RTD vs Thermocouple

What is the difference between and a resistance temperature detector (RTD)  and a thermocouple?  Both RTDs and thermocouples are sensors used to measure heat in scales such as Fahrenheit and Kelvin. Such devices are used in a broad range of applications and settings, often presenting people with the dilemma of choosing to use either RTDs or thermocouples. Each kind of temperature sensor has its own advantages and disadvantages that make it suitable for certain conditions and circumstances.

Resistance Thermometer Detectors

The electrical resistance of metals rises as heat increases and the metals become hotter, while their electrical resistance falls as heat decreases and the metals become colder. RTDs are temperature sensors that use the changes in the electrical resistance of metals to measure the changes in the local temperature. For the readings to be interpretable, the metals used in RTDs must have electrical resistances known to people and recorded for convenient reference. As a result, copper, nickel, and platinum are all popular metals used in the construction of RTDs.

Thermocouples

Thermocouples are temperature sensors that use two different metals in the sensor to produce a voltage that can be read to determine the local temperature. Different combinations of metals can be used in building the thermocouples to provide different calibrations with different temperature ranges and sensor characteristics.

RTD vs Thermocouple

Because the terms encompass entire ranges of temperature sensors tailored for use under a range of conditions, it is impossible to conclude whether RTDs or thermocouples are the superior option as a whole. Instead, it is more useful to compare the performance of RTDs and thermocouples using specific qualities such as cost and temperature range so that users can choose based on the specific needs of their organizations.

In general, thermocouples are better than RTDs when it comes to cost, ruggedness, measurement speed, and the range of temperatures that can be measured using them. Most thermocouples cost 2.5 to 3 times less than RTDs and although RTD installation is cheaper than thermocouple installation, the savings in installation costs are not enough to tip the balance. Furthermore, thermocouples are designed to be more durable and react faster to changes in temperature because of that same design. However, the main selling point of thermocouples is their range. Most RTDs are limited to a maximum temperature of 1000 degrees Fahrenheit. In contrast, certain thermocouples can be used to measure up to 2700 degrees Fahrenheit.

RTDs are superior to thermocouples in that their readings are more accurate and more repeatable. Repeatable means that users reading the same temperatures produce the same results over multiple trials. RTDs producing more repeatable readings means that their readings are more stable, while their design ensures that RTDs continue producing stable readings longer than thermocouples. Furthermore, RTDs receive more robust signals and it is easier to calibrate RTD readings due to their design.

Conclusion

In brief, RTDs and thermocouples each have their own advantages and disadvantages. Furthermore, each make of RTDs and thermocouples possesses its own advantages and disadvantages. Buyers should base their purchasing decisions on the specific needs and capabilities of their organizations matched to the specific capabilities of the brands available to them. In general, thermocouples are cheaper, more durable and can measure a bigger range of temperatures, while RTDs produce better and more reliable measurements.

THERMOCOUPLES IN GAS TURBINES

A complex engine like a gas turbine needs to be thoroughly instrumented in order to be safely and correctly operated: the most important parameter to be monitored is definitely temperature.

The operation of a two shaft jet engine, schematically shown in Figure8, can be simply described with the following steps:

• air is aspirated from the atmosphere (station 0) to the low pressure compressor inlet (station 2) and compressed till the low pressure compressor exit (station 2.5); an increase in air temperature and pressure results;
• air is compressed from the high presssure compressor inlet (station 2.5) to the high pressure compressor exit (station 3); an increase in air temperature and pressure results;
• in the combustor compressed air is mixed with fuel and combustion takes place: the gas mixture exits the combustor (station 4) at higher temperature than at the combustor inlet (station 3) and with almost the same pressure;
• combustion gases are expanded in the high pressure turbine from station 4 to station 4.5 with a reduction in pressure and temperature: the high pressure turbine drives the high pressure compressor as they share the same shaft;
• combustion gases are expanded in the low pressure turbine from station 4.5 to station 5 with a reduction in pressure and temperature: the low pressure turbine drives the low pressure compressor as they share the same shaft; part of the power produced in the turbine is used to drive the compressor and part is converted in useful work, that is thrust in the jet engine;
• gases are then released to the atmosphere through a diffuser (from station 5 to station 8).

Figure8: Schematic drawing of a two shafts gas turbine

The following stations are commonly instrumented with thermocouples:

• station 2
• station 2.5
• station 3
• station 4
• station 4.5
• station 5

and usually each station has at least 8 thermocouples at different angular locations. All of these temperature measurements aregas temperature measurement.

