United States of America, United States Army Corps of Engineers, Movant-Appellant v. City of Detroit , 329 F.3d 515 ( 2003 )

KEITH, J., delivered the opinion of the court, in which BOYCE F. MARTIN, Jr., C.J., DAUGHTREY, COLE, CLAY, and GILMAN, JJ., joined. MOORE, J. (pp. 527-530), joined in J. KEITH’s opinion as to Parts I, II, and III.A., and also delivered a separate opinion concurring in part and concurring in the judgment. ALAN E. NORRIS, J. (pp. 531-534), delivered a separate opinion concurring in part and dissenting in part, in which BOGGS, BATCHELDER, GIBBONS, and ROGERS, JJ., joined.

OPINION

KEITH, Circuit Judge.

In this case, the district court issued an order under the All Writs Act, 28 U.S.C. § 1651(a), directing the United States Army Corps of Engineers (“Corps”) to accept dredged waste material in order to prevent the frustration of a consent judgment designed to address water pollution problems in the greater Detroit area. For reasons set forth below, we hold that the All Writs Act provides district courts with the authority to bind nonparties in order to prevent the frustration of consent decrees that determine parties’ obligations under the law. However, we find that remand is necessary in this ease because of the district court’s failure to consider the following two issues: (1) whether Detroit could have brought suit under the Administrative Procedure Act (“APA”), and if so, whether this case presents the type of “exceptional circumstances” that would render the APA inadequate; and (2) whether the Corps’ determination that another Environmental Assessment (“EA”) was necessary was “arbitrary and capricious” under Marsh v. Oregon Natural Resources Council, 490 U.S. 360, 109 S.Ct. 1851, 104 L.Ed.2d 377 (1989). Accordingly, we REMAND this case to the district court for further consideration consistent with this opinion.

I. BACKGROUND AND FACTUAL SUMMARY

In 1977, the United States brought suit against the City of Detroit (“Detroit”), the Detroit Water and Sewage Department (“DWSD”), and the State of Michigan (“the State”), alleging that the Detroit wastewater treatment system was operating in violation of the Clean Water Act, 33 U.S.C. §§ 1251-1387, and its National Pollutant Discharge Elimination System (“NPDES”) permit. That year, the parties signed a consent judgment setting a schedule to bring the treatment plant into compliance. Detroit’s failure to comply led to a party-negotiated amended consent judgment in 1981 that contained a revised compliance schedule.

The State revised the NPDES permit in July of 1997, and Detroit fell out of compliance. In 1998, the State issued a notice of violation to Detroit, and the parties thereafter entered into negotiations for a proposed administrative consent order. In August of 2000, Detroit and the State negotiated a second amended consent judgment, which was approved by the district court, to bring Detroit into compliance.¹ The second amended consent judgment required Detroit to dredge and dispose of 146,000 cubic yards of sediment from Conner Creek, a channel connected to the Detroit River. Discharges from Detroit’s sewage treatment plant had contaminated the creek. Under the agreement, this dredging was to be completed “as soon as possible,” and definitely before the completion of a combined sewer overflow basin that Detroit was building in the vicinity. The sediment was to be disposed of “in accordance with state and federal requirements.” The United States Environmental Protection Agency (“EPA”) refused to participate in the negotiation of the agreement, explaining that it had not been part of the administrative proceedings.

Detroit planned to dredge Conner Creek before the project was added as a requirement in the consent judgment. In 1998, Detroit asked the Corps if it could dispose of the sediment at the Confined Disposal Facility (“CDF”) at Pointe Mouillee, which is a wetlands area on the western shore of Lake Erie. Pointe Mouillee, operated by the Corps on bottomland owned by Michigan, includes a state game area and a 3.5-mile dike built to contain dredged material from the Detroit and Rouge Rivers. The CDF, which has a capacity of 18,600,000 cubic yards, was constructed in 1981 under the authority of a statute on soil disposal facilities, 33 U.S.C. § 1293a.

The Corps refused to accept the Conner Creek sediment, citing the elevated concentrations of lead and cadmium in the material. Detroit next explored the idea of dewatering the dredged sediment at the edge of the creek and then transporting the sediment to a landfill. Vigorous community opposition to the prospect of a malodorous dewatering along the creek led Detroit to table this plan. Detroit then returned to the Corps and suggested putting the sediment in a containment cell at the Pointe Mouillee facility; the cell would be covered with clean material to prevent contamination of the environment. The Corps expressed concern over the level of contaminants but agreed to work with Detroit and the State to find a solution. Negotiations ensued. The Corps requested and received the State’s approval for the use of Pointe Mouillee for the sediment. The Corps then required the State to obtain the approval of the EPA and the United States Fish and Wildlife Service and to agree to hold the federal government harmless from liability arising out of the Conner Creek disposal. The Corps also insisted upon an environmental assessment to determine whether the disposal would comply with the National Environmental Policy Act of 1969 (“NEPA”), 42 U.S.C. § 4382 et. seq., requirements. The Corps also directed Detroit to obtain a dredging permit, a permit already required by the Rivers and Harbors Appropriation Act of 1899, 38 U.S.C. § 403. The State responded to the Corps’ demands by refusing to obtain the concurrence of either EPA or the Fish and Wildlife Service on the ground that these approvals would take too much time.

