Opinion
No. 91-1210.
Argued February 19, 1992.
Decided May 1, 1992.
Stephen M. LeGendre, Annapolis, Md., for petitioner.
Joshua M. Levin, Attorney, Dept. of Justice, with whom Barry M. Hartman, Acting Asst. Atty. Gen., and Raymond Ludwiszewski, Acting Gen. Counsel, Earl Salo, Asst. Gen. Counsel, and George Wyeth, Atty., U.S. Environmental Protection Agency, Washington, D.C., were on the brief, for respondent. Michael A. McCord, Atty., Dept. of Justice, Washington, D.C., also entered an appearance for respondent.
Petition for Review of an Order of the United States Environmental Protection Agency.
Before: WALD, WILLIAMS and D.H. GINSBURG, Circuit Judges.
Opinion for the Court filed by Circuit Judge WALD.
Petitioner Anne Arundel County petitions for review of a rulemaking pursuant to which the Glen Burnie Sanitary Landfill ("Landfill"), which petitioner owns, was placed on the National Priorities List ("NPL") of known releases or threatened releases of hazardous substances. In particular, petitioner claims that it was arbitrary and capricious for respondent Environmental Protection Agency ("EPA") to use only unfiltered groundwater samples to measure purported releases at the Landfill and to calculate the "distance to nearest well" factor on the basis of a private drinking well the existence of which was never subjected to public comment and that has since been abandoned.
Because we believe that the EPA's reliance on data collected only from unfiltered groundwater samples was arbitrary and capricious and because the EPA was obliged to give notice of the well on which it relied, we grant the petition for review and remand the case back to the EPA for further proceedings.
I. BACKGROUND
A. Statutory/Regulatory Framework
In order to implement the Comprehensive Environmental Response, Compensation, and Liability Act, 42 U.S.C. §§ 9601- 9675 (1988) ("CERCLA"), the EPA is responsible for the promulgation and revision of the National Contingency Plan ("NCP"), which sets forth guidelines and procedures to respond to releases and threatened releases of hazardous substances, pollutants, and contaminants. The NCP was required, under CERCLA, to include "criteria for determining priorities among releases or threatened releases throughout the United States for the purpose of taking remedial action and, to the extent practicable taking into account the potential urgency of such action, for the purpose of taking removal action." CERCLA § 105(a)(8)(A) (codified at 42 U.S.C. § 9605(a)(8)(A)). Based on these criteria, the EPA has created a list of national priorities representing the sites subject to CERCLA-funded remedial action.
The primary purpose of the NPL is informative,
identifying for the States and the public those facilities and sites or other releases which appear to warrant remedial actions. Inclusion of a facility or site on the list does not in itself reflect a judgment of the activities of its owner or operator, it does not require those persons to undertake any action, nor does it assign liability to any person. Subsequent government action in the form of remedial actions or enforcement action will be necessary in order to do so, and these actions will be attended by all appropriate procedural safeguards.
S.REP. No. 848, 96th Cong., 2d Sess. 60 (1980), reprinted in 1 A LEGISLATIVE HISTORY OF THE COMPREHENSIVE ENVIRONMENTAL RESPONSE, COMPENSATION, AND LIABILITY ACT OF 1980 (SUPERFUND), PUBLIC LAW 96-510, at 308, 367 (Comm.Print 1983). The original NPL, promulgated in September 1983, contained 406 sites, see 48 Fed.Reg. 40,658, 40,660 (1983); by February 1991, the total number of NPL sites had grown to 1,189, see 56 Fed.Reg. 5598, 5598 (1991).
A hazardous waste site is placed on the NPL only after rulemaking by notice and comment, see Administrative Procedure Act ("APA"), 5 U.S.C. § 553(c) (1988). The EPA makes this determination on the basis of the Hazard Ranking System ("HRS"). The HRS is a set of criteria which measure the risk of harm to the environment from the migration of hazardous substances from the site by way of three routes: groundwater, surface water, and/or air. The HRS score is derived from the score for each route; these intermediate scores are calculated on the basis of factors related to actual or threatened releases of hazardous substances from the site. Once all of the factors are evaluated and combined for each of the three separate routes, a final numerical value for the HRS score is assigned. If this score is 28.5 or greater, the site is placed on the NPL. See 40 C.F.R. pt. 300, app. A (1990).
