# News

### Industrial Temperature Measurement

Industrial Temperature Measurement

The term temperature supposedly originated from the Latin word ‘tempera’, which means ‘moderate or soften’.  Technological processes need accurately controlled temperatures. Physical parameters and chemical reactions are temperature dependent, and hence temperature control is of major importance. Temperature measurement in the present day industrial environment encompasses a wide variety of needs and applications. Temperature is a very critical and widely measured variable for several technological processes. Many processes need to have either a monitored or controlled temperature, and for accurate temperature control, its precise measurement is needed.

Temperature is the measure of average molecular kinetic energy within a substance. It is defined as the degree of heat or cold which an element exhibits. This concept comes from practical evaluation of the thermal behaviour of substances, most commonly water. A more sophisticated definition for the measurement of temperature is the measurement of the average thermal energy per molecule contained in a material. The unit used to describe the thermal energy per molecule of a certain material is the degree of temperature.

Temperature is a fundamental physical concept which consists of the three basic quantities of mechanics namely (i) mass, (ii) length, and (iii) time. Temperature is an expression which denotes a physical condition of matter. Yet, the idea of temperature is a relative one, arrived at by a number of conflicting theories. Classic kinetic theory depicts heat as a form of energy associated with the activity of the molecules of a substance. These minute particles of all matter are assumed to be in continuous motion which is sensed as heat. Temperature is a measure of this heat. To standardize on the temperature of objects under varying conditions, several scales have been devised.

The decisive starting point for a general temperature scale is the indispensable requirement for a reproducible scale, independent of the special characteristics of the materials used. In addition, the entire temperature range is to be applicable without restrictions, actually, from the lowest to the highest temperatures. This is the only way to ensure the transferability of measurement results.

Today, the applicable scale is the 1990 international temperature scale (ITS-90). It is the result of improved knowledge of thermometry from the first scale, dating from 1927, through to the present. It is based on fixed temperature points (themselves based on the phase transitions of pure substances), instruments (thermometers) and formulae for interpolation between the fixed points or for extrapolation. This scale necessarily evolves over time due to the improved accuracy of the fixed-point temperatures, bringing the scale value closer to the thermodynamic temperature. It is possible to identify two categories of temperature measurement units. These are (i) absolute units, and (ii) relative units.

Absolute units – Absolute units start from absolute zero, theoretically the lowest temperature possible. It corresponds to the point where the molecules and atoms in a system have the lowest possible thermal energy. Kelvin (international system) represented by the letter K without any degree symbol is the thermodynamic temperature unit and is defined on the basis of the triple point of water, 273.16 K (or 0.01 deg C).

Relative units – Relative units are compared with a physical and chemical process which always produces the same temperature. Degrees Celsius (international system), also called degrees centigrade and represented by the symbol ‘degree C’. This measurement unit is defined by assigning the value 0 degree to the freezing point of water and the value 100 degree to the boiling point of water when both measurements are taken at a pressure of one atmosphere. The scale is then divided into 100 equal parts in which each corresponds to 1 degree. Degrees Fahrenheit (international system) measurement unit is based on divisions between the freezing and evaporation points of ammonium chloride solutions. In this scale, the zero and hundred corresponds to the freezing and evaporation temperatures of ammonium chloride in water. As per this scale 32 degree F corresponds to the melting point of ice and 212 degree F corresponds to boiling point of water. The difference between the two points is 180 degrees which is divided into 180 equal portions, determines the degree Fahrenheit.

ITS-90 – It is defined for temperatures above 0.65 K and upto the highest temperature measurable according to Planck’s law for monochromatic radiation. The temperature measured with this scale (T90) is the closest to the thermodynamic temperature. This means it is universal. ITS-90 covers several temperature ranges. For each temperature range, it therefore defines fixed temperature points and a specific instrument for measurement and interpolation between these fixed points. The fixed temperature points correspond to phase transitions in pure substances. For example, the freezing points of zinc (Zn), tin (Sn) or silver (Ag), the melting point of gallium (Ga) or the triple points of oxygen (O2), mercury (Hg) or water.

The SI (Systeme International d’Unites) primary standard for temperature is the triple point of pure water. The triple point of a pure substance is defined as that temperature and pressure at which all three phases (solid, liquid, and vapour) are in equilibrium in a closed vessel. The triple point of pure water occurs at +0.0098 deg C and 4.58 mm Hg pressure. This is a single, unique point in the pressure-temperature phase diagram for water. The transition (melting) temperature between solid (ice) and liquid water above 4.58 mm Hg pressure is a decreasing function of pressure. Hence the melting point of ice is slightly pressure dependent at atmospheric pressure. The pressure dependence is sufficiently small, so that an ice bath made with double distilled water (both ice and water) can serve as a secondary standard for 0 deg C.

Temperature is one of seven basic values in the current SI-System of Units and at the same time, probably the most important parameter in measurement technology. Temperature measurements can be roughly divided in three application categories (i) precision temperature measurements for scientific and basic research, (ii) technical temperature measurements for measurement and control technology, (iii) temperature monitoring using temperature indicators. The goal of the technical temperature measurement is to strive for a practical solution for every application requirement, which is to be an optimum for the required measurement accuracy at acceptable costs.

Reproducible temperature points established by physical constants of readily available materials define the scale. Interpolation between these fixed points is made by platinum (Pt) resistance thermometers when the temperature is below 1,000 deg C and by Pt-Pt and 10 % rhodium (Rh) thermocouples (TCs) when it is higher. The national level Standards Bureaus have capability for calibrating temperature- measuring devices against these primary temperature points. These devices are secondary standards which are then used by manufacturers and users to calibrate other equipment. The capability of the national level Standards Bureaus for calibrating temperature-measuring devices is illustrated in Fig 1. The figure also shows the error (uncertainty) of other thermometers at different temperatures.

