Shieldalloy Metallurgical Corp. v. Nuclear Regulatory Commission

ROGERS, Circuit Judge,

concurring in part and dissenting in part.

I join the court in deferring to the interpretation of the Nuclear Regulatory Commission of its authority under the Atomic Energy Act, 42 U.S.C. § 2021 , to transfer jurisdiction over Shieldalloy’s Newfield site to the State of New Jersey. As the court explains, the NRC’s statutory reconciliation is plausible, and also consistent with legislative history indicating Congress’s desire to recognize states’ interests as well as the NRC’s prior state transfer approvals, including to Oklahoma. Op. at 375-76. I also join the court in concluding that the NRC’s interpretation of Criterion 25 of its guidance on state transfer agreements is not arbitrary or capricious or contrary to law. Op. at 375-77.

To the extent the court concludes, however, that the NRC’s transfer to New Jersey is arbitrary and capricious, and again remands the case, I respectfully dissent. The court inexplicably excuses Shieldalloy from two fundamental requirements: first, to raise its challenges to agency action with the agency so it has an opportunity to respond, and second, to state its challenges on appeal in more than a skeletal *384 way. Because Shieldalloy has done neither, despite a remand, and because, in not deferring to the NRC’s reasonable interpretation of its regulations, the court has injected a textual analysis of its own, a second remand is unwarranted. A review of the NRC’s analysis on remand demonstrates, moreover, that it has offered a reasoned response to Shieldalloy’s challenges, and I therefore would deny the petition for review.

I.

In remanding the case to the NRC for a second time, the court has concluded that the NRC’s response upon remand fails for lack of clarity. Op. at 373. Indeed, the court cloaks its disposition vacating the transfer of authority to New Jersey in concern that the NRC has somehow jeopardized public safety. See id. Yet the administrative record before the court indicates that any lack of clarity arises not from the NRC’s lack of articulation, or evidence it has failed to protect public safety, but from Shieldalloy’s repeated failure to set forth its arguments with sufficient clarity so that the NRC could respond to them.

When the court initially remanded the case to the NRC, it noted that Shieldalloy had failed to raise in its comments to the NRC that removal of Shieldalloy’s radioactive materials from the Newfield site in New Jersey to a facility in Utah would result in greater harms to public health and the environment than onsite disposal. See Shieldalloy Metallurgical Corp. v. NRC, 624 F.3d 489, 496 (D.C.Cir.2010). The court then observed that “the unacknowledged source of Shieldalloy’s criticisms regarding ALARA [i.e., the As Low As Reasonably Achievable principle], restricted use, and various standards for decommissioning” might be the “odd” fact that the License Termination Rule, 10 C.F.R. §§ 20.1401-06 , which prescribes conditions for decommissioning licensed facilities, “does not feature in the Criteria Document” that the NRC uses to assess the compatibility of state and federal regulatory programs. Id. at 496-97 . The court today again recalls that in the first appeal Shieldalloy raised a “claim pertain[ing] to a cost-benefit analysis principle embedded in the NRC’s regulatory program called the ALARA principle,” Op. at 374-75, but that the court did not address it “because Shieldalloy had not raised it sufficiently with the Commission.” Id. at 374 .

The court has long instructed what it reaffirmed in ExxonMobil Oil Corp. v. FERC, 487 F.3d 945 (D.C.Cir.2007):

A party must first raise an issue with an agency before seeking judicial review. This requirement serves at least two purposes. It ensures “simple fairness” to the agency and other affected litigants.

It also provides this Court with a record to evaluate complex regulatory issues; after all, the scope of judicial review under the APA would be significantly expanded if courts were to adjudicate administrative action without the benefit of a full airing of the issues before the agency.

Id. at 962 (citations omitted); see also Advocates for Hwy. & Auto Safety v. Fed. Motor Carrier Safety Admin., 429 F.3d 1136, 1148-50 (D.C.Cir.2005); United Transp. Union v. Surface Transp. Bd., 114 F.3d 1242 , 1244-45 (D.C.Cir.1997).

