An RTD tells temperature by its resistance, which is a lovely, precise idea until you remember that the wires running out to the sensor also have resistance, and that resistance sits in series with the very thing you are trying to measure. On a long field run, the copper leads can add enough resistance to read as several degrees of error if nothing is done about it. Lead-wire compensation is the set of wiring schemes that cancel or eliminate that lead resistance so the instrument sees the element and not the copper going to it. This page drills into how the three-wire and four-wire methods actually cancel the leads, what they assume, and where they fail.
RTD Lead-Wire Compensation in one line: RTD lead-wire resistance compensation cancels the resistance of the copper leads that otherwise adds directly to the element resistance the instrument measures. A three-wire connection assumes the two current-carrying leads are equal in resistance, measures one lead's worth, and subtracts it, which works only if the leads are matched. A four-wire Kelvin connection drives a known current through two leads and senses the voltage across the element through two separate leads carrying almost no current, so the lead resistance produces no measurable voltage drop and is eliminated rather than merely estimated.
An RTD measurement is fundamentally a resistance measurement: the instrument passes a small known current through the element and reads the voltage across it, and from that resistance it looks up the temperature. The problem is that with a simple two-wire connection, the voltage the instrument reads is the drop across the element plus the drop across both leads, because the current has to travel out along one lead, through the element, and back along the other. The instrument cannot tell which part of that voltage came from the element and which came from the wire, so it attributes all of it to the element and reports a higher resistance, and therefore a higher temperature, than the element actually has.
The size of that error scales with the lead length and gauge. Copper has a definite resistance per unit length, and on a long run of thin wire to a remote sensor, the round-trip lead resistance can be a meaningful fraction of an ohm or more. Since a platinum RTD changes resistance by a fixed, fairly small amount per degree, even a fraction of an ohm of extra lead resistance translates into a real temperature offset. This is why a two-wire RTD is only acceptable for short runs where the lead resistance is negligible against the required accuracy; stretch the run out and the two-wire error grows with it.
There is a second, nastier feature of the two-wire error: it is not constant. The lead resistance itself changes with temperature, because copper's resistance rises as it warms, so a two-wire RTD run through an area whose temperature swings will show an offset that drifts with the ambient conditions along the cable, not just with the process. That means you cannot even cleanly calibrate the error out, because the amount you would subtract keeps moving. Compensation schemes exist precisely because the lead resistance is both significant and variable, and the only robust fix is a wiring arrangement that removes its influence rather than a fixed correction factor.
The three-wire connection is the common industrial compromise. It adds a third wire to the sensor so the instrument can measure the resistance of one lead separately and subtract it. In the classic arrangement, the measurement circuit is set up so that one path sees the element plus two leads while another path sees essentially one lead on its own, and by subtracting the single-lead measurement, doubled, from the element-plus-two-leads measurement, the instrument cancels the lead contribution and is left with the element resistance. The elegance is that it gets most of the benefit of full compensation with only one extra conductor over the two-wire case.
The catch is right there in the method: it works only if the leads are matched, meaning the leads carrying the current are equal in resistance to each other. The three-wire scheme measures one lead and assumes the other current-carrying lead is identical, then subtracts on that basis. If the two leads are the same gauge, the same length, and at the same temperature, that assumption holds well and the compensation is accurate. That is why RTD leads are specified to be matched conductors run together in the same cable, so they share length and thermal environment and stay equal.
When the assumption breaks, the compensation breaks with it. If one lead develops extra resistance, from a corroded terminal, a partly failed splice, a different gauge on a repair, or one leg of the cable running through a hotter region than the other, the three-wire scheme subtracts the wrong amount and leaves a residual error. Because it only measures one lead and trusts the other to match, an imbalance between the two current-carrying leads passes straight into the reading. This is the characteristic three-wire failure mode: not a gross fault that alarms, but a subtle offset that appears when the leads stop being equal, which is why unequal lead resistance is a real concern on long or aging RTD runs.
The four-wire, or Kelvin, connection eliminates lead resistance rather than estimating it. It uses two leads to force a known measuring current through the element and two entirely separate leads to sense the voltage across the element. The trick is that the two sensing leads connect to a high-impedance voltage measurement, so almost no current flows in them. Since the voltage drop across a wire is its resistance times the current through it, and that current is essentially zero in the sense leads, the sense leads develop no meaningful voltage drop no matter what their resistance is. The instrument therefore reads the true voltage across the element itself, and the lead resistance of all four wires simply drops out of the answer. This makes four-wire measurement independent of lead length, lead matching, and lead temperature.
Because it does not rely on the leads being equal, the four-wire method also sidesteps the three-wire failure mode entirely. Unequal leads, one corroded terminal, a lead running through a hot spot, none of it matters, because the measurement never assumes anything about the leads, it just refuses to let their resistance into the sensed voltage. The cost is one more conductor and slightly more instrument channel complexity, which is why four-wire is the choice for the most accurate laboratory and reference measurements and for long or demanding industrial runs where lead matching cannot be trusted, while three-wire remains the practical default for ordinary process points where matched leads in a single cable are good enough.
In field operations, the choice of scheme has consequences that outlive commissioning, and monitoring makes those consequences visible. A three-wire RTD whose leads slowly diverge, as a terminal corrodes or a splice ages, produces a temperature that creeps off true in a way that is easy to miss on a single reading but stands out as a slow drift or a growing disagreement with a neighboring sensor when trended over months. When these points are monitored in a platform such as Merobix, an RTD that starts wandering relative to its history or its neighbors flags a likely lead-compensation problem, pointing maintenance at the wiring and terminals rather than assuming the process actually changed. On critical measurements, that failure mode is one of the arguments for spending the extra conductor on four-wire from the start.
In a two-wire RTD the measuring current travels out one lead, through the element, and back the other lead, so the voltage the instrument reads includes the drop across both leads plus the element. The instrument cannot separate the lead drop from the element drop, so it reports a higher resistance and therefore a higher temperature than the element actually has. The error grows with lead length and also drifts, because copper lead resistance itself changes with temperature.
A three-wire connection measures the resistance of one lead separately and subtracts it, assuming the two current-carrying leads are equal. The circuit compares an element-plus-two-leads measurement against a single-lead measurement and cancels the lead contribution mathematically. This works well when the leads are the same gauge, length, and temperature, which is why RTD leads are run as matched conductors in one cable, but it leaves a residual error if the two leads become unequal.
Four-wire sensing drives the measuring current through two leads and reads the element voltage through two separate high-impedance sense leads that carry almost no current. Because voltage drop equals resistance times current, and the sense-lead current is essentially zero, the sense leads develop no voltage drop regardless of their resistance, so the instrument reads the true element voltage. This eliminates lead resistance entirely rather than estimating it, so it does not depend on the leads being matched or at the same temperature.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.