Lead 90th Percentile - Two High Results and a Councilmember
A community water system is required to collect 20 lead samples per six-month monitoring period from tier sites. The results come back in mg/L, highest to lowest: 0.024, 0.019, 0.017, 0.014, 0.012, 0.011, 0.010, 0.009, 0.009, 0.008, 0.007, 0.007, 0.006, 0.006, 0.005, 0.004, 0.004, 0.003, 0.002, 0.002. All copper results are well below the copper action level. No site exceeded a lead action level in the previous two monitoring periods. The two highest results came from homes built in 1952 and 1948, both with lead service lines; the lab has already confirmed the results with its quality control samples and nothing about the analyses is in question.
1. Calculate the 90th percentile lead level and determine whether the lead action level has been exceeded.
Read the worked answer
Method: rank the results from highest to lowest and find the 90th percentile - the value that 90% of the samples fall below. With 20 samples, 0.9 × 20 = 18, which is an integer, so the 90th percentile is the average of the 18th and 19th values counted from the lowest - equivalently, the average of the 2nd- and 3rd-highest results.
Counting down from the top:
- 1st highest: 0.024
- 2nd highest: 0.019
- 3rd highest: 0.017
90th percentile = (0.019 + 0.017) ÷ 2 = 0.018 mg/L (18 ppb).
The lead action level is 0.015 mg/L. Because 0.018 > 0.015, the action level has been exceeded for this monitoring period.
What did not decide it: the single 0.024 result. A 90th percentile is a population statistic, not a maximum. A system can have individual results well above 15 ppb and still not exceed the action level, and vice versa if enough mid-range results cluster near the level.
Sensitivity worth noting: had the third-highest result been 0.011 instead of 0.017, the 90th percentile would be (0.019 + 0.011) ÷ 2 = 0.015 mg/L, which is exactly at - not above - the action level and would not be an exceedance. This is a close call, and it needs to be calculated carefully and documented, not eyeballed.
2. The 90th percentile exceedance is confirmed. What does the system owe the state, the customers, and the corrosion control program?
Read the worked answer
State (primacy agency): report the exceedance on schedule and keep the records - all 20 results, site identification and tier classification, the 90th percentile calculation, and the laboratory's QC confirmation. Site selection matters here: if any site was incorrectly classified or not a valid tier site, the calculation has to be redone before anything else proceeds.
Customers:
- Public education on lead delivered to all bill-paying customers within the timeframe the state sets - commonly 60 days from the end of the monitoring period. Use the approved template and health-effects language; it is not optional and not editable in substance.
- Individual results to the people who took part in the monitoring, including the two homes with the highest results, with an explanation of what their number means.
- Language access - the two high homes are pre-1960 construction, and the notice must reach non-English-speaking residents in the affected area.
Corrosion control program:
- Conduct or update the water quality parameter monitoring and confirm optimal corrosion control treatment is actually being achieved in the distribution system, not just at the plant. Check pH and alkalinity stability, and whether the two high sites sit at a dead end or beyond a change in water chemistry.
- If optimal corrosion control is in place and the action level is still exceeded, the state will require re-evaluation and may require lead service line replacement, with the two LSL sites as the first candidates.
- Return to routine monitoring only as the state allows; normally two consecutive six-month periods at or below the action level.
Do not treat this as a one-sample problem or as something to fix by resampling until the numbers look better.
3. A councilmember demands a list of which homes have unsafe water and asks why the system is not issuing an immediate do-not-drink order. How do you answer?
Read the worked answer
Answer directly and in plain terms:
- There is no lead MCL. Lead is regulated by a treatment technique with an action level, not a maximum contaminant level. The action level is a trigger for additional treatment and outreach; it is not a line between safe and unsafe water at a single tap.
- The action level is a system-wide statistic. Our 90th percentile of 18 ppb means the top 10% of sampled homes are at or above that range - it does not mean 90% of homes are safe and 10% are dangerous. Individual tap results vary with plumbing, stagnation, and whether the service line is lead.
- We are not issuing a do-not-drink order because the state framework does not call for one at this 90th percentile level, and because the effective advice is different: flushing before use, cold water only for formula and cooking, cleaning aerators, and using a certified filter if a resident is concerned or pregnant. That advice goes out with the public education, and the two homes with the highest results get a direct conversation and a filter.
- What we are doing instead of a list: accelerating corrosion control evaluation, confirming the service line material inventory so we know where lead lines actually are, and prioritizing replacement at those addresses.
Tone matters: do not minimize. Acknowledge that 18 ppb is above where we want to be, commit to a date for the public education mailing, and offer the councilmember the same plain-language fact sheet customers will receive.