Return of liquid in discharge line 1

[EngMecBra]

Hi,

Here we have a pumping system that sends a specified volume of water to a tank, it works by batch.
In discharge line we have a valve and a flowmeter, when the required volume of water is achieved at the flowmeter the valve closes and the transfer is completed.
There is a plate with a lof of connections between flowmeter and tank to we choose which tank will receive the water during that batch.
But a problem is happening: the liquid in discharge piping after valve isn´t going to tank, is returning!
What could we do to avoid this return and always get the specified volume of water at tank?


See attached file. Please let me know if you need more information.

BR

 

[crshears]

Not allowed to open your file; firewall issues...

Maybe you need a tradesman and not an engineer; check valve malfunction, perhaps?

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[EngMecBra]

Correcting: just the water between plate and tank is returning, not the water between valve and plate!

 

[AWloo]

Perhaps installing a check valve so that flow can only go through plate to tank and not return.

 

[EngMecBra]

AWloo,
A check valve after plate could stop the returning flow, but the water would be stopped after this check valve, not going to tank... We need that all water flowed through flowmeter arrive at tank...

 

[crshears]

...sounds like a simple physics problem; but I've been wrong before...

Dumb question, maybe, but since I can't open the file, is there a possibility that siphoning is occurring?

Are the receiving tanks at different elevations? Is water therefore flowing from a higher tank to a lower one?

Should there be vacuum breaking facilities in the delivery lines? Or perhaps these are present but not functioning correctly...

I'm having a tough time visualizing your set-up, but from your description it appears that you are having water escape from the line whenever you alter the configuration of the diverter plate.

Would it be feasible to use a "Christmas tree," or header, at the output of the flow meter and use a series of gate valves to choose the destination tank instead of a diverter plate? That way, any water that's in the line will just sit there and should not pose an issue, unless it's drinking water and shouldn't be allowed to stagnate...but you haven't said...



CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[AWloo]

Move the check valve closer to the tank. This would require you to adjust the shut off point on the pump to account for the volume of piping between the flowmeter and tank.

Sounds like you have back pressure in the tank which is what is causing back flow.

 

[crshears]

A check valve after plate could stop the returning flow, but the water would be stopped after this check valve, not going to tank... We need that all water flowed through flowmeter arrive at tank...

What am I missing?

A check valve allows flow through it in one direction but prohibits return flow. How would "the water be stopped" after a check valve?

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[LittleInch]

Based on your description I can only assume that this "plate" is a multi hose connection type of thing where usually the hose connection is a self sealing type in the same manner as hydraulic hoses on tractors and trucks. Is this system new?

Another way would be to have the plate higher than the liquid level and allow the pipe or hose to drain into the tank when you disconnect then add the volume of the pipe / hose to the volume you send.

Is this some sort of lab type set-up? Please define the size and if you can send a picture it would help a lot.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[EngMecBra]

LittleInch,
"Is this system new?" Yes.
Your comprehension is correct, "plate" is a multihose connection that we use to send water to some different tanks just changing the hoses, connecting a hose in the specific "plate" connection.
It seems the problem is regarding the levels, our idea here now is put this "plate" in a higher level to be sure the water will be drained to tank.
Thank you!

 

[EngMecBra]

crshears,

Check valve after "plate" (multihose connection) wouldn´t make water flows, just would avoid the return to "plate".

If the water between "plate" and tank isn´t going to tank (it´s my problem) and check valve would avoid the return of this liters of water to "plate", where would water be? Between "plate" and tank, of course.

I hope you understand now...

 

[LittleInch]

You just need to make sure there is a constant fall from the plate to the tank to avoid water gathering in low spots.

My motto: Learn something new every day

Also: There's usually a good reason why everyone does it that way

 

[crshears]

Ah.

Gotcha.

I was confused by the word 'returning,' which I took to mean there was an actual flow in the line toward the source once the filling action was completed.

So if what you meant was that you did not want to see standing water remaining in the lines but wanted it ALL to flow to the tank, your solution of pitching the lines toward the tanks / elevating the 'plate' is the right one.

What is the approximate volume of water per batch?

How tight are your tolerances, in other words plus or minus one per cent, +/- 2%, 5%, or what? Your description is that when the [integrating?] flowmeter indicates that the correct amount of water has been transferred, the water supply valve closes...but how rapidly? Will this contribute to metering inaccuracies?

And is the water pump dedicated to this one purpose or is the water being supplied from a larger system? In the first case you will not want to operate the pump at zero flow for an extended period of time and it would be wise to include an automatic pump shut-down feature in the control circuitry.

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[chicopee]

If the inlet piping to the tank under is under higher pressure than the plate outlet piping, then the water may be forced back to the plate.

 

[MatthewL]

Another option would be to put the shutoff valve AFTER the flow meter and tie in a purge valve just after the shutoff. Once the charge is finished and the shutoff closes, the purge valve opens and blows N2/air through the line to push the residual water to the tank. If the valve is automated, the control system can look for a pressure rise on the destination vessel to determine when the blowdown is complete, otherwise if manual, you can trend several blowdowns to determine a minimum amount of time to purge. I'm surprised you don't have metering issues already, since the meter could be empty at the start of the charge, which can lead to metering errors. Don't try to tie the purge line in with the system setup you have now, since the purge will "spin" the meter if a rotometer style or give errors if a coriolis style. I have seen this setup used several times without issues when there is a low spot in the piping between the meter and the destination vessel. It also helps to ensure that the hose connection is empty for the operator when they go to change the connection.

Regards,

Matt

Quality, quantity, cost. Pick two.

 

[bimr]

The best method to do this is to use a piping loop with the return going to the feed tank. Install an orifice on the return at the feed tank discharge to control the pressure in the piping loop. The pump should be sized for a greater flow than all of the takeoff flows.

Install an automatic takeoff valve on the loop directly above each injection point where you want water. You can use a timer to meter the flow into the tank or measure the flow with a totalizer.

If you want extreme accuracy, install a scale on each tank.