In the meantime, Detroit agreed to undertake another project that became linked to the dredging — the construction of a thirty-million gallon settling basin at Conner Creek to contain the combined sewer overflow (“CSO”) from industrial and sanitary sewage and stormwater runoff. The basin was required by Detroit’s NPDES permit for its sewage treatment plan. Under the state-issued permit, Detroit was to begin construction of the basin by January 1, 2001, and to complete it by January 1, 2005. The high cost of the project prompted Detroit to apply to the State Revolving Loan Fund, which required a project description and environmental studies. When Detroit presented its basin project description, the State required that Detroit include in the proposal its plans for the Conner Creek dredging (even though the dredging is not being funded by the state revolving fund, and Detroit maintained that the two projects were separate). As a result, Detroit could not get funding for the basin unless it had secured a place to put its Conner Creek sludge. The city stated that, if it completed the basin project without state funding, there would be an additional cost to ratepayers of $40,000,000.

On October 12, 2000, the State and Detroit filed a motion seeking an order to show cause as to why the Corps should not be ordered to accept the Conner Creek sediment. The district court concluded that the Corps was frustrating the August 2000 consent judgment, and it issued an order that “[t]he ACE [Army Corps of Engineers] accept dredged materials from Conner Creek for disposal at the Pointe Mouillee Confined Disposal Facility.” United States v. Michigan, 122 F.Supp.2d 785, 793 (E.D.Mich.2000). The district court continued by rejecting the Corps’ conditions on the acceptance of the sediment:

I find that, pursuant to the Agreement Between the United States of America and the State of Michigan Acting Through the Michigan State Department of Natural Resources for Local Cooperation at Detroit and Rouge Rivers, Michigan, dated May 10, 1974, the ACE has obtained from the State of Michigan the statutorily required liability protection language to which it is entitled under Section 123 of the River and Harbor Act of 1970, 33 U.S.C. § 1293a, and I ORDER that no further liability protection language from the State of Michigan is required or authorized by law.

Since I find that the proposed disposal at the Pointe Mouillee CDF [Confined Disposal Facility] of dredged materials from Conner Creek does not constitute a new use of the facility, I ORDER that it is not necessary that the ACE conduct a review of such disposal under the National Environmental Policy Act, or have developed a new EIS [Environmental Impact Statement] or EA [Environmental Assessment] by NEPA.

Id. at 793-94 (paragraph numbers omitted). The “new use” issue related to the Corps’ assertion that the CDF had been used previously for “navigational dredging” and the Conner Creek sludge was “environmental dredging,” a new use prompting a need for supplemental environmental studies. The district court denied the Corps’ motion for reconsideration. The Corps timely appealed the district court’s judgment to our Court, asserting that sovereign immunity on the part of the Corps barred the injunctive order issued by the district court and that the All Writs Act did not permit the issuance of an order requiring a third party to take actions necessary to effectuate a settlement to which it is not a party.

On January 11, 2002, our Court issued an unpublished opinion in which a divided three-judge panel vacated the district court’s order. The panel majority found that the waiver of sovereign immunity as codified in the Administrative Procedure Act (“APA”), 5 U.S.C. § 702, applied to this case, but that a consent judgment is not a legal obligation authorizing the district court to issue orders under the All Writs Act, 28 U.S.C. § 1651(a), that impose obligations on nonparties. In sum, the panel majority concluded that the district court lacked the authority to compel the Corps to accept the dredged Conner Creek materials, and that it abused its discretion in issuing an injunction which ordered the Corps to do so.

Upon the issuance of this unpublished judgment by our Court, the City of Detroit, the State of Michigan, and the Counties of Wayne, Oakland, and Macomb all filed petitions for en banc rehearing. The full court voted in favor of rehearing the case en banc. Supplemental briefs having been filed, and the case having been rear-gued, this appeal is now ready for decision by the full Court.

II. STANDARD OF REVIEW

We review a district court’s grant of an injunction for an abuse of discretion. See In re Dublin Sec., Inc., 133 F.3d 377 (6th Cir.1997); Golden v. Kelsey-Hayes Co., 73 F.3d 648, 653 (6th Cir.1996). A court abuses its discretion in granting an injunction if it incorrectly applies the law or relies on clearly erroneous findings of fact. See Golden, 73 F.3d at 653. Thus, in reviewing a lower court’s grant of an injunction, our Court examines the district court’s conclusions of law de novo and its findings of fact for clear error. See id.

III. ANALYSIS

A. Sovereign Immunity Defense

The Corps claims that its sovereign immunity precludes this action. Under the doctrine of sovereign immunity, the United States is immune from suit unless it consents to be sued. See Hercules, Inc. v. United States, 516 U.S. 417, 422, 116 S.Ct. 981, 134 L.Ed.2d 47 (1996). The government has waived its immunity with respect to non-monetary claims. See Administrative Procedure Act (“APA”), 5 U.S.C. § 702. The APA provides that

[a]n action in a court of the United States seeking relief other than money damages and stating a claim that an agency or an officer or employee thereof acted or failed to act in an official capacity or under color of legal authority shall not be dismissed nor relief therein be denied on the ground that it is against the United States ... Provided, That any mandatory or injunctive decree shall specify the Federal officer or officers (by name or by title), and their successors in office, personally responsible for compliance.

