo of the contaminated Silica Gel. Your submission of the photo is great because it is the next best thing to being there.

You have not answered all my questions and inquiries, but the information is now clearer. A sketch like the one I attach would have been helpful. Additional information is now more revealing and the answer to your dilemma is getting closer. Your application, in my opinion is very critical. Instrument air is not as critical as process air requirements due to the higher quality specifications on the latter. You must resolve this issue because it is probably causing high operating costs or losses.

You left out a lot of background information:

• Type of adsorption unit (2-tower, 3-tower, etc.) and length of NEMA cycle;
• Operating pressure in the adsorber;
• Flow rates and pressures of feed air and regeneration gas;
• Type of regeneration system (dry gas or wet gas? Cool down?);
• Flow direction for the feed air and regen gas within the adsorbent bed;
• Temperature and pressure of regeneration gas and duration of regen cycle;
• Size (diam. & height) of adsorption beds and quantity of adsorbent in each;
• Manufacturer of adsorption unit and design capacity;
• Service life of adsorption unit;
• Type of feed air blower and whether it is oil lubricated;

Nevertheless, thanks to your photo I believe you’ve hit close to one of the major causes for your unit’s lack of performance. I suspect you are using a positive-displacement type of air blower (possibly a “Roots-Connersville” type) to generate around 5- 15 psig air. In doing so, I believe the blower is either being oil-lubricated in the compression chamber or an oil seal has ruptured and is leaking timing gear lubricating oil into the compression chamber and contaminating your air. I suspect that it is this contamination that is making itself present in your silica gel adsorbent bed. You have contaminated (“poisoned”) silica gel adsorbent in your dryers and it is obviously coming in with the air – or possibly with the regen gas stream. Your silica gel should not be discolored or darkened in the manner that it shows up in your photo. You have obviously lost sorptive capacity due to this poisoning of your adsorbent. This is costing money and time and has to be resolved quickly and effectively. The fact that you now report that you are not achieving the design regeneration gas temperature is another negative factor that prevents you from obtaining the maximum available sorptive capacity from your silica gel beds.

I don’t know if you can produce (or have ever produced) the desired dew point with the silica gel adsorbent in your present beds; however, if properly designed and operated, your system should not give you any problems whatsoever. The type of regeneration system and the regen temperatures achieved and sustained during regeneration are important and, if possible, you should sketch the system out much as I have in the attached workbook. Do not expect to achieve 22% sorptive capacity (22 kg of water adsorbed on 100 kg of silica gel) with a dew point of -60 ^oF. This is just not practically possible with silica gel (or with any other known industrial adsorbent). I still don’t know your adsorption process conditions, but normally I would expect a design sorptive capacity of approximately 8% (maximum) for silica gel for a period of up to 2 years of operation. This, of course, is assuming that there is absolutely no contamination or poisoning of the adsorbent. You can (& should) check out my capacity comments with your silica gel manufacturer. If they are knowledgeable and honest with you I am sure their opinions will concur with my comments.

Replacing your silica gel isn’t going to resolve your present problem. It will only compound it with other costs. You will continue to contaminate your adsorbent until you get to the root cause of that problem. You should also get to the root of why you aren’t achieving the design regeneration temperatures (approximately 350 to 500 ^oF). Without identifying the cause of these two problems, you will continue to suffer inefficiencies and the related higher production costs and/or losses. I strongly recommend you investigate your blower and any other possible source of your contaminated air feed as well as your regeneration heater problem. I would normally recommend you employ Activated Alumina as your adsorbent rather than Silica Gel; however, this is not the time to switch types of adsorbent. First, you must find the cause(s) that are behind your present bad drying operation. When you have resolved them, then you can think of improving your present operation and its cost efficiency.

Let us know if we can help – but furnish us ALL of the basic data and background information available. Otherwise, you force us all to guess at what is happening and that is not being effective.
 Silica_Gel_Adsorption_Dryer.xls   3.25MB   173 downloads

[dodong]

Dear Art,

Yes, this problem is giving us headaches.

Please see attached file for background of the system.