Fig 1 Calibration of temperature measuring devices

Temperature is a more easily detected quantity than either thermal energy or heat. In fact, when there is a need to measure either thermal energy or heat, it is met by measuring temperature and then inferring the desired variable based on the laws of thermodynamics. There are many different ways to measure temperature, from a simple glass-bulb Hg thermometer to sophisticated infra-red (IR) optical sensor systems. Like all other areas of measurement, there is no single technology which is best for all applications. Each temperature measurement technique has its own strengths and weaknesses.

The development of temperature measurement has and is occurring in parallel with the technological developments. Thereby only a portion of the new measurement methods have replaced the older ones. They have actually expanded their scope allowing temperature measurements to be made in areas where in the past none or only very restricted ones were possible.

Many physical and chemical phenomena and physical constants are found to be functions of temperature and thus, can be used to measure temperature. Temperature dependent properties and constants include resistance, dielectric constant, and the magnetic permeability and susceptibility (of paramagnetic salts). Other temperature sensitive phenomena include linear and volume expansion of solids and gases, generation of the Seebeck (thermoelectric electro-motive force) effect by thermocouples and the generation of Johnson (thermal) white noise by resistors.

Temperature measuring devices

There are several methods of measuring temperature which can be categorized as (i) expansion of a material to give visual indication, pressure, or dimensional change, (ii) electrical resistance change, (iii) semiconductor characteristic change, (iv) voltage generated by dissimilar metals, and (v) radiated energy. Tab 1 gives some major types of temperature measuring devices along with their range and error limits.

 Tab 1 Major types of temperature measurement devices Sl. No. Measurement methods Range in deg C Error limits From to 1 Mechanical thermometers Liquid filled glass thermometer Non-wetting liquid -38 630 As per standards Wetting liquid -200 210 2 Indicator thermometers Bimetal thermometer -50 400 1 % to 3 % of the indicator range Rod expansion thermometer 0 1,000 1 % to 2 % of the indicator range Liquid filled spring thermometer -30 500 1 % to 2 % of the indicator range Vapour pressure spring thermometer -200 700 1 % to 2 % of the scale length 3 Thermocouples Cu-CuNi, Type U, T -200 600 0.75 % of the reference value of the temperature Fe-CuNi, Type L, J -200 900 NiCr-Ni, Type K, NiCrSi-NiSi, Type N 0 100 PtRh-Pt, Type R, S                                            10 % Rh (S); 13 % Rh (R) 0 1,600 0.5 % of the reference value of the temperature Pt Rh30-PtRh6, Type B 0 1,800 As per standards 4 Resistance thermometers with metal resistors Pt-resistance thermometer -200 1,000 0.3 deg C to 4.6 deg C depending on the temperature Ni-resistance thermometer -60 250 0.4 deg C to 2.1 deg C depending on the temperature 5 Semiconductor resistance thermometers Hot wire resistance thermometer, thermistor -40 180 0.1 deg C- 1 deg C                                0.5 deg C- 2.5 deg C                                     depending on the temperature -60 200 -100 400 Cold wire resistance thermometer 200 2 deg C- 10 deg C Silicon measurement resistor -70 175 0.2 deg C- 1 deg C Semiconductor diodes/integrated temperature sensor 160 0.1 deg C- 3 deg C                                 depending on the temperature 6 Radiation thermometers Spectral pyrometer 20 5,000 0.5 % -1.5 % of the temperature, but at least 0.5 deg C to 2 deg C in the range from -100 deg C to 400 deg C Infrared radiation pyrometer -100 2,000 Total radiation pyrometer -100 2,000 Ratio pyrometer 150 3,000 Thermography instrument -50 1,500 Quartz thermometer -80 250 Resolution 0.1 deg C Thermal noise thermometer -269 970 0.1 % Ultrasonic thermometer 3,000 Around 1 % Gas thermometer -268 1,130 Depending on design 7 Optical methods Fiber optic luminescence thermometer 400 0.5 deg C Fiber optic measurement system based on Raman-Radiation 600 1 deg C

Temperature sensors

It is believed that Galileo invented the liquid-in-glass thermometer around 1592. Thomas Seebeck discovered in 1821 that an electric current flows between different conductive materials which are kept at different temperatures, known as the Seebeck effect. The same year, Sir Humphry Davy noted the temperature dependence of metals. Fifty years later, William Seamens used Pt in a resistance thermometer. This favourable choice standardized all future resistance thermometers with Pt being the key element for high precision temperature measurement. The ‘Platinum Resistance Temperature Detector’ is now used to measure from the triple point of hydrogen (H2) (-259.34 deg C) to the freezing point of Ag (961.78 deg C). Pt is particularly convenient for this type of temperature range, as it maintains an excellent stability which hardly alters after repeat use.

In 1932, CH Meyers proposed construction of a ‘Platinum Resistance Temperature Detector’ composed of a Pt wire wrapped around a mica support core inside a glass tube. This type of construction minimizes the wire tension and maximizes resistance. Even though, it is a very stable assembly, the thermal contact between the Pt and the measurement point is weak and a diminished temperature response time is the result. Due to the structure’s fragility, this type of RTD (resistance temperature detector) is used mainly in laboratories today.