5 U.S.C. § 702.

The Corps contends that section 702 does not serve as a general waiver of sovereign immunity for suits against the United States seeking non-monetary relief. Rather, the Corps believes the waiver apples only to complaints filed under the APA. In this case, Detroit and Michigan did not file an APA complaint, and the district court did not rely on the APA in issuing its order.

We reject the Corps’ restrictive reading of section 702. This circuit has applied the waiver of sovereign immunity in section 702 in cases brought under statutes other than the APA. See A.E. Finley & Assocs., Inc. v. United States, 898 F.2d 1165, 1167 (6th Cir.1990) (noting, in a case arising under the Contract Disputes Act, that “the Administrative Procedure Act, specifically 5 U.S.C. § 702, waives sovereign immunity under § 1331”); Newsom v. Vanderbilt Univ., 653 F.2d 1100, 1107 (6th Cir.1981) (applying the APA waiver to a claim brought under the Hill-Burton Act on free hospital care). Furthermore, the other circuits that have addressed the issue agree that “the waiver of sovereign immunity contained in § 702 is not limited to suits brought under the APA.” Specter v. Garrett, 995 F.2d 404, 410 (3d Cir.1993), rev’d on other grounds, 511 U.S. 462, 114 S.Ct. 1719, 128 L.Ed.2d 497 (1994). See also Up State Fed. Credit Union v. Walker, 198 F.3d 372, 375 (2d Cir.1999) (describing section 702 as a “general waiver,” but noting that relief is forbidden if another statute forbids it); Black Hills Inst. of Geological Research v. South Dakota Sch. of Mines & Tech., 12 F.3d 737, 740 (8th Cir.1994) (“Section 702 waives the federal government’s sovereign immunity in cases challenging agency action ... and seeking relief other than money damages.”); Panola Land Buyers Ass’n v. Shuman, 762 F.2d 1550, 1555 (11th Cir.1985) (“The defense of sovereign immunity is waived in actions against federal government agencies seeking non-monetary relief if the agency conduct is itself subject to judicial review.”); see also Tucson Airport Auth. v. Gen. Dynamics Corp., 136 F.3d 641, 645 (9th Cir.1998) (adding an additional requirement that there be “no other adequate remedy in a court” before sovereign immunity is waived).

Because the motion in the instant case did not seek monetary damages, the waiver of sovereign immunity in section 702 applies.

B. Applicability of the APA to the Claims Presented by Detroit

We turn now to the question of whether the APA applies to the facts of this case. Specifically, we must ascertain whether the district court should have properly relied upon the APA, rather than the All Writs Act, as the basis for asserting its jurisdiction in this matter. In a related query, we also find it appropriate to investigate whether Detroit should have utilized the APA, and not the All Writs Act, as the proper avenue to federal court. With regard to this issue, the dissent notes that “[w]here a statute specifically addresses the particular issue at hand, it is that authority, and not the All Writs Act, that is controlling.” Syngenta Crop Prot, Inc. v. Henson, 537 U.S. 28, 123 S.Ct. 366, 369, 154 L.Ed.2d 368 (2002) (alteration in original) (quoting Pennsylvania Bureau of Corr. v. United States Marshals Serv., 474 U.S. 34, 43, 106 S.Ct. 355, 88 L.Ed.2d 189 (1985)). The dissent contends that because Detroit could have brought suit against the Corps pursuant to the APA, it could not instead resort to the All Writs Act as a means to secure a remedy. However, the precise method by which Detroit should have employed the APA in this case is not elucidated by the dissent’s commentary on this subject. Interestingly enough, Detroit appears to have conceded the applicability of the APA by stating that “[s]uch claims fall foursquare within both sections 702 and 706 of the APA.” (Appellee’s Supp. Br. at 6).

We find the dissent’s argument on this issue to be overly broad. Immediately after announcing the rule quoted above, the Supreme Court cautioned that it did not need to “categorically rule out reliance on the All Writs Act” under such circumstances as we find in the instant case. United States Marshals, 474 U.S. at 43, 106 S.Ct. 355. In point of fact, the Supreme Court proceeded to elaborate on the subject by clearly stating that “[t]here may be exceptional circumstances in which a district court can show clearly the inadequacy” of the statute that specifically addresses the particular issue at hand. Id. The Supreme Court in United States Marshals concluded its discussion of this issue by stating that “[wje therefore leave open the question of the availability of the All Writs Act to authorize such an order where exceptional circumstances require it.” Id.

Based on the above, we find that remand on this issue of potential APA applicability is the proper course of action. We consider it necessary, under the circumstances, for the district court to undertake an analysis as to whether Detroit should have brought suit against the Corps pursuant to the APA in order to compel its acceptance of the dredged material. Next, if the district court concludes that Detroit could have used the APA, the district court should then issue a ruling as to whether this case presents the type of “exceptional circumstances” that would render the APA inadequate, and thus allow for the utilization of the All Writs Act as a means to compel acceptance of the waste material by the Corps. As the case now stands, the district court appears to have undertaken neither inquiry. Remand is therefore the appropriate resolution here, given that the question of whether “exceptional circumstances” exist is precisely the type of highly fact-bound determination that the district court is meant to decide.

