Annual Water Treatment System Service Standard

Purpose

This standard guarantees a whole-house treatment system is tested and adjusted once a year against this year's actual water, not last year's settings. Without it, capacity loss is silent: resin fouls a little more each season, a UV lamp keeps glowing past the point its dose is adequate, a filter channels instead of treating, and nobody finds out until laundry stains, an inspection flags hard water, or a lab result comes back positive on something the system was supposed to handle. A visit run to a checklist of glances catches none of that; one run to measured acceptance criteria catches all of it.

Scope

Covers annual and preventive service visits on residential and light-commercial installations carrying any combination of a softener, an iron or manganese filter or neutralizer, a carbon filter, UV disinfection, and a point-of-use RO. It does not cover physical installation (owned by the softener and RO installation SOPs), the baseline captured on day one (owned by the new system startup and water quality baseline standard), well bacterial sampling technique (owned by the well-water bacterial testing protocol SOP), or cooling-tower Legionella service (owned by that SOP). Where resin capacity looks degraded rather than simply due for a scheduled regeneration, this visit hands off to the regenerate-versus-rebed-versus-replace decision matrix rather than making that call itself.

Roles and responsibilities

Safety before opening anything

Unplug a control valve's low-voltage transformer at the wall before opening it or wiring in an accessory such as a dose pump or alarm; a plug is not a formal lockout but it is a real, verifiable isolation point, so verify it rather than trust it was pulled. Where a booster or well pump circuit is hardwired, treat that as electrical work: isolate at the panel, prove dead at the terminals, qualified-person practices under 29 CFR 1910.333(b)(2). Relieve pressure at a bleed valve or a nearby tap before breaking any fitting on a pressurized housing (RO tank, carbon vessel, filter housing); a housing opened under pressure can eject a cap or an O-ring. Potassium permanganate regenerant, used in some iron filters, is a strong oxidizer and irritant on contact; goggles and gloves before handling it, rinse a spill rather than sweep it dry.

Procedure

1. Review last visit's record and brief the customer. Pull the prior baseline or annual record: raw and treated numbers, resin and salt notes, UV lamp install date. Ask what has changed since: a new complaint, a remodel, well work, or anyone newly pregnant, immunocompromised, or an infant in the household. Acceptance: every open item from last visit is closed or carried forward with a stated reason. Wrong looks like walking straight to the equipment with the prior ticket unread, which means the same complaint gets diagnosed fresh twice a year forever. Stop rule: a health-relevant household change on a well with bacterial history moves sampling ahead of every other step, per the bacterial testing protocol. Hazard: none, this step is record review and conversation.

2. Isolate before opening anything. Confirm the transformer unplugged or the hardwired circuit proved dead, and pressure relieved at a bleed valve or a nearby tap before any fitting is broken. Acceptance: both isolations confirmed before the first fitting is touched. Wrong looks like a fitting cracked loose with system pressure still behind it; it will let go, not ease open. Stop rule: a bleed valve that will not open, or a circuit that will not isolate cleanly at the panel, stops work on that component until the isolation itself is fixed; do not work around a bad isolation point. Hazard: stored pressure ejecting a fitting, or live conductors on a mis-labeled circuit; both are why the isolation is proved, not assumed.

3. Test raw water ahead of every treatment device, full panel. Match the parameter list to what is already on file: hardness, iron and manganese, pH, and anything flagged at the last visit. Acceptance: every parameter on the baseline record gets a current reading; none skipped. Wrong looks like testing only hardness because "that is what the softener is for" and missing a rising iron level fouling the resin ahead of it. Stop rule: a raw result crossing a health-based threshold pauses equipment service and moves to the customer results-explanation-and-handoff standard before any wrench touches the system. Hazard: none.

4. Service the softener: salt, bridge check, forced regen, output test. Probe the salt physically for a crusted bridge rather than eyeballing the level through the lid; break one with a broom handle or hot water if found. Force a regeneration and test output hardness afterward. Acceptance: output reads 0 to 1 GPG after regen, the controller's hardness setting matches today's raw reading rather than last year's, and salt is in usable form, not solid mush at the bottom. Wrong looks like output holding at 3 to 4 GPG right after a fresh regen with clean resin and confirmed-correct settings, which points at capacity, not settings. Stop rule: a capacity shortfall with settings and salt both confirmed correct hands off to the regenerate-rebed-replace decision matrix rather than getting force-regenerated a second time on the spot. Hazard: brine fumes concentrate under a closed tank lid; ventilate before leaning in to probe the salt.

5. Service the filtration media: backwash function, air check, media depth, channeling. Run the backwash, watch it complete its full programmed cycle, check media depth against the fill line recorded at install, retest the treated result. Acceptance: backwash completes without an air-charge failure, media loss is within about an inch of the recorded fill line, and treated iron, manganese, or pH is back in band. Wrong looks like a filter that backwashes on schedule while the treated result has not moved, which is exhausted or channeled media letting water pass around itself. Stop rule: media loss beyond that inch, or a treated result that will not clear on a second backwash, means the bed is refilled or rebedded, not run another year on faith. Hazard: this media is a strong oxidizer that stains and irritates on contact; goggles and gloves before handling it, never let concentrated product touch anything organic.

