Pump suction pipe sloped upwards?? 1

[pszor]

Hello,

Most engineering guides (e.g. Shell DEP) advise that pump suction piping shall be horizontal or sloped downward to the pump.
I understand this might be a good practice but I can't see it is written as a requirement.

I'll go straight to my case.
We'd like to increase the emptying rate of a storage tank and it would be of great interest to use a specific pump. However, that pump is installed 1 meter higher than the tank nozzle so the suction pipe will have to be sloped upwards to the pump instead of downwards.

I am having trouble identifying what risks may arise from this design:
[ul]
[li]Suction pipe will be sloped upward from the tank to the pump.[/li]
[li]Suction pipe will have no pockets.[/li]
[li] Low-level switch will be set at or above pump elevation to avoid reverse flow from the pump suction into the tank when the pump stops.[/li]
[li] No NPSH problems.[/li]
[/ul]

What are the possible problems I might have with such design(sketch attached)?

Thank you everyone!

 

[LittleInch]

If NPSH isn't an issue then your only issue would be lack of prime, but that doesn't look to be an issue. The upwards pipe could result in any gas gathering in the pump but again so long as you don't start the pump until the liquid level is above pump elevation then you should be OK.

Why not just install a check valve and then you can set the low level trip lower?



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

 

[Artisi]

Drain back to the tank has nothing to do with the pump location, it's a function of the discharge head on the pump, if the discharge point is higher than the pump it will drain back,it will require a non return valve on the pump discharge, if on the inlet side it will ensure the pump is alway primed.

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.)

 

[MJCronin]

I agree with the comments above, and particularly with LittleInch ... I believe that he wisely touches on a gray area in the design ..

He states:

The upwards pipe could result in any gas gathering in the pump but again so long as you don't start the pump until the liquid level is above pump elevation then you should be OK.

We do not know anything about your re-used pump but we do know that you are aware of the loss of tank capacity as the cost of the new design.

But, as LittleInch aludes, you should be aware that you might be making system venting problems for yourself IF the pump develops a gas pocket within the casing ... We just do not know

Some pump/impeller combinations can clear internal gas pockets on startup, some cannot.

Does the pump have a vent plug in the upper casing ?

Please post more details and pictures about the pump ... What is the make and model ? .... What type of pump is it ????

Importantly, is the largest impeller installed in the pump (if it is centrifugal, of course)

More details please...

MJCronin
Sr. Process Engineer

 

[pierreick]

Hi,
Why don't you relocate the pump to make a better utilization of the capacity of your tank and correct the wrong design?
Note : How do you prevent gas accumulation at the inlet of the pump ? Any pipe reduction?
It seems you have guidelines to perform the task.
My view
Pierre

 

[pszor]

Thanks for your valuable input.

MJCronin, thanks for pointing that out, yes the pump is centrifugal and it has a vent plug in the casing.
I'll try to get more details/photos during the next week and post them.

Artisi, suction pipe will drain back to the tank whenever the liquid level in the tank is below the pump so I'd say the pump elevation is a factor here.
You are right that a check valve in the suction pipe will help ensure the pump is primed but I'd rather avoid installing those in the suction.
Setting the LSLL at or above pump elevation level will ensure (at least I hope) that the pump suction is always full of liquid.

Pierreick, relocating the pump is not an option right now but we'll do it in the future.
"Gas accumulation" is something that I've heard before but I can not see how it can happen in this scenario. If the pump and piping are properly vented during start-up and the low level switch is set at or above pump elevation, how can gas even enter the system?

I am not that experienced so I might not be seeing the full picture.

 

[mk3223]

Agreed with the risk and issue as mentioned in the responses.
IMO, to mitigate the operational risk, this pump may set the low-level switch above pump elevation and with a "safe marge" of NPSHa to turn off the pump.
Of course, in the idea case, there is no gas to be trapped in the pump suction as following the proper operation procedure, tightened the venting connection, enabled level switch, etc.

