Tex Tin Corporation (Tex Tin or petitioner) petitions for review of the Environmental Protection Agency's (EPA) listing of Tex Tin on the National Priorities List (NPL), pursuant to Sec. 105 of the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) and pertinent regulations. See 42 U.S.C. Sec. 9605(a)(8)(B); 40 C.F.R. Pt. 300, App. A. A hazardous waste site is placed on the NPL for potential corrective action under CERCLA by informal notice and comment rulemaking on the basis of the site's numerical "score" under the Hazard Ranking System (HRS). The HRS is a set of criteria which measures "the potential for harm to humans or the environment from migration of a hazardous substance away from the facility by routes involving ground water, surface water, or air. It is a composite of separate scores for each of the three routes." 40 C.F.R. Pt. 300, App. A Sec. 1.0.1 The final, composite score is derived from the score for each route or pathway which is in turn derived from a series of subscores, based on factors related to actual or threatened releases of hazardous substances from the site. Any site scoring 28.50 or above is placed on the NPL. The composite HRS score thus represents an estimate of "the probability and magnitude of harm to the human population or sensitive environment from exposure to hazardous substances." 47 Fed.Reg. 31,187 (1982). Having received an overall score of 38.43, Tex Tin was placed on the revised NPL on August 30, 1990. See 55 Fed.Reg. 35,502, 35,508.

Tex Tin's challenge to its NPL listing focuses on the criteria EPA used for the air route score. Tex Tin contends that EPA's errors in computing the air route score led to an overall score above the cutoff for listing on the NPL. The air route score is composed of several components, the scores for two of which petitioner challenges--the observed release score and the toxicity score. See 40 C.F.R. Pt. 300, App. A Secs. 5.1-5.2. The latter is a subpart of the waste characteristics score. Id. at Sec. 5.2.

We agree with the EPA that petitioner has waived its challenge to the validity of the observed release of copper, which is the basis of the observed release score. Absent special circumstances, a party must initially present its comments to the agency during the rulemaking in order for the court to consider the issue. See Eagle-Picher Indus. v. EPA, 822 F.2d 132, 146 (D.C.Cir.1987). Tex Tin failed to object to the wind direction data on the date in question when it made its comments to the agency, although it raised this very objection regarding wind direction on other test dates, which EPA then chose not to rely upon. An objection must be made with sufficient specificity reasonably to alert the agency. See Northside Sanitary Landfill, Inc. v. Thomas, 849 F.2d 1516, 1519 (D.C.Cir.1988), cert. denied, 489 U.S. 1078, 109 S. Ct. 1528, 103 L. Ed. 2d 833 (1989). Raising an objection as to a different test date is insufficient to alert the agency; it is more likely to exclude the date not objected to from the agency's focus. Tex Tin's explanation of its failure timely to object to the copper data rings hollow. Since it did make comments critical of samples collected on this date when copper was the exclusive toxic material found, petitioner will not be heard to suggest that it did not object because it did not consider copper to be at issue in the rulemaking.

Tex Tin's challenge to the toxicity scoring raises more troubling questions about the way in which the agency proceeded. Petitioner's objection to EPA's use of arsenic as the basis for the toxicity score appears to be well-taken. Petitioner contests EPA's conclusion that arsenic, present in the tin slag at the site, poses a real threat to migrate into the air. Petitioner contends that EPA has no basis for believing that the arsenic will be released into the air since there has been no reliable observation of arsenic release into the air nor sufficient explanation of why the agency believes that there is a potential for release. Petitioner vigorously contests that the arsenic in the slag will escape into the atmosphere, stating that it is bound in the rock and can only be released at high temperatures, and that such a release would only occur in a ratio consistent with the other metals with which it is bound in the rock matrix, a pattern of release which EPA has never documented.

EPA does not respond directly to this "metallurgical impossibility" argument, merely stating that a threat of release is enough. That is true if the threat meets the standard of "reasonably expected" which EPA has itself set. As petitioner argues, the EPA's own regulation requires the agency to base its toxicity score on "the most toxic of the substances that can reasonably be expected to be transported away from the facility via the air route." 40 C.F.R. Pt. 300, App. A Sec. 5.2 (emphasis added). But the agency offers no explanation of the basis for its judgment that the arsenic is reasonably likely to be transported via the air route, merely directing the court's attention to its conclusory statements in the agency's response to comments. There the EPA concedes that the agency does not rely on any actual detection of a release of arsenic into the air, but on its presence in the slag, and on the following syllogism: "Since the arsenic is available to the air pathway (the slag is not contained), arsenic thus represents a potential threat and EPA has considered the possibility of an arsenic release when it concluded under section 105(g), that the site should be added to the final NPL." J.A. at 282. Counsel points to no other evidence in the record.

Our diligent search of the record reveals only a bare reference, without explanation, to the lack of containment. In the discussion of the hazardous waste quantity (another of the subcategories on which the air score is based) there is a notation that there are three slag piles (composition unspecified) and that "[t]hese piles have no cover or containment/diversion system." J.A. at 310. This may provide enough factual evidence for the agency's conclusion, but without enlightenment from the EPA, we cannot say whether the agency was reasonable in relying on this slim datum. The court is not expert in the scientific theory by which the potential for migration is measured and cannot defer to agency expertise that was never explained. See Sierra Club v. Costle, 657 F.2d 298, 410 (D.C.Cir.1981). We will affirm a decision of the agency "of less than ideal clarity if the agency's path may reasonably be discerned." Bowman Transp. v. Arkansas-Best Freight Sys., 419 U.S. 281, 286, 95 S. Ct. 438, 442, 42 L. Ed. 2d 447 (1974). But the record is silent on why the lack of containment leads EPA reasonably to expect that the arsenic will be transported via the air pathway. The agency has "cross[ed] the line from the tolerably terse to the intolerably mute." Greater Boston Television Corp. v. FCC, 444 F.2d 841, 852 (D.C.Cir.1970), cert. denied, 403 U.S. 923, 91 S. Ct. 2233, 29 L. Ed. 2d 701 (1971).

At oral argument, EPA counsel invited us to "use our common sense" to conclude that slag containing arsenic (at "concentrations of 1,000 to 2,000 mb/kg"), see J.A. at 282, can be "reasonably expected" to migrate. Common sense cannot answer that question; petitioner is entitled to a decision based on the expertise of the agency. Where the agency has failed to exercise its expertise or to explain the path that it has taken, we have no choice but to remand for a reasoned explanation for the conclusion that the arsenic is reasonably likely to be transported via the air route.

It is so ordered.