Adair v. EQT Production Co.

320 F.R.D. 379

• Court: District Court, W.D. Virginia
• Decided: March 29, 2017
• Docket: Case No. 1:10CV00037, Case No. 1:10CV00041, Case No. 1:11CV00031, Case No. 1:10CV00059, Case No. 1:10CV00065
• Status: Published

Opinion

OPINION

James P. Jones, United States District Judge

Table of Contents

I. Background.. .387

II. The Parties’ Arguments.. .389

A. Hale, Addison, Adair, and Kiser.. .389

1. Proposed Class Definitions.. .389

2. Specific Legal Claims.. .390

3. Defendants’ Response.. .391

B. Adkins v. EQT. .. .393

1. Proposed Class Definitions.. .394

2. Specific Legal Claims.. .394

3. Defendants’ Response.. .395

III. Discussion.. .396

A. Hale, Addison, Adair, and Kiser.. .396

1. Numerosity, Typicality, and Adequacy. . .396

2. Ascertainability.. .399

3. Commonality and Predominance.. .400

4. Superiority.. .417

5. Fail-Safe Classes.. .420

6. Statutes of Limitations.. .420

7.Class Action Fairness Act.. .421

B. Adkins v. EQT. .. .422

1. Numerosity, Typicality, and Adequacy. . .422

2. Ascertainability.. .423

3. Commonality and Predominance.. .424

4. Superiority.. .428

5. Fail-Safe Class.. .429

IV.Conclusion.. .429

In these related cases, the plaintiffs’ latest motions for class certification are before the court, having been fully briefed and argued.¹ In summary, I grant certification, at least in part, for the Hale, Adair, and Adkins classes, and deny certification for the Addison and Kiser classes.

I. Background.

These cases have a long history—unusually long, at least for this court. As I earlier related,

The five cases involve two coalbed methane [gas] (“CBM”) producers, EQT Production Company (“EQT”) and CNX Gas Company LLC (“CNX”), with the Adair, Adkins, and Kiser cases concerning EQT and the Hale and Addison cases involving CNX. The earliest case— Adair—was filed in this court on June 15, 2010, and the other four following thereafter, with the last—Kiser—being filed on April 20, 2011. Numerous hearings have been held in the five cases, and dozens of orders and opinions entered, either by me or U.S. Magistrate Judge Pamela Meade Sargent, covering a wide range of procedural and substantive issues. After substantial briefing and argument, I certified classes in all of the eases on September 30, 2013. The defendants then sought interlocutory appeals of the certifications pursuant to Federal Rule of Civil Procedure 23(f). After briefing and argument, the court of appeals granted the appeals and remanded the cases by opinion dated August 19, 2014.

In its opinion, the court of appeals summarized the classes certified by this court as follows:

“Four of the five classes—-Adair, Addison, Hale, and Kiser—consist of persons who have never received CBM royalties for a CBM interest they claim to own. As defined by the district court, the classes include (1) all persons or their successors, (2) whom EQT or CNX have identified as being the owners of the gas estate in a tract underlying a CBM drilling unit, (3) whose interest in the CBM is ‘in conflict’ because a different person owns the coal estate in the same tract.

“The ownership classes can be further broken down. In two cases (the ‘force pooled’ classes)—Adair and Hale—the plaintiffs’ purported CBM interests have been force pooled by a [Virginia Gas and Oil] Board order.

“In the other two ownership cases (the ‘voluntary lease’ classes)—Kiser and Addison—the defendants entered voluntary lease arrangements with the putative class members. Nonetheless, the class members’ CBM interests have been subject to pooling, and their royalties have either been paid into Board escrow accounts or internally withheld by EQT and CNX.

“The primary object of the ownership classes is to obtain the release of es-crowed or suspended royalties. To that end, they seek a declaratory judgment that: (1) the ownership conflict EQT and CNX identified between gas estate owners and coal estate owners is ‘illusory’; (2) as gas estate owners, the class members are entitled to the CBM royalties withheld; and (3) any royalties held in escrow or internally suspended by EQT and CNX as a result of the ‘illusory’ ownership conflict must be paid to the class members.

[[Image here]]

“The fifth class—Adkins—is unique, as it consists of persons whose CBM ownership interest is not disputed. Instead, the putative class includes persons who have received a royalty from EQT at some point since January 1, 1995. The Adkins plaintiffs allege that EQT has systematically underpaid CBM royalties. The four other classes make similar claims against the defendants. Each of the classes seek a complete accounting of the royalties EQT and CNX have remitted to class members, paid into escrow, or internally suspended.”

Adair v. EQT Prod. Co., Nos. 1:10CV00041, 1:11CV00031, 2015 WL 505650, at *1-2 (W.D. Va. Feb. 6, 2015) (quoting EQT Prod. Co. v. Adair, 764 F.3d 347, 355 (4th Cir. 2014)) (footnotes omitted).

The court of appeals remanded the case and instructed this court to conduct “a more rigorous analysis as to whether the requirements for class certification have been satisfied.” EQT Prod. Co., 764 F.3d at 352. In the meantime, however, two events occurred that the plaintiffs contend has simplified the court’s task, As a result, the plaintiffs no longer seek a determination of CBM ownership. Instead, they seek an accounting of royalty payments by the defendants, as well as a determination of the propriety of certain practices relating to the calculation of those royalties.

The first of these simplifying events was the Virginia legislature’s enactment of House Bill 2058, which amended Virginia Code Ann. § 45.1-361.1 to direct the release of money held in CBM escrow accounts to the gas interest owner, absent a coal interest owner’s proceeding against or agreement with the gas interest owner. Because this amendment required operators to identify CBM gas owners by the end of 2015, the plaintiffs assert that concerns regarding the ascertainability of class members are now moot.

The second simplifying event was the issuance of an opinion by the Supreme Court of Virginia in Swords Creek Land Partnership v. Belcher, 288 Va. 206, 762 S.E.2d 570 (2014). Swords Creek affirmed that court’s earlier holding in Harrison-Wyatt, LLC v. Ratliff, 267 Va. 549, 593 S.E.2d 234 (2004), and clearly established that CBM is a mineral estate separate and distinct from the coal estate. Swords Creek Land P’ship, 762 S.E.2d at 572 (“CBM is not a constituent part of coal at any time but rather is a separate mineral estate”). As a result of this determination of Virginia law, the coal owners who previously laid claim to CBM royalties have relinquished those claims. The plaintiffs argue that these events have simplified both the nature of the primary dispute and the questions raised by the court of appeals regarding class certification.

II. The Parties’ Arguments.

A. Hale, Addison, Adair, and Kiser.

In light of these two events, the plaintiffs in Hale, Addison, Adair, and Kiser have framed their cases as actions for an accounting. The plaintiffs represent that after the classes have been certified, they will move for summary judgment as to questions of “pure law” regarding the liability of the defendant CBM operators. Once those questions have been answered, the plaintiffs say, the accounting can take place and damages can be calculated as to each subset of class members. In short, the plaintiffs believe this action should consist of three “phases” as follows:

Phase 1-Answer legal questions via summary judgment.

Phase 2-Perform requested accounting.

Phase 3-Calculate the plaintiffs’ damages. Such calculations will be performed as to specific subsets of class members, and in some instances may involve calculating damages for individual class members,

During oral argument on their motions for class certification, the plaintiffs argued that other courts have favored this bifurcated approach. See, e.g., McReynolds v. Merrill Lynch, Pierce, Fenner & Smith, Inc,, 672 F.3d 482, 488-89 (7th Cir. 2012) (allowing limited class action treatment regarding issue of whether employer’s practices had a disparate impact on African-American employees), abrogated on other grounds by Phillips v. Sheriff of Cook Cty., 828 F.3d 541, 559 (7th Cir. 2016).

1. Proposed Class Definitions.

The Hale and Adair plaintiffs propose nearly identical class definitions. Where the Hale proposed class refers to CNX, the Adair proposed class refers to EQT. In addition, both class definitions encompass gas claimants who received royalties between a given date and January 1, 2016. For the Hale class, that date is September 23, 2010; for the Adair class, that date is June 15, 2010. In all other respects, these plaintiffs’ class definitions are identical:

All gas claimants who were identified by [CNX/EQT] in filings with the Board as “unleased” owners of gas estate interests and for whom [CNX/EQT] has applied, as of the later of the date of this Court’s class certification order or the resolution of an interlocutory appeal of such order, pursuant to Virginia Code § 45.1-361.22:2(A), for the release of funds held in escrow or internally, and all such gas claimants who have received distributions from escrow or directly from [CNX/EQT] as a result of a judicial determination of ownership or agreement between [September 23, 2010/ June 15, 2010] and January 1, 2016. “Gas claimants” is defined by Virginia Code § 45.1-361.1. The Class excludes (a) the Defendant, (b) any person who serves as a judge in this civil action and his/her spouse, (c) any individuals who have received a Court-supervised accounting of [CNX’s/EQT’s] payments into escrow or internal suspense, and (d) any person who operates a gas well in Virginia and any person who holds a working interest in a well operated by [CNX/EQT] in Virginia.

Hale v. CNX Gas Co., 1:10CV00059, Pl.’s Revised Class Definitions 1-2, ECF No. 481; Adair v. EQT Prod. Co., 1:10CV00037, Pl.’s Revised Class Definitions 1-2, ECF No. 576,

The Addison and Kiser plaintiffs also propose nearly identical class definitions. Where the Addison proposed class refers to CNX, the Kiser proposed class refers to EQT. In addition, both class definitions encompass gas claimants who received royalties between a given date and January 1, 2016. For the Addison class, that date is November 9, 2010; for the Kiser class, that date is April 20, 2011. In all other respects, these plaintiffs’ class definitions are identical:

All gas claimants who voluntarily leased gas estate interests and for whom [CNX/EQT] has applied, as of the later of the date of this Court’s class certification order or the resolution of an interlocutory appeal of such order, pursuant to Virginia Code § 45.1-361.22:2(A) or (C), for the release of funds held in escrow or held internally, and all such gas claimants who have received distributions from escrow or directly from [CNX/EQT] as a result of a judicial determination of ownership or agreement between [November 9, 2010/ April 20, 2011] and January 1, 2016. “Gas claimants” is defined by Virginia Code § 46.1-361.1. The Class excludes (a) the Defendant, (b) any person who serves as a judge in this civil action and his/her spouse, (c) any individuals who have received a Court-supervised accounting of [CNX’s/EQT’s] payments into escrow or internal suspense, and (d) any person who operates a gas well in Virginia and any person who holds a working interest in a well operated by [CNX/EQT] in Virginia.

Addison v. CNX Gas Co., 1:10CV00065, Pl.’s Revised Class Definitions 2, ECF No. 399; Kiser v. EQT Prod. Co., 1:11CV00031, Pl.’s Revised Class Definitions 2, ECF No. 360.

2. Specific Legal Claims.

As noted, the plaintiffs seek, first and foremost, an accounting. During oral argument, the plaintiffs asserted that they are entitled to this accounting as a matter of right because the defendants owe them a fiduciary duty. The plaintiffs characterize the accounting question as being one of six questions of law or fact that are common to the class.

In Virginia, one who is owed a fiduciary duty is entitled to an accounting if there is an allegation that the fiduciary is not living up to its duty. See Va. Code Ann. § 8.01-31 (“An accounting in equity may be had against any fiduciary ... for receiving more than comes to his just share or proportion.”). An action for an accounting must normally be predicated on some wrong. Accordingly, at oral argument, the plaintiffs acknowledged that an accounting would not be available if, at the summary judgment phase, the court decided the five other common questions in favor of the defendants. According to the plaintiffs, those common questions are as follows.

First, whether the defendants’ cost deductions are excessive. The plaintiffs argue both that the costs are excessive and that the excessive nature of the deductions can be evidenced by examining the defendants’ common practices. These alleged practices include CNX’s decision to fabricate gathering rates by manipulating accounting data, rather than basing gathering rates on postpro-duction costs. The plaintiffs also claim that CNX improperly charged the gas owners for the depreciation of capital assets and facilities and that EQT improperly charged the gas owners for gathering, compression, property taxes, depreciation, and fuel costs. They further argue that the defendants improperly passed selling, general, and administrative charges to them. Because the defendants’ respective courses of conduct were the same for all class members, the plaintiffs assert, this issue can be adjudicated as to the entire class.

Second, whether the defendants paid royalties based on improperly low prices. The plaintiffs argue that the defendants used swap contracts and other financial instruments to pay royalties based on prices lower than those realized by the defendants. The plaintiffs further argue that the defendants paid royalties based on prices below the published index prices. Again, they assert that these practices affected the royalties of all class members.

Third, whether the defendants improperly delayed making payments into escrow. The plaintiffs argue that the defendants have consistently made late payments and that, in some cases, payments were delayed by years. They also assert that this practice raises further common questions, such as whether these late payments provide the plaintiffs with a cause of action and, if so, the extent to which the plaintiffs are entitled to damages.

Fourth, whether the defendants may permissibly deduct severance taxes from the lessors’ royalties. The plaintiffs in the Hale and Adair classes argue that this practice is unlawful. See Hershey v. ExxonMobil Oil Corp., No. 07-1300-JTM, 2011 WL 1234883, at *5 (D. Kan. Mar. 31, 2011) (holding that whether lessee made proper deductions for severance taxes was a common question that justified certification). Importantly, such taxes appear to be deducted from the royalties of all class members.

Fifth, whether the defendants improperly force-pooled, drilled, completed, and produced CBM wells before a pooling order had been entered. This claim applies only to some of the Hale putative class members.²

3. Defendants’ Response.

a.Ascertainability.

While the defendants’ briefs address ascer-tainability, their discussions contemplate only the classes that the plaintiffs originally named in their motions to certify the classes. At oral argument, the plaintiffs proposed new class definitions that narrow the classes and appears to make them more ascertainable. As a result, many of the arguments made in the defendants’ briefs no longer apply. Nonetheless, the defendants consistently argue that not all gas claimants will have been identified by January 1, 2016.³ They cite the revised version of Virginia Code Ann. § 45.1-361.22:2(A), which allows the Virginia Gas and Oil Board to extend the time by which gas well operators must file their applications for payment and delay the release of funds in escrow. The defendants further note that the plaintiffs make no attempt to explain how unidentified gas claimants will be ascertained. The plaintiffs suggest that those claimants who have not yet been identified will identify themselves, but the defendants argue that class certification cannot be based on such speculation. See Marcus v. BMW of N. Am., LLC, 687 F.3d 583, 592-93 (3d Cir. 2012) (“Many courts and commentators have recognized that an essential prerequisite of a class action, at least with respect to actions under Rule 23(b)(3), is that the class must be currently and readily ascertainable based on objective criteria.”

b. Commonality.

The defendants also argue that the plaintiffs’ claims do not present common questions suitable for class adjudication because they “do not generate common answers apt to drive resolution of this case.” Kiser v. EQT Prod. Co., 1:11CV00031, Def.’s Mem. Opp’n Class Certification 6, ECF No. 352. They further note that the question of whether an accounting is appropriate cannot be considered a common question for purposes of class certification unless and until the court has found some wrongdoing by the defendants.⁴ In addition, EQT argues that the questions regarding its deduction of severance taxes are not of sufficient importance to justify certification of the class due to the relatively low dollar amounts at issue. Finally, both EQT and CNX allege that, contrary to the plaintiffs’ allegations, their royalty payment practices are not uniform across the classes and that, as a result, questions regarding those payment practices are neither common questions nor have common answers.⁵

c. Predominance.

The plaintiffs argue that while individual determinations of damages will eventually become necessary, the “pure legal questions” presented above represent the fundamental issues in these cases. They assert that once these questions are resolved, damages will largely be calculated using basic formulas. Some of the legal questions apply only to certain subgroups within the proposed classes, but the plaintiffs argue that such a bifurcation or division of pai’ties does not detract from the fact that the above questions are the predominate issues in this matter. See McReynolds, 672 F.3d at 491 (“If there are genuinely common issues ... identical across all the claimants, ... the accuracy of the resolution of which is unlikely to be enhanced by repeated proceedings, then it makes good sense, especially when the class is large, to resolve those issues in one fell swoop while leaving the remaining, claimant-specific issues to individual follow-on proceedings.” (citation omitted)).