C. The District Court’s Authority Under the All Writs Act

Another essential issue before us revolves around the district court’s issuance of an injunctive order compelling the Corps to accept the dredged Conner Creek materials for disposal. As the basis for its authority to issue the injunctive order, the district court relied upon 28 U.S.C. § 1651(a), commonly referred to as the All Writs Act. The All Writs Act states that: “[t]he Supreme Court and all courts established by Act of Congress may issue all writs necessary or appropriate in aid of their respective jurisdictions and agreeable to the usages and principles of law.” 28 U.S.C. § 1651(a). Further, the Supreme Court has clearly annunciated that the All Writs Act specifically authorizes a federal court “to issue such commands ... as may be necessary or appropriate to effectuate and prevent the frustration of orders it has previously issued in exercise of jurisdiction otherwise obtained.” United States v. New York Tel. Co., 434 U.S. 159, 98 S.Ct. 364, 54 L.Ed.2d 376 (1977); United States Marshals, 474 U.S. 34, 106 S.Ct. 355, 88 L.Ed.2d 189.

The All Writs Act permits courts to issue orders to nonparties in certain situations. See New York Tel., 434 U.S. at 174, 98 S.Ct. 364 (“The power conferred by the Act extends, under appropriate circumstances, to persons who, though not parties to the original action or engaged in wrongdoing, are in a position to frustrate the implementation of a court order or the proper administration of justice, and encompasses even those who have not taken any affirmative action to hinder justice.”) However, courts may issue orders to non-parties only when acting pursuant to a legal authority. See Syngenta, 537 U.S. 28, 123 S.Ct. 366, 370, 154 L.Ed.2d 368 (finding that the All Writs Act could not justify a removal order when there was no basis for removal jurisdiction). Therefore, while some consent judgments impose obligations pursuant to law and some impose obligations pursuant to contract, see Local No. 98 Int’l Ass’n of Firefighters v. City of Cleveland, 478 U.S. 501, 519, 106 S.Ct. 3063 (1986) (referring to consent judgments as a “hybrid” of judgments and contracts), only consent judgments that impose obligations pursuant to law can serve as the basis for an injunctive order to a nonparty under the All Writs Act.

In this case, the district court issued the injunctive order with the specific intent of preventing the Corps from frustrating the obligations provided for by the Second Amended Consent Judgment, which was entered on August 3, 2000. The question is whether this consent judgment imposed obligations pursuant to law for purposes of the All Writs Act. We conclude that it did.

A consent judgment is

a contract between the parties to the agreement, operates as an adjudication between them and, when the court gives the agreement its sanction, becomes a judgment of the court. The fact that the judgment is by consent gives it neither greater nor less force than if rendered after protracted litigation. It has the same weight and effect as any other judgment and, unless vacated or set aside, stands as a final determination of the rights of the parties.

BlaCk’s Law Dictionary 842 (6th ed.1990) (quoting Traveler’s Ins. Co. v. United States, 283 F.Supp. 14, 28 (S.D.Tex.1968) (internal citations omitted)). The force of a consent judgment is well settled within our judicial system. See, e.g., Rufo v. Inmates of Suffolk County Jail, 502 U.S. 367, 378, 112 S.Ct. 748, 116 L.Ed.2d 867 (1992) (stating that although a consent decree “in some respects is contractual in nature[,] ... it is an agreement that the parties desire and expect will be reflected in, and be enforceable as, a judicial decree that is subject to the rules generally applicable to other judgments and decrees”); Vanguards of Cleveland v. City of Cleveland, 753 F.2d 479, 484 (6th Cir.1985) (“A consent decree, although founded on an agreement of the parties, is a final judgment.”).

Here, all parties and the district court agree that Detroit violated the Clean Water Act. Presumably, the consent judgment was agreed upon, in part, to save the time and expense of litigation required to “prove” this uncontested point. The district court’s sanctioning of the agreement was clearly more than just a “stamp of approval.” Rather, the consent judgment operated as broadly as a ruhng after a full trial. Therefore, the consent judgment was a sufficient basis for invoking the All Writs Act.

The All Writs Act makes no distinction between consent judgments and court orders. Rather, the relevant distinction is between writs issued in aid of the court’s jurisdiction pursuant to a legal obligation and writs issued in aid of the court’s jurisdiction pursuant to private contract. Again, the writ issued in this case falls into the former category. For these reasons, we find that the district court had the authority under the All Writs Act to issue the injunctive order.

The next issue we are presented with pertains to whether the district court abused its discretion in ordering the Corps to accept the dredged sediment irom Conner Creek. As stated previously, we review a district court’s grant of an injunction for an abuse of discretion. See In re Dublin Sec., Inc., 133 F.3d 377; Golden, 73 F.3d at 653.

The Corps contended that the district court erred because: (1) the Corps has not frustrated the consent judgment; (2) the Corps is not part of the underlying controversy; (3) the district court’s order imposes a substantial burden on the Corps and may have significant adverse impacts on the environment; (4) Detroit failed to show the absence of feasible alternatives for the disposal of the sediment; and (5) the district court’s order is not “agreeable to the usages and principles of law.” With regard to the first four matters, we conclude that the district court’s determinations were within the purview of its discretion.

