Can a Water System Pass Its Lead Test While One Tap Reads High?

Suppose a household reads 48 ppb of lead at its kitchen tap, and in the same six-month window its utility reports no exceedance of the lead action level. Both can hold at once, and neither has to be a clerical error.

For most systems the compliance decision is not made on any single tap, nor on an average: it is made on one ranked position in a short ordered list of results; results ranked above that position do not enter the comparison. Where a state has approved fewer than five samples in writing, the highest result is the answer.

The Chain

  1. The pool is drawn first. Under 40 CFR 141.86(a) a system evaluates the materials in its distribution system and identifies a pool of targeted sampling sites.
  2. The pool is ranked by plumbing, not by geography.
  3. The sample is defined by standing time and by tap. A first-draw litre after at least six motionless hours.
  4. The laboratory returns one number per site, in milligrams per litre.
  5. Results are sorted ascending and numbered 1 through n.
  6. n is multiplied by 0.9. The product is a rank.
  7. The value at that rank is the 90th percentile for the system.
  8. That single value is compared with the action level.
From a pipe to one compared number 1. Materials evaluation Pool of targeted sampling sites 2. Tier ranking Tier 1 first, filled down tier by tier 3. First-draw sample One litre, motionless at least six hours 4. Laboratory result One value per site, in mg/L 5. Sort ascending Number the results from lowest upward 6. Multiply by 0.9 The product is a rank, not a value 7. Read that rank Its value is the 90th percentile 8. One comparison That value against the action level Steps read from 40 CFR 141.86 and 40 CFR 141.80(c), Subpart I, as codified July 1, 2020.

The arithmetic reads position only, so a result already above the deciding position can rise without limit and the answer does not move.

Which Taps Are Allowed Into the List

In the operative text, a community water system's Tier 1 sites are single family structures that "Contain copper pipes with lead solder installed after 1982 or contain lead pipes; and/or (ii) Are served by a lead service line. When multiple-family residences comprise at least 20 percent of the structures served by a water system, the system may include these types of structures in its sampling pool." A system that cannot fill its quota from Tier 1 completes its pool with tier 2 sampling sites, "consisting of buildings, including multiple-family residences that: (i) Contain copper pipes with lead solder installed after 1982 or contain lead pipes; and/or (ii) Are served by a lead service line."

Two exclusions in the same section narrow what can be measured. First, "Sampling sites may not include faucets that have point-of-use or point-of-entry treatment devices designed to remove inorganic contaminants." The sentence says nothing about devices designed for other contaminants. Second, the collection paragraph fixes both the standing time and the tap: "Each first-draw tap sample for lead and copper shall be one liter in volume and have stood motionless in the plumbing system of each sampling site for at least six hours. First-draw samples from residential housing shall be collected from the cold water kitchen tap or bathroom sink tap." For a nonresidential building it requires an "interior tap from which water is typically drawn for consumption." It also points to a substitution, at 141.86(b)(5): "A non-transient non-community water system, or a community water system that meets the criteria of § 141.85(b)(7), that does not have enough taps that can supply first-draw samples, as defined in § 141.2, may apply to the State in writing to substitute non-first-draw samples. Such systems must collect as many first-draw samples from appropriate taps as possible and identify sampling times and locations that would likely result in the longest standing time for the remaining sites. The State has the discretion to waive the requirement for prior State approval of non-first-draw sample sites selected by the system, either through State regulation or written notification to the system."

40 CFR 141.86(f) puts the decision with the state — "The State may invalidate a lead or copper tap water sample at least if one of the following conditions is met." — and leaves the system a duty, not a veto: "The system must report the results of all samples to the State and all supporting documentation for samples the system believes should be invalidated."

The Rank, and Where It Stops

In the rule's words, "The number of samples taken during the monitoring period shall be multiplied by 0.9."

Consider a system serving between 101 and 500 people, which collects samples from 10 sites. Ten multiplied by 0.9 is 9, so the ninth-lowest result decides. Suppose the ordered list ends 9, 12, 48 ppb. The 48 ppb at rank 10 sits above the deciding 12 ppb, and the system's 90th percentile stays at 12 ppb, below the 15 ppb action level.