On remand from this court, the NRC “decided to examine anew all of the issues surrounding transfer of the Newfield site to New Jersey and afford Shieldalloy a fresh opportunity to comment on New Jersey’s agreement-state application.” Shiel-dalloy Metallurgical Corp., CLI-11-12, 74 NRC — , Memorandum and Order at 6-7 *385 (Oct. 12, 2011) (“Mem.”). Both Shieldalloy and New Jersey submitted new comments. Thereafter, the NRC reinstated New Jersey’s authority to regulate the Newfield site finding it “adequate” and “compatible” with the NRC’s regulatory program “within the meaning of section 274d and our implementing agreement-state policies.” Id. at 20 (citing 42 U.S.C. § 2021 (d)). 1 Among other things, the NRC described the regulatory framework, ALARA, the License Termination Rule, and New Jersey’s license termination program. See id. at 21-28 . It recounted Shieldalloy’s “misunderstandings regarding [the NRC’s] regulatory approach to license termination and ALARA principle,” id. at 42 , before considering, “in the proper context,” id., Shieldalloy’s position, “belatedly raised before the court but not as a comment on the New Jersey agreement,” id. at 34 , that “New Jersey’s license termination regulations are not as protective as [the NRC’s],” id. at 42 .

Further examination of the NRC’s response on remand is discussed in Part II, infra. What is significant at this point is Shieldalloy’s procedural default. On remand the NRC observed:

Despite the open-ended opportunity we provided in this remand proceeding for Shieldalloy to fully articulate its position on this and other issues, it has presented its “comparative dose” position, and its related argument as to ALARA, in summary and conclusory fashion, leaving us largely to guess at the technical rationale and underlying foundation for its position. This is unfortunate, given the highly complex and technical nature of our license termination regulations. While we endeavor to respond fully to Shieldalloy’s comparative dose and related ALARA argument based on our understanding of them, we are mindful of the admonition that “the dialogue between administrative agencies and the public is a two-way street.”

Id. at 35-36 (quoting Northside Sanitary Landfill, Inc. v. Thomas, 849 F.2d 1516, 1520 (D.C.Cir.1988)) (emphasis added and quotation marks omitted). The record in this court confirms that on remand Shieldalloy commented, in two short paragraphs, without providing technical support, that New Jersey’s “[f]ailure to implement the ALARA standard would allow New Jersey to reject the decommissioning option for the Newfield Facility that would result in the lowest doses to the public and the environment.” Response to NRC’s Jan. 3, 2011 Order at 15-16. It cited only 10 C.F.R. § 20.1003 , which defines ALARA. Id. at 15 .

The court today is in the same predicament as the NRC on remand as a result of Shieldalloy’s conduct. The court states: “In other words, if we understand Shiel-dalloy correctly, the proper application of the emphasized language [in 10 C.F.R. § 20.1403 (a) 2 ] would entail a comparison *386 between restricted and unrestricted release, and the former would win when it yielded lower risks than unrestricted.” Op. at 378 (emphasis added). The record shows that on remand Shieldalloy never cited 10 C.F.R. § 20.1403 (a), nor argued that the regulation, part of the License Termination Rule, either requires a “comparative dose” analysis or the NRC to approve a decommissioning option (either unrestricted or restricted) based on the outcome of that analysis.

Even now, in this second appeal challenging the NRC’s transfer of authority to New Jersey, it is at best doubtful that Shieldalloy has properly presented the issues regarding comparative dose and ALARA. In its opening brief, Shieldalloy contends, in conclusory fashion, leaving the court to guess at its rationale, that “there is a need to compare the radiological doses that would result from the decoinmission-ing of a facility under unrestricted release and restricted release approaches and to apply the ALARA principle to select the option that results in the lowest doses.” Pet’r’s Br. 43. Although not referencing 10 C.F.R. § 20.1403 (a) in its comments on remand, Shieldalloy now contends, summarily, that the regulation requires a comparison between the two decommissioning approaches. At most it offers only a con-clusory textual analysis of § 20.1403(a) and never explains why the regulation requires the NRC to approve one release option over another as a result of such an analysis. In similar fashion, Shieldalloy has provided no supporting technical rationale for its ALARA contention. This court has instructed: “ ‘It is not enough merely to mention a possible argument in the most skeletal way, leaving the court to do counsel’s work, create the ossature for the argument, and put flesh on its bones.’” Schneider v. Kissinger, 412 F.3d 190 , 200 n. 1 (D.C.Cir.2005) (citation omitted); see also Carducci v. Regan, 714 F.2d 171, 177 (D.C.Cir.1983).