The development of temperature sensors was a slow process until the middle of the 20th century. Presently over 20 different types of thermometers are available. In addition, the old practice of using only filled system thermometers, RTDs, or TCs throughout a particular industrial plant is giving way to the practice of selecting each temperature sensor for a particular application, just as each level or flow meter is individually selected.

Non-electric temperature sensors

The important non-electric temperature sensors are described below.

Liquid-in-glass thermometers – The major types have used Hg or alcohol as the liquid. The element Hg is liquid in the temperature range of around -38.9 deg C to 356.7 deg C. As a liquid, Hg expands as it gets warmer. Its expansion rate is linear and can be accurately calibrated. Because of toxicity of Hg and the strict governing laws, the use of the Hg-in-glass thermometer has declined. However, for high accuracy applications, laboratory grade and reference standard models are available with calibration certification to standards.

Bi-metallic thermometers – Bonding two dissimilar metals with different coefficients of expansion produces a bi-metallic element. These are used in bi-metallic thermometers, temperature switches, and thermostats having a range of -70 deg C to 540 deg C. When manufactured as a helix or coil, its movement with a change in temperature can move a pointer over a dial scale to indicate temperature. Dial thermometers ranging from pocket size to 125 mm dials are available with a variety of local and remote mounting configurations.

Many process applications require use of a thermal well to allow for the removal or replacement of the thermometer while the process is pressurized. Other designs include switches for on-off control which range from the simple wall thermostat to more rugged industrial models for simple process control or over-temperature protection. Other configurations include snap disk switches often used for over temperature alarm and control. Rugged units are used in industrial machinery as over-temperature limits.

Filled system thermometers – Filled system thermometers have been used for decades. They have a useful range of 200 deg C to 650 deg C. Applications vary in sophistication from those which are used in heating, ventilation, and air conditioning (HVAC), to rugged industrial units suitable for a variety of applications. The filled system element can operate switch mechanisms for industrial shut-down controls. Some models have an analog output which can be connected to remote locations. Thermal systems filled with Hg or solid materials have all but disappeared since in many countries mandate the removal of any Hg filled devices due to its extreme toxicity.

There are some models which use a vapour fill, but these suffer in applications where the temperature crosses the vapour/liquid point and causes liquification and loss of performance (typical is 0.25 % per 14 deg C). Liquid filled units are the most popular, but consideration is to be given to offsets due to the weight of the liquid head and compensation for capillary length. Thermocouples and RTDs are replacing filled systems in industrial process control applications. The low cost of electronic devices to read the output of TCs and RTDs and to indicate or control, together with the ability to locate the sensor independently of the receiving device, has made electronic methods more attractive.

Bi-state/phase change sensors – These low cost non-electric sensors are made from heat-sensitive fusible crystalline solids which change decisively from a solid to a liquid with a different colour at a fixed temperature depending on the blend of ingredients. They are available as crayons, lacquers, pellets, or labels over a wide range of temperatures from 38 deg C to 1,650 deg C. They offer a very inexpensive method for surface temperature visual verification within around 0.5 deg C. Monitoring minimum and maximum temperatures during shipment of perishable goods is a common application

Electronic thermometers/sensors

There are many types of electronic thermometers as described below. Fig 2 shows the characteristics of these sensors.

Fig 2 Characteristics of electronic thermometers

Thermocouples (TC) – A TC is an assembly of two wires of unlike metals joined at one end designated the hot end. At the other end, referred to as the cold junction, the open circuit voltage is measured. Called the Seebeck voltage, this voltage known as electro-motive force (EMF) depends on the difference in temperature between the hot and the cold junction and the Seebeck coefficient of the two metals. There is a relationship between temperature and output signal. For most industrial applications, the TC has been the popular choice over the years for a variety of reasons. TCs are relatively inexpensive and can be produced in a variety of sizes. They can be of rugged construction, can cover a wide temperature range from 260 deg C to 2760 deg C, and are available in both standard and premium grade models.

Every credible temperature transmitter, indicator, controller, or data logger accepts a direct TC input. For many applications, this is a viable solution. However, TCs produce a very small micro-volt output per degree change in temperature which is very sensitive to environmental influences. Electro-magnetic interference (EMI) from motors and electrical distribution and especially radio frequency interference (RFI) from walkie-talkies can produce dramatic errors in measuring circuits in these instruments. The user is to insist on a noise reduction specification and an RFI immunity specification which minimizes these effects. A top quality instrument offer common mode noise rejection of 100 dB (decibel), normal mode rejection of around 70 dB, and RFI immunity of 10 V/m (volts per meter) to 30 V/m.

There are some applications where a bare TC with an exposed junction can be used either by itself or inserted into a protective well. For most process applications, the TC is produced with a protective outer sheath which uses an insulating material to electrically separate the TC from the sheath and provide mechanical and environmental protection. In some cases, the TC junction is placed in direct contact with the tip of the sheath to increase speed of response. These sensors demand the use of an electrically isolated measurement circuit. Even insulated TCs eventually suffer from a breakdown of the insulation and the TC tip contacts the sheath and associated well. It is virtually assured that a ground loop will be present which causes measurement errors. These errors are usually dangerous in that they normally vary over time and can go unnoticed. Recommended practice is to always use an instrument with full isolation to eliminate this concern. Another consideration for applying TCs is that their low output (about 40 micro V/deg C) limits the minimum span of even the best transmitters to around 35 deg C. RTDs are the usual choice for narrow span applications.