Ten sites: the ninth result is the answer Rank 9 decides because 10 multiplied by 0.9 is 9. Rank 10 sits above it and is not read. Action level 15 ppb 12 ppb 48 ppb 1 2 3 4 5 6 7 8 9 10 Rank of the tap result, lowest to highest Illustrative values, not measurements. Method: 40 CFR 141.80(c)(3), as codified July 1, 2020.

For five samples, the rule reads: "For water systems serving fewer than 100 people that collect 5 samples per monitoring period, the 90th percentile is computed by taking the average of the highest and second highest concentrations." Nothing is set aside, and the single worst tap carries half the weight of the deciding number.

Below five samples the ranking stops. The site-table paragraph as codified July 1, 2020 tells a system with fewer than five qualifying taps to sample each tap it has and then resample them on different days to reach the required number of sites, and then adds a route: "Alternatively the State may allow these public water systems to collect a number of samples less than the number of sites specified in paragraph (c) of this section, provided that 100 percent of all taps that can be used for human consumption are sampled. The State must approve this reduction of the minimum number of samples in writing based on a request from the system or onsite verification by the State." Where a state has approved it in writing, 141.80(c)(3)(v) of the same codification supplies the result: "For a public water system that has been allowed by the State to collect fewer than five samples in accordance with § 141.86(c), the sample result with the highest concentration is considered the 90th percentile value." Nothing sits above it: one tap reading high is then the 90th percentile.

How Many Taps Decide a System

List length is fixed by population served. The standard-monitoring site table at 40 CFR 141.86(c) steps down from 100 sites to 5 as population served falls, and the same six figures stand in the 2020 codification. A reduced column sits beside them there, and it halves most of those counts. Entry differs by size: the two-consecutive-periods route in 141.86(d)(4)(i), as codified July 1, 2020, is written for "A small or medium-size water system", defined in the rule as one serving 50,000 persons or fewer. A larger system falls outside that sentence and reaches reduced monitoring through 141.86(d)(4)(ii) of the same codification.

A query of EPA's SDWIS Federal Reporting Services water system summary, filtered to active community water systems in the 2026 quarter 2 submission and retrieved on 2 September 2026, returns 49,378 of 142,891 active public water systems. Sorted into the rows of the site table:

  • 100 or fewer people served: 11,392 systems, 5 sites each. The 2020 averaging sentence quoted above reaches only systems serving fewer than 100, so nothing is set aside there; the table row is drawn at 100 or fewer, and a system serving exactly 100 sits inside the row and outside that wording.
  • 101 to 500: 15,074 systems, 10 sites, deciding rank 9, one result above it.
  • 501 to 3,300: 13,293 systems, 20 sites, deciding rank 18, two results above it.
  • 3,301 to 10,000: 5,012 systems, 40 sites, deciding rank 36, four results above it.
  • 10,001 to 100,000: 4,122 systems, 60 sites, deciding rank 54, six results above it.
  • More than 100,000: 485 systems, 100 sites, deciding rank 90, ten results above it.

Two consequences follow. First, 26,466 of the 49,378 community water systems, 53.6 percent, decide lead compliance on ten taps or fewer, and 11,392 of them, 23.1 percent, on five. Second, the systems with the most set-aside room are the rarest: the 485 serving more than 100,000 people are 0.98 percent of the count.

Multiplying each row by its site count sizes one national standard monitoring period: 56,960 samples from the smallest band, 150,740 from the next, 265,860, 200,480, 247,320, and 48,500, for a total of 969,860 tap samples. Multiplying instead by the number of positions above the deciding rank gives 0, 15,074, 26,586, 20,048, 24,732 and 4,850, for a total of 91,290 results. That is 9.4 percent of the samples collected.

These two totals are computed here. EPA publishes the system counts in SDWIS and the site table in the Code of Federal Regulations; it does not publish their product, and the 969,860 and 91,290 figures are not agency statistics. Two assumptions sit inside them. Every system is treated as running one standard monitoring period at the minimum site count, which no calendar period looks like and which a state may lower in writing for a system with fewer than five qualifying taps. And the whole 100-or-fewer row is treated as averaging rather than ranking. No equivalent total is offered for the reduced column: 5 multiplied by 0.9 is 4.5, and the 2020 paragraph that resolves a five-site list is written for systems serving fewer than 100 people.

