Perforated Pipe Distributor Design for diffusion of gas in a aqueous solution

[Luisa89]

Hi everyone.
please, I need some help for a project I am working on. My collegues and I are designing a pharmaceutical vessel for the preparation of injectable solutions. Now we need to design a gas sparger (in particular, Nitrogen sparger). The only data we have are the gas flow rate (25 Nm3/h), the nitrogen conditions (Patm and T=20 °C) and size of the pipe feeding the nitrogen (3/4"). Our unknown values are the length of the sparger, the number of holes and the diameter of holes. I have searched everywhere, first of all in Perry's, but I have found only the formula for DeltaP calculation through a perforated-pipe.
Please, could you help me for solving this problem and optmize the design of the sparger?

Thanks for your kindness and your help.
Luisa

 

[Latexman]

Clean or dirty service?

Good Luck,
Latexman

 

[Luisa89]

Dear Latexman,
clean service. Nitrogen is used to deoxygenate the aqueous solution during its preparation. I have found in Internet this procedure

https://www.cheresources.com/invisi...forated-pipe-distributor-sizing-calculations/

, but dimensionally I think something is wrong.

Thanks for your help. So grateful for your kindness and help.

Luisa

 

[pierreick]

Hi,
Consider this brochure , you may contact this vendor .
Good luck
Pierre

 

[Luisa89]

Dear pierreick,
thanks for the brochure. Anyway, I would need a procedure or a formula for dimensioning this kind of sparger.
We have just seen on the vendor website. Something useful is written, but at the end it is necessary to decide the sparger configuration according to Mott sparger types.

Thanks again,
Luisa

 

[Latexman]

So, stripping out dissolved O2?

Being Pharma and clean service, I'd think some over design would not break the budget. Stack the deck for success, kind of thing. Have you looked into sintered metal tubes? Those give smaller bubbles than holes in tubes/pipes.

If that's not in the cards, use Search near top, left between Forum and FAQs with "maldistribution" or "distributor pipe" as keywords to find old threads. Open the search up to the whole site. Some of the good threads are not in this forum. A design that has < 5% maldistribution is usually close enough.



Good Luck,
Latexman

 

[EdStainless]

To strip oxygen you need large bubbles. It is the concentration gradient that drives the stripping.
If you were injecting a gaseous reactant you would want very fine bubbles to assure complete reaction.
You N2 pressure isn't atmospheric, unless the tank is under vacuum.
The needs to be some delta P.
The hole trick to these is to keep the holes small enough so that there is a little back pressure, keeping the pressure inside the manifold roughly equal along its full length. But making the holes large enough to maximize flow.
In many services this is done with a series of nozzles around the wall near the bottom of the vessel.
That keeps all of your supply piping outside the process.
It also makes it easier to build a ring header to help keep the pressure uniform.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[LittleInch]

Luisa89,

You seem to be doing this the wrong way around. As Ed says, first define your process conditions and bubble sizes.

Your gas pressure will be more than atmospheric to force the gas through the holes and into the liquid which is at ?? pressure / height above the sparger.

Sparger design is a complex thing and not easily "calculated".

I suspect your process conditions of 20C and Patm simply refer to the volume flow of the nitrogen.

What size of tank are you sparging into?

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Compositepro]

Bubbling N2 through a fluid is a slow and crude way to remove dissolved O2. It relies on diffusion though the liquid and the gas, interfacial contact area, and gross mixing. Diffusion is a relatively slow process.

Boiling the liquid using heat or a vacuum will almost instantaneously remove dissolved gasses from liquids.

 

[Latexman]

The "length of sparger" and shape of sparger should be made to mostly fill in the superficial area of the vessel/tank so the rising bubbles can contact the majority of the superficial volume of the liquid.

It's a surface area of bubble(s) to volume of liquid thing. So ask yourself - 1 big bubble or a gazillion little bubbles?

If it's batch, the time domain can be used too. Just purge it longer to get less O2.

As LI says, More details = better answers. You really haven't given us much. Will heat or vacuum affect the product? How big is the vessel? Cotinuous or batch? Lot of unknowns.

Good Luck,
Latexman

 

[Luisa89]

Dear all,
I ask you pardon for my inadequate information. This is my fist approch to a pure engineering design. For this reason, I ask you a little piece of patience, if I am not very clear during my explanation. Please, I try to write you all the details I have. You can find a P&ID detail in attachment referring to the preparation tank I have spoken. @Compositepro clearly explain the rule of the nitrogen for the pharmaceutical process. I am sorry, but previously I was wrong. Nitrogen pressure inlet is 3 barg. The line at 50 mbarg is used for the inertization of the tank. Unfortunately, I have no more information.
You can see the idea of shape of the sparger from the P&ID detail. It must be an horizontal and linear perforated-pipe.
Please, write me all the information you need. I will try to reply you in the best way I can.

