RTD vs Thermocouple: Temperature Sensor Selection for BMS and Industrial Applications

Temperature sensors are specified on almost every instrumentation project — and yet the choice between an RTD and a thermocouple is often made by habit rather than by engineering judgment. Both measure temperature. Both are available in a wide range of probe styles. But they differ fundamentally in how they work, how accurate they are, what signal conditioning they need, and what temperature ranges they cover.

For building automation and HVAC, the choice is usually straightforward. For industrial and process applications, it requires more thought.

How RTDs Work

An RTD (Resistance Temperature Detector) uses the principle that the electrical resistance of a metal increases predictably with temperature. Platinum is used because its resistance-temperature relationship is highly linear, stable over time, and well-characterised by international standards.

The most common RTD is the Pt100 — a platinum element with a resistance of exactly 100 Ω at 0°C. The Pt1000 has 1000 Ω at 0°C and is increasingly preferred in BMS applications because its higher resistance makes lead resistance less significant and allows longer cable runs without correction.

RTDs require an excitation current — typically 1 mA — which the measurement instrument or transmitter provides. This current through the element produces a voltage that is measured to determine resistance and hence temperature. The passive RTD element itself is just a precision resistor; all signal conditioning is done by the measurement instrument.

How Thermocouples Work

A thermocouple consists of two dissimilar metal wires joined at one end (the measurement junction). When the measurement junction is at a different temperature from the reference junction (at the instrument), a small voltage — typically in the range of microvolts to millivolts — is generated. This is the Seebeck effect.

Different thermocouple types use different metal combinations, each with a different voltage-temperature curve:

Type	Metals	Range	Output at 100°C
K (most common)	Nickel-Chromium / Nickel-Aluminium	-200 to +1260°C	~4.1 mV
J	Iron / Copper-Nickel	-210 to +760°C	~5.3 mV
T	Copper / Copper-Nickel	-270 to +370°C	~4.3 mV
N	Nickel-Chromium-Silicon / Nickel-Silicon	-270 to +1300°C	~2.6 mV
S, R, B	Platinum alloys	Up to +1760°C	< 1 mV

The millivolt signal from a thermocouple requires amplification and cold junction compensation (correcting for the temperature at the reference junction) before it can be converted to a temperature reading. This is done inside the measurement instrument or signal conditioner.

Accuracy Comparison

This is where the choice often becomes clear for BMS applications:

Parameter	Pt100 RTD (Class B)	Pt100 RTD (1/3 DIN)	Type K Thermocouple (Class 1)
Accuracy at 0°C	±0.30°C	±0.10°C	±1.5°C
Accuracy at 100°C	±0.80°C	±0.27°C	±1.5°C
Accuracy at 500°C	±1.30°C	±0.43°C	±2.5°C
Accuracy at 1000°C	N/A (above range)	N/A	±2.5°C or ±0.4%
Long-term stability	Excellent (±0.02°C/year)	Excellent	Fair (drift due to metallurgical changes)
Repeatability	±0.05°C	±0.02°C	±0.5°C

For BMS and HVAC applications — where temperatures are typically 0–130°C and measurement accuracy of ±0.5°C is required for control and energy metering — the RTD wins unambiguously. The thermocouple's inaccuracy of ±1.5°C is a significant fraction of the ΔT being measured in heat meters (typically 5–10 K) and would introduce substantial energy metering errors.

When Thermocouples Are the Right Choice

Thermocouples have three advantages that make them the right choice in specific applications:

1. High temperature range: RTDs are limited to approximately 850°C (Pt100 Class B limits). Thermocouples measure up to 1760°C (platinum types) — essential for furnaces, kilns, combustion chambers and turbines.
2. Fast response: A thermocouple junction can be made extremely small (bare wire, no sheath), giving response times below 100 ms. Sheathed RTDs are slower due to thermal mass. For rapidly changing temperatures in combustion or extrusion applications, thermocouples respond faster.
3. Self-generating signal: Thermocouples generate their own voltage — no excitation current required. This makes them suitable for hazardous area applications where current injection is restricted, and for some battery-powered instruments.

2-Wire, 3-Wire and 4-Wire RTD Connections

One practical disadvantage of RTDs is lead resistance error. The excitation current flows through the measurement leads as well as the element, and the resistance of long cable adds to the apparent element resistance — causing a reading error of approximately +0.25°C per 1 Ω of total lead resistance.

Three connection methods address this:

• 2-wire: Simplest. Only appropriate for cable runs below 5 m where lead resistance is negligible, or for Pt1000 elements where lead resistance is a smaller proportion of total resistance.
• 3-wire: The standard for BMS applications. A third wire allows the measurement instrument to compensate for lead resistance — accurate to cable runs of 50–100 m with standard 1.5 mm² cable. Requires a 3-wire input card on the BMS or transmitter.
• 4-wire (Kelvin): Highest accuracy. Two wires carry excitation current; two separate wires measure voltage. Lead resistance is completely eliminated. Used in laboratories and high-accuracy process applications. BMS input cards rarely support 4-wire RTD.

Active Transmitters: The Practical Solution for Long Runs

For cable runs above 50 m, or where the BMS analogue input card does not have RTD conditioning, a head-mounted or rail-mounted temperature transmitter is the practical solution. The transmitter is located at the sensor, converts the RTD signal to a 4–20 mA output (spanning, say, 0–100°C across 4–20 mA), and transmits this current signal to the BMS input card over a two-wire loop that can run hundreds of metres without accuracy degradation.

Active transmitters also simplify wiring: a standard 4–20 mA analogue input card handles all sensor types, no specialist RTD input cards required.

BMS Compatibility

Most modern BMS controllers (Reliable Controls, Siemens, Honeywell, Schneider, Johnson Controls) have universal analogue input cards that accept:

• Pt100 3-wire (direct)
• Pt1000 2-wire (direct)
• 0–10 V (from active transmitter)
• 4–20 mA (from active transmitter)

Thermocouple input cards are available as accessories for most platforms but are rarely standard. If your BMS uses universal input cards, an RTD or active 4–20 mA transmitter will always integrate more cleanly than a thermocouple.

Recommended Models

For HVAC duct temperature measurement (supply and return air, AHU sections), the AG-RTD-100 provides Pt100 Class B or Pt1000 elements with optional active 4–20 mA output, in probe lengths from 100–500 mm. For pipe immersion measurement (chilled water, hot water, condenser water), the AG-RTD-200 includes a SS316 thermowell — the sensor is removable for replacement without draining the pipe. Both models are certified to CE and RoHS and support direct connection to BMS analogue input cards or loop-powered 4–20 mA transmission over long cable runs.