Where the 969,860 samples would land Tap samples in one standard monitoring period, counted by system size 56,960 none above 100 or fewer 150,740 15,074 above 101 to 500 265,860 26,586 above 501 to 3,300 200,480 20,048 above 3,301 to 10,000 247,320 24,732 above 10,001 to 100,000 48,500 4,850 above over 100,000 Samples collected Ranked above the deciding sample Computed from EPA SDWIS Federal Reporting Services water system summary (2026 Q2, active) and 40 CFR 141.86(c).

Two Action Levels, One Operative in September 2026

The current text of 40 CFR 141.80(c)(1) reads: "The lead action level is exceeded if the 90th percentile concentration of lead as specified in paragraph (c)(3) of this section is greater than 0.010 mg/L." That is lower than the 0.015 mg/L used above, and it is what eCFR displays today.

The dates sit two paragraphs earlier, and they separate two things. Paragraph (a)(2) reads: "The requirements of this subpart are effective as of December 30, 2024." Effectiveness is one date; compliance is another. Paragraph (a)(3) sets the second date, and names five sections that can specify otherwise: systems "must comply with the requirements of this subpart no later than November 1, 2027, except where otherwise specified in §§ 141.81, 141.84, 141.85, 141.86, and 141.90, or where an exemption in accordance with 40 CFR part 142, subpart C or F, has been issued by the Administrator." Section 141.86, which carries the site table above, is one of them, and its current text reads: "Standard monitoring consists of six-month tap monitoring periods that begin on January 1 and July 1." November 1 is not one of those start dates. Paragraph (a)(4)(i) names what applies in the meantime: "Between October 30, 2024, and November 1, 2027, community water systems and non-transient non-community water systems must comply with 40 CFR 141.2, 141.31(d), and 141.80 through 141.91, as codified on July 1, 2020, except systems must also comply with 40 CFR 141.84(a)(1) through (10) (excluding paragraphs (a)(6) and (7)), 141.85(a)(1)(ii) and (e), 141.90(e)(1) and (13), (f)(4), and (h)(3), 141.201(a)(3)(vi) and (c)(3), and 141.202(a)(10); 40 CFR part 141, appendix A to subpart Q, entry I.C.1 (excluding § 141.90, except paragraphs (e)(1) and (13) and (f)(4)) and entry I.C.2; 40 CFR part 141, appendix B to subpart Q, entry D.23; and 40 CFR 141.31(d)(2), as codified on July 1, 2024." Neither the action level in 141.80(c) nor tap sampling is among the provisions listed there.

So the operative action level in September 2026 is the one in the July 1, 2020 codification, 0.015 mg/L, and EPA's own explanation page still states the threshold as "0.015 mg/L (15 ppb)".

Under the current 141.80(c)(3), systems with enough Tier 1 and Tier 2 sites use a different input: "For lead, water systems must include the higher of the first-liter and fifth-liter lead sample results at each Tier 1 and 2 site (or first-liter lead sample if tiering is based on premise plumbing) ... Lead or copper sample results from Tier 3, 4, or 5 sites cannot be included in this calculation." Two other branches cover a system with only Tier 3 to 5 sites and one with too few Tier 1 and 2 sites.

The multiplication by 0.9 reads the same in both codifications and in all three. The averaging sentence does not. The 2020 text limits it to systems "serving fewer than 100 people that collect 5 samples per monitoring period", while the current text drops it in all three, and (c)(3)(ii)(E) reads: "For water systems that collect samples from five sites per tap sampling period, the 90th percentile concentration is the average of the highest and second highest concentration..." The gap at exactly 100 people closes on the date systems must comply with that sentence. A fewer-than-five sentence sits in each of the three branches, at (i)(E), (ii)(F) and (iii)(F).

What the Household Is Told Regardless

Under 40 CFR 141.85(d) every system must deliver the individual tap result to the people served at the address sampled: "A water system must provide the consumer notice as soon as practical, but no later than 30 days after the system learns of the tap monitoring results." The notice carries that tap's result, the health effects of lead, and both the action level and the maximum contaminant level goal.