That aside, on appeal Shieldalloy points for the first time to the NRC’s comprehensive guidance document on decommissioning, NUREG-1757, and to NRC staff requests for additional information purportedly in support of its comparative dose/ALARA contention. Additionally, only in its reply brief does Shieldalloy suggest that the NRC’s explanation of § 20.1403(a) — as requiring a licensee to demonstrate that remediation to the level of adequate protection to allow unrestricted release without institutional controls would not be cost-beneficial, see Resp’ts’ Br. at 55-56, 61 — “ignores the second half of 10 C.F.R. § 20.1403 (a) after the word ‘or’, that calls for an ALARA analysis of the restricted release option.” Reply Br. 15-16. Ordinarily, the court will not address arguments first raised in a reply brief. See United States v. Wilson, 605 F.3d 985, 1035 (D.C.Cir.2010); Students Against Genocide v. Dep’t of State, 257 F.3d 828, 835 (D.C.Cir.2001).

For purposes of this court’s review, one problem is that identified by the NRC on remand: Shieldalloy failed to provide the NRC with a textual analysis of § 20.1403(a) or a basis for concluding that the regulation calls for a comparative ALARA analysis or requires the NRC to approve a decommissioning option based on the outcome of that analysis. It also failed to provide technical support for its claim — a significant failure because § 20.1403(a) entails the use of ALARA, infra Part II.B. To that extent it has deprived the court of a record addressing *387 fully explicated and supported objections to the NRC’s transfer order. Having had its comparative dose interpretation twice rejected by the NRC, see Mem. at 35, much less the court’s observation that Shieldalloy had “not raised [its cost-benefit claim] sufficiently with the Commission,” Op. at 374 (citing Shieldalloy, 624 F.3d at 495-97 ), Shieldalloy had an obligation on remand to fully present argument and data in response to the NRC’s reactions to its approach. The second problem is that on appeal Shieldalloy has not meaningfully cured these deficiencies. So neither the court nor the NRC is clear about Shieldal-loy’s position (beyond opposing the transfer of authority and the unrestricted release option as less safe) even though Shieldalloy is a sophisticated corporate litigant that has had four opportunities to present its arguments — twice before the NRC and twice before this court, in addition to opportunities during the decommissioning proceedings since the early 1990s. Against this backdrop, the court’s decision to vacate the transfer order and remand the case again to the NRC is an unwarranted “windfall” for Shieldalloy.

II.

All is not lost, however; or at least it should not be. “[E]ndeavor[ing] to respond fully to Shieldalloy’s comparative dose and related ALARA argument based on [its] understanding of them,” Mem. at 35-36, the NRC stated on remand: “Shiel-dalloy apparently construes our license termination regulations as calling for a licensee to compare doses of the restricted-release and unrestricted-release decommissioning options and to choose the option that affords the lowest dose.” Id. at 36 (emphasis added). On appeal, Shieldal-loy does not dispute this interpretation of its position and the NRC addressed the issue comprehensively, providing a detailed and reasoned explanation as to how ALARA is used in the license termination regulations and why New Jersey’s program is compatible with the federal decommissioning standards. Id. at 34-44.

Upon reviewing the NRC’s response on remand, Shieldalloy’s contentions on appeal fail to present grounds upon which the court can conclude that the NRC failed to offer a reasoned explanation of its complex regulatory scheme, see Thomas Jefferson Univ. v. Shalala, 512 U.S. 504, 512 , 114 S.Ct. 2381 , 129 L.Ed.2d 405 (1994); Gen. Elec. Co. v. EPA, 53 F.3d 1324, 1327 (D.C.Cir.1995), or a reasonable interpretation of its regulations, see Auer v. Robbins, 519 U.S. 452, 461-63 , 117 S.Ct. 905 , 137 L.Ed.2d 79 (1997), or that the transfer of authority to New Jersey is arbitrary and capricious or contrary to law, see Motor Vehicle Mfrs. Ass’n of the U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 , 103 S.Ct. 2856 , 77 L.Ed.2d 443 (1983); 5 U.S.C. § 706 (2)(A). Our review is narrow, and the agency action being challenged is entitled to a presumption of regularity, see Citizens to Preserve Overton Park, Inc. v. Volpe, 401 U.S. 402, 415 , 91 S.Ct. 814 , 28 L.Ed.2d 136 (1971), abrogated on other grounds by Califano v. Sanders, 430 U.S. 99 , 97 S.Ct. 980 , 51 L.Ed.2d 192 (1977). To appreciate the vacuous nature of Shieldalloy’s challenge requires an overview of the NRC’s responses to Shieldalloy’s comparative dose and ALARA assertions.

