Opinion for the Court filed by Circuit Judge ROBERTS.
Concurring opinion filed by Senior Circuit Judge WILLIAMS.
ROBERTS, Circuit Judge.
An Oklahoma agency licensed by the Federal Energy Regulatory Commission to operate a hydroelectric project applied for a license amendment to permit the expansion of a commercial marina located on project property. FERC approved the [3]*3amendment over the opposition of petitioners — two lakefront homeowners — who argued that the expansion was not in the public interest. In particular, they maintained that the cove where the marina is located had exceeded its “carrying capacity,” precluding further growth. They also contended that FERC could not approve the expansion in the absence of. comprehensive guidelines for shoreline development on the project, and that FERC erred in approving the state agency’s waiver of one of its regulations.
Our review of FERC’s decision is limited, and under that limited review we conclude that FERC considered the impact of the proposal on a variety of non-development objectives, as required by the applicable statute, and that FERC’s treatment of the carrying capacity issue was not arbitrary or capricious or lacking in substantial evidence. We also find that FERC adequately explained its decision to proceed in the absence of comprehensive guidelines for shoreline development, and its determination to approve the departure from the state agency’s regulations. We deny the petition.
I.
The Pensacola dam in northeast Oklahoma impounds the waters of the Grand or Neosho River, creating the Grand Lake O’ the Cherokees. The dam and resulting lake are regulated by the Grand River Dam Authority (GRDA), a state agency, pursuant to a license issued by FERC. See Grand River Dam Authority, Order Issuing New License, 59 FERC ¶ 62,073, 1992 WL 510727 (1992).
FERC’s licensing authority for such hydroelectric projects is governed by Section 4(e) of the Federal Power Act, 16 U.S.C. § 797(e). That section provides, in pertinent part:
In deciding whether to issue any license under this subchapter for any project; the Commission, in addition to the power and development purposes for which •licenses are issued, shall give equal consideration to the purposes of energy conservation, the protection, mitigation of damage to, and enhancement of, fish and wildlife (including related spawning grounds and habitat), the protection of recreational opportunities, and the preservation of other aspects of environmental quality.
Further guidance is found in Section 10(a)(1), 16 U.S.C. § 803(a)(1), which conditions all licenses on a requirement that
the project adopted ... shall be such as in the judgment of the Commission will be best adapted to a comprehensive plan for improving or developing a waterway or waterways for the use or benefit of interstate or foreign commerce, for the improvement and utilization of waterpower development, for the adequate protection, mitigation, and enhancement of fish and wildlife (including related spawning grounds and habitat), and for other beneficial public uses, including irrigation, floód control, water supply, and recreational and other purposes referred to in section 797(e) of this title.
GRDA’s license calls on that agency to develop a long-term recreation management plan for Grand Lake, including an estimate of the lake’s “carrying capacity” — defined as “the level of use ... that would begin to detract from a safe or enjoyable recreation experience” — and “measures for managing lake use if it exceeds the carrying capacity.” Order Issuing New License, 59 FERC at 63,230. Regarding the agency’s oversight of boat docks, the license permits GRDA to authorize the construction of “noncommercial” structures “that can accommodate no more than 10 watercraft at a time.” Id. at 63,231. Larger undertakings require the [4]*4agency to apply to FERC for a license amendment.
GRDA filed a recreation plan in 1997. Appended to the plan was a study of carrying capacity conducted by faculty and students at Oklahoma State University. See Pensacola Recreation Management Plan, App. F. That study, the plan noted, did not “defin[e] a ‘magic number’ that represented the statistical carrying capacity of Grand Lake.” Rec. Plan at 15-16. The plan also included a separate memorandum by the study’s principal author, Professor Lowell Caneday, which did set forth a specific capacity figure for the lake: 0.13 boats per acre. Id., App. I at 3. Based on aerial photographs, estimated boating densities for the lake as a whole were below Dr. Caneday’s figure in 1996, but above it in 1997. See id., App. G & H.