The most misunderstood drawback of using TCs is their inherent drift. The junction of the two dissimilar metals begins to degrade to some degree immediately after production. For some types, used at low temperatures, this can only be a few degrees per annum and can be calibrated out of the system. Other types used at higher temperatures degrade much more quickly. Consideration is also to be given to any TC extension wire which is used for long wiring runs from the field location to the measurement instrument. Not only is its accuracy only about half as good as a TC, but also it is often subjected to harsh environmental conditions as it passes through the plant which causes significant degradation and drift. Some plants replace their extension wire on a regular basis to minimize this effect. Installing a top quality transmitter in the connection head of the TC minimizes many of the problems and concerns described above.

Resistance temperature detector (RTD) – RTDs are constructed of a resistive material with leads attached and usually placed into a protective sheath. The resistive material can be Pt, Ni (nickel), or Cu (copper) with the most common by far being Pt. The relationship between the resistance change of the RTD and the temperature is referred to as its alpha curve. The instrument used with the RTD is to be configured to use the same alpha curve as the RTD or significant errors occur.

As with TCs, there are some applications, where an exposed sensor is suitable and serves the purpose. Normally RTDs are manufactured with a protective sheath which provides a hermetic seal to protect the sensor from moisture and/or contamination. These protective sheaths are offered in a variety of lengths to provide the proper insertion into the process to obtain a representative measurement. The sheathed elements are often installed into a protective well to isolate the sensor from the process.

There are very few applications for a 2-wire RTD since the error introduced by the leads can be a significant error. Measurement circuits which accept 3-wire inputs include a method of minimizing the effects of lead wire resistance as long as the outer legs are equal. However, factors such as terminal corrosion and loose connections can create significant differences between the lead resistances seen by the measurement circuit. A single ohm of difference between the legs is reflected as a 2.6 deg C error. Using a 4-wire measuring circuit eliminates this problem. Direct connection to remote devices with 3-wire extension cable often produces errors which can be significant and which vary with environmental conditions.

Cu RTDs are most commonly used to sense the winding temperature of motors, generators, and turbines. Connecting them to an alarm trip provides an over-temperature shut-down function. Historically, 10 ohm Cu RTDs were the norm and accurate measurements of the small change in resistance change with temperature limited the accuracy to around +/- 1.6 deg C. Many users have now opted for 100 ohm or even 1,000 ohm units to get higher resolution. Such units have a useful range of -50 deg C to 250 deg C. Ni RTDs have declined in use over the years primarily due to their limited range and the more popular Pt RTDs. The useable range is -80 deg C to 320 deg C. Most transmitters and alarm trips still offer the capability to accept Ni RTD inputs.

A concern common to all RTDs is that of error produced by self-heating. RTD measurement circuits measure the voltage across an RTD produced by passing a precise current flow through the RTD. A current flowing through a resistance produces heat which appears as a positive offset to the actual process temperature at the RTD. The lower the measuring current, the less is this heating effect. It is to be minimized by good thermal contact to the process fluid and is less of a concern at higher temperatures. A measure of the quality of an RTD measuring circuit is the amount of measuring current used. Circuits in better transmitters use around 250 micro amperes. This current is typically higher for Ni and Cu RTDs.

Thermistors – Like the RTD, the thermistor is also a resistive device which changes its resistance predictably with temperature. Its benefit is a very large change in resistance per degree change in temperature, allowing very sensitive measurements over narrow spans. Due to its very large resistance, lead wire errors are not significant. However, thermistors have some disadvantages namely (i) it is a very non-linear device and reasonable accuracy is obtained only over narrow spans, (ii) it is quite small and shows errors due to self heating, and (iii) exposure to high temperature causes a dramatic and permanent shift in its output characteristics. Most applications of the thermistor are in commercial and laboratory applications. Few are used in industrial process control. Thermistors are designated by their resistance at 25 deg C with the most common value being 2252 ohms.

Optical pyrometers – These instruments provide a no-touch means of estimating the surface temperatures of hot objects in the range of 780 deg C to 4200 deg C, such as metals being hot-worked, liquid metals, gas plasmas, and furnace interiors. Optical pyrometers make use of the fact that all objects at temperatures above 0 K radiate heat in the form of broadband, electro-magnetic energy. The range of the electro-magnetic spectrum, generally considered to be thermal radiation, lies in the range 0.01 mm to 100 mm wavelength. Objects which are radiating heat are characterized by three parameters which describe what happens to long wave, electromagnetic radiation (heat) at their surfaces. This relation is ‘e = a = 1 – r’ where e is the surface emissivity, which is always equal to its absorbtivity, r is the reflectivity of the surface. An ideal black-body radiator has e = a =1 and r =0 (that is, all radiant energy striking its surface is absorbed, and none is reflected). For a non-ideal radiator, a finite fraction of the incident energy is reflected. The same properties exist for emitted radiation from a black-body.

Practical hot surfaces have non-unity emissivities which are generally a function of wave-length. Thus, the practical, spectral emittance curve for a hot object can have many peaks and valleys. If these irregularities are averaged out, it is often fit a scaled down, black-body spectral emittance curve to the practical curve so that their peaks occur at the same temperature, T. Such a scaled black-body is called a gray-body curve.

In one form of optical pyrometer, shown in Fig 3 (a), a human operator makes a subjective colour comparison of a glowing tungsten filament with the hot surface under measurement. The colour comparison is made easier by optically super-imposing the image of the filament on that of the hot object. When the filament’s colour, determined by its current, matches that of the object, it disappears on the background of the object, and the filament current is read. Since the filament can be considered to be a black-body, its spectral emittance closely follows that of the object when its temperature is the same as that of the object. Often, a red filter is used to convert the colour matching task to one that involves brightness matching. The filament ammeter is calibrated in temperature.