1.) Our dehumidifier is composed of 2 beds, just like other beds, one is being used and at the same time the other bed is being regenerated. Our regeneration cycle is an 8-hour cylce. 5 hours heating of Silica, 3 hours cooling. In heating, we use Steam at 6 bars, so the maximum temperature of heating reaches only up to 130 C.

2.) Our adsorber is 0.4 - 0.6 bars. Outlet pressure of 0.5 bar currently.

3.) Air pressure after process blower is 0.65 bar. After glycol heat exchanger, the temperature of chilled air is 6 C. This is 6930 kg/hr of chilled air. For regeneration gas, there is no measurement of flow for that. This is atmospheric air, heated with steam.

4.) The regeneration process is this: We heat atmosheric air with steam. Cooling process is a closed loop. We cool the atmospheric air(same air used in heating) but closed loop, with heat exchanger (cooling water medium).

5.) Flow direction of air, from top to bottom, set up of bed is layer. With 200 kg activated alumina on top, then silica at the bottom.

6.) The regeneration gas (I assume its the gas that comes in contact with the silica) - pressure is not measured, but temperature reaches up to 130 C due to steam.

7.) geometric capacity of 30540 L. I do not have the dimension of the bed with me. Its been running for 27 years already. But silica use: we sieved last year and added new silica to compensate for the undersize. Manufacturer: Ballestra. Current silica gel volume is 2200 kg. If we want to change silica (since its poisoned), how do i compute the amount of silica needed? I am contacting our supplier but doesnt get response yet.

8. Feed blower is a positive displacement, screw type. Robuschi.

We have tested the silica and it gave us an acidic solution. We were suspecting its poisoned with SO2 (so the acid is sulfuric), however qualitative analysis doesnt give presence of Sulfate (no precipitate),or maybe the concentration is too small.
. Air is used to burn sulfur, we suspect there was a backflow happened sometime before the problem occured that poisoned our silica.

Yes, we used to get the dew point of -65 C.

Please also see attached excel, our graph of regeneration, upon reaching 130 C we then cooled our silica, which doesnt give us confidence that the bed is fully regenerated. BUT this is a chicken and egg problem, where we cannot use anymore the running bed because it is giving us high moisture content of air. We have to shift to the next bed.



Many Thanks Art,

DODONG

Attached Files

•  Bed_Performance_Monitoring.xls   59.5KB   82 downloads

[Art Montemayor]

DODONG:

Please always label your pressure units as absolute (bara) or gauge (barg). Otherwise, there is a confusion created when a need arises to identify the precise pressure condition. An example is your steam supply temperature. If you have 6 bara steam supply, the steam temperature is 158.8 ^oC; if it is 6 barg, the temperature is 165 ^oC.

You state “adsorber is 0.4 - 0.6 bars. Outlet pressure of 0.5 bar”. I have to assume you mean gauge pressure. You also say “Air pressure after process blower is 0.65 bar”. If you are being consistent in your terms, then your steam pressure is 165 ^oC and you should be able to obtain regeneration air at 155 ^oC very easily. However, you only obtain “up to 130 ^oC”. You should try to maximize your regen temperature – I would strive for 175 ^oC (this would require you employ an electric heater instead of a steam heater).

The quality of your flow diagram sketches photos are just the opposite of the Silica Gel photos you previously attached and of the “Regeneration Temperature Profile” graph you attach here. Since I am unable to make out or read your flow diagram(s), I am forced to generate my own – based on your description. It is important to read every detail in the diagrams because there are a lot of important factors that make a large difference in the performance of an adsorption unit. For example, now you are revealing that there are actually two different type of beds in each adsorber vessel – one top bed of Activated Alumina and a bottom bed of Silica Gel. This makes a big difference. You also now say that your regeneration cycle is a closed loop. In order to establish a closed loop you must have a driving force to recirculate the regeneration air. This is often a blower. Are you using another blower for this duty? Another blower means another potential source of oil to contaminate the beds.

The Robuschi blower you are using seems to be a type very much like the one I’ve guessed you were using. It seems to have only one oil seal between the timing gear section and compression section – something I am very suspicious about.

