A thermoelectric generator makes DC power by exploiting a temperature difference across a stack of semiconductor junctions called a thermopile, and the size of that difference is the single number that decides how many watts come out. One face of the thermopile is pressed against a hot shoe heated by a small gas burner, and the opposite face is bonded to a finned cold side that dumps heat to the surrounding air. The generator does not care how hot the burner is in absolute terms; it cares about the gap between the hot shoe and the cold side. Understanding what raises and lowers each of those two temperatures is the key to understanding why a TEG that delivers rated power on a cold winter day can quietly sag below its budget on a hot, still summer afternoon.
TEG Hot Shoe and Cold Side in one line: A TEG's DC output is driven by delta-T, the temperature difference across its thermopile, not by the burner temperature alone. The hot shoe is held near a fixed temperature by the burner, while the finned cold side floats above ambient air, so anything that warms the cold side, such as high ambient temperature, blocked fins, or no breeze, shrinks delta-T and reduces the watts delivered. Monitoring both the hot-shoe and cold-side conditions is how you tell whether a TEG will meet its power budget in the season that stresses it.
The physics under a thermoelectric generator is the Seebeck effect: when the two junctions of a thermoelectric couple sit at different temperatures, a voltage appears across the couple, and stacking many couples in series into a thermopile multiplies that small voltage into something useful. The magnitude of the voltage, and therefore the power the device can push into a battery, scales with the temperature difference between the hot face and the cold face. This is why the governing variable is written as delta-T, the difference across the thermopile, rather than either temperature by itself. A hot shoe at a high temperature delivers no power at all if the cold side happens to be at the same temperature, because with no gradient there is no Seebeck voltage.
In a field TEG the hot shoe is held close to a target temperature by the burner and its flame management, so it behaves as a roughly fixed anchor. The variable end is the cold side. The cold side is a heat sink, a set of aluminum fins whose job is to reject the heat that flows through the thermopile into the surrounding air. Those fins cannot be colder than the air around them; at best they sit some tens of degrees above ambient, because heat only flows from the fins to the air when the fins are warmer than the air. So the cold-side temperature is essentially ambient temperature plus whatever rise the fins need to shed their heat load. As ambient climbs, the cold side climbs with it, and delta-T shrinks from the cold end even though the burner has not changed at all.
This is the counterintuitive part that trips people up. Operators reasonably assume a TEG delivers constant power because the burner runs constantly, but the output actually breathes with the weather. On a cold, breezy day the cold-side fins run cool, delta-T is wide, and the generator makes rated power or more. On a hot, windless day the fins run warm, delta-T collapses toward the low end of its range, and the same generator burning the same fuel delivers noticeably fewer watts. Nothing has failed. The device is simply obeying the temperature difference it is given, and the worst case for output coincides with the hottest, stillest conditions rather than the coldest.
Because the cold side is the swing variable, most of the practical output problems on a TEG trace back to things that let it run hotter than it should. The first and unavoidable one is ambient temperature. A TEG is usually rated at a moderate ambient, and its output curve slopes downward as the surrounding air warms, so the summer worst-case output can be materially below the nameplate figure. This is why a TEG's power budget has to be checked against the hot season, not the mild spring day it was commissioned on, and why the honest number to design around is the rated output at the highest ambient the site sees.
The next set of culprits are things that reduce the fins' ability to reject heat, which forces them to run at a higher temperature above ambient to shed the same watts. Fins caked with dust, mud, insect nests, or paint overspray lose surface area and airflow. Fins mounted where nothing moves the air, tucked against a wall or inside a tight enclosure, lose the convection that normally cools them. Snow packed against the cold side, or a nearby heated vessel radiating onto the fins, warms them from outside. Every one of these raises the cold-side temperature, narrows delta-T, and drops the output, and none of them announces itself as a fault. The burner is fine, the fuel flow is fine, and the generator simply makes fewer watts than its rating implies.
There is a subtler failure worth naming: degraded thermal contact inside the module. The thermopile has to be tightly clamped to both the hot shoe and the cold-side plate, with thermal interface material filling the microscopic gaps, so that heat flows efficiently through the semiconductor rather than piling up. If that clamping loosens or the interface material dries out and voids, heat cannot cross the joints cleanly. The hot face runs hotter and the cold face runs cooler than the temperatures the thermopile actually feels, so the internal delta-T the device experiences is smaller than the external temperatures suggest, and output falls even though the burner and the fins look correct. This is a real aging mechanism in TEGs and one reason long-serving units slowly lose capacity.
Because a TEG's power breathes with ambient and slowly declines with age, its charge current and battery voltage are exactly the kind of slow-moving signals that a remote monitoring platform is built to trend. A TEG powers a site precisely where nobody drives out to check it, so the only way to know whether delta-T is holding up is to watch what the generator delivers into the battery over time. Trending charge current against ambient temperature turns the invisible physics into a visible curve: the current should sag predictably in summer heat and recover in winter, and that predictable breathing is the signature of a healthy unit.
The value of watching it in a cloud SCADA system such as Merobix is that it separates the normal seasonal sag from a genuine problem. If charge current drops in a heat wave and comes back when the weather cools, the delta-T is behaving and the design simply needs enough battery to ride through the hot spells. If instead the charge current is trending steadily downward across seasons, running lower this summer than last summer at the same ambient, that is the fingerprint of fouled fins, a loosening thermal joint, or a burner slowly losing its target temperature. The trend distinguishes the two, and it does so months before the battery finally fails to recover overnight.
That early warning is the whole point at a remote site. A TEG whose output has quietly drifted below the site's load will not fail on the day it drifts; it will fail on the first stretch of hot, cloudy, still weather when the shrunken delta-T and the depleted battery finally cannot carry the load through the night. Watching the charge current and battery voltage on a monitoring platform lets an operator schedule a fin cleaning or a module reseat as routine maintenance in mild weather, rather than dispatching an emergency truck roll to a dead RTU in the middle of a heat wave. The temperature difference is the physics; the trended output is how you keep an eye on it from a hundred miles away.
A TEG's output tracks the temperature difference across its thermopile, and the cold side is a finned heat sink that can only run a certain amount above the surrounding air. When ambient temperature rises, the cold-side fins run hotter, so the temperature difference across the thermopile shrinks even though the burner has not changed. Less temperature difference means less Seebeck voltage and fewer watts, which is why a TEG's worst-case output is on a hot, windless day.
The hot shoe is the metal component that transfers heat from the gas burner into the hot face of the thermopile, holding it near a target temperature. It acts as a roughly fixed thermal anchor, so in practice the variable end of a TEG is the finned cold side rather than the hot shoe. The generator's output depends on the gap between the hot-shoe temperature and the cold-side temperature, not on the hot shoe alone.
The cold-side fins reject the heat that flows through the thermopile into the air, and they can only do that when they are warmer than the air. Dust, mud, insect nests, or snow on the fins cut their effective surface area and airflow, so they have to run at a higher temperature to shed the same heat. That warmer cold side narrows the temperature difference across the thermopile and lowers the watts delivered, without any fault appearing in the burner or fuel system.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.