The plaintiffs assert that their approach is consistent with the Fourth Circuit’s directive to “take full advantage” of Rule 23(c)(4), which “permit[s] class treatment of separate issues in the case.” Gunnells v. Healthplan Servs., Inc., 348 F.3d 417, 441 (4th Cir. 2003) (internal quotation marks and citation omitted). “Neither Rule 23 nor any gloss that decided cases have added to it requires that every question be common. It is routine in class actions to have a final phase in which individualized proof must be submitted.” Suchanek v. Sturm Foods, Inc., 764 F.3d 750, 756 (7th Cir. 2014).

The defendants assert that an analysis of the above questions does not show that class-wide issues predominate. While the plaintiffs frame the central issue in this case as being whether they are entitled to an accounting, the defendants frame the central issue as being whether they, the defendants, underpaid royalties. According to the defendants, the plaintiffs have not set forth class-wide questions capable of addressing this central issue that also predominate over individual issues. See Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 350, 131 S.Ct. 2541, 180 L.Ed.2d 374 (2011) (“What matters to class certification ... is not the raising of common ‘questions’—even in droves—but, rather the capacity of a elasswide proceeding to generate common answers apt to drive the resolution of the litigation. Dissimilarities within the proposed class are what have the potential to impede the generation of common answers.” (citation omitted)). Citing Dukes, the defendants contend that, by focusing on the accounting action, the plaintiffs are stating an overly-broad, overly-inclusive issue that does not truly address the subject matter of this case and, therefore, does not satisfy the predominance requirement.

The defendants argue both that individual issues predominate as to each of the five legal questions posed by the plaintiffs and that the plaintiffs have not offered the “evi-dentiary proof’ necessary to satisfy the predominance requirement. Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp’n Class Certification 36, ECF No. 463; see also Comcast Corp. v. Behrend, 569 U.S. 27, 133 S.Ct. 1426, 1428, 185 L.Ed.2d 515 (2013). They address each of the plaintiffs’ stated legal questions in turn.

First, the defendants argue that individual issues predominate over the question of whether their cost deductions are “unreasonable.” Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp’n Class Certification 23, ECF No. 463. They assert that the question of whether their payment practices violated the duties they owed will vary from class member to class member. In support of this contention, the defendants point to the Fourth Circuit’s finding that the deductions taken from the putative class members were not uniform because the defendants took different deductions depending on where they sold the CBM. The Fourth Circuit further noted that deduction calculations can vary, both between different wells and within the same well over time. See EQT Prod. Co., 764 F.3d at 367. Determining the cost deduction for any given class member depends on a number of factors, including the applicable time frame, well location, place of sale, gathering rates, and the amount of gas transported. The defendants contend that the plaintiffs have offered no specific formula, calculation method, or expert analysis that can uniformly evaluate these factors.

Second, the defendants argue that individual issues predominate over any common question regarding the reasonableness of the prices used to calculate royalty payments. They assert that, contrary to the plaintiffs’ claim, they are not required to pay royalties at the “highest price obtainable” and that instead, they are required only to act as a “reasonably prudent operator,” or to pay royalties based on a reasonable price. Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp’n Class Certification 28, ECF No. 463. The defendants further argue that whether a price is reasonable will depend on numerous individualized facts.

Third, the defendants argue that individual issues predominate over the question of whether they delayed making payments into escrow. Again, they first note the plaintiffs’ concession that this practice has not affected all class members. The defendants contend that in order to determine whether late payments have caused injury, the court will need to review the timeliness of every payment as to every class member, based on the date of each specific pooling order. This analysis will cover a twenty-two year span. In short, they say, this fact-intensive process alone will constitute a much larger portion of this action than any common issue.

Fourth, the defendants contend that the question of whether they may deduct severance taxes from class members’ royalties is not of “sufficient importance” to justify certifying the class. Kiser v. EQT Prod. Co., 1:11CV00031, Def.’s Mem. Opp’n Class Certification 7, ECF No. 352. They appear to concede that the severance tax issue presents a common question, but they also note that any damages that might result from this issue would be relatively small.⁶ The defendants thus argue that “the costs associated with litigating [this] claim threaten to outweigh the potential recovery” and that the severance tax issue is not of central importance. Id. at 8.

Fifth, the defendants argue that individual issues predominate over any common question regarding whether they force-pooled and produced gas from CBM wells before a pooling order was entered. They first note that this issue does not apply to all putative class members and that individual review would therefore be necessary to determine whether the putative class members were actually injured. The defendants also contend that individual issues will be raised by affirmative defenses. Whether or not the defendants committed any wrongdoing, they contend, will depend on the specific permits and facts that relate to their conduct at a given well.

At bottom, the defendants contend that the individual analyses that must take place as to each putative class member and each issue constitute a greater proportion of this action than the common “pure legal questions” cited by the plaintiffs. “[I]ndividu-alized damage determinations cut against class certification under Rule 23(b)(3).” Ward v. Dixie Nat’l Life Ins. Co., 595 F.3d 164, 180 (4th Cir. 2010). The defendants also note the plaintiffs’ failure to propose any expert or scientific model that might be used to calculate damages, which the defendants point to as evidence of how difficult it would be to adjudicate each individual claim.

Sixth, the defendants assert that individual trials are superior to a class action suit in this ease, that there is sufficient financial incentive for the plaintiffs to litigate these trials individually, and that mechanisms provided by Virginia’s regulatory scheme can further help resolve these matters. Thus, the defendants argue that this court should abstain from hearing these cases under the doctrine of Burford v. Sun Oil Co., 319 U.S. 315, 63 S.Ct. 1098, 87 L.Ed. 1424 (1943). See also Johnson v. Collins Entm’t Co., 199 F.3d 710, 719 (4th Cir. 1999) (“Basie abstention doctrine requires federal courts to avoid interference with a state’s administration of its own affairs.”).

B. Adkins v. EQT.

Unlike the plaintiffs in Hale, Adair, Kiser, and Addison, the plaintiff in Adkins does not seek an accounting. Instead, Adkins raises claims of breach of contract and conversion.

1. Proposed Class Definitions.

Adkins proposes the following class definition:

Each person to whom EQT Production Company (‘EQT’) has paid since June 8, 2005, or is currently paying royalties under a lease(s) on gas produced by EQT from the Nora Field in the Commonwealth of Virginia and whose leases do not contain language expressly authorizing the lessee to deduct or expressly precluding the lessee from deducting the cost of gathering, treating, compression, dehydration, processing, and/or transportation when calculating royalty payments, according to business records maintained by EQT.

The Class excludes (a) EQT; (b) the federal government; (c) any person who serves as a judge in this civil action and his/her spouse; (d) any person who operates a gas well in Virginia; and (e) any person who holds a working interest ownership in a well operated by EQT in Virginia.

Adkins v. EQT Prod. Co., 1:10CV00041, PL’s Mem. Supp. Class Certification 7-8, ECF No. 357.

Adkins argues that this revised class definition is more precise and, thus, cures the deficiencies noted by the court of appeals.

The plaintiff in Adkins has restricted the scope of her case, such that she now raises only two claims: (1) breach of contract and (2) conversion. These claims stem from deductions that EQT has allegedly taken from the plaintiffs royalty payments. Adkins states that such deductions are for the purpose of making the gas marketable and argues that her royalty payments cannot properly be reduced for this reason. She also asserts that EQT has applied a “uniform accounting methodology” that has led to the systematic underpayment of royalties to members of the putative class. Id. at 6.

The Adkins plaintiff asserts that application of the First Marketable Product Rule to the putative class members’ leases constitutes a question common to the class. In its opinion, the Fourth Circuit expressed concerns about the commonality of the class based on variations among the gas owners’ leases with the defendant, and it “invited plaintiffs to show that ‘there are a limited number of lease forms, such that the validity of the defendants’ conduct can be assessed on a subclass basis.’” Id. at 10-11 (quoting EQT Prod. Co., 764 F.3d at 369). In response to this invitation, plaintiffs counsel reviewed many of the leases in the class. Although some leases require payment of royalties based on proceeds and others require payment based on market value—a variation noted by the Fourth Circuit—Adkins argues that this variation does not affect the question of whether EQT violated the First Marketable Product Rule, and thus does not preclude a finding of commonality. In addition, although some leases specify that the price of CBM must be determined at the gas well and others permit calculation at the point of sale—another variation noted by the Fourth Circuit—Adkins argues that this variation is similarly immaterial because EQT admits in its discovery responses that the gas is being sold at the well.

At bottom, plaintiffs counsel state that their lease review did not reveal any material lease term that would affect the common question of whether EQT violated the First Marketable Product Rule. They now argue that this review, along with their new proposed class definition, establishes commonality, and they identify three common questions in this matter, as follows.

First, what deductions EQT may take from the putative class members’ royalties.

Second, whether EQT has obtained the highest possible price for the CBM, in accordance with its duty to market the gas. Adkins asserts that this is a common question because there is no evidence that the prices EQT uses to calculate royalties vary among the putative class members.

Third, whether EQT’s gas is first marketable when it has entered an interstate pipeline. Adkins asserts that the question of whether the First Marketable Product Rule applies turns on the answer to this question.

Adkins argues that in order to receive a favorable adjudication on these issues, she will need to show that EQT’s royalty calculations, which are based on uniform sales prices and affect all class members, constitute breaches of the class members’ leases by resulting in underpayment of royalties. While there may be individual differences in the amounts of damages, Adkins argues that under the limited class definition proposed, the above common questions predominate over any such individual issues.

a. Commonality and Typicality.

The defendant asks the court to assess the elements of the claims made in Adkins’ case. As the plaintiff concedes, the ease “turns on the application of [the first marketable product] rule to all Class members.” Id. at 2. As such, the defendant suggests, the court must rule on the merits of the claims—and specifically the application of the First Marketable Product Rule—in order to decide the issue of class certification.

The defendant contends that the version of the First Marketable Product Rule on which the case turns is an “extreme version of the rule” that has not been adopted in Virginia. Id. Def.’s Mem. Opp’n Class Certification 11, EOF No. 361. As a result, it asserts, the class should not be certified, because the plaintiff will be unable to show that there is a common question as to this issue.

Furthermore, the defendant argues, even if the First Marketable Product Rule does apply, and the plaintiff can show that there is a common question—that is, whether EQT’s gas is first marketable when it enters an interstate pipeline—the court should deny class certification because the plaintiff cannot show that this question has a common answer. According to the defendant, in order for this question to have a common answer, “all of the gas from all the wells must be marketable at the same location all of the time.” Id. at 29. Under Virginia law, however, this is not necessarily the case, and the question of marketability will therefore vary on a well-by-well basis. As a result, the defendant asserts, the plaintiff cannot show commonality on the issue of gas marketability.

Next, the defendant contends that the plaintiffs newly proposed class definition does not remedy the problems created by variations among the class members’ leases. It notes that although the plaintiffs expert identified twenty standard lease forms during her review, the review encompassed less than half of the actual leases at issue. The defendant also references language from individual leases in an effort to show how that language varies from lease to lease. Ultimately, the defendant asserts, the lease variations preclude the plaintiff from fulfilling the commonality and predominance class certification requirements.

Citing Wallace B. Roderick Revocable Living Trust v. XTO Energy, Inc., 725 F.3d 1213 (10th Cir. 2013), the defendant argues that the plaintiff, in order to prevail, must prove that no single lease in the class negates the First Marketable Products Rule. It further asserts that, contrary to the claims of the plaintiffs expert, “nearly all” of the leases in the class contain language negating the First Marketable Product Rule. Adkins v. EQT Prod. Co., 1:10CV00041, Def.’s Mem. Opp’n Class Certification 20, ECF No. 361. In support of this assertion, and in an attempt to undermine the plaintiffs expert, the defendant calls attention to specific areas of contradictory language in the leases and disputes the admissibility of the plaintiffs expert’s declaration. It also argues that Adkins’ claims are not typical of the proposed class because her particular leases correspond to only two of the twenty lease forms identified by her expert.

b. Predominance.

EQT contends that Adkins cannot fulfill the predominance requirement because of the myriad of individual issues in this case. It notes that because the course of performance between two parties controls the meaning of the putative class members’ leases, such course of performance must be analyzed with regard to each individual lease. There are thousands of leases at issue in this case. In addition to undermining any showing of commonality, the defendant asserts, the necessary individual analyses will predominate over any common questions regarding the meaning of the leases. To support its argument, the defendant provides several specific examples of individual courses of performance and asserts that these different courses of performance must result in different interpretations for each individual lease, thus affecting the outcome of each class member’s claim. In addition, the defendant asserts that individual issues regarding damages will predominate over any common questions that might be raised in this case.

c. “Fail-Safe” Class and State Interests.

Finally, EQT characterizes the plaintiffs proposed class as an “impermissible fail-safe” class that violates Rule 23(c)(3)’s requirement that there be a “possibility of an adverse judgment against the class.” Id. at 33 (citing Randleman v. Fid. Nat’l Title Ins. Co., 646 F.3d 347, 352 (6th Cir. 2011)). It also argues that the involvement of “substantial state interests” and questions of state law that have never been addressed by a Virginia court support federal abstention in this case. Id. at 33-34.

III. Discussion.

1. Numerosity, Typicality, and Adequacy.

a. Numerosity.

Rule 23(a) requires a prospective class to comply with certain prerequisites, including numerosity, commonality, typicality, and adequacy of representation.⁷ Numer-osity requires that a class be so large that “joinder of all members is impracticable,” Fed. R. Civ. P. 23(a)(1). I previously found that the plaintiffs satisfied the numerosity requirement, and the Fourth Circuit did not address this issue on appeal. Although the plaintiffs have since proposed new class definitions, I do not believe that these revisions require a different finding on this issue. The plaintiffs plead that CNX and EQT have 529 and 375 units in escrow, respectively, “with multiple claimants in each unit.” Hale v. CNX Gas Co., 1:10CV00059, PL’s Mem. Supp. Class Certification 27, ECF No. 459 (for the Hale and Addison plaintiffs); Kiser v. EQT Prod. Co., 1:11CV00031, Pl.’s Mem. Supp. Class Certification 20, ECF No. 350 (for the Adair and Kiser plaintiffs). These alleged numbers, which the defendants do not dispute, would certainly result in classes large enough to make joinder of all members impractical. See, e.g., Lilly v. Harris-Teetar Supermarket, 720 F.2d 326, 333 (4th Cir. 1983) (finding that a class of 229 was “easily enough” to fulfill the numerosity requirement). I therefore find that the plaintiffs have satisfied the numerosity requirement of Rule 23(a)(1).⁸

b. Typicality and Adequacy.

Typicality requires that the claims raised by the class representative be “typical of the claims ... of the class.” Fed. R. Civ. P. 23(a)(3). “‘The premise of the typicality requirement is simply stated: as goes the claim of the named plaintiff, so go the claims of the class.’ ” Broussard v. Meineke Discount Muffler Shops, Inc., 155 F.3d 331, 340 (4th Cir. 1998) (quoting Sprague v. Gen. Motors Corp., 133 F.3d 388, 399 (6th Cir. 1998)). In order for a class representative’s claims to be considered “typical” of the class, his “interest in prosecuting his own case must simultaneously tend to advance the interests of the absent class members.” Deiter v. Microsoft Corp., 436 F.3d 461, 466 (4th Cir. 2006). However, the claims of the plaintiff and the other class members need not be “perfectly identical or perfectly aligned.” Id. at 467. Instead, a typicality analysis requires the court to examine the elements of the claims, examine the facts on which the plaintiff would rely to prove those claims, and determine “the extent to which those facts would also prove the claims of the absent class members.” Id.

Adequacy of representation requires that a proposed class representative will “fairly and adequately protect the interests of the class.” Fed. R. Civ. P. 23(a)(4). The primary purpose of the adequacy requirement is to “uncover conflicts of interest between named parties and the class they seek to represent. ‘[A] class representative must be part of the class and “possess the same interest and suffer the same injury” as the class members.’” Amchem Prods., Inc., 521 U.S. at 625-26, 117 S.Ct. 2231 (citations omitted). Because these requirements “tend to merge,” I will address them both under the umbrella of the typicality requirement.⁹ Broussard, 155 F.3d at 337 (quoting Gen. Tel. Co. of Sw. v. Falcon, 457 U.S. 147, 157 n.13, 102 S.Ct. 2364, 72 L.Ed.2d 740 (1982)).

i. The Addison and Kiser Classes.

