## Featured Articles

Check out the latest featured articles.

## New Article

Product Viscosity vs. Shear

## Featured File

Vertical Tank Selection

## New Blog Entry

Low Flow in Pipes- posted in Ankur's blog

3

# Flare Kod Liquid Transfer Pump Available Npsh Calculation

13 replies to this topic
|

### #1 eimonthet

eimonthet

Brand New Member

• Members
• 9 posts

Posted 10 September 2022 - 02:59 AM

Hello,
I hope you would be having a good day. I have two questions to ask with relation to flare knock out drum (KOD) liquid transfer pumps in an oil separation train project.

There are two condensate (mainly crude oil and water) transfer pumps connected to KOD which are pumping out condensate liquid from KOD to other vessel.

The elevation difference between flare KOD liquid level and pump suction nozzle centreline is 3.45 meters. Pump suction piping head loss is 0.9 meters. The minimum operating pressure of flare KOD is 0.25 barg as per Hysys simulations.

The objective is to optimise flare header elevation. The flare header is about 1 km long from KOD to flare stack and it is supported at 75 locations along the route. If we use actuals values of KOD operating pressure (1.25 bara @ operating temperature) and condensate vapor pressure (0.9 bara @60 degree celcius) in NPSH calculation then the available NPSH is 5.5 meters.

But, our contractor has used 1 atmospheric (a) pressure for KOD and 1 atmospheric (a) for the condensate vapor pressure which cancels each other out in NPSH equation. The resulting availbal NPSH is 2.5 meters which is nothing but the elevation difference between vessel and pump after deduction of frictional head losses in pump suction line.

I want to use actual design parameters in the NPSH calculation and based on that want to lower KOD elevation from 4 meters to 2.5 meters with an availble NPSH of equal to 4.0 meters. This will result in saving of 1.0 ton structural steel at each support location and would save about USD 1.0 million in terms of material and construction cost. We do not want a conservative design. We just want an optimised safe and reliable design meeting minimum requirements.

The questions are on the calculation of available NPSH for these pumps.
(1) Why minimum operating pressure of KOD not considered in pump NPSH calculation? Is there any basis which support such assumption? Is there any reference such as code or any client specification? What is industry practice while calculating NPSH for KOD pumps?
(2) What is the value of vapour pressure for KOD condensate (crude oil and water) at 60 degree centigrade? looking forward for the answerees.
Regards

### #2 Pilesar

Pilesar

Gold Member

• Members
• 1,044 posts

Posted 10 September 2022 - 07:00 AM

The elevation of a 1 km long flare header should not rely on the suction head of a knock-out drum pump. Surely there are more important criteria for flare header design.

### #3 eimonthet

eimonthet

Brand New Member

• Members
• 9 posts

Posted 10 September 2022 - 08:15 AM

Dear Pilesar, Thanks for your comment. Flare KOD pumps suction head govern the elevation of the KOD. The flare header connect with KOD top nozzle with an 90 degree elbow to keep the elevation to minimum. The flare header routed with flare stack with a positive slope (free draining into KOD). In this case only KOD pump suction pressure governs elevation of flare header starting from KOD to the flare stack. Please provide your feedback and answer the questions I asked in my original question.

### #4 Bobby Strain

Bobby Strain

Gold Member

• Members
• 3,301 posts

Posted 10 September 2022 - 09:13 AM

Bobby

### #5 eimonthet

eimonthet

Brand New Member

• Members
• 9 posts

Posted 10 September 2022 - 02:58 PM

Dear Bobby Strain

Ours is a below case. Please let me know why contractor is correct?

(1) LIQUIDS BELOW BUBBLE POINT TEMPERATURE
In the situation where a fluid BELOW its bubble point temperature is being pumped, the reservoir pressure is, by definition, higher than the vapor pressure and the vapor space must contain light gases whose partial pressure makes up the difference between vessel pressure and fluid vapor pressure.  This difference provides a positive NPSHa.  The net NPSHa at pump suction is somewhat lower, however, because of friction losses.  In this situation, increasing reservoir pressure will increase NPSHa, by an equivalent amount, for a fixed system geometry.  Doubtless, it is this experience that leads novices to suggest the same solution for saturated liquids.