6. Service UV and RO where installed. UV: confirm lamp age against the one-year replacement standard, since a lamp on its second year is out of spec whether or not it still glows, and test the alarm by shading or disabling the sensor to confirm it fires. RO: test TDS in versus out, check tank precharge with the tank drawn empty. Acceptance: UV lamp within its one-year window with the alarm confirmed live; RO TDS reduction in the 90 to 95 percent range, tank precharge at spec (commonly around 7 psi empty). Wrong looks like a UV unit glowing into year two with an owner who has no idea its dose has fallen below the certified minimum. Stop rule: an alarm that does not fire when tested is a disabled safety device, flagged for repair before the visit closes, not noted for next year. Hazard: never look into an open UV chamber with the lamp powered; UV-C burns the cornea in seconds with no pain to warn it happened, so treat the chamber as live until the sleeve and housing are fully closed.

7. Verify treated water at point of use and walk the customer through the record. Draw a sample downstream of every device at an actual fixture and compare it to the raw baseline from step 3. Acceptance: treated results in band for every parameter, shown to the customer side by side with the raw numbers, not summarized verbally alone. Wrong looks like handing over a clean-sounding verbal summary the customer cannot check again in six months. Stop rule: a parameter that does not clear at the tap despite its device testing correctly at its own outlet sends that device back to step 2 before close-out; a fault between the device and the tap is a plumbing problem, not an equipment problem, and gets found before the visit ends. Hazard: none.

The record this produces

This visit's raw and treated results for every tested parameter, salt and media condition notes, UV lamp install date and alarm test result, RO TDS in and out with tank precharge, any item handed off and to which reference or decision matrix, and next year's date. Filed against the equipment record so next year's technician compares against this year, not against install day. Without last year's numbers on file, "still working" is an opinion; with them, it is a trend line.

One run of this standard, filled in

A four-year-old well-served install: 32,000-grain softener, birm iron filter, POU RO at the kitchen. Four-person household, last year's raw baseline 14 GPG hardness and 0.6 mg/L iron.

• Step 1: prior record reviewed, no open items, no household change. Pass.
• Step 2: transformer unplugged; pressure bled at the RO tank valve before disconnecting it for the precharge check. Pass.
• Step 3: raw retest reads 15 GPG hardness, 0.7 mg/L iron, pH 7.1. No health-based threshold crossed. Pass.
• Step 4: salt level fine, no bridge. Controller's hardness setting still read 14 from last year; updated to 15. Forced regen run. Output test immediately after: 3 GPG. Fail. Settings and salt were both confirmed correct going in, so this is capacity, not adjustment. Stop rule taken: handed off to the resin decision matrix rather than run a second regen; softener bypassed so it does not confound the rest of the visit's readings, customer briefed that a follow-up recommendation is coming, not a same-visit resin swap.
• Step 5: iron filter backwash completes fully, media depth within an inch of the fill mark, treated iron 0.05 mg/L, in band. Pass, and it shows the step 4 fault is isolated to the softener, not the whole system.
• Step 6: no UV on this install, noted not applicable. RO TDS in reads 210 ppm, matching this well's mineral load, TDS out reads 14 ppm: a 93.3 percent reduction, inside band. Tank precharge reads 7 psi, at spec. Pass.
• Step 7: RO tap matches step 6. Iron-filtered bath tap reads clear, matching step 5. Softener-fed taps read hard, expected with the softener in bypass pending the resin decision. Findings shown against last year's numbers. Pass for everything in service; softener flagged not-cleared.

When the visit does not run as written

Customer must leave before completion. Finish every testable step and the record; reschedule the walkthrough and any sign-off decision rather than skip it or write it up as if it happened.

A component is out of scope on arrival, such as an undocumented pressure tank or an addition installed since last visit. Note it, test what is on file, quote the new item separately rather than absorb it into the standard visit's time.

No prior record exists, on a first annual visit after a change of ownership. Treat step 3's reading as the new baseline and say so explicitly, rather than comparing against nothing and calling it a trend.

A health-based finding turns up at step 3. Hold equipment service and move to the results-explanation-and-handoff standard before continuing.

References

• See related: the water softener installation SOP and the reverse osmosis system installation SOP, for the physical install steps this standard assumes are already in place.
• See related: the regenerate versus rebed versus replace resin decision matrix, for the call this visit hands off rather than makes.
• See related: the UV disinfection systems reference, for dose math and the lamp and sleeve replacement procedure.
• NSF/ANSI 55, Ultraviolet Microbiological Water Treatment Systems, source of the one-year lamp replacement convention cited above.
• 29 CFR 1910.333(b)(2), electrical work practices for qualified persons, for any hardwired pump circuit isolated during this visit.