Chromium-6 in U.S. Tap Water: Study Methodology
City Selection: EWG targeted 35 cities in 23 states and the District of Columbia for tap water testing. We chose large cities as well as cities whose water utilities reported frequent detections of total chromium, based on our review of state records compiled in EWG's national tap water database (EWG 2009) and on annual water quality reports published by water suppliers.
Sample Collection: EWG recruited water collectors via its staff and their contacts. Tap water samples were collected from unfiltered taps in homes or in public buildings such as hospitals, libraries and malls. Utility bills were typically reviewed to verify the water source of each sample.
All volunteers used a standardized sample collection protocol. Samplers ran the cold-water tap for two minutes to clear pipes of standing water and then collected approximately 100 mL of tap water in a 125 mL HDPE container. Samples were packed in coolers with chilled freezer packs and immediately shipped to the laboratory for analysis. With few exceptions, samples arrived within 24 hours of collection.
Hexavalent Chromium Analysis: Hexavalent chromium levels in tap water samples were measured by Exova (Santa Fe Springs, Calif.; www.exova.com), an ISO/IEC 17025-accredited analytical laboratory, using EPA method 218.6. Samples were prepared through adjustment to pH 9.0-9.5 and filtration. Then a 1,200 microliter portion of the sample was introduced into an ion chromatograph. A guard column removed organics from the sample before hexavalent chromium as CrO42- was separated on an anion exchange separator column. Post-column derivatization of the hexavalent chromium with diphenylcarbazide was followed by detection of the colored complex at 540 nm. This method has a detection limit of 0.02 parts per billion.
Exova’s procedures for quality assurance and quality control include use of duplicate and matrix spike analyses (or matrix spike & matrix spike duplicate analyses) for 5 percent of each batch of samples. The Relative Percent Difference (RPD) between duplicates should fall within the control limit of 13 maximum. Spike recovery can range from 74-to-117 percent.
Exova also measured total chromium levels in tap water samples using EPA method 200.8; these results are not reported here because the detection limit was five times higher than that for the hexavalent chromium measurements. As a result, for 11 of 35 samples no total chromium could be detected using this method. Hexavalent chromium was the dominant form of chromium present in 21 of 24 samples (88 percent) for which total chromium could be quantified.
Chromium Mapping: The maps of population-adjusted average total and hexavalent chromium by county were constructed using the EWG tap water database (EWG 2009). Averages were computed by summing the population served times the average chromium level for each water supplier serving the county, then dividing by the total population served by the county’s water suppliers. Average levels account for variations in testing frequency.