Tutorial: Understanding multivariable sensors
Many smart instrumentation devices can provide more than just one process variable. Is it information you can really use?
One of the benefits of the growing sophistication of instrumentation devices is that they can often give you more than one variable. These bonus measurements are essentially free in that they don't require any additional sensors or penetrations into your process. All they require is a way for you to extract the information.
Multivariable approaches fall into three categories depending largely on the needs of the primary variable.
Most electronic sensors are influenced to some extent by more than one variable. For example, pressure sensors that use capacitive or strain gage technologies are affected by temperature. Consequently, the transmitter for such a device takes a temperature measurement and uses that data to correct the primary reading. Since that measurement is there in the transmitter, it is usually a simple matter to add a way to send it to the control system.
The caution of using corrective measurement data is making sure you understand where it comes from. The temperature in this example is taken where it is needed for the sensor correction and may not reflect the process at all; it just may be reflecting the ambient temperature around the transmitter or the electronic devices. Make sure you understand what it is before you use such data.
One of the most common flow measurement methods is using an orifice plate and differential pressure gage . There are many implementation variations, but the basic concept calculates flow based on pressure readings on both sides of a known obstruction. While the flow measurement only needs the differential pressure value, it isn't difficult to extract the line pressure measurements as well.
Another example is a corrosion measuring sensor . These typically provide outputs for both general corrosion and local or pitting corrosion. It is also simple to get a conductivity reading with such devices.
With the growing sophistication of transmitter electronics, adding calculated values to measured process variables has become far simpler. Coriolis flowmeters use this technique, and can calculate a range of variables from the three that are actually measured (see diagram). Probably the most common example of this is setting your Coriolis device to read in gallons per minute, since the device does not provide a measured volume. It calculates volume based on its measurements of mass flow and density. The transmitter can be setup to provide whichever of the available values you need most as the primary variable.
Bear in mind that all the measurements from a multivariable device are connected in one way or another. These are not clusters of unrelated instruments. You won't likely find a spool section of pipe that has a magnetic flowmeter, temperature sensor, and pressure sensor all inserted. The specificity of process needs precludes the possibility of making something like that with wide enough applicability to be commercially viable.
Extracting the data
Most devices are designed to provide the primary reading via an analog signal (4-20 mA) or a digital output. However, if more information is available, you have to find a way to get at it.
• A few devices offer multiple (usually two) analog outputs. This approach certainly works, but requires as many cables as variables.
• The most common method for sending the secondary variable is via a HART signal on top of the primary analog variable. If you use a HART interface or have HART I/O with your control system, you can capture the secondary measurements and use them in any way that's valuable to the process. Complex devices such as Coriolis flowmeters, allow you to choose which output comes over the analog signal and which others are overlaid. There are various types of HART reading devices. Some translate the secondary variables into their appropriate engineering units for display. Others convert them into a second or third 4-20 mA signal for input into a DCS. There are even wireless approaches for capturing the information.
Getting your HART interface setup properly can be a challenge since manufacturers use this mechanism differently. You'll probably need to use the device description to sort through the data available for any given device. If you are counting on using some specific information, make sure it is available, since there is little consistency between different manufacturers even for similar devices.
(Opinions differ on using secondary HART information for anything important. Control system builders generally agree that it is valuable for monitoring, but are more cautious when using it for critical measurements. At the same time, any measurement that is truly critical will likely have its own device anyway.)
• Fieldbus protocols make multiple variables very simple, if you use that networking approach and have suitable devices. Using fieldbus requires minimal setup since all variables, primary and secondary, will be available in the appropriate engineering units. Moreover, they can all be handled with the same importance.
Multivariable sensors can be very useful in the right contexts, but using them to your best advantage does require some homework. As your best first step, make sure you know what your process needs.
-Edited by Peter Welander, process industries editor, PWelander@cfemedia.com ,
Control Engineering Process Instrumentation & Sensors Monthly
Register here to select your choice of free eNewsletters .
Annual Salary Survey
After almost a decade of uncertainty, the confidence of plant floor managers is soaring. Even with a number of challenges and while implementing new technologies, there is a renewed sense of optimism among plant managers about their business and their future.
The respondents to the 2014 Plant Engineering Salary Survey come from throughout the U.S. and serve a variety of industries, but they are uniform in their optimism about manufacturing. This year’s survey found 79% consider manufacturing a secure career. That’s up from 75% in 2013 and significantly higher than the 63% figure when Plant Engineering first started asking that question a decade ago.