OPINION AND ORDER
SHIRA A. SCHEINDLIN, District Judge:
I. INTRODUCTION
This is a consolidated multi-district litigation (“MDL”) relating to contamination — actual or threatened — of groundwater from various defendants’ use of tie gasoline additive methyl tertiary butyl ether (“MTBE”) and/or tertiary butyl alcohol, a product formed by the breakdown of MTBE in water. In this case, the New Jersey Department of Environmental Protection (“NJDEP”), the Commissioner of the NJDEP, and the Administrator of the New Jersey Spill Compensation Fund allege that Defendants’ use and handling of MTBE has contaminated, or threatened to contaminate groundwater at service stations, refineries, and terminals throughout [261]*261New Jersey. Familiarity with the facts of this case is presumed for the purposes of this Order.
Currently before the Court is Plaintiffs’ motion for judicial approval of the Judicial Consent Order (“JCO”), which recites the terjns of their settlement with Citgo Petroleum Corporation (“Citgo”). The JCO resolves all claims against Citgo for $23.25 million. Several Defendants (“Non-Settling Defendants”)1 oppose the JCO on the grounds that it does not account for Cit-go’s proportionate share of liability.2 For the reasons stated below, Plaintiffs’ motion is DENIED.
II. BACKGROUND
A. Plaintiffs’ Claims
Plaintiffs filed this lawsuit to recover damages for alleged MTBE contamination in groundwater at 5,045 sites throughout New Jersey. Plaintiffs’ Fourth Amended Complaint (“Complaint”) alleges two statutory claims under the New Jersey Spill Compensation and Control Act (“Spill Act”) and four common law claims against all Defendants.3 Pláintiffs request (1) the costs of restoring MTBE-contaminated groundwater (“restoration costs”), (2) the costs of past and future MTBE testing of all public water supplies, (3) the costs of past and future treatment of all drinking water supplies containing detectable levels of MTBE, (4) the costs of past and future monitoring of other waters to detect MTBE, (5) the costs of past cleanup and removal costs, and (6) attorneys’ fees and costs.4
B. The Focus Sites
The parties have not conducted site specific discovery at each of the 5,045 sites at issue.5 Instead, at the start of discovery, Plaintiffs and Defendants each selected ten focus sites.6 Plaintiffs’ experts submitted reports offering damages opinions for the ten Plaintiff-selected sites and one Defendant-selected site.7 As to the remaining Defendant-selected sites, Plaintiffs’ experts concluded that the costs of assessing damages would exceed the damages in[262]*262curred.8
Plaintiffs allege that Citgo is responsible for MTBE contamination at four of the ten Plaintiff-selected sites: (1) the Skyline Service Center, (2) the Maple Shade Citgo, (3) the 5-Points Site, and (4) the HP Delta Service Station.9 As such, Citgo is targeted at more Plaintiff-selected sites that any other Defendant.10
C. The Terms of the JCO
After six months of negotiation, Plaintiffs settled with Citgo. Under the JCO, Citgo would pay $23.25 million for contribution protection and a release from liability at all 5,045 sites in this case.11 The JCO also entitles Non-Settling Defendants to a settlement credit to be applied at trial.12 The JCO adopts the Spill Act’s credit scheme for Plaintiffs’ Spill Act claims and the common law credit scheme for Plaintiffs’ common law claims.13 In addition, the JCO grants Citgo contribution protection under the Joint Tortfeasors Contribution Law, the Comparative Negligence Act, and common law principles.14
D. Non-Settling Defendants’ Objections
Non-Settling Defendants’ objections are based on the- unique way that settlements are treated under the Spill Act.15 Under the Spill Act credit scheme, non-settling defendants receive a credit in the dollar amount of the settlement.16 However, under the common law credit scheme, non-settling defendants receive a credit based on the settling party’s percentage of liability as determined by the trier of fact.17
Hypothetically, a jury could award Plaintiffs three hundred million dollars for their common law claims and ten million dollars for their Spill Act claims. If the jury finds Citgo to be responsible for one-third of the total damages, Citgo would be liable for $103.3,3 million. Non-Settling Defendants would pay $210 million — the entire $10 million for the Spill Act claims and $200 million for their combined share of the common law claims. But they would receive a settlement credit of $23.25 million. In that scenario, Non-Settling Defendants would have no objection because the $23.25 million settlement credit exceeds the $10 million they would have to pay for the Spill Act claims.
