January cold snaps surface every weakness in a heat trace install. The cable that worked fine at -10 trips its GFCI at -22. The eaves trough cable that was supposed to clear the ice dam is melting nothing while water backs up under the shingles. The controller is showing power but the heater stays cold. Each of these is a different failure mode and each one has a different fix. Here is the pattern we see on heat trace calls every January and how we sort it out without guessing.
How heat trace cable is supposed to work
Modern heat trace cable for pipes, eaves, and roof valleys is almost all self-regulating. The cable contains a conductive polymer core between two bus wires. The polymer's resistance varies with temperature - cold polymer is low resistance and lets more current flow, generating more heat; warm polymer is high resistance and throttles the current down. The result is a cable that puts out more heat where it is cold and less where it is warm, without a thermostat doing the math.
The cable runs on a dedicated 120V or 240V GFCI-protected circuit, often through a controller (Raychem EMDR, Heat-Line, Easyheat ADKS or similar) that adds a thermostat sensor and sometimes a moisture sensor to switch the cable on only when needed. The cable is rated for a service life of roughly 10 years on pipe trace, often less on roof and eaves applications because of UV exposure.
Failure mode one: GFCI trips at low temperature
The most common January call. The cable ran fine in November and December; the first night the temperature drops below -20, the GFCI breaker on the heat trace circuit trips. The owner resets it; it trips again within minutes or hours.
What is happening: as a self-regulating cable ages, the polymer matrix degrades and develops microscopic cracks and pathways for moisture. At warmer temperatures the leakage is below the GFCI threshold (5 mA). At very low temperatures the polymer contracts, the cracks open further, the cable resistance changes, and the leakage current creeps above 5 mA - the GFCI trips. The cable is at end of life.
The diagnostic is straightforward: a megger test (insulation resistance) on the cable end to bus and bus to bus. A new healthy cable reads in the hundreds of megohms. An end-of-life cable reads in the single digits or worse. Once we have that reading, the cable is replaced. No amount of resetting will give you another season.
Failure mode two: cable looks fine, controller is dead
The cable tests good on the megger, the GFCI is not tripping, and the pipe is still freezing. Walk to the controller. Most controllers have an LED or display that should show power and a state ("heat on" or "standby"). Common controller failures:
- Temperature sensor failed open. The controller cannot read the pipe or ambient temperature and stays in standby. Sensor swap, often a 10-minute fix.
- Internal relay or contactor failed. The controller is signalling the cable on but no current is actually getting through. Replace the controller.
- Moisture in the controller enclosure. Field-mounted controllers can lose their seal over years. Water shorts the logic board.
- Battery-backed program lost. Some older controllers had memory backed by a coin cell. Battery dies, settings reset to defaults, which often means "off." A wattmeter reading on the cable, or a re-programming pass, fixes it.
Failure mode three: cable is on but the pipe still freezes
The cable is energized and putting out heat - you can feel it warm to the touch through the pipe insulation - and the pipe still froze. This is almost always an installation issue, not a cable failure.
- Cable not wrapped on the cold side of the pipe. The cable has to run on the surface of the pipe in continuous contact, not floating inside the insulation. A cable taped to the top of the pipe with a 3-cm air gap is doing very little.
- Insulation missing or compressed. The pipe insulation does the bulk of the work; the cable is there to make up the last few degrees. A pipe with insulation that has slumped, fallen off, or been compressed by a contractor's foot in the attic is losing heat faster than the cable can replace.
- Cable spec undersized. Heat trace cable comes in wattages (W/ft at 50F is the spec). A 3 W/ft cable on a 1-inch pipe in an unheated crawl space at -25 is not enough. The replacement is a higher-wattage cable.
- Cable run length too short. The owner ran cable on the easy-to-reach section, but the pipe also passes through an exterior wall stud bay where the cable does not reach. The cold side freezes.
Eaves and roof heat trace - a different conversation
Roof and eaves heat trace serves a different purpose than pipe trace. The goal is to keep a meltwater path open through the ice that builds at the eaves so water from any roof melting can drain instead of backing up under the shingles and into the soffit. It is not supposed to keep the roof dry. The right install is a specific zigzag pattern in the eaves trough and along the lowest 30-60 cm of the roof edge, sized by the roof area and the local snowfall.
Roof and eaves trace fails for the same self-regulating-cable reasons, plus UV degradation. Cable jacket exposed to sun for 7-8 years loses elasticity, cracks, and lets water into the polymer core. UV is also why we run lifetime expectations a bit lower on roof installs - 7-10 years is realistic.
The other roof-trace gotcha: ice dam patterns change as a house ages. Attic insulation settles, the dam pattern shifts, and the cable that used to be in the right place is now wrapped around an inch of clear shingle while the dam forms 30 cm uphill of it. A walk after the first big snow tells you whether the cable is in the right place this year.
What we install when the cable needs to be replaced
For pipe trace at residential and cottage scale: Heat-Line Carapace or Easyheat ADKS self-regulating cable, properly wrapped on the pipe in contact, fully covered by sized pipe insulation, on its own GFCI-protected circuit. Cable wattage matched to the pipe size and the worst-case ambient at the location.
For roof and eaves: Heat-Line Roof Trace or equivalent self-regulating cable, anchored with cable clips (never screwed through the shingle), sized to the dam pattern and the roof area, on its own controller with a temperature and moisture sensor so it only runs when both are saying "yes."
For cottage and Muskoka use specifically, a remote-monitoring controller is often worth the upgrade - we install models that text the owner when the cable trips or the temperature falls below a threshold, so a winter failure does not become a thawed-pipe disaster.
The temptation to skip GFCI
Every January at least one customer asks if the cable would stop tripping if we wired it on a regular breaker instead of a GFCI breaker. The honest answer: yes, and you would have a fire risk we will not install. Heat trace cable is required by the OESC to be on GFCI protection because the failure mode is leakage to ground through a damp environment. The GFCI is doing exactly what it was designed to do when it trips. The fix is a new cable, not a worse circuit.
When to call us
If your heat trace circuit is tripping, your eaves are damming, or you are not sure whether the install you have ever worked the way it should, we troubleshoot and replace residential and cottage heat trace across both regions. For Muskoka cottages, this often pairs with a winter shutdown review. Request a visit and bring the cable spec sheet if you can find it.