### #6 Bobby Strain

Bobby Strain

Gold Member

• Members
• 3,301 posts

Posted 10 September 2022 - 03:49 PM

Your operators never know what is in the drum. Nobody would design the pump for anything but saturated liquid.

Bobby

### #7 eimonthet

eimonthet

Brand New Member

• Members
• 9 posts

Posted 10 September 2022 - 04:00 PM

Ours is a simple oil, gas and water separation train. The liquids handelled are crude oil, produced water and associated gas which is being flared. Are you suggesting we should not optimise even it is inhently safe design as level rises pump suction head will increase and during emergeny liquid will reach KOD at higher pressure. The design pressure for KOD is 10.6 barg, maximum operating pressure is 1.64 barg, normal operating pressure is 0.84 barg and minimum operating pressure is 0.4 barg.

Just would like to add below safegaurds for the facility. Flare KOD pump will trip on low low liquid level (LLLL=150 mm)Flare KO Drum Pump will start on high liquid level 1(HLL1=500mm), standby pump will start on high liquid level 2 (HLL2=2925mm). Flare KOD pumps are designed to empty knock out drum vessel from HHLL to LLL in 2 hours. Second pump will start when additional 2.7 meter head is added to the NPSH due to rise in level. It is inherently safe design. Do you still see a reason not to calculate NPSHA for normal operating condition as the liquid being pumped in not saturated.

### #8 Bobby Strain

Bobby Strain

Gold Member

• Members
• 3,301 posts

Posted 10 September 2022 - 04:44 PM

A flare KO drum has no "normal operation condition".

Bobby

### #9 eimonthet

eimonthet

Brand New Member

• Members
• 9 posts

Posted 10 September 2022 - 10:23 PM

Okay. Than lets consider worst condition which is minimum operating condition but in any scenario liquid will not be boiling inside KOD. That assumption is not realistic.

### #10 Bobby Strain

Bobby Strain

Gold Member

• Members
• 3,301 posts

Posted 10 September 2022 - 11:06 PM

Saturated liquid is not boiling. But boiling liquid is saturated.

Bobby

### #11 eimonthet

eimonthet

Brand New Member

• Members
• 9 posts

Posted 10 September 2022 - 11:30 PM

Saturation temperature of water is 100°C at atmospheric pressure.

What is saturation temperature of crude oil at 0.25 kg/cm2g (1.25 kg/cm2a)?

That answer will help me in decision making.

### #12 breizh

breizh

Gold Member

• 5,633 posts

Posted 10 September 2022 - 11:34 PM

Hi,

I agree with Pilesar's and Bobby comments, you should stay on the safe side for the design.

At the end of the day is your responsibility.

Note: nobody in this forum can answer the question about characteristic of the crude oil, you should ask people from the plant or R&D based on a representative sample.

Good luck

Breizh

### #13 eimonthet

eimonthet

Brand New Member

• Members
• 9 posts

Posted 11 September 2022 - 12:00 AM

Thanks for providing crutial insight. Just would like to add that our iquid is sub-cooled not saturated liquid.The maximum operating temperature is 70 degree celcius and minimum operating presuure is 0.4 kg/cm2g. The liquid will be a mixture of cruid oil, water. The HC vapor will be on top of the liquid hydro carbons. There would not be dissolved gases in liquid as vessel presuure is low.

### #14 eimonthet

eimonthet

Brand New Member

• Members
• 9 posts

Posted 12 September 2022 - 03:18 PM

Dear Bobby and others, Thanks a lot for your valuable answeres. I agree what you said in all your posts. Its a saturated liquid. Thanks again.