All references are to the original HRS promulgated in 1982. The substantially revised version of the HRS currently in place did not become effective until March 14, 1991, see 55 Fed.Reg. 51,532 (1990); 40 C.F.R. pt. 300, app. A (1991), and the EPA was entitled to list sites on the NPL pursuant to the old HRS until that date. See Linemaster Switch Corp. v. EPA, 938 F.2d 1299, 1304-05 (D.C. Cir. 1991).
The original HRS scoring system is explained in Tex Tin Corp. v. EPA, 935 F.2d 1321, 1322-23 (D.C. Cir. 1991) (per curiam), and Eagle-Picher Industries v. EPA, 759 F.2d 905, 909-11 (D.C. Cir. 1985) ( "Eagle-Picher I").
B. Factual Background
Until 1970, the Glen Burnie Sanitary Landfill was a privately owned, open dump site. It was uncontrolled, and no records of disposal were maintained. Between 1970 and 1982, petitioner operated a dump site at the Landfill under a Maryland state refuse permit. During this period, at least 100 tons of inorganic solids and salts were deposited there. The Landfill was closed in 1982.
In August 1983, samples taken of the groundwater from 11 monitoring wells placed around the perimeter of the Landfill detected elevated levels of both organic and inorganic contaminants. Specifically, the EPA found trichloroethylene, 1,1-dichloroethylene, boron, vanadium, and arsenic in the groundwater at the monitoring wells at significantly higher levels than those measured in the background wells. The EPA concluded that releases from the Landfill had been observed, and it performed remedial preliminary assessments and remedial site inspections to gather further information on the releases. It is undisputed that the EPA based its conclusion that a release had been observed by only testing groundwater samples that were unfiltered; no filtered samples were taken. Because arsenic was the most hazardous substance observed at the Landfill, the EPA calculated the "toxicity/persistence" factor to be "18," the highest possible under the regulations.
The "toxicity/persistence" score is derived by evaluating the most hazardous substance at the site that could migrate into groundwater. 40 C.F.R. pt. 300, app. A § 3.4 (1990). Of the substances identified at the Landfill, arsenic was the most hazardous. The EPA chose not to analyze the samples for either boron or vanadium, the other inorganics observed at the Landfill. See EPA's Response to Comments at 1-11.
In addition, the EPA believed that, on the basis of information provided by the Maryland Geological Survey, a domestic well ("Well 32") located approximately 4,600 feet from one of the Landfill's monitoring wells was still in use. Because Well 32 was located between 2001 feet and one mile from the Landfill, the EPA assigned the value of "3" to the "distance to nearest well" factor.
The two factors making up the "target" category — the "distance to nearest well" and the size of the "population served" — are combined in a matrix to yield a numerical value. 40 C.F.R. pt. 300, app. A § 3.5 (1990). With a "population served" value of "5" (greater than 10,000 people) and a "distance to nearest well" value of "3," the Landfill's target score was "35."
In October 1987, after internal review within the EPA, the Landfill received a total HRS score of 37.93. Because this was in excess of the 28.5 cut-off point for inclusion on the NPL, the EPA proposed that the site be included on the list. See 53 Fed.Reg. 23,988, 23,996 (1988). During the comment period, petitioner challenged the EPA's use of only unfiltered samples, arguing that because the practice of the agency was to use both filtered and unfiltered samples, it had acted arbitrarily in relying exclusively on unfiltered samples.
Furthermore, petitioner informed the EPA that Well 32 was no longer in use. In verifying petitioner's claim, the EPA learned that another property owner (the son of the owner of Well 32) was served by a different private well. The EPA did not, however, reopen the comment period or otherwise inform petitioner. Because this second well was also located between 2001 feet and one mile of the Landfill, the EPA did not modify the "distance to nearest well" score. Petitioner claims that the EPA's reliance on this second well without notice deprived petitioner of the opportunity to close it down and to comment on that fact prior to the promulgation of the final rule.
According to petitioner, this second well has since been abandoned, see Brief of Anne Arundel County, Maryland (filed Nov. 27, 1991) ("Petitioner's Brief") at 12, and the only remaining wells that tap into the aquifer are located more than one mile from the Landfill.
Petitioner seeks review of the EPA's rulemaking on two grounds: First, by failing to filter the samples taken from the monitoring well, the EPA acted arbitrarily by violating its own procedures; second, by failing to disclose the existence of the second well prior to promulgating the final rule, the EPA violated the APA by depriving petitioner of notice and of the opportunity to comment.
