4 replies to this topic

[paul88]

Hi friends from Cheresources,

 

In designing a crude oil -water separator vessel, some calculation procedures state the following:

A maximum droplet settling velocity of 10 inch/min is used regardless of the computed settling velocity.

 

The issue is:  If we use a coalescer to increase the size of the liquid droplets  ¿ Could we then use a droplet settling velocity higher than 10 inch/min to reduce the length of the separator?  

 

If the computed setling velocity give say like 40 inch/min ¿ can we use let`s say 80% of 40 inch/min to calculate the lengh of the separator vessel? 

 

I appreciate some lights on this subject please.

[PingPong]

I don't know which formula you use but usually such formula is based on Stokes Law, which is only appicable for laminar flow. For turbulent flow such formula gives too optimistic results, hence the (rather arbitrairy) limit of 10 inch/min.

 

If you want to calculate the settling of a water droplet in oil more accurately you should use the Drag Coefficient C_D for spheres, which is a function of the Reynolds Number of the settling (spherical) droplet:

 

 

As can be seen in the graph: for laminar flow (Re < 1) the C_D is 24/Re, but for higher Re numbers the C_D is always higher than 24/Re

 

(In reality a settling droplet is not quite spherical, and moreover there will occur a circulation within the droplet due to the wall friction, but for your purpose this approach should be accurate enough.)

Edited by PingPong, 03 February 2015 - 11:02 AM.

[paul88]

PingPong, thank you for answering.

 

I already know the 3 equations to determine the Liq-Liq settling velocity, their application depends on the value of the Reynolds number.

 

But the issue was:

 

Can we use a Liq- Liq settling velocity higher than 10 inch/min to reduce the size of a separator?

 

I got the answer from two equipment suppliers: Costacurta  and Sulzer

They use coalescers to reduce the length of a liquid-liquid separator.

 

This implies that the water droplet settling velocity they calculated and used was higher than 10 inch/min. (They guarantee the separation operation)

 

So 10 inch/min is not a Holy velocity that can not be surpassed as some made me belief.

[Bobby Strain]

Not only is it not "Holy", its origin is unknown.

 

Bobby

[PingPong]

Can we use a Liq- Liq settling velocity higher than 10 inch/min to reduce the size of a separator?

Yes, if you calculate the settling velocity using a C_D as function of Re as I described above.

No, if you use an equation that is actually based on Stokes Law.

 

They use coalescers to reduce the length of a liquid-liquid separator.

A coalescer increases the droplet size, and as we all know: bigger droplets settle faster than small ones.