Q: Are piezoelectric switches really safe in hazardous areas?

I understand that in the vibrating, tuning fork-type level switch, a piezoelectric crystal is generating an oscillating frequency. We have a Class A product stored in a tank, which requires a Zone 0 installation, and I need a level switch that is certified for Zone 0. My question is, even when it's certified for Zone 0, is the use of such a frequency-generating crystal design good practice? Why does a frequency-generating crystal in not represent a risk in a hazardous area? What type of protection has been provided to achieve that?

I understand that intrinsic safety (IS) applies to wiring and terminals associated with the enclosure—is the section between the tuning fork and the enclosure intrinsically safe? Is the transmission of frequency from a piezoelectric crystal safe when the switch is used in an application involving a Class A product? How is the safety achieved?

M. Ulaganathan, instrumentation design engineer / ulaganathan.inst@gmail.com

----------

A: IS certification looks at the energy of the whole circuit, including wiring and the end device. All components have to be assessed.

Essentially, the piezoelectric crystal is assessed as an energy storage device (i.e. inductance and/or capacitance). The energy storage equations are (1/2)LI2 and (1/2)CV2. The energy storage of the end device is added to that from the wiring calculation. The total energy storage has to be less than that required for the specific IS calculation. A piezo crystal would not have any inductance, so all the energy would be stored in the equivalent capacity.

This is exactly the same as for any end device (e.g. transmitter, positioner, solenoid). Any end device has to be certified to be IS (except for simple apparatus that do not generate more than 1.5 V, 100 mA and 25 mW). The certification then provides the energy ratings for that device, which is used in the IS calculation. In practice, this information can come in a variety of ways. There is a lot of information from the barrier vendor websites on certification, what the data means, and how it is used.

Note that the energy is coming from the power supply in the equipment room. In this application, the piezo is not generating power; it is only storing energy. It is moving in response to applied electrical energy. We are not vibrating it externally to produce energy.

The most likely hazardous area protection technique for Zone 0 is IS. Intrinsically safe designs ensure that the energy level of the instrument and associated wiring and power supplies is below the level that would ignite the potentially flammable atmosphere.

IS is an established design code and practice. It has been around for more than 40 years and is recognized by the various electrical standards (ATEX, IEC, NEC, etc). The latest edition of Liptak’s handbook has a chapter written about the use of instrumentation in explosive atmospheres.

Vendors also provide a lot of literature explaining various explosion-proofing techniques. For example, see www.intec-automatizari.ro/custom_images/Resurse/Principles_of_Explosion_Protection_2012.pdf. Also look at the Pepperl+Fuchs website for their literature.

Simon Lucchini, chief controls specialist and Fluor Fellow in safety systems / Simon.Lucchini@Fluor.com

----------

A: The non-scientific response is, if you're not comfortable with something, don't use it. A device rated for an application should be safe. A rated device probably requires special installation requirements for it to be safe. Installation details must be followed and the installation inspected by the proper person. There are plenty of level switches based on other technologies; you could look at them and find something you're comfortable with.

Cullen Langford / CullenL@aol.com

----------

A: If you're interested, Micronor offers an inherently safe fiber optic microswitch, model MR386. The switch is an entirely passive, all optical, and non-electrical sensor. The remote controller outputs are inherently safe optical radiation. A suitable, inherently safe level limit switch can be constructed using such a fiber optic microswitch.

Dennis Horwitz / dennis@micronor.com

----------

Q: How can we use a level gauge to set a flow alarm?

What is the easiest way to implement a level decline alarm? We have a tank that supplies caustic injected to a process. The caustic flow doesn't have a flowmeter, but the level decline is fairly consistent. We want to alert the operator if the caustic pump suddenly trips and there is no caustic flow to the process by using the level decline. How do we implement an alarm to let the operator know that the level drop is abnormal? Note that the level doesn't stay constant if a pump trips, as two pumps are operating in parallel most of the time. We want to alert the operator when at least one pump has tripped.