II. DISCUSSION
A. Standard of Review
The applicable standard of review is whether the EPA's actions were "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2)(A) (1988). When this court reviews the EPA's decision to include a particular site on the NPL, it must determine whether the decision is consistent with CERCLA and with the policies underlying the NCP, NPL, and HRS. See City of Stoughton, Wis. v. EPA, 858 F.2d 747, 749 (D.C. Cir. 1988); Eagle-Picher Indus., Inc. v. EPA, 822 F.2d 132, 137 (D.C. Cir. 1987) ( "Eagle-Picher III").
Specifically, this court has recognized that inclusion on the NPL is not in itself a remedial action:
Instead, the NPL is simply a rough list of priorities, assembled quickly and inexpensively to comply with Congress' mandate for the agency to take action straightaway. Utilizing the NPL, EPA will thereafter perform in-depth examinations of each site on the list to determine whether remedial action is necessary. If EPA determines at that later stage that the release of a "pollutant or contaminant" at a particular site does not present "an imminent and substantial danger," then no remedial action will be taken.
Eagle-Picher Indus., Inc. v. EPA, 759 F.2d 922, 932 (D.C. Cir. 1985) ( "Eagle-Picher II") (quoting CERCLA § 104(a)(1) (codified at 42 U.S.C. § 9604(a)(1))).
B. Unfiltered Samples
The EPA concluded that a release of arsenic had occurred at the Landfill, justifying a "toxicity/persistence" score of 18. Petitioner argues that this conclusion was based solely on sampling groundwater that was unfiltered, contrary to EPA policy.
When petitioner itself took filtered samples in 1987, it recorded levels of arsenic at the monitoring wells that were significantly below background levels. See Glen Burnie Sanitary Landfill, Hydrogeologic Evaluation Phase IIB, tbl. 6. Of course, the fact that arsenic was not detected at the site in subsequent testing is irrelevant for purposes of the NPL. As this court has held,
[t]he fact that a later test yielded different results does not require that EPA remove its score of an observed release. "[N]egative results during one or more sampling intervals cannot refute a finding, when based on valid sampling and analyses, that an observed release has occurred."
City of Stoughton, 858 F.2d at 756 (quoting from preamble to the 1984 NPL final rule, 49 Fed.Reg. 37,070, 37,078 (1984)).
Our resolution of this issue is governed by the analysis of the identical question in an opinion issued today by a separate panel of this court. See Kent County, Del. Levy Court v. EPA, 963 F.2d 391, 396-398 (D.C. Cir. 1992). We agree that the EPA's use of unfiltered samples alone is inconsistent with its own policy — as articulated in several internal EPA documents — of preferring both filtered and unfiltered samples. Although the EPA is entitled to deference from this court on technical questions, see Chemical Waste Management, Inc. v. EPA, 869 F.2d 1526, 1539 (D.C. Cir. 1989); MCI Cellular Tel. Co. v. FCC, 738 F.2d 1322, 1333 (D.C. Cir. 1984), it has offered no explanation or justification for its use of unfiltered samples alone in direct conflict with the stated policy of Region III. Of course, as we make clear in Kent County,
Kent County concerns a challenge to the listing of a Delaware landfill on the NPL. The EPA calculated the landfill's toxicity score on the basis of having found arsenic in the groundwater, but, just as in this case, it based its conclusions entirely on unfiltered samples. Both Kent County and Anne Arundel County have challenged the EPA's exclusive use of unfiltered samples, arguing that internal EPA documents from Region III (which includes both Delaware and Maryland) explicitly require that both filtered and unfiltered samples be taken to measure the presence of metals in groundwater. The documents relied upon by the petitioners in both cases are substantially identical. The opinion in Kent County discusses in considerable detail the conclusions to be drawn from these documents, and we see no reason to embark on a duplicative analysis in this case.
Without retracing the analysis in Kent County, we note simply that the principal Region III documents on which both Kent County and Anne Arundel County rely include the following:
(1) A 1987 memorandum in which Region III recognized that the EPA's policy under CERCLA differs slightly from its policy under the Resource Conservation and Recovery Act, 42 U.S.C. §§ 6901- 6992k (1988) ("RCRA"), but that the best policy under either Act is to take both filtered and unfiltered samples. See Memorandum on Filtered vs. Unfiltered Groundwater (Aug. 24, 1987).