The goal in 40 CFR 141.51(b) is zero. The action level is something else, and EPA says so in as many words: "The lead action level is a measure of the effectiveness of the corrosion control treatment in water systems. The action level is not a standard for establishing a safe level of lead in a home."

Where This Doesn't Apply

The framing above describes a compliance instrument, not a health measurement.

  • It says nothing about any address that was not sampled. The pool is drawn from tiered sites with known lead plumbing, and the system-level 90th percentile is not an estimate for a tap outside it.
  • The set-aside results are set aside only for this calculation. They are reported to the state under 141.86(f)(2) and reach the household under 141.85(d).
  • Two systems with the same 90th percentile are not comparable. One may have drawn 100 samples and one 5, or fewer, and for a system serving fewer than 100 people the five-sample rule averages the two highest concentrations.
  • Reduced monitoring changes the list without changing the rule. The reduced column halves most rows, but the 100-or-fewer row reads 5 in both, so for those 11,392 systems the list does not shorten.
  • Not every system is a community water system. The rule also covers non-transient non-community systems, such as schools with their own supply; the counts here are community systems only.

What Would Change These Numbers

Four inputs govern everything above. The site counts sit in the standard-monitoring table at 141.86(c), identical in both codifications. The system counts come from a live SDWIS query and move every quarter. The deciding rank follows from multiplying by 0.9, worded identically in both, and from the averaging sentence for five-sample systems, which is not: the 2020 version carries a population condition the current version drops. The action level is 0.015 mg/L until November 1, 2027, and 0.010 mg/L in the text systems must comply with from that date.

Only the multiplication by 0.9 is fixed. The rest, the averaging sentence included, are wordings and values with dates attached.

Comments

Post a Comment

Popular posts from this blog

The Repair Takes Hours: Clearing a Boil Water Notice Runs on Laboratory Time

A boil water notice usually arrives with a cause attached: a broken main under an intersection, a pump that lost power, a pressure zone that fell below the level the operator treats as safe. The cause can be repaired within a day. The notice can remain in force well after the repair is finished. That gap is not slow paperwork. Issuing the notice and withdrawing it run on two different mechanisms: one has a federal deadline written into it, and the other is a sequence of physical and biological steps, several of which cannot run in parallel and one of which is a waiting period inside a warm cabinet. The distance between them is what a household experiences as an advisory that will not lift. Two clocks start at the same moment One is fixed by federal rule. The other is a sequence of physical and biological steps. NOTICE CLOCK — bounded at 24 hours CLEARANCE CLOCK — set by steps, not a deadline Hour 0 The system learns of the situation No later than 24 hours Tier 1 public no...
Read more

Twenty-Three Days Before an Outbreak Is Visible, and Longer Before a Recall Notice

A recall notice arrives as a single event: a brand, a lot code, a date range, a list of states. Inside the machinery that produces it, the notice is one of the last steps. Before a product can be named in public, several separate systems have to finish running, in order. A person's immune system, a clinical laboratory, a courier, a state public health laboratory, a national genomic database, a round of interviews, and finally a paper trail held by firms with no direct relationship to one another. The Centers for Disease Control and Prevention publishes an illustration of the first half of that sequence. In it, a person eats a contaminated food on Day 1, and the outbreak is identified on Day 23. That interval is not a measure of effort. It is a sum of waiting periods belonging to different institutions, none of them working to a shared deadline. What the First Nine Days Consume In CDC's published timeline, nothing in the first nine days is decided by a food company or by...
Read more

A Landslide Dam Forms in Seconds and Fails on Its Own Schedule

A Flood With No Storm Attached Strong ground shaking is over in well under a minute. The flood it sets up can arrive a week later, or two months later, in clear weather, with nothing in the rainfall record to account for it. The object standing between the two events is a pile of rock and soil that fell into a river channel during the shaking and stayed there. Geologists call it a landslide dam , or a barrier lake when the emphasis falls on the water rather than the debris. The U.S. Geological Survey studied the general case in The Formation and Failure of Natural Dams , Open-File Report 87-392, which compiled landslide dams worldwide and tabulated how long they lasted. Its central finding is uncomfortable: these structures do not fail on any schedule that can be pinned down, and among those that do fail, most fail well after the response has wound down. The failure mode is not the landslide. The landslide is the visible event, the one that gets counted. The failure mode is the i...
Read more