OT: Keeping water out of impulse line

TheWaterboy · May 1, 2023

I have an issue where I have a DP transmitter on a horz tank reading liquid level.
Due to odd and unforeseen circumstances the tank filled to the top above where the transmitter is connected and so it allowed water into the low side impulse line to the Transmitter which threw everything off of course.

Is there a readily available solution to this? I came up with the idea of using a hydrophobic filter, a membrane style gauge isolator, a water trap like those on compressed air lines, and simply taking the impulse line above the top of the tank to prevent the siphon from ever occurring, like an inverted P-Trap.

I can’t change the pressure tap location so I have to deal with what I have.

But before I fall back on my training from the Rube Goldberg School of Engineering, I wonder if there is already a bolt in solution available for such conditions?

drbitboy · May 1, 2023

Yield to the force: fill the upper tap with water and leave it that way. You will have to recalibrate of course.

(edit: tap not trap, and added the need to recal )

TheWaterboy · May 1, 2023

Don't think I want to trap water up there, just need it to not be able to go into the low side. So an upward loop would work, and is the only viable option I can think of.

drbitboy · May 1, 2023

TheWaterboy said:
Don't think I want to trap water up there, just need it to not be able to go into the low side. So an upward loop would work, and is the only viable option I can think of.

Not sure what you mean by low side. Can you make a sketch?

Even if you do everything you can to prevent liquid water out, depending on the conditions, water vapor may condense and you could get water in there and not know it, which will throw off the calibration. This is a more common problem in plants with steam, but it could anywhere. Ensuring the upper tap is full of water to a known elevation is the solution: it is better to know the densities and heights of the fluids on the traps, water or air, than to not be sure

My dad ran large stream turbine thermodynamic heat rate acceptance tests to an accuracy of a quarter of a percent, and pressure measurements were a critical part of that, and this was the solution.

TheWaterboy · May 1, 2023

Like this. Tank too full lets water fill that reference tube (Low Side of DP Xmtr)

OkiePC · May 1, 2023

P-trap the dry leg up top and put in a blow down valve at the bottom.

Similar to this:
https://www.validyne.com/blog/tank-level-measurement/

TheWaterboy · May 1, 2023

That's what I was thinking. Didn't even consider the condensate leg though, nice touch.
Thanks guys.

danw · May 1, 2023

You already got it, an upward loop would work.

That along with a drop leg below the low side port elevation with a drain valve.

drbitboy · May 2, 2023

What is the pressure in the vessel space above the water?

Is it a vacuum, a positive pressure, or open to the atmosphere?

Tom Jenkins · May 2, 2023

You could have the upper tap terminate with a float valve inside the tank. That would close off the impulse tube.

drbitboy · May 2, 2023

If you go with the condensate leg below the DP unit to keep condensate from affecting the measurement, then that will require a regular maintenance task to open the valve daily or weekly or monthly or after an event.

Or we could steal @Tom Jenkins' idea and put the float valve after the condensate leg to drain it.

Or put another DP sensor on the impulse tube to generate an alarm when it gets water in it.

To be fair, you mentioned Rube Goldberg first

TheWaterboy · May 3, 2023

It's in a pressurized tank (130-ish psi) so internal maintenance of a float would be a problem, plus vapor condensation would still be a potential issue.

Emerson sent me a flyer for a set of transmitters that are DP but connected by wire so no tubing at all. Interesting option.

I went with the high loop and condensate drain, It will be tested today.

g.mccormick · May 3, 2023

You could also of course result to 2 gauge transmitters and a bit-o-math.

That is what the emerson device would do for you.

The issue with 2 transmitters vs 1 is that 2 will possibly less accurate than one due to summation of errors. It really depends on how accurate you need and if you can easily get there with 2.

TheWaterboy · May 3, 2023

the tank is about 4 feet high so I'm measuring around 0-2psi full scale. Not a lot of room for errors to add up.

OkiePC · May 3, 2023

VEGA Makes an electronic DP setup too. I'd shy away from rolling your own with two sensors and PLC math unless your tolerance for error is large.

OT: Keeping water out of impulse line

TheWaterboy

Lifetime Supporting Member + Moderator

drbitboy

Lifetime Supporting Member

TheWaterboy

Lifetime Supporting Member + Moderator

drbitboy

Lifetime Supporting Member

TheWaterboy

Lifetime Supporting Member + Moderator

OkiePC

Lifetime Supporting Member

TheWaterboy

Lifetime Supporting Member + Moderator

danw

Lifetime Supporting Member

drbitboy

Lifetime Supporting Member

Tom Jenkins

Lifetime Supporting Member

drbitboy

Lifetime Supporting Member

TheWaterboy

Lifetime Supporting Member + Moderator

g.mccormick

Lifetime Supporting Member

TheWaterboy

Lifetime Supporting Member + Moderator

OkiePC

Lifetime Supporting Member

Similar Topics