Sizing a priming unit

[vinhermes]

Hello to all,

I am currently sitting on a problem.

On a ship currently under trials, we are facing seawater cooling pumps that are de-priming on the main engines. The pumps level is exactly at the seawater level and they are fitted with priming unit (rubber turbine in parallel of the pump). The issue comes from the fact that the ship is designed as a DP unit, DP meaning Dynamic Positioning. In other words, it has the ability to remain static at a given location using its waterjets and transversal thrusters. By doing this, this makes a very bubbly water and generates the de-priming.

You can see an example of the media underneath the hull at this video (for instance between 30 to 45 seconds):

Link

So the idea is to add additional self-priming centrifugal pumps that would bring water at the inlet of the main pumps. By having their own inlet way forward, this should allow getting unaerated water.

My question is: how to size their flowrate? 10% of the main pumps flowrates? Is there any norm or reference when sizing a priming unit?

I hope that I am clear. If not, I can provide a quick sketch. Thanks by advance, Vincent

 

[LittleInch]

Sketches are always good to help understand the issue.

I'm struggling to understand what these priming units actually do. And what on earth is a rubber turbine??

Also what sort of size and flow are we talking about here?

It sounds like you're trying to pump aerated water which isn't easy for centrifugal pumps.

I think fundamentally you have the wrong sort of pump for this fluid. Any manner of priming units will only solve the problem if the fluid you're pumping doesn't contain a lot of bubbles.

If you can add pumps which have a different inlet away from all this churned up water then why not just use that for the cooling water supply??

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

 

[vinhermes]

Cheers LittleInch.

The rubber turbine is only a flexible impeller helping to prime the main engine-driven pump but that shares the same inlet.

Sorry for not mentioning the flowrate. It is 100 m3/h max.

See the attached sketch. The potentially added parts are highlighted in red. The idea is to limit the flow rate of the added pumps as much as possible, this is why I would like to stick to engine-driven for the main one.

Thanks again for your interest in this.

 

[LittleInch]

Are you sure about 100m3/hr in a 3" pipe. Or is the pipe bigger than the connection? What is the distance from hull to pump inlet?

That's over 6m/second. With what looks like a strainer in the way as well that's looking at some big pressure losses there so I'm not surprised a bit of air is going to lead to big problems.

I really think this won't be solved by just piping in a bit of water into the inlet line.

When this thing is working without air in the inlet water, what's the pressure you're seeing at the inlet into the pumps?

Where exactly is the inlet piing on the hull? Is this some sort of sea chest?

Does the air congregate there?

I still don't understand how piping water into the inlet at some fraction of the required flow is actually going to do you any good?

It still sounds to me like your inlet is in the wrong place, your pipes are too small, your filter is in the wrong place and you have the wrong sort of pump.

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

 

[goutam_freelance]

Even though the scheme is complicated with lot of pumps I understand that 100 m^3/h pump needs to be primed. Priming is considered available when the pump and suction piping is full of water.

Features to be added:

1. There should be a foot-check valve(submerged) just at the inlet of existing rubber turbine.
2. Depending on pump and suction piping volume up to discharge of 100 cum/h pump and required time of priming a suitable pump capacity may be selected. Say, if pump/piping volume is 0.5 m^3 and required time is 1 min, the required pump capacity is 0.5/60x3600=30 m^3/h.
3. A level switch needs to be at pump discharge to indicate completion of priming.

 

[vinhermes]

Cheers for that. I'll review this Goutam

 

[Artisi]

You are going to need either a-pump primed by a vacuum pump or a venturi primed unit---the problem you have is the inlet pipework and potentially the pump is full of entrained air - you have to remove this air from the system prior to the impeller inlet.
Alternately the inlet pipework can be equipped by a self-priming tank, either vacuum pump equipped or venturi to handle the incoming air.
Nothing special here or difficult, there are thousands of vacuum assisted self-priming pump in operation around the world.

Start by looking at Godwin self-priming pumps and search the net for other manufacturers.

Edit :

http://www.xylem.com

(Godwin pumps)

http://www.allightsykes.com

(Sykes pumps)

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.)

 

[LittleInch]

vinhermes,

I really think you're missing the point here. This has nothing to do with "priming" and everything to do with trying to pump water with entrained air with a centrifugal pump.

The advice by goutam would work for a simple pump in a well or sump, but not for this. As soon as you fill up the inlet pipe (where does the air go?) and start the main pump then it will draw in lots or airy water again and loose prime again.

Address the root cause, not some little band aid plaster.

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

 

[Artisi]

vinhermes / LittleInch
See my post - solves the problem of entrained air.

The misunderstanding is that you don't require a self-priming pump, you need the air removal system incorporated into vacuum assisted self-priming pumps.

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.)

 

[goutam_freelance]

Of course, a vent valve at 100 m^3/h pump needs to be opened to pass water. This is to be closed when pipe is full of water. I hoped this valve is existing since the pump is being primed by the existing system now.

 

[LittleInch]

artisi,

vacuum priming is, of course, a good solution for the initial prime and depending on the amount of entrained air can be good during operation.

I just don't think anyone has an idea about how much air is being entrained that needs to be got rid of on a continuous basis.

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

 

[TugboatEng]

Gilkes makes pumps specifically for your application.

https://www.gilkes.com/cooling-lubrication-pump-system-suppliers/self-priming-technology

 

[Artisi]

Tugboat: good link, I've added it to my reference material.

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.)

 

[TugboatEng]

FYI, we had a vessel that used these pumps. We still had loss of prime issues. Previous operators had removed the discharge check valve, I had to purchase and reinstall it. We still had some issues. The best pump can't compensate for bad sea chest design. The tugboat industry is obsessed with 1" vent piping using non-corrosion resistant materials despite rules indirectly alluding to 2" minimum.

 

[Artisi]

LittleInch / vinhermes,
The amount of entrained air is /could be a problem but should it exceed a reasonablely sized inlet pipe to assist air seperation by low velocity plus a correctly sized tank that is constantly being evacuated of air, then I don't see much chance of the main pumps ever being correctly operated.
Guess the physical size of this equipment could be a restraint depending on the space available.

Just food for thought.

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.)