To read an RTD, an instrument has to push a small current through its platinum element - and that current, flowing through a resistance, produces heat. The element warms itself slightly, and it cannot distinguish that self-generated warmth from the process temperature it is supposed to measure, so the reading comes out a touch high. This is RTD self-heating error, and while it is small enough to ignore in many cases, it becomes a real, standing bias in still air, low-flow gas, and other conditions where the element cannot shed the heat it makes.
RTD Self-Heating in one line: RTD self-heating error is the measurement bias that occurs because the excitation current used to read the element also warms it through resistive heating. The element reads its own slightly elevated temperature rather than the true process value, and the size of the error depends on how much power is dissipated and how effectively the surrounding medium carries that heat away.
An RTD is measured by passing a known current through it and reading the voltage that develops, which reveals the resistance and therefore the temperature. But any current flowing through a resistance dissipates power as heat, following the familiar relationship that power equals current squared times resistance. That heat is generated right inside the platinum element, the very thing whose temperature the instrument is trying to read. The element inevitably runs a little warmer than its surroundings, and it faithfully reports that warmer temperature as if it were the process.
Because the heating goes with the square of the current, self-heating is very sensitive to how hard the element is excited. Doubling the excitation current roughly quadruples the self-heating power. This is the core reason well-designed RTD transmitters use a deliberately small excitation current, often a fraction of a milliamp, rather than a larger current that would give a stronger, cleaner signal. The design is a compromise: enough current to measure the resistance reliably, but little enough that the self-heating stays negligible under expected conditions.
The element resistance matters too, which is one subtle difference between PT100 and PT1000. For the same excitation current, the higher-resistance PT1000 dissipates ten times the power of a PT100, so it self-heats more per unit of current - a factor to weigh when a design pushes current for signal reasons. In every case, self-heating is not a defect or a fault; it is an unavoidable physical consequence of measuring temperature by driving current through a resistor, and the goal is to manage it, not eliminate it.
How much the element actually warms for a given amount of self-heating power depends on how easily it can shed that heat, and that is captured by the dissipation constant. The dissipation constant expresses how many milliwatts of power raise the element temperature by one degree in a specified medium and flow condition. A high dissipation constant means the medium carries heat away efficiently, so a given power produces only a tiny temperature rise; a low dissipation constant means the same power produces a much larger, more troublesome rise.
This is why the same RTD carrying the same excitation current can show negligible self-heating in one application and a meaningful error in another. Immersed in flowing water or moving liquid, the element sheds heat readily and self-heating is often insignificant. Sitting in still air, in a slow-moving or low-pressure gas, or in a poorly coupled thermowell with an air gap, the same element cannot get rid of its self-generated heat, its effective dissipation constant collapses, and the error grows. Low-flow gas measurement is a classic setting where self-heating quietly biases readings high.
Installation details feed directly into this. Good thermal contact between the element and the process - proper immersion depth, a snug thermowell fit, sometimes a thermally conductive filler - raises the effective dissipation constant and suppresses self-heating. Air gaps, shallow immersion, and stagnant media do the opposite. So self-heating is partly a sensor-and-transmitter matter and partly an installation matter, and a self-heating problem can sometimes be fixed by improving how the element is coupled to the process rather than by changing the electronics.
In practice, self-heating is managed at the transmitter by keeping the excitation current low, and most quality RTD transmitters are designed so that under normal immersed conditions the self-heating error is well within the accuracy budget. The place to pay attention is the application that undermines heat dissipation: low-flow or static gas, still air, shallow or poorly coupled installations, and any measurement demanding tight accuracy. In those settings it is worth confirming the transmitter's excitation current and the sensor's dissipation constant support the accuracy you are claiming, rather than assuming self-heating is always negligible.
The error tends to be a consistent offset under steady conditions, which is precisely what makes it easy to overlook. It does not spike or alarm; it simply sits a fraction of a degree to a couple of degrees high, and it can even change with process conditions - a temperature that reads slightly higher when flow drops, because the element sheds less heat at low flow, can be self-heating masquerading as a process effect. Seeing a temperature that correlates oddly with flow rate in a historized SCADA trend is a hint worth remembering.
For a cloud monitoring platform such as Merobix, self-heating lives entirely upstream in the sensor and transmitter; the platform receives a compensated, scaled temperature and cannot see the excitation current or the dissipation directly. Its value is in the trend. When a temperature point that ought to be steady drifts in a way that tracks flow, ambient conditions, or a change in installation, the historized record makes that correlation visible and points an engineer toward causes like self-heating that a single spot reading would never reveal. Getting the sensor excitation and installation right at the source is what keeps that trend trustworthy.
Under good conditions - an element well immersed in flowing liquid, driven by a low excitation current - self-heating is usually small enough to sit within the transmitter's accuracy budget and can be ignored. It grows when the surrounding medium carries heat away poorly, such as still air or low-flow gas, where it can become a standing bias of a fraction of a degree up to a couple of degrees. The exact figure depends on the excitation current and the dissipation constant.
The dissipation constant tells you how much self-heating power raises the element temperature, expressed as milliwatts required per degree of temperature rise in a stated medium and flow condition. A high dissipation constant means the medium removes heat efficiently and self-heating is small, while a low value means the same power causes a larger error. It is specified for particular conditions because moving liquid, still air, and low-flow gas differ dramatically.
The main lever is keeping the excitation current small, because heating rises with the square of the current, so a modest reduction in current sharply reduces self-heating. Many RTD transmitters use only a fraction of a milliamp for this reason, and some pulse the excitation rather than driving it continuously. Good installation - proper immersion and thermal contact - also helps by letting the element shed its self-generated heat more effectively.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.