The Addison and Kiser plaintiffs both assert that they satisfy the typicality and adequacy of representation requirements. See Addison v. CNX Gas Co., 1:10CV00065, Pl.’s Mem. Supp. Class Certification 29-30, ECF No. 377; Kiser v. EQT Prod. Co., 1:11CV00081, Pl.’s Mem. Supp. Class Certification 22-23, ECF No. 350. CNX, the defendant in the Addison case, has asserted that the Addison plaintiff has failed to satisfy these requirements. Addison v. CNX Gas Co., 1:10CV00065, Def.’s Mem. Opp’n Class Certification 19 n.19, ECF No. 381. EQT, the defendant in the Kiser case, has not addressed the issue. However, as I explain below, see infra at III.A.3, neither the Addison nor the Kiser class has fulfilled the commonality and predominance requirements. Because I deny certification of these classes on this ground, and because the commonality requirement typically merges with the typicality and adequacy requirements, I do not find it necessary to address the question of whether these plaintiffs have satisfied the typicality and adequacy requirements.

ii. The Hale and Adair Classes.

The Hale and Adair plaintiffs both assert that they satisfy the typicality and adequacy requirements because “[ljike the rest of the Class members they seek to represent, they stand to receive escrowed and suspended funds, and share Class members’ interests in being paid properly now and in the future.” Hale v. CNX Gas Co., 1:10CV00059, Pl.’s Mem. Supp. Class Certification 29, ECF No. 459; Adair v. EQT Prod. Co., 1:10CV00037, PL’s Mem, Supp. Class Certification 22, ECF No. 566. Because the Hale and Adair plaintiffs raise the same claims—breach of fiduciary duty, conversion, and unjust enrichment— I will address them together.¹⁰

A claim for breach of fiduciary duty requires proof of (1) the existence of a fiduciary relationship, (2) breach of a duty arising from relationship, and (3) damages. See Carstensen v. Chrisland Corp., 247 Va. 433, 442 S.E.2d 660, 666 (1994). A fiduciary relationship exists where “special confidence has been reposed in one who in equity and good conscience is bound to act in good faith and with due regard for the interests of the one reposing the confidence.” H-B Ltd. P’ship v. Wimmer, 220 Va. 176, 257 S.E.2d 770, 773 (1979). In other words, “[a] fiduciary owes total fidelity to the interests of his principal” and “may engage in no self-dealing which may have any adverse effect on the interests of his principal.” State Farm Mut. Auto. Ins. Co. v. Floyd, 235 Va. 136, 366 S.E.2d 93, 97 (1988). Fiduciary relationships can arise by contractual provision, common law, or statute. See, e.g., Remora Invs., L.L.C. v. Orr, 277 Va. 316, 673 S.E.2d 845, (2009) (statute); Augusta Mut. Ins. Co. v. Mason, 274 Va. 199, 645 S.E.2d 290, 295 (2007) (contract); Williams v. Dominion Tech. Partners, L.L.C., 265 Va. 280, 576 S.E.2d 752, 757 (2003) (common law).

Here, the plaintiffs allege that the fiduciary relationship between the force-pooled gas owners and the defendants arose in at least one of three ways: (1) by virtue of the defendants’ “status as the Board-appointed operator of CBM Units created through ... forced pooling”; (2) by virtue of the defendants’ “control over and handling ... of CBM production and sales proceeds for the benefit of’ the gas owners; and/or (3) by virtue of the defendants’ “undertakings to act as agent and/or joint venturer ... for the benefit of[] other [gas] interest owners.” Hale v. CNX Gas Co., 1:10CV00059, PL’s Am. Compl. ¶ 140, ECF No. 166; Adair v. EQT Prod. Co., 1:10CV00037, Pl.’s Am. Compl. ¶ 131, ECF No. 330. Because the members of the putative Hale and Adair classes are all gas owners whose interests are governed by the applicable pooling orders, they are all subject to the same entitlements. This means that any evidence Hale and Adair might use to prove the existence of the fiduciary relationships would necessarily prove the existence of that relationship as to all members of the classes as well.¹¹ Likewise, because the defendants engaged in certain practices common to all putative class members, see infra at III.A.3, the evidence adduced by Hale and Adair to prove the existence of these common practices, and any breach of fiduciary duty arising from them, would likewise apply to the claims of the rest of the class. I therefore find that the Hale and Adair plaintiffs have fulfilled the adequacy and typicality requirements as to their claims for breach of fiduciary duty.

A claim for conversion requires proof of an “act of dominion wrongfully exerted over property in denial of, or inconsistent with, the owner’s rights.” Simmons v. Miller, 261 Va. 561, 544 S.E.2d 666, 679 (2001). A defendant is liable for conversion where its conduct breaches a duty, or infringes on an owner’s rights, arising under the common law. Condo. Servs., Inc. v. First Owners’ Ass’n of Forty Six Hundred Condo., Inc., 281 Va. 561, 709 S.E.2d 163, 171 (2011). All gas owners possess the same rights of ownership under the common law. Although the putative class members’ ownership interests are not all within the same gas field, they are governed by pooling orders that appear to be substantially identical. The facts on which Hale and Adair would rely to prove their entitlements and the defendants’ common practices and, therefore, their claims for conversion would thus be either identical or substantially similar to the facts needed to prove the claims of the class members as well. Moreover, as I note above, typicality does not require that the class representative’s and class members’ claims be “perfectly identical.” Deiter, 436 F.3d at 467. I therefore find that the Hale and Adair plaintiffs have fulfilled the adequacy and typicality requirements as to their claims for conversion.

A claim for unjust enrichment— which amounts to a claim under a theory of implied contract—requires proof (1) that the plaintiff “conferred a benefit on” the defendant; (2) that the defendant “knew of the benefit and should reasonably have expected to repay” the plaintiff; and (3) that the defendant “accepted or retained the benefit without paying for its value.” Schmidt v. Household Fin. Corp., II, 276 Va. 108, 661 S.E.2d 834, 838 (2008). Again, because these plaintiffs’ and class members’ interests are governed by pooling orders, they are subject to the same entitlements. Hale and Adair would need to use their pooling orders, along with other evidence, to prove both that they conferred a benefit on the defendants and that the defendants knew of the benefit and should have expected to repay them. Because their respective class members’ interests are governed by the same pooling orders, the same evidence would likewise prove these elements for the other members of the classes. Furthermore, Hale’s and Adair’s evidence of the defendants’ common practices, which by definition are the same across the classes, would similarly prove that the defendants accepted the benefits without fully paying for them—that is, that the defendants underpaid royalties—for all members of the classes. I therefore find that the Hale and Adair plaintiffs have fulfilled the adequacy and typicality requirements as to their claims for unjust enrichment.

CNX contends that the Hale plaintiff cannot fulfill the adequacy and typicality requirements for two reasons.¹² First, it asserts that because Hale “ha[s] no judicial determination of ownership” over his gas interest, he is “subject to standing challenges that will not be faced by class members who already have received determinations of ownership.” Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp’n Class Certification 19 n.19, ECF No. 463. Second, it asserts that the Hale plaintiff cannot fulfill the typicality requirement because “[t]he costs and deductions relevant to each class member’s claims vary over time and by unit.” Id¹³

These arguments do not preclude a finding of typicality and adequacy. The specific costs and deductions associated with class members’ gas interests do not alter whether CNX’s conduct was improper. The fact that class members may need to adduce individual evidence of damages does not preclude a finding of typicality. Indeed, were this the case, no class could ever be certified where the class members sustained different amounts of damages; the law clearly states otherwise. See, e.g., Stillmock v. Weis Mkts., Inc., 385 Fed.Appx. 267, 273 (4th Cir. 2010) (unpublished) (holding that individual damages issues do not necessarily defeat class certification). It is sufficient that the facts on which Hale will rely to prove his entitlements under the pooling orders and common law, and the facts on which he will rely to prove that CNX’s common practices breached those entitlements, will tend to likewise prove the claims of his fellow class members.

In addition, I decline to alter my findings on the basis of a potential, speculative argument regarding Hale’s standing. This issue has not been briefed; CNX merely suggests in a footnote that arguments against Hale’s standing “could” exist, Hale v. CNX Gas Co., 11:10CV00059, Def.’s Mem. Opp’n Class Certification 19 n.19, ECF No. 463. “Standing is determined at the commencement of a lawsuit.” Pashby v. Delia, 709 F.3d 307, 316 (4th Cir. 2013). Although subsequent events can moot a plaintiffs claim, id. an argument as such is not properly raised in a footnote in a brief opposing class certification. I therefore conclude that Hale and Adair satisfy the Rule 23(a) typicality and adequacy of representation requirements.

2. Ascertainability.

Classes certified in accordance with Federal Rule of Civil Procedure 23 must meet the “implicit threshold requirement that the members of [the] proposed class be ‘readily identifiable.’” EQT Prod. Co., 764 F.3d at 358 (quoting Hammond v. Powell, 462 F.2d 1053, 1055 (4th Cir. 1972)). The Fourth Circuit concluded that the ascertaina-bility requirement had not been met in the present cases because the classes were “defined to include both former and current gas estate owners,” and many gas estate owners were not readily identifiable. Id. at 359. Now, however, the plaintiffs argue that the issues identified by the Fourth Circuit have been resolved in light of changes in the law, see supra at I, and that the classes as defined by the plaintiffs now satisfy the ascertainability requirement.

In its opinion, the court of appeals specifically instructed me to “give greater consideration to the administrative ehal-lenges it will face when using land records to determine current ownership” and to “determine whether it is possible to adjust the class definitions” in order to avoid such administrative challenges. Id. at 360. The court of appeals further suggested that verifying ownership of the gas estates at issue was a “prerequisite to identifying the class.” Id. at 369-60. Although the defendants had prepared schedules identifying owners of gas estates, these schedules were out of date, and updating them based on land records would have been a complicated endeavor, which the court of appeals considered a “significant administrative barrier” to verifying ownership. Id. at 369. However, as the plaintiffs have noted, a 2015 amendment to Virginia Code Ann. § 46.1-361.22:2 necessarily required gas well operators such as EQT and CNX to identify CBM gas owners before January 1, 2016.¹⁴ This amendment, and specifically, the requirements it imposes on gas well operators, greatly reduces the administrative challenges that previously plagued the aseertain-ability analysis in these cases.¹⁵

CNX protests that “[rjegardless of the class definition, individualized analysis must take place to determine the current gas claimant” and that “[s]uch a process is precisely what the ascertainability requirement is intended to avoid.” Hale v. CNX Gas Co., No, 1:10CV00059, Def.’s Mem. Opp’n Class Certification 14, ECF No. 463. It also notes that “[v]arious situations may prevent [it] from completing an ... application” for release of funds from escrow, including a lack of cooperation from identified claimants. Id. at 10. It is undoubtedly true that gas well operators have been required to undertake individual analyses in order to identify the gas claimants, and it is undoubtedly true that the gas well operators have faced challenges in complying with the new law. These challenges, however, are now part of the well operators’ obligations under the revised statute—not simply part of the process required in order to identify members of the classes. The proposed revised class definitions take advantage of work the defendants are already required to do, and they eliminate any Rule 23-specific administrative challenges by defining the classes based on what the defendants have accomplished as of specific dates.

“A class cannot be certified unless a court can readily identify the class members in reference to objective criteria.” EQT Prod. Co., 764 F.3d at 368. Here, those objective criteria are drawn directly from the requirements imposed on the defendants by law. For the reasons described above, I believe the plaintiffs’ revised proposed class definitions, combined with the change in the law, adequately address the administrative challenges of determining class membership.¹⁶

3. Commonality and Predominance.

“Rule 23(a)(2) requires a plaintiff to show that ‘there are questions of law or fact common to the class.’ ” Id. at 360 (quoting Fed. R. Civ. P. 23(a)(2)). Specifically, “what matters to class certification ... [is] the capacity of a classwide proceeding to generate common answers apt to drive the resolution of the litigation.” Id. (quoting Wal-Mart Stores, Inc. v. Dukes, 664 U.S. at 360, 131 S.Ct. 2541). Importantly, “[dissimilarities within the proposed class ... have the potential to impede the generation of common answers.” Dukes, 564 U.S. at 350, 131 S.Ct. 2541 (internal quotation marks and citation omitted). This is because “[a] question is not common ... if its resolution turns on a consideration of the individual circumstances of each class member.” Thorn v. Jefferson-Pilot Life Ins. Co,, 446 F.3d 311, 319 (4th Cir. 2006) (internal quotation marks and citation omitted). Furthermore, because the plaintiffs here seek certification under Rule 23(b)(3), any common questions must predominate over individual or other questions, or the class cannot be certified. EQT Prod. Co., 764 F.3d at 366 (citing Fed. R. Civ. P. 23(b)(3)).

As the Fourth Circuit noted in its previous opinion, the commonality requirement may be satisfied by a showing that “the defendants engaged in numerous common practices.” Id. at 366. However, because the predominance requirement of Rule 23(b)(3) is “more demanding” than the commonality requirement of Rule 23(a)(2), “the mere fact that the defendants engaged in uniform conduct is not, by itself, enough to satisfy” the predominance requirement. Id.; see also Comcast Corp. v. Behrend, 569 U.S. 27, 133 S.Ct. 1426, 1432, 185 L.Ed.2d 516 (2013); Lienhart v. Dryvit Sys., Inc., 255 F.3d 138, 146 n.4 (4th Cir. 2001) (quoting Amchem Prods., Inc., 521 U.S. at 609, 117 S.Ct. 2231). The more stringent predominance standard further requires both that this uniform conduct somehow “relate to the controversy at the heart of the litigation” and that “individual issues will not predominate.” EQT Prod. Co., 764 F.3d at 366. In order to meet the commonality and predominance requirements, the plaintiffs must therefore demonstrate both that a given practice was applied to all class members and that the class members are similarly situated, such that questions about the propriety of that practice can be answered on a classwide basis.

Ultimately, then, my inquiry is threefold. First, I must determine whether the defendants engaged in practices that are common to all class members or, alternatively, whether the plaintiff has identified a question capable of generating a common answer. If so, I must next determine whether these common practices or questions bear on the ultimate question of whether the defendants underpaid royalties. If the answer to this second question is “yes,” then finally, I must determine whether these common practices or questions are “sufficient to ensure that the class members’ common issues [will] predominate over individual ones.” Id. As part of this analysis, I must “analyze each of the plaintiffs’ claims to determine whether any of the distinct elements of those actions might affect the predominance of common questions.” Id. at 367 n.19.

The Hale, Addison, Adair, and Kiser plaintiffs identify three specific practices related to the defendants’ calculation and payment of royalties that they assert were applied to all class members. These practices are as follows: (1) taking excessive deductions; (2) calculating royalties based on improperly low prices; and (3) making payments into escrow after the deadlines imposed by the pooling orders. Hale v. CNX Gas Co., 1:10CV00069, Pl.’s Mem. Supp. Class Certification 11-21, ECF No. 459 (describing practices applied by CNX on behalf of the Hale and Addison classes);¹⁷ Kiser v. EQT Prod. Co., 1:11CV00031, Pl.’s Mem. Supp. Class Certification 11-14, ECF No. 350 (describing practices applied by EQT on behalf of the Adair and Kiser classes).¹⁸ The Hale and Adair plaintiffs, who represent deemed lessors in the so-called “force pooled” cases, further identified a fourth practice common to their classes: (4) deducting severance taxes from deemed lessors’ royalties. Id. at 12. In addition, the Hale and Addison plaintiffs identify an additional practice that is allegedly common to them classes: (5) calculating royalties based on physical sales, rather than realized profits.¹⁹ Hale v. CNX Gas Co., 1:10CV00059, Pl.’s Mem. Supp. Class Certification 16, ECF No. 469. The Hale plaintiff also identifies a sixth practice, which he concedes applies only to some members of the class: (6) drilling, completing, and producing CBM wells and selling gas from those wells before a pooling order was entered. Id. at 17. Finally, all four plaintiff classes seek an accounting, the availability of which they frame as an additional common question. I address each practice and question in turn.

a. Excessive Deductions.

i. CNX and the Hale and Addison Classes.