GRDA explained in the recreation plan that, in light of the carrying capacity information, it had adopted the goal of making certain changes in the management of Grand Lake. Rec. Plan at 16. Notably, these changes did not include a halt to shoreline development, although a management proposal featured in the OSU study called for such a step. Com/pare id., App. J (GRDA management strategy) (“Grand Lake will provide ... (F) Shoreline development following best management practices and existing regulations”) with id., App. F at 83 (OSU proposal) (“Grand Lake ... will provide ... (f) restriction of further shoreline development”).
The recreation plan also addressed the license requirement that it include recommended measures for managing lake use if carrying capacity were exceeded:
If the use of the project exceeded its documented carrying capacity, GRDA could implement seasonal rules targeted at the activity. More seasonal patrolmen could be hired to enforce existing and future restrictions. Special events could be limited to certain days of the week or the months they would be allowed.
Id. at 17. Finally, the plan featured proposed rules and regulations governing shore and water use — including a rule that docks may not extend into the water the longer of 125 feet or one-third of the distance to the opposite shore. Id., App. B at 6.
FERC approved the recreation plan. See Grand River Dam Authority, Order Modifying and Approving Long-Term Recreation Plan, 84 FERC ¶ 62,144, 1998 WL 481851 (1998). In particular, the Commission endorsed GRDA’s capacity management approach, including its decision not to implement the strategy outlined in the OSU study that would be more restrictive with respect to development. See id. at 64,229 (“While [the OSU strategy] includes greater restrictions, we believe the licensee’s proposal is reasonable at this time given the project’s slower than expected growth in recreation uses and the currently underutilized recreation facilities”). Anticipating that the lake would experience increased shoreline development in the future, however, and noting GRDA’s express willingness to do so, FERC stated that GRDA “should continue its efforts to work with project stakeholders on the development of a comprehensive [shoreline] management plan [CSMP] for the project.” Id. at 64,231. FERC also directed the agency to file periodic reports monitoring recreation use and shoreline development, as well as progress updates on the development of a CSMP. Id. at 64,232.
Duck Creek cove is an arm of Grand Lake described as “one of the lake’s most frequently used boating areas.” Final Environmental Assessment at 27. In 2001, the owner of Arrowhead North Marina, a commercial facility located on Duck Creek [5]*5cove, asked GRDA for permission to expand and reconfigure the facility. The expansion sought to add 64 boat slips, many for large boats. GRDA endorsed the plan, and, as required by its license, filed an application with FERC for a license amendment permitting the expansion.
Numerous area residents, as well as the U.S. Department of the Interior, submitted comments regarding the proposal. In-tervenors opposing the expansion included Mike Brady and Cheryl Creekmore, owners of property on Duck Creek. While proceedings were pending before the Commission, GRDA filed its first monitoring report on recreation use and shoreline development, which showed a continuing rise in boating densities on the lake. See 2003 Recreation Plan Monitoring Report at 3 & Attachment C.
FERC approved the marina expansion, over the dissent of one Commissioner. Grand River Dam Authority, Order Approving Non-Project Use of Project Property, 105 FERC ¶ 61,100, 2003 WL 22422344 (2003). The Commission reasoned that the proposed expansion would have “beneficial impacts on employment, tax revenues and tourism,” offsetting “some adverse impacts on the visual character and scenic quality of the landscape.” Id. at 61,506-07. The Commission further found that the expansion “would result in only moderate long-term adverse impacts to boating use and navigational safety.” Id. at 61,507. The Commission also held that it could approve the proposal even though the reconfigured marina exceeded GRDA’s dock length restrictions. Id.
Turning specifically to the issue of carrying capacity, FERC found that “while there is evidence that boating densities in the cove are increasing, such evidence does not confirm that the cove has reached its carrying capacity.” Id. In a final environmental assessment accompanying the order, FERC staff added that “the resources of Grand Lake’s Duck Creek arm are approaching carrying-capacity limits for boating-related facilities and activities” and promised to “intensively scrutinize any future non-project proposals involving this cove.” Final Environmental Assessment at 51. Lastly, although GRDA had not yet developed a CSMP, the Commission refused to delay consideration of the proposal until such a plan was submitted. ' Order Approving Non-Project Use, 105 FERC at 61,507.