First, the NRC explained that Shiel-dalloy’s apparent argument “is a fundamentally inaccurate understanding of our license termination requirements and appears to lie at the heart of Shieldalloy’s claim that New Jersey’s program is not as protective of the public health and safety as our program with respect to the Newfield site.” Mem. at 36. The NRC noted, citing a 44-page staff re *388 quest, that “this very misunderstanding ... was the subject of a number of requests for additional information by the staff on Shieldalloy’s 2006 decommissioning plan.” Id. at 36 n. 115. Stating its regulations “neither explicitly nor implicitly require a comparison of the levels of protection afforded by the unrestricted and restricted decommissioning options,” the NRC pointed out that “the levels of protection of unrestricted release and restricted release are simply not susceptible to being compared meaningfully” because each option has different methods and attendant risks. Id. at 37. Given the “inherent complexities and uncertainties associated with restricted release,” the NRC’s stated preference is for unrestricted release. Id. at 39. Moreover, the NRC found Shieldalloy’s own dose projections show the difficulty of meaningfully comparing doses.

Second, the NRC explained that the ALARA analysis in its regulation on restricted release, 10 C.F.R. § 20.1403 (a), does not “compel the selection of one decommissioning option over another,” id. at 43 , but instead acts as an eligibility requirement to screen out sites that should remediate to unrestricted release, in furtherance of the NRC’s preference. Id. at 25-26, 39-43 . Because, under its regulatory scheme, adequate protection of health and safety is accomplished “through satisfaction of the dose criteria and other conditions for [a] chosen decommissioning option,” id. at 41 , and New Jersey protects health and safety through a dose threshold that is more stringent than the NRC’s, the NRC concluded that New Jersey’s program is compatible with the federal regulatory scheme. I address each of these aspects of the NRC’s analysis.

A.

The NRC provided a reasoned explanation as to why its regulatory scheme does not envision a dose comparison between restricted and unrestricted decommissioning options. See Mem. at 36-39. The NRC explained that its “regulations neither explicitly nor implicitly require a comparison of the levels of protection afforded by the unrestricted and restricted decommissioning options” “because the levels of protection of unrestricted release and restricted release are simply not susceptible to being compared meaningfully.” Id. at 37 . In the NRC’s view, dose comparison is not meaningful because “[e]ach option uses significantly different methods to achieve adequate protection and has significantly different risks and uncertainties associated with it.” Id.

The NRC elaborated that restricted release is “far more complex and involves significantly greater uncertainties than off-site disposal,” because it “relies on the sustained effectiveness of institutional controls over a 1000-year compliance period to restrict future access and use to meet the 25 mrem per year dose requirement.” Id. To wit: Engineering controls must perform numerous complex functions; monitoring and maintenance are required; sufficient long-term funding from an independent third party is also required. Id. For this reason, the NRC explained, its stated preference is for unrestricted release. Id. at 39 . “Unrestricted release requires the removal of contamination on-site to the extent necessary to comply with the dose criteria of 25 mrem per year and transportation of the contaminated material to an isolated and regulated long-term disposal site.” Id. at 37 . Acknowledging that “[sjome uncertainties are inherent in these activities,” id. at 37-38 , the NRC explained, however, that they “generally involve!] well-known and quantifiable handling and associated radiological impacts on workers and the public over a short *389 time period (one to two years),” id. at 38 . Whereas, “dose estimates from contaminated slag left onsite [under the restricted release option] are subject to limitations in understanding the performance of a disposal system and its institutional and engineering controls over the course of the 1000-year compliance period.” Id.

For restricted use, the NRC explained, § 20.1403(a) provides a licensee with either of two cost-benefit approaches to demonstrate its eligibility, see Mem. at 25-26, what the NRC describes on appeal as either “a full cost-benefit analysis that compares all potential benefits to all potential costs (the ‘ALARA’ analysis), or an abbreviated cost-benefit analysis that compares all potential benefits to only a subset of potential costs that excludes the out-of-pocket costs of soil removal, transportation, and disposal (the ‘net public or environmental harm’ analysis).” Resp’ts’ Br. 59; see infra Part II.B. No such demonstration is required for unrestricted release, 10 C.F.R. § 20.1402 . 3 This contrast confirms the NRC’s interpretation of its regulations as not envisioning a dose comparison or that such a comparison dictates the choice of decommissioning method. While the NRC’s interpretation accounts for the contrasting text, Shieldalloy’s position does not. Instead, in reply Shieldal-loy takes the anomalous position that, in order to protect public safety, § 20.1403(a) requires the NRC to allow the decommissioning option resulting in the lowest ALARA-produced dose level — but “only when the licensee seeks to implement the restricted release option,” Reply Br. 19, not if it seeks to pursue unrestricted release, because the NRC “will allow th[at] option to be selected without further inquiry,” id. Shieldalloy ignores that the applicable dose threshold must be met, in order to provide adequate protection to the public, regardless of any ALARA analysis. Mem. at 40-41.

