21 replies to this topic

[mikkelpoulsen]

Hi 

I'm working on a design for a gas sparger in a water column, and so far I have not been able to really understand how to design this in correct way.

I have been reading a lot on sparger online (and in Perry hand book edition 7) but i have not fully understand how to get it to work like I want to.

 

I have attached some pictures of my test stand, this has the following specs:

1 water tank ø490mm (inside diameter).

1 inlet pipe for the gas (test stand it is air or nitrogen) ø33,7mm

Water height is 1,5m static water.

Outlet pipe is the same as inlet pipe.

 

The sparger is designed as a hole plate sparger - mounted on a box that is mounted on the inlet pipe and then submerged to the bottom of the tank.

The specs for the plate is a outside diamter of ø470

 

The specs for the gas is the following:

Inlet speed of gas/air is: 15-16m/s

Inlet pressure from the pressure regulator right now is around 1bar

Inlet pressure on the column when running 750lpm is 0,35bar (taken from the manometer)

Volume flow is up to around 50m3/hr or around 750 lpm

I want the outlet speed out of plate in bottom (superficial gas velocity) total out the sparger to be as low as possible, around 1m/s. 

Current plate is 1,64m/s -> this is 741 holes, ø3,5mm evely distrubuted in 13 lines (16 holes inner circle and 98 outer circle) 

 

Here is my problems that I do not know how to calculate on:

• How to get a even distrubution out of all the holes? Right now there is not, i have tried many thing larger holes, different sizes of holes etc. But i'm pretty sure this has something to do with the counter pressure of the sparger plate, and I do not know what i should aim for? So far I have just been testing different plates and patterns to get some understanding of the system. 
  Only calculating on the total area of all the holes. 
  The problem is, that at low volume flow only the inner holes are used, and even at the full load I do not see a even distribution out of all the holes.
• Second problem, I can see that the water level as a impact on the system, when the column is tilting forward must of the gas is coming out of the holes where the water level is lower. (less static pressure on top of these) 

So my question is, does anybody has some advice on how to calculate on such a gas sparger box with a plate on top, what parameters to I need to tweak on? 

Is the box design also something I need to look into? The size of this could have an impact on the amount of water that need to get pushed out before the gas is coming out of all the holes I assume? 

 

I'm trying to get this "small" scale to work, so I can make my larger water tank of a diamter of 1,5m - but right now I do fill confident enough to upscale my system and still believe I can get the distrubution to be on as much surface as possible.

Attached Files

•  Presentation1.jpg   64.35KB   0 downloads

Edited by mikkelpoulsen, 12 September 2023 - 07:59 AM.

[Bobby Strain]

What are you attempting with this operation? You should consider an eductor instead.

 

Bobby

[latexman]

Here is some information I have on pipe distributors:

 

Flow distribution and pressure drop are two primary considerations in design of distributors.

Flow distribution is dependent on Ar, which is the ratio of total perforation or orifice area to pipe

cross sectional area. To a lesser extent, it also depends on the distributor length-to-diameter ratio

(L/D). FIGURE 16 shows the effect of these parameters on maldistribution. It is evident from

the graph that if Ar = 0.5, the distribution is good and above Ar = 1 it is poor. Note that a positive

maldistribution indicates higher orifice velocity and pressure near the end of the distributor.

Pressure drop through a distributor pipe can be assumed to be 2.8 velocity heads based on the

orifice velocity. For Ar > 1, the velocity head should be based on the inlet pipe velocity.

 

 Capture.JPG   60.76KB   1 downloads

 

I would think this information can give some insight here.  It appears you have too much hole area (too many holes and/or holes too large), and you have a high degree of maldistribution.

 

What is your total hole area/feed pipe cross-sectional area?  About 8?

 

Does water leak into the sparger chamber?  Is it a lot?