In another form of optical pyrometer, shown in Fig 3 (b), the filament is run at a constant current and brightness. The intensity of the image of the hot object is then varied with a neutral density wedge. Again, a red filter is used to make the null process an exercise in monochromatic intensity matching. The wedge position is calibrated in object temperature, assuming blackbody emission.

Fig 3 Types of optical and radiation pyrometers

Radiation pyrometers – Total radiation pyrometer (Fig 3) is also known as radiometer. Radiation pyrometers are electronic instruments which measure the integral of the gray-body curve. In one form of the radiation pyrometer, the detector is a thermopile in which half of the thermocouple junctions are to be kept at a reference temperature. The thermopile is designed to be a nearly 100 % black-body absorber. Thus, regardless of the shape of the gray-body curve, all of the incoming energy is captured and converted to a temperature rise of the sensing junctions. Total radiation pyrometers can be used with black-body radiation, as well as coherent sources such as lasers.

Other detectors used in radiation pyrometers can include photo-conductors, photo-diodes and pyro-electric detectors. Pyro-electric detectors absorb thermal energy and generate electrical signals. One example of a pyro-electric detector material is the polarized polymer film, poly-vinylidene difluoride (PVDF). PVDF film absorbs strongly in the IR, and a free mounted, 28 mm PVDF film has a response time constant of around five seconds to a step temperature change. All metallic radiation detectors have spectral response characteristics.

An infrared (IR) thermometer is a non-contact radiant energy detector. The amount of radiant energy emitted is proportional to the temperature of the object. Non-contact thermometers measure the intensity of the radiant energy and produce a signal proportional to the target temperature. The physics behind this broadcasting of energy is called Planck’s law of thermal radiation. This radiated energy covers a wide spectrum of frequencies, but the IR spectrum is most commonly used for temperature measurement. IR thermometers capture the invisible infrared energy which is naturally emitted from all objects warmer than absolute zero (0 K). Infrared radiation is part of the electro-magnetic spectrum which includes gamma rays, x-rays, micro-waves, ultra-violet, visible light, and radio waves. IR falls between the visible light of the spectrum and radio waves. IR wavelengths are usually expressed in microns with the infrared spectrum extending from 0.65 micro meters to 1,000 micro meters.

In practice, the 0.65 micro meters to 14 micro meters band is used for IR temperature measurement over a range from -50 deg C to 3000 deg C. IR technology has become a viable and cost effective alternative to TC and RTD measurements in hostile environments like furnaces. The TC or RTD can only measure the temperature of its immediate surroundings and hence cannot measure the actual product temperature. The actual temperature of the product changes due to variations such as line speed, thickness of the product, colour, or roughness. The TC or RTD does not respond to these temperature changes quickly enough to permit close control. The IR thermometer instantly measures the actual product temperature, not the environment surrounding the product.

IR thermometers are ideal for moving targets. They do not interfere with the process. They are also ideal for measuring products with very high temperatures or hostile environments. They can see through windows to measure products in a vacuum furnace or a semi-conductor reactor. IR thermometers can measure targets as small as 0.6 mm in diameter and can respond in 10 milli-seconds (ms) to a temperature change. IR instruments operate at various wave-length bands. Each band represents a series of instruments. The shortest wave-length band utilized is 0.65 micro meters and the longest is 8 micro meters to 14 micro meters. An instrument using a 0.65 micro meters wavelength can measure the higher range temperatures from 700 deg C to 3600 deg C. Instruments with sensitivity to the 8 micro meters to 14 micro meters range can measure down to -45 deg C.

Distance does not affect the measurement. Radiation pyrometers are available which can measure from 0.3 meters (m) to 91 m. However, IR sensors measure the energy from a circular spot on the target, and the size of that spot is a function of the distance between the sensor and target. The farther away from the target the sensor is, the larger the spot. Consequently, distance is limited by the size of the object which is to be measured. Some pyrometers have a low power laser which facilitates proper aiming. Where a direct line of sight to the target is not possible, pyrometers with a fiber-optic connected sensor system often can solve the problem. Fiber optics allows installations into difficult locations upto 9 m from the instrument. They can withstand ambient temperature upto 200 deg C without cooling. The lowest target temperature for these systems is around 700 deg C. For application in steel plants, where the sighting is to be made through dirty small windows, and where there are obstructions in the sight path, or dusty atmospheres, the two-colour or ratio thermometer provides a good solution. This instrument utilizes two detectors operating at two wavelengths to measure one hot target. They can measure temperatures from 250 deg C upto 3500 deg C and, by using sensors with two different wavelengths, can eliminate the interference problems. In addition, they can measure targets which do not completely fill the optical spot size of the instrument.

The installation of an IR thermometer requires attention to detail to ensure successful operation. Ideally, a clear unobstructed line of sight is needed and the target has to be large enough to fill the cone of vision. The spot size required is related to the distance to the sensor.

Solid obstructions have to be removed or eliminated from the field of view by possibly aiming the instrument at a different angle or by using a fiber optic remote sensor system. For applications with sighting windows, the instrument is to be selected to use a wavelength which passes through the window material unchanged. The window is also to be large enough so as not to obstruct the cone of vision. It also has to be kept clean by possibly using an air purge. Smoke, steam, and dust all cause temperature fluctuations.

Most IR thermometers have an electrical feature called a ‘peak picker’. This is a simple circuit which picks the peak temperature and does not allow the signal to decay when dust obstructs the field of view. Flames are also a consideration. Clean gas flames are transparent to most thermometers while in the case of opaque flames, no thermometer can see through them. The actual temperature of opaque flames can usually be measured with the two-colour instrument. Most sensors can operate in environments upto 65 deg C. For hotter ambient environments, water or air cooling is needed. IR instruments are to be calibrated once a year using a certified black-body.

