Automated Chemical Controllers and Feed Systems

Automation can adjust disinfectant or pH more consistently than occasional manual dosing. It cannot decide whether its own sample is representative, whether the probe is clean, or whether the chemical reached the intended venue.

Treat the controller, sample loop, sensors, flow switches, output wiring, pumps or valves, chemical supply, injection points, alarms, and recirculation system as one control loop. One weak link can make a polished display dangerously misleading.

What the controller actually knows

A controller reacts to signals. Depending on the installed system, those signals may include pH, oxidation-reduction potential, direct disinfectant measurement, temperature, tank level, and sample flow.

The controller does not directly observe the whole pool. Its probes see the water moving through a small sample chamber. Stagnant sample water, a restricted strainer, air in the line, a fouled probe, or an injection point too close to the sample pickup can create a reading that does not represent bathers' water.

ORP also is not a free-chlorine test. It is an electrical measurement affected by disinfectant activity and several water conditions. Use the controller value for the purpose approved in the facility plan. Use an independent test to determine the actual bulk-water chemistry.

Trace the control loop

Before changing anything, identify:

• the venue and recirculation system being controlled;
• the sample pickup and return locations;
• every probe, flow switch, and sample-line component;
• each controller output and the feeder it commands;
• the product connected to each feeder;
• the injection location and required separation from other chemicals;
• recirculation, no-flow, tank-level, door, or other interlocks;
• local and remote alarms;
• approved setpoints, limits, delays, and operating mode.

Follow labels and wiring. A handwritten pipe label is useful evidence, but it is not enough by itself. Compare it with the permit records, facility diagram, and exact equipment manuals.

Verify the sample stream first

Look through the sample chamber. Water should move steadily without trapped air, heavy debris, scale, algae, or obvious chemical staining. Confirm the sample valve position, strainer condition, tubing direction, and return path.

Check that the flow switch recognizes a real loss of sample flow. Do not defeat a switch to keep the display active. A controller that cannot detect stagnant sample water may continue making feed decisions from an old reading.

The sample pickup must represent treated venue water without seeing a concentrated chemical slug. If readings swing when a feeder pulses, inspect the hydraulic arrangement and compare it with the approved design.

Compare with independent water tests

Use Testing and Recording Public Aquatic Water Quality to collect and record the required independent sample. Compare disinfectant, pH, temperature, and any other controlling parameter with the controller at the same actual time.

Do this before adjusting a setpoint. A low display could mean real demand. It could also mean a fouled sensor, bad calibration, lost sample flow, exhausted chemical supply, failed feeder, or an invalid independent test. Each cause needs a different response.

When the values conflict, repeat the independent test with a valid method, then inspect and calibrate the sensor exactly as its manual directs. Do not turn a setpoint until two disagreeing numbers happen to meet.

Check calibration without hiding drift

Record the reading before calibration. Inspect the sensor and sample cell. Clean only with the manufacturer-approved method, use valid standards when required, and allow the value to stabilize.

Record the adjustment and post-calibration result. A large or repeated correction is diagnostic evidence. It can point to an aging sensor, poor maintenance, sample contamination, unstable water, grounding problems, or a method error. Escalate the cause rather than normalizing constant recalibration.

Verify feed response and safeguards

Observe each output separately. Confirm the controller calls for the correct feeder, the feeder responds, and the intended product moves toward the correct injection point. Check tubing, check valves, fittings, tanks, suction assemblies, and containment for leakage or crystallized residue.

Then verify the safe stop. Under the authorized test method, loss of required recirculation or sample flow should stop chemical feed when the design requires that interlock. Confirm alarms appear where the operating plan expects them.

Never test a safeguard by creating an uncontrolled chemical release. Use the manual's test mode, an approved simulation, or a qualified service procedure. Electrical, programming, and feeder repairs stay within the boundaries in Recurring Maintenance, Repair, and Licensed-Trade Boundaries.

Manual mode is still controlled operation

Manual feed is not a shortcut around a failed controller. If the facility plan permits temporary manual operation, follow its testing frequency, dosing authority, staffing, communication, and documentation requirements. Keep failed automation clearly identified so another operator does not assume it is protecting the water.

If you cannot maintain the required chemistry and circulation under the approved fallback, restrict use and close under the facility procedure.

Chemical-room checks

Inspect ventilation, access control, lighting, spill containment, labels, Safety Data Sheets, emergency equipment, and separation of incompatible products before handling the feed system. Use Chemical Labels, SDSs, PPE, Transport, and Storage and Incompatible Pool Chemicals and Mixing Prevention.

Never swap a product because the container fits. Concentration, formulation, tubing material, pump output, controller configuration, and injection compatibility all matter.

Operating record

Record the independent result beside the controller value, not in a separate story that must be reconstructed later. Include setpoints, mode, flow, alarms, chemical levels, calibration, interlock checks, adjustments, retests, restrictions, and the person making the decision.

That record should answer three questions fast: What did the water measure? What did the system believe? What did the operator do next?