[Pilesar]

Spargers like this are often needed for bubble column reactors. See this reference for ideas, equations, and experimental discussion: https://www.scienced...385894707001337 (Chemical Engineering Journal, 15 September 2007, Pages 173-186)

Your sieve plate will tend to favor the center in a level vessel due to the column hydraulics. Once the bubbles begin, the mixed fluid is less dense in that vertical column therefore the pressure drop is less so more bubbles flow into that region. Perhaps eliminating the center holes would help -- then your sparger is more like a ring. Consider whether using random packing inside the vessel would suit your needs. The packing should tend to redistribute the bubbles for a more even flow pattern. Consider also Bobby Strain's suggestion of using an eductor. A properly designed eductor would help thoroughly mix the air and water and probably reduce the bubble size as well.

Edited by Pilesar, 12 September 2023 - 12:25 PM.

[breizh]

Hi,

Attached a document about sparger from Mott corporation.

You may want to drop them a mail for your specific application.

Breizh

[mikkelpoulsen]

Hi,

Attached a document about sparger from Mott corporation.

You may want to drop them a mail for your specific application.

Breizh

HI

I have been in contact with these guys - and have a offer for my small scale system from them that we are discussion here.

 

My problem with these porous sparges is the hole size, i tend to think that these will clock. I can not assure that the gas i'm purging into the tank are "clean" 

[mikkelpoulsen]

Spargers like this are often needed for bubble column reactors. See this reference for ideas, equations, and experimental discussion: https://www.scienced...385894707001337 (Chemical Engineering Journal, 15 September 2007, Pages 173-186)

Your sieve plate will tend to favor the center in a level vessel due to the column hydraulics. Once the bubbles begin, the mixed fluid is less dense in that vertical column therefore the pressure drop is less so more bubbles flow into that region. Perhaps eliminating the center holes would help -- then your sparger is more like a ring. Consider whether using random packing inside the vessel would suit your needs. The packing should tend to redistribute the bubbles for a more even flow pattern. Consider also Bobby Strain's suggestion of using an eductor. A properly designed eductor would help thoroughly mix the air and water and probably reduce the bubble size as well.

That makes sense, and is also what i'm seeing, that it is in the middle of the column, especial in low loads. 

I have tried to make one pipe sparger also, with 4 arms - my problem with this is that I can not have enough holes in releation to the area I need to get down into speed.

 

I'm currenty testing with random packing also, and yes this has a influence where it distributes the gas over the area I can see. 

This was also seen when i tested with multiple sieves up in the column. 

 

It is a bubble column kind of system correct, i'm trying to catch ammonia gas in a water tank when it is purged out with nitrogen

Edited by mikkelpoulsen, 13 September 2023 - 07:32 AM.

[mikkelpoulsen]

Here is some information I have on pipe distributors:

 

Flow distribution and pressure drop are two primary considerations in design of distributors.

Flow distribution is dependent on Ar, which is the ratio of total perforation or orifice area to pipe

cross sectional area. To a lesser extent, it also depends on the distributor length-to-diameter ratio

(L/D). FIGURE 16 shows the effect of these parameters on maldistribution. It is evident from

the graph that if Ar = 0.5, the distribution is good and above Ar = 1 it is poor. Note that a positive

maldistribution indicates higher orifice velocity and pressure near the end of the distributor.

Pressure drop through a distributor pipe can be assumed to be 2.8 velocity heads based on the

orifice velocity. For Ar > 1, the velocity head should be based on the inlet pipe velocity.

 

Capture.JPG

 

I would think this information can give some insight here.  It appears you have too much hole area (too many holes and/or holes too large), and you have a high degree of maldistribution.

 

What is your total hole area/feed pipe cross-sectional area?  About 8?

 

Does water leak into the sparger chamber?  Is it a lot?

My area of the tank is:  0,18m2 (ø490 inside diamter) 

My total are of the holes in orifce is: 0,0071m2 on the case I presented in this (i have teste different areas)

There is water in the sparger chamber - beacuse the water is standing still so the chamber is submerged and there is water up in the inlet pipe also.