(2) A 1990 Region III Quality Assurance Directive which "formally state[s] Region III RCRA and CERCLA policy which requires both filtered and unfiltered groundwater samples for metal analyses." Field Filtration Policy for Monitoring Well Groundwater Samples Requiring Metals Analysis, Bulletin No. OAD009 (Apr. 23, 1990) at 1. According to the Directive, "it is necessary to take both filtered and unfiltered samples to fully characterize the distribution of metals at a given site." Id.
In addition to these two documents, Anne Arundel County also relies on an earlier 1987 Memorandum in which the Chief of the Quality Assurance Section of Region III of the EPA sought to clarify "any misconceptions or interpretations of proper procedure for collection of metals samples" by explicitly providing that two samples shall be collected from groundwater monitoring wells — one unfiltered and one filtered. See Memorandum on Proper Procedure for Collection of Metals Samples (May 28, 1987) at 1.
The EPA argues that documents declaring national EPA policy should trump those announcing the policy of a single region. But the only document in our record purporting to announce a national policy — the so-called Memorandum # 15 — is a RCRA document which, by definition, could not be a statement of national CERCLA policy. Furthermore, Memorandum # 15 recommends, on the one hand, that an unfiltered sample be the "standard technique in determining the concentration of metals in groundwater" while, on the other hand, suggesting that "it may be worthwhile to perform both the dissolved [filtered] and total metals [unfiltered] analyses." Id. at 1-2. We agree with our colleagues that Memorandum # 15 is "somewhat equivocal" as a statement of national policy. Kent County, 963 F.2d at 397. For this reason, we do not address — because we do not have before us — a case where there is a conflict between unambiguous national and regional policies. (An additional document apparently dealing with national policy — Robert W. Puls Michael J. Barcelona, Superfund Ground Water Issues: Ground Water Sampling for Metals Analyses (1989) — was contained in the Kent County record but not in ours. The absence of the document from our record is not significant, first, because it actually supported Kent County's (and, therefore, Anne Arundel County's) argument in that it recommended using both filtered and unfiltered samples; and, second, because it was an advisory document written by scientists representing regional offices to the national office and was not an articulation of national policy by the EPA itself.)
[w]e do not suggest that the EPA must use both tests in all future listing decisions. There may well be instances where the agency could justify using only unfiltered tests; for example, at a particular site where wells have been inspected and properly purged to ensure that a groundwater sample is not tainted with soil particles.
Kent County, 963 F.2d at 397. But the EPA has not attempted to justify its use of only unfiltered samples by reference to the unique soil chemistry of Anne Arundel County or to any factor that would set the Landfill apart from other sites. The EPA acted arbitrarily and capriciously when it used only unfiltered samples to test the groundwater at the Landfill. We shall grant the petition for review with respect to the EPA's calculation of the Landfill's "toxicity/persistence" score.
The Landfill's current HRS score of 37.93 places it in the middle of Group 12, with an NPL rank of 575. Should the EPA choose not to conduct another test using both filtered and unfiltered samples, it has the option of recalculating the "toxicity/persistence" score on the basis of 1,1-dichloroethylene, the compound with the next highest matrix value. According to the undisputed calculations of the parties, this would lead to a revised HRS score of 31.94, placing the Landfill at the top of Group 19 with a ranking of approximately 900. See 40 C.F.R. pt. 300, app. B at 208 (1991) (National Priorities List (by Rank) February 1991); see also Petitioner's Brief at 12; EPA's Brief at 43. (Because we also remand for reconsideration of the "distance to nearest well" issue, see infra Part II.C., we do not reach the question of whether we would have granted a petition for review simply for the purposes of recalculating a site's HRS score when the score as revised would still exceed the 28.5 threshold for inclusion on the NPL.)
C. Nearest Well
During the notice-and-comment period, petitioner informed the EPA that all properties in the vicinity of the Landfill had been connected to the County's central water system and that, in particular, Well 32 was no longer in use. See Letter from petitioner to EPA (dated June 30, 1989) at 7 ("June 30 Letter"). In its response to comments, the EPA made the following statement:
The Agency has also verified the well identified as the nearest well in the HRS documentation record, located opposite Furnace Creek, is no longer in service, [and] the owner now receives water from a public system. However, the Agency has verified that another property owner on the street is still served by a private well. This well is also within one mile of the landfill. Thus, the value of 3 for the distance to the nearest well remains unchanged.