The Hale and Addison plaintiffs allege that CNX took deductions at unjustifiable fixed rates, which in turn resulted in deductions that were excessive. Although CNX did not take deductions at a single fixed rate applicable to every member of the Hale and Addison classes, it does appear to have taken deductions at two fixed rates within each class. These rates were applied to both deemed lessors (the Hale class) and voluntary lessors (the Addison class). Id., Mumford Dep. 65-56, Ex. EE, ECF No. 232-31. CNX takes deductions of either $1.35 or $0.97 per unit of gas, depending on whether the gas is sold at one of three internal sales points or one of two external sales points. Id., Mumford Aff. 2-3, ECF No. 281-5. The gas produced from a given well can be sold at both internal and external sales points, for a total of five possible sales points per well. Id. Although the deductions taken from each class member’s royalties fell into one of two categories, and thus were not numerically common across the class, CNX’s application of these two standard deductions nevertheless support a finding of commonality under Rule 23(a)(2).²⁰ Because the entire class was subjected to this standardized deduction scheme, application of this scheme is a common practice.

The next question I must address is whether this common deduction scheme bears on the question of whether the deductions taken by CNX were “excessive” and thus resulted in underpayments into escrow. The obvious answer is that it does. The fact that CNX appears to have applied one or two common deductions to a group of gas owners is certainly relevant to whether CNX underpaid those gas owners. If the gas owners are similarly situated with regard to their entitlements, application of a common scheme could mean that the group as a whole was underpaid (or not). Even if the gas owners are not so similarly situated—as CNX claims—application of a common scheme despite the owners’ differing entitlements could mean that some were underpaid, while others were not. Determining whether CNX actually did take deductions that were excessive as to a given gas owner, of course, requires further analysis.

I must finally consider whether the common question of the propriety of CNX’s deductions predominates over any individual issues. On this issue, the nature of the plaintiff classes is largely dispositive. The putative Hale class members, who are deemed lessors, are all subject to the same pooling order. The question of whether CNX’s deduction practices were improper as to these gas owners can thus be answered for the entire class in one fell swoop. The Addison plaintiffs, by contrast, are voluntary lessors bound by individual leases. Because they are subject to different entitlements, conduct that violates one set of leases may not violate another. The question of whether CNX’s deduction practices were improper as to these gas owners would therefore require individualized analysis.

The Hale Class.

In conducting a predominance inquiry, I must consider the elements of the plaintiffs claims. As to this particular issue, the Hale plaintiff raises claims for breach of fiduciary duty, conversion, and unjust enrichment. A claim for breach of fiduciary duty requires proof of (1) the existence of a fiduciary relationship and (2) a violation of a duty arising from that relationship. See, e.g., Remora Invs., L.L.C., 673 S.E.2d at 849; Williams, 576 S.E.2d at 757. A fiduciary relationship can arise by contractual provision, common law, or statute. See, e.g., Remora Invs., L.L.C., 673 S.E.2d at 849 (statute); Augusta Mut. Ins. Co., 645 S.E.2d at 295 (contract); Williams, 576 S.E.2d at 757 (common law).

Because the members of the Hale class are deemed lessors, their relationships with CNX are all governed by the same pooling orders. See, e.g., Hale v. CNX Gas Co., 1:10CV00059, Exs. AA & BB, ECF Nos. 232-27 & -28. This means they are all subject to the same entitlements. Whether CNX’s alleged fiduciary relationship arises by virtue of the pooling orders, common law, or statute is of no moment; if such a duty is owed, it is owed all the members of the putative Hale class. The answer to the question of whether CNX’s deduction practices constituted a breach of a fiduciary duty will likewise be the same across the class. I therefore find that the Hale plaintiff has fulfilled the predominance requirement on the breach of fiduciary duty claim as to the question of whether CNX took excessive deductions.

A claim for conversion requires proof of an “act of dominion wrongfully exerted over property in denial of, or inconsistent with, the owner’s rights.” Simmons, 544 S.E.2d at 679. A defendant is liable for conversion where its conduct infringes on an owner’s rights arising under the common law. Condo. Servs., Inc., 709 S.E.2d at 171. Because all gas owners possess the same rights of ownership under the common law, the putative class members are subject to the same entitlements. The answer to the question of whether CNX’s deductions were proper under the common law will therefore be the same for all members of the class. Thus, I find that the Hale plaintiff has met the predominance requirement on the conversion claim as to this issue.

The elements of a claim for unjust enrichment are (1) the plaintiff “conferred a benefit on” the defendant; (2) the defendant “knew of the benefit and should reasonably have expected to repay” the plaintiff; and (3) the defendant “accepted or retained the benefit without paying for its value.” Schmidt, 661 S.E.2d at 838. Such a claim arises under a theory of implied contract. Because the deemed lessors’ relationships with CNX are governed by pooling orders, the lessors are all subject to the same entitlements. The answer to the question of whether CNX’s deduction practices constituted unjust enrichment will therefore be the same across the class. I thus find that the Hale plaintiff has satisfied the predominance requirement on the unjust enrichment claim as to this issue.

The actual royalties to which each gas owner is entitled will vary depending on their ownership percentages. However, because I adopt a bifurcated approach to liability and damages, I believe the common question of whether the deductions taken by CNX were proper predominates over these individual damages calculations. See infra at III.A.3.f.ii. I therefore conclude that the Hale plaintiff has satisfied Rule 23(b)(3)’s predominance requirement on all claims as to the question of whether CNX properly calculated the deductions it took from class members’ royalty payments.

The Addison Class.

The Addison plaintiff raises claims of breach of contract, conversion, and, as to the class members whose royalties are es-crowed, breach of fiduciary duty. The elements for a claim for breach of contract are “(1) a legal obligation of a defendant to the plaintiff, (2) a violation or breach of that right or duty, and (3) a consequential injury or damage to the plaintiff.” Brown v. Harms, 251 Va. 301, 467 S.E.2d 805, 807 (1996) (internal quotation marks and citations omitted). Unlike the members of the putative Hale class, who are deemed lessors subject to the same pooling orders, the members of the putative Addison class are subject to individual leases into which they voluntarily entered.

The plaintiff has presented evidence that CNX uses a standardized “form lease” and that any departures from the terms of that form lease require special approval. Hale v. CNX Gas Co., 1:10CV00059, Scott Dep. 24-28, Ex. Y, ECF No. 232-25; Adair v. EQT Prod. Co., Nos. 1:10-CV-00037, 1:10-CV-00041, 1:11-CV-00031, 1:10-CV-00059, 1:10-CV-00065, 2013 WL 5429882, at *39 (W.D. Va. Sept. 5, 2013). However, Addison has not demonstrated that all members of the class are bound by this form lease. Indeed, as the Fourth Circuit specifically pointed out in its opinion, “the fact that CNX now uses a form lease ... does not establish that all of the Addison class members’ leases are uniform. CNX has inherited a large number of leases from predecessor companies, many of which contain different royalty provisions.” EQT Prod. Co., 764 F.3d at 369. If the Addison class was limited only to gas owners who were bound by CNX’s standard lease, then the question of whether CNX’s deductions constituted a breach of that lease would be the same across the entire class. As the class is currently defined, however, I agree with the Fourth Circuit that the “variable [lease] terms will make it difficult, if not impossible, ... to assess the validity of the defendants’ royalty payment practices on a classwide basis.” Id. at 867-68. I cannot find that the Addison class has satisfied the predominance requirement on the breach of contract claim as to this issue.

As I note above, a claim for conversion requires proof of an act of dominion wrongfully exerted over property in denial of the owner’s rights under the common law. Simmons, 544 S.E.2d at 679; Condo. Servs., Inc., 709 S.E.2d at 171. Because all gas owners possess the same rights under the common law, a finding that CNX’s common deduction practices violate those rights would apply across the entire Addison class. Thus, I find that the Addison plaintiff has satisfied the predominance requirement on the conversion claim as to this issue. However, for the reasons described below, see infra at III.A.4.b, I do not find that the Addison plaintiff has satisfied the superiority requirement on the conversion claim.

Finally, a claim for breach of fiduciary duty l’equires proof of (1) the existence of a fiduciary relationship and (2) a violation of a duty arising from that relationship. However, because the Addison class members’ relationships with CNX are governed by individual leases, they are not all subject to the same entitlements. Determining whether a fiduciary relationship exists between CNX and the putative class members would require individualized analysis as to each lease. The answer to the question of whether CNX’s deduction practices constituted a breach of fiduciary duty will therefore not be the same across the class. I therefore find that the Addison plaintiff has not fulfilled the predominance requirement on the breach of fiduciary duty claim as to this issue.

ii. EQT and the Adair and Kiser Classes.

The Adair and Kiser plaintiffs assert that, “[w]hen calculating its royalty payments” for gas produced from the Nora Field, EQT “took deductions of various ‘charges’ ” that were excessive and resulted in underpayments of royalties. Kiser v. EQT Prod. Co., 1:11CV00031, PL’s Mem. Supp. Class Certification 12, ECF No. 350. However, unlike the Hale and Addison plaintiffs, the Adair and Kiser plaintiffs do not allege that these deductions were uniform across the classes. Moreover, EQT operates three gas fields in Virginia, of which the Nora Field is only one, and the deductions it applies to the gas owners’ royalties vary by field. See id. Def.’s Mem. Opp’n. Class Certification 8-9, ECF No. 352 (citing Adair v. EQT Prod. Co., 1:10CV00037, Bergonzi Deck 2-3, Ex. 6, ECF No. 493-6). From their briefing, it is unclear whether the plaintiffs seek to establish as a common practice EQT’s deduction practices generally, or EQT’s deduction practices for the Nora Field. However, because I decline to certify these classes on either interpretation, this difference is of no import.

The plaintiffs cannot establish commonality based on EQT’s deduction practices generally because they have not shown that EQT has a common deduction practice. Between the three fields it operates in Virginia, EQT has employed four different practices -with regard to post-production deductions. For one field (Roaring Fork), EQT does not take post-production deductions. For the second field (Pilgrims Knob), post-production deductions are taken at a rate that is pre-deter-mined and uniform within a given year. For the third field (Nora), postproduction deductions were previously taken at the same rate as the Pilgrims Knob field; beginning in 2007, deduction rates have been calculated based on actual costs. Id. The plaintiffs do provide evidence that EQT “calculate^ all royalty payments using the same methodology” based on net proceeds. Adair v. EQT Prod. Co., 2013 WL 5429882, at *23 (citing Bergonzi Dep. 25, 27, ECF No. 399-20). However, the question they seek to certify as “common” addresses EQT’s deduction practices, not its royalty calculation methodology. EQT may very well calculate all of its royalty payments the same way, but if it applies different deduction practices to those royalty payments, there is no common question regarding those deduction practices,²¹

Similarly, the plaintiffs cannot establish commonality based on EQT’s deduction practices with regard to Nora Field specifically. They do not allege, and there is no evidence suggesting, that the members of these classes own gas interests only in the Nora Field, Because the plaintiffs bear the burden as to class certification, I assume that the members of these classes collectively own gas interests in all three fields. Assuming that the members of the Adair and Kiser classes own gas interests in three different fields, then an objection to EQT’s practices with regard to only one of those fields cannot serve as the basis for class certification,²² Because the Adair and Kiser plaintiffs have not sufficiently demonstrated that EQT’s deduction practices pose a question common to the classes, I find that they have not satisfied the commonality requirement of Rule 23(a)(2).

Moreover, even if a common question existed, the Kiser plaintiff would be unable to satisfy the predominance requirement as to her claims for breach of contract and fiduciary duty.²³ EQT uses multiple different lease forms in Virginia, and these leases vary as to their language governing the payment of royalties and post-production deductions. Kiser v. EQT Prod. Co., 1:11CV00031, Hoge Deck 2-3, Ex. O, ECF No. 351-19 (finding that “there are 20 different variations of royalty payment language” among the leases encompassed by the Kiser class).²⁴ “[S]ome leases expressly limit deductions,” while others don’t. Adair v. EQT Prod. Co., 1:10CV00037, EQT Interrog. 5, ECF No. 399-21. For the reasons described above regarding the Addison class, see supra at III.A.3.a.i, the varied leases held by the members of the putative Kiser class make it impossible for any common questions to predominate over individual inquiries. See EQT Prod. Co., 764 F.3d at 368.

b. Royalties Based on Improperly Low Prices.

The Hale and Addison plaintiffs allege that CNX is obligated to pay royalties based on the highest price obtainable, that it failed to meet this obligation, and that it therefore underpaid royalties. Specifically, they assert that CNX engaged in two improper common practices. The first is that CNX based its royalty payments on the prices it realized from physical sales, rather than the higher prices it realized through hedging. The second is that CNX based its royalty payments on prices “lower than published index prices.” Hale v. CNX Gas Co., 1:10CV00059, PL’s Mem. Supp. Class Certification 16-17, ECF No. 459. I will address each practice in turn.

Royalties Eased on Prices From Physical Sales.

First, the plaintiffs allege that it was improper for CNX to base its royalty payments on the prices it realized from physical sales, rather than from “prices [it] has realized through hedging or derivative transactions.” Id. at 17. They assert that CNX is not “entitled to ignore those revenues in paying royalties.” Id. In response, CNX asserts that “there is no connection between any hedge and the gas of any class member” and that its profits realized through hedging are therefore irrelevant to the calculation of royalty payments. Id. Def.’s Mem. Opp’n Class Certification 29, ECF No. 463.

The fact that CNX bases its royalty payments on the lower prices (and, thus, lower profits) from the sale of actual gas, rather than on the higher prices (and higher profits) realized from hedging, appears to be undisputed. See id. at 36-37 (citing id., England Report 3, Ex. 7, ECF No. 463-7); Def.’s Mem. Supp. Mot. to Dismiss 20-22, ECF No. 185 (arguing that CNX is entitled to pay royalties based on proceeds from the sale of gas rather than market speculation). This method of calculating royalty payments is therefore a common practice applicable to all members of the putative classes. In addition, it is clear that this common practice bears on the question of whether CNX underpaid royalties. If CNX should have been basing its royalty payments on the higher realized prices, rather than the lower sale prices it has been using, then CNX has been underpaying royalties; likewise, if CNX was entitled to use the lower sale prices, its calculations were proper. Finally, I must consider whether this common question—whether CNX should base its royalty payments off of physical sales prices or prices realized from hedging—predominates over the individual analyses required to actually calculate royalty entitlements for each gas owner. Again, the nature of the plaintiff classes is disposi-tive.

I must first consider the elements of the plaintiffs claims. As to this particular issue, the Hale plaintiff raises claims for breach of fiduciary duty and unjust enrichment. For the reasons described above, see supra at III.A.3.a.i, I find that the Hale plaintiff has fulfilled the predominance requirements for both of these claims as to the question of whether it is proper for CNX to calculate royalties based on its profits realized from physical sales. Because the members of the putative Hale class are all subject to the same pooling orders, the obligations imposed on CNX are uniform across the class. The question of whether CNX’s common royalty calculation practice constitutes a breach of fiduciary duty or unjust enrichment can therefore be answered for the entire class in a single stroke. If CNX is obligated to make royalty payments based only on proceeds from the sale of the gas—as it argues—this obligation will be the same as to all class members. Likewise, if CNX is obligated to make royalty payments based on the profits it realized from financial transactions—as the plaintiffs argue—this obligation will be the same as to all class members. Determining the actual royalties to which each class member is entitled will require some individual analysis, but because I adopt a bifurcated approach to liability and damages, I believe the common question presented by this practice predominates over these individual calculations. See infra at III. A.3.f.ii. I therefore conclude that the Hale plaintiff has satisfied Rule 23(b)(3)’s predominance requirement on all claims as to the question of whether CNX properly calculated royalty payments based on prices from physical sales, rather than on prices realized through hedging.

The Addison plaintiff raises claims for breach of contract, breach of fiduciary duty, and unjust enrichment. For the reasons described above, see supra at III.A.3.a.i, I cannot find that the Addison plaintiff has fulfilled the predominance requirements for the claims of breach of contract or breach of fiduciary duty as to the question of whether it is proper for CNX to calculate royalties based on its profits realized from physical sales. As voluntary lessors holding different individual leases, the members of the putative Addison class are subject to varying entitlements. Because the leases binding the putative class members are not uniform, the individual analyses required to determine CNX’s obligation with respect to every class member would “make it difficult, if not impossible,” to answer this question on a class-wide basis. EQT Prod. Co., 764 F.3d at 367-68.

