Priming a Pump in Suction Lift Service

[RefineryRR]

Problem Statement: The pump (shown in the crude sketch attached) is a basic overhung, centrifugal pump which is required to pump out salt water from a pond (at approximately 20 m3/h). Due to the suction line extending over the bund wall to enter the pond, there is an accumulation of air in the suction line between the water level and the top of the pipe, when the pump is stopped. When the pump is put online, the air is drawn into the pump and causes seal failure due to dry-running.

Additional Info: The suction line is approximately 100m in overall length.

Request: What options are available to evacuate the suction line of air? Currently, I am investigating the feasibility of having the suction line drilled right through the bund wall horizontally. What other options are available?

 

[LittleInch]

Generally you either introduce a vacuum line or an air educator / venture using compressed air to generate a lower than atmospheric pressure and induce flow in the pipe. Connect via a float system to the inlet line and fill it nearly all the way before turning on your pump.

Either that or stop it draining out in the first place with a foot valve

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

 

[pumpingtips]

Normally we would use a priming tank at the pump suction. Just needs to be big enough to cover the volume of air in the pipe (times two for margin). The priming tank will evacuate all the start up air once it gets going and will run full of water usually. You cant stop/start too many times in a row or you need to keep filling the tank with initial water charge. Be careful of syphoning - i mention that because your diagram shows that the pump is lower than the dam water level. Good luck

 

[LittleInch]

I was wondering about that as well. I used one years ago on a seawater pump which was above water level, but always thought it was a bit odd. At some point during start up you have a mixture of water and air going through the pump for short periods. So long as your downstream equipment can permit quantities of air being emitted from the far end then it is simple and needs no air or power.

This is an example

http://www.builditsolar.com/Projects/WaterHeating/PrimingTank.htm

However given you have quite a long line the vessel might need to be excessively large....

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

 

[Artisi]

Do you require a fully auto system or a manual operating method, it makes a large difference to the best approach?

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[RefineryRR]

An excellent suggestion, LittleInch. A foot valve at the suction line entry may be the simplest, and most cost-effective approach. There is minimal DP across the theoretical check valve; the only consideration will need to be the head loss across is, as the NPSHA is already very close to the NPSHR for this pump.

Artisi - I am open to any, and all proposals, but obviously, cost efficiency will be a major driver.

 

[Artisi]

The simplest is an air ejector / eductor fitted onto the pump casing with means of isolation
as suggested by LittleInch, with a gate or butterfly valve on the discharge of the pump.
Operation: close the discharge valve, run the air ejector until the casing in full, start the pump unit and open the discharge valve, and close the isolation between ejector and case, shut off the air supply to the ejector.
The simplest trouble free method.

It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts. (Sherlock Holmes - A Scandal in Bohemia.)

 

[pumpingtips]

The reason foot valves are not ideal, particularly with salty water or solutions with particulates, is that they need to be air tight. Over time they build up with encrustation, or corrode. The advantage of priming tanks is they are reliable once sealed properly. But littleinch is right about the 100m length. Its a lot. So best option would be put pipe through wall. Or maybe the system lends itself to manually start/stopping the pump up at start up to allow the air to escape on the discharge side of the pump.

 

[LittleInch]

What is downstream of your pump? Can you flow backwards for a short while at velocity > 1-2m/sec in your inlet line for a short time to flush the air out and then switch over to the pump?? You probably need to work on a volume of your inlet line x 1.5 or 2 to fully charge the line.

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

 

[georgeverghese]

There are centrifugal pumps meant for this type of service with built in suction aux. startup priming pump - this was described by an OP in this forum some months ago - you could check this option out with vendors if interested. Agreed a suction foot valve in high concentration brine service may be problematic.

 

[cneberly]

You might consider a submersible pump system in lieu of a self-priming pump. I found this website online that may be of some help:

http://parkusa.com/index.php/products/waste-water/lift-stations-2