Coriolis meter in Demin Regeneration 1

[Ranito]

Hello,

I am looking at revamping my demin plant. Part of the project is to fully automate the entire process. With regards to regeneration control, I'm looking at coriolis flow meters to control hidrocloric ácid and soda injection. I know that the norm is to use conductivity metering, however the robust nature of coriolis meters as well as flow, density and temperature readings in one instrument is quite atractive (although more expensive). Does anyone have experience/opinions?

 

[JLSeagull]

One domestic US supplier provided Hastelloy C-22 meters for a low flow rate HCl vapor application.

 

[bimr]

It will work. The drawback is that you are talking about an expensive meter, either exotic metals or lined.

Not sure that the increased accuracy will have a payback either.

 

[gedomin]

I installed a Coriolis meter for measuring NaOH injection into a caustic wash system. It works fairly OK.
The solution density-concentration curve is in the process computer, so operators set the injection rate in lbs/hr of pure chemical NaOH. A Math block translate flow and density readings into NaOH mass flow rate. Typical concentration ranges 10 to 20 Baumé.

However, some troubleshooting was required with this NaOH application. We had to introduce a correction factor for density reading. Problem source has been not yet identified. Density signal dropped to values, say water density = 0.97. So we adjusted the density value in the DCS Math Block. That happened after a few weeks of commissioning. Now it´s OK, after a year.
Water hardness have something to do with signal drift?... It would be good to hear someone else´s opinion.

I don´t know what to say regarding HCl. We could foresee corrosion concerns.

Good Luck

 

[bimr]

One would suspect that you would have to do some correction if Baumé of the solution changed.

10.2 Baumé has a s.g. of 1.0758
19.2 Baumé has a s.g. of 1.153

Most of the hardness in the water will precipitate into solids at the pH values expected.

The error may be caused by the device used to measure the Baumé of the caustic.

 

[gedomin]

We test the density reading by injecting fresh water for a while. Water temp = typicaly 30F. Expected reading should be very close to 1. This is our reference for setting a correction factor, in the DCS. Correction factor option was cheaper than running a rigorous calibration of the meter.

Baumé value is calculated from the same density reading, and displayed, just for operators comfort.
NaOH density-concentration curve works as an internal algorithm that outputs a setpoint value, for a chemical injection pump stroke adjustment. This is a flow control loop... and it works, basically.

By now we continue living with the question in mind. Why the density error? This was a recently commissioned small unit.

 

[bimr]

Perhaps you are using the incorrect compensation factor due to temperature effects. The factor is different for caustic than water.

 

[jmw]

I shall be commissioning a densitometer for dilution control duty next week. It replaces an earlier coriolis meter.

I have chosen a vibrating tube density meter which has bellows at either end of the tube (stainless steel, triple ply). The bellows ensure that the temperature effects are minimised (as temperature changes it would otherwise introduce added stress into the tube as well as increase the contained volume of fluid).

Having said that and having had it calibrated traceable back to national standards per the end users request, during commissioning it can only be tested against the operators local standard. I have a suspicion that this will probably be a hydrometer though if I is an Anton Paar hand-held density meter I will be happy.

The point is that in some process applications it is not the absolute accuracy that is important but the repeatability and the ability to fit the measurement to the existing off-line sample testing regime that has been in place for some years and around which plant management will have been optimised.

In such cases a simple density off-set value is quite acceptable if appropriately applied.
In my case the frequency is converted to density using:
Density = K₀ + K₁T + K₂T²
K_n are constants and T is the periodic time (1/f).
This calculation is then corrected for Youngs Modulus and various other small but measurable effects to arrive at the density at the process temperature.
A matrix is then used to convert the density at process temperature to the density at a reference temperature. It is then optional to use a special expression or to simply scale the resultant density to obtain the concentration (concentration may not be linear with density).

When being commissioned the concentration determined by the density meter will be compared to that determined by the local off-line method.
To make a simple density off-set correction it is necessary to add or subtract the density difference from or to K₀ i.e. not to adjust the reference temperature density but the density determined at the operating temperature.

If it is a simple offset error then this is OK but if the error varies with concentration or with temperature then it is the referral matrix that must be adjusted i.e. the method by which the density at the process temperature is corrected to the density at a reference temperature. In the case of NaOH the source data is easily obtained but it may also depend on the method used in the software.

Different manufacturers use different techniques.

(See the FAQ section for caustic dilution control.)


JMW

http://www.ViscoAnalyser.com