Ice dams
An ice dam is not really a roofing failure. It is heat escaping into an attic, melting snow high on the roof, and that meltwater refreezing at the cold eave.
- 1 Ridge caps. Thin, fully exposed, and the first course to lift.
- 2 The field. Granule loss is the honest measure of how much life is left.
- 3 Valleys. One channel taking the runoff of two slopes, all year round.
- 4 Flashing. At the chimney, and where the garage roof meets the wall of the house.
- 5 The vent boot. Small, rubber, and on a much shorter clock than everything around it.
- 6 Eaves. Where water that has backed up at the edge gets under the courses.
How the water gets in
Melt runs down under the snow until it reaches the overhang, which is not warmed from below and stays at outdoor temperature. It refreezes there and builds a ridge of ice. Water backing up behind that ridge sits on the roof rather than running off, and shingles are designed to shed water flowing downhill, not to hold standing water. It works its way under the courses and into the deck.
What actually fixes it
Air sealing and insulation at the attic floor, so heat stays in the house. Ventilation, so the attic runs close to outdoor temperature and the roof deck stays uniformly cold. Ice-and-water shield membrane at the eaves during a replacement, which does not prevent the dam but does prevent the backed-up water from reaching the deck.
What does not fix it
Chipping at the ice with anything sharp damages the shingles underneath and is dangerous from a ladder in winter. Salt products stain and corrode. Both treat the symptom on a roof that will form the same dam at the next thaw, because nothing about the attic has changed.
The two winters this market actually has
Something worth settling before anyone diagnoses a dam here: the area this site covers contains two different winters. Over the last decade NOAA’s records put the Baltimore side at roughly eighty-five nights a year at or below freezing with about sixteen inches of snow, and the Arlington and Alexandria side closer to fifty nights and twelve inches. The southern half is not a milder version of the northern one. It is a different problem, because a run of shallow snowfalls that melt and refreeze builds more ice at an eave, inch for inch, than one deep cover that simply stays frozen.
It is also why dams here often form without a snowfall anyone remembers. A few inches on a Sunday, a bright Tuesday in the forties, a hard freeze that night, and the same again the following week: meltwater at the eaves on every cycle, refreezing on every cycle, working a little further under the courses each time. Anyone hunting for one event to blame will not find it, which is the whole point. What changed is inside the attic, not in the forecast.
The northern edge of the coverage is where the harder winter is the ordinary one. Frederick and Hagerstown sit high enough and far enough inland to keep snow on a roof for days rather than hours. Up the bay, Havre de Grace and Aberdeen rarely hold it that long, but the water keeps the air just wet enough that the freeze-and-thaw runs again and again, which does the same work more slowly.