What Whole House Water Filters Actually Remove: A Contaminant-to-Technology Matrix

How to read this matrix

Rows are contaminants commonly tested in residential drinking water. Columns are the six filtration technologies a whole house system might include. Each cell answers a single question: is this technology recognised as effective against this contaminant? "Yes" means the technology is recognised by EPA, NSF, or both, and a properly certified product is independently verified. "Partial" means the technology can reduce the contaminant under some conditions but is not the primary recommended treatment. "No" means the technology is not effective. "n/a" means the contaminant does not apply to that technology.

Three caveats matter. First, removal claims require certification. A carbon block can pass NSF/ANSI 53 lead reduction or it can be a hardware-store cartridge with no testing - the technology is the same, the product testing is not. Second, removal effectiveness depends on contact time, bed depth, influent concentration, and pH. A GAC bed with 30 seconds of contact time will outperform the same bed with 5 seconds. Third, a single technology rarely handles every contaminant in real water; the typical whole house train combines two to four technologies in series.

Five things this matrix does not tell you

The matrix is a starting point, not a specification. Five caveats apply to every cell.

1. NSF certification is required for the "Yes" to be verified. A carbon block exists in two regulatory states: NSF/ANSI 53 certified for a specific contaminant claim, or not. The technology can adsorb lead in either case; only the certified product has been tested in an accredited lab to demonstrate it actually does. See NSF standards for the full list and verification procedure.

2. Influent concentration matters. A GAC bed treating 0.1 mg/L lead at 1 GPM will perform very differently from the same bed treating 5 mg/L lead at 8 GPM. Certification testing is done at defined challenge concentrations and flow rates that reflect realistic residential conditions; results outside those bounds may not hold.

3. Bed depth and contact time matter. Activated carbon adsorption is a contact-time phenomenon. A whole-house GAC tank with 30 seconds of empty bed contact time will outperform a small inline cartridge with 3 seconds of contact time, even if both nominally contain the same media. Carbon block cartridges achieve more contact time per unit volume than GAC because the block forces water through a denser matrix.

4. Some contaminants enter the water after the POE filter. Lead is the canonical example. Lead service lines and lead-tin solder joints in pre-1986 plumbing leach into water on the premises side of the meter. A carbon filter at the meter does nothing about lead released between the filter and the kitchen tap. EPA's Lead and Copper Rule treats lead primarily as a corrosion problem requiring corrosion control treatment at the utility and lead service line replacement, not as a source-water problem. POU filters at the kitchen sink are the right intervention. See our lead page.

5. Some technologies require pretreatment. A UV chamber cannot disinfect water that contains visible turbidity or iron precipitates, because UV-C cannot penetrate them. EPA's UV Disinfection Guidance Manual specifies a 5-micron sediment polish upstream of any UV stage. A whole-house RO membrane will foul rapidly without sediment and chlorine pretreatment; chlorine destroys most polyamide membranes. Stage order in a multi-stage train is a real constraint.

The standards behind each "Yes"

Each "Yes" in the matrix corresponds to a specific NSF/ANSI standard, EPA recognition, or CDC guidance document. The cluster looks like this:

• NSF/ANSI 42 covers aesthetic effects: chlorine, taste, odour, particulates Class I to VI.
• NSF/ANSI 53 covers health effects: lead, mercury, arsenic V, cysts, VOCs, MTBE, herbicides, pesticides, DBPs.
• NSF/ANSI 58 covers reverse osmosis system performance and specific contaminant claims.
• NSF/ANSI 401 covers emerging compounds: pharmaceuticals, herbicides, pesticides not previously regulated.
• NSF P473 covers PFOA and PFOS reduction; the protocol is being incorporated into NSF/ANSI 53.
• EPA UV Disinfection Guidance Manual (2006) defines 40 mJ/cm² at 254 nm as effective for bacteria, viruses, and cysts.
• WQA hardness scale is the industry reference for grain-per-gallon thresholds and softener sizing.

Each subpage in the filter types section links the technology to the standards that apply to it. Each contaminant page in the contaminants section cites the EPA regulation or CDC guidance directly.