New System Startup and Water Quality Baseline Standard

Purpose

This standard guarantees every install leaves with a proven-working system and a written baseline the shop can stand behind at the first future dispute: a warranty claim, a "this stopped working" call, or a slow drift nobody notices until a symptom shows up. Every one of those gets judged against a number captured while the crew was still on site. Without a baseline written down before the truck leaves, every future service call starts by guessing what normal was for this house.

Scope

Runs immediately after any completed residential or light-commercial water treatment installation, whatever combination of softener, iron or manganese filter, neutralizer, carbon filter, UV, or point-of-use RO was installed, before the crew leaves the first visit. It does not cover the physical plumbing and mounting steps (owned by the softener and RO installation SOPs), the ongoing annual visit (owned by the annual water treatment system service standard), or well bacterial sampling technique (owned by the well-water bacterial testing protocol SOP), beyond confirming a bacterial sample was scheduled where the system's design assumes clean raw water.

Roles and responsibilities

Safety before energizing or testing

A forced regeneration or flush cycle discharges water, and on a softener concentrated brine, to a drain; confirm the drain path is connected and clear before the first forced cycle, and stand clear of a discharge hose under initial flow, it moves. A housing hand-tightened for the dry fit (an RO membrane housing, a carbon vessel, a filter housing) gets brought to pressure slowly at the supply valve, watching every joint, not opened to full pressure at once; a joint that lets go under a sudden full-pressure test can spray. Never look into an open UV chamber with the lamp powered; treat the chamber as live until the sleeve and housing are fully closed. Where a booster or well pump circuit was wired as part of this install, prove it dead before the final terminal connections under 29 CFR 1910.333(b)(2), qualified persons only, then prove it correctly loaded once restored rather than assume the wiring diagram was followed without error.

Procedure

1. Confirm the install itself is complete before any water-quality test begins. Pull the install SOP's own acceptance criteria, leak-free fittings, correct bypass direction, drain air gap where required, and confirm each is met on this visit's own record. Acceptance: every install-SOP acceptance item checked and initialed here, not assumed from having just finished the work. Wrong looks like moving straight to a baseline test on a system with a fitting still weeping, which puts a false signature into the baseline that has nothing to do with treatment performance. Stop rule: any unresolved leak or plumbing defect stops the baseline sequence until fixed; a baseline captured over a live defect is worse than no baseline, because it gets trusted later. Hazard: none beyond the install's own, already covered.

2. Bring the system up slowly and run its first full cycle unassisted. Open the supply valve gradually, watch every joint disturbed during install, and run the softener's first regeneration, or the RO's first flush, or the UV's first proving run, to completion. Acceptance: no weep or spray at any joint through slow pressurization, and the first automatic cycle completes on its own without a fault code or manual intervention. Wrong looks like a cycle that needs the tech to manually advance it past a stuck step, a controller or valve problem the customer will meet alone later. Stop rule: a cycle that will not complete unassisted is not left running while the crew packs up; resolve it, or reschedule with the customer told plainly why. Hazard: a discharge hose moves under initial flow, keep it secured and stand clear; do not stand over a housing you just brought to pressure for the first time.

3. Sample raw water ahead of every device, full panel. Draw from a point before any treatment reaches it, the same tap convention used for annual testing, so this year's and every future year's numbers are comparable. Acceptance: every parameter the system was sold to address, plus the general panel, hardness, iron or manganese, pH, TDS, has a raw number on the record. Wrong looks like testing only the parameter the sales visit flagged and missing that the well's iron has moved since that quote was written, which mis-sizes everything downstream. Stop rule: a raw result crossing a health-based threshold not already accounted for in the system's design pauses startup and moves to the results-explanation-and-handoff standard before the system is declared complete. Hazard: none.

4. Sample treated water at point of entry and point of use, same panel as step 3. Acceptance: every device meets its own design target at its own outlet, softener output 0 to 1 GPG, RO TDS reduction in the 90 to 95 percent range on matched raw and treated samples, iron or manganese and pH at the filter or neutralizer outlet back in band. Wrong looks like testing hardness at the kitchen tap and calling the softener proven, when that tap draws through fixtures downstream of the softener's own sample port. Stop rule: any device missing its own target at its own outlet does not get baselined as passing; troubleshoot before writing down a number that will be trusted for years. Hazard: none.