Brady and Creekmore filed requests for rehearing, which were denied without further opinion. Grand River Dam Authority, Notice of Denial of Rehearing, 105 FERC ¶ 61,310, 2003 WL 22987231 (2003). They now petition for review of FERC’s order approving the marina expansion.
II.
In reviewing licensing proceedings, we will uphold FERC’s conclusions unless they are “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” 5 U.S.C. § 706(2)(A). The Commission’s findings of fact are conclusive if supported by substantial evidence. 16 U.S.C. § 825i (b). In both regards, the scope of our review is quite limited; “[s]o long as the Commission has examined the relevant data and provided a reasoned explanation supported by a stated connection between the facts found and the choices made, we will defer to the agency’s expertise.” North Carolina v. FERC, 112 F.3d 1175, 1189 (D.C.Cir.1997); see also U.S. Dep’t of Interior v. FERC, 952 F.2d 538, 543 (D.C.Cir.1992) (“In a licensing decision such as this, where few explicit statutory provisions govern, our role is narrowly circumscribed.”).
A. Brady and Creekmore begin with the Federal Power Act. They assert that, in approving the marina expansion, [6]*6FERC abdicated its duty to give “equal consideration” to the non-development purposes specified in Section 4(e) of the FPA, including protection of fish and wildlife, recreational opportunities, and other aspects of environmental quality, and failed to ensure that the project is “best adapted to a comprehensive plan” that incorporates such non-development purposes, as required by Section 10(a)(1) of the FPA. 16 U.S.C. §§ 797(e), 803(a)(1). See Pet. Br. at 26. Congress directed that FERC consider a variety of factors in making licensing decisions under the FPA, but the decision about how to weigh those factors remains with FERC. The judicial role is “narrowly circumscribed,” and “[t]he statutory ‘equal consideration* requirement does not change the standard of review that we, as an appeals court, apply.” U.S. Dep’t of Interior, 952 F.2d at 543, 545. FERC decides how to balance the various elements listed in the FPA; our task is to ensure that it considers the statutory factors and that its decisions meet the threshold standards of non-arbitrariness and substantial evidence.
Contrary to petitioners’ suggestion, FERC did not fail to consider non-development public uses. See Order Approving Non-Project Use, 105 FERC at 61,506-07 (discussing, in addition to “visual character and scenic quality,” impacts on “fish and wildlife, air and water quality, ambient noise levels, [and] shoreline access,” as well as “boating use and navigational safety”). Nothing on the face of FERC’s order suggests a disregard of the considerations listed in the FPA. While Brady and Creekmore would strike the balance differently, against further development, the policy decision about how to weigh the pertinent factors is FERC’s to make. The amendments to the FPA that added the “equal consideration” language “were aimed primarily at increasing FERC’s sensitivity to environmental concerns,” but those amendments “do not give environmental factors preemptive force.” U.S. Dep’t of Interior, 952 F.2d at 545.
In making the foregoing determination, we note that petitioners have only raised FERC’s duty to balance development with competing considerations. They have not challenged — either in their petitions for rehearing before FERC or in the briefs submitted to us — the propriety of the Commission’s reliance on “employment, tax revenues and tourism” as the development benefits of this proposal. Accordingly, we do not address whether the Commission may consider these particular benefits, either generally or under the particular circumstances of this case. See 16 U.S.C. § 825i (b) (“No objection to the order of the Commission shall be considered by the court unless such objection shall have been urged before the Commission in the application for rehearing unless there is reasonable grounds for failure to do so.”); compare Nat’l Wildlife Fed’n v. FERC, 912 F.2d 1471, 1482-83 (D.C.Cir.1990) (presenting a challenge to FERC’s reliance on “water supply” benefits under FPA Section 4(e)). All we conclude is that the Commission satisfied the statutory requirement to consider a variety of specified public uses, including those militating against development.
