Leveraging Multivariable Sensing

Long ago in the realm of process control, engineers had few options for gathering process variable data, especially if information had to be monitored within a confined area. Sensors and their associated electrical wiring and componentry were physically large, cumbersome, and not terribly sophisticated.

By Dick Johnson, CONTROL ENGINEERING June 1, 1999


Process control & instrumentation

Multivariable sensing

Multivariable final control elements

Sidebars: Loop monitor opens multivariable transmitter doors

Long ago in the realm of process control, engineers had few options for gathering process variable data, especially if information had to be monitored within a confined area. Sensors and their associated electrical wiring and componentry were physically large, cumbersome, and not terribly sophisticated.

As technology advanced into the age of electronics and, as sensors and microprocessors shrank, the possibility of getting “10 lb of instrumentation into a 5 lb bag” became reality. But, where have these technological advances taken process control?

The evolution of the so-called “smart” instrument has given control engineers a better look at and, in many cases, feel for the process. However, attributes such as self-diagnostics, remote calibration and reranging, and digital communication capability have not expanded measuring capability, it took the development of multivariability to do that.

Massaging numbers

Ability to collect more than one measurement and transform them through algorithmical manipulation into the required control attribute is one key to these instruments’ usefulness. For instance, to get the tight control of natural gas and oxygen flow level necessary to fuel its production furnace, The Corning Co.’s (Fall Brook, N.Y.) glassware plant replaced two pressure transmitters and an RTD wired to a flow compensating PLC with a Fisher-Rosemount (Austin, Tex.) Model 3095 MV multivariable transmitter.

Computing mass flow from a single instrument, allowed Corning to perform a full dynamic compensation calculation which resulted in greater flow accuracy and better process control. According to Tom Heath, Corning’s project engineer, the multivariable instrument can compensate not only for changes in density but also changes in thermal expansion factor, pipe ID, and gas expansion factor. “The result is less gas and oxygen fluctuation and elimination of furnace damping due to quicker and more accurate flow information. The result is finer control of the burners and improved quality of Corning’s glass,” Mr. Heath explains.

Using multivariable capability to compensate a pressure measurement signal for both ambient and process temperature changes is important in many processes. Because of this need, Dynisco Instruments (Sharon, N.Y.) originally developed such an instrument for use by DuPont in several of its fiber spinning plants in the South East U.S. The device’s onboard computational ability allows the IPX Series transmitter to compensate for both fluctuating ambient and process temperatures. Pressure output is self-corrected maintaining 0.15% accuracy for an operating range of 25-350ÞC.

During the continuous polymerization process used in the fiber-spinning operation, control of both process pressure and temperature are critical to fiber quality. For a polymerization operation optimized at 290ÞC, process conditions can change temperatures up to 30 ÞC, a totally unacceptable variation. With accurate compensation, resin melt pressure could be accurately controlled with the result of improved production quality and yields.

Control and the ‘complicated’ variable

Not all onboard multivariable transmitter computing power must deal with simple compensation or straightforward flow calculations. In applications such as food processing, control situations that involve unusual process variables often arise.

One such application is at a sugar cane manufacturer, Açucar Guarany, located in Olimpia, State of São Paulo, Brazil. At this operation, accurate control of sugar concentration in processed syrups and juices was necessary to maintain concentration limits, thus lowering costs and improving product quality.

Smar International’s (Houston, Tex.) DT301 Touche intelligent concentration/density transmitter uses a capacitive-type differential pressure transmitter coupled to a pair of pressure repeaters immersed in the process. A temperature sensor located between repeaters compensates for temperature variations. Dedicated software that leverages the device’s internal computing power uses special algorithms and the various inputs to calculate fluid density expressed (depending on the process) in degrees Brix, degrees Gay-Lussac, degrees Baume, precent concentration, etc.

With the good, comes some bad

Industrial Thermal Systems (ITS, Cincinnati, O.) is a Honeywell Inc. (Phoenix, Ariz.) OEM. ITS builds combustion systems (burners) for many process industries. ITS uses Honeywell SMV 3000 multivariable transmitters with orifice plates and Venturi meters to measure both combustion air and fuel flowrates, accurately controlling air/fuel ratio and reducing both NO x and fuel consumption.

According to Robb Jackson, president, ITS recently used SMV 3000s for this task on two separate aluminum melters that the company manufactured. At 200,000 lb/hr throughput, ITS was able to decrease a unit’s natural gas consumption up to 29% through better control of the air/fuel ratio. Mr. Jackson says, “The only problem with using the multivariable transmitters has been the ‘learning curve’ associated with new technology. Our field technicians understand pressure and temperature compensation and how flow rate is proportional to the square root of differential pressure. However, it just takes some more time to understand the specific flow calculation inside the SMV 3000 and how to configure it—a small price to pay for such increased performance.”

End of the line

Final control elements like valves and motors also have benefited from the adaptation of multivariablity. Additionally, the miniaturization of sensors has helped to solidify the worth of this technology.

Case in point, Dodge, a business unit of Rockwell Automation (Greenville, S.C.), offers EZLink, an online, real-time, networked monitoring system that incorporates an integral measurement and communication module that can be mounted on bearings, speed reducers, pumps, and motors to check them for temperature, speed, and vibration. According to Mike Kretzschmar, electrician at Vulcan’s Cherokee Limestone Quarry in Cherokee, Ala. which uses makes extensive use of these devices, “The maintenance crew no longer depends on timing-dependent ‘walk by’ checks and off-line analysis. We now can accurately predict and prevent problems in the conveyor drives at this 800 ton/hr facility before they cause major disruptions.” Devices can be connected via DeviceNet and configured using Microsoft Windows-based proprietary software. A dedicated monitor can be adapted to the system.

So called “smart” technology has given control engineers the ability to track multiple variables and manipulate outputs into many more. As sensors become smaller and onboard computers become more powerful, the outcome will undoubtly be expanded variable monitoring. Can an omniscient system be far behind?

Loop monitor opens multivariable transmitter doors

One difficulty when migrating to HART-based multivariable instruments is not all distributed control systems (DCSs) can take advantage of the digital data that “ride” on the 4-20 mA loop. According to Moore Industries’ (Sepulveda, Calif.) Matt Moren, senior applications specialist, these problems often occur in plant upgrades.

Users are often faced with situations in which—say a multivariable mass flow transmitter—can send its mass flow analog signal to the DCS but needs an alternate variable such as density measurement. Additionally, the user may want contact closure alarm backups should process variables go high, or if the transmitter was not behaving properly. Often budgets (no surprise) do not allow a DCS upgrade.

One solution is a “smart” transmitter interface monitor, such as Moore Industries’ SPA, that can extract digital information from the 4-20 mA smart transmitter loop. These devices mount transparently on a multivariable transmitter’s 4-20 mA loop. The devices can break out required digital information and send a 4-20 mA signal proportional to the DCS. It can also be be configured to provide alarm outputs for various process variables as well as set to monitor the “health” of the multivariable transmitter. Mr. Moren continues, ” By using the SPA to access HART digital data, users are able to get the additional information and protection they needed, yet keep their existing DCS in place.”