Choose a Water Filter From Your Test Results

Start with the result, not the product aisle

Write one treatment requirement for each confirmed problem: contaminant, untreated concentration, unit, target, required household flow, and sample date. Then decide whether the target applies to drinking water at one tap or to every gallon entering the house.

Do not choose a system from taste, odor, color, a salesperson's free screening alone, or a high TDS number. Those observations help select tests. They do not identify every health contaminant, and TDS does not tell you which dissolved substances are present.

Separate urgent health findings from aesthetic problems

If multiple problems coexist, treatment order matters. For example, sediment may foul carbon or UV equipment; iron can foul softener resin; pH can affect several processes.

Understand what certification says

NSF/ANSI 42 covers selected aesthetic effects such as taste, odor, chlorine, and particulate claims. NSF/ANSI 53 covers specific health-effect reduction claims. NSF/ANSI 58 applies to reverse-osmosis systems, and other standards apply to UV, softeners, and microbiological devices.

The number is not a universal badge. A model certified under NSF/ANSI 53 for one claim is not automatically certified for lead, arsenic, cysts, or every other claim. Verify the exact manufacturer, model, replacement cartridge, standard, and contaminant-reduction claim in the certification body's current listing.

Also read the certified conditions: challenge concentration, capacity, flow, pressure, temperature, and replacement interval. Certification cannot compensate for use outside those limits.

Choose point-of-use or point-of-entry deliberately

Point-of-use treatment serves a drinking tap, refrigerator, or small group of outlets. It often makes sense when the target is ingestion and whole-house treatment would be wasteful. Point-of-entry treatment serves the building and may be justified for sediment, hardness, staining, odor throughout the home, or a contaminant whose exposure route requires broader control.

Compare:

• Gallons per day that actually need treatment.
• Peak flow needed without unacceptable pressure drop.
• Whether showers, laundry, appliances, or plumbing need protection.
• Wastewater or reject flow.
• Drain, electrical, bypass, and space requirements.
• Failure mode and alarm behavior.
• Annual consumables, sanitation, service, and verification tests.

Do not install a drinking-water-only device as if it certifies every tap.

Convert the lab report into vendor questions

Send the same brief to every vendor:

1. Which exact contaminant and incoming concentration is each stage designed to reduce?
2. What independent certification or published performance data supports the claim?
3. What pH, iron, manganese, hardness, turbidity, temperature, and pressure limits apply?
4. What service and peak flow can the system deliver at an acceptable pressure loss?
5. What pretreatment is required and why?
6. What is the rated capacity before cartridge, media, membrane, resin, or lamp service?
7. What happens when the device reaches exhaustion or loses power?
8. How will performance be verified after installation and over time?
9. What waste, backwash, brine, or reject stream is produced?
10. What exact annual ownership cost should be expected?

Reject proposals that cannot connect equipment to your measured result.

Common treatment branches and their limits

Sediment cartridges remove particles within their rating but do not remove dissolved nitrate or arsenic. Activated carbon can address selected organic compounds, taste, odor, or disinfectants when certified for them; it is not universal disinfection. Reverse osmosis can reduce specific dissolved contaminants at a drinking tap when the exact system is certified and maintained, but it produces reject water and needs pretreatment under some conditions.

Ion-exchange softeners target hardness and may handle limited dissolved iron within design conditions; they do not make microbiologically unsafe water safe. UV can inactivate microorganisms when water clarity, dose, flow, lamp condition, and pretreatment meet design requirements; it does not remove nitrate, metals, sediment, or many chemicals.

Verify the installed system

Collect an untreated sample and a treated sample at locations selected with the laboratory. Test after commissioning at the time recommended for the contaminant and treatment. Save model numbers, settings, installation readings, certification listings, and replacement parts.

Create a maintenance record containing pressure drop, meter totals, alarms, cartridge or lamp dates, media or membrane service, sanitizing, and follow-up results. A clear glass of water is not a performance test.

Build a treatment train in the right order

Some reports point to one treatment process. Others require a sequence. The sequence matters because one device can protect—or interfere with—the device after it. Sediment may need to be controlled before a fine membrane. Hardness or iron may create scale on equipment that was selected for a different contaminant. A disinfecting stage may require water to have enough clarity for the process to work as designed. The correct order comes from the complete water chemistry and the exact equipment limits, not from arranging products by price.