The main reasons to monitor continuously the gas turbine temperature are:

• performance evaluation: the knowledge of inlet and exit temperatures allows performance engineers to calculate the efficiency of compressors and turbines;
• engine control: the maximum power from the jet engines is not always required during the flight of an airplane. Maximum power is required at take off, but lower power level are required for instance during cruise. The control of the jet engine through all the different operational conditions is a complex engineering problem where temperature monitoring at the different stations plays a major role;
• health monitoring of high temperature components: a temperature increase at the turbine inlet (T₄) would increase the efficiency of the engine and, as a result, reduce the fuel consumption at the same power or increase the available power with the same fuel consumption. However, increasing the temperature T₄ is not easy: components already face very demanding temperatures at the high pressure compressor exit and the high pressure turbine inlet and serious components damage and failure can occur for the turbine blades and last stages of the high pressure compressor blades if the temperature is above the capability of the used materials. Temperature limits are implemented in the control of the engine to avoid temperatures above the melting points of alloys and to have limited creep deformation for rotating blades. Furthermore the temperature history of the components can be used to estimate their residual life.

The acquisition of the temperature measurement from a thermocouple immersed in a flowing gas through the walls of the gas duct is not trivial. In fact the temperature sensed at the junction of the thermocouple, T_J, is different from the total temperature of the gas, T_T, which is the useful thermodynamic parameter for the engine. T_J is a result of the thermal equilibrium between the thermocouple and the environment around the thermocouple. In particular heat transfer occurs:

• through conduction along the wires and the sheath of the thermocouple,
• through radiation to/from the walls and the blades/vanes surfaces,
• through convection at the boundary layer around the thermocouple.

Conduction and radiation give rise to two measurement errors called conduction error and radiation error respectively.

Furthermore, obtaining the total temperature of a flowing gas is technically difficult. This would require to stop the gas adiabatically in order to convert the kinetic energy of the flowing gas in a temperature increase: so the static temperature of the flowing gas, T_S, would be converted in the total temperature T_T.

The gas flowing around a thermocouple immersed in the gas is slowed down, but not stopped and not adiabatically: for this reason a measurement error arises, called velocity error.

The recovery factor α , defined as

Equation3

takes into account conduction error, radiation error and velocity error.

A good thermocouple design is achieved if the recovery factor is very close to 1.

A shield around the thermocouple, as shown in Figure9, is a common measure to increase the recovery factor: in fact it reduces the radiation and velocity errors.

Figure9: Schematic diagram showing a shield around a thermocouple.

Furthermore, it should be mentioned that the thermocouple always lags behind the real gas temperature during transient: the thermocouple has its own time response, which can be improved through careful design.

THE OPERATING PRINCIPLE OF THERMOCOUPLES

A thermocouple is a device made by two different wires joined at one end, called junction end or measuring end. The two wires are called thermoelements or legs of the thermocouple: the two thermoelements are distnguished as positive and negative ones. The other end of the thermocouple is called tail end or reference end (Figure1). The junction end is immersed in the enviroment whose temperature T₂ has to be measured, which can be for instance the temperature of a furnace at about 500°C, while the tail end is held at a different temperature T₁, e.g. at ambient temperature.

Figure1:Schematic drawing of a thermocouple

Because of the temperature difference between junction end and tail end a voltage difference can be measured between the two thermoelements at the tail end: so the thermocouple is a temperature-voltage transducer.

The temperature vs voltage relationship is given by:

Equation1

where Emf is the Electro-Motive Force or Voltage produced by the thermocople at the tail end, T₁ and T₂ are the temperatures of reference and measuring end respectively, S₁₂ is called Seebeck coefficient of the thermocouple and S₁ and S₂ are the Seebeck coefficient of the two thermoelements; the Seebeck coefficient depends on the material the thermoelement is made of. Looking at Equation1 it can be noticed that:

1. a null voltage is measured if the two thermoelements are made of the same materials: different materials are needed to make a temperature sensing device,
2. a null voltage is measured if no temperature difference exists between the tail end and the junction end: a temperature difference is needed to operate the thermocouple,
3. the Seebeck coefficient is temperature dependent.

In order to clarify the first point let us consider the following example (Figure2): when a temperature difference is applied between the two ends of a single Ni wire a voltage drop is developed across the wire itself. The end of the wire at the highest temperature, T₂, is called hot end, while the end at the lowest temperature, T₁, is called cold end.

Figure2: Emf produced by a single wire

When a voltmeter, with Cu connection wires, is used to measure the voltage drop across the Ni wire, two junctions need to be made at the hot and cold ends between the Cu wire and the Ni wire; assuming that the voltmeter is at room temperature T₁, one of the Cu wires of the voltmeter will experience along it the same temperature drop from T2 to T1 the Ni wire is experiencing. In the attempt to measure the voltage drop on the Ni wire a Ni-Cu thermocouple has been made and so the measured voltage is in reality the voltage drop along the Ni wire plus the voltage drop along the Cu wire.

The Emf along a single thermoelement cannot be measured: the Emf measured at the tail end in Figure1 is the sum of the voltage drop along each of the thermoelements. As two thermoelements are needed, the temperature measurement with thermocuoples is a differential measurement.

