Inverted U pipe in pump suction side. 6

[mod3302]

Scenario
In one of the ships, ballast system for the vessel has routed with an inverted U in pump suction line. This is not an avoidable situation due to space constraints and room arrangements in the vessel.
Refer drawing TANK802P-SUCTION LINE.pdf, here this is typical line from one tank, the function is to transfer sea water from this tank (802P) to overboard (to sea, outside the ship). this is basically to adjust the vessel stability based on various conditions. The vessel has several such tanks. But pump suction is considered only one at a time. In this present situation we are having difficulties in getting suction from some tanks (geometry is similar to attached one for most of the tanks). We have checked for air leaks but seems no such leaks existing.

We have also tried to operate the pump by fill the vent pipe with water, but that was also not successful, pump run for some time. Water height was around 14m (sensor is place 50mm above the tank bottom) including the air vent height. The pump was started after closing the overboard discharge valve. (Suction pressure 1.03 bar, pump discharge pressure 6.58 bar). The overboard valve was half opened then, causing continues flow for some time. Then the discharge pressure reduced to approx. zero bar and suction pressure of (-) 0.22 bar. Tank water height was around 6m at this time.
Pump capacity is 200m3/hr at 5 bar.

Kindly review and advice.

 

[1503-44]

Nothing mysterious going on there. Not eough suction pressure to maintain the siphon, pipe too small, too high and too long. Extremely bad design that will cause any pump to fail from seal failure and or cavitation related damages. A new pump may run for awhile, but not ong. How old is the pump?

You really need to reroute the pipe, or relocate the pump closer to the tank, or provide some kind of feed pump at the tank's location. You cannot avoid the physics of the installation.

“What I told you was true ... from a certain point of view.” - Obi-Wan Kenobi, "Return of the Jedi"

 

[katmar]

I doubt whether air leaks are your problem. Even at half of your rated 200 m³/hr the water velocity would be enough to flush the air out of the 200 mm pipe.

Having the water boil and vaporize at the top of the inverted U is the more likely problem. Especially as you said that it worked with the air vent pressurised. Lifting the water 8.2 m would not cause it to boil at temperatures below 60°C (140°F) but if you have a foot valve or strainer (or both) in the tank that would add pressure drop and could cause boiling at the high points.

As an aside, be careful of filling the air vent with water. I am not familiar with marine specs, but land based prismatic tanks do not take kindly to even a few inches of head above the top of the tank.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[mod3302]

Dear ax1e and katmar,

Many thanks for your feed back.

Due to unavoidable circumstances we had to route this pipe as it is. Some critical rooms are arranged below the valve manifold rooms for this pump. Hence pipe had routed in such a way.

Can I use an external priming pump to clear this issue, that is if I use a liquid ring vacuum pump (LRVP) for priming , will that be suitable for smooth operation of the existing pump?. It is noted that use of LRVP can give vacuum of 33mbar.

 

[HTURKAK]

The maximum height the water can raise at suction for the siphon theoretically 10.30 m. However, this height reduces for the friction loss, density of saline water, steam pressure of the water.

Just a preliminary calculation Hmax= 10.30-0.45-0.3=9.5 meters. That is, the pressure at highest point could be reduced to -9.5 m of water.However, in this case ,the dissolved air in the water will come out and will be collected at summit and the flow will stop.

The common practice, to provide safety margin of 2.5-3.0 water pressure to avoid the air build up at summit and keep the flow.

That is H max = 9.5- 2.7= 6800 mm

In your case , H= 8226 mm. !!!!

The options are, lower the summit ht from 8226 to 6800 mm or, provide another feed pump around TANK 802.
I must add that , the piping is POORLY DESIGNED !!.

 

[1503-44]

I agree with HTURKAK and Katmar Those calculations are correct as is the 2.5 to 3m safety reduction. For practical reasons mentioned, with a suction head loss of more than 7m, flow is very difficult to maintain for any length of time and you will most likely have much cavitation. I appreciate your routing constraints, however the limits of physics are even more severe. I think you have no practical alternative, other than perhaps considering installing a submersible pump into the tank itself, provided that you can get electrical power cables routed into that area.