The Corps contended that it was not frustrating the Judgment or any other orders of the district court. It noted that it had not denied Detroit permission to dispose of the sediment at the CDF, but rather was in the process of negotiating with the defendants when Michigan and the DWSD moved for an order compelling the Corps to accept the sediment. As a result, the Corps asserted that their failure to agree was not sufficient justification to trigger implementation of the All Writs Act. Moreover, the Corps noted that the various deadlines in effect could be relaxed by either Michigan or the district court.

The Corps’ contention that Michigan or the district court could extend the various deadlines was unrealistic, and disregarded the district court’s findings that time was of the essence. Indeed, nothing in New York Telephone requires parties seeking a writ to seek the extensions that the Corps suggested. In issuing the writ, the district court’s main concern was that Detroit needed to have a named disposal site for the Conner Creek dredged materials finalized by November 22, 2000, in order to obtain funding from the state. Moreover, the court noted that if the disposal site was not finalized, DWSD ratepayers would incur an estimated $40,000,000 in additional interest charges as a result of missing the deadline. Therefore, the Corps frustrated the consent decree when it failed to accept the dredged materials in a timely manner.

In New York Telephone, the Supreme Court considered four major factors for determining whether a writ issued to a nonparty under the All Writs Act was within the district court’s discretion. There, the district court issued an order authorizing the FBI to use pen registers on two phone lines operated by the New York Telephone Company. See 434 U.S. at 174, 98 S.Ct. 364. After the New York Telephone Company refused to assist the FBI, the district court directed the company to provide all information, assistance, and facilities necessary for the FBI to install and use pen registers pursuant to the All Writs Act. See id. In concluding that the district court had authority to issue an order against the New York Telephone Company, the Court first examined the nonparty’s relationship to the controversy. See id. (concluding that the phone company was not “a third party so far removed from the underlying controversy that its assistance could not be permissibly compelled”). Second, the Court considered the burden that cooperation would impose on the nonparty. See id. Third, the court examined the nonparty’s interest in not providing assistance both from an ideological and financial perspective. See id. at 174-175, 98 S.Ct. 364 (concluding that because the telephone company had used pen registers in the past and because the FBI would compensate the company for its assistance, the company had no interest in not assisting). Lastly, the Court looked at the importance of the non-party’s assistance to fulfilling the goals of the order. See id. at 175, 98 S.Ct. 364.

With respect to the first and third factors outlined in New York Telephone, we conclude that the Corps was not “so far removed from the controversy that its assistance could not be permissibly compelled,” and that the Corps did not have a substantial interest in not accepting the dredged materials from Conner Creek. See 434 U.S. at 174, 98 S.Ct. 364. In New York Telephone, the Court concluded that the New York Telephone Company was not so far removed from the controversy and that the Company did not have a substantial interest in noncompliance with the writ because: (1) the company was in a position to assist; (2) the company had a “substantial interest” in providing assistance; and (3) the company regularly installed pen registers in its own course of dealings. Id. Similarly here, the Corps was in a position to assist, and it had a substantial interest in ensuring that the environmental problems created by the sewage runoff were dealt with in a timely manner. Moreover, the CDF’s sole purpose was to accept dredged materials from local rivers. Accordingly, the Corps was not “so far removed from the underlying controversy that its assistance could not be permissibly compelled,” nor did it possess a substantial interest in noncompliance with the writ. Id.

Although the Corps suggested that under New York Telephone, it must be actively facilitating criminal activity to be subject to an order under the All Writs Act, this assertion lacks merit. In New York Telephone, the Court did not suggest that criminal conduct is required before a nonparty can be subject to an order under the All Writs Act. Instead, the Court required only that the Company not be “so far removed from the underlying controversy that its assistance could not be permissibly compelled.” 434 U.S. at 174, 98 S.Ct. 364. Consequently, the fact that the Corps was not facilitating criminal activity does not preclude the Corps from being subject to a writ under the All Writs Act.

With respect to the second factor, we conclude that compliance with the writ would not impose a substantial burden upon the Corps. The Corps argued that the court’s order imposed a substantial burden, and may have significant adverse impacts on the environment. The Corps also stated that the sediment from Conner Creek had elevated contaminant levels, and that special procedures would be necessary to dispose of it properly. Further, the Corps suggested that the sediment could adversely impact the environment, but this would not be ascertainable until it had reviewed the potential adverse impacts of such disposal.

Although the Corps contended that the district court erred in determining that its order would not significantly burden the Corps, the Corps faded to point to any evidence in the record that supports their argument. The City agreed to pay all costs associated with the disposal of the Conner Creek sediment, and the disposal would not have exposed the Corps to any additional liability for which it would not be indemnified by the State, as per the CDF agreement. Consequently, we find no basis for concluding that the district court erred when it determined that its order would not significantly burden the Corps.

With respect to the fourth factor, we conclude that the Corps’ assistance was necessary given the exigent circumstances and the lack of feasible alternatives. Although the Corps argued that the district court erred by finding that Detroit and Michigan proved the absence of feasible alternatives and that the CDF was feasible, Detroit clearly investigated the option of on-site dewatering of the sediment, but members of the public residing in proximity to the site strenuously opposed that course of conduct. Detroit also submitted evidence demonstrating that it had considered several alternatives. The Corps never suggested any alternatives. Therefore, we find that the district court was also within its discretion to determine that Detroit had exhausted all feasible alternatives.