 

[georgeverghese]

a)There must be some means to prime both pump casing and suction line / clear out any gas pockets during startup.
b)In many cases, the liquid feed to the pump is not entirely degassed for one or more reasons ( unless this is some large tank with very low downward liquid phase velocity that will allow for complete degassing, and there is minimal risk of the feed fluid TO the tank short-circuiting to the EXIT from the tank to pump) - the suction line will lose prime eventually. The high point vent line on the suction line(returning to tank) to pump may have to be kept open all the time to keep this line primed for this case.
c)If there are solids in the liquid feed, these can accumulate at the suction line low point and eventually block the line. Even if the fluid is clean, corrosion debris can accumulate at this low point.
d)In addition to NPSH requirements, low level trip must also be high enough to prevent vortexing and gas carry under to pump suction.

 

[pierreick]

Hi,
We have no idea about the material pumped , no detailed description of your system , no data about the pump , nothing !

https://www.theprocesspiping.com/pumps-suction-piping-eccentric-reducers-straight-lengths/

Anyway you and others may be interested with the article underneath .
Good luck
Pierre

 

[pszor]

Thanks for your answers!

 

[LittleInch]

pszor,

There are a few puzzling things.

It depends on your aim and operation. If the aim is to be ab le to provide liquid flow to a downstream user at any time, then yes, you should keep the liquid level in the tank higher than the pump inlet. If it is to remove all the tank contents, then if you have no NPSH issues, then you pump will easily be able to cope with a 1m inlet level "lift". So long as the tank is refilled to above the pump inlet you should be fine, even if there are a few pockets of gas there, but a simple pump vent or auto air bleed would do the trick.

You could put the NRV on the discharge side ans unless you have a vacuum breaker somewhere, when the fluid flow stops, the pipe should remain full. You can also always tie in an actuated on off valve to close when the pump stops or trips to preserve the line full of liquid. I don't understand why you think this line will empty on pump stop - "to avoid reverse flow from the pump suction into the tank when the pump stops."??



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

 

[shvet]

What are the possible problems I might have with such design(sketch attached)?

Design you provided is not recommended by industry practices:
- ANSI/HI 9.6.6-2009
"9.6.6.3 ... For pumps operating with a suction lift, the inlet (suction) line should be slope constantly upward towards to the pump..." and particularry fig. 9.6.6.3
- API 686-2009
"3.1.2.3 ... When the liquid source is located above the pump centerline, the suction piping shall be sloped toward the pump ... Sloping the pump suction line toward the pump is preferred as it allows any gas or vapor to escape back to the suction tank ...

So in your case if bubbles accumulation&coalescence can not cause problems piping is designed well enough. Bubbles accumulation can cause e.g. loss of priming or gas plug (especially dangerous for wetted parts of shaft seal and bearings). Note that starting conditions (including intermittent start/stop) should be considered along with normal conditions as surprisingly large amount of gas bubles can come&accumulate in pump case during such conditions. The problem is more dangerous if there is a source of tiny bubbles e.g. a free-falling stream which are hard to pop up.

Discharge piping can be rather long and twisted/complex so gas plug can cause inadvertent problems in random sections of a hydraulic system.

 

[pszor]

LittleInch Sorry for not being clear about my thoughts.

I didn't mean to say the pipe will be fully emptied.
What I was trying to say is that if the pump stops when the liquid level in the tank is below the pump inlet, the suction pipe will partially drain back to the tank.
In such scenario, the suction pipe will be no longer full of liquid and the pump will lose its prime.

 

[LittleInch]

I can see that, but why will the pump drain backwards? Where is the air coming from?

a max 1m head won't cause a vacuum to form.

Thing is we don't know if this is significant or not, but if its a big tank you don't want to lose ~1m of your pump out capacity for no valid reason. So installation of a non return valve on the discharge line or an actuated valve linked to pump stop would prevent both of those things.

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

 

[georgeverghese]

Also note that a pocketed pump suction line is a bad idea for liquids operating close to or at bubble point pressure. Saturated liquids in such lines will be "simmering" all the time while in transit to the pump due to to minor fluctuations in pressure control at the source vessel. Hence the requirement for an unpocketed suction line operating at low velocity that will allow bubbles to escape in the reverse direction to the source vessel.
@LI,
Where is the air coming from?
Ambient air could leak into the pump from shaft seals, even if it is only (-)1m of liquid head in the pump. This would not occur if the shaft seal remains pressurised ( mechanical seal API plan 5253B or higher??) on pump stop.

 

[bimr]

We have no idea about the material pumped , no detailed description of your system , no data about the pump , nothing !

It would help to have these details.

Note that self-priming pumps may have this piping arrangement.