The elements of a claim for unjust enrichment are (1) the plaintiff “conferred a benefit on” the defendant; (2) the defendant “knew of the benefit and should reasonably have expected to repay” the plaintiff; and (3) the defendant “accepted or retained the benefit without paying for its value.” Schmidt, 661 S.E.2d at 838. Because the members of the putative Addison class have entered into a variety of leases, it would be necessary to examine thp details of each lease in order to determine whether CNX’s royalty calculation practice constitutes unjust enrichment as to that leaseholder. The amount of money CNX should reasonably have expected to repay may very well be different for gas owners with different leases. Because the question of whether CNX’s royalty calculation practice constituted unjust enrichment cannot be answered without extensive individualized analysis, and because the answer to this question may not be the same across the class, crampon issues do not predominate.²⁵ I therefore conclude that the Addison plaintiff has not satisfied Rule 23(b)(3)’s predominance requirement on any of her claims as to the question of whether CNX properly calculated royalty payments based on prices from physical sales.

Royalties Based on Prices Below the Published Index.

The Hale and Addison plaintiffs also allege that it was improper for CNX to base its royalty payments on prices below the published index prices because CNX is obligated to pay royalties based on the highest price obtainable. CNX asserts in response that it has no such obligation, and instead has a duty only to act as a “reasonably prudent operator in marketing the CBM.” Hale v. CNX Gas Co., 1:10CV00069, Def.’s Mem. Opp’n Class Certification 22, EOF No. 463 (internal quotation marks omitted). It further asserts that this duty obligates it only to obtain the “best price reasonably attainable” at the time of the sale and that this price cannot be based on a “hypothetical” index price. Id. at 29 (internal quotation marks and emphasis omitted).

CNX’s practice of basing its royalty payments on prices from the actual sale of gas, rather than on published index prices, appears to be undisputed. The issue is whether this common practice is proper, which involves two questions. The first is whether CNX has an obligation to pay royalties based on the highest price obtainable (as the plaintiffs claim) or whether CNX has a duty to act as a reasonably prudent operator (as CNX claims). If CNX has an obligation to use the highest price obtainable, the second question is whether its current practice fulfills that obligation. Alternatively, if CNX has a duty to act as a reasonably prudent operator, the second question is whether it did, in fact, act as a reasonably prudent operator. Clearly, CNX’s royalty calculation practices and obligations to gas owners bear on the ultimate issue of whether it underpaid royalties. Because there is a common question that bears on the ultimate issue, I must consider whether the common questions above predominate over the individual analyses required to calculate each gas owner’s royalty entitlements. For the reasons described above, the nature of the plaintiff class is again dispositive.

As to this particular issue, the Hale plaintiff again raises claims for breach of fiduciary duty and unjust enrichment. For the reasons described above, see supra at III.A.3.a.i and III.A.3.b.i, I find that the Hale plaintiff has fulfilled the predominance requirements for both of these claims as to the question of whether it is proper for CNX to calculate royalties based on prices below the published index prices. The obligations imposed on CNX are uniform across this putative class of deemed lessors. As a result, the answer to the first question of whether CNX has an obligation to (a) pay royalties based on the highest price obtainable or (b) act as a reasonably prudent operator will be the same across the class. Furthermore, if the answer to this question is (a)—that CNX has an obligation to pay royalties based on the highest price obtainable—the answer to the second question of whether its current practice fulfills that obligation will also be the same across the class. If actual sales prices are considered the highest prices obtainable, then CNX has neither breached its fiduciary duty nor been unjustly enriched. By contrast, if actual sales prices are not the highest prices obtainable, CNX may have breached its fiduciary duty and/or been unjustly enriched. Either way, the questions can be answered for the entire class in one fell swoop.

CNX correctly notes that if the answer to the first question is (b)—that CNX has a duty to act as a reasonably prudent operator—the answer to the second question cannot be answered on a classwide basis. To fulfill a duty to act as a reasonably prudent operator, CNX must have obtained the “best price reasonably attainable.” Id. (internal quotation marks and emphasis omitted). Thus, the question of whether CNX did, in fact, act as a reasonably prudent operator requires the factfinder to determine the “best price reasonably attainable” for every individual CBM transaction. However, I do not believe this additional analysis precludes a finding that common questions predominate. Because the class members’ CBM has been pooled, the CBM sold in each transaction is traced back to the pool, not to individual class members. The analysis required to answer this question as to a single class member would therefore be as onerous as the analysis required to answer this question for the entire class. Because individualized analysis would be required whether the question was litigated on a classwide or member-by-member basis, I do not believe such analysis precludes a finding that the classwide nature of the question predominates. I therefore conclude that the Hale plaintiff has satisfied Rule 23(b)(3)’s predominance requirement on all claims as to the questions of (1) whether CNX has an obligation to use the highest price obtainable or to act as a reasonably prudent operator and (2) whether CNX’s practice met this obligation.

As above, the Addison plaintiff raises claims for breach of contract, fiduciary duty, and unjust enrichment.²⁶ For the same reasons previously stated, see supra at III. A.3.a.i and III.A.3.b.i, I cannot find that the Addison plaintiff has fulfilled the predominance requirements for her claims as to these questions. These class members are subject to varying entitlements; theoretically, one gas owner’s lease could require CNX to use the highest price obtainable, another may use the reasonably prudent operator standard, and still another might be silent on the topic. Because answering this question requires examination of individual leases, there is no way to answer this question on a classwide basis. Thus, I find that the Addison plaintiff has not satisfied Rule 23(b)(3)’s predominance requirement on any of her claims as to the questions of (1) whether CNX has an obligation to use the highest price obtainable or to act as a reasonably prudent operator and (2) whether CNX’s practice met this obligation.

The Adair and Kiser plaintiffs allege that EQT failed to meet its obligation to pay royalties based on the highest price obtainable because its “sales prices were consistently less than reported index prices.” Kiser v. EQT Prod Co., 1:11CV00031, Pl.’s Mem. Supp. Class Certification 13, EOF No. 360 (citing id. Adair Interrog. Resp. 6-7, Ex. C, ECF No. 361-3). Like CNX, EQT has responded that it has no such obligation and that it was obligated only to act as a reasonably prudent operator. Id. Def.’s Mem. Opp’n Class Certification 10, ECF No. 362. It further asserts that regardless of which standard applies, “plaintiffs will have to look behind each and every transaction to determine if it was appropriate given the market and other circumstances at the time and place of the transaction” and that the plaintiffs have not shown “how this can be done on a classwide basis.” Id. at 10-11.

As above, the issue here can be divided into two questions: (1) which standard applies—“highest price obtainable” versus “reasonably prudent operator”—and (2) whether EQT’s practice for calculating royalty payments meets that standard. Furthermore, as above, the outcome of the commonality and predominance inquiries turns on the nature of the plaintiff classes.

The Adair Class.

It appears to be undisputed that EQT bases its royalty payments to deemed lessors on sales prices rather than reported index prices. Id. Adair Interrog. Resp. 6-6, Ex. C, EOF No. 351-3 (noting that EQT’s discovery materials show that “the royalty payments made by EQT into [escrow] were calculated based on sales prices” and showing a table comparing EQT sales prices to three index prices). Use of sales prices, rather than index prices, in calculating royalty payments is therefore a practice common to all putative Adair class members.²⁷ It is also clear that both common questions bear on the ultimate issue of whether EQT underpaid royalties. The answer to the first question defines EQT’s obligation to gas owners and helps set the standard by which the propriety of its royalty payments will be evaluated; the answer to the second determines whether EQT did, in fact, underpay royalties.

I must next determine whether these common issues predominate over individual issues, and in so doing, must consider the elements of the plaintiffs claims. As to this particular issue, the Adair plaintiff raises claims for breach of fiduciary duty, conversion, and unjust enrichment. For the reasons described above as to the Hale class, see supra at III.A.3.a.i and III.A.3.b.i, I find that the Adair class fulfills the predominance requirements for all of these claims as to the question of whether it is proper for EQT to calculate royalties based on sales prices rather than published index prices. Because the members of the putative Adair class are deemed lessors subject to the same pooling orders, the obligations imposed on EQT are uniform across the class. The answer to the first question of whether EQT has an obligation to (a) pay royalties based on the highest price obtainable or (b) act as a reasonably prudent operator will therefore be the same across the class. Furthermore, if the answer to this question is (a), the answer to the second question of whether EQT’s current practice fulfills that obligation will also be the same across the class.

Like CNX, EQT contends that if the answer to the first question is (b), the answer to the second question cannot be adjudicated on a classwide basis because answering the question would require the plaintiffs to “look behind each and every transaction to determine if it was appropriate.” Id. Def.’s Mem. Opp’n Class Certification 10-11, ECF No. 352. Although I agree, for the same reasons described above as to the Hale class, see supra at III.A.3.b.i, I do not believe this precludes a finding that common issues predominate. Because the gas owners’ interests are pooled, the determination of propriety for each individual transaction will affect the claims of all gas owners with an interest in the well in question, if not the entire class. I therefore conclude that the Adair class satisfies Rule 23(b)(3)’s predominance requirement as to all claims based on the questions of (1) whether EQT has an obligation to use the highest price obtainable or act as a reasonably prudent operator and (2) whether EQT’s practice met this obligation.

The Kiser Class.

By contrast, the Kiser plaintiff has not shown that EQT employed a royalty calculation practice common to all class members. Although there is evidence that EQT calculated its royalty payments based on sales prices for members of the Adair class, see Id. Adair Interrog. Resp. 5-6, Ex. C, ECF No. 351-3, the Kiser plaintiff has not submitted evidence—and EQT has not conceded— that EQT uses this same practice for its voluntary lessors. I therefore conclude that the Kiser plaintiff has not satisfied the commonality requirement of Rule 23(a)(2).

Even assuming that EQT did engage in such a common practice for its voluntary lessors, however, for the reasons stated above as to the Addison plaintiff, see supra at III.A.3.b.i, I could not find that the Kiser plaintiff has satisfied the predominance requirement as to her claims for breach of contract, unjust enrichment,²⁸ or breach of fiduciary duty.²⁹ Because the members of the putative Kiser class are voluntary lessors with their own individual leases, the obligations imposed on EQT are not uniform across the class. The Kiser plaintiffs own expert sorted class members’ leases into categories based on the different obligations the leases impose on EQT regarding royalty payments. Eight different lease variations require EQT to calculate royalties based on proceeds; five require calculation based on market value; four require calculation based on both proceeds and market value; one requires calculation based on the wellhead price; one requires calculation based on the sales price; and one requires calculation based on both proceeds and maximum lawful price. Id., Hoge Deck 12-18, Ex. O, ECF No. 351-19. With six categories and twenty lease variations on this specific issue, it is obvious that the question of whether EQT must pay royalties based on the highest price obtainable versus act as a reasonably prudent operator cannot be answered on a classwide basis. I therefore find that the Kiser class does not satisfy Rule 23(b)(3)’s predominance requirement on the breach of contract, unjust enrichment, or breach of fiduciary duty claims as to this issue.

c. Late Payments into Escrow.

The Hale and Addison plaintiffs allege that CNX failed, on “dozens” of occasions, to make timely payments into escrow after obtaining a pooling order or producing CBM from class members’ wells. Hale v. CNX Gas Co., 1:10CV00059, Pl.’s Mem. Supp. Class Certification 21, ECF No. 459. The fact that CNX has made some delayed payments appears to be undisputed. Id., Duty Deck 6, Ex. 1, ECF No. 463-1 (describing various reasons why payments are delayed). However, the plaintiffs have failed to show that making delayed payments is a common practice. Indeed, they concede that “not all Class members [were] affected by this particular practice,” id. PL’s Mem. Supp. Class Certification 21, ECF No. 459, and they do not allege that CNX has adopted a policy of delaying payments. Because CNX does not employ a common practice of making late payments into escrow, the fact that it has occasionally done so cannot serve as a basis for class certification.

The plaintiffs nevertheless assert that CNX’s actions raise a common question that can produce a common answer: “whether a failure to make timely payments into escrow makes CNX liable for” breach of fiduciary duty, conversion, and unjust enrichment. Id. I disagree, however, that this question has a common answer that would apply across these classes. The question proposed by the plaintiffs cannot be answered without examining every late payment, determining whether the delay was justified, and then determining whether an unjustified delay constitutes proof of these claims. As CNX points out, a payment could be delayed as a result of a number of circumstances beyond its control, including waiting on action by the Gas and Oil Board or an escrow agent. Id. Def.’s Mem. Opp’n Class Certification 32-33, ECF No. 463; Duty Deck 6, Ex. 1, ECF No. 463-1. A payment that appeared to be delayed or outstanding could also turn out not to have been required in the first place. Id. at 6-7. Because there are possible justifications for any given delayed payment, and because there are multiple instances of delayed payments within the plaintiff classes, the question of whether delayed payments render CNX liable for breach of fiduciary duty, conversion, or unjust enrichment cannot be answered for the entire classes in a single stroke.

The fact that the plaintiffs’ question cannot generate a single answer applicable across the classes means that it is not a common question. Moreover, the fact that the question cannot be answered without delving into the details of individual transactions means that, even if it were common, it would not predominate over individual issues. I therefore conclude that neither the Hale class nor the Addison class satisfies the commonality and predominance requirements as to the issue of CNX’s delayed payments into escrow.

The Adair and Kiser plaintiffs likewise allege that EQT “consistently failed to make payments into escrow” after obtaining a pooling order or producing CBM from class members’ wells, with “dozens” of delays on the order of years. Kiser v. EQT Prod. Co., 1:11CV00031, Pl.’s Mem. Supp. Class Certification 13-14, ECF No. 350. The fact that EQT has failed to make timely payments into escrow is supported by a 2009 newspaper investigation, and EQT does not dispute these allegations. See id. Adair Interrog. Resp. 4, Ex. C, ECF No. 351-3. However, like the Hale and Addison plaintiffs, the Adair and Kiser plaintiffs concede that “[n]ot all Class members ... have been affected by this particular practice,” and they do not allege that EQT has adopted a policy of delaying its payments. Id. Pl.’s Mem. Supp. Class Certification 14, ECF No. 350. Therefore, as above, there is no common practice here.

The Adair and Kiser plaintiffs pose the same purportedly common question as the Hale and Addison plaintiffs: whether EQT’s “failure to malee timely payments into escrow makes [it] liable” for breach of fiduciary duty. Id. For the reasons described above, see supra at III.A.3.C.Í, I must find that neither the Adair class nor the Kiser class satisfies the commonality and predominance requirements as to this issue.

d. Deduction of Severance Taxes.

i. CNX and the Hale Class.³⁰

The members of the putative Hale class are deemed lessors whose gas interests were force-pooled by the Virginia Gas and Oil Board. They allege that “CNX deducts severance taxes from all deemed lessors’ royalties” in violation of the pooling orders that govern their interests. Hale v. CNX Gas Co., 1:10CV00059, Pl.’s Mem. Supp. Class Certification 11, ECF No. 459. The fact that CNX has, in fact, deducted severance taxes from the Hale class members’ royalties is undisputed. See id. Def.’s Mem. Supp. Mot, to Dismiss 22, ECF No. 185 (arguing that severance taxes are “post-production costs” and are therefore properly deducted from royalties). Furthermore, CNX appears to deduct severance taxes from the royalty payments of all deemed lessors and, therefore, all members of the putative Hale class. See id. Def.’s Suppl. Resp. Interrog. 11, Ex, W, ECF No. 232-23 (stating that CNX “has been deducting [severance and/or license] taxes when calculating royalty payments ... since January 1, 2004”); id., Mumford Dep. 58-59, Ex. EE, ECF No. 232-31 (testifying that deductions for severance or licensing taxes were “part of the deduction calculation” and that, “absent some exception to the ordinary rule,” CNX takes deductions for severance taxes). The deduction of severance taxes is therefore a common practice.

In addition, this common practice clearly bears on the ultimate question of whether CNX underpaid royalties. If severance taxes are not an appropriate deduction under the terms of the governing pooling order, then by taking them as a deduction, CNX has been underpaying royalties. By contrast, if severance taxes are an appropriate deduction, then CNX has not been underpaying royalties in this regard.

I must finally determine whether this common question predominates over any individual issues. In so doing, I must consider the elements of the plaintiffs claims. As to this particular issue, the Hale plaintiff raises claims for breach of fiduciary duty, conversion, and unjust enrichment. For the reasons described above, see supra at III.A.3.a.i, I find that the Hale plaintiff has fulfilled the predominance requirements for all of these claims as to this question. Because the putative class members are governed by the same pooling orders, the answer to the question of whether deduction of severance taxes is proper will be the same across the class.