If, however, the hypothetical is reversed, and the jury awards Plaintiffs ten million dollars for their common law claims and three hundred million dollars for their [263]*263Spill Act claims, damages for the Spill Act claims would far outweigh the settlement amount. Even though Citgo is found to be liable for one-third of the total damages, again $103.33 million, Non-Settling Defendants would have to pay $306.67 million dollars of the $310 million verdict. Even with the $23.23 million settlement credit, Non-Settling Defendants would be out of pocket $283.42 million dollars, which is far more than two-thirds of the total damages award.18
E. Plaintiffs’ Damages Calculations
On October 21, 2013, Plaintiffs published notice of the JCO in the New Jersey Register.19 Non-Settling Defendants then provided comments to the NJDEP. They objected to the JCO on the grounds that Plaintiffs had failed to disclose (1) their total alleged damages, (2) how they determined Citgo’s fair share of liability, and (3) how Citgo’s payment would be allocated among the four identified Citgo sites or any of the other sites.20
On December 20, 2013, Plaintiffs formally responded to the comments. First, Plaintiffs estimated their total damages for the 5,045 sites to be between $1.99 and $3.32 billion, plus out-of-pocket costs.21 To arrive at this estimate, Plaintiffs’ experts first calculated the average cost of restoration at the ten Plaintiff-selected sites to be $4,657,608 per site.22 Plaintiffs then multiplied that cost by 498 because — according to Plaintiffs’ recent groundwater sampling — 498 of the 5,045 sites contained MTBE above 700 parts per billion (“ppb”).23 Thus, damages at the 498 sites totaled approximately $2.32 billion.24 Plaintiffs then assigned an average damages value of $50,000 to each of the remaining 4,547 sites, resulting in an additional $227.35 million in damages.25 Plaintiffs further estimated that compensatory damages at these sites — outside of restoration costs — would total approximately $104 million.26 Finally, Plaintiffs assigned a twenty-five percent contingency factor to “account for uncertainty.”27
Second, Plaintiffs explained how they determined that $23.25 million represented Citgo’s fair share of liability. Plaintiffs found Citgo to be liable at only 128 of the 5,045 sites.28 They also considered Citgo’s limited business operations in New Jersey, the number of Citgo-branded sites in New Jersey, the quantity of MTBE-containing gasoline in New Jersey, Citgo’s potential defenses, Citgo’s willingness to settle ear[264]*264ly, the time-value of money, and the public interest in settlement.29 The final JCO was approved by the Commissioner of the NJDEP and the New Jersey Attorney General.30
III. APPLICABLE LAW31
Under New Jersey law, a court will not “second-guess those judgments of an administrative agency,” like the NJDEP, which “fall squarely within the agency’s expertise.”32 In fact, a court “will reverse the decision of an administrative agency only if it is arbitrary, capricious, or unreasonable, or if it is not supported by substantial credible evidence in the record as a whole.”33 Despite this deference, the court’s review of an agency decision is “not simply a pro forma exercise in which [the court] rubber stamp[s] findings that are not reasonably supported by the evidence.” 34
A court will approve an agency’s decision to settle as long as the settlement is “fair, reasonable, and adequate.”35 The standard to be applied “is not whether the settlement is one which the court itself might have fashioned, or considers as ideal, but whether the proposed decree is fair, reasonable, and faithful to the objective of the governing statute.”36 The court cannot simply “rubber stamp” a settlement based on the arguments and recommendations of counsel.37 Rather, “[t]o make this determination, the factual record before the district court must be sufficiently developed.” 38
When a court reviews a settlement that involves a Spill Act claim, it should look to the federal case law involving the Comprehensive Environmental Response, Compensation, and Liability Act (“CERLA”) for guidance.39 In fact, CERCLA has been called the “federal analogue to the Spill Act.”40 Under [265]*265CERCLA case law, “[a] court should approve a consent decree if it is fair, reasonable, and consistent with CERCLA’s goals.”41
“In evaluating the fairness of a consent decree, a court should assess both procedural and substantive considerations.”42 “Procedural fairness requires that settlement negotiations take place at arm’s length.”43 “A court should ‘look to the negotiation process and attempt to gauge its candor, openness and bargaining balance.’ ”44 “Substantive fairness requires that the terms of the consent decree are based on ‘comparative fault’ and apportion liability ‘according to rational estimates of the harm each party has caused.’ ”45 The court must “compare the proportion of total projected costs to be paid by the settlors with the proportion of liability attributable to them, and then to factor into the equation any reasonable discounts for litigation risks, time savings, and the like that may be justified.”46 “As long as the measure of comparative fault on which the settlement terms are based is not arbitrary, capricious, and devoid of a rational basis, the district court should uphold it.” 47
IY. DISCUSSION
Non-Settling Defendants do not dispute the procedural fairness of the JCO, which was the product of six months of arms-length negotiation.48 Instead, they object to its substantive fairness on the grounds that Plaintiffs have failed to adequately evaluate Citgo’s proportionate share of liability.49 Defendants object to Plaintiffs’ calculation of total damages and to Plaintiffs’ calculation of Citgo’s share of those damages. I will discuss each objection in turn.
