26 documented events, written the way an incident investigation is written: what happened, which barrier failed, how many people were affected, and what would have prevented it. Figures are quoted from official reports and peer-reviewed studies. No victims are named. There are no products and no links to shops on this page.
Most of these are rare. They are here to show where the real risk sits, not to frighten: for a typical US household on municipal water, lead from the home's own plumbing and Legionella in hot-water systems come first; the big outbreaks came from private wells and from utility failures.
What happened: Oocysts passed through filtration at Howard Avenue Water Purification Plant; turbidity rose; 'most likely the result of human error'
Which barrier failed: Coagulation/filtration (turbidity control) at utility
Ill: ~403,000 (estimate) · Deaths: At least 69 (mostly elderly or immunocompromised)
In the source's words: 'About 403,000 of an estimated 1.61 million residents became ill' (estimate from MacKenzie et al., NEJM 1994); deaths 69 per Corso et al. 2003
What would have prevented it: Acting on turbidity spikes (shutdown/optimized coagulation), multi-barrier treatment (later IESWTR/LT2: UV/ozone)
Source: en.wikipedia.org · pubmed.ncbi.nlm.nih.gov · Primary NEJM abstract could not be loaded; figures taken via Wikipedia summary of peer-reviewed sources. Hoxie et al. AJPH 1997 mortality figure not verified.
Walkerton outbreak
Walkerton, Ontario, Canada · Contamination on/after 12 May 2000 · E. coli O157:H7 + Campylobacter jejuni
What happened: Manure from nearby farm entered shallow Well 5; inadequate chlorination; adverse results not disclosed
Which barrier failed: Source protection, chlorination (residual), monitoring and reporting
Ill: More than 2,300 · Deaths: 7
In the source's words: 'more than 2,300 became ill'; seven people died (Walkerton Inquiry, O'Connor 2002, via CBC summary)
What would have prevented it: Inquiry: continuous chlorine residual and turbidity monitors at Well 5 'would have prevented' it; disclosure of 15 May results would have triggered boil advisory 19 May, avoiding 300-400 illnesses; mandatory lab reporting
Source: cbc.ca · Official inquiry findings, secondary summary (CBC). Not US but canonical case.
Genesee County Legionnaires' outbreak (Flint water crisis)
Flint / Genesee County, Michigan · 2014-2015 · Legionella
What happened: After switch to Flint River water (Apr 2014), chlorine residual loss and corrosion in distribution/premise plumbing
Which barrier failed: Disinfectant residual and corrosion control
In the source's words: '92 cases of Legionnaires' disease were confirmed in residents Genesee County from 2014 to 2015' with 12 fatal (Becker's); 87 cases/12 deaths in other reports incl. Zahran et al. PNAS 2018 coverage
What would have prevented it: Maintaining residual chlorine, corrosion control, timely public notification and Legionella testing of pneumonia cases
Source: beckershospitalreview.com · wgbh.org · SOURCES DISAGREE on case count (87 vs 92); experts suspect undercount. Primary MDHHS report not fetched.
Flint, Michigan · Source switch 25 Apr 2014; switched back 16 Oct 2015 · Lead
What happened: Switch to Flint River without corrosion control -> lead release from service lines and plumbing
Which barrier failed: Corrosion control treatment; LCR sampling/regulatory oversight
Ill: Elevated child blood lead proportion rose (see figures) · Deaths: None attributed to lead specifically
In the source's words: CDC MMWR: BLL>=5 ug/dL in children <6: 3.1% before vs 5.0% early switch (aOR 1.46; 95% CI 1.06-2.01), 1.4% after. Hanna-Attisha AJPH 2016: 'increased from 2.4% to 4.9%'
What would have prevented it: Orthophosphate corrosion control from day one; proper LCR sampling; early public warning
Source: cdc.gov · hero.epa.gov · Two official/peer-reviewed analyses with different periods and denominators; both show increase.
Washington DC lead in drinking water
Washington, DC · 2001-2004 · Lead
What happened: Switch from chlorine to chloramine (to cut DBPs) altered corrosion; partial lead service line replacements
Which barrier failed: Corrosion control during disinfectant change; LSL replacement practice; public health reporting
Ill: Not counted as illness; elevated blood lead risk · Deaths: None documented
In the source's words: Fall 2003: 2,287 of 6,118 homes >50 ppb; 157 homes >300 ppb. CDC 2010 notice: children in homes with lead pipes 'more than twice as likely' to have BLL >=10 ug/dL; 2004 MMWR statement called 'misleading'
What would have prevented it: Corrosion study before changing disinfectant; full (not partial) LSL replacement; transparent blood lead reporting
Source: cdc.gov · en.wikipedia.org · CDC 2010 Notice is official; home-sampling counts via Wikipedia citations (not re-verified at primary source).
What happened: Tube wells drilled into naturally arsenic-rich groundwater; water not tested
Which barrier failed: Source selection/testing
Ill: Millions exposed · Deaths: Not quantified (projected cancer mortality)
In the source's words: 'the largest poisoning of a population in history, with millions of people exposed'; '1 in 10 people who drink water containing 500 [ug] of arsenic per litre may ultimately die from cancers caused by arsenic' (abstract prints 'mg' — apparent typo)
What would have prevented it: Testing wells before use; deeper/alternative safe sources
Source: who.int · WHO Bulletin 2000 (Smith et al.). Projections, not counted deaths.
