Liquid Level Sensors

Thanks for the info Massey!

Do you have any idea if we could install such a system on a particular tank? The vessel we're measuring liquid level in is part of a larger system and therefore immobile, so we would not be able to place it on a scale.

Build in load cells that constantly monitor the weight into the support structure of the tank, no need to place it on a scale.

Do you have any recommendations of suppliers to look at? in addition, if the load cells are to be within the tank I assume they would need to be compatible with our liquid?

Sounds like you may be a little unfamiliar with load cells. The load cells can be external the tank even under the structure supporting the tank. You can even put them under the floor that supports the structure that supports the tank.

Right, I guess I should probably explain the unit in more detail; the vessel is part of a larger unit, and is suspended in the frame (about a foot or so off the ground). So, my question is, where would the load cells go? If the load cells are mounted on bottom of the vessel (externally), would they have the capacity to sense changes in weight due to filling without the vessel contacting the ground?

Thanks for all the help guys!

Load cells may be the right choice, but without knowing what else is being supported along with the tank, it could be hard to get into the right range. The most obvious placement is right under a nearby foot. Unfortunately that means that the said foot must not be bolted or welded to the floor.

I'm not sure what it takes to contain 4500 (psi?) but even some scuba divers hose is good to 3000 psi, so I guess there is some serious metal reinforcement involved in the hose. A mesh is what comes to mind, but it all depends on the inside diameter of the plumbing.

A flexible coupling capable of supporting the tank would be a good start. Then support the tank by a floating foot to the floor. Your load cell would be most accurate this way. But if that is not permitted, well then, never mind...

The frame supports a few motors, various cooling components, a buffer tank, and a second tank identical to the one we have been discussing (we have to monitor the level in both). Therefore, the frame design is not particularly flexible, which seems to make load cells a bit impractical in this case. I will look into them though as it still appears to be an interesting option. Steel piping is typically what we use in the system, but some heavy duty CNG hose is used in places

The hoses used on firefighters SCBA's SELF CONTAINED BREATHING APPARATUS have a working press in excess of 4,000psi. They connect the cylinder on the back to the pressure regulator at the chest which feeds the air mask at about 15-20" of water. The big air cylinders used to refill them in a cascade system have a working rating of well in excess of 7,500 psi. OP should contact a hi press hose manufacturer. If no info there call Scott Safety Products (years ago was Scott Aviation for the breathing air systems used on war planes) or MSA. Ask them where they get their hi press hoses.. Also if the tank is hung load cells could be installed within the hanging gear that is suspended with. This is used sometimes in the chemical industry for the same problem as OP has. Good Luck, Old Salt

Thanks for the tip on contacts for hose; we had been having a bit of issues finding good hose as well. Are load cells particularly susceptible to vibrations? Since there are motors connected to the same frame, this was a worry of one of our engineers.

If not with the load cell it can be done with the electronics before the load cell display. Almost anything can e done if you are willing to pay for it or develop it. Contact a load cell scale manufacturer or supplier. If they can weigh 234,000 lbs rail cars they can handle the design you are working with. Good Luck, Old Salt

Build in load cells on the tank support, silly man.

What really is this magic liquid that has a vapor pressure of greater than 4500 psig at 90C? Why is this kept so hot? Something not sitting right on this end, as this thread appears to be looking for a place to detonate. If the temperature swings as wildly as you suggest, this thing is less of a storage vessel, and more of a B.L.E.V.E. waiting to happen. I think someone has latched onto a very bad idea.

GA.

I have experience with fire-suppressant gas cylinders..same thing. Nothing like a good scale!

GA and probably the cheapest and most practical solution.

Now, to find a skyhook...

DP

>We do not want to use differential pressure as the tank (12" diameter, 6' height) fills rapidly and contains a gas at a high pressure, which would lead to potential errors in such an indirect measurement.

How quickly is this vessel filling? Current DPs sample from 5-10x per second.

Does the liquid vary greatly in density over the temperature range? The DP-to-level calculation assumes a given density. Any density change from the assumed density is a proportional error in level conversion.

4500 psig is near the maximum allowable working pressure, 90 Deg C near the max process temp limit, but unless the tank was filling in 2 seonds I don't see why DP wouldn't work (assuming instrument handles the process conditions).

Load cells

Weight is also subject to level change with density change. It's great as a mass measurement, X.xx pounds/Kg, but suffers like DP does with level varying with density.

Using flexible connections to the vessel is key, because rigid connections (electrical, piping, supports other than load cells) can be a source of error. with the choice tension or compression load cells, a tank can be suspended and weighed.

Nuclear

Shoot some gamma rays through the vessel. Ronan is one I spec'd once.

DP

The tank fills in about 7.5 seconds, but this is not the main problem with dp; the main issue is that based on tank dimensions, we would only have around 8200in^3 of liquid. Since the pressure of the gas above the liquid maxes out at 4500psi, we are worried that the added pressure of the liquid will be insignificant in comparison, leading to erroneous readings. If this is not the case though, I'd love to hear a counter argument.

Load Cells

Due to our frame being essentially unchangeable and vibrations throughout the unit caused by the motor, we'd like to avoid this if possible.

Nuclear

We got a quote from Berthold on a gamma sensor; that bad boy was >$7000 just for point level detection!

Davis Instruments, 0-5000psi, SS316 wetted parts, digital display + 4-20mA output + 2x pnp digitally set switched set points, accuracy 1% of full range. $414. You will need an adaptor to fit 1/4"M NTP switch to 3/4" socket. Check on web for further details.

