Level Sensor in Sanitary Slurry Tank with Diaphragm Pump

  • Thread starter supervisedautomation
  • Start date

Thread Starter


Thank you in advance.

I have a process where water and flour are mixed together in one tank in batches, before being pumped out of the tank for further processing. The tank is conical.

The original design was to use a level transmitter to indicate two conditions. One, when enough water has been added to the tank. Two, when the tank was empty.

I tried using an ultrasonic but it failed often at telling when the tank was empty. I tried using a laser but that fails at both.

The tank is sanitary, meaning it has to be cleaned between batches. I fill it with cleanser and run the mixer.

I considered putting the tank on load cells but that's expensive, and I'm worried the vibration from the diaphragm pump will make the signals too noisy to determine when it is empty.

Thank you very much.

You can spend the same amount or more messing around with other technologies that don't do the job well (like you've already discovered). Or you can go to what you've already identified as the proper technology, load cells, and get the brand that'll give you best 'weigh' around a potential issue with load cells - dealing with mixer motion.

That brand is Minebea-Intec (formerly Sartorius, formerly Global Weighing). It is a German engineered product in a class by itself in search of a vendor who can promote superior performance.

These load cells
- are matched (so don't require a summing box)

- use a unique mounting jig that for practical purposes eliminates the effect of agitator/mixer motion

- have the strain gage etched into the base metal eliminating the bonding that is a huge contributor thermal and aging error.

Spend a brief two minutes and watch the Youtube video (made back when it was Sartorius)

Watch weight response from a high shear mixer on loads cells - and note not only how stable the reading is, but how quickly it responds that addition of 0.5Kg weight (on 500Kg load cells).

If you think the run-of-the-mill load cells are expensive, you'll gasp at the 15-20% premium for Minebea Intec system over its competitors, but Sartorius load cell systems are designed to do EXACTLY what you want do and be the best at it. That premium disappears when you discover how little effort goes into a periodic calibration: a mere cal check with no needed tweaks.

Their web site is here:
Have you considered a DP cell with flush mounted diaphragm seal mounted on the cone at the bottom?

Variation on the DP cell would be a bubble tub with regular DP cell

Cone bottom tank with an agitator seems like a simple application for an ultrasonic I must say, or is there a foam layer on top?
The Load cells would be a good solution for this. I would be apprehensive on DP cell though the cost is always an attraction. My concern with DP cells are that mounting them low enough for precision is always a bit tricky. I find they tend to have slightly larger error too. Quite regularly some tanks I've used that are calibrated for 5-10 MT often still show +/-50kg. If this doesn't matter than perhaps that's fine. Using a DP cell for level indication and perhaps a vibrating fork level sensor right in the gullet of the vessel may be ideal.

I've recently started using Vega Radars for food applications. The Vegapuls64 is super accurate +/- 2mm in height and can be setup to ignore agitators. Also in the setup application you can program in the dimension of the tank so that it will naturally output a linear signal. Also set signal dampening for any vibration.

The DP cell measures Head not weight.

On the cone as far as equivalent weight goes, it's more accurate as the level gets lower. If you want to measure the weight, you also have to think how you are going to remove any pipe stress.

I agree load cells are the best for mass, but the original post asked for level.
Since you are only monitoring two conditions you should consider using level switches, one for full and one for empty. There are several brands that feature sanitary CIP connections.