U.S. DC Circuit Court of Appeals
             United States Court of Appeals

  Argued October 20, 2003   Decided December 9, 2003
                      No. 02-1326
           Safe Food and Fertilizer  v. Environmental Protection Agency,

Williams, Senior Circuit Judge: Zinc fertilizers can be produced either from virgin materials or recycled byproducts of
certain industrial processes. In the rule under review here, the Environmental Protection Agency resolved that Subtitle
C of the Resource Conservation and Recovery Act ("RCRA"), 42 U.S.C. § 6901, would not apply to the recycled
materials used to make zinc fertilizers, or to the resulting fertilizers themselves, so long as they met certain handling,
storage and reporting conditions and (in the case of the fertilizers themselves) had concentration levels for lead,
arsenic, mercury, cadmium, chromium, and dioxins that fall below specified thresholds. Petitioners claim that both the
materials and the fertilizer are "hazardous wastes" and that therefore the EPA must regulate them under RCRA's
Subti- tle C.

A material is a "hazardous waste" under RCRA if it is a "solid waste" as defined in 42 U.S.C. § 6903(27) and is
"hazardous" as defined in 42 U.S.C. § 6903(5). Both parties agree that the materials are "hazardous" as that
word of
is used under RCRA, although (as we shall see) the EPA does not in fact regard them as posing any material
hazard if they comply with the conditions specified by the rule. The issue is whether the materials in question are
"solid waste." The EPA has concluded that they are not-that so long as they satisfy the stated conditions,
they have
not been "discarded"
as RCRA's definition of solid waste uses the term.

The analysis is somewhat different, however, for chromium. While the original Fertilizer Institute study apparently
included chromium, see letter from Fertilizer Institute to EPA, dated February 26, 2001, at 9 n.9, the summary of the
study's results actually submitted to the EPA did not include a proposed chromium risk threshold comparable to those
for lead, arsenic, mercury, and cadmium, id. at 11, and the EPA did not report any such risk threshold in its final
rulemaking notice, 67 Fed. Reg. at 48,405/1. Though the EPA risk study included an analysis of
chromium, the
results of this study are not easily translatable by lay judges
into a form comparable with the proposed
exclusion ceiling.

Furthermore, the difference between the EPA's proposed chromium threshold and the chromium concentrations
found in the existing products reported by the EPA is particularly striking. The EPA set its
chromium exclusion
level at 21.3 ppm
for fertilizer with 35.5% zinc content, 67 Fed. Reg. at 48,403/3. Of the twenty commercial virgin
fertilizer samples reported by the EPA, six included test results for chromium. Of these six, one had a chromium
concentration of 8 ppm; the other five all had chromium concentrations of less than 1 ppm. Data Summary. The EPA's
chromium exclusion level thus appears to be more than double the highest virgin commercial sample, ten times the
commercial sample mean, and twenty times the commercial sample median. Given that the EPA has not pointed to
anything in the record indicating that these differences in chromium concentrations are trivial from a health and
environment perspective, we cannot affirm on the basis of the identity principle. We therefore remand to the agency
for an explanation as to whether the differences that its rule allows are irrelevant when considered in light of possible
effects on human health or the environment.