Irrigation Controller and Backflow Service

Purpose

Service the two devices where a lawn service touches something bigger than the lawn. The backflow assembly is the only thing between an irrigation system full of fertilizer, pet waste and standing water and the drinking water in the house, and in most jurisdictions it is a regulated device with a certified annual test and a report to the water purveyor. The controller is where a season's water is spent or wasted, and where the startup audit's measured numbers either get used or ignored.

Both fail quietly. A fouled relief valve looks identical to a working one from the driveway, and a controller running a rain sensor bypass looks like one that is simply set wrong.

Scope

Covers the annual service visit to the point of connection and the controller: identifying the assembly and its test status, verifying the installation, the certified test or the handoff to a tester, controller power and memory, sensor verification, solenoid and field wiring faults, and entering the season's program.

Does not cover pressurization, head repair or the catch-can audit, which are the LawnCare irrigation startup and audit SOP and normally run first on the same visit. Does not cover programming theory: run-time and cycle-and-soak concepts live in the LawnCare irrigation controller programming HowTo, and this SOP consumes them. Does not cover repair inside the assembly beyond the manufacturer's field kits, or a system with chemical injection, which needs a reduced pressure principle assembly and a purveyor conversation first.

Roles and handoffs

The handoff that fails is the due date: a tester booked after the purveyor's late notice means the customer has already had a letter about their own water service.

Procedure

Step 1 - Identify the assembly and pull its test history before touching it

Record what is actually installed and when it was last tested. Acceptance: assembly type (pressure vacuum breaker, double check, or reduced pressure principle), make, model, size, serial, location, and last test date against the purveyor's due date. Wrong looks like a work order that says "backflow" with no type, which matters because the pass criteria, the elevation requirement and the freeze exposure all differ by type. If the shop does not hold current tester certification, do not open a test cock: one on a pressurized assembly releases a hard stream, and an untrained reassembly can leave the potable supply unprotected while looking finished.

Step 2 - Verify the installation before verifying the device

Check the assembly is installed the way its type requires before any test is run. Acceptance for a pressure vacuum breaker: at least 12 in above the highest downstream outlet, upright, canopy vent clear. Wrong looks like a vacuum breaker set below a head on a rise at the back of the property, which is not a functioning device no matter how well it tests, because its protection depends on that elevation. The height requirement and its measuring point come from the plumbing code your jurisdiction adopted, an IPC or UPC edition rather than a federal rule, so confirm the measuring point in your own code. If the installation is wrong, the finding goes on the report as a defect.

Step 3 - Run the certified test against the form, not from memory

The certified tester runs the field test procedure the purveyor has adopted, with a calibrated differential gauge, filling the form as the readings come. Acceptance: every reading transcribed on the purveyor's own form, gauge calibration current, and each criterion met for that assembly type. A pressure vacuum breaker's air inlet must open at 1.0 psid or more and its check hold at 1.0 psid or more; a reduced pressure principle assembly's relief valve must open at 2.0 psid or more, with its check criteria stated on the same form. Wrong looks like criteria recalled across assembly types, because they differ: transcribe them off the form in front of you. Relieve pressure through the test cocks in the order the procedure gives, and keep your face out of the relief port line.

Step 4 - Take the stop rule on a failed assembly

A failing assembly does not get adjusted until it passes and then reported as a pass. Acceptance: the failure recorded with its actual reading against the criterion, the purveyor notified inside the program's deadline, and either a manufacturer repair kit fitted and a full re-test recorded, or the system left off at the isolation valve. Wrong looks like a system left running on a failed assembly through a repair lead time, which leaves the potable supply unprotected for the whole period. Tell the customer in person what is off and why, and put it in writing the same day. Who carries liability for a cross-connection event is a question for the owner and the shop's insurer, not something settled at the door.

Step 5 - Service the controller and confirm what survives a power loss

Power the controller, check the display, and confirm what the unit retains when power drops. Acceptance: display lit with no error, correct current date and time, and the program present after a deliberate 60-second power interruption at the transformer. Wrong looks like assuming the backup battery holds the program: on most modern controllers the schedule sits in nonvolatile memory and the battery only keeps the clock, so a dead battery gives you a correct program running at the wrong hour, which shows up as watering at noon in August. Kill the branch circuit at the panel and prove the terminals dead with a meter, live-dead-live per NFPA 70E-2021, 120.5, before landing or removing any wire; 29 CFR 1910.333(b)(2) is the controlling practice for that work.

