10 replies to this topic

[xmb09129]

Hi folks

 

I have just finished designing a horizontal gravity decanter, the light phase is made up of biodiesel (90%), methanol (4%) and small amounts of other components, the heavy phase is made up of glycerol (55%), methanol (35%) and the rest made up of water, catalyst and soaps. 

The volumetric flow rate of the feed is 18.6 m³/ hour, the light phase is 16m³/hour and heavy phase is 2.6 m³/hour. 

 

I found the average densities of light phase to be 834 kg/m3 and heavy phase at 940 kg/m3, the decanter is at 60^oC and 1 atm. 

 

My real question is that i calculated the residence time to be 1.8 hours which i believe to be rather quick as i got the impression from the internet it would require a lot longer than this.

 

Any advice on what i could of done wrong or if anyone thinks this seems a reasonable time would be very much appreciated.

 

Regards

[breizh]

Let us know how did you get this result , based on experimentation or some calculation .

 

Perry's chemical engineering handbook should be a good resource to consult.

 

Breizh

[xmb09129]

Thank you for your fast response, its much appreciated.

 

I got my results through calculations, i have been using both Perrys and an Applied process design book by E. Ludwig, i first determined what the dispersed phase was and then calculated the settling velocity using stokes law. I then estimated the interface to be just below the feed with its width being 80% of the diameter and L/D ratio = 5. I found the diameter of the cylinder to be 2m from D=(Qc/4.Vs)^0.5 and thus a length of 10m. After this i continued to find the reynolds number were sufficient to expect good separation and the entrained droplet size was at 50micrometers. 

From these calculations i found the residence time by volume of tank / total volumetric flow rate.

 

I am unsure whether or not my calculations are correct and i just have a residence time that is low or if i have made an error.

 

I understand this is very difficult to provide help without actually seeing the calculations but any advice on probable areas i have made errors or even if anyone has an estimate of how long i should be looking at for a separation of this kind.

 

Regards

[breizh]

http://www.koch-glit...s_HP_June09.pdf

 

Consider this resource too . 

 

Breizh 

[xmb09129]

thank you, i will look into this now and see if i can spot what errors i have made.

 

regards

[Bobby Strain]

I suggest that you contact one of the many suppliers of separator internals. You can find some in the links at my website. Just google my name.

 

Bobby Strain

[xmb09129]

Cheers Bobby

 

Your website is excellent, it has a lot of useful links and the online sizer for a separator looks excellent.

 

I wont bother contacting suppliers as this design is only for a project at uni so il keep looking and hopefully determine what my mistake is, the only thing i can think of at the moment is my viscosities must be incorrect, but thanks for directing me to your website, very helpful

 

regards

[Bobby Strain]

You can use one of the rating programs that provide some calculations for water settling. If you adjust the light phase density, you should approximate the settling velocity of the heavy phase.

 

Bobby

[Pingue2008]

Xmb09129,

 

Whether your calculations are right or not the rest of the forum should be able to see what you have done and comment it. we are all learning. i am also glad you brought up this topic. I was working on that a couple of months ago before  we change our mind about using a decanter. so the attached is fairly complete. you can enter the true viscosities on and see what you get. Notice that have used your flow and densities and L/D ratio on the second worksheet and got a slightly differend diameter and Height (notice that this a vertical not hozontal decanter so this might not be of any use for you). did you consider both layouts?. The comments of others on the attached spreadsheet will really be appreciated for, this can help  later on.

 

Thank you,

Attached Files

•  Sizing decanters.xlsx   45.05KB   425 downloads

Edited by Pingue2008, 18 March 2013 - 09:35 AM.

[Bobby Strain]

So, did you complete your exercise? If you provide the viscosity for each fluid, we can comment on your work.

 

Bobby

[xmb09129]

Pingue2008

 

Thank you for getting back to me, i found your excel sheet very useful. I did actually originally design a vertical decanter but i was advised to change this to horizontal, by my advisor of studies, due to my high flow rate. I was wondering though when i changed values in the sheet titled LL-separators why the settling velocity came out with a negative number as my dispersed phase is the heavy phase and thus is sinking so should be positive, i assumed it was just due to me not entering data correctly, otherwise i thank you for the excellent help you have been. 

 

Bobby,

i have to an extent completed my work but i am yet to type it up for the report and once i do i will post it on this page. I calculated my viscosities of each phase by using the Viscosity Blending Index calculation and was hoping someone could advise me whether or not this is a satisfactory method for determining the viscosity? I calculated my viscosity of light phase to be 0.003582 Ns/m² and heavy phase of 0.007996 Ns/m² which did result in my other answers changing such that my residence time is around 3.4 hours and dimensions of 2.5m diameter and 12.5m in length. My biggest concern at the moment of using the viscosity blending index is that it is used for the mixing of liquids so i am unsure how i can include for the viscosity change caused by sodium methoxide that is soluble in the methanol, is it possible to use the viscosity of liquid sodium methoxide in the equation? it is a very small mass percentage anyway so would i be better just neglecting it.

 

Regards