Tap and WellTreasure Valley water, from the public record

Methodology

How every number is computed: Idaho DEQ and EPA sources, 36-month windows, flags, well-zone confidence grades, and the filter matching engine.

Sources

Public system inventory, type, population served and violations come from EPA SDWIS through the Envirofacts API. Sample concentrations come from Idaho DEQ Drinking Water Watch, crawled one system at a time at a polite rate and stored as raw HTML with a fetch date. PFAS and lithium results from EPA’s UCMR 5 study are loaded alongside compliance results and labeled as a monitoring study. Hardness, sodium and TDS are not regulated, so they do not appear in DWW; they come from each system’s Consumer Confidence Report, extracted from the PDF and approved by a human reviewer before publication. Groundwater samples for the well model come from the USGS and EPA Water Quality Portal, the Idaho State Department of Agriculture, and IDWR well records.

Every number on a system page links back to the raw file and row it came from. Raw files are never edited.

Public system profiles

For each contaminant we take all results at entry points and in the distribution system over the most recent 36 months and report the latest result and the highest result in that window. Flags use the highest result, which is what Consumer Confidence Reports also show. Lead and copper use the 90th percentile of samples taken at customer taps, compared with the federal action level, because that is how the Lead and Copper Rule works. Those values are from sampled homes in the system, not from your tap. Home plumbing can add lead and copper. First-draw samples run higher than flushed ones. Where a contaminant appears in both DWW and the CCR, DWW wins for regulated contaminants and the CCR wins for hardness.

Flags

  • Above limit: result above the MCL or action level.
  • Elevated: result at half the limit or above, or above a stricter published health guideline where one exists. Any PFAS detection is at least elevated, because the federal MCLs for PFOA and PFOS are 4 parts per trillion.
  • Aesthetic issue: iron above 0.3 mg/L or manganese above 0.05 mg/L (secondary standards). Not a health flag.
  • Detected: found above the laboratory reporting limit but below elevated.
  • Not detected: tested, nothing found above the reporting limit.
  • Not tested: no result in 36 months.

Hardness is never a health flag. It is classified as soft (under 3.5 grains per gallon), moderately hard (3.5 to 7), hard (7 to 10.5), or very hard (over 10.5). We suggest a softener at 7 gpg and above.

A system’s verdict is “above limit” if any contaminant is above its limit or a health-based violation is open; “watch” if anything is elevated or a health-based violation was recorded in the last two years; otherwise “meets”.

The verdict

Every system page opens with one of four verdicts, decided by rule, not by judgment:

  • Meets all federal standards: no contaminant’s highest result in the last 36 months is above its federal limit, and EPA records no open health-based violation.
  • Meets federal limits, worth watching: nothing above a limit, but at least one result is at half its limit or higher, or a health-based violation was recorded in the last two years.
  • Above a federal limit: at least one result in the last 36 months exceeded its MCL or action level, or a health-based violation is open. A single high sample is not always a legal violation, because compliance for many contaminants is judged on a running annual average; the violations section shows what EPA recorded.
  • No recent results: Idaho DEQ shows no chemical or radiological samples in the window. Small systems sample on multi-year schedules.

For arsenic, uranium, radionuclides, fluoride, other inorganic chemicals, and disinfection byproducts EPA judges compliance on a running annual average at each entry point, so we compare the average of the last 36 months of detected results and show the highest single result next to it. Nitrate, PFAS, microbial results, and aesthetic standards are compared on the highest result, because a single high sample matters for those. Raw well samples taken before treatment are shown in the source section but never used for the verdict. Gross alpha is compared after subtracting uranium (0.67 pCi/L per ug/L), the same adjustment Idaho DEQ applies for compliance. Lead and copper compare the 90th percentile of home tap samples to the action level. Sodium is never flagged; the 20 mg/L figure shown is an advisory for restricted-sodium diets.

Well zones

The region is divided into hexagonal cells about 1.2 km across (H3 resolution 7). For each cell and contaminant we compute the sample count, median, 90th percentile, and share of samples over the federal limit, using samples since 1995. Confidence grades: A means 10 or more samples inside the cell; B means 3 to 9 in the cell, or 10 or more including the six neighbouring cells; C means fewer than that but at least one nearby; D means no nearby samples, county median only. A cell is flagged “above limit” only when 25 percent or more of samples exceed the limit with grade A or B confidence; otherwise a detection is “possible”. ZIP well pages are published only when at least two Tier 1 contaminants have grade A or B; other ZIPs point to the county page.

A well zone estimate describes the aquifer near you. It is not a test of your well.

Matching engine

Every flagged contaminant gets a weight: above limit 3.0, elevated 2.0, possible 1.5, aesthetic issue 1.0, detected 0.5. Every published product gets coverage for each contaminant: 1.0 for a third-party certified claim (0.8 when the certification covers arsenic V only, since the species in Treasure Valley groundwater varies), 0.6 for a manufacturer claim stating 90 percent reduction or better, 0.4 for a manufacturer claim without a number, 0 otherwise. Score is the sum of weight times coverage. Penalties: minus 0.5 if the product needs a prefilter and the water has an iron or manganese issue; minus 1.0 for a reverse osmosis unit with no softener pairing guidance when the water is hard. Products are ranked within three categories (drinking water, whole house, softener); ties break on first-year cost, then offer priority. Renters never see point-of-entry equipment. The “why” line lists what each pick is certified for, what it addresses on manufacturer data only, and what it does not address at all.

A golden test set of real systems and well zones with hand-verified expected flags and top picks runs on every change to the engine.

Softener sizing

Daily grains = hardness (gpg) x household size x 75 gallons per person per day, plus 3 gpg per 1 mg/L of clear water iron. Weekly grains = daily x 7. We recommend the smallest of 24,000, 32,000, 48,000 or 64,000 grains where the weekly load is at or below 70 percent of rated capacity, so the unit regenerates about weekly and does not run at the salt-inefficient top of its range.

Certified

The word “certified” appears on this site only when a certification record exists from NSF, WQA, or IAPMO for that product and that specific claim. Manufacturer claims are labeled as such. Uranium, iron and manganese claims are rarely certified; where the only evidence is a data sheet, the page says so.