The level is detected by an electronic guided wave radar (GWR) transmitter. There is no pressure transmitter on the pump discharge, only a pressure gauge. The level decline (rate of level dropping) is pretty constant, with the rate being low if one pump is running, and about double if two are running.

Yee Kiat / tanykiat@gmail.com

----------

A: If you have a DCS/PLC control system, the control circuits in the motor control center (MCC) for the pumps probably have auxiliary contacts for remote indications. These can generate alarms to signal if pumps stop.

You can also use an algorithm in the DCS/PLC to convert a change of level into a flow rate, both when one and when two pumps are running, as volume rate flow. Alarms could be set for these.

H.S.Gambhir / Harvindar.S.Gambhir@ril.com

----------

A: Level is unlikely to be a sensitive method of detection, as you will have to compare an “old” reading against a “current” reading of the level sensor. Assuming the tank is of constant cross-section, you can store a value and convert it to volume (level*area). Store this value for (say) one hour, and repeat the measurement and calculation. Difference between the two will be an average flowrate over the past hour. Store the current value in place of the old value and repeat. As I said, insensitive.

An alternative, faster method is to fit the tank with a much smaller diameter gauge tube, and feed the pump(s) from the gauge tube, measuring the level from the gauge tube. Intermittently close the connection from the tank to the gauge tube and time the drop in level, then reopen the tank connection. This does give you an increased risk of failure, however, as the valve may fail to open.

Why not bite the bullet and fit a magnetic flowmeter—they are not all that expensive.

Ian H. Gibson, process, control and safety engineering consultant / gibs0108@optusnet.com.au

----------

A: One of the key rules of measurement is that indirect measurements are to be avoided whenever possible. You're looking for flow, not level change. If you want to alert the operator if a pump has tripped, then alarm on the run status from the motor. The power/current measurement of the motor will give a reasonable indication of flow from the pump.

I would not use the rate of change; it will be problematic to set the right time intervals to do the rate of change. It can either not alarm when needed or give spurious alarms.

A better solution is to add a power or current meter to the motor starter. This will give an indication of the caustic flow. Since you're looking for a change, the absolute accuracy will not be a problem. The advantage over the motor run status is that it will detect a blocked discharge from the pump (i.e. power will go down). A power meter would be better than a current meter since it's independent of the current vs. phase angle characteristics of the motor (for example, some motors have strange current characteristics when consuming lower power).

Either a power or current meter are relatively inexpensive, and don't involve process plant modifications for installing a flowmeter. It is also a direct measurement of flow.

Since the rate of change of level is a programming exercise, you could just try it. You can then confirm for yourself if it gives a reliable/consistent result or not. It would be a quick and cheap change.

Simon Lucchini, chief controls specialist and Fluor Fellow in safety systems / Simon.Lucchini@fluor.com

----------

A: The best way to do it may be to program a rate-of-change logic. Some DCSs have embedded the rate-of-change variable out of the analogic variable block (PCS7, for instance). When I do this, I typically capture several samples in a period of time (for instance, 10 samples every six seconds), and then do a minute average with the corresponding time and unit conversion.

If we're talking about a slow ramp, this can be done using two samples in a minute. The program has to consider the pumps running, and also hold the last value if the tank is being filled. The tricky part is to have an accurate calculation of the volume based on the tank geometry, which can be verified using a calibration column if available. If the value needs to be obtained as mass flow, a density sensor is also needed to convert volume to mass.

F. Alcala / AlcalaF@cdmsmith.com

----------

A: Connect the level transmitter to a controller and generate a calculation of how fast the level changes based on one or two pumps operating. That way, the change in level will be measured against the setpoint, which is based on the calculated change in level when the pumps are operating. Set an alarm setpoint to this rate of level change.

Alex (Alejandro) Varga / vargaalex@yahoo.com 

----------

This column is moderated by Béla Lipták, automation and safety consultant, who is also the editor of the Instrument and Automation Engineers’ Handbook (IAEH). If you have an automation related question or if you feel qualified to contribute to one of the volumes of Mr. Lipták's handbooks, send a letter to: liptakbela@aol.com.