B. Petitioners attempt to buttress their statutory argument by focusing on the concept of carrying capacity. They maintain that the evidence before the Commission indicates that the lake’s capacity has been exceeded, and that in approving further development, FERC necessarily proceeded on the erroneous assumption that it had not.
We agree that the record contains evidence of crowded conditions at Duck Creek. Aerial photographs taken in 1997 show that boating densities in the cove were significantly higher than those for the lake as a whole. See Rec. Plan, App. H. In surveys conducted as part of the [7]*7OSU study, more people listed Duck Creek as a place to avoid than any other area of the lake, citing “too many boats” and “unsafe or too dangerous” conditions. Id., App. F at 72. The 2003 monitoring report indicates that, prior to the institution of directional traffic lanes, the cove was the site of more boating accidents than any other arm of the lake. See 2003 Recreation Plan Monitoring Report, Attachment B.
FERC, however, did not ignore this evidence. In the order approving the marina expansion, the Commission expressly acknowledged that “boating densities in the lake and the cove continue to increase,” and that “boating densities in the cove are increasing.” Order Approving Non-Project Use, 105 FERC at 61,507. FERC noted the new boat-traffic lanes and speed limits GRDA had implemented to accommodate the increasing number of boats in the cove, and required a slight change in the location of several of the proposed slips to ensure that boats entering and leaving the marina did not present a safety concern. Id. The issue of crowding was plainly recognized and addressed by FERC.
FERC also specifically addressed the issue of crowding relative to the concept of carrying capacity. The Commission concluded that “while there is evidence that boating densities in the cove are increasing, such evidence does not confirm that the cove has reached its carrying capacity.” Id. Despite the acknowledged crowded conditions in Duck Creek, this conclusion was a permissible one for the Commission to reach:
First, the concept of “carrying capacity” does not present the sort of bright-line limit that would mandate a particular finding on the facts before the Commission. The recreation plan does not treat capacity in quantitative terms. As noted in the text of the plan, the appended OSU study did not set forth a “magic number” of boats above which the collective lake experience becomes unsafe or unenjoyable. See Rec. Plan at 15-16. Dr. Caneday’s separate memorandum, which does set forth such a value, concedes that a “concern with this number is that it reflects merely the standard ‘quantity’ of watercraft without consideration for the ‘quality’ of experience for the visitor or ‘quality’ of the lake environment.” Id., App. I at 3. What is more, the OSU study makes clear that “[tjhere are many types of carrying capacity,” id., App. F at 82, and that the “management-objective recreation carrying capacity” is “undefined,” because it depends on the goals and objectives of the managing agency. Id. at 80, 90. The amorphous nature of the carrying capacity concept (or concepts) as presented by the OSU study makes it difficult for us to fault FERC’s conclusion that the record evidence did not confirm that the capacity of the cove had been exceeded.
Second, Dr. Caneday’s estimate speaks to the capacity of the lake as a whole, whereas FERC’s order focuses on the capacity of the cove. This difference is significant: according to the OSU study, “[mjeasurements of recreational carrying capacity must be site specific.” Rec. Plan, App. F at 81. The record amply demonstrates that Duck Creek has characteristics not shared by the lake as a whole: for instance, boats must proceed through a portion of the cove at “no wake” speed, and no waterskiing or swimming is permitted in the directional traffic lanes at any time. See 2003 Recreation Plan Monitoring Report, Attachments D & E. Dr. Cane-day’s estimate for the lake, by contrast, factors in space for activities such as unlimited powerboating and waterskiing. See Rec. Plan, App. I at 3.1 Thus, what [8]*8limited tangible guidance is provided by the record does not speak to the particular issue before the Commission.2
Petitioners broaden their argument to challenge not only FERC’s dismissal of their evidence on carrying capacity, but more generally FERC’s decision to approve the license amendment on what petitioners regard as the necessary “assumption” that the cove has not exceeded carrying capacity. See Reply Br. at 1-2, 9. Nothing, however, requires FERC to conduct its statutorily-mandated consideration and balancing of the pertinent factors under the “carrying capacity” framework. Certainly nothing in the FPA does so — the statute never mentions “carrying capacity.”
