OT: Measuring Level In A Tank

Off topic, but I know you guys deal with all sorts.

Basically, I have got a tank, that I need to measure level in.

I don't really want to put a radar, or ultrasonic probe in the top, I would prefer to use hydro-static pressure.

That means having a pressure transmitter in the top, and one in the bottom.

Now the problem is the one in the bottom. I would prefer to not have to cut into the tank, so that means installing the bottom transmitter in the outlet pipe.

I have tried this, and when I start to empty the tank (30,000 l/hr) i am pulling a vacuum on the pressure transmitter meaning I am getting a 0 reading for my level.

Has anyone successfully fitted a pressure transmitter into a pipe and got it to measure level in a tank?

The transmitter I am using is an IFM PI2794.

Cheers

Mark

I see exactly what you describe often here where people have fitted pressure transmitters onto water reservoir outlet pipes without consideration for the headloss that occurs during high flows. No way i know of to avoid that problem short of a new fitting on the side of the tank.

Is the tank open to atmosphere? They do make hydrostatics that will sit inside the tank, however I'm used to using them with applications where the free space inside the tank is at atmospheric pressure. Have a look at the Endress waterpilots. Also chemical compatibility may be an issue depending on what's in that tank.

Mark,
sounds like you have a sealed tank, does that rule out using radar or ultrasonics? They would be ideal, depending on the material inside the tank, and what properties it shows when filling and emptying the tank.

I would be tempted to call in the experts from Siemens or Endress & Hauser, explain the situation and see what they come up with.

What about a submersible pressure transducer like this one? http://www.transducersdirect.com/submersible-pressure-transducer-tda85

We use all sorts of instruments in the water industry. "Drop in" or submersible transducers eliminate the need to tap into a port or the tank itself but they cost quite a bit more and can be more trouble to replace depending on the size of the tank. We have one customer who uses them in 100' water "stand pipes"

What is the height of the tank?

Is it open to atmosphere (or well vented)?

What is the fluid?

Thanks for all of your answers so far.

I had looked at ultra-sonic, but I have had bad luck in the past, and still are having bad luck with them now.

I have had Siemens in, to look at them, but when the tanks get low level, I get false readings.

The tanks are 3.8 metres tall, and are pre-purged with CO2.

The tanks hold cider, and then are cleaned using hot caustic, and then sterilised using acid.

Cheers

Mark

sounds like a job for hydrostatic load cells and some math...

Avoid the gas, the liquid and most of the chemistry.

maybe expensive way -GAMMA

Maybe a couple of these would do the trick:

http://www.jowa-usa.com/industrialMetritape.cfm

Any appreciable percentage of CO2 will give any ultrasonic problems. CO2 blanket atmospheres are a no-no for ultrasonics. Radar could care less.

Dished bottoms are problematic for both ultrasonic and radar.

A submersible pressure sensor diaphragm (and cabling) will not be happy with acid wash. The cabling might not like the high pH caustic.

I've said it before, pulsed radar baby. Not cheap, but with proper configuration, it can handle odd shaped tanks, multiple liquid boundaries, foam, mist, sludge, dip tubes, agitators, dished bottoms, etc. They make them with sensor heads designed to withstand almost anything, and for almost any type of process connection.

I have to agree with OkiePC, load cells are the way to go. That is what we used when I worked at a large commercial bakery for our egg tans, as well as for measuring water, flour, sugar, etc at the mixers. They work great and don't require any contact with the product or the cleaning solutions. All you need is the density of the fluid and the tank geometry to do some relativly simple math to deduce level.

Another option - if it's an open/vented tank - is a bubble tube. Piece of pipe run vertically all the way to the bottom of the tank (leave it about an inch off the bottom), and then pipe a very slow flow of compressed air into it, so that it releases one bubble every couple of seconds. Then measure the compressed air pressure in the pipe - this will correspond to your tank level.

Obviously, this only works if your tank is vented/open as I said, and if your process is OK with the bubbles. It's best to cut a small v-notch in the bottom of the pipe, so you only get small bubbles squeezing out of the notch, instead of a big one the size of the whole pipe each time. But this means that the only part in contact with the nasty chemicals is a bit of 1" pipe. I've used these in caustic and acid tanks before with very good results, and it's much cheaper than putting the tank on load cells.

In theory (or so I was taught at instrumentation school), the pipes are self cleaning due to the constant flow of bubbles, so they're maintenance free. In reality, this is not always the case - the caustic pipe I mentioned before had to be cleaned of crystallized caustic every few months. Otherwise it'll eventually block up and your tank will read one million percent full

There are also the strain gauge type. Just bolt it to a leg and reading changes with stress in the leg. No the most accurate though. We use them a lot on grain silos.

Omron 61F GP N + what ever probe is needed