A. Plaintiffs’ Calculation of Total Damages Is Arbitrary, Capricious, and Unreasonable
First, Non-Settling Defendants object that Plaintiffs’ total damages estimate of $1.99 to $8.32 billion is unreasonable because of its large range.50 Non-Settling Defendants assert that Citgo’s proportionate share of liability would be vastly different if the judgment were $1.99 billion than if it were $3.32 billion.51 Even though Citgo is identified at more Plaintiff-selected sites than any other defendant, Citgo’s payment of $23.25 million would amount to only 0.7% of a $3.32 billion judgment.52
Second, Non-Settling Defendants object to the methodology Plaintiffs used to calculate total damages. Plaintiffs’ calculation is based on a list of 498 sites that is [266]*266filled with inconsistencies.53 For example, the list includes only three of the ten Plaintiff-selected trial sites.54 But it also includes one Defendant-selected site where Plaintiffs’ experts declined to conduct discovery.55 Furthermore, the list includes only sites were MTBE was supposedly detected above 700 ppb during-the latest round of groundwater sampling, but Plaintiffs fail to disclose when that round occurred, the concentration of MTBE at any of the sites, or the pervasiveness of the contamination.56
Aside from the list, Plaintiffs fail to support their assumption that the restoration costs at the remaining 4,547 sites — where MTBE was detected below 700 ppb, if at all — would be $50,000 per site.57 Because Plaintiffs have not even calculated damages at nine of the ten Defendant-selected sites, Non-Settling Defendants assert that Plaintiffs offer no basis for this assumption.58
Third, Non-Settling Defendants argue that Plaintiffs ignore the damages calculations of one of their own experts, Kevin Boyle.59 Boyle used a different methodology than Plaintiffs’ other experts and calculated damages to be, on average, $7.32 million per site for the ten Plaintiff-selected sites.60 However, Plaintiffs’ other experts, Anthony Brown and Robert Uns-worth, calculated damages at those sites to be $4.66 million per site.61 Thus, Non-Settling Defendants argue that Plaintiffs have underestimated their damages by relying on Brown and Unsworth, rather than Boyle.62
Fourth, Non-Settling Defendants contend that Plaintiffs’ total damages estimate is based solely on “restoration costs” at the ten Plaintiff-selected trial sites.63 Plaintiffs have failed to account for all other categories of damages alleged in the Complaint, including (1) the costs of past and future MTBE testing of all public water supplies, (2) the costs of past and future treatment of all drinking water supplies containing detectable levels of MTBE, (3) the costs of past and future monitoring of other, waters to detect MTBE, (4) the past cleanup and removal costs, and (5) attorneys’ fees and costs.64 .