The NRC also addressed Shieldalloy’s dose comparison position on its own terms, pointing out its flawed factual premise. The NRC found, first, that Shieldalloy’s statements — that license termination using onsite disposal for the Newfield site would result in lower doses to the public than offsite disposal — “are inaccurate.” Id. at 34-35. Shieldalloy claimed that this proposition had not been controverted, but the NRC pointed out that insofar as Shieldal-loy’s comparative dose position was “reflected in its proposed 2005 decommissioning plan,” that plan was rejected by NRC staff as not being in compliance with the license termination regulations, and that the NRC staffs request for additional information on Shieldalloy’s proposed 2006 plan likewise indicated rejection of Shiel-dalloy’s comparison approach and identified related technical concerns. Id. at 35.

The NRC found, second, that “Shieldal-loy’s own dose estimates for the Newfield site reflect that it is meaningless to compare the level of protection between unrestricted release and restricted release.” Id. at 38 (emphasis added). Shieldalloy asserted on remand that “terminating the license under restricted conditions by the *390 method described in the [2009 decommissioning plan] (i.e., isolating the materials onsite under an appropriate cover) would result in doses to the decommissioning workers and the general public that are lower than those that would result from” decommissioning by Shieldalloy’s proposed method of unrestricted release. Response to Order at 13. While acknowledging that Shieldalloy’s 2009 “plan calculates an infinitesimally small dose” for restricted release assuming all controls hold, the NRC found that “when institutional controls are assumed to fail and the engineered cover is assumed to degrade, Shieldalloy’s filing shows that the dose estimate would be far greater, up to a bounding dose of 86 mrem per year” and “well in excess of Shieldal-loy’s dose estimates for unrestricted release.” Mem. at 38-39. In other words, Shieldalloy’s “own dose estimates for on-site disposal assuming the uncertainty and potential failure of controls over the long term in actuality show a higher dose.” Id. at 39. On appeal, Shieldalloy rejects the NRC’s interpretation of its regulations, summarily asserting that dose comparisons are possible and meaningful, see Reply Br. 19-20, but it does not refute the NRC’s finding that by Shieldalloy’s own estimates, were the cover to degrade, restricted release would expose the public to a higher dose of radiation than is reflected in its unrestricted release plan.

B.

The NRC reasonably explained how ALARA functions under its decommissioning regulations. Noting that “Shieldalloy has not set forth or explained the basis for its apparent position,” the NRC concluded, “perhaps it is alluding to our ALARA-based eligibility criterion for restricted release.” Mem. at 41 (emphasis added). Further noting “its submission is hardly clear on this point,” the NRC concluded “Shieldalloy apparently believes that our ALARA principle compels us to compare decommissioning options and to allow a licensee to select the lowest-dose option.” Id. at 40 (emphasis added). The NRC responded: “This is a fundamental misconception of our ALARA principle and appears to be the root of Shieldalloy’s misunderstanding of our approach to license termination.” Id. Indeed, the NRC explained that “the very premise of Shieldal-loy’s position on ALARA — that our license termination rule requires a choice to be made between a higher or lower dose option — is erroneous.” Id.

First, the NRC explained that the ALARA principle, “either as a general regulatory principle or as used in our license termination rule, [does not] incorporate or call for any comparative analysis of doses from restricted and unrestricted release.” Id. at 40. Rather, under the rule, the NRC explained, ALARA has two purposes in license termination: to reduce doses as low as achievable below applicable dose limits (not to “compar[e ] between achievable doses”), and to provide a criterion to limit the use of restricted release. Id. at 40-41. The latter purpose, “effectively, to screen out sites that should be removing contamination to achieve unrestricted use,” the NRC continued, is achieved in section § 20.1403(a) through the use of a cost-benefit analysis to determine initial eligibility for restricted release. Id. at 41. This eligibility criterion “was intended to support [the NRC’s] preference for the unrestricted-release decommissioning option.” Id. Licensees must act to remediate their sites to the dose threshold in order to terminate their licenses under either restricted or unrestricted release. Id. at 24-26. If the threshold is met by limiting access to the site and the ALARA analysis demonstrates that it would not be cost-beneficial to remove radioactive materials below the *391 dose threshold so that institutional controls are not required, then the site will be eligible for restricted release decommissioning; to qualify for unrestricted release, a licensee would need to undertake sufficient remediation or removal of radioactive materials so access to the site need not be limited or controlled. Id. at 24-26, 37, 39, 41; see 10 C.F.R. § 20.1003 . Thus, the NRC explained:

The ALARA analysis for restricted-release eligibility purposes does not and was never intended to demonstrate whether one decommissioning option affords greater protection than another. In fact, because an ALARA analysis focuses on dose reductions below what we have determined to be necessary for adequate protection of the public health and safety, that analysis does not go to adequate protection at all. A licensee’s demonstration of adequate protection is accomplished, instead, through satisfaction of the dose criteria and other conditions for its chosen decommissioning option.