Step 6 - Prove every sensor by making it act, not by reading its light

Test each sensor by forcing its condition rather than trusting an indicator. Acceptance: a rain sensor manually depressed or wetted causes an active zone to shut down and the controller to show the interrupt; a flow sensor shows a rising count on the display with a zone running; a master valve opens and closes with the zone. Wrong looks like a rain sensor bypass switch left on from a service call two seasons ago, which is the single most common cause of a system running in the rain, and it is invisible unless somebody forces the condition. If a sensor does not act, disable the feature in the program and tell the customer, rather than leaving a sensor the customer believes is protecting them.

Step 7 - Chase a dead zone with a meter, at both ends of the wire

For any zone that will not open, measure the solenoid circuit at the controller terminal and again at the valve. Acceptance: resistance at the controller consistent with the other working zones and with the solenoid manufacturer's figure, commonly in the range of roughly 20 to 60 ohms for a 24 volt alternating current irrigation solenoid; an open circuit reads over range and a shorted coil reads near zero. Wrong looks like replacing a solenoid that measures fine at the valve, when the fault is a nicked common wire in a box. Take the reading with the controller output off and the field wire disconnected from the terminal, since measuring resistance on a live circuit gives a meaningless number and can damage the meter.

Step 8 - Enter the program from the measured numbers and print what you set

Enter the season's schedule from the precipitation rate and uniformity the startup audit measured, then capture the finished program. Acceptance: run times derived from the measured precipitation rate per zone, cycle-and-soak split where the run time exceeds the soil's intake, start times set so the cycle finishes before the household's morning demand, and a photograph of every program screen attached to the record. Wrong looks like a run time carried over from last season on a zone whose heads were replaced this spring, which is watering by habit. If no audit numbers exist for a zone, do not invent a run time: schedule the audit and set that zone conservatively with a note.

The record this produces

Two records leave the property. The backflow test report is the purveyor's form: assembly type, make, model, size, serial, location, gauge serial and calibration date, every reading, the verdict, and tester name and certification number. It goes to the purveyor inside the deadline, with copies to the customer and the property card, and it is what keeps the customer's own water service in good standing.

The service record is the shop's: controller make and model, program retained after the power-loss check, sensor results by sensor, solenoid resistance per zone with both reading points for any zone chased, every zone's run time with the precipitation rate it came from, and the program screenshots. The next tech reads it to see whether a run time was measured or guessed, and the office reads the sensor results, because a disabled rain sensor is a customer conversation rather than a note in a truck.

One pass through this procedure

Property: reduced pressure principle assembly at the point of connection, twelve zones, annual test due this month under the purveyor's cross-connection program. The startup audit ran the same morning and handed over measured precipitation rates.

Steps 1 and 2 pass: assembly type, serial and last test date recorded, and the assembly sits above grade with clear access to the relief port.

Step 3 fails. The relief valve opens at 1.6 psid against the 2.0 psid minimum on the purveyor's form, 20 percent below the criterion, with both checks holding. Step 4 takes the stop rule: the reading and the criterion go on the form as a fail rather than the tester nudging the relief and re-running it, the purveyor is notified that day, and because the repair kit is a next-day part the isolation valve is closed and the customer told in person that the irrigation is off until the assembly passes. The kit is fitted next morning and the full test re-run: relief opens at 3.1 psid, checks hold, pass.

Step 6 fails on the rain sensor. The indicator light is on, but wetting the sensor does not interrupt a running zone, and the bypass switch on the controller face is in the bypass position, left by whoever serviced the system previously. Switched back, the test repeats and the zone shuts down within seconds. The customer had been watering through every storm for at least a season, which is the finding that pays for the visit.

Step 7 clears zone 5, which would not open. At the controller terminal, zones 1 through 4 read between 32 and 36 ohms and zone 5 reads over range. At the valve box the solenoid itself reads 31 ohms, inside the same band as the working zones, so the fault is in the field wire between those two points, and the splice is found corroded in a box the winter frost had heaved. Replacing the solenoid would have been the expensive way to learn nothing.

Step 8 enters run times from the audit numbers, splits two zones into cycle-and-soak, and photographs each program screen into the record.

References

• Your water purveyor's cross-connection control program, which sets the test interval, who may test, the form and the reporting deadline used in steps 1, 3 and 4; it is administered by the purveyor under your state drinking water rule, not by a federal standard reaching your shop
• The plumbing code your jurisdiction has adopted, an IPC or UPC edition, for the vacuum breaker elevation and its measuring point in step 2
• NFPA 70E-2021, 120.5 for live-dead-live, and 29 CFR 1910.333(b)(2) as the controlling OSHA practice for the electrical work in steps 5 and 7
• The assembly manufacturer's field repair kit instructions, the only in-field repair this SOP permits
• See related: the LawnCare irrigation startup and audit SOP for the numbers step 8 consumes, and the irrigation controller programming HowTo for the run-time method