Then i the water is pushed out of the pipe when the gas is purged into the tank.

I do not know if all water is pushed out of the chamber, maybe that could be a problem? that the chamber is to large?

 

i'm not sure how to read the table.

[mikkelpoulsen]

What are you attempting with this operation? You should consider an eductor instead.

 

Bobby

 

I'm not sure how to use a eductor in this case.

 

The operation is the following:

I have a water tank with water (static no movement) where i then are purging ammonia gas with the use of nitrogen - the idea is to catch the ammonia in the water so only nitrogen is leaving. So i kind of absorbed operaition, and to do this i need to have a good contact between the gas and the water.

[Bobby Strain]

Eductors would be more effective. The best solution is to use a contact column with packing, operating with a continuous vapor phase as suggested.

 

Bobby

[latexman]

My area of the tank is:  0,18m2 (ø490 inside diamter) 

My total are of the holes in orifce is: 0,0071m2 on the case I presented in this (i have teste different areas)

There is water in the sparger chamber - beacuse the water is standing still so the chamber is submerged and there is water up in the inlet pipe also.

Then i the water is pushed out of the pipe when the gas is purged into the tank.

I do not know if all water is pushed out of the chamber, maybe that could be a problem? that the chamber is to large?

i'm not sure how to read the table.

 

 

The 33.7 mm dia. air feed pipe (1" or 1-1/4" ?) is about 0.00089 m2 in cross-sectional area.  So your A_R = 8.  This A_R = 8 is indicative of extremely poor distribution of air in a pipe sparger.  Your application is not 100% apples-to-apples to the correlation I provided, but as I said, I think this information can give some insight here. You have a pipe and holes, there is just a large chamber between the two.

 

I would think the presence of water in the chamber would contribute to maldistribution of air.  Sloshing water will affect the air a lot.

 

Good distribution has an A_R = 0.5, or about 0.000446 m2, or about 46 holes.  However, with only 46 holes your chamber will have more pressure in it than with 741 holes, and you need to be sure it can handle that air pressure safely before proceeding.  If it cannot handle the same air pressure as in the upstream section of that pipe, then you may never be able to get to good air distribution in this column.  Then, listen to the folks suggesting alternative technologies.

[Pilesar]

I don't know your process but suspect there are better ways to scrub ammonia than the path you are taking. You might want to review some environmental emissions control equipment for ideas. The US Environmental Protection Agency has some good resources for an overview. For example, see chapter five of https://airknowledge...udentManual.pdf on absorption.

[breizh]

hi,
To me this is an absorption process, a simple column with packing should do the job. Use your existing tank to collect the ammonia solution to be recirculated or not (based on calculation) somewhere on the top of the column. On very top you need to spray clean water to clean the residual ammonia.

Breizh

[mikkelpoulsen]

 

My area of the tank is:  0,18m2 (ø490 inside diamter) 

My total are of the holes in orifce is: 0,0071m2 on the case I presented in this (i have teste different areas)

There is water in the sparger chamber - beacuse the water is standing still so the chamber is submerged and there is water up in the inlet pipe also.

Then i the water is pushed out of the pipe when the gas is purged into the tank.

I do not know if all water is pushed out of the chamber, maybe that could be a problem? that the chamber is to large?

i'm not sure how to read the table.

 

 

The 33.7 mm dia. air feed pipe (1" or 1-1/4" ?) is about 0.00089 m2 in cross-sectional area.  So your A_R = 8.  This A_R = 8 is indicative of extremely poor distribution of air in a pipe sparger.  Your application is not 100% apples-to-apples to the correlation I provided, but as I said, I think this information can give some insight here. You have a pipe and holes, there is just a large chamber between the two.

 

I would think the presence of water in the chamber would contribute to maldistribution of air.  Sloshing water will affect the air a lot.

