Draw off height 2

[hanon]

We are trying to revamp a tower and as a result of this changes a high flow increase in a partial draw-off is required. I am interested in any correlation to predict a safe height to prevent flashing in the nozzle. Xurrently we are operating with 40 m3/h and a draw-off height of 220 mm but we are moving to 90 m3/h . Any suggestions are welcome: Is it ok? Another extraction is required? Lifting the tray to increase draw-off height?

Thansks

 

[reena1957]

DearBanon,
As you are practically doubling the flow, the head requirement will go up 4 times which might not be practical to implement. Better to double the area of x-section of draw nozzle OR providing additional equal size draw off nozzle symmetrically placed and ne line meeting the older draw line about 1.0 meter below the draw level.
Best wishes.

 

[25362]

You might need about five times the head you needed before the change. What is the size of the present nozzle, and what is the "reserve" you actually have now ?

The formula for the hydraulic head in inches of hot -low viscosity- liquid to push it through a clean (not plugged) draw-off nozzle is [Δ]H ~ 0.34 V², where V is in feet per second.

Then the external line must also be checked for friction drop, in particular if it goes to a pump requiring a given NPSH.

Can you provide more details ?

 

[hanon]

Thanks for your help. The actual pipe is 6" and we want to extract through the nozzle heavy naphta at 150ºC a 1.4 Kg/cm2 . I am not afraid for NPSH problems because the draw-off is located 20 meter above the pump. Q = 90.8 m3/h --> V = 1.4 m/s = 3.66 ft/s

I have used two correlations from Lieberman:

In "Working Guide to Process Equipment":
?H = 0.34·V^2, where V is velocity in ft/s. =...= 187 mm
Im "Process Design for Reliable Operations":
?H = 15·10^(-6)·G^2/(D^4). G flow in gph; D inlet diameter, inches,?H height in incles of liquid = ... = 169 mm

Currently we have a height of 220 mm. Specialists say that there are more conservative correlations. Besides, Lieberman recomends for the 2nd correlation to double the height value to manage disturbances inside the column. If there is no other solution we´ll make a new simetrical draw-off (to simplify the project) but first we want to be sure if we really need it.

Any suggestions are always welcome. Thanks

 

[25362]

Some comments for your consideration:

[•] The conversion from m/s to ft/s is incorrect.
[•] From tables by the Hydraulic Institute, a 6-in sch 40 pipe would result in a velocity of 3.44 ft/s for a flow rate of ~ 400 gpm.
[•] An 8-in sch 40 pipe, with such a flow rate, would result in 2.57 ft/s.
[•] Minimum residence times in the liquid draw-off sumps are designed to enable complete V/L disengagement.
[•] A 6-in line may not be self-venting even for the original flow rates.

 

[hanon]

Sorry for the writting error. The right conversion is 1.4 m/s = 4.66 ft/s instead of the previous 3.66. It was a numeric typing error. The calculation dH = 0.34·V^2 = 7.38" = 188 mm was right. Where can I find references about the minimmun residence time to V/L disengagement that you suggest. I didn´t know this criteria to design draw-off sumps. I thought the only real variable to satisfy it was the height in order to avoid cavitation dued to fricction loss. Please give me some references to find this criteria.

About real design (40 m3/h) for the draw-off we have not any problem with the 6" pipe. In fact, we have reached 60 m3/h without any trouble

Thanks for your help

 

[25362]

The book of interest is DISTILLATION-OPERATION- by Henry Z. Kister - McGraw-Hill

 

[25362]

Do you measure the 90.8 m³/h at 15^oC and convert to actual conditions, or is this the actual liquid flow rate ?

 

[hanon]

The 90 m3/h is supposed to be the flowrate at liquid condition ( 56000 Kg/h with 630 Kg/m3 at P,T ) We don´t measure it because these conditions are for the future project. Currently our maximun tested draw-off flowrate is 60 m3/h measured at P,T : 1.4 Kg/cm2, 150 ºC). We can not try a test to go futher (and test the sump) because actually at 60 m3/h we reach a hydralic limitation and the control valve goes fully open.

What about lifting the draw-tray (modifying the sump weldings) to have a higher height in the sump in case of not being valid the current height? We could takes some centimeters ...

 

[25362]

Does it mean the tower diameter, the pumps, heat exchangers, etc., are all suitable for the expanded capacity designed in the revamp ?

 

[reena1957]

Dear henon,
See the link

http://www.koch-glitsch.com/koch/faq/faq.asp.

Especially Q13. Read others also.
Best wishes

 

[25362]

The recommended residence times for non-foaming liquids is ~ 1 minute when feeding a pump on flow control. It may be less, according to site experience.

This is what Kister says:

Under some conditions, it may be impractical or unattractive to provide sufficient residence time in the liquid draw-off sump. Then separate drawoff drums or surge drums that vent back to the column are added outside the column to provide the required residence time.

I still recommend to read the chapter titled "Bottom section and column outlets".
It shows various trapout pan arrangements and the heights on each one.

 

[hanon]

Residence time of about 1 minute is huge for a partial draw-off. The current sump was designed for a calculated residence time of 2.6 sec.!! ( is a box of 250x300x320 mm for design flows of about 30 m3/h). To satisfy this design criteria we should make a huge sump!. I will try to get Kister´s book because I have noticed that is highly recommended. I have also checked Koch FAQ page is quite useful. Thanks

 

[25362]

Agreed. Good luck.

 

[reena1957]

Dear hanon,
If the sump is at the botom of a downcomer, add 50% of downcomer volume ( expected foam free liquid level)to the the sump volume to calculate the residence time. Also use the total liquid flow in stead of the partial draw flow in the calculation. May be this will be close to your requirement! Tell us the results.
Best wishes

 

[hanon]

The sump is at the bottom of a central downcomer. Tower diameter = 2300 mm and tray spacing = 610 mm. For the design flow (around 30 m3/h) and taking into account 50% of downcomer volume the residence time is 30 secs., closer to the recommendation.
In other hand I have read that typical drawn-off velocities are about 0.5 to 1.0 m/s.