There are line scanners which can produce a two-dimensional thermal image. This type of IR thermometer is used to measure wide web such as hot strip steel. This instrument utilizes one detector and two 45-degree mirrors with one mirror rotating and scanning over a 90 degree angle. For moving targets, the instrument uses a software technique to create two-dimensional thermal images of the moving web. The software provides temperatures at any location on the web and can provide output signals which can be used for closed loop control or stored for future review.

Thermal imagery is a rapidly growing application of IR technology for quality control and many other uses. This imaging system is an IR thermometer which uses a detector called a focal plane array instead of a single sensor. Its functionality is similar to a digital camera, except that, instead of capturing photographic images, each pixel measures temperature. A two-dimensional image is created using software resident within the system. The detector can have as many as 76,000 pixels. The image of the temperature profile can be used for closed loop control of processes or stored away for future analysis. There is one pyrometer model which has bi-directional digital networking communications. Multiple sensors can connect to host systems for monitoring, control, and diagnostics.

Solid-state sensors – These are also known as the electronic IC (integrated circuit) temperature sensors. These are specialized integrated circuits used for sensing temperatures in the -55 deg C to 150 deg C range. The small solid-state sensor converts a temperature input into a proportional current output over a range of -55 deg C to 150 deg C. It is especially suited for PC (personal computer) boards or heat sink mounting for special temperature measurement and control applications where solid-state reliability, linearity, and accuracy are needed. They can be used to determine minimum, average, and differential temperatures, in addition to being used for TC cold junction compensation and temperature control applications. These sensors having low price are gaining popularity. Diodes are typically used in cryogenic applications over a range of -271 deg C to 202 deg C and can be accurate to 0.05 deg C when properly calibrated.

Heat-flow and thermal-conductivity sensors – The accurate measurements of heat flow through thermal insulators and of the thermal conductivity of construction materials are both important. Such measurements are of interest for the purpose of safety and energy conservation. A common heat flow meter design involves the placing of a thin plate of known thermal conductivity on a heat radiating surface. It has been found that the heat flow through these elements is directly related to the temperature difference through them. This temperature difference is often detected by thermopiles, a large and even number of TCs connected in series in such a manner that their high-temperature junctions are on the inside and their low-temperature junctions are on the outside surface of the sensing element. The heat flows which are encountered in different processes range from about 10 kcal/sqm h through freezer walls to about 100,000 kcal/sqm h through the shells of water-cooled electric furnaces. The thickness of the sensor plates is a few millimeters, and the plates are made of rubber, organic materials, or other heat-resistant materials, sometimes contained in a thin, stainless steel disk case.

The heat-flow distribution frequently varies with the direction of heat flow. For example, the heat loss from the top of a steam pipe is found to be much more than through the bottom surface. Such findings are generally explained by noting that heat flow from a surface is not only a function of the surface temperature, but also of the effects of coating. It has been found that if a surface is coated with bright and glistening aluminum (Al) paint, it radiates much less heat at the same temperature as a surface where the coating has worn off. Sensor elements can measure the variations in heat flow at different points on many shells ranging from liquefied petroleum gas (LPG) tank walls to electric and blast furnace shells, and their readings can reveal the erosion of linings as well as other hard-to-detect phenomena.

In other processes, the interest is in measuring the thermal conductivity of heat insulating substances. Thermal conductivity instruments are designed to measure thermal conductivities of solid materials in the range of 0.001 W/mK to 10 W/mK. Typical materials with conductivities in this range are foam, insulation, polymers, composites, glass, silicon, natural fibers, and rock. Higher thermal conductivity samples produce a lower rate of temperature rise because the heat is being conducted away from the interface. The heating element of the probe provides a one-dimensional heat flow. The entire element is to be covered during testing, establishing the minimum flat surface area of 5 mm × 25 mm. The probe calculates the value of thermal conductivity given known values of heat capacity and density.

The hot wire method of thermal conductivity measurement involves the stretching of a thin heating wire through a sample and applying a constant amount of power. The higher the thermal conductivity of the sample, the lower is the resulting surface temperature of the heater wire. Hence, it is possible to read the surface temperature of the heater wire and interpret from that reading the thermal conductivity of the sample.

The thermal conductivity of an unknown substance can be determined by first recording the surface temperature response curve (time vs. temperature) of the wire while the wire is surrounded by a known thermal conductivity material. After that, half of the known sample can be replaced by a material having an unknown conductivity. After repeating the test, the difference in the response curves can be correlated to the thermal conductivity of the unknown substance.

Intelligent transmitters and remote input/output (I/O) – Microprocessor-based temperature transmitters have continued to evolve in sophistication and capability. In present day environment all the credible transmitters use this technology. A transmitter includes an input circuit referred to as an analog-to-digital (A/D) converter which converts the sensor input signal from its analog form into a digital representation for presentation to the micro-processor. The micro-processor performs all of the mathematical manipulations of ranging, linearization, error checking, and conversion. The output stage accepts the resultant digital representation of the current value of the measurement and converts the signal back to an analog signal (D/A) which is typically a 4-20 mA DC current. For some special applications, 0 VDC to 1 VDC or 0 VDC to 10 VDC signals can be used and in others the signal is transmitted digitally using either an open or proprietary protocol. Normally 0-20 mA is used for the standard transmitted signal.