Individualized analysis as to damages would be necessary only if such deductions are found to constitute a breach of fiduciary duty, conversion, or unjust enrichment. If such deductions are proper, as CNX asserts they are, then there will be no need for class members to adduce individualized proof of damages. Moreover, if the deductions are found to be improper and it becomes necessary to calculate individual damages, such calculations will not be complicated or overly burdensome given that CNX retains records of the amounts deducted for severance and license taxes. See id, Hodges Letter 3, Ex. M, ECF No. 232-13. Furthermore, because I adopt a bifurcated approach to liability and damages, such individualized damage calculations will not preclude a finding of predominance. See infra at III.A.3.f.ii. The common question regarding the propriety of CNX’s severance tax deduction practice therefore predominates over individual issues. I thus conclude that the Hale plaintiff has satisfied Rule 23(b)(3)’s predominance requirement on all claims as to the question of whether CNX properly deducted severance taxes from deemed lessors’ royalty payments.

ii. EQT and the Adair Class.³¹

Like the Hale plaintiffs, the Adair plaintiffs are deemed lessors whose gas interests were force-pooled by the Virginia Gas and Oil Board. They allege that “EQT deducts severance taxes from all deemed lessors’ royalties” in violation of the pooling orders. Kiser v. EQT Prod. Co., 1:11CV00031, Pl.’s Mem. Supp. Class Certification 12, ECF No. 350. The fact that EQT does deduct severance taxes from the royalties paid to all of its deemed lessors appears to be undisputed. See Adair v. EQT Prod. Co., 1:10CV00037, Bergonzi Dep. 235-36, Ex. T, ECF No. 399-20 (testifying that a deduction for severance taxes is “taken against all deemed lessors”). Deduction of severance taxes from deemed lessors’ royalties is therefore a common practice. As above, this common practice clearly bears on the ultimate question of whether EQT underpaid royalties.

As to this issue, the Adair plaintiff raises claims of breach of fiduciary duty, conversion, and unjust enrichment. For the reasons described above as to the Hale class, see supra at III.A.3.d.i, I conclude that this common question predominates over any individual issues. Although EQT employs different general postproduetion deduction practices across its three Virginia gas fields, see supra at III.A.3.a.ii, by its own admission, it applies the same severance tax deduction practice to all its deemed lessors, regardless of field.³²

EQT asserts that this particular question is not of “sufficient importance” to satisfy Rule 23(a)(2). Kiser v. EQT Prod. Co., 1:11CV00031, Def.’s Mem. Opp’n Class Certification 7, ECF No. 352. It estimates that the amount at issue for these particular claims is approximately $192,000 and asserts that “[a] claim of this magnitude does not support the time and expense of a class action, given the cost of class notice and other administrative expenses.” Id. at 8 (citing Adair v. EQT Prod. Co., 1:10CV00037, Def.’s Mot. for Protective Order 3, Ex. 3, ECF No. 333-3). I express no opinion as to whether this amount alone is of “sufficient importance” to justify class certification, however, because I grant certification to the Adair class on another issue as well. See supra at III.A.3.b.ii. I therefore find that the Adair plaintiff has satisfied Rule 23(b)(3)’s predominance requirement on all claims as to the question of whether EQT properly deducted severance taxes from deemed lessors’ royalty payments.

e. Production Prim to Pooling Orders.

i. CNX and the Hale Class.³³

The Hale plaintiff has alleged that, for certain force-pooled wells, “CNX drilled, completed, and produced the wells, and took possession of and sold CBM from the wells before a pooling order had been entered,” Hale v. CNX Gas Co., 1:10CV00059, Pl.’s Mem. Supp. Class Certification 17, ECF No. 459 (emphasis omitted). A pooling order authorizes an operator to begin drilling and operating a well. Va. Code Ann. § 45.1-361.21(C)(1). The fact that CNX did begin drilling and operating some wells prior to obtaining the necessary pooling orders appears to be undisputed. Hale v. CNX Gas Co., 1:10CV00059, Pl.’s Mem. Supp. Class Certification 19, ECF No. 459; id. Seitz Deck, Ex, N, ECF No. 460-14 (table showing gas units in which CNX began producing CBM prior to receiving a pooling order). However, it does not appear that early production of CBM was standard practice for CNX. Indeed, although the plaintiff asserts that CNX engaged in this “fraudulent” practice “with shocking regularity,” id. Pl.’s Mem. Supp. Class Certification 18-19, ECF No. 459, he also acknowledges that it did not do so as a matter of course. Id. at 19 (noting that one review showed that CNX had prematurely produced CBM in twenty-two out of sixty-seven cases). Because CNX does not employ a common practice of drilling wells and producing CBM before obtaining a pooling order, the fact that it has occasionally done so does not satisfy the commonality requirement.

The plaintiff nevertheless asserts that CNX’s actions raise a common question that can produce a common answer: whether CNX’s production of CBM prior to obtaining a pooling order renders CNX liable for trespass, conversion, breach of fiduciary duty, or unjust enrichment. See id. at 20. However, I disagree that this question is capable of being answered on a classwide basis. On the face of the statute, it is far from clear that producing CBM prior to obtaining a pooling order is per se improper. See Va. Code Ann. § 45.1-361.21(0(1). The Virginia Gas and Oil Board has established its own procedures governing operators’ pooling and drilling permit applications. The question of whether CNX’s conduct renders it liable for Hale’s claims cannot be answered without examining these regulations and CNX’s compliance with them as to each individual gas owner.

More importantly, the Hale plaintiff cannot satisfy Rule 23(b)(3)’s two-pronged predominance requirement: the issue must “relate to the controversy at the heart of the litigation,” and “individual issues [must] not predominate.” EQT Prod. Co., 764 F.3d at 366. As part of my predominance inquiry, I must ask whether this issue bears on the ultimate question of whether CNX underpaid royalties, and I conclude that it does not. Even if it is improper for CNX to produce CBM before obtaining a pooling order, the timing of the production does not necessarily impact its royalty payments. Indeed, CNX has presented evidence that it escrowed royalties even where the CBM was produced before a pooling order had been issued. Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp’n Class Certification 32, ECF No. 463 (citing id., Hr’g Tr. 67:23-68:1-8, Feb. 2, 1999, Ex. J, ECF No. 174-10). Furthermore, Hale has not alleged that CNX’s premature production of CBM has impacted CNX’s payment of royalties. He asserts that premature production deprives gas owners of the opportunity to make an informed decision regarding their participation in the well, but this assertion, even if true, has no bearing on the ultimate question of whether CNX has underpaid the gas owners’ royalties.

Finally, it is far from clear that common issues would predominate over individual issues. As I note above, whether CNX’s conduct constituted trespass, conversion, breach of fiduciary duty, or unjust enrichment as to a given gas owner may very well depend on whether CNX complied with the Board’s regulations as to that gas owner. More importantly, the claims of each individual gas owner could be subject to certain equitable defenses, depending on the circumstances surrounding the permitting process for that owner. See id. at 31. The validity of such defenses could only be assessed on an individual basis. CNX also appears to have entered into prior agreements with some gas owners—who subsequently became deemed lessors and, thus, putative class members— that gave CNX the right to “explore for CBM and drill and operate CBM wells” prior to obtaining a pooling order. Id. at 31-32. Determining whether CNX’s conduct was in fact improper would therefore require an inquiry into its history with each gas owner. For the reasons stated, I conclude that the Hale plaintiff has not satisfied the predominance requirement as to the question of the propriety of CNX’s production of CBM prior to obtaining pooling orders.

f Availability of Accounting and Bifurcated Proceedings.

Finally, the plaintiffs have characterized as common the question of whether they and their putative class members are entitled to an accounting. They assert that the answer to the question of whether they are entitled to an accounting is “a simple yes or no” and would apply equally to all class members. Id. PL’s Mem. Supp. Class Certification 10, ECF No. 459; Kiser v. EQT Prod. Co., 1:11CV00031, PL’s Mem. Supp. Class Certification 10, ECF No. 350. They further assert that employing a bifurcated strategy—that is, by determining liability on a classwide basis and damages on an individual basis— they can satisfy the predominance requirement. Hale v. CNX Gas Co., 1:10CV00059, PL’s Mem. Supp. Class Certification 22-25, ECF No. 459; Kiser v. EQT Prod. Co., 1:11CV00031, Pl.’s Mem. Supp. Class Certification 15-18, ECF No. 350. The defendants contend that the individualized proof of damages required in these cases defeats the predominance requirement. Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp’n Class Certification 35-40, ECF No. 463; Kiser v. EQT Prod. Co., 1:11CV00031, Def.’s Mem. Opp’n Class Certification 11-14, ECF No. 352. I find the nature of the plaintiff classes dispositive.

For the putative Addison and Kiser classes, the question of whether they are entitled to an accounting is not a common question capable of classwide resolution. In Virginia, a plaintiff may seek an accounting in equity against his fiduciary “for receiving more than comes to his just share or proportion.” Va. Code Ann. § 8.01-31. Because the putative class members are voluntary lessors subject to different entitlements, a practice that results in the defendant “receiving more than ... [its] just share” with regard to one lease may well be perfectly proper with regard to a different lease. Id. It is impossible to determine, on a classwide basis, whether a given practice produces entitlement to an accounting. I therefore must conclude that the Addison and Kiser plaintiffs have not satisfied the commonality requirement as to their request for an accounting.

For the force-pooled classes, the answer to the question of whether the class members are entitled to an accounting will be the same across the classes. Once an accounting has been awarded, the fiduciary will have the burden of proving that it properly paid and otherwise handled the plaintiffs funds. See Bain v. Pulley, 201 Va. 398, 111 S.E.2d 287, 291 (1959). Before an accounting can be awarded, however, the plaintiff must provide a “valid justification for [his] accounting request[ ].” Vaughan v. Wells Fargo Bank, N.A., No. 6:15-CV-00038, 2016 WL 2901752, at *5 (W.D. Va. May 18, 2016).

Because the members of the putative Hale and Adair classes are deemed lessors subject to the same pooling orders, the question of whether these plaintiffs have provided such a justification can be answered for their classes in one fell swoop. If the defendants’ conduct constituted receipt of more than their “just share or proportion,” Va. Code Ann. § 8.01-31, as defined by their obligations under the law and pooling orders, the plaintiffs will be entitled to an accounting. Likewise, if the defendants’ conduct did not constitute such a breach, the plaintiffs will not be entitled to an accounting. The plaintiffs therefore satisfy the commonality requirement of Rule 23(a)(2).

Rule 23(b)(3)’s predominance requirement requires that common questions predominate over individual issues. In order to satisfy the predominance requirement with respect to their request for an accounting, the plaintiffs must base their “valid justification” only on the evidence adduced in support of the other questions that have been certified for class treatment. The classes were certified as to those questions specifically because they present common questions that predominate over individual issues; as a result, the plaintiffs will be presenting evidence as to those questions independent from their request for an accounting. To the extent that such evidence is probative of both their claims on the certified issues and their request for an accounting, the plaintiffs satisfy the predominance requirement as to the question of whether they are entitled to an accounting.

Predominance would be defeated, however, by the introduction of any evidence that is not also adduced in support of the certified issues. For example, neither the Hale nor the Adair class met the Rule 23 requirements as to the issue of the defendants’ alleged late payments into escrow. As such, the plaintiffs may not present evidence of such alleged late payments in support of their request for an accounting. Because the issue is not appropriate for class treatment, the plaintiffs would need to present individualized evidence, separate from the evidence adduced in support of the certified issues. Opening the door to such individualized evidence would prevent common issues from predominating.

In addition, implementation of a bifurcated proceeding, similar to the one proposed by the plaintiffs, is essential to Hale and Adair’s satisfaction of the predominance requirement. Rule 23(c)(4) permits a district court to certify a class “with respect to particular issues.” Fed. R. Civ. P. 23(c)(4). Although bifurcated liability and damages proceedings are most common in the context of Title VII discrimination class action suits, they have also been employed in other contexts. See, e.g., Henley v. FMC Corp., 20 Fed.Appx. 108, 111, 118 (4th Cir. 2001) (unpublished) (in personal injury case, finding no problem with district court’s contemplated bifurcated approach, but holding that court had erred in failing to properly instruct the jury as to the bifurcation of classwide and individual issues). Indeed, the Advisory Committee specifically suggests such an approach, noting that Rule 23(c)(4) “recognizes that an action may be maintained as a class action as to particular issues only” and that following a classwide adjudication of liability, “the members of the class may thereafter be required to come in individually and prove the amounts of their respective claims.” Fed. R. Civ. P. 23(c)(4) advisory committee’s note to 1966 amendment.

The defendants object that the plaintiffs have failed to “establish that damages are susceptible of measurement across the entire class,” as required by the Supreme Court in Comcast Corporation v. Behrend, 569 U.S. 27, 133 S.Ct. 1426, 185 L.Ed.2d 515 (2013). Hale v. CNX Gas Co., 1:10CV00069, Def.’s Mem. Opp’n Class Certification 35, EOF No. 463 (quoting Comcast Corp., 133 S.Ct. at 1433). In Comcast, the Supreme Court held that a class action was improperly certified where “[questions of individual damage calculations [would] inevitably overwhelm questions common to the class.” 133 S.Ct. at 1433. However, I believe the bifurcated approach proposed by the plaintiffs mitigates this concern.

In deciding Comcast, the Court did not consider the role a bifurcated approach would play in resolving the predominance problem presented by individual damage calculations. In Comcast, the district court was asked to determine both liability and damages on a classwide basis, see Comcast Corp,, 133 S.Ct. at 1431; here, the plaintiffs have requested classwide determinations only as to questions of liability. Multiple courts of appeals have since concluded that “[w]here determinations on liability and damages have been bifurcated [pursuant to Rule 23(c)(4)], the decision in Comcast—to reject certification of a liability and damages class because plaintiffs failed to establish that damages could be measured on a classwide basis—has limited application.” Glazer v. Whirlpool Corp. (In re Whirlpool Corp. Front-Loading Washer Prods. Liability Litig.), 722 F.3d 838, 860 (6th Cir. 2013); see also In re Deepwater Horizon, 739 F.3d 790, 817 (5th Cir. 2014) (holding that “the rule of Comcast is largely irrelevant [w]here determinations on liability and damages have been bifurcated ... and the district court has reserved all issues concerning damages for individual determination” (internal quotation marks and citation omitted)). Indeed, even the Comcast dissenters, arguing that the Court’s writ of certiorari was improvidently granted, recognized the role bifurcation may play in a predominance inquiry: “[A]t the outset, a class may be certified for liability purposes only, leaving individual damages calculations to subsequent proceedings.” Comcast Corp., 133 S.Ct. at 1437 n.* (Ginsberg, J. & Breyer, J., dissenting) (citation omitted).

“Even after Comcast, the predominance inquiry can still be satisfied under Rule 23(b)(3) if the proceedings are structured to establish liability on a class-wide basis, with separate hearings to determine—if liability is established—the damages of individual class members.” In re Deepwater Horizon, 739 F.3d at 817 (internal quotation marks and citations omitted). I agree, and so structure the proceedings here. The Hale and Adair plaintiffs have fulfilled the certification prerequisites with respect to certain issues. It would be far more efficient to litigate these common issues of liability on a classwide basis than it would be to require each claimant to relitigate the same common issues, using the same proof as his putative fellow class members, on an individual basis. Following the elasswide resolution of these issues, in accordance with the Advisory Committee’s suggested procedure, class members can come in individually or in homogenous subclassses to prove their damages.

The defendants also object that the plaintiffs cannot “ ‘establish that damages are susceptible to measurement across the entire class’ ” because they have proposed no model with which to calculate damages. Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp’n Class Certification 35, ECP No. 463 (quoting Comcast Corp., 133 S.Ct. at 1433). I disagree. I believe the accounting requested by the plaintiff, combined with the transaction and deduction records kept by defendants, offer a sufficient means of calculating damages. In Comcast, the damages model proposed by the plaintiffs “failed to measure damages resulting from the particular ... injury on which [the defendants’] liability ,.. [wa]s premised,” which in turn precluded a finding that damages were “susceptible of measurement across the entire class.” 133 S.Ct. at 1433. In Hale and Adair, by contrast, the requested accounting and records would make it easy to measure damages for all class members. The injury alleged is underpayment of royalties, and there are clear records of the methods by which royalties were calculated, the deductions taken from said royalties, and the royalties actually paid. Any damages to which the class members are entitled will be easily measured.