Unregulated water sources, Navajo Nation
Navajo Nation, USA · Sampling published c. 2017 · Arsenic + uranium
What happened: Natural geology and mining legacy; livestock/unregulated wells used for drinking
Which barrier failed: No regulated supply/treatment
Ill: Not quantified · Deaths: Not quantified
In the source's words: 'arsenic and uranium concentrations exceeded national drinking water standards in 15.1 % (arsenic) and 12.8 % (uranium)'
What would have prevented it: Piped regulated water; source testing; treatment
Source: heronetnext.epa.gov · Exposure study, not outbreak; sample size not verified.
Mid-Ohio Valley PFOA contamination / C8 Science Panel
Parkersburg area, West Virginia & Ohio · Exposure decades to 2000s; Panel 2005-2012 · PFOA (C8)
What happened: Emissions/discharges from DuPont Washington Works plant reached public and private water supplies
Which barrier failed: Source control/unregulated chemical; no drinking water standard at the time
What happened: Naegleria fowleri in home plumbing supplied by treated municipal water
Ill: 2 · Deaths: 2
In the source's words: "In 2011, 2 adults died in Louisiana hospitals of infectious meningoencephalitis after brief illnesses." "N. fowleri were identified in water samples from both homes."
What would have prevented it: Distilled/sterile or boiled-and-cooled water (or 1-µm absolute filter) for nasal irrigation
What happened: Presumed N. fowleri in RV tank/campground water system; suboptimal disinfection and high turbidity found
Ill: 1 · Deaths: 1
In the source's words: "previously healthy woman aged 71 years"; "subsequently died 8 days after symptom onset"; "No N. fowleri DNA or viable ameba were detected in any environmental samples" (sampled 23 days later)
What would have prevented it: Distilled, sterile or boiled-and-cooled water in the squirt bottle; never untreated RV/campground water
Fatal Acanthamoeba encephalitis in immunocompromised patient using tap water in an electronic nasal irrigation device and a CPAP humidifier
New Mexico, USA · 2023 · Devices (nasal rinse, CPAP, humidifier, dialysis)
What happened: Acanthamoeba from household tap water via nasal device and/or CPAP
Ill: 1 · Deaths: 1
In the source's words: 66-year-old "died" about 3 weeks after hospitalization; cited series: ten nonkeratitis cases all performed nasal irrigation, "many using tap water"; "a majority die"
What would have prevented it: Distilled/sterile water in both nasal device and CPAP chamber
PAM case possibly linked to sinus rinsing with tap water (Florida DOH report)
Charlotte County, Florida, USA · Feb 2023 · Devices (nasal rinse, CPAP, humidifier, dialysis)
What happened: Naegleria fowleri; FDOH: "possibly as a result of sinus rinse practices utilizing tap water"
Ill: 1 · Deaths: not confirmed in fetched text (press reported a death)
In the source's words: CDC counts cited by CBS: 157 US cases 1962-2022; Florida 37; two prior infections after sinus rinsing with contaminated tap water
What would have prevented it: Boil ≥1 min and cool, or distilled/sterile water for sinus rinses
Legionnaires' disease from a private-use hot tub at a vacation rental (private well supply)
Western New York, USA · Oct 2024 – Apr 2025 (investigation) · Legionella
What happened: Legionella pneumophila sg1 in under-maintained hot tub; tubs run 100–104°F
Ill: 2 (1 ventilated) · Deaths: 0
In the source's words: "L. pneumophila was detected in all three hot tub samples" at "134 to 13,677 MPN/mL"; culture "3,700 CFU/mL" at air-water interface; isolate differed by "two to three single nucleotide polymorphisms"
What would have prevented it: Automatic disinfectant feed, routine professional servicing/testing; drain and dry between guests
Sporadic community-acquired Legionnaires' disease traced to home electric hot-water tanks
Quebec, Canada · 2010-2012 study · Legionella
What happened: Legionella colonising electric water heaters with cold tank bottoms (mean 31.4 °C at drain valve)
Ill: 36 patients studied; 5 with matching home isolates · Deaths: not reported
In the source's words: "Legionella was recovered in 12/36 (33%) homes"; isolates "microbiologically related in 5/36 (14%) patients"
What would have prevented it: Higher setpoint (≥60 °C storage) with anti-scald mixing valve; periodic tank flushing; attention to electric-tank cold zones
US drinking-water-associated outbreaks 2015–2020 (aggregate): Legionella dominates deaths
United States (28 states reporting) · 2015–2020 · Legionella
What happened: Legionella in building/premise plumbing (biofilm), plus enteric pathogens mostly in wells
Ill: 2,140 · Deaths: 88
In the source's words: "214 outbreaks"; "563 hospitalizations (26% of cases), and 88 deaths (4% of cases)"; Legionella = "786 (37%) of all illnesses, 544 (97%) hospitalizations, and 86 (98%) of all deaths"
What would have prevented it: Premise-plumbing temperature control and flushing; well testing/disinfection for enteric outbreaks
What happened: Mineral aerosol from ultrasonic humidifier (dust contained calcium, magnesium, fluoride, nickel)
Ill: 1 · Deaths: 0
In the source's words: Weeks of rapid breathing and low blood oxygen; resolved after IV steroids and removal of humidifier (NPR account of Pediatrics case report)
What would have prevented it: Evaporative humidifier or no humidifier in infant rooms; mineral-free fill water; stop at first sign of white dust