Just as microprocessors have evolved in sophistication, so have A/D converters. Eight-bit resolution devices common in the 1960s provided a resolution of around +/- 0.4 %. In 2000, the first 21-bit resolution A/D was used in a temperature transmitter providing a resolution of +/- 0.00005 %. D/A converters have also evolved with resolutions increasing from 8-bit upto the 18-bit versions used in the better transmitters beginning in 2000. The result of combining these technologies is a universal transmitter which accepts inputs from any TC, RTD, mV, resistance, or potentiometer signal, checks its own calibration on every measurement cycle, has minimal drift over a wide ambient temperature range, incorporates self-diagnostics, and is configured using push-buttons or simple PC software.

The reconfiguration process is quick and convenient, and it tends to allow for lower inventories by making the transmitters inter-changeable. Some transmitters are capable of handling dual RTD elements. This allows for temperature averaging, temperature difference measurement, or automatic RTD sensor switchover if the primary sensor fails in a redundant installation. Due to terminal limitations, these models can only accept dual 3-wire RTDs. Caution is needed to minimize lead resistance differences to reduce the error.

Fieldbus structures – In the 1980s, another level of capability was added to field devices. The ‘Highway Addressable Remote Transducer’ (HART) protocol was added to enable detailed information about the setup and operation of the device to be superimposed onto the 4-20 mA signals. The benefits of remote configuration and access to diagnostics have encouraged the dramatic increase in use of HART enabled instruments. During the same time period, a variety of proprietary protocols emerged supported by many of the larger manufacturers which provided comparable benefits of remote setup and diagnostics. Unlike the HART protocol, these products were limited to use within the manufacturer’s system. In the 1990s, a trend emerged for more open protocols to enable plug-and-play of instruments from varying manufacturers to work as part of a fieldbus structure. Two which have emerged as leaders are Foundation Fieldbus and Profibus. Temperature transmitters, as well as other field and control room devices, are to incorporate the specific fieldbus technology to be used in these systems.

In 1979, Modicon introduced the MODBUS protocol as a means of exchanging data between field devices and controllers. This protocol serves as a means to share data among multiple devices. There are several temperature multiplexers available today which can communicate with a host system using MODBUS, MODBUS Plus, or MODBUS TCP/IP protocols.

Ethernet communication has been used for interconnection of digital equipment in offices and control rooms for many years. Its use is now migrating out onto the shop floor. There are a variety of products introduced beginning in 2002 which interface process measurements, including temperature, over high speed Ethernet links to host systems using OPC (object linking and embedding for process control) servers. There is more to making a measurement than the transmitter itself. The weakest link in virtually all measurements is the temperature sensor. The vast majority of temperature measurements are made with either a TC or an RTD. As with the electronic devices described above, sensors have also established dramatic increases in precision and reliability over the years. Higher purity materials and improved manufacturing processes have provided sensors which more closely match theoretical curves and show lower drift than sensors of the 1980s and the 1990s. It has been generally accepted that the more closely a sensor matched its ideal characteristics, the more accurate the measurement will be since transmitter measuring circuits refer to the ideal data to make the measurement.

Advanced transmitters – More advanced transmitters incorporate the ability to match the sensor to the transmitter to minimize this error. One method uses the Callander van Deusen method, which defines three experimentally determined constants which define the temperature/resistance relationship specific to an RTD. By entering these data from the RTD tag into the transmitter’s firmware, the sensor and transmitter become a calibrated system. Typical accuracies for this technique are around +/- 0.2 deg C. For higher precision, a bath calibration technique is used which allows the transmitter to capture actual values output by the RTD or TC at specific temperatures. This method provides system accuracies of around 0.1 deg C for RTDs and around +/- 1 deg C for TC measurements. Maximum performance is gained by selecting the trim points to bracket the operating point.

Direct connection of temperature sensors to input subsystems of distributed control systems (DCSs) or programmable logic controllers (PLCs) is an alternative to using a temperature transmitter for each measurement. It is suitable for less demanding data acquisition or control applications where wider variations in the measurement can be tolerated. The benefits of using transmitters include higher precision, sensor-transmitter systems calibrated to the range of interest, better RFI immunity and noise rejection, transmitter diagnostics, lower wiring costs, less expensive input/output (I/O) cards, faster loop checks, and shorter start-ups.

The types of intelligent temperature transmitters available today are almost endless. Some common features of the leading brands are namely universal inputs from any TC, RTD, mV, resistance or potentiometer, loop-powered with 4-20 mA output, digital outputs, and configuration with pushbuttons, PC software, or a hand-held configurator. Choices are made depending on the protocol is required (HART, Foundation Fieldbus, Profibus, vendor proprietary, Ethernet, or just 4-20 mA).

Application of temperature measurements

High temperature measurement – There are two viable methods for measuring temperatures upto 1,100 deg C. These are special high-temperature TCs and IR pyrometers. At high temperatures TCs are installed into protective wells or protection tubes. When installed horizontally, wells tend to droop causing binding on the TC element when it is to be removed for replacement. The latest design of a TC incorporates a 25 mm sheath with a flexible cable which can be easily inserted into even badly drooping wells. Upper limit for this sensor is around 1,100 deg C. Ceramic wells do not suffer from droop but have other limitations of low surface strength, brittleness, and low erosion resistance. IR pyrometers offer a very viable non-contact method to measure temperatures all the way upto 3600 deg C and are the best choice for most applications.

Speed of response – The fundamental problem of measuring the temperature of a fluid is one of assuring strong thermal coupling. For a fluid temperature measurement to have meaning, the sensor is to come to equilibrium with the temperature of the fluid. The difference between the equilibrium temperature of the sensor and the fluid temperature is a direct error. The most common process temperature measurements are made with TC and RTD sensors.

