Tradeoffs The Post-Mortem John Diefenbach

The Rain Barrel Plan That Drained Itself

I built a water system nobody could touch — then watched it fail for exactly the reason I told everybody else to watch out for.

What Was Supposed to Happen

The plan was clean. I mean genuinely clean — the kind of clean where you draw it out on paper and it holds up every time you look at it. Four 55-gallon drums daisy-chained off the back roof, gravity-fed to a spigot at the garden line. No pump. No electric. No meter, no utility company, no monthly bill with some fee on it that nobody can explain. Just rain, barrels, and physics doing what physics does for free.

I'd been watching the county water rates for two years before I built it. That's exactly what they want you to think — that the increases are about infrastructure, about pipe maintenance, about whatever word they're using that month. I wasn't buying it then and I'm not buying it now. The rate goes up every year, right on schedule, like somebody set an alarm. So I set my own plan on a schedule: collect the rain, use the rain, owe nobody.

I figured I'd have the whole thing running before the first hard rain of fall. Enough capacity to water the garden through a dry stretch, maybe supplement the indoor line if things got tight. I'd tested the fittings. I'd read three different manuals — not the manufacturer manuals, those are written by lawyers — actual how-to documents from people who'd done it. I pressure-checked the connections with a hand pump before I ever attached a barrel. I was not winging this. I was prepared. I had thought of everything that could go wrong.

That last sentence is the one that should've scared me.

How the Systems Around You Work

Clear explanations of government, business, technology, finance, healthcare, and everyday bureaucracy.

Learn more

What Actually Happened

The first rain filled the barrels in about six hours. I stood out there watching the downspout diverter kick in and felt something I don't feel that often — satisfied. Not excited, not relieved. Just satisfied. The system worked exactly the way the paper said it would. I went inside. I slept fine.

The second rain came four days later. Heavier. I didn't stand outside and watch that one because I didn't need to. The system was proven. I had other things to do.

Somewhere in that second rain, barrel number three developed a slow weep at the lower coupling. Not a burst. Not a flood. A weep — maybe a tablespoon of water every few minutes running down the side of the barrel and pooling against the foundation. I didn't find it for eleven days. By then the ground against the east wall was saturated, the water had been wicking under the sill plate, and I had a four-foot section of floor joist that had started to soften.

I caught it before it became structural. Barely. I pulled the sill plate section, let it dry out over three weeks, treated it, sistered a new joist alongside the soft one. The repair took longer than the original build. Cost more in materials than two years of the water rate I'd been trying to avoid.

The barrel system still works. I fixed the coupling. The weep is gone. But the four barrels I was so proud of now sit in a configuration I check obsessively — twice a week, sometimes more — looking for the next thing I won't see until it's already doing damage. The system I built for peace of mind became the thing I can't stop thinking about. The utility company doesn't keep me up at night. My own barrels do.

Where It Turned

The coupling on barrel three was a standard threaded brass fitting. I'd hand-tightened it and then gone a quarter-turn with channel-locks, which is what every source I read said to do. What those sources did not say — or what I skimmed past — was that brass threads on plastic barrel bungs need PTFE tape on every pass, not just when you think you might have an imperfect thread. I thought I had a perfect thread. I thought I could feel the difference.

That's where it turned. Not the second rain, not the eleven days I didn't check — the moment I decided my hands knew something that tape couldn't improve on. It was a five-second decision. Not even a decision, really. More like an absence of one. I just didn't reach for the tape because nothing told me I needed to.

And here's the part that sits with me: the fitting looked right. It seated clean. It passed the hand-pump pressure test. For the first rain, it held. It held well enough that I had evidence — actual evidence — that it was fine. That evidence was wrong, but it was real evidence. I didn't ignore a warning sign. There wasn't one. The warning sign was the absence of a warning sign, and I had no system for reading that.

The plan failed in the exact space where it looked most solid. Not at the joints I'd worried over. Not at the diverter I'd rebuilt twice before I was happy with it. At the fitting I'd stopped thinking about because it had already passed a test. That's the thing about a near-miss. It doesn't announce itself. It just quietly becomes your floor joists.

What Changes

I don't trust the first test anymore. That's the concrete thing. Not as a philosophy, not as a lesson I learned — as a rule I actually follow now, with my hands, in the physical world.

Every fitting on every water system I build gets taped. Full stop. Not the ones I'm uncertain about. All of them. Because I cannot trust the feeling of certainty, and that's not a character flaw I'm working on — it's a material property of the situation. The feeling of certainty and the presence of certainty are not the same thing, and I have the joist sister work to prove it.

I also built a check interval into the system itself. Not a mental note — an actual physical trigger. I tied a strip of bright orange surveying tape to the lower coupling on each barrel. Every time I walk past, the tape is either dry or it isn't. I don't have to remember to check. The system checks for me. I am not relying on my own diligence, because my own diligence already failed once when I thought I had everything handled.

The thing I haven't fixed — and I'll say this plainly because this column doesn't do redemption arcs — is the reflex to skip the last ten percent of a job when the first ninety looks clean. I know it's there. I know it cost me three weeks and a sistered joist. But knowing something is there and having it gone are two different inventories, and right now I only have one of those. The tape is orange. I check the fittings. The floor is dry. That's where it stands.