Level Applications from the Dark Side

You'd think that since level is a measurement we've been making since we were still using stone tools, that we'd have figured out how to measure level easily and simply and with high accuracy in every possible application.

Well, we haven't. And there are reasons for this. Mostly, the reasons are that there is no one kind of level, so there are many kinds of level measurement.

Wait! Level is just the top of a liquid or solid or powder in a tank or vessel, right? Not really.

Level measurements span a continuum that runs from "too easy to bother with" all the way to "this is too hard to do" and even to "this is impossible." Where a specific application lands on the continuum depends on how important a measurement it is, how necessary and how difficult technically it is to accomplish.

The right hand side of the chart is where things get really interesting. The chart is based on technical difficulty. That is, how hard it is to acquire an accurate measurement and do it repeatably.

What the chart doesn't show you is whether you should make the measurement or not. That's a decision that is based on several factors.

First, do you absolutely need that level measurement to control your process? If the answer is yes, it doesn't matter how technically difficult it is to make the measurement, you will figure out some way to do it.

If the measurement isn't exactly necessary, but it would be very nice to have it for control or optimization purposes, it comes down to how much it will cost to make the measurement and maintain the instrumentation over time. If the cost is very high, you may decide it isn't worth making the measurement based on the available technology.

In either case, there will be some level measurements that are either absolutely necessary or very nice to have that will be on the extreme left of the chart. At this point you have to decide if an instrumented measurement is necessary, or just having Jack from Maintenance drop by the vessel once a shift and look inside will do. "Yup, the level is fine." That's a measurement, too.

We're going to look at a few examples from the "Dark Side" of the Level Measurement Continuum.

Tickle.

That's what they call it. It is one of the most dangerous industrial chemicals. Its real name is titanium tetrachloride, TiCl₄ and it is used to make titanium metal and white pigment. Titanium tetrachloride is a yellowish liquid, but it hydrolyzes into titanium dioxide (TiO₂), a thick white goop, and hydrogen chloride gas (HCL g) in the presence of water. By water, I mean water vapor, humidity, the sort of water that can be found in an industrial vessel. Titanium tetrachoride also fumes. The hydrogen chloride quickly becomes hydrochloric acid (HCI).

You would think that the level of tickle in a vessel, being a clear-to-yellowish liquid, should be a relatively easy measurement. But it is not.

If tickle stayed a clear-to-yellowish liquid, the measurement would be simple. Unfortunately, it is not possible to remove all the water in the air inside the vessel, and what water there is, even humidity, quickly hydrolizes the titanium tetrachloride into TiO₂ and HCL gas. What water is left quickly turns the HCL into hydrochloric acid. The TiO₂ is thick, sticky and ropey. It clings to surfaces and hardens on them. Standpipes fill quickly with TiO₂ and become useless. Sight glasses fail almost instantly. The HCL gas and hydrochloric acid corrode the interior of the vessel and make pressure taps unworkable.

Capacitance or RF admittance devices coat up, and even those that are supplied with anti-coating circuitry can have real problems making the level measurement. The same thing is true with time domain reflectometry (TDR) radar devices. Ultrasonic level devices are unreliable because the liquid fumes and the ultrasonic signal can be lost in the vapor or see a thermocline in the vapor space as the level rather than the actual level.

Frequency Modulated Continuous Wave (FMCW) or pulse non-contact radar has some chance of working, provided the TiO₂ doesn't coat the sensor horn or process seal.

Weighing the tank isn't an option either; a load cell can't tell the difference between the coating in the tank and the level itself because the coating isn't necessarily stable. The more throughput through the tank, the more likely it is that the coating will increase as the level goes up and down.

Even nuclear level gauges aren't a panacea for this measurement, because they need to be able to read through the tank wall and through the coating, through the air, through the coating again, and then the other tank wall. This may require a nuclear source that is significantly larger than plant safety policy may allow.  Also, the coating changes will cause apparent "drift" in the measurement, adding to the difficulty of making the measurement at all.