Ask the designer to draw the proposed path from the well connection to every treated outlet. The drawing should name each stage, its purpose, its rated service flow, the pressure it consumes, the waste or drain connection it needs, and the condition that triggers maintenance. If a stage has no defined job tied to a test result or to protecting another stage, ask why it exists.

Also decide where untreated water should remain available. Outdoor irrigation often does not need the same treatment as drinking and bathing water. Treating unnecessary flow can enlarge equipment, increase backwash or replacement frequency, and make troubleshooting harder. That is a design decision, not a reason to bypass treatment required for health protection.

Size for peak flow, not only daily gallons

A household can use a moderate amount of water per day and still have a demanding short peak when showers, laundry, fixtures, or appliances overlap. Treatment that looks adequate by daily volume can create disappointing pressure if its service-flow capacity is too small. Ask for the design flow used in the proposal and how it relates to the house, pipe size, pump system, and likely simultaneous demand.

Capacity also needs a time dimension. A cartridge has a replacement condition. Media has a service or regeneration interval. A membrane has a production rate and recovery behavior. A UV system has lamp, sleeve, electrical, and pretreatment requirements. The proposal should explain what happens between service events, how the homeowner knows capacity is being approached, and whether the system fails visibly or can lose performance without an obvious change at the tap.

Record pressure before and after the complete treatment train when it is new and operating normally. That baseline makes a future restriction much easier to locate. Pressure loss should be interpreted at a stated flow; a no-flow gauge reading cannot describe how the equipment behaves during household demand.

Calculate the full ownership burden

The purchase price is only one line. Build an annual operating view that includes cartridges, media, membranes, lamps, salt or chemicals, laboratory verification, electricity, service visits, sanitizing, water sent to drain, and any pretreatment required to protect the main device. Include the owner tasks and the consequence of missing them. A cheap unit that depends on frequent specialty cartridges may cost more—and be less reliable—than a more expensive design with a clear maintenance path.

Ask whether replacement parts are standard, proprietary, locally stocked, or available only through one dealer. Document the exact part numbers before buying. If an alarm, meter, or controller is essential to safe operation, find out what the alarm actually detects. A light that confirms power is not necessarily proof that a contaminant is being reduced.

For each recurring item, write down both a calendar interval and a condition-based trigger. The condition may be a pressure drop, treated volume, alarm, water-quality result, or measured loss of performance. Calendar-only replacement can waste usable capacity; symptom-only replacement can allow performance to fail before anyone notices.

Require a commissioning record

Installation is not finished when water flows through the equipment. A useful commissioning record identifies the installed model and serial numbers, valve positions, settings, initial meter values, untreated and treated sample locations, pressure readings at a defined flow, drain operation, and the first verification-test plan. It should also state which outlets receive treated water and which remain untreated.

Keep photographs of the finished piping and readable labels. Mark the normal valve positions without covering manufacturer information. Store manuals and certification listings with the laboratory report that justified the design. If the installer changes the proposed model, require the same contaminant claim and sizing check for the substituted model; similar-looking equipment is not automatically equivalent.

Plan the first follow-up before the installer leaves. The timing should match the contaminant, treatment process, and laboratory guidance. Compare untreated and treated samples where that comparison is meaningful. If the result does not meet the agreed target, investigate setup, sampling, flow, pretreatment, and equipment selection instead of simply adding another device.

Recognize when the design needs to be revisited

Treatment is based on a water condition measured at a point in time. Revisit the design after a major change in the source, flooding, well repair, sustained changes in taste or appearance, a new health-related result, unusual pressure loss, or repeated early exhaustion of consumables. A system that worked for the original report may not address a new contaminant or a changed concentration.

Use a simple decision record:

1. What changed in the untreated water, operating pressure, or household demand?
2. Which stage was expected to address it?
3. What measurement shows that stage is or is not performing?
4. Is the cause maintenance, hydraulic sizing, changing chemistry, or a different source problem?
5. What test will confirm that the correction worked?

Do not solve recurring failure by stacking products without answering those questions. More equipment adds valves, pressure loss, maintenance, and failure points. The best treatment train is the smallest one that addresses the measured problem at the required flow and can be verified over time.

Safety boundary

If a health-related result exceeds guidance, use another safe source and involve the local health or environmental department. Do not rely on a newly purchased unit until the treatment has been matched to the exact contaminant, installed correctly, and verified. Do not bypass health-protective treatment, open pressurized housings, handle UV electrical equipment, or mix treatment chemicals without the manufacturer's safe procedure and appropriate qualifications.