Id. at 41 . Notably ALARA “does not compare or explicitly analyze any of the uncertainties that affect the level of protection afforded by a particular disposal option.” Id. at 42 .

Second, the NRC interpreted the ALARA test in § 20.1403(a) not to call for comparing doses between release options, “[n]or [to] compel the selection of one decommissioning option over another.” Id. at 43. In its opinion, “the ALARA requirement is irrelevant to whether Shiel-dalloy may pursue restricted release over unrestricted release in New Jersey.” Id. Omission of an ALARA-based criterion for restricted-release eligibility is, the NRC opined, “immaterial to [the] adequacy or compatibility” of the New Jersey program because adequate protection is accomplished through the dose threshold. Id. at 42. The NRC pointed out that New Jersey “accomplishes th[e] same objective” as the federal scheme does by the restricted-release eligibility criterion — namely, “to limit the use of restricted release in license termination” — “by adopting more stringent ... as well as more conservative criteria.” Id. at 42-43; see Radiological Criteria for License Termination, 62 Fed. Reg. 39,058 , 39,069 (July 21, 1997). The NRC thus reasonably concluded that New Jersey’s license termination program, which must meet only a Category C level of compatibility, 4 was adequate and compatible.

C.

Notwithstanding the NRC’s reasoned explanation of its regulatory scheme and reasonable interpretation of its regulations, including what ALARA is, the court concludes that a remand is required, principally in light of its own textual analysis of 10 C.F.R. § 20.1403 (a). See Op. at 388-94. Referencing the NRC’s brief on appeal, the court states that it is “baffled by the NRC’s current interpretation of § 20.1403(a)” and that it “do[es] not find the meaning of § 20.1403(a) self-evident,” but rather concludes that § 20.1403(a) neither precludes Shieldalloy’s reading “nor, at least without exegesis,” supports the *392 NRC’s interpretation. Op. at 379-80 (citing Resp’ts’ Br. 55, 59).

Undeterred by Shieldalloy’s defaults, the court engages in a textual interpretation of § 20.1403(a), Op. at 379-81, noting that on appeal the NRC discussed ALARA’s role in § 20.1403(a) as applying only to unrestricted release, not restricted release. Op. at 380. Section § 20.1403(a) provides — following the “or” — that a site is acceptable for license termination under restricted release if “[t]he licensee can demonstrate that further reductions in residual radioactivity necessary to comply with ... § 20.1402 [unrestricted release] ... were not being made because the residual levels associated with restricted conditions are ALARA.” (emphasis added); see supra note 2. In light of the italicized text, the court concludes that “the availability of restricted release under § 20.1403 would appear to have nothing to do with whether unrestricted release can be attained in a cost-beneficial manner, and everything to do with some property of restricted release.” Op. at 380. Hence the court finds confusion in the text, see id. at 379-81, implying that the final clause should read un restricted instead of restricted.

Ordinarily the court will defer to an agency’s reasonable interpretation of its regulations, see Auer, 519 U.S. at 461-63 , 117 S.Ct. 905 , and there is every reason to do so here where the complex regulatory scheme draws on the NRC’s expertise. The court acknowledges that its own interpretation of § 20.1403(a) “seems in tension with [the] second sentence of the provision, which states that the ALARA analysis referenced in the first sentence must account for ‘consideration of any detriments, such as traffic accidents, expected to potentially result from decontamination and waste disposal.’ ” Op. at 378 (quoting § 20.1403(a)). That tension does not exist under the NRC’s interpretation, whereby its regulations “demand that licensees ... demonstrate that remediation to the level of 25 mrem per year for unrestricted release would not be beneficial from a cost standpoint before allowing them to pursue restricted-use (onsite) disposal.” Resp’ts’ Br. 55-56; see also Mem. 25-26, 41. In other words, the licensee must meet the applicable dose threshold regardless of the ALARA analysis; that analysis, however, may reveal that release of the site without institutional controls (i.e., unrestricted release) is cost-beneficial. This accords with the NRC’s interpretation that its regulations do not imply dose comparisons, and instead only § 20.1403(a) sets an eligibility requirement because restricted release “is far more complex and involves significantly greater uncertainties than offsite disposal.” Mem. at 37.

