Controller Concepts: That Critical Sample Stream

There’s a lot said in the common phrase “automation system”. That box on the wall needs more than a power cord and mounting screws to smooth out your water’s chemical management. After you’ve installed more-than-adequate feed systems for your controller (PrP 13), the next critical item in the system is the sample source. Establishing a representative sample stream is critical to its effective operation.

A typical sample is no more than five minutes “behind” the pool, and the pool, even a big one, responds to chemical treatment far quicker than one might guess. If there’s a good circulation system, especially with wall and floor inlets, the pool might reflect the latest feed cycle in ten or fifteen minutes. If the pool is smaller, especially if it’s in use, the electrodes in the sample water could sense that chemical change in half that time. The “sample delay” is the time between the controller’s shut off ? having just achieved the desired set-level value ? and the reversal of that rising value rounding downward again. The “pool response time” is the time between the controller’s command to “go” and a reversing of the chemical value that had just passed through set level downward. This time includes the sample delay, as well as the consumption-verses-recovery rate of the pool water as the feeders dose the busy pool. The whole automation curve looks a lot like a sine wave, the flatter it is the better the sample, the feed rate, and the circulation.

What can cause large peaks and valleys in pH or sanitizer residual? Sample lines too long, circulation too slow or poorly distributed, sample-loop lines too big… That’s right; cheap one-half-inch tubing is about six times better (faster) than a one-inch, schedule-80, beautifully installed PVC sample loop. Most samples like 2 to 5 GPM, which scoots through tubing with little delay.

Here are a couple of great examples of how not to do it. A pool with a huge problem automating ? lagging response followed by massive overshoots ? was found to have the sample tap taken from the end of an eight-foot-long piece of six-inch PVC ? stubbed out for a never-installed heater. That huge piece of pipe became part of the sample loop! At .03 feet per second flow velocity, 3 GPM took four and one-half hours to make it through that hunk of “sample line”! Then there was the skimmerless 100-foot pool with one main drain and one, just one, four-inch inlet at the shallow end. Now the whole length of the pool became part of the loop, with disastrous results when a controller was installed!

~ kw