Plaintiffs respond that their total damages calculation is reasonable. Under the “bellwether” approach to discovery, the parties focused on nineteen trial sites as a [267]*267basis for their damages calculations.65 Plaintiffs contend that it was impractical and unnecessary to conduct site-specific discovery at all 5,045 sites, given that CERCLA case law does not require precise damages estimates as a condition to settlement.66
Next, Plaintiffs assert that they were allowed to ignore Boyle’s damages estimate because it merely provided a context for their claims.67 Because Plaintiffs do not intend to rely on his estimate at trial, they did not rely on it during the settlement process.68 Finally, Plaintiffs contend that it was unnecessary to consider all potential damages categories listed in the Complaint before estimating total damages.69 Plaintiffs explain that most of the categories are “complementary” to restoration damages.70 For example, if Defendants clean up and restore MTBE-contam-inated groundwater, the cost of testing and treating the drinking water is diminished.71 Similarly, if Defendants restore the groundwater, the cost of monitoring is reduced.72
As an initial matter, Plaintiffs are free to ignore Boyle’s damages calculations, especially since they will not rely on them at trial.73 Plaintiffs are also correct that they need not calculate damages that fall into overlapping or redundant categories for settlement purposes. At trial, Plaintiffs’ experts may not testify beyond the scope of their reports, which include calculations for restoration damages only.74 Because [268]*268they are precluded from introducing new calculations, Non-Settling Defendants will not be prejudiced.75
However, Non-Settling Defendants’ remaining objections raise serious concerns about the fairness and reasonableness of the JCO. While total damages estimates need not be precise, Plaintiffs’ estimate spans $1.33 billion. More troubling are the inconsistencies with Plaintiffs’ list of 498 sites, which serve as the basis for their damages estimate. Seven of the ten Plaintiff-selected sites are excluded from the list.76 Meanwhile, the list includes at least one site where Plaintiffs’ experts conducted no discovery.77 Plaintiffs offer no explanation for these inconsistencies, which raise doubts about the reliability of the entire list. Nor do they explain how they determined that the average restoration costs at the remaining 4,547 sites to be $50,000. Based on the record, it is unclear how many of these sites — if any — are contaminated at all. As a result, the $50,000 estimate appears arbitrary. With such an incomplete record, it is impossible for me to determine whether the settlement is fair and reasonable.
B. Plaintiffs’ Calculation of Citgo’s Share of Liability Is Arbitrary, Capricious, and Unreasonable
Non-Settling Defendants also contend that the $23.25 million settlement amount does not fairly or reasonably account for Citgo’s liability.78 First, they assert that Plaintiffs’ list of 128 sites associated with Citgo is inconsistent and unreliable.79 For example, the list includes only one of the four Plaintiff-selected sites associated with Citgo.80 As such, Plaintiffs assessed Cit-go’s liability without considering damages at the Skyline Service Station site, the HP Delta site, or the Five-Points site. This is especially problematic because Citgo’s liability at those three sites alone ranges from $16 million to $24.5 million according to Plaintiffs’ experts.81 Therefore, Citgo’s liability at just those three sites could potentially exceed the $23.25 million settlement amount.
Second, BP contends that Plaintiffs used inappropriate databases ta generate their 128 site list.82 Specifically, they relied on an Environmental Management System (“ÉMS”) database, which is designed to allow “mapping” of release sites, and summarize the site’s regulatory status, “discharge parameter,” “program interest,” and regulatory violations.83 The EMS database does not identify historic suppliers of gasoline to sites or link those suppliers with MTBE releases.84 Similarly, Plaintiffs’ HazSites database merely stores soil and groundwater laboratory test results.85 It does not store information that identifies parties responsible for MTBE releases [269]*269or identify the stations’ suppliers.86 BP argues that reliance on these databases caused deficiencies in Plaintiffs’ list.87 Plaintiffs should have prepared a complete list by examining the NJDEP’s site files and identifying MTBE releases and the suppliers during those release periods.88
Plaintiffs respond that they relied on “site-specific discovery” as well as information from the databases to create their list.89 Plaintiffs conducted discovery on the ten Plaintiff-selected sites and one Defendant-selected site. They then “ac-eount[ed] for such information on a statewide basis [ ] in the absence of site-specific discovery on the 5,000+ other sites” by considering Citgo’s “market share” for gasoline sold in New Jersey.90 Plaintiffs argue that this approach was a reasonable alternative to conducting site-specific discovery at all 5,045 sites.91
Next, Plaintiffs argue that Non-Settling Defendants improperly assumed that Citgo is one hundred percent liable at all 128 sites.92 In fact, many of the sites are associated with both Citgo and other defendants.93 For example, because BP and Citgo both supplied MTBE at the Five Points Site, Plaintiffs suggest that they should share liability there.94