5. Test every alarm and safety function once, deliberately. Trip the UV's low-dose or no-flow alarm per the manufacturer's test method, trip a high-salt or brine-fault alarm if the controller has one, confirm a booster pump's low-pressure cutoff actually cuts off. Acceptance: every alarm fires when its trigger condition is created on purpose, and clears when the condition is removed. Wrong looks like assuming an alarm works because the unit is new; a wiring fault at install disables an alarm as completely as age does. Stop rule: an alarm that does not fire is fixed before the visit closes; a system commissioned with a dead safety alarm fails silently, on the customer's schedule, not the shop's. Hazard: this step deliberately drives a fault condition; for a UV low-dose test, disturb only the electronic sensor with the chamber sealed, never open the chamber to interrupt the lamp physically.

6. Build the baseline record and walk the customer through it. Show raw and treated numbers side by side for every parameter, in plain terms, and explain this specific set of numbers is what every future visit gets compared against. Acceptance: the customer can repeat back, in their own words, what the baseline means for their water, and signs acknowledging they were shown it. Wrong looks like handing over a printed report with no walkthrough, filed unread and useless six months later when a complaint comes in. Stop rule: no signature, no completed startup; if the customer is unavailable, this step is rescheduled, not skipped, because an unwitnessed baseline is weaker in a dispute. Hazard: none.

7. Schedule the first follow-up and file the warranty registration. Set the interval the equipment and the water call for, commonly a first check at 60 to 90 days for a new well or a system correcting a significant problem, annual after that, and register the warranty using the baseline record's install date and serial numbers before leaving. Acceptance: a follow-up date is on the calendar, not left as "call if there's a problem," and the warranty is registered before the crew leaves the property. Wrong looks like a claim eighteen months later the manufacturer cannot find because it was never registered. Stop rule: a registration portal that is down, or a required serial photo that could not be captured, gets a same-day office follow-up task, not an indefinitely open item. Hazard: none.

The record this produces

Baseline raw and treated results for every tested parameter, install-SOP acceptance confirmation, every alarm's deliberate test result, RO tank precharge and TDS in and out, warranty registration confirmation, signed customer acknowledgment, and the first follow-up date. Filed as day one, the number every later comparison in the annual standard, and every troubleshooting call, starts from.

One run of this standard, filled in

A new well, softener and iron filter installed the same visit, no UV. Sale-visit test: 16 GPG hardness, 0.4 mg/L iron, coliform negative.

• Step 1: install SOP's acceptance items all confirmed, no weeping fittings, bypass correct, drain air gap present. Pass.
• Step 2: supply opened gradually, no weep on any joint; softener's first regen and iron filter's first backwash both complete unassisted. Pass.
• Step 3: same-day raw retest reads 17 GPG hardness, 0.5 mg/L iron, pH 7.3; coliform not retested this visit, already negative within the sampling protocol's window. Pass, no health threshold crossed.
• Step 4: softener output after first regen tests 4 GPG against a 0 to 1 target. Fail. Cause traced to the bypass valve, opened partway during step 2's gradual pressurization and never returned fully to service position, blending raw water in at the softener's own outlet regardless of resin condition. Valve seated fully to service, output retested: 0 GPG. That is the number baselined, with the first reading logged too so the record shows the bypass was the fault, not the resin.
• Step 5: controller's display fault flag confirmed by killing power briefly and restoring it; no other alarms on this equipment set, no UV, no booster pump, noted not applicable.
• Step 6: numbers shown side by side, raw 17 GPG and 0.5 mg/L iron against treated 0 GPG and 0.05 mg/L; customer restates it correctly and signs.
• Step 7: follow-up scheduled at 60 days given the new well; warranty registered same day with serial numbers and install date.

When the startup does not run as written

Customer not present at completion. Complete every step and the record; reschedule the walkthrough and signature within a defined short window, such as before the follow-up call, rather than skip or backdate it.

A parameter the system was not sized for turns up at step 3, such as a well testing positive for something outside the quoted scope. Stop, do not declare startup complete, move to the results-explanation-and-handoff standard to scope the conversation about what the system was and was not sold to fix.

A specified part is missing at final assembly, such as a backordered fitting or accessory. Document what is substituted or deferred, and do not baseline a device running on a substitute as though it were the final configuration; flag it for a return visit when the correct part arrives.

A registration portal is down. A same-day office follow-up task, not an indefinitely open item, per step 7.

References

• See related: the water softener installation SOP and the reverse osmosis system installation SOP, for the physical steps this standard picks up after.
• See related: the annual water treatment system service standard, which this baseline exists to be compared against every year after.
• NSF/ANSI 55, for the UV alarm and dose-verification test method referenced in step 5.
• 29 CFR 1910.333(b)(2), electrical work practices for qualified persons, for any pump circuit proved dead before final connection.