D. Whether the Writ was Agreeable to the Usages and Principles of Law

Finally, the Corps claimed that the district court’s order was not “agreeable to the usages and principles of law” as required by the All Writs Act. 28 U.S.C. § 1651(a). The Corps contended that the district court failed to confront the inconsistency of its order and the law governing the Corps’ operation of the CDF. The Corps argued that “ordering the Corps to accept the Conner Creek sediment at Pointe Mouillee, without allowing the Corps to first evaluate the impacts of such disposal on the environment and to determine whether such disposal would be in the public interest, directly contravenes the National Environmental Policy Act and the Clean Water Act.” The Corps also asserted that, pursuant to NEPA and agency regulations, it was obligated to conduct additional NEPA review before it decided whether or not to accept the sediment in question.

The district court found that the Corps did not need to conduct additional NEPA review prior to accepting the dredged materials. United States v. Michigan, 122 F.Supp.2d at 792. The district court based this conclusion upon the applicable laws and the CDF Agreement as well as the record evidence submitted by the parties. The district court noted that the Corps had previously agreed that it was not legally precluded from accepting the materials, and that the toxicity of the materials was below hazardous levels. The district court also took into consideration the desire of the State of Michigan to have the sediment housed in the CDF.

NEPA requires an agency to prepare an environmental impact statement (“EIS”), or a supplemental EIS (“SEIS”), only when it is proposing to take “major federal action significantly affecting the quality of the human environment.” 42 U.S.C. § 4332(2)(c). Citing Township of Ridley v. Blanchette, 421 F.Supp. 435, 446 (E.D.Pa. 1976), Detroit argues that the proposed deposit of sediment at the CDF is not a “major federal action” requiring an EIS or SEIS. The deposit of the dredged material, according to Detroit, would simply be an ongoing use of the CDF. See, e.g., Upper Snake River v. Hodel, 921 F.2d 232 (9th Cir.1990) (holding that the continued operation of projects for which an original EIS was conducted did not constitute a “major federal action,” or otherwise require an SEIS).

The Corps, however, has taken the position that some sort of environmental study is required by NEPA. A court reviewing the Corps’ determination must do so under the “arbitrary and capricious” standard. Marsh v. Oregon Natural Res. Council, 490 U.S. 360, 377, 109 S.Ct. 1851, 104 L.Ed.2d 377 (1989). In Marsh, the Supreme Court reviewed a decision by the Corps to proceed with a dam construction project without preparing a second supplemental EIS to consider information that was not available at the time of a previous EIS. 490 U.S. at 368, 109 S.Ct. 1851. The Supreme Court stated that because an agency’s decision whether to perform additional environmental analysis requires “substantial agency expertise,” courts must defer to the agency’s “informed discretion.” Id. at 376-77, 109 S.Ct. 1851.

The district court in the instant case does not seem to have identified or applied the “arbitrary and capricious” standard to the Corps’ decision. For example, although the court found that “the levels of cadmium and lead are not beyond the site’s capability,” United States v. Michigan, 122 F.Supp.2d at 792, it did not analyze whether the Corps could reasonably conclude that those levels amounted to significant new circumstances with respect to its federal action. Accordingly, we remand this case to the district court for an initial determination of whether the Corps’ decision to perform another EA was “arbitrary and capricious.”

IV. CONCLUSION

In conclusion, we hold that the All Writs Act provides district courts with the authority to bind nonparties in order to prevent the frustration of consent decrees that determine parties’ obligations under the law. However, for the reasons discussed above, we REMAND the case to the district court for further consideration of the following two issues: (1) whether Detroit could have brought suit under the APA, and if so, whether this case presents the type of “exceptional circumstances” that would render the APA inadequate; and (2) whether the Corps’ decision to perform another EA was “arbitrary and capricious” under Marsh.

MOORE, Circuit Judge, concurring in part and concurring in the judgment.

I write separately because, although I agree with Judge Keith’s analysis of the All Writs Act and I agree that this case should be remanded, I think that the scope of the remand is broader than necessary.¹ Specifically, because NEPA analysis was plainly required, I do not think it is necessary for the district court to determine whether the Corps’s insistence on a NEPA analysis was arbitrary or capricious. Under the district court’s order, the Army Corps of Engineers was required to accept at its Pointe Mouillee Confined Disposal Facility sediment that is polluted at levels not only higher than the levels in sediments usually accepted at the facility, but at levels up to ten times higher than those contemplated when the facility was first built. No one contends that the law forbids the Corps from accepting contaminated sediments at Pointe Mouillee. However, the law does require that federal agencies consider the potential environmental consequences of their actions. The National Environmental Policy Act (“NEPA”), Pub.L. No. 91-190, 83 Stat. 852 (1970) (codified as amended at 42 U.S.C. § 4321 et seq.), requires that the Corps analyze the environmental consequences of the proposed action before proceeding.

The Conner Creek sediment is not Pointe Mouillee’s everyday sludge. In fact, the Conner Creek sediment contains levels of lead and cadmium much higher than the sediment previously considered or usually accepted. In the Pointe Mouillee CDF’s original 1974 EIS, the Corps considered accepting sediment with lead levels up to 110 parts per million (“ppm”), 1974 EIS at 188, and in actual practice the Corps accepts sediment with lead levels in the range of 233 ppm, J.A. at 285. The Conner Creek sediment contains lead levels of 1,100 ppm, J.A. at 285 — ten times that initially considered and more than four times that normally accepted.

