How Automated Chemical Dosing Works

Why timers stop being adequate

A timer-based dosing regime assumes the system is always in the same condition. It never is. Production load changes, ambient temperature changes, make-up water quality changes, and the demand for treatment chemical changes with all of them.

A fixed dose is therefore either too much or too little most of the time. Too much wastes chemical and can cause its own problems; too little leaves the system unprotected, which with a corrosion inhibitor can be worse than no programme at all.

How the feed is driven

• Proportional to make-up volume. A meter on the make-up line drives the pump, so chemical enters in proportion to the water it needs to treat. The standard method for inhibitor feed.
• Proportional to blowdown. Feed follows what is being lost from the system rather than what is entering it.
• On a measured parameter. Conductivity, pH or ORP drives the output directly, holding a condition rather than delivering a quantity.
• Timed with a limit. Some biocides are dosed on a schedule by design; the automation adds feed limits, lockouts and verification rather than replacing the schedule.

What makes an automatic system stable

A setpoint alone produces an unstable system. Three settings turn an automatic system into a controlled one.

1. Deadband. Without it the output switches constantly around the setpoint, wearing the pump or valve and destabilising the reading.
2. Feed limit. A maximum run time per period. If a sensor fails or a leak develops, this is what stops the system emptying a drum into the water.
3. Lockout and interlock. No dosing without flow; no dosing during a backwash; no dosing when the drum is empty. Each removes a specific way the system could dose into nothing or into the wrong place.

The injection point decides the outcome

A correctly chosen chemical injected in the wrong place is a correctly chosen chemical that does not work. Three requirements have to be met at once.

• Turbulent, well-mixed flow, so the chemical disperses rather than travelling as a concentrated slug.
• Away from surfaces the chemical could attack at neat strength — concentrated acid or caustic against a fitting is a common cause of unexplained localised damage.
• Far enough upstream of the measurement point that the sensor reads the mixed system rather than the injection itself.

Chemical compatibility applies to the injection assembly too. Quill, non-return valve, tubing and any isolating valve all see the chemical undiluted, and the wrong material fails quietly.

The gap automation does not close by itself

This is the point most often missed. A controller calling for a dose is not evidence that a dose was delivered. The pump may be airlocked, the suction line degassed, the injection quill blocked or the drum empty — and in every one of those cases the controller reports normal operation.

Feed verification closes that gap: flow verification on the discharge line, pulse feedback from the pump, or level tracking on the drum compared against calculated consumption.

A separate chemical check completes it. A residual test, an inhibitor level or a corrosion coupon confirms that the chemical not only left the drum but is doing what it was chosen to do.