“What I told you was true ... from a certain point of view.” - Obi-Wan Kenobi, "Return of the Jedi"

 

[katmar]

The priming pump will of course help with startup, but I suspect your real problem is when trying to run at full rated flow and the pressure at the high point is lowest because of static and frictional losses. The priming pump won't help in this situation.

You need to calculate the pressure at the high point. Start with the pressure that you know, i.e. at the surface of the water in the ballast tank. Calculate the frictional and static losses and subtract them from the known pressure. That will tell you how close to boiling you are getting.

In the end, ax1e's suggestion of using a submersible pump in the tank is probably the most practical solution if you can't reroute the pipe.

Katmar Software - AioFlo Pipe Hydraulics

http://katmarsoftware.com

"An undefined problem has an infinite number of solutions"

 

[LittleInch]

I agree with all of my fellow posters above.

You may indeed have space and routing issues, but in the end gravity will always win.

You are fairly fortunate I think that your attempt to start this pump by pressurising the tank to about 1 bar by filling the vent with water hasn't resulted in failure of the tank.

As I see it these are the possible ways out of your current routing and issues.

1) See if the tank will actually withstand a pressure of about 0.5 bar or say 10 psi. If it can accept this (I highly doubt it), then you would need to set a pressure regulating valve on the vent and blow air or another gas into the tank to maintain it at the saifd pressure.
2) As Ax1e says, install a submersible pump into your tank.
3) Find a routing which means the highest part of your pump inlet pipework isn't more than 7m above your lowest liquid level
4) Move the pump into that first vertical leg out of the tank in the side void.



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

 

[mod3302]

Dear
HTURKAK
ax1e
katmar
LittleInch

Many thanks for your responses.

we normally test all these tanks for its structural strength, hydraulically by filling tanks till top of the vent (alternate tanks covering all tanks boundaries)

 

[georgeverghese]

The pump is pulling in air through the open air vent, which is why the pump loses prime and hence cannot deliver pressure and flow. With this arrangement, you have to manually fill the pump suction line through the air vent till water dribbles out of it. (Am assuming you have a working foot valve at the pump suction bellmouth inside the tank?? - Else this manual priming procedure wont work). Then manually close off the vent. The diagram shows the air vent height at just the same as the highest point in the suction line. It preferably should be a little higher to ensure the high section of this suction line is also liquid filled.

The current location of the air vent appears to be wrong - it should be on the top of the high section of the suction pipe; just upstream of the current location, and oriented upward.

 

[mod3302]

Many thanks GeorgeVerghese
Air vent height is approximately 13.706 m and top most point of suction line from pump level is 8.226 m. Air vent is provided for the tank

 

[LittleInch]

So what are you planning to do?

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

 

[mod3302]

LittleInch

Many thanks

Not concluded yet. We are now thinking of using a external vacuum pump for trying the system before doing any modifications on pipes(since modification is a huge rework).

 

[LittleInch]

mod3302,

Where are you placing your vacuum pump connection?

I'm 95% certain this won't work I'm afraid.

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

 

[mod3302]

LittleInch,

This unit will be placed near existing pump. something similar shown below

 

[LittleInch]

You need to place the suction point of the vacuum at the highest point in the system and also make sure it doesn't suck loads of water into your system or your vacuum pump will go pop. So you need a float type system to block off the water coming in one it fills up and then allows the valve to open again when the level falls.

Normally the pump is the highest point in something like a water lift from a well or a river, but not here.

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

 

[mod3302]

LittleInch,

Many thanks for your response.

We are planning to use a float switch or level switch in discharge line of main pump, when water present in that discharge line of main pump, will send signal to a remote operated valve placed at air line of LVRP so that it closes.

 

[LittleInch]

OK,

I don't think you're really getting what I'm saying, so best to go ahead and give it a go and then let us know if/when it doesn't work.

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