The levels of cadmium in the proposed sediment are similarly unusual. Whereas the 1974 EIS mentions nothing higher than the 7.2 ppm average level of cadmium apparently found throughout the Detroit River system, 1974 EIS at 47, and the sediments usually accepted at Pointe Mouillee generally contain cadmium at a level of 5.3 ppm, J.A. at 285, the Conner Creek sediment has cadmium levels reaching 31 ppm, J.A. at 285 — more than four times the level originally contemplated and more than five times the level usually accepted.

These exceptionally high levels of contaminants trigger the procedural requirements of the National Environmental Policy Act. Under 42 U.S.C. § 4332(2)(C)(ii), federal agencies contemplating “major Federal actions significantly affecting the quality of the human environment” must provide an Environmental Impact Statement (“EIS”) assessing the effects that the proposed action will have on the environment, discussing alternatives, and identifying the best way to undertake the action. Although NEPA’s substantive environmental standards are flexible and leave room for agency discretion, NEPA’s procedural requirements demand strict compliance. Calvert Cliffs’ Coordinating Comm., Inc. v. United States Atomic Energy Comm’n, 449 F.2d 1109, 1112 (D.C.Cir.1971). The EIS requirement is an “action-forcing” requirement designed to “assure consideration of the environmental impact of [agencies’] action” and ensure that the project, when undertaken, is handled responsibly. Kleppe v. Sierra Club, 427 U.S. 390, 409, 96 S.Ct. 2718, 49 L.Ed.2d 576 (1976) (quotations omitted). The question here is whether NEPA prohibits the Corps from accepting the Conner Creek sediment without first assessing the environmental impact of acceptance. NEPA may require either a full-fledged EIS or an Environmental Assessment (“EA”), “a preliminary document which ‘[b]riefly provide[s] sufficient evidence and analysis for determining whether to prepare an environmental impact statement or a finding of no significant impact.’ ” Southwest Williamson County Cmty. Ass’n, Inc. v. Slater, 243 F.3d 270, 274 n. 3 (6th Cir.2001) (quotation omitted).

In order to determine whether a proposed action is a “major Federal aetion[ ] significantly affecting the quality of the human environment,” 42 U.S.C. § 4332(2)(C)(ii), agencies first look to whether the proposed action is a type of action that normally requires an EIS. That is, an agency determining whether an EIS is necessary must

(a) Determine under its procedures ... whether the proposal is one which: (1) Normally requires an environmental impact statement, or (2) Normally does not require either an environmental impact statement or an environmental assessment (categorical exclusion).

40 C.F.R. § 1501.4(a). Unless the proposed action is one that does not normally require either an EIS or an EA, the agency must undertake a NEPA analysis. 40 C.F.R. § 1501.4(b). If the action normally requires an EIS, the agency prepares an EIS; if the action normally requires an EA, the agency prepares an EA, and based on that assessment determines whether to proceed with a full EIS or issue a finding of no significant impact (“FONSI”). 40 C.F.R. § 1501.4(c)-(e). Even if the agency prepared an EIS when the project was first initiated, it must prepare a supplemental EIS (“SEIS”) if the agency intends to make “substantial changes” to the action or if “[tjhere are significant new circumstances,” and those changes or new circumstances would relate to the project’s effects on the environment. 40 C.F.R. § 1502.9(c)(1). NEPA makes no distinction between initial actions and subsequent changes to initial actions, and “the decision whether to prepare a supplemental EIS is similar to the decision whether to prepare an EIS in the first instance.” Marsh v. Oregon Natural Res. Council, 490 U.S. 360, 374, 109 S.Ct. 1851, 104 L.Ed.2d 377 (1989). That is, if the change itself constitutes a “major Federal action” that will “significantly affect[ ]” the environment, the agency must prepare a SEIS. 42 U.S.C. § 4332(2)(C)(ii).

The NEPA regulations’ default rule is that federal actions, unless shown to be to the contrary, require preparation of an EA. In other words, unless a proposed action appears on an agency’s list of categorical exclusions, the environmental analysis must be performed. See 40 C.F.R. § 1501.4(b). The action proposed here does not appear on the Corps’s list of categorical exclusions from NEPA analysis. See 33 C.F.R. § 230.9.

Indeed, the governing authorities explicitly indicate that this action normally does require an EA. Pursuant to 40 C.F.R. § 1507.3(b)(2), which requires federal agencies to define those actions that normally require an EIS, those that normally do not require an EIS or an EA, and those that normally require an EA but not necessarily an EIS, the Corps has identified actions requiring NEPA analysis. The proposed action clearly requires an EA, for in identifying actions that normally require an EA but not necessarily an EIS, the Corps has listed, “Changes in environmental impacts which were not considered in the project EIS or EA.” 33 C.F.R. § 230.7(d). Here, the record unambiguously indicates that this criterion is met: The 1974 EIS considered lead levels of 110 ppm and cadmium levels of 7.2 ppm; the Conner Creek sediment contains lead levels up to 1,100 ppm and cadmium levels of 31 ppm.² The syllogism is quite simple: Actions that an agency says ordinarily require an EA must receive EAs, see 40 C.F.R. § 1501.4(b), and this is an action that the Corps has said ordinarily requires an EA, 33 C.F.R. § 230.7(d).³ Therefore, this action must receive an EA.