Note: if the wire in Figure2 was a Cu wire a null voltage would have been measured at the voltmeter.

The temperature measurement with thermocouples is also a differential measurement because two different temperatures, T₁ and T₂, are involved. The desired temperature is the one at the junction end, T₂. In order to have a useful transducer for measurement, a monotonic Emf versus junction end temperature T₂ relationship is needed, so that for each temperature at the junction end a unique voltage is produced at the tail end.

However, from the integral in Equation1 it can be understood that the Emf depends on both T₁ and T₂: as T₁ and T₂ can change indipendently, a monotonic Emf vs T₂ relationship cannot be defined if the tail end temperature is not constant. For this reason the tail end is mantained in an ice bath made by crushed ice and water in a Dewar flask: this produces a reference temperature of 0°C. All the voltage versus temperature relationships for thermocouples are referenced to 0°C.

The resulting measuring system required for a thermocople is shown in Figure3.

Figure3: A measuring system for thermocouples

In order to measure the voltage at the tail end, two copper wires are connected between the thermoelements and the voltmeter: both the Cu wires experience the same temperature difference and as a result the voltage drops along each of them are equal to each other and cancel out in the measurement at the voltmeter.

The ice bath is usually replaced in industrial application with an integrated circuit called cold junction compensator: in this case the tail end is at ambient temperature and the temperature fluctuations at the tail end are tolerated; in fact the cold junction compensator produces a voltage equal to the thermocouple voltage between 0°C and ambient temperature, which can be added to the voltage of the thermocouple at the tail end to reproduce the voltage versus temperature relationship of the thermocouple.

A sketch of a thermocouple with cold junction compensation is reported in Figure4.

Figure4: An example of Cold Junction Compensation

It should be underlined that the cold junction compensation cannot reproduce exactly the voltage versus temperature relationship of the thermocouple, but can only approximate it: for this reason the cold junction compensation introduces an error in the temperature measurement.

Figure4 shows also the filtering and amplification of the thermocouple. Being the thermocouple voltage a DC signal, removal of AC noise through filtering is beneficial; furthermore the thermocouples produce voltage of few tens of mV and for this reason amplification is required. The small voltage range for some of the most common thermocouples (letter designated thermocouples) is shown in Figure5, where their voltage versus temperature relationship is reported.

Type R, S and B thermocouples use Pt-base thermoelements and they can operate at temperatures up to 1700°C; however they are more expensive and their voltage output is lower than type K and type N thermocouples, which use Ni-base thermoelements. However, Ni base thermocouples can operate at lower temperatures than the Pt-base ones. Table1 reports the approximate compositions for positive and negative thermoelements of the letter designated thermocouples.

Figure5: Voltage vs Temperature relationship for letter-designated thermocouples

Thermocouple type Positive Thermoelement Negative Thermoelement B Pt-30%Rh Pt-6%Rh R Pt-13%Rh Pt S Pt-10%Rh Pt K Ni-10%Cr Ni-5% other elements N Ni-14%Cr-1.5%Si Ni-4.5%Si-0.1%Mg E Ni-10%Cr 45%Ni-55%Cu J Fe 45%Ni-55%Cu

Table1: Approximate composition for thermoelements of letter-designated thermocouples

All the voltage-temperature relationships of the letter designated thermocouples are monotonic, but not linear. For instance the type N thermocouple voltage output is defined by the following 10 degree polynomials, where t is the temperature in degree Celsius:

Equation2

The coefficients Ci are reported in Table2.

In order to have a linear voltage-temperature relationship the Seebeck coefficient should be constant with temperature (see Equation1); however the Seebeck coefficient is temperature dependent, as shown for instance for type K thermocouple in Figure6. Additional details on the voltage-temperature relatinships for letter designated thermocouple can be found at:

http://srdata.nist.gov/its90/main

Coefficient Temperature range: (-270°C,0°C) Temperature range: (0°C,1300°C) c0 0.000000000000 x100 0.000000000000 x100  c1 0.261591059620 x10-1 0.259293946010 x10-1 c2 0.109574842280 x10-4 0.157101418800 x10-4 c3 -0.938411115540 x10-7 0.438256272370 x10-7 c4 -0.464120397590 x10-10 -0.252611697940 x10-9 c5 -0.263033577160 x10-11 0.643118193390 x10-12 c6 -0.226534380030 x10-13 -0.100634715190 x10-14 c7 -0.760893007910 x10-16 0.997453389920 x10-18 c8 -0.934196678350 x10-19 -0.608632456070 x10-21 c9 - 0.208492293390 x10-24 c10 - -0.306821961510 x10-28

Table2: Type N thermocouple coefficents

Figure6: Type K Seebeck coefficient vs Temperature