One case from the Ninth Circuit, Leyva v. Medline Industries, 716 F.3d 510 (9th Cir. 2013), is particularly on point in this regard. In Leyva, the putative plaintiff classes alleged that their defendant employer violated various state labor codes. The district court denied certification of the classes under Rule 23(b)(3) because although “common questions exist with respect to Defendant’s liability ... the damages inquiry will be highly individualized.” Id. at 513 (citation omitted). The Ninth Circuit reversed, finding that the district court had abused its discretion in denying certification on this basis and concluding that “the presence of individualized damages cannot, by itself, defeat class certification under Rule 23(b)(3).” Id, at 514 (citing Dukes, 564 U.S. at 362, 131 S.Ct. 2541). It further explained that the putative class members’ damages would be calculated based on the defendant’s own records and databases, which “would enable the court to accurately calculate damages ... for each claim.” Id. Because this method would allow damages to be “feasibly and efficiently be calculated once the common liability questions [were] adjudicated,” it reasoned, class certification was appropriate. Id. Likewise, I believe the method proposed here—calculating damages based on the accountings and the defendants’ records—would likewise permit a feasible and efficient calculation of damages following any adjudication of liability.

In adopting a bifurcated approach, I note that the plaintiffs’ request for an equitable accounting does not strip the defendants of their right to a jury trial on any legal issues in the case. See Official Comm. of Unsecured Creditors v. Schwartzman (In re Stansbury Poplar Place, Inc.), 13 F.3d 122, 125 (4th Cir. 1993) (citing Dairy Queen, Inc. v. Wood, 369 U.S. 469, 477-78, 82 S.Ct. 894, 8 L.Ed.2d 44 (1962)). Thus, at the first stage of the proceeding, the defendants have a right to have their liability decided by a jury. If the defendants are found not to be liable for the plaintiffs’ claims, the second stage of the proceeding—the requested accounting and, of course, damages—would not take place. Indeed, the plaintiffs themselves have acknowledged that an accounting will not be necessary if the defendants are found not to have engaged in improper practices. Adair v. EQT Prod. Co., 1:10CV00037, Hr’g Tr. 109:20-24, Sept. 18, 2015, ECF No. 581; see also supra at II.A.3.b. n.4.

For the reasons explained above, I find that the Hale and Adair plaintiffs have satisfied the commonality and predominance requirements with regard to their request for an accounting, on the condition that any evidence used to support this request be limited to the evidence adduced in support of their other certified claims. I also find that a bifurcated approach to liability and damages in this case is appropriate. I will therefore certify these classes, pursuant to Rule 23(c)(4), as to the common issues identified above.

4. Superiority.

a. The Hale and Adair Classes.

In addition to predominance, Rule 23(b)(3) requires that plaintiffs demonstrate that “proceeding as a class ⅛ superior to other available methods for fairly and efficiently adjudicating the controversy.’ ” EQT Prod. Co,, 764 F.3d at 371 (quoting Fed. R. Civ. P. 23(b)(3)). The plaintiffs assert that “a class action remains the superior method of adjudicating these claims.” Hale v. CNX Gas Co., 1:10CV00059, Pl.’s Mem. Supp. Class Certification 34, ECF No. 459; Kiser v. EQT Prod. Co., 1:11CV00031, Pl.’s Mem. Supp. Class Certification 26, ECF No. 360. In response, CNX argues that the plaintiffs have failed to carry their burden of “prov[ing] that class action proceedings would be superior to individual actions.” Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp’n Class Certification 41, ECF No. 463.1 disagree. I previously concluded, and the Fourth Circuit agreed, that multiple factors weigh in favor of class action litigation in these cases. See EQT Prod. Co., 764 F.3d at 371 (citing Adair, 2013 WL 5429882, at *40). The removal of ownership claims from these cases does not change this calculus.

On remand, the Fourth Circuit has instructed me to additionally “consider how the dominance of state law issues may affect the suitability of this litigation in a federal forum and what state-law mechanisms may be available to resolve the underpayment claims as an alternative to a class action.” Id. These are diversity cases, and as a result, state law issues certainly dominate. However, for the reasons described below, I do not believe this dominance precludes a finding of superiority.

EQT asks this court to abstain from hearing these cases under the doctrine of abstention outlined by the Supreme Court in Burford v. Sun Oil Co., 319 U.S. 315, 63 S.Ct. 1098, 87 L.Ed. 1424 (1943). See Kiser v. EQT Prod. Co., 1:11CV00031, Def.’s Mem. Opp’n Class Certification 19-20, ECF No. 352. I decline to do so. It is well-established that a federal court may abstain from hearing a case “only where the relief being sought is equitable or otherwise discretionary.” Quackenbush v. Allstate Ins. Co., 517 U.S. 706, 731, 116 S.Ct. 1712, 135 L.Ed.2d 1 (1996). Where the action is one for damages, Bmford abstention is unwarranted. Id. The plaintiffs here seek a combination of equitable relief—in the form of an accounting—and money damages. As such, Bmford abstention is unavailable in these cases. Moreover, in passing House Bill 2058, the General Assembly specifically declined to divest the courts of jurisdiction over gas owners’ claims for escrowed funds, including equitable claims for an accounting. Va. Code Ann. § 45.1— 361.22:2(F) (stating that “[t]his section shall not operate to extinguish any other right or cause of action ... that may exist including ... claims for an accounting”).

In addition, as the plaintiffs point out, Virginia law does not permit class actions. This means claimants must either bring individual claims or seek certification of a federal class action, and the evidence on this point weighs heavily in favor of the class action. As the Fourth Circuit noted, “for many of these claimants, collective action may offer the only realistic opportunity recover.” EQT Prod. Co., 764 F.3d at 371. Many claimants “own only a fractional interest in a 12.5 percent royalty,” a fact which has “no doubt ... resulted in the sparse number of individual eases filed to date.” Id. (citing Adair, 2013 WL 5429882, at *40). Furthermore, “concerns of judicial economy support[ ] a finding of superiority because a collective action would allow a court to resolve all of the royalty owners’ claims in a single forum and lessen the risk of inconsistent judgments against the defendants.” Id. (citing Adair, 2013 WL 5429882, at *40).

The Fourth Circuit also specifically noted that “the extent of the defendants’ efforts to resolve and pay undisputed claims” is relevant to the superiority analysis. Id. “A finding that the defendants have not acted in good faith toward that end,” it held, “may weigh strongly in favor of a finding of superiority of a class action.” Id. The Hale plaintiff asserts that CNX has not acted in good faith, alleging that although CNX made payments to companies following another lawsuit, it did not adjust those payment practices as they applied to lessors with smaller intei'ests. Hale v. CNX Gas Co., 1:10CV00059, Pl.’s Mem. Supp. Class Certification 35, ECF No. 459. The fact that CNX changed its practices for companies but not for individual lessors, the plaintiff asserts, demonstrates “the relative powerlessness of [its] lessors, and the need for a class action to aggregate their claims.” Id. I express no opinion as to whether either defendant has acted in good faith regarding their efforts to resolve and pay undisputed claims. However, as I have explained, I agree with the Hale plaintiff that the financial situations of the individual lessors, and the fact that “for many of [them], collective action may offer the only realistic opportunity to recover,” EQT Prod. Co., 764 F.3d at 371, weighs in favor of a finding of superiority.³⁴

b. The Addison and Kiser Classes.

The analysis above, supra at III.A.4.a, is equally applicable to the Addison and Kiser plaintiffs. However, as discussed above, see supra at III.A.3,1 found that these plaintiffs were unable to fulfill the commonality and/or predominance requirements for all but one of their claims. Absent those prerequisites, I cannot certify those classes as to those claims. It is therefore unnecessary for me to consider whether they have satisfied the superiority requirement.

I did find that these plaintiffs satisfied, or could conceivably satisfy, the commonality and predominance requirements for their conversion claims. However, I find it unlikely that these claims will prevail at the merits stage of the litigation. A plaintiff may properly plead both a claim for conversion and a claim for breach of contract. Legard v. EQT Prod. Co., No. 1:10CV00041, 2011 WL 86598, at *14 (W.D. Va. Jan. 11, 2011) (citing Combined Ins. Co. of Am. v. Wiest, 578 F.Supp.2d 822, 833 (W.D. Va. 2008)). However, “[t]o recover for ... conversion, ‘the duty ... breached must be a common law duty, not one existing between the parties solely by virtue of the contract.’ ” Condo. Servs., Inc., 709 S.E.2d at 171 (quoting Dunn Constr. Co. v. Cloney, 278 Va. 260, 682 S.E.2d 943, 946 (2009)). Thus, a plaintiff cannot maintain a claim for conversion if the act of dominion exerted by the defendant was wrongful only because it breached the defendant’s duty under a contract.

Here, the plaintiffs allege that the defendants’ “conduct constitutes an unlawful conversion of [their] royalties.” Addison v. CNX Gas Co., 1:10CV00065, Am. Compl. ¶ 86, ECF No. 113; Kiser v. EQT Prod. Co., 1:11CV00031, Am. Compl. ¶ 86, ECF No. 139. The defendants’ duty to pay royalties exists only under their leases with the plaintiffs; they have no such duty under the common law. Thus, if the defendants have, indeed, breached this contractual duty to pay royalties, the plaintiffs’ proper recourse is in an action for breach of contract, not an action for conversion.

Although a district court may not “engage in free-ranging merits inquiries at the certification stage,” it may nevertheless consider merits questions to the extent “they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied.” Amgen Inc., 133 S.Ct. at 1195. The conversion claims are the only claims on which the Addison and Kiser plaintiffs satisfy the commonality and predominance requirements. Thus, if I were to certify these classes, I could do so only as to their claims for conversion.

Rule 23(b)(3) requires me to consider whether a class action is the best method for “fairly and efficiently adjudicating the controversy.” Fed. R. Civ. P. 23(b)(3). I do not believe certifying the putative Addison and Kiser classes as to their claims for conversion would be efficient. Although their conversion claims could be adjudicated in a class proceeding, the class members would still need to bring individual lawsuits in order to be heard on their claims for breach of contract, breach of fiduciary duty, and unjust enrichment. If the conversion claims were the plaintiffs’ primary claims, then elasswide adjudication of those claims might well be an efficient method of resolving these cases. A likelihood of success on the merits might also shift the calculus in favor of certification. However, where the only claim available for class certification is one that is unlikely to prevail on the merits, and where the class members will need to bring individual suits regardless, it would be a waste of time and resources to proceed with a class action for that single, questionable claim. I therefore conclude that the Addison and Kiser plaintiffs have not satisfied Rule 23(b)(3)’s superiority requirement.³⁵

5. Fail-Safe Classes.³⁶

The Fourth Circuit instructed me to consider whether it would be possible to define the plaintiffs’ classes without creating fail-safe classes. EQT Prod. Co., 764 F.3d at 360 n.9. A fail-safe class is a class “that is defined so that whether a person qualifies as a member depends on whether the person has a valid claim.” Messner v. Northshore Univ. HealthSystem, 669 F.3d 802, 825 (7th Cir. 2012). Put another way, a fail-safe class exists where “[e]ither the class members win or, by virtue of losing, they are not in the class and, therefore, not bound by the judgment.” Randleman, 646 F.3d at 352. CNX argues that the class proposed by the Hale plaintiff is a fail-safe class. I disagree,³⁷

The proposed class includes two groups of CBM claimants, identified by the defendants as deemed lessors: (1) those for whom the defendants have applied for the release of funds, and (2) those who actually received distributions of funds between the plaintiffs’ respective dates of filing and January 1, 2016. CNX argues that the Hale class is a fail-safe class because “by assuming that [House Bill] 2058 confers CBM ownership rights on every putative class member, Plaintiffs’ proposed class definitions allow each class member to prevail on his or her ownership claims (which claims remain in the ... complaint[ ]).” Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp’n Class Certification 44 n.39, ECF No. 463. However, even if ownership claims remain in the complaint,” id. the class has not been certified as to these ownership claims. The Hale plaintiff is seeking class treatment only of his claims for underpayment of royalties, so the classwide litigation will occur only with respect to these claims.

I need not consider whether House Bill 2058 does, in fact, “confer[ ] CBM ownership rights on every putative class member,” id. because the answer has no impact on the fail-safe question. The class is defined to include gas claimants whom CNX has “identified ... in filings with the [Virginia Gas and Oil] Board as ‘unleased’ owners of the gas estate interests.” Id. Pl.’s Revised Class Definitions 1, ECF No. 481. It is therefore theoretically possible that a person identified as an un-leased owner—and therefore included in the class—is not actually a CBM owner under the law. In that case, the person would lose on the ownership claims, but—because she was identified as an unleased owner—would still be bound by the judgment, On the other hand, if House Bill 2058 does confer CBM ownership rights on the unleased owners identified by CNX, their ownership claims would now be moot, and would not properly be a part of the litigation. In any event, CNX has not briefed this argument, and I find it neither necessary nor appropriate to decide this merits question here. Most important, as I note above, is the fact that the plaintiff has not sought classwide treatment of the ownership claims, I therefore conclude that the Hale and Adair classes do not constitute impermissible fail-safe classes.

6, Statutes of Limitations.³⁸

EQT argues that, if the Adair class is certified, it must be defined to exclude claims for which the applicable statute of limitations period has expired. Kiser v. EQT Prod. Co., 1:11CV00031, Def.’s Mem. Opp’n Class Certification 20-23, ECF No. 352. Adair filed his lawsuit on June 15, 2010. Thus, EQT asserts, with respect to any claims for breach of fiduciary duty, the class must be defined to exclude claims arising more than two years prior to the date of filing; with respect to claims for conversion, more than five years prior; and with respect to claims for unjust enrichment, more than three years prior. Id. at 21.

CNX argues that the Hale plaintiff cannot satisfy the predominance requirement because “[flor each class member, CNX is entitled to present evidence on when the injury .,. occurred, which will require separate analyses to determine whether each action is time barred under the relevant statute of limitations.” Hale v. CNX Gas Co., 1:10CV00069, Def.’s Mem. Opp’n Class Certification 34, ECF No. 463. I disagree with both of these assertions.

As I note above, the Hale and Adair plaintiffs have proposed class definitions that encompass two groups of CBM claimants: (1) those for whom the defendants have applied for release of funds, pursuant to Virginia Code § 46.1—361.22:2(a), and (2) those who actually received distributions of funds between the plaintiffs’ respective dates of filing and January 1, 2016. There is no statute of limitations issue with respect to either group. Group (1) includes only those claimants for whom the defendants applied for release of funds pursuant to a statute that did not take effect until July 1, 2015—well after either suit was filed. Group (2) includes only those claimants who actually received royalty payments after the dates of filing (for the Hale class, September 23, 2010; for the Adair class, June 16, 2010). Because the proposed classes are defined to include only those claimants whose claims accrued after the original lawsuits were filed, no further modification is necessary to comport with the statutes of limitations. Moreover, because the classes include only gas owners whose claims are not time-barred, I decline to find that the plaintiffs have failed to satisfy the predominance requirement due to the defendants’ need to adduce individualized evidence.

7. Class Action Fairness Act.³⁹

The Class Action Fairness Act (“CAFA”) provides that a district court has original jurisdiction over class actions where the class has more than one hundred members, the parties are minimally diverse, and the amount in controversy exceeds five million dollars. Standard Fire Ins. Co. v. Knowles, 568 U.S. 588, 133 S.Ct. 1345, 1348, 185 L.Ed.2d 439 (2013) (citing 28 U.S.C. § 1832(d)(2) & (d)(5)(B)). The CAFA is not the exclusive source of a district court’s jurisdiction over a class action suit, however; a district court may also hear a class action by way of its diversity jurisdiction. 28 U.S.C. § 1332(a)(1) (providing that district courts have original jurisdiction over all civil actions where the amount in controversy exceeds $75,000 and the parties are completely diverse). The Hale plaintiff alleges that this court has jurisdiction over these cases pursuant to both § 1332(a) and § 1332(d)(2). Hale v. CNX Gas Co., 1:10CV00059, Am. Compl. ¶ 13, ECF No. 166.