The NRC’s “comprehensive guidance document” on decommissioning, id. at 24, supports the “exegesis” offered by the NRC. In describing the equation to be used to calculate doses that are ALARA, Appendix N states that “[t]he residual radioactivity level that is ALARA is the concentration ... at which the benefit from removal equals the cost of removal.” NU-REG-1757, Vol. 2, Appendix N at N-10 (Rev.l, Sept. 2006). The calculation of these benefits includes collective doses averted from a given action and the costs include the monetary costs to the licensee. Id. at § 6.3 at 6-3 to 6-4. So understood in light of NUREG-1757, the reference in § 20.1403(a) to a licensee demonstrating that “further reductions ... were not being made because the ... levels associated with restricted conditions are ALARA” means that no further radioactive materials can be eost-beneficially removed, washed away, or the like so the site can be decommissioned under § 20.1402 (unrestricted release). See id. at N-l to N-12; see also Mem. at 25-26, 41; Resp’ts’ Br. *393 55-56, 61. This understanding of ALARA was embodied in the NRC staff requests to Shieldalloy for additional information. See e.g., Request for Additional Information on Shieldalloy’s 2006 proposed decommissioning plan at 21 ¶ 31. That is, the “levels associated with restricted conditions are ALARA,” 10 C.F.R. § 20.1403 (a) (emphasis added), when no more actions can be cost-beneficially taken to meet the unrestricted use criteria.

Consequently, the confusion the court identifies arising from passages of other NRC regulations and statements is dispelled. Op. 380-82 (citing the definition of ALARA, NUREG-1757, and the staff Request for Additional Information). 5 In concluding these materials “can reasonably be read to call for precisely the kind of comparative dose analysis that Shieldalloy claims is contemplated by § 20.1403(a),” Op. at 381, the court engages in a recalibration of § 20.1403(a)’s eligibility requirement. That is not the same as demonstrating that the NRC’s interpretation is “baffling” or lacking needed “exegesis,” much less unreasonable and not entitled to deference. The suggestion of “prior inconsistent language,” id. at 381, fails for the same reason. Because the interpretation of § 1403(a) depends on the use of ALARA, including the equation set out in NUREG-1757, Shieldalloy’s cursory argument on remand was inadequate, despite the court’s assertions to the contrary, Op. at 383.

Inexplicably the court protests a lack of textual analysis by the NRC and the dissent, Op. at 379, 382, when, as discussed here, in fact the NRC explained, based on the plain text, that § 20.1403(a) affords a licensee two alternative ways to demonstrate its initial eligibility for restricted release. Mem. at 25-26. Further, it explained, licensees that do not demonstrate either initial eligibility or satisfaction of the remaining criteria of § 20.1403 “will not ‘be considered acceptable for license termination under restricted conditions’ ” and so must prepare their sites for unrestricted release, the NRC’s preferred decommissioning option because of its far more well-known and quantifiable radiological impacts. Id. at 26 (quoting § 20.1403); see also id. at 37-43. What is more, by explaining the use and function of ALARA in its license termination regulations, id. at 34-43, the NRC provided additional exposition of the text of § 20.1403(a), specifically what “restricted conditions are ALARA ” means. The court ignores that the formula and related guidance in NUREG-1757, which is used to assess if a remediation activity “is ALARA,” also elucidates the clause in § 20.1403(a) that the court finds “bafflfing],” Op. at 379; see supra note 2. And, in examining Shieldalloy’s own dose estimates for decommissioning, the NRC *394 demonstrated why the text cannot reasonably be read to call for the dose comparison envisioned by Shieldalloy.

Moreover, even assuming, for purposes of argument, that the NRC’s interpretation of § 20.1403(a) as not contemplating a comparative dose analysis is unreasonable, which it is not, neither the court nor Shiel-dalloy offers a basis on which to conclude that the NRC’s conclusion that the regulation does not “compel the selection of one decommissioning option over another,” Mem. at 43, based on the outcome of an ALARA analysis, is unreasonable. Instead, as the NRC explained, a licensee must meet the dose threshold regardless of the outcome of the ALARA analysis. Id. at 39, 41, 43. So the NRC could reasonably conclude that the NRC’s program and New Jersey’s program are not meaningfully different, let alone incompatible in regard to safety and the use of ALARA in § 20.1403(a): the New Jersey program is simply more stringent in using a 15 mrem instead of a 25 mrem threshold, which is permissible for Category C elements. See supra note 4.