Finally, Plaintiffs contend that the settlement amount is also based on relevant factors, such as the relatively small number of sites associated with Citgo, Citgo’s minor market share in New Jersey, Citgo’s willingness to settle early, the liability of remaining Defendants, litigation risks, the certainty provided in settlement, the time value of money, the public interest in settlement, and Citgo’s potential defenses.95 As for its potential defenses, Citgo argues that unlike other Defendants, it did not own, lease, or operate any service station in New Jersey.96 Instead, it merely supplied MTBE-containing gasoline that was refined by others.97
On the record before me, I cannot determine whether the $23.25 million settlement amount represents Citgo’s fair share of liability. Plaintiffs reached this amount based on a 128 site list that is replete with problems. First, the list omits three of the four Plaintiff-selected sites where Plaintiffs claim Citgo is liable. Because Citgo’s liability at these three sites alone could exceed $23.25 million, Plaintiffs have severely underestimated Citgo’s liability. Second, the list was generated from databases incapable of identifying historic sup[270]*270pliers of MTBE at the time of the releases. Plaintiffs attempt to compensate for their lack of site-specific knowledge by relying on Citgo’s reported market share of four percent. But as Citgo admits, if its share of liability was four percent of the total damages, it would be liable for $88 million.98
Plaintiffs insist that they gave Citgo a discount because “other defendants in addition to Citgo” were identified at some of the 128 sites.99 But they fail to identify which sites Citgo supplied at the time of the MTBE releases, which other Defendants supplied those sites at those times, or how to allocate liability among Defendants. Plaintiffs also claim that they considered the number of sites associated with Citgo, Citgo’s market share, Citgo’s willingness to. settle early, the liability of remaining defendants, litigation risks, the certainty provided in settlement, the time value of money, the public interest in settlement, and'Citgo’s arguments regarding its potential defenses.100 While these are all reasonable factors to consider, the Court cannot approve the settlement without first evaluating “the measure of comparative fault on which the settlement terms are based.”101 Without a reasonable measures of both total damages and Citgo’s share, I cannot determine whether the settlement is fair and reasonable.
C. Plaintiffs Cannot Justify the Settlement as an “Aggregate Settlement”
Finally, Plaintiffs insist that the Court can still approve their settlement as an “aggregate settlement,” even without site-specific information on over 5,000 sites.102 Plaintiffs note that this Court previously approved a multi-defendant, multi-site settlement as an “aggregate settlement” in In re MTBE.103 There, I found that “the estimate of each defendant’s share of liability is- not relevant where allocation of liability under the settlement was done on an aggregate and not on a case-by-case basis, and where any non-settling defendant that sustains an adverse judgment at trial will receive the benefit of the entire aggregate amount paid in settlement as a setoff.”104 Plaintiffs insist that the JCO is similar because Non-Settling Defendants-will be entitled to prove how the Citgo settlement proceeds should be allocated at trial.
In fact, the two settlements are dramatically different. The settlement in In re MTBE arose out of global settlement discussions that included approximately sev[271]*271enty percent of the named defendants.105 This Court held that plaintiffs did not need to estimate each settling defendant’s share of liability, as long as the aggregate settlement amount reasonably represented “the settling defendants’ estimated combined share of liability.”106 Here, however, Cit-go is the only settling defendant, and Plaintiffs are required to fairly estimate Citgo’s share.
Furthermore, in In re MTBE, plaintiffs used reasonable and reliable methods to calculate the settling defendants’ propor- ■ tionate share. For example, during settlement negotiations, plaintiffs provided detailed, site-specific explanations regarding their damages estimates.107 They also explained how their settlement proceeds would be allocated across the sites.108 In calculating total damages, Plaintiffs first relied on a study by the American Petroleum Institute (“API”) that calculated “high, low, and mean costs of treating wells contaminated with MTBE for a ten-year period.”109 They then tailored the API estimates to the specific contaminated wells in these cases, using “a standard linear regression analysis to derive the estimated API cost for treatment at each well’s flow rate.”110 Finally, plaintiffs assessed the characteristics of each well, and assigned each a “grade” based on a number of site-specific factors, including flow rate, level of detections, the length of time over which the detections occurred, how recently detections have occurred, and the relationship of the detections to applicable regulatory standards.111
Here, Plaintiffs have conducted no site-specific analysis beyond eleven trial sites. Instead, they created an unreliable list of 498 sites to determine total damages. Without conducting any analysis of the remaining 4,547 sites, they assigned a an arbitrary $50,000 damages value. Plaintiffs have sued Defendants at 5,045 sites about which little or nothing is known. The fact that Plaintiffs consider it “impractical” to conduct site-specific discovery at these sites does not excuse them from ensuring that the JCO is fair and reasonable.
V. CONCLUSION
For the foregoing reasons, Plaintiffs’ motion is DENIED. The Clerk of the Court is directed to close this motion (Doc. No. 344). Plaintiffs may resubmit the JCO for approval when they develop and present a more complete record supporting the settlement.
SO ORDERED.