Why global comparisons are hard

Multiple countries · tap water · 2018

5.45 particles/L (mean); mostly particles above 100 µm; microscopy with partial FTIR confirmation. Kosuth, Mason & Wattenberg, 159 tap-water samples.

Global brands · bottled water · 2018

325 particles/L (mean); approximately 6.5–5,000 µm. Mason, Welch & Neratko. The workbook treats this as a sensitivity study rather than silently mixing it with smaller-particle methods.

Germany · single-use PET water · 2018

2–44 particles/L; 5–100 µm; micro-Raman spectroscopy. Schymanski et al.

Germany · groundwater-derived water · 2019

0–0.007 particles/L; 20–500 µm; FTIR imaging. Mintenig et al.

Hong Kong · tap water · 2020

0–8.605 particles/L; 50–4,830 µm. Lam et al.

United States · bottled water · 2024

110,000–400,000 particles/L, mean 240,000; approximately 0.1–5 µm targeted by stimulated Raman scattering microscopy. Qian et al.

Lower size limit

A method looking down to 0.1 µm can count particles that a 100 µm method cannot see. More detected particles do not necessarily mean a more contaminated sample.

Identification method

Visual microscopy, FTIR, Raman and stimulated Raman methods do not identify the same size ranges with the same confidence.

Sample and place

Tap water, groundwater, packaged water and individual brands represent different matrices. One city or product set is not a country average.

Units and summary

Means, medians and ranges answer different questions. Particle count also cannot be converted to polymer mass without size, shape and density data.

The calculator uses country only to choose among eligible studies and food-consumption defaults. It prefers an exact-country study, then a study from the same named region, then a global study. The result labels the tier actually used; country is never a numeric multiplier.

Country-level modeled uptake and the personal estimate remain separate when their foods, particle sizes, assumptions, or units differ. This prevents an attractive but invalid global comparison.