Nor does the recreation plan mandate that FERC discharge its responsibilities by assessing carrying capacity; GRDA’s license required it to submit a plan including an estimate of carrying capacity, but the recreation plan does not state that the estimate would or should become the benchmark for FERC under the FPA. Under the recreation plan, a finding that carrying capacity has been exceeded does not trigger a ban on commercial development. In the section of the plan titled “Carrying Capacity Management Methodology,” GRDA commits to monitoring future recreation use in consultation with other state agencies. See Rec. Plan at 16-17. The agency then proposes the following methods of dealing with an exceeded capacity:
If the use of the project exceeded its documented carrying capacity, GRDA could implement seasonal rules targeted at the activity. More seasonal patrolmen could be hired to enforce existing and future restrictions. Special events could be limited to certain days of the week or the months they would be allowed.
Id. at 17. Plainly, these remedies do not include the suspension of commercial development.
With respect to development, the recreation plan seems to contemplate a different approach: “The construction of commercial facilities will be controlled by demand and the decreasing availability of suitable locations.” Id. at 22. FERC recognized this, noting in its order approving the amendment that “[ujnder the approved recreation plan for the project, shoreline development is currently controlled by demand and site availability.” Order Approving Non-Project Use, 105 FERC at 61,505; see also id. at 61,507 (noting, with respect to earlier expansion of another marina, that “under the project’s recreation plan approved in 1998, shoreline development is currently controlled by demand and site availability”).
This is not to say that FERC may rest on “demand and site availability” in approving license amendments; on the contrary, the FPA requires the Commission to weigh a number of factors. It does make clear, however, that the recreation plan itself cannot be read as imposing a duty on FERC to block development solely in light of any particular level of carrying capacity, or to assess proposals in light of a carrying capacity analysis.
[9]*9And that is in fact not how FERC read the plan when it approved it. FERC’s order approving the plan noted that GRDA “does not specify a specific number for carrying capacity but proposes to use the previously prepared recreation study to assist in its determination and implementation of future actions at the project to protect and promote recreation usage on Grand Lake.” Order Modifying and Approving Long-Term Recreation Plan, 84 FERC at 64,227-28. FERC recognized GRDA’s commitment that “if recreation use exceeds documented carrying capacity” GRDA could take a number of specified steps — none of which included a ban on development. Id. at 64,228. Petitioners rely to a large extent on the evidence in the 1997 recreation plan to establish that carrying capacity has been exceeded, but FERC obviously had that evidence before it when it approved the plan in 1998, and in particular approved GRDA’s decision to adopt a less restrictive approach to development than that recommended by the OSU study. See id. at 64,229. It was obvious from the get-go, therefore, that neither the recreation plan nor FERC’s approval of it interpreted carrying capacity as a necessary limit on development, as petitioners now envisage it, and certainly not as a limit on how FERC analyzes proposed license amendments.
Consistent with the foregoing, FERC’s precedents since the approval of the recreation plan do not dictate a “carrying capacity” analysis. Of the many prior decisions approving dock expansions on Grand Lake, only one squarely engaged the concept, and it failed to give carrying capacity any discernible weight. See Order Approving Non-Project Use of Project Lands and Waters, 95 FERC ¶ 62,184, 64,268-70, 2001 WL 586944 (2001) (noting the shortcomings of Dr. Caneday’s numeric estimate); compare, e.g., Order Approving Non-Project Use of Project Lands, 97 FERC ¶ 62,132, 2001 WL 1379467 (2001) (no discussion of carrying capacity); Order Approving Now-Project Use of Project Lands, 94 FERC ¶ 62,263, 2001 WL 323208 (2001) (same); Order Modifying and Approving Non-Project Use of Project Lands and Waters, 92 FERC ¶ 62,235, 2000 WL 1349361 (2000) (same); Order Approving Non-Project Use of Project Lands and Waters, 92 FERC ¶ 62,238, 2000 WL 1358717 (2000) (same); Order Approving Non-Project Use of Project Property, 89 FERC ¶ 61,139, 1999 WL 989535-(1999) (same).