In sum, the regulatory scheme administered by the NRC is complex, and further explanation is welcome as a general principle, but the issues raised by Shieldalloy have been adequately addressed by the NRC. This is so even though, because Shieldalloy never referred to § 20.1403(a) in comments on remand, the NRC had no occasion to explain more than its purpose and method, but see Op. at 379. Even on appeal Shieldalloy did not make the textual argument in its opening brief now provided by the court as the basis for another remand; neither does it offer that analysis in its reply brief. Although the court forswears that it is “substituting [its] judgment for that of the agency,” Op. at 382 (citation and quotation marks omitted), it clearly is substituting its own textual analysis for the bare assertions by Shieldalloy, see Schneider, 412 F.3d at 200 n. 1. These circumstances — namely, Shieldalloy’s default on remand, the NRC’s reasonable explanation of its regulations on remand, and the court’s acknowledgment that its reading of the NRC’s brief and its own interpretation of § 20.1403(a) creates an internal tension, Op. at 380-81 — present an odd basis for another remand much less a determination that the NRC’s transfer order is arbitrary and capricious.

. See Criteria for Guidance of States and NRC in Discontinuance of NRC Regulatory Authority and Assumption Thereof by States Through Agreement, 46 Fed.Reg. 7540 (Jan. 23, 1981), as amended by 46 Fed.Reg. 36,969 (July 16, 1981) and 48 Fed.Reg. 33,376 (July 21, 1983); Statement of Principles and Policy for the Agreement State Program; Policy Statement on Adequacy and Compatibility of Agreement State Programs, 62 Fed.Reg. 46,517 (Sept. 3, 1997) (the latter, "1997 Policy Statement”).

. Section 20.1403(a) provides that "[a] site will be considered acceptable for license termination under restricted conditions if:”

The licensee can demonstrate that farther reductions in residual radioactivity necessary to comply with the provisions of § 20.1402 [standards for unrestricted release] would result in net public or environmental harm or were not being made because the residual levels associated with restricted conditions are ALARA. Determi *386 nation of the levels which are ALARA must take into account consideration of any detriments, such as traffic accidents, expected to potentially result from decontamination and waste disposal.

10 C.F.R. § 20.1403 (a) (emphasis added).

. S ection 20.1402 provides:

A site will be considered acceptable for unrestricted use if the residual radioactivity that is distinguishable from background radiation results in a TEDE [total effective dose equivalent] to an average member of the critical group that does not exceed 25 mrem ... per year, including that from groundwater sources of drinking water, and that the residual radioactivity has been reduced to levels that are as low as reasonably achievable (ALARA). Determination of the levels which are ALARA must take into account consideration of any detriments, such as deaths from transportation accidents, expected to potentially result from decontamination and waste disposal.

. Pursuant to the 1997 Policy Statement, supra note 1, Agreement States may adopt programs that provide a level of protection that is "equivalent to, or greater than, the level provided by the NRC program.” 62 Fed.Reg. at 46,524. Category C elements of Agreement State programs are required "in order to avoid conflicts, duplications, gaps, or other conditions that would jeopardize an orderly pattern in the regulation of agreement material on a nationwide basis." Id.; see also Radiological Criteria, 62 Fed.Reg. at 39,079-80, 39,086.

. Neither the passages in NUREG-1757 nor the staff Request for Additional Information support Shieldalloy's position that § 20.1403(a) requires a comparative dose analysis. Instead, those passages relate to the required ALARA analysis of restricted release, which when properly conducted, will reveal if additional remediation to meet the requirements of unrestricted release is cost-beneficial. The court’s conclusion that under the NRC’s view ALARA is used in § 20.1403(a) "to assess the cost-efficiency of attaining radiation levels at those [dose ] limits,” Op. at 380 (emphasis in original) overlooks that the NRC has consistently stated that ALARA is used in § 20.1403(a) to assess the cost-efficiency of remediating below the applicable dose threshold for restricted release in order to meet the criteria for unrestricted release. Mem. at 40-41; Resp’ts' Br. 55-56. This accords with the NUREG-1757 guidance in Appendix N at N-l ("In order to terminate a license, a licensee should demonstrate that the dose criteria ... have been met, and should demonstrate whether it is feasible to further reduce the levels ... below those necessary to meet the dose criteria (i.e., to levels that are ALARA).”).