We quantified it the only honest way: by metered flow. With the ridge valve open and my house isolated, we ran the pump into a calibrated test meter for fifteen minutes. The flow averaged 68 gallons per minute, sustained, with no fixture open on my property. Sixty-eight gallons a minute, drawn through a main that should have been serving no one west of the ridge. Over an hour, that was more than four thousand gallons. Over a day, close to a hundred thousand.
I wrote the numbers in the logbook under Ray's readings, and Ray signed each page. We photographed the gauges at each stage, with the logbook and a timestamp card in frame. We did not estimate. We did not round up to make it sound worse. Sixty-eight gallons per minute was bad enough on its own, and it had the virtue of being true.
When we were done, I closed the ridge valve again and restored my house service. Pressure came back to 56 psi and stayed there. My house had water. Ridgecrest, west of that valve, did not, because the water it had been drawing was not water it had a right to draw through my main.
Ray packed his gauges and looked at the ridge. "You know what you've got here," he said. "You've got a system that's sound until you open it to a load it was never designed to carry, and never permitted to carry. That's not a failure. That's a tap."
Ray ran the test a second time, shorter, to confirm the first run had not been a fluke of a pump cycle or a gauge sticking. Same setup, same isolation, same result: pressure held with the ridge valve closed, fell within minutes when it was opened, recovered when it was closed again. Repeatability is what separates a reading from an anecdote, and Ray knew it without me saying so. He logged the second run on a fresh page and signed that one too.
We also checked my house fixtures during the open-valve run, one by one, to rule out a draw on my side. Every tap was off. The stock tank float was closed. The barn line was valved out. There was no other demand on my system except the one west of the ridge. I wrote that down as well, because a good test does not just prove what is happening. It rules out what is not.
I nodded. I had the numbers now, in a signed logbook, with a witness who knew what they meant. Numbers do not shout. They do not need to.
The dye test was Tom Alvarez's idea, in part, because he understood what a neighborhood needs to see to believe a thing.
"A pressure reading is a number," he had said, when I told him about the test with Ray. "People argue with numbers. They don't argue with color in a glass."
We set it up on a clear morning, with witnesses, because a test done alone is a story, and a test done in front of witnesses is a record. Tom was there. Ray Delgado was there. I had also invited a county environmental technician, a woman named Susan Pruitt, who had agreed to observe and to take independent samples, because a county witness carries weight that a neighbor and a friend do not.
The plan was simple, and I explained it before we started, so no one could later say they had not understood it. We would introduce a measured dose of fluorescein tracer dye into my six-inch main at station four, just east of the ridge, with the ridge valve open and my house isolated. Fluorescein is non-toxic, approved for water tracing, and visible at very low concentrations as a bright green-yellow color under normal light. If the main fed Ridgecrest through a tap under the berm, the dye would appear in Ridgecrest samples within a predictable travel time. If it did not, the dye would stay in my system and flush through my own fixtures, and I would have learned something else.
We closed my house valve, opened the ridge valve, and injected the dye at station four at 9:14 a.m. Ray logged the time, the dose, and the injection point. Susan took a baseline sample from my wellhead, clear, and a baseline sample from a Ridgecrest hose bib that Tom had arranged access to, also clear. Both samples were labeled, sealed, and photographed.
Then we waited.
At 9:41 a.m., twenty-seven minutes later, Tom called from the Ridgecrest sample point. "You need to see this," he said.