Finally, the absence of support for petitioners’ broader argument should come as no surprise. A carrying capacity analysis would simply determine whether approval of a proposal entails a net detriment with respect to recreational opportunities. See Order Approving New License, 59 FERC at 63,230 (defining carrying capacity as “the level of use ... that would begin to detract from a safe or enjoyable recreation experience”) (emphasis added). But just as we have held that environmental factors do not have “preemptive force” in the weighing process, U.S. Dep’t of Interior, 952 F.2d at 545, neither do recreational factors. FERC recognized that approving the proposal would result in “moderate long-term adverse impacts to boating use and navigational safety,” 105 FERC at 61,507, but determined that the perceived benefits — not challenged here — were nonetheless worth it. Our “narrowly circumscribed” role, U.S. Dep’t of Interior, 952 F.2d at 543, does not include second-guessing that judgment.
C. Petitioners next turn to what they decry as the absence of any definitive standards for development on Grand Lake. According to Brady and Creekmore, FERC should have imposed a moratorium on development pending the creation of a CSMP, or at least required the Arrowhead marina to conform to the length restric[10]*10tions in GRDA’s rules and regulations. Without a CSMP, petitioners argue, the rules are the only possible standards to apply.
When it approved GRDA’s recreation plan in 1998, FERC noted that the state agency should continue its efforts to develop a CSMP, and directed the agency to file periodic reports documenting its progress. Order Modifying mid Approving Long-Term Recreation Plan, 84 FERC at 64,-231. The state agency has yet to complete the CSMP, and FERC has accordingly been compelled to consider license amendments on a case-by-case basis. Such an approach does not provide a clear roadmap for ordered development of the Pensacola project.
FERC staff, however, found that a CSMP “will require a substantial and uncertain amount of time to develop.” Final Environmental Assessment at 52. Under these circumstances, it was within the Commission’s authority to act on the expansion proposal instead of instituting a moratorium. At any rate, neither the text of the FPA nor FERC’s prior orders relating to the project preclude such a course. See, e.g., 16 U.S.C. § 803(a)(2)(A) (Commission shall consider “[t]he extent to which the project is consistent with a comprehensive plan (where one exists)”) (emphasis added); Order Modifying and Approving Long-Term Recreation Plan, 84 FERC at 64,231-32 (requiring GRDA to submit progress reports but failing to establish a deadline for development of the plan itself).
As for the departure from GRDA’s regulations, FERC noted in its decision that the regulations had not been adopted by FERC as license conditions. Order Approving Nonr-Project Use, 105 FERC at 61,507. The FPA sets forth the criteria that FERC is to consider in approving a proposed license amendment, and that process cannot be preempted or displaced by local regulations. As FERC explained, “we may consider a proposal’s consistency with these regulations, but we are not bound by them.” Id.
D. Brady and Creekmore raise two other objections that merit only brief discussion. First, they contend that the Commission improperly relied on the lack of county land use requirements in deciding that lakefront homeowners did not have a “realistic expectation that their environs would remain undisturbed.” Id. at 61,507. Second, they argue that FERC disregarded two alternate configurations for the Arrowhead marina proposed by petitioner Brady.
As for the first objection, petitioners rightly maintain, as we just noted, that local land use requirements cannot supplant the Commission’s duties under the FPA. But FERC does not suggest otherwise; rather, the Commission explained that it simply noted expectations affected by local land use- — in addition to the many other factors it examined — as part of the overall statutory balancing process.
The second objection is simply incorrect. The final environmental assessment considered, and rejected, both of Brady’s alternate designs. See Final Environmental Assessment at 81.
III.
The Federal Power Act requires FERC to balance various public uses in making licensing decisions. In its oversight of the Pensacola project, the Commission to date has plainly seen the balance of statutory factors as tipping in favor of development. Brady and Creekmore make a vigorous case for the opposite outcome, but so long as FERC’s decisions are not arbitrary or capricious, or lacking in substantial evidence, this court does not have the power [11]*11to alter the Commission’s judgment. The petition for review is denied.