Anyway, if I consider the formula for the calculation of flow through an orifice (see 6-22 of Perrys')is an error? And what do you think about the procedure in the link?
Sorry again for my inexperience.
So grateful for all your help and kindness.

Luisa

 

[Luisa89]

Please, try now and refer to N2 lines.

Luisa

 

[Latexman]

In the 3/4" Sch. 40 pipe at 25 Nm3/hr (at 4 bara and 20 C) the velocity is << Mach 0.06, therefore it's incompressible flow.

For a sparger:
[ul]
[li]The orifice velocity should be greater than 9 m/s at the minimum flow rate.[/li]
[li]A large number of small orifices are needed to achieve high heat or mass transfer rates.[/li]
[li]Spargers are particularly susceptible to vibration. They should be constructed of sturdy materials and anchored securely.[/li]
[li]Orifice holes to be arranged opposite of one another to balance the forces (thrust).[/li]
[li]Typically, superficial velocities in bubbling contactors are in the range of 0.03 to 0.6 m/s.[/li]
[/ul]

What is the depth and density of liquid over the sparger? I'd guess about 1 m depth and s.g. = 1.0?
What is the smallest orifice you can drill into the pipe? Too small and the drill bit breaks frequently.

Good Luck,
Latexman

 

[pierreick]

Hi Luisa,
The document underneath should help you in a better way.
Good luck.
Pierre

 

[Luisa89]

Dear all,
in detail,
- Our perforated-pipe distributor, or sparger, must have pharmaceutical charateristics. So it must be in AISI 316L stainless steel, not in carbon steel.
- The depth must be defined. Regarding this argument, please what are the parameters to be considered for having the optimal installation? Probably, it seems to be under the magnetic stirrer, so that mixing contributes to nitrogen diffusion.
- The density of the product can be assumed similar to the water. The product is an aqueous solution, so we can consider a density of 1000 kg/m3.
-The dimension of the holes/orifice is one of the unknown value. Looking for information in Internet, I have read the minimum hole diameter is 1/2" (13 mm). Anyway, it could be calculated through DeltaP formula, but what is the optimal DeltaP to be considered? Please, see this link:

https://www.tlv.com/global/ME/calculator/air-flow-rate-through-orifice.html?advanced=on

. In your opinion, formulas are correct?

Thanks again for your help.
L

 

[pierreick]

Hi,
Another document to support your work.
Good Luck
Pierre

 

[Luisa89]

Dear Pierre,
thank you for the document you have sent to me. It is very insteresting. The only formula I have to use is (5), but I have n°2 unknown values (the diameter of the single hole, D0, and the number of the holes, N). It is necessary to fix at least one of the unknown values, right? Please, how I could choose the adequate D0 or N?

Thank you again,
Luisa

 

[pierreick]

Hi,
I think it's more less based on experience , take a look at the example (last sentence) .
I believe that you need to fix the number of holes ,calculate the D0 max , choose a D0 # 1/2 D0 max , test on field , basically error and trial .
hope this is helping you a bit

Pierre

 

[Luisa89]

Dear Pierre,
if I consider the following formulas:
- DeltaPp=pressure drop in pipe caused by friction=alfa*rho*Vp^2/2, see formula (1), and I explicit Vp as a function of the volumetric flow rate Qp, what I obtain is = alfa*rho*(Qp/Ap)^2/2, where Qp=know value (25Nm3/h to convert in Nm3/s), Ap=pi*Dp^2/4, alfa=1,05 for turbolent regime, rho=density of nitrogen. I have all know value, so I can calculate DeltaPp.

- DeltaPo=2,6*rho*Vo^2/2*[1-(Ao/Ap)^2], see formula (3), and I explicit Vo as a function of the volumetric flow rate Qo, what I obtain is = 2,6*rho*(Qo/Ao)^2/2*[1-(Ao/Ap)^2]; but if I consider the fluid as incompressible fluid, Qo=Qp, so I have = 2,6*rho*(Qp/Ao)^2/2*[1-(Ao/Ap)^2].

- If I impose DeltaPp/DeltaPo <= 0,1, the only unknown value is Ao.

- Ao is = N*pi*Do^2/4, so I have still n°2 unknow value (N and Do).

- If I impose a system of two equations with two unknowns, Ao = N*pi*Do^2/4 and Do < 0,7 Dp/sqrt N (see formula (5)), is this the right way to solve the problem?

Thank you so much.

Luisa

 

[Luisa89]

At the end, I have found that 1/N <= 0,7^2 * Dp^2 *pi^2 / (4*Ao).
Hope this is the right way to solve.
Plaese, give me your feedback, if you want.

Thank for all,
Luisa