Diff Pressure xmtr location in Liquid CO2 Tank

[danw]

The reason the transmitter appears to be tapped 3 feet above the bottom is that the exterior bottom is the outside of the 'vacuum' insulating shell, not the bottom of the liquid tank. The green arrow shows the elevation of the outside insulating shell.

The bottom of the interior, liquid CO2 tank is shown by the blue arrow. The high side DP pressure transmitter pressure tap, E, (circled in red) to LIT-100A (the DP xmtr, circled in red), is at the bottom elevation of the interior liquid tank, like it should be.

The low side tap, F, is at the top of the liquid tank and measures the vapor pressure.

Looks kosher to me.

 

[RET]

So, the differential pressure measured by your transmitter is the height of the liquid above the sensing point. If there's only gas in the lower leg (since it's outside the tank and warm), then the pressure sensed by the transmitter is the hydrostatic pressure of the entire height of liquid in your tank, since the gas is a much lower density than the liquid, then it's contribution either way + or - is negligible.

But if that lower leg is full of liquid, now you have a dense liquid in that leg, which will subtract from the hydrostatic pressure of the height of liquid in the tank from the sensing point, so now your differential pressure will be lower (i.e. lower indicated level) even though with the same tank level.

But that looks like a pretty typical installation, and I'm sure that your low level sensing capability isn't that critical, if you get down to wherever the transmitter is located, then it's time to schedule a delivery anyway.

 

[MasterBlaster]

I think you're right - if a pressurized water tank were the issue, AND there was a diaphram at the water tap, the transmitter could be anywhere. So, in the case of the LOX tank, I think the installation is good. I think.

 

[danw]

If there's only gas in the lower leg (since it's outside the tank and warm), then the pressure sensed by the transmitter is the hydrostatic pressure of the entire height of liquid in your tank, since the gas is a much lower density than the liquid, then it's contribution either way + or - is negligible.

In cryogenics, there are 'noise' issues where the liquid boils to a vapor state and in so doing creates pressure pulses which appear as noise.

Although pressure is supposed to be equal in all closed systems, there's something about gas in a liquid measuring tube that can create errors at low pressures, as seen on bellows or bourdon tube mechanical gauges used for low levels on water. I'd like to know how this system performs.

But if that lower leg is full of liquid, now you have a dense liquid in that leg, which will subtract from the hydrostatic pressure of the height of liquid in the tank from the sensing point, so now your differential pressure will be lower (i.e. lower indicated level) even though with the same tank level.

Nope, the lower leg connects to the high side port, not the low side port.
A DP internal measurement is High side minus Low side, so the lower leg pressure component is NOT subtracted from the liquid column, it is a component of the high side pressure; the other components are the liquid column hydrostatic pressure and the vapor pressure.

If the actual piping were as shown, with the DP transmitter at a much lower elevation than the exterior shell, then the transmitter would have to be ranged to zero out and eliminate the liquid head pressure component of the bottom leg impulse tubing.

It appears that the transmitter is mounted at an elevation the same as the bottom of the interior vessel; so the lower leg impulse tubing is probably level with the high side port connection in the actual installation.

 

[TheWaterboy]

DanW - It is without question NOT mounted at the same elevation as the bottom the interior tank. The size relationship in the schematic drawing is purely for clarity. There is a few inches difference between the inner and outer tank sizes.

Now the boiling in the lower (Higher pressure)side of the DP makes complete sense thus far. At some point in that tube there is a transition between liquid and vapor and how that would ever remain stable is interesting.

RET - yea, that's what I can't grasp. If liquid is in the tube then the mounting is wrong. But since its a cryogenic liquid, there couldn't be liquid in that tube. So the liquid in the tank would increase the gas pressure in the tube to the point where it remains a stable gas. But as the pressure increases more liquid it pushed into the tube and boils.
I wish it was glass, I'd like to visualize this.

As for not worrying about the level below it, that's just the wrong way to think.

I found the calibration procedure for the gauge. It says to open both sides of the gauge to atmosphere and make it read zero. Which tells me there is no offset put into it.

I sure will be happy to figure this out.

 

[RET]

RET - yea, that's what I can't grasp. If liquid is in the tube then the mounting is wrong. But since its a cryogenic liquid, there couldn't be liquid in that tube. So the liquid in the tank would increase the gas pressure in the tube to the point where it remains a stable gas. But as the pressure increases more liquid it pushed into the tube and boils.
I wish it was glass, I'd like to visualize this.

Yea, there's no real certain way to know if the CO2 is going to be liquid or a vapor. Chances are that at your tank that you said would be around 400 psig, it's likely that in cold weather it will be liquid and warm weather it would be gas. You could look up a phase chart to see what it would be and what quality it would be, and that just goes back to the uncertainty of the measurement, and how accurate it really needs to be.

 

[RET]

Nope, the lower leg connects to the high side port, not the low side port.
A DP internal measurement is High side minus Low side, so the lower leg pressure component is NOT subtracted from the liquid column, it is a component of the high side pressure; the other components are the liquid column hydrostatic pressure and the vapor pressure.
.

Just think about jumping in the liquid CO2 tank and diving down to the bottom, where the hydrostatic pressure is the highest, then swimming up to the tubing to the d/p transmitter, the pressure will reduce, the higher you go up the tubing, until you get to the transmitter, which is still below the level of the C02 tank level, so the pressure there is just due to the height of the liquid in the tank above you (relative to the gas pressure above of course).

Now imagine that there is a nice warm gaseous atmosphere, with a little beach just at the bottom of the tank where the tubing connects, like a little diving bell with a beach. As you dive down through the freeze ocean of carbon dioxide, the pressure on your eardrums is unbearable, then you reach the little beach and you expect some relief, but when you get out of the carbolic ocean, the pressure is still the same, then you try to climb up the tubing to the shiny little transmitter diaphragm way up in the sky thinking that you can get some relief, and you realize that pressure isn't changing that much because the gaseous atmosphere is not near as dense as the ocean was.

So, you see, if the transmitter sensing point is located at a higher elevation than the bottom of the tank, pressure felt by the transmitter will be lower with a leg full of fluid, and higher with a leg full of gas.

 

[TheWaterboy]

OK I follow that. And I'll stipulate that's what going on here as I have verified the premise with a tank filled with water.

I wonder what it looks like in this impulse tubing at the boundary layer between gas and liquid and is that boundary layer located within the liquid storage, the vacuum space or exterior. It doesn't really matter but its interesting to think about.

The Engineer should be in the area in the next couple days. Perhaps he can share some insight.

I did hear some interesting stuff from a natural gas tank system where a multitude of instruments are needed just to determine level in that tank. I count myself lucky I don't have to deal with that one.

 

[shooter]

IT will work , and when starting open this crossvalve and no overpressure on one side of the sensor.

 

[TheWaterboy]

Shooter, I don't understand the comment.

I got some more info from the tank engineer. There is the equivalent of a P-Trap in the internal tubing that acts just like the P-Trap under a sink. The jagged lines show this in the schematic. This simple arrangement allows the material in the tubing to the outside to transition to a gas within the inner tank so the transmitter sees only gas and so can be mounted almost anywhere.

The space between the inner and outer tank wall is only a few inches in each direction. Enough for a mylar like wrapping to reflect radiant heat, a ribbon like inner tank support and the space needed for the vacuum.

I asked him to send me some pics from the factory and when he does I'll post them.

 

[MasterBlaster]

They could certainly enhance the usefullness of the schematic by adding to the legend the fact that those squiggly lines represent the equivalent of traps. But, now that we know that, it is elegantly simple and easy to understand! Looking forward to the pics.