It may be true that in the course of routine management of an ongoing operation, agencies may cause various environmental measurements to reach levels higher or lower than previous levels without triggering a NEPA analysis. These actions, such as raising or lowering water flow from a dam "within levels previously considered, do not necessarily require NEPA analysis. See Upper Snake River Chapter of Trout Unlimited v. Hodel, 921 F.2d 232, 235 (9th Cir.1990). However, this is not a case in which the agency is doing “nothing new, nor more extensive, nor other than that contemplated when the project was first operational,” id. at 235, but a case in which the proposed change is “incremental” but “of major proportions,” id. (quotation omitted). If an increase of less than 200% over the environmental impact considered in the original EIS requires NEPA analysis, see Northwoods Wilderness Recovery, Inc. v. United States Forest Serv., 323 F.3d 405, 411 (6th Cir. 2003), surely an increase of 1000% warrants no less.

I would therefore remand the case to the district court, with instructions that it amend its order to permit the Corps to comply expeditiously with NEPA’s requirements. NEPA’s demands may be achieved forthwith, and “under circumstances that ensure an objective evaluation,” Metcalf v. Daley, 214 F.3d 1135, 1146 (9th Cir.2000), but they may not be ignored.

1 . Wayne, Oakland, and Macomb Counties were also parties to the consent judgment.

1 . In addition to agreeing with Judge Keith that the All Writs Act authorizes a district court to issue writs to nonparties to prevent frustration of a consent judgment so long as the consent judgment was entered pursuant to law and not pursuant to private contract alone, and that a district court's discretion in issuing such an order should be evaluated using the four factors considered in United States v. New York Telephone Co., 434 U.S. 159, 174-75, 98 S.Ct. 364, 54 L.Ed.2d 376 (1977), I also join in Parts I, II, and III.A. of Judge Keith's opinion. I do not join Part III.B, which instructs the district court to determine whether the Administrative Procedure Act "specifically addresses the particular issue at hand,” in which case "it is that authority, and not the All Writs Act, that is controlling.” Syngenta Crop Prot., Inc. v. Henson, 537 U.S. 28, 123 S.Ct. 366, 369, 154 L.Ed.2d 368 (2002) (quotation omitted). The APA is a procedural device that enables courts to "compel agency action unlawfully withheld or unreasonably delayed,” and to set aside actions that are "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706. So far as I know, it is undisputed here that until the district court issued its All Writs Act order, the Corps was in no way obligated to accept the Conner Creek sediment. I thus see no agency action that, absent the All Writs Act order, could have been "unlawfully withheld," "unreasonably delayed,” or "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law,” under § 706 of the APA, and I see no reason that the district court would be better suited to make the initial, purely legal determination that is required in this analysis.

2 . That the Conner Creek sediment is not classified as "toxic” under the Toxic Substances Control Act, Pub.L. No. 94-469, 90 Stat.2003 (1976) (codified at 15 U.S.C. § 2683) ("TSCA”), or under state environmental regulations, is wholly irrelevant in determining whether NEPA requires analysis. The decision whether to prepare a NEPA analysis does not depend on whether the proposed action will be hazardous; if it did, federal agencies would have to consider the environmental consequences of their actions only when the action was prohibited by federal law. Laws governing toxic substances aim to "establish cleanup criteria" for, Mich. Comp. L. § 324.20120a(1), or "eliminate," 42 U.S.C. § 4851a(l), risks deemed wholly intolerable. In contrast, NEPA aims not simply to prevent actions that are prohibited, but to "assure consideration of the environmental impact of [agencies'] actions.” Kleppe v. Sierra Club, 427 U.S. 390, 409, 96 S.Ct. 2718, 49 L.Ed.2d 576 (1976) (quotations omitted). The purpose of an EIS would be to ensure discussion of the proposal's environmental impacts and to "inform decisionmakers and the public of the reasonable alternatives which would avoid or minimize adverse impacts or enhance the quality of the human environment." 40 C.F.R. § 1502.1. By requiring such analyses, NEPA aims not to identify actions that would violate environmental standards, but to force consideration of consequences and alternatives, see 40 C.F.R. § 1502.2(d), and thus ensure that, even if a proposed action is legal— as the Pointe Mouillee action appears to be— it will done in a way that minimizes unnecessary negative environmental effects.

3 . Even apart from the explicit regulatory classification of this kind of project as one normally requiring an EA, reference to the Corps's usual practice confirms that EAs are necessary in these circumstances. In the late 1990s, the Corps considered whether to accept environmental dredging from Black Lagoon. In that case, like this one, the proposed sediments contained contaminants at levels lower than those that would require regulation under the TSCA but higher than those that Pointe Mouillee had previously accepted: whereas Pointe Mouillee had previously accepted sediments containing mercury concentrations of 1-3 ppm, the average level of mercury in the Black Lagoon sediment was 4.3 ppm. Because the proposed disposal would involve slightly elevated levels of contaminants, and because the Black Lagoon sediment was the product of non-navigational dredging, which Pointe Mouillee had not previously accommodated, the Corps prepared a separate NEPA document. J.A. at 397 (April 7, 2000 Corps Memo).