I am assuming you have vertical, cylindrical adsorber vessels. If so, the quantity of adsorbent in each vessel is the internal diameter of the cylindrical bed squared multiplied by (Pi/4) and by the height of the bed. That is why I asked you for the internal diameter and the height of the bed. Now we find out that you actually have TWO beds in each vessel. We need the height of each bed - both Activated Alumina and Silica Gel.

You tell me the quantities of “Current silica gel” (2,200 kg) but you fail to say if that is for each vessel or for both vessels. The same applies to the 200 kg of Activated Alumina.

[dodong]

Dear Art,

1.) Yes, the pressure I mentioned is in terms of barg. Yes, i also have the same question why we only reached 130 C instead of 150 C we are using 6 barg steam.

2.) Closed loop, yes we also have the another blower to drive the regeneration air into the beds. We are checking on possible leaks, thank you for your inputs. This gives us idea, how critical this is.

3.) We have 1 vessel. Vertical vessel containing 2 beds. One bed is regenerated while the other is being used. Each bed contains 2200 kg Silica. + 200 kg Alumina. Internal Diameter of each bed is 2.2 m, with the height of 0.81 m.


Thanks,
dodong

[dodong]

Dear Art,

Good News! We were able to pin point he source of our problem. The silica gel was indeed poisoned and at the same time very inefficient due to smaller granules.

When tested into solution, it gave us an acidic solution of pH 2. Our team indentified the source, a backflow of process air containing SO2 and SO3 during power failure/shutdown. Also, a change happened in the plant 6 months ago where hot air reaching to 200 C is used in regenerating silica, this we identified made the silica smaller than the expectd size.


Currently, we are looking at possible corrossion in the system due to oleum generation casued by high moisture in process air combined with SO3.


Thanks,
Dodong

[Mazfar]

We are also facing problems not meeting the dew point in our dryers.
The cooler temperature even after chemical cleaning is > 50 C.
I beleive its loading the dryers and not removing the moisture.
The cooler is a high performance tubes where area credit of 4 times was taken.The actual tube (finned) area is 1/4th the required for a normal cu-ni tubes.

[srtools1980y]

hi!

all of u using silica gel:
don't believe the supplier. as soon as u receive new supply, get it analysed & evaluated by a reputed lab.

we do this regularly for activated carbon.

goodday
srt.

[Art Montemayor]

Mazfar:

I was prompted to this thread because I’m one of the original respondents. Now I find that you are having a problem with dryers – but you don’t seem to have adsorbers. Also, srtools1980y is alerting us all to phony and flawed silica gel out in the marketplace – which is good advice, but we don’t seem to have an adsorber with silica gel in your application.

This thread was started on the basis of problems with a fixed-bed adsorption type of dryer. If you don’t have that type, please tell us ALL of your basic data – including the type of dryer, of course.

The only cooler an adsorption dryer has is the one used if you recirculate process gas as regeneration gas which has to be cooled after exiting the adsorption bed and rejoining the inlet stream. Is this your case? Even if it is, this cooler has little or nothing to do with the performance of the adsorption beds. Please give us all of the basic data requested in order to help you out. Also, generate and submit an accurate and legible schematic drawing of your system. This doesn’t have to be a P&ID, complete with all instrumentation and piping shown. It can be a simple PFD (Process Flow Diagram). This will avoid a lot of confusion and non-applicable advice and comments that only lead to further confusion and frustration in trying to understand just exactly what you are confronting.

From what little basic data you offer, I suspect you are using a refrigeration type of dryer – which doesn’t remove the moisture directly as an adsorption type. The refrigeration type merely cools the wet gas stream close to the freezing point of water (2-3 ^oC) and drains out the condensed water from the chilled stream. The same chilled stream is subsequently warmed up to original process conditions by using a heat exchanger and exchanging heat with the incoming wet gas stream. Is this what you are using?

Please start a new thread for your problem. We can deal and discuss the subject of your problem much simpler and directly if we dedicate one thread and only one thread to it. A chaotic and confusing condition is created when you try to resolve several different problems in one single thread.

I will await your reply in the form of a new thread in the Industrial Professional Forum – complete with basic data and a drawing.