EPA's Response to Comments at 1-21.
According to the telephone record, an EPA representative called the owner of Well 32 to verify petitioner's claim:
The originator called to determine if a homewell was still in use at Bx [sic] 83, Marley Creek Dr. in Glen Burnie, Md. owned by Roland Witt. This well is no longer in use, but the sons [sic] family, living at 1630 Witt Drive in Point Pleasant, is still using a homewell. A water main is in place on the street, but they are not hooked up to it. This homewell is between Furnace Creek [and] Marley Creek, which is within 1 mile of the Anne Arundel Landfill.
EPA Telephone Conversation Record (dated Dec. 19, 1989).
Petitioner learned of the second well, for the first time, after reading EPA's Response to Comments, which was made available to the public only after the final rule was published on February 11, 1991. See 56 Fed.Reg. 5598, 5599 (1991). Obviously, petitioner had no opportunity to comment on the status of this second well, even though its existence was known to the EPA since December 1989. Petitioner claims that the EPA violated the APA in not giving notice of this second well and in failing to reopen the comment period so that petitioner could have the opportunity "to address the newly discovered well." Petitioner's Brief at 14.
1. Notice-and-Comment
The EPA argues that it did not need to notify petitioner of the existence of a second well because the final rule was a "logical outgrowth" of the proposed rule — only the identity of the nearest well changed. See City of Stoughton, 858 F.2d at 753; Natural Resources Defense Council, Inc. v. Thomas, 838 F.2d 1224, 1242 (D.C. Cir.), cert. denied, 488 U.S. 888, 109 S.Ct. 219, 102 L.Ed.2d 210 (1988); Small Refiner Lead Phase-Down Task Force v. EPA, 705 F.2d 506, 546-47 (D.C. Cir. 1983). Relying on Rybachek v. EPA, 904 F.2d 1276, 1286 (9th Cir. 1990), and on BASF Wyandotte Corp. v. Costle, 598 F.2d 637 (1st Cir. 1979), cert. denied, 444 U.S. 1096, 100 S.Ct. 1063, 62 L.Ed.2d 784 (1980), the EPA argues that new data "collected by the agency in a continuing effort to give the regulations a more accurate foundation" need not be subjected to a new comment period. Id. at 644. "If data used and disclosed for the interim regulations presented the issues for comment, then there is no need to go seek new comment even though significant quantitative differences result." Id. at 645; see also Rybachek, 904 F.2d at 1286 ("Nothing prohibits the Agency from adding supporting documentation for a final rule in response to public comments.").
In considering whether a final rule is the logical outgrowth of the proposed rule, "the key focus is on whether the purposes of notice and comment have been adequately served." Fertilizer Inst. v. EPA, 935 F.2d 1303, 1311 (D.C. Cir. 1991); Small Refiner Lead, 705 F.2d at 547. A final rule will be deemed to be the logical outgrowth of the proposed rule if a new round of notice and comment would not provide commenters with "`their first occasion to offer new and different criticisms which the agency might find convincing.'" Fertilizer Inst., 935 F.2d at 1311 (quoting BASF Wyandotte, 598 F.2d at 642). The essential inquiry "focuses on whether interested parties reasonably could have anticipated the final rulemaking from the draft [rule]." Natural Resources Defense Council, Inc. v. EPA, 863 F.2d 1420, 1429 (9th Cir. 1988) (citing Small Refiner Lead, 705 F.2d at 547-49); see also American Medical Ass'n v. United States, 887 F.2d 760, 768 (7th Cir. 1989) (proper inquiry is "whether or not potential commentators would have known that an issue in which they were interested was `on the table' and was to be addressed by a final rule").
The EPA argues that it does not matter which well was ultimately used to support its final calculation of the Landfill's "distance to nearest well" score. According to the EPA, petitioner was on notice that the issue of a drinking well's proximity to the Landfill was "on the table" by virtue of the EPA's prior (albeit erroneous) identification of Well 32. But the EPA misunderstands the special character of an NPL rulemaking: The identity of the nearest well in this case is not like other data commonly used in APA rulemaking to support a new standard or guideline. In contrast to traditional rulemaking under the APA, the procedure for placing a hazardous waste site on the NPL consists of the application of a detailed scoring system to particular facts. The identity of the nearest well is not merely one fact among many invoked to support a final rule; it is, in this case, the only fact supporting a "distance to nearest well" value of "3."
