7 tips for implementing sensors in processing facilities

Editor’s note: This article is part of the State of Smart Manufacturing theme week.

Manufacturing facilities across a broad range of industries rely on sensors and data acquisition (DAQ) systems to keep a close eye on their operations. From manufacturing the tiniest stainless steel washers to producing high purity bulk chemicals, industrial processing and manufacturing facilities rely on sensors to “see” into their operations.

There are a variety of common challenges factory and manufacturing professionals may encounter when it comes to adding sensors to their machinery and equipment, either as an expansion or initial deployment. Considering some of the common challenges for networks like this can expedite technology selection and deployment. Of course, no two facilities are the same, so expect some additional considerations based on unique application or industry variables.

1. Determine the parameters to monitor

Engineers need to first understand which parameters need to be measured. This varies greatly, depending on the task at hand.

Flow sensors measure flow rates of gases, liquids, slurries and particulates. For gases, the volumetric flow rate of the gas is calculated based on the temperature and pressure of the gas being monitored. For most liquids, factoring pressure to determine volumetric flow rate is not as important as it is for gases. For slurries and particulates, the sensors need to be calibrated based on density, in order to obtain accurate readings.

Temperature sensors are also used to monitor fluids inside process piping, storage vessels and more. This greatly affects media density and viscosity. In addition, cold spots in process piping can indicate plugged lines, whereas hot spots may indicate that an unwanted reaction is taking place. Temperature sensors are also used at the inlet and outlet of heat exchangers to confirm proper operation.

Pressure sensors monitor the pressure inside a vessel or piping system. Pressure sensors can help identify runaway reactions or overpressure and can also be used in place of, or with, level sensors to monitor feedstock inventory. A variety of level sensors and transducers are available, some of which are non-contact, such as ultrasonic sensors. A leak detector outside of the storage tank can determine if there has been a breach to the wall of the storage tank causing its contents to be released.

For mechanical equipment such as the pumps and compressors that enable material movement and control, vibration sensors can be useful to ensure the equipment is running to spec.

2. Determine the metrics needed and expected ranges

Equipment operating out of range for an extended period of time could potentially damage components, creating manufacturing bottlenecks and repair needs.

The anticipated operating range of the machinery needs to be considered, along with high and low tolerances that might indicate machinery malfunctions. Data sheets supplied by the equipment manufacturer should provide intended operating specs.

A mismatch in the units of measure between the sensor and equipment may lead to meaningless data, if the mismatch is not identified and addressed. In some cases, sensors will need to be recalibrated for the correct units — converting between measurements may not correct the issue.

The sensor also needs to be responsive to small as well as large changes within its detection span. This ability of the sensor to accurately detect changes at the low end of its detection span to the upper end of its detection span is known as the turndown ratio.

3. Identify alerts needs for integrated machinery

Once the proper machinery operating range has been established, users can program alarm setpoints for the sensor and system. If the data sheet is unavailable, or if the application is unique, the system can help establish a machinery profile of normal operation, and then compare future measurements versus baseline.

The sensor sends data back to the DAQ system, and depending on edge computing power, the sensor DAQ can trigger an alert. Critically, process engineers need alerts to go off before the sensor measures defective equipment. This could be if a threshold reached is multiple times within an interval, or if parameters indicate a trend that could lead to malfunction.

The response to the alert setpoint being met can be configured at the sensor itself or at the DAQ system. Audible and visual alarm settings are sometimes prebuilt into the sensor. More commonly, alerts are issued digitally, available to users as an email or test message via digital devices.

4. Ensure environmental compatibility

Sensor materials and construction need to be compatible with the content of its operating environment. Where the sensor is being installed will ultimately determine how rugged it needs to be.

In chemical and process environments, sensors need to be inert in the presence of chemicals, fluids, gases, and cleaning agents and processes. Some sensors are specially designed for liquid contact, such as sump level sensors and tank level transmitters.

Other sensors are designed to handle and detect high vibrations from equipment without damage. Intrinsically safe sensors are designed to safely operate under conditions that could potentially lead to explosive conditions, such as Class 1, Division 1 environments. The external environment is also another consideration If the sensor is being installed outdoors, it will be susceptible to the elements, temperature swings and potentially even fouling from animals or insects.

5. Determine how frequently the data needs to be reported

For mission critical systems, such as reactors, data needs to be generated instantaneously. Any lag in data transfer and collection could result in failing to identify process errors or malfunctions, leasing to potentially catastrophic results. For other systems, such as storage vessels or conveyor belts, data transfer may not be as critical.

The manufacturing and process equipment needs to be evaluated to ensure that the data transmission and communication process aligns with the critical nature of the equipment. A network of dozens of sensors can create thousands or even millions or data points each day. It is an ocean of data that can be overwhelming for human analysis. In addition to the data storage needs, right-sizing sensor reporting is an important consideration.

6. Estimate the needs for interoperability

Many industrial facilities have accumulated a collection of different types of sensors and DAQ systems over the years. If possible, transitioning to a single system that communicates across multiple sensors and a single DAQ service can be highly effective. However, it is also costly and not necessarily pragmatic for manufacturers that are trying to quickly adapt their operations without high disruption.

A potentially cheaper workaround is to retain existing sensors and the DAQ systems that they are linked to as-is, but to implement an interface that can translate multiple types of data into a format that is of value to the organization. It is also possible to purchase sensors and systems that support multiple communication standards. This enhances interoperability in internet of things networks by enabling products to communicate with a broader range of devices. Dynamic multi-protocol support allows a device to switch between different standards as needed, ensuring seamless communication across devices using various protocols.

7. Evaluate potential signal interference

It’s important to build modular systems that can grow as the facility grows when completing hard wire installations. For simple systems, signal interference between sensors may not be an issue. However, for more complex systems, signal interference needs to be factored in if sensors are installed too close together, or if the environment includes other wireless technologies. This interference will corrupt the data being collected by the system. Sensor data sheets provide valuable information on spacing requirements for certain types of sensors to avoid interference. It is also possible that modular radio frequency shield solutions could be used, such as peel-and-stick shielding materials.

Another way to overcome signal interference is to hard wire all sensors back to the DAQ system. Although complex, this will reduce the occurrence of signal interference as additional sensors are added to the process, although it may inhibit system growth and adaptability.

Summary

Sensors and the data they provide offer critical insight into many industrial processing facilities. Often, engineers are forced to integrate machinery ad-hoc, as new technologies become available or efficiency demands additional data.

With a little forethought and considerations for metrics needed, environmental and system compatibility, along with many of the other factors listed above, the sensor selection and installation process can be a little bit easier.

It’s important to plan ahead for a successful installation of these sensors and the associated DAQ system. As discussed above, there are several tips to keep in mind to ensure a successful implementation of the sensors and DAQ system in industrial processing facilities. By following these tips, these facilities can ensure that the data being transmitted and collected is accurate and timely,

About the author

Simone Ammons has led a 15-plus year career as a process engineer spanning the oil and gas, chemical and pharmaceutical industries. She now leverages her extensive engineering background as a technical writer to concisely articulate complex concepts into a format that is easy to understand.