 

Good distribution has an A_R = 0.5, or about 0.000446 m2, or about 46 holes.  However, with only 46 holes your chamber will have more pressure in it than with 741 holes, and you need to be sure it can handle that air pressure safely before proceeding.  If it cannot handle the same air pressure as in the upstream section of that pipe, then you may never be able to get to good air distribution in this column.  Then, listen to the folks suggesting alternative technologies.

 

Hmm, yes i understand this calculation now - but ad you say it is difficult maybe to convert it directly to how i have made it.

 

With this calculation of hole area, the total speed out of the holes will actually increase to the double, where my idea and wish (espcecially in a bubble column) is to decrease the speed of the gas. To give it more time in the water and not to spray it out of the tank i want to decrease the velocity.

[mikkelpoulsen]

I don't know your process but suspect there are better ways to scrub ammonia than the path you are taking. You might want to review some environmental emissions control equipment for ideas. The US Environmental Protection Agency has some good resources for an overview. For example, see chapter five of https://airknowledge...udentManual.pdf on absorption.

 

You are right it is not the most effecient way, but the choosen way for other reasons for this project this is how is has to be done.

[mikkelpoulsen]

hi,
To me this is an absorption process, a simple column with packing should do the job. Use your existing tank to collect the ammonia solution to be recirculated or not (based on calculation) somewhere on the top of the column. On very top you need to spray clean water to clean the residual ammonia.

Breizh

 

You are right it is not the most effecient way, i'm fully aware of these absorbers for my project here - i can not use this system. This system has to work without the use of pumping water for other reasons. It will not be the most effecient way, but that is not the goal.

You are right it is not the most effecient way, but the choosen way for other reasons for this project this is how is has to be done.

[latexman]

With this calculation of hole area, the total speed out of the holes will actually increase to the double, where my idea and wish (espcecially in a bubble column) is to decrease the speed of the gas. To give it more time in the water and not to spray it out of the tank i want to decrease the velocity.

 

I understand you are stuck with this column.  Just a thought.  The air distribution is so bad you may be fooling yourself when looking and discussing average velocity through the holes.  Some holes are actually zero velocity, or negative (water getting in).  And, some holes are very high velocity, to make the mass balance work with all the non-bubbling holes.

 

I'm also not sure that low speed of bubble rise should be your objective function in a bubble column.  My recollection and experience points to maximizing bubble surface area to increase mass transfer.  This means lots of small bubbles, and not large bubbles coming out of a few holes.  My $0.02.  Good luck with your project, and enjoy!

[breizh]

Hi,

In addition to my previous reply, let you consider the document attached.

Breizh

 

[Pilesar]

For sieve plate sparger design, Froude number seems key. See the attached.   Design of Sieve Plate Spargers for Bubble Columns - Joshi.pdf   373.33KB   30 downloads

[mikkelpoulsen]

For sieve plate sparger design, Froude number seems key. See the attached.  Design of Sieve Plate Spargers for Bubble Columns - Joshi.pdf

That is very interesting and good article this one. 

I'm not sure exactly if i get all the numbers and what the froude means, but i will try to study it a littel more.

 

But i'm pretty sure that my proplem is weeping now, which means that there is water in my chamber and in the holes and that is properly why my distribution is poor.

but the article also claims in the beginning that the window of controlling the weeping of a bubble column could be very narrow.

[latexman]

But i'm pretty sure that my proplem is weeping now, which means that there is water in my chamber and in the holes and that is properly why my distribution is poor.

but the article also claims in the beginning that the window of controlling the weeping of a bubble column could be very narrow.

 

One of the classic solutions to widen the operating window of a sieve tray is to switch to valve trays.  There are other solutions too.  IIRC, using bubble cap trays is an older, but more expensive, way to widen the operating window of a sieve tray.

[Bobby Strain]

It appears that you have only a small vapor stream, insufficient to properly distribute with a sparger. So, you might consider using a liquid recirculation via an eductor to aspirate the vapor. If you share more, we can be more helpful. Now, we are simply guessing.

 

Bobby