Petitioner had no reason to anticipate, on the basis of the proposed rule alone, that the EPA would continue to look for a well to substitute for Well 32 once petitioner had informed the EPA that Well 32 was no longer in service. Indeed, petitioner represented to the EPA that, to the best of its knowledge, "no active wells appear to exist within 2,000 feet to one mile of the Site." June 30 Letter at 8 n. 24.
In contrast to City of Stoughton, where the EPA was permitted to rely (without giving prior notice) on a consultant's report to support a different theory for the rule, the EPA in this case entirely abandoned one factual predicate (that Well 32 was the nearest well) and substituted for it another (that the second well was the nearest well). See Reply Brief of Anne Arundel County, Maryland (filed Jan. 17, 1992) ("Petitioner's Reply") at 13. Petitioner informed the agency that local ordinances required all owners of real property within the County, "abutting a public right-of-way in which a public water main or wastewater main lies," to connect directly to that public water main. See Anne Arundel County Code, art. 27, § 3-213(d) (1985). The EPA knew that the petitioner would have required the owner of a newly-discovered well to close the well and apply for a permit to connect to the public water system, but the agency intentionally concealed the existence of the well from the petitioner.
The EPA discovered the existence of the second well on December 19, 1989, well over a year before the final rule was promulgated. We are rather dismayed that, in order to ensure a sufficiently high HRS score, the EPA would permit a citizen to continue using a well located within one mile of a landfill which it believed sufficiently dangerous to the public health and environment to justify considering it a "national priority." The EPA neither notified the petitioner nor the well user that the well had been identified for the purposes of calculating the Landfill's HRS score. Although there is no evidence (at least in this record) that the well user was ever in danger, the EPA's behavior was wholly at odds with both the spirit and letter of the APA's notice requirement.
2. Harmless Error
The EPA argues that even if it erred in failing to give notice of the second well, the error was harmless. According to the EPA, if petitioner had closed the second well prior to the promulgation of the final rule, the HRS score would not have been modified because the score does not take into account response actions. See 48 Fed.Reg. 40,658, 40,664 (1983). The EPA believes that, even where response actions occur before the listing process begins, these actions should not be considered when scoring the site of the NPL. Without this rule, serious underlying health risks might go "unremedied." For example, it has reasoned that where a population is provided with alternative drinking water supplies after a site's inclusion on the NPL has been proposed, "rescoring in this situation could penalize residents for securing alternative supplies by lowering the priority of the site or deleting it from the list and thereby precluding completion of proper remedial actions." Id. at 40,665; see also Eagle-Picher III, 822 F.2d at 149 (upholding EPA's decision not to consider effects of remedial measures in HRS scoring).
The one exception to this rule is that the EPA, in calculating a site's HRS score, will consider well-closings which are unrelated to the site in question such that the closings are not, in fact, "response actions" at all. Petitioner points to two examples in which the EPA has recalculated the "distance to nearest well" score based on well-closings for reasons unrelated to the site. See EPA's Response to Comments filed by Davidson Lumber Co., Fla. at 4-4 (strong taste and odor of gasoline prompted closing of well, and because no evidence could link contamination of well to site, HRS score was modified because closure was not response action); EPA's Response to Comments filed by Comet Oil Co., Mont. at 5-4 (wells were closed for high nitrate concentrations not attributable to site; "[s]ince these wells were closed for reasons unrelated to the contamination being evaluated for the HRS, the wells should not be included in determining the population served").
Petitioner claims that it closed the second well in May 1991 because the local ordinance requires all owners whose property abuts a public water main to connect to that system within six months of receiving notice. See Public Water and/or Sewer Connection Permit issued to Roland Witt (Apr. 30, 1991). Because the EPA erroneously concluded that it had no obligation to give notice of the second well, it has not explicitly addressed the question of whether petitioner's closing of that well would have been considered "remedial action." This determination is for the EPA in the first instance, and we shall remand for consideration of whether petitioner's closing of the second well would have been "remedial action." If it would have been, then the EPA's error in failing to give notice would have been harmless, and the "distance to nearest well" score would not need to be recalculated.
III. CONCLUSION
For the reasons stated above, the petition for review is granted. We vacate the EPA's order placing the Landfill on the NPL and remand the case to the EPA for further proceedings consistent with this opinion.
It is so ordered.