There are very few industrial measurements where an exposed sensor is to be used. This is since the process of taking the sensor out of service releases pressure or product from the pipe or vessel where the measurement is being made. Therefore, most applications use thermal wells to isolate the sensor from the process. Accordingly, the mass of the well and the piping into which it is inserted are the dominant causes of thermal lag and conduction errors. These errors can be almost insignificant for processes with stable temperatures and rapid flow. However, for dynamic temperature fluctuations or where there is little flow across the sensor, the errors can be large.

There are different designs of high speed of response sensors but most are limited by being produced with a protective sheath which runs the length of the well providing a long and massive path for thermal losses. A new design uses low mass RTDs or TCs in a 25 mm sheath to minimize this effect. To further increase the response, the sensor is spring loaded against the tip of the well and inserted with thermally conductive grease. Fast response is especially valuable for gas applications. Some of the better transmitters update their output many a times per second and therefore are rarely the limiting factor in the measurement.

Surface measurement – Measuring the surface temperatures of moving objects (such as rolls, rotary kilns, or rotating drums) needs special consideration. IR pyrometers offer a cost effective method. They offer many advantages, including high accuracy and fast response, and are especially suited to moving surfaces. However, for stationary surfaces, often a less sophisticated and less costly method is called for. The phase shift products offer a visual measurement but have no other output. There are a variety of mounting methods where either a TC or an RTD can be affixed to the outside of a pipe or vessel and be connected to a transmitter or data logger to get continuous information. The design of the mounting hardware and its proper installation are critical in obtaining a representative surface temperature. The thermodynamics of the application are complicated by the thermal losses to the surrounding atmosphere, thermal lag of the wall of the vessel or pipe, and the rate of change of the medium. A properly designed system uses a low mass sensor with high speed of response inserted into a fixture which places the sensor tip as close as possible to the surface being measured.

Measuring the temperature of solids – Determination of the allowable size and configuration of the sensor needs some knowledge of the heating or cooling conditions together with an estimate of the magnitude of the temperature gradients which are expected to exist in the region in which the measurement is to be made. A simple rule-of thumb indicator to determine if significant gradients are likely to be present is the magnitude of the Biot modulus (hL/K), where h is the surface heat transfer coefficient, L is the smallest dimension of the solid, and K is the thermal conductivity of the solid. If this modulus is over 0.2, significant temperature gradients are likely to exist in the solid, and care is to be taken in choosing the size, location, and orientation of the sensor within the solid. If the Biot modulus is less than 0.2, no significant gradient is expected and a measurement anywhere on or within the solid is to give identical results regardless of size or configuration of the sensor. If significant gradients are likely to exist, the maximum rate of heat transfer to the surface of the solid is to be known or estimated, and the maximum gradient at the point measurement is to be determined. The relationship ‘delta T/ delta X = q/K’ allows the maximum gradient at the surface of a solid is to be calculated. In this relationship delta T/ delta X is the temperature gradient at the surface, q is the heat transfer rate per unit area at the surface, and K is the thermal conductivity of solid.

Under certain conditions of heating or cooling, if measurements at points other than the surface are important, it can be necessary to evaluate anticipated heat transfer conditions and resulting temperature gradients. On the basis of this gradient, it is possible to establish limits on the size of the sensing device.

It is assumed that the sensors are in satisfactory thermal coupling with the process material, which is not always the case. If the thermal coupling is poor, the sensor does not reflect the true temperature history which is experienced by the solid. This condition can produce dynamic errors. The best thermal coupling is achieved by direct bonding of the sensor, such as welding a TC to the solid surface or into a cavity within the solid. The bond line between the sensor and the solid is to be kept as thin as possible and is not to fracture or fail during thermal cycling. Different epoxy and ceramic cements, with fillers to improve their conductivity, have been successfully used for such bonding.

Averaging measurements – There are different considerations depending on the application. The solution is different for liquids, gasses, or solids and also depends on the accuracy and speed of response needed required and what is to be done with the measurement results. For a temperature profile in a storage tank that is to be used for volume correction for inventory, an range of high accuracy RTDs can be the appropriate choice.

For a high temperature reactor or furnace, the sensing rage is to use TCs to generate the signal which can then be averaged in software by a receiving device. A highly accurate average temperature is obtained by using an RTD sensor which is constructed with a continuous resistive element so as to eliminate hot spots. This special RTD is to be a 4-wire construction and be used with a 4-wire measuring circuit like what is generally found in a high quality transmitter to obtain an accurate temperature measurement. For gas flow in pipes an RTD and transmitter assembly is an appropriate choice. If the application is for temperature compensation for a mass flow calculation, a 4-wire RTD trimmed to match a quality transmitter provides the highest accuracy. Such systems are available with accuracies of +/- 0.01 deg C. For stationary or moving surfaces of almost any size, the IR pyrometer is the best choice over any range.

Narrow span measurements – When the operating point of an application is stable, higher accuracy can be achieved by reducing the span of the measuring instrument. For example, when measuring temperature of a process which operates at 125 deg C with a variation of +/-10 degrees, a range of 0 to 150 degrees is well suited. An instrument ranged from 0 degrees to 600 degrees has 4 times less resolution. TCs are generally limited to a minimum span of around 35 deg C to 82 deg C, depending on type, while spans using 1,000 ohms Pt RTD with a high quality transmitter can be as low as 1 deg C. Matching a sensor to its transmitter offers the highest possible accuracy.