My worry here is that in our 6 ft tank, since the density of the fluid is very similar to water, we have about 72" WC. This is only about 3psi; 3psi is only 0.067% of 4500psi.

Try Endress & Hauser or Krone Instruments.

You give up too easy...

I had long conversations with both Endress and Huaser and Krohne. From both, there were only two options, Krohne's OptiFlex and E&H's LevelFlex. I was advised by both of their engineering department to look elsewhere for fear of losing calibration due to rapidly changing conditions since both systems are GWR. You may be too quick to judge

Sorry to hear that.

You may find something suitable here:

www.google.co.za/search?q=liquid+level+probe&rlz=1C1VFKB_enZA613ZA615&espv=2&biw=1680&bih=949&tbm=isch&tbo=u&source=univ&sa=X&ved=0CEEQ7AlqFQoTCJvX_fPdh8YCFeIo2wod3VQAXQ#imgrc=MjkOA59oSO7qvM%253A%3Bj_Wbi9Y-5NcnlM%3Bhttp%253A%252F%252Fi01.i.aliimg.com%252Fphoto%252Fv0%252F314306461%252FUltrasonic_Liquid_Level_Sensor_Ultrasonic_Level_Sensor.jpg%3Bhttp%253A%252F%252Fcnakm.en.alibaba.com%252Fproduct%252F314306461-210060811%252FUltrasonic_Liquid_Level_Sensor_Ultrasonic_Level_Sensor_Liquid_Level_Probe_.html%3B1142%3B886

Not too quick..you only mentioned two manufacturers that you tried !

Nevertheless, I want to alert you to the fact that if you are working with a flammable gas, all your electrics will need to be certified Exd (certified flame/explosion proof).

Caution is the word here, else your start-up may well shut down at first batch.

Just for interest there are hand-held level detectors available. Look here:

www.google.co.za/webhp?sourceid=chrome-instant&rlz=1C1VFKB_enZA613ZA615&ion=1&espv=2&ie=UTF-8#q=coltraco%20portalevel

I suggest you contact all of the following vendors and get their input on what they can do for you:

Emerson Rosemount

www2.emersonprocess.com/en-us/brands/rosemountanalytical

Omar VEGA

www.vega-americas.com/en/index.htm

Siemens

www.industry.usa.siemens.com/.../level_guide_en.pdf

These are all great companies, I contacted all of them and met with both Emerson and VEGA personally. Unfortunately they and Siemens could not provide anything under about $4500. Siemens had a capacitance switch for $2200, but the specs weren't quite up to par; Ametek Drexelbrook offered a cheaper and more robust system. However, all of these capacitance systems seem to have issues with the rapidly cycling; replacing an expensive unit like this every year is not very efficient.

Yea, given those internal temperatures & pressures it may be very volatile inside the tank so Weighing the tank via load cell or some other means seems like the best approach.

Another issue is that similarly to DP, the weight of the tank is about 1600lbs (has to be sturdy to handle the high pressures), so unless the load cells are >=0.01% accurate, it'd be pretty tough to pick up changes in liquid level as the weight of the liquid would contribute so little to the overall weight

DP

There's no reason why DP wouldn't provide a head pressure measurement.

DP is designed to 'see' relatively small differential pressures in the presence of relatively large static pressures.

An industrial DP transmitter mechanically subtracts the low side pressure (vapor) from the high side pressure (vapor + liquid head pressure) so the sensor 'sees' the differential pressure of the liquid head pressure only, which it converts to an electrical signal.

DP transmitters are sized according the range of DP that is desired. 10 feet, (120 inches of water column), is easily within the range of a typical DP cell.

You'd have to consider the MAWP though. If it were me, I'd get the 6000psi option:

But there's still the open question about the stability of the liquid density.

Load cells

Sartorious has load cells whose mounts negate the effects of vibration.

Video shows turbulence due an agitator, but the effect from external vibration is the same, which throws off load measurements that don't negate the effect:

http://tinyurl.com/lq57wvz

Davis switches

I didn't look it up, but is the medium sticky or gooey that will cause switches to 'hang' over time due to build up. Good price if the technology is compatible.

With DP, the sensors are typically 0.1-0.25% accurate, so I'll assume best case and say one we were to use is 0.1% accurate. Lets say the tank is empty, and the gas is at 4500psi. If the low side somehow gets this reading spot on (0.0% error), it will read 4500psi. If the high side reads this with say a 0.05% error, pretty good based on the maximum variance of the sensor, it will read 4502.25 psi, meaning a 2.25psi liquid head pressure, which would be about 60" WC. That's a problem; am I missing something on how these operate?

I will look into those load cells, as if they negate motor vibrations they could prove useful. once again though they are non ideal due to the frame build.

Fantastic explanation, this clears up the issue a lot. DP now sounds like possibly the most viable option for us. Thanks a lot Carl

Yes, load cell or DP is a good answer, for pressurized vessels as this. Since you are filling the vessel so rapidly, then your system will have to be tuned for extreme rapid response.

Does the vessel remain in place, or is it filled and removed for next vessel? Is the contents of the vessel corrosive? flammable? All considerations, especially safety must be taken into account prior to final answers.

Try looking for rf sensors. They can be used for high pressures, temps, and various materials.

http://www.sensorsmag.com/sensors/leak-level/a-dozen-ways-measure-fluid-level-and-how-they-work-1067

This web page has some good info about different ways to measure tank levels.

It seems Massey got the link before me. Good luck.

Look through K-tek Magnetostrictive Level Switches