Process manufacturing usually requires temperature measurements. These range from finding the temperature of fluid (liquid or gas) within a pipe or vessel to determining how hot an equipment surface is. Either contacting or non-contacting methods can provide readings. So, let’s look at them and their appropriate use.
A direct measurement requires immersing a temperature sensor in the medium, allowing heat to flow through a thermowell or other component to the sensing element. The most common temperature sensors used for direct measurements at plants are thermocouples (TCs) and resistance temperature detectors (RTDs). Various references and articles, e.g., “Select the Most Suitable Temperature Sensor,” compare and contrast these two technologies. For the purposes of this discussion, we’ll lump them together and simply refer to them as temperature sensors.
Typically, these temperature sensors are protected by a stainless steel sheath and mounted permanently in a thermowell to get a specific measurement from the process. The most appropriate location and mounting depends on the application. The signal from the sensor can go to a local display, an automation system or both.
An alternative approach involves contact but not directly with the fluid itself. Instead, it measures the surface temperature of a pipe and infers how hot the fluid within is.
Most non-contact measurement methods determine temperature by analyzing the infrared (IR) radiation thrown off by a surface using a combination of optics and detectors to zero-in on the area of interest. Depending on its sophistication, a device can take readings across a great distance or of a very small and inaccessible area. Portable devices are very common and useful for a wide range of applications. However, sensors also can be mounted permanently and send data to the plant automation system.
Non-contact measurement has a number of attractions:
• Taking a reading can be extremely fast, effectively instantaneous.
• It can measure surfaces that are poor heat conductors.
• No marring or contamination of the surface being measured occurs.
• The subject can be moving.
A rotary kiln in a lime or cement plant exemplifies an ideal application. A non-contact instrument can monitor the outside jacket at the hot end to watch for degradation of the refractory lining. It can do this from a safe distance while the kiln is moving.
Of course, there are limitations:
• Non-contact technologies can’t measure liquid and gas temperatures directly.
• Some substances have peculiar emissivity characteristics that can cause incorrect measurements.
• Highly polished and shiny surfaces can pose difficulties in getting accurate readings.
• Airborne contaminants can obstruct the view, particularly over long distances.
Now, let’s look at how contact and non-contact measurement methods apply to specific situations.
Direct Measurement Of Fluids
This method probably is the most common at chemical plants. It uses a sensor to measure the temperature of a liquid or gas — getting the sensor to the desired point in the process is essential but not necessarily easy. The method is intrusive; it requires inserting the sensor into the product stream, typically using a thermowell (Figure 1). When designing a thermowell installation, mechanical issues become the most problematic elements for a variety of reasons.
First, a thermowell is directly exposed to the process and, therefore, is part of the larger process containment strategy. If the process liquid or gas is corrosive or abrasive, the thermowell must tolerate the aggressiveness of the medium; its material of construction must match or exceed the specifications of the piping and other wetted equipment. The thermowell also must withstand the process pressures and temperatures; this requires use of appropriate material cross-sections and mounting techniques to ensure leak-free service, to avoid contamination or an environmental incident.
Second, the thermowell insertion point determines where the actual measurement is made in the process. To get a reading at some specific spot within a pipe or reactor deemed critical, the thermowell must extend to that point. However, that can pose issues. For instance, mechanical or process constraints may complicate obtaining a crucial temperature measurement at the bottom of a reactor. They may necessitate installing the thermowell from a side some feet away and result in a long unsupported structure that can invite problems.
Third, mechanical issues to a large extent determine the response time of the measurement. If the process undergoes fast temperature swings, a thermowell can significantly affect how quickly the sensor sees the difference. Thick thermowell walls reduce the speed at which the sensor discerns a temperature change but may be necessary for structural reasons.
Finally, the thermowell faces stresses caused by fluid movement. Its protrusion into a flowing stream creates alternating pressure cells called vortices. The thermowell experiences vortex-induced vibration. If the frequency of the vortices matches the natural frequency of the thermowell, the resulting side-to-side movement eventually can lead to thermowell fatigue failure. If these vibrations aren’t mitigated and the thermowell isn’t strong enough to withstand them over the long term, it can fracture, causing an opening to the process.