I used to think leaks would show up right below the problem spot, but nope. Sometimes it’s halfway across the roof, following seams or little dips you can’t even see.
That’s the part that gets people every time. Water’s got a mind of its own—give it a tiny path and it’ll zigzag all over before finally showing up somewhere totally random. I’ve seen leaks start at a vent pipe on one end of a flat roof and end up dripping into a closet on the opposite side. Drives folks nuts.
Thermal cameras are definitely handy, but I’ve had mixed luck depending on how wet things are. If the insulation’s already soaked, sometimes the temp difference isn’t as obvious. Still, way better than just poking holes and hoping for the best.
One thing I always check is the flashing around anything that sticks through the roof—vents, skylights, satellite dishes, you name it. Even a tiny crack or loose sealant can let water sneak in. And with green roofs, roots are like little escape artists. I’ve seen them worm their way into seams you’d swear were sealed tight.
Honestly, sometimes it feels like you need a PhD in “leakology” to figure these things out. But hey, at least the gadgets make it a bit more fun... or at least less miserable.
I get the love for gadgets, but honestly, I’ve had more luck with the old-fashioned water hose test than with thermal cameras. Maybe it’s just my luck, but every time I try to go high-tech, I end up chasing phantom leaks or getting weird readings because the insulation’s already soaked. Sometimes you just need two people—one on the roof with a hose and one inside with a flashlight and a lot of patience. Not glamorous, but it’s found me more leaks than any fancy tool so far. And yeah, flashing is always suspect, but don’t sleep on those seams either... especially after a few freeze-thaw cycles.
I hear you on the hose test—sometimes the simplest methods really do cut through the noise. I’ve had mixed results with thermal cameras too, especially if there’s already a lot of moisture in the insulation or decking. The readings can get confusing fast, and it’s easy to chase your tail if you’re not careful. That said, I still think there’s a place for tech, but only as a supplement, not a replacement for hands-on inspection.
On flat roofs, especially older ones or those with a green roof system like mine, leaks can be maddeningly hard to track down. Seams are definitely high on my list of suspects, right up there with flashing and any roof penetrations. After a couple freeze-thaw cycles, even a tiny gap can open up enough to let water sneak in and travel way farther than you’d expect before it shows up inside.
One thing I’ve found helpful is to check for ponding water after heavy rain. Even shallow puddles can stress seams and membranes over time. If you’re seeing the same spots staying wet longer than the rest, that’s usually a red flag. I’ve also started using a moisture meter on the underside of the roof deck when possible—sometimes you can trace the gradient back toward the source, though it’s not foolproof.
I’m all for patience and methodical testing, but I do think there’s value in combining approaches. Hose test for sure, but maybe also try isolating sections with temporary barriers or even colored water (food dye works) to see if you can narrow it down further. Just be ready for some trial and error.
And yeah, don’t underestimate how much water can travel horizontally between layers before it finally drips through. I once spent weeks chasing a leak that turned out to be coming from a vent pipe ten feet away from where the water was showing up inside. Flat roofs are just tricky like that...
I get the logic behind combining methods, but honestly, I’m still not sold on the whole “hose test” thing. Tried it last fall and just ended up with water everywhere, no clearer on where the leak was actually coming from. Maybe I did it wrong, but it felt like I was making more mess than progress. For me, crawling around in the attic after a rainstorm and looking for fresh drips or stains has been way more helpful than any gadget or water test so far. Could just be luck, but sometimes low-tech is less confusing...
Tried it last fall and just ended up with water everywhere, no clearer on where the leak was actually coming from.
That’s pretty much how my first hose test went, too. I had my son up on the roof with the garden hose while I watched for drips in the hallway ceiling. We managed to soak half the siding and still couldn’t pin down the leak. In hindsight, I think we just flooded everything too quickly—didn’t give the water a chance to follow its usual path.
I get what you mean about crawling around after a rainstorm. That’s been more productive for me as well, especially since our flat roof is old torch-down and has a few “mystery spots.” Once, after a heavy spring rain, I found a fresh drip right above an old light fixture—turned out there was a tiny puncture under some gravel that only let water in when it pooled just right. No gadget would’ve caught that.
Low-tech works, but it’s definitely more time-consuming. Still, sometimes you just need to see where nature takes the water... even if it means a few trips into the attic at 2am.