CNX now asserts that “the citizenship of the putative class members likely would mandate or permit the Court to decline jurisdiction under CAFA’s exceptions.” Id. Def.’s Mem. Opp’n Class Certification 16, ECF No. 463. CNX notes two exceptions that it asserts would “likely” apply: the “home-state” exception and the “discretionary” exception. Id. at 16-17. Under the “home-state” exception, a district court must decline to exercise jurisdiction over a class action in which at least two-thirds of the putative class members, as well as the primary defendants, are citizens of the state in which the action was originally filed. 28 U.S.C. § 1332(d)(4)(B). Under the “discretionary” exception, a district court may decline to exercise jurisdiction over a class action in which “greater than one-third but less than two-thirds” of the putative class members, as well as the primary defendants, are citizens of the state in which the action was originally filed. Id. at § 1332(d)(3). One or both of these exceptions are “likely” to apply, CNX contends, because “citizenship of the class[ ] likely is at least one-third Virginia citizens ... and potentially ... two-thirds.” Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp’n Class Certification 18, ECF No. 463.

I decline to decide, on the present motion, whether either of these exceptions apply. The Hale plaintiff has met his burden of fulfilling the requirements for class certification under Rule 23. If CNX wishes to challenge this court’s subject matter jurisdiction over this class action suit, it is free to do so in a separate motion. However, I will not decline to certify the class based on the defendant’s speculation that an exception to CAFA might possibly apply.

As I have previously noted, Rule 23(a) requires a prospective class to satisfy four prerequisites: numerosity, commonality, typicality, and adequacy of representation.⁴⁰ Fed. R. Civ. P. 23(a). The defendant has not objected to, and I find no problem with, the numerosity of the proposed class. Numerosity requires that a class be so large that “joinder of all members is impracticable.” Fed. R. Civ. P. 23(a)(1). The class as proposed by the plaintiff includes all gas owners who are party to 3,591 individual leases. Even where I have modified the class, see infra at III.B.l.b, the modified class includes all gas owners who are party to 3,388 individual leases. The plaintiff has clearly fulfilled the numerosity requirement.

As stated earlier, I will address both typicality and adequacy under the umbrella of the typicality requirement.⁴¹ Typicality requires that the claims raised by the class representative be “typical of the claims ... of the class.” Fed. R. Civ. P. 23(a)(3). I find that Adkins satisfies the typicality requirement as to her claims for conversion, but not as to her claims for breach of contract. That said, I believe the proposed class definition can be modified so that this class would be certifiable under a different class representative.

As I note above, see supra at IILAl.b, a typicality analysis requires the court to examine the elements of the claims, examine the facts on which the plaintiff would rely to prove those claims, and determine “the extent to which those facts would also prove the claims of the absent class members.” Deiter, 436 F.3d at 467. A claim for conversion requires proof of an “act of dominion wrongfully exerted over property in denial of, or inconsistent with, the owner’s rights.” Simmons, 544 S.E.2d at 679. A defendant is liable for conversion where its conduct breaches a duty, or infringes on an owner’s rights, arising under the common law. Condo. Servs., Inc., 709 S.E.2d at 171. Because all gas owners possess the same rights of ownership under the common law, and because all the putative class members own gas interests in the same gas field, the facts on which Adkins would theoretically rely to prove her claim of conversion would necessarily prove the claims of the class members as well. Adkins therefore satisfies the typicality requirement with respect to her claims for conversion.

The elements of a claim for breach of contract are “(1) a legal obligation of a defendant to the plaintiff, (2) a violation or breach of that right or duty, and (3) a consequential injury or damage to the plaintiff.” Brown, 467 S.E.2d at 807 (internal quotation marks and citation omitted). Adkins contends that her claims are typical because “her leases contain the same royalty payment terms as the overwhelming majority of the Class leases.” Adkins v. EQT Prod. Co., 1:10CV00041, PL’s Mem. Supp. Class Certification 14, ECF No. 357. The defendant, by contrast, argues that Adkins cannot fulfill this requirement because she holds only two of the multiple lease types encompassed by the proposed class. Id. Def.’s Mem. Opp’n Class Certification 23, ECF No. 361.1 agree.

Plaintiffs “cannot advance a single collective breach of contract action on the basis of multiple different contracts.” Broussard, 155 F.3d at 340. Here, the putative class members and Adkins hold multiple different contracts. The plaintiffs expert has identified 3,591 leases that fall within the class definition proposed by the plaintiff. Of those, 3,388—94.3 percent—are “Type A” leases that contain identical royalty payment language. Adkins v. EQT Prod. Co., 1:10CV00041, PL’s Mem. Supp. Class Certification 10, ECF No. 357; id., Hoge Deck 3, Ex. O, ECF No. 358-19. The remaining leases are organized into Types “B” through “T” and contain variations on this royalty payment language. Id. Adkins herself holds two types of leases: “Type A” and “Type C.” Id. Def.’s Mem. Opp’n Class Certification 23, ECF No. 361.

The plaintiff characterizes the variations among the leases as “non-material,” id. PL’s Mem. Supp. Class Certification 10, ECF No. 357, but after reviewing the report prepared by the plaintiffs expert, I do not agree. Although some variations are substantially similar, others are not. The expert categorized the class leases based on their different prescribed methods of calculating royalties: some provide for calculation of royalties based on the defendant’s “proceeds,” some provide for calculation based on “wholesale market value,” and some provide for calculation based on “sales price,” to name just a few examples. Id., Hoge Deck 12-18, Ex. 0, ECF No. 358-19. The appropriate basis for royalty calculations is an issue on which the plaintiff seeks certification, but the leases that fall within the proposed class contain different provisions governing this issue. Where the putative class members have signed contracts containing “materially different ... language,” class certification is inappropriate. Broussard, 155 F.3d at 340. Here, I believe the differences between the putative class members’ leases raise the possibility that EQT’s conduct constituted a breach of contract with some class members, but not with others. See id. I therefore find that Adkins’ claim for breach of contract is not sufficiently typical of the claims of the class as proposed by the plaintiffs.

However, I believe a modified version of the proposed class could be certified on the breach of contract claim under a different class representative. As I explain above, the class as proposed by the plaintiff is problematic because it includes different types of leases, which could result in class members being subject to different entitlements. A class consisting only of the holders of the 3,388 identical “Type A” leases, however, would avoid the problems identified above. In addition, Adkins’ service as a class representative is problematic because she holds two types of leases. This problem would be solved by identifying a representative who holds only “Type A” leases or, alternatively, by Adkins choosing to drop from this particular proceeding any breach of contract claims arising under her “Type C” leases. I therefore find that the Rule 23(a) requirements are satisfied by a class defined to include only the 3,388 identical “Type A” leaseholders and a class representative who raises a breach of contract claim only under a “Type A” lease.

The Fourth Circuit previously concluded that the ascertainability requirement had not been met with respect to the Adkins class. The class was defined to “include only those gas owners whose leases are ‘silent’ with respect to the deduction of costs,” and the court found this “silence” requirement to be unclear and open to interpretation. EQT Prod. Co., 764 F.3d at 368. The plaintiff now asserts that she has addressed the court’s concerns by defining the class with greater specificity, making its putative members “readily identifiable” as required by the Fourth Circuit. Id. at 358 (internal quotation marks and citation omitted). I agree.

Unlike the other four proposed classes, the Adkins class consists only of known, leased gas owners who have actually received royalty payments. The updated definition further limits the class to include only those owners whose leases neither expressly authorize nor expressly prohibit deductions for the costs of “gathering, treating, compression, dehydration, processing, and/or transportation.” Adkins v. EQT Prod. Co., 1:10CV00041, PL’s Mem. Supp. Class Certification 7, ECF No. 357. Although identifying the individual gas owners whose leases fall under this specific definition might have been onerous, the plaintiffs expert appears to have already done this. Id. at 6.1 believe the detailed definition, particularly in light of the work done by the plaintiffs expert, adequately resolves the concerns raised by the court of appeals regarding the ascertainability of the class.

Like the four plaintiffs discussed above, Adkins seeks certification of her class under Rule 23(b)(3). My inquiry as to each issue will therefore be the same as above. See supra at III.A.3. First, I must determine whether the defendant engaged in practices that are common to all class members or, alternatively, whether the plaintiff has identified a question capable of generating a common answer. If so, I must next determine whether these common practices or questions bear on the ultimate question of whether the defendant underpaid royalties. If the answer to this second question is “yes,” then I must determine whether these common practices or questions are “sufficient to ensure that the class members’ common issues [will] predominate over individual ones.” EQT Prod. Co., 764 F.3d at 366. As part of this analysis, I must “analyze each of the plaintiff’s] claims to determine whether any of the distinct elements of those actions might affect the predominance of common questions.” Id. at 367 n.19.

The Adkins plaintiff identifies two specific practices related to the defendant’s calculation and payment of royalties that she asserts are common to all putative class members: (1) deducting certain “marketability” costs from class members’ royalties and (2) calculating royalties based on improperly low prices. Adkins v. EQT Prod. Co., 1:10CV00041, PL’s Mem. Supp. Class Certification 12-13, ECF No. 367. I address each practice in turn.⁴²

a. Deduction of Marketability Costs.

The members of the putative Adkins class include gas owners in the Nora Field who entered into individual leases with EQT and who have received payments since 2005. The class members’ leases neither expressly permit nor expressly prohibit EQT from taking deductions for the costs of gathering, treating, compressing, dehydrating, processing, and transporting the gas. The plaintiff contends that EQT’s deductions for these costs—“the costs it incurs in placing the gas into a marketable location and condition”—is improper. Id. at 13.

The fact that EQT employs a uniform deduction scheme to all members of the putative class appears to be undisputed. The class is limited to lessors who own interests in the Nora Field. Although EQT has applied two different deduction schemes to the Nora Field since 2006, both the prior and the current schemes were generally uniform across the field. See id. at 6 (citing id., Bergonzi Dep. 25-26, Ex. L, ECF No. 221-12). EQT did employ different deduction schemes where individual leases so required, but the Adkins class has been defined to eliminate these individuals. The evidence indicates, and EQT does not assert otherwise, that unless a lease expressly permitted or expressly prohibited certain deductions, EQT applied a uniform deduction scheme. See id at 7 (citing id., Atkinson Dep. 150-51, Ex. A, EOF No. 221-1). The taking of these deductions is therefore a common practice.

This common practice clearly bears on the ultimate question of whether EQT underpaid royalties. If these deductions were not appropriate under the terms of the lease or governing law, then by taking them, EQT has been underpaying royalties. By contrast, if these deductions are appropriate, then EQT has not been underpaying royalties. I must finally determine, then, whether this common question—whether it is proper for EQT to take such deductions from the members of the putative Adkins class—predominates over any individual issues. I conclude that the common question does not predominate over individual questions as to the claim for breach of contract, but that it does predominate as to the claim for conversion.

In conducting a predominance inquiry, I must consider the elements of the plaintiffs claims. The Adkins plaintiff raises claims for breach of contract and conversion. As I have noted above, the elements of a claim for breach of contract are (1) a legal obligation by the defendant to the plaintiff, (2) a breach of that duty, and (3) a consequential injury to the plaintiff. Brown, 467 S.E.2d at 807. But for the fact that the leases included in the putative Adkins class are silent as to the deduction of costs, I believe the common question of the propriety of those deductions would predominate over individual issues. Because the leases in the putative class are silent as to deductions, however, I must conclude that individual issues would predominate.

Where an agreement is silent as to a certain issue, parol evidence may be admitted to “prove the existence of additional terns to an agreement,” so long as such terms are “not inconsistent with the express terms of the written instrument.” Jim Carpenter Co. v. Potts, 255 Va. 147, 495 S.E.2d 828, 833 (1998). This exception to the parol evidence rule, called the partial integration doctrine, recognizes that “the final form of a contract between parties may not ... accurately reflect the course of dealing between parties based on their complete agreement.” Id. (citing High Knob, Inc. v. Allen, 205 Va. 503, 138 S.E.2d 49, 52 (1964)). Here, the leases encompassed by the class definition are silent as to the deduction of costs, and deducting “marketability” costs is therefore not inconsistent with the express terms of the leases. Course of performance evidence is thus admissible to ascertain the parties’ intentions regarding deductions and, ultimately, to determine whether EQT’s deduction practices violated its obligations to its lessors. See also Video Zone, Inc. v. KF & F Props., L.C., 267 Va. 621, 594 S.E.2d 921, 924 (2004) (“[Uncertain rights of parties may be determined and fixed by their practical dealings with each other.”).

The plaintiff argues that because “EQT responded to [lessors’] inquiries [about deductions] ... in categorical terms, with no indication that its treatment of one lessors’ [sic] deductions was any different from its treatment of another,” its course of performance can be assessed on a classwide basis. Adkins v. EQT Prod. Co., 1:10CV00041, PL’s Mem. Supp. Class Certification 17, ECP No. 357. However, I find this to be an untenable reach and unsupported by the evidence. Even if EQT did have a stock response to the lessors’ questions about deductions, this does not necessarily mean that its subsequent correspondence with each lessor was identical or even substantially similar.

It thus appears that course of performance evidence can only be adequately adduced on a “lease-by-lease basis.” Id. Def.’s Mem. Opp’n Class Certification 27, ECP No. 361. This individualized evidence may very well show that EQT had different obligations to each class member, which, in turn, means there is no single, classwide answer to the question of whether EQT’s deduction practices were proper. Because I believe the need for individualized proof predominates over the common question, I cannot find that the Adkins plaintiff has satisfied Rule 23(b)(3)’s predominance requirement on this issue as to her claim for breach of contract.

As I have noted above, a claim for conversion requires proof of an “act of dominion wrongfully exerted over property in denial, or inconsistent with, the owner’s rights.” Simmons, 544 S.E.2d at 679. A defendant is liable for conversion where its conduct breaches a duty, or infringes on an owner’s rights, arising under the common law. Condo. Servs., Inc., 709 S.E.2d at 171. Because the rights of the putative class members arise under the common law, all members of the class are subject to the same entitlements, regardless of the terms of their leases. Moreover, because EQT’s duty arises under the common law rather than the leases, course of performance evidence is not relevant in determining the parties’ rights or obligations. The answer to the question of whether EQT’s deductions are proper under the common law will therefore be the same for all members of the class.

Although the actual royalties to which each gas owner is entitled will vary, because I adopt a bifurcated approach to liability and damages, I believe the common question of whether the deductions taken by EQT were proper predominates over these individual damages calculations. See supra at III. A.3.f.ii. Moreover, “the need for individualized proof of damages does not necessarily preclude class certification so long as common issues continue to predominate over individual issues.” Leinhart, 255 F.3d at 147. While it is true that individual damage calculations may defeat predominance where “proof of damages is essential to liability,” id. or “where individual damages issues are especially complex or burdensome,” Stillmock, 385 Fed.Appx. at 273, neither of these scenarios apply in this case. Instead, the “overarching issue by far is the liability issue,” id.: if EQT is, in fact, liable for taking improper deductions, all class members were injured by this conduct. Moreover, if EQT retains records of the royalty payments it has made and the deductions it has taken—which it appears to have done, see Adkins v. EQT Prod. Co., 1:10CV00041, Atkison Dep. 40, Ex. A, ECF No. 221-1—calculating damages for each leaseholder should be a relatively simple matter. I therefore find that the Adkins plaintiff has satisfied the commonality and predominance requirements as to her conversion claim on the question of whether EQT properly deducted “marketability” costs from the class members’ royalty payments.⁴³

The Adkins plaintiff also alleges that EQT has failed to “obtain[ ] the highest price obtainable, as it is required to do pursuant to its duty to market the gas,” and that it therefore underpaid royalties. Id. PL’s Mem. Supp. Class Certification 14, ECF No. 357. Specifically, she asserts that EQT improperly based its royalty calculations on sales prices that were lower than the published index prices. EQT has not responded to this argument as raised by this plaintiff, but it has elsewhere asserted in response to the same argument—by a different plaintiff in a related case—that it has no obligation to pay royalties based on the highest price available and that it is obligated only to act as a reasonably prudent operator. See Kiser v. EQT Prod. Co., 1:11CV00031, Def.’s Mem. Opp’n Class Certification 10, ECF No. 352. The fact that, since 2007, EQT has calculated royalties for its Nora Field lessors based on sales prices, rather than published index prices, appears to be undisputed. Adkins v. EQT Prod. Co., 1:10CV00041, Bergonzi Dep., Ex. 4, ECF No. 227-5; see also id. Def.’s Mem. Opp’n Class Certification 5, ECF No. 361; id. PL’s Mem. Supp. Class Certification 6, ECF No. 357. Because all members of the putative class are Nora Field lessors, EQT’s use of sales prices in calculating royalty payments is a common practice across the class.

The issue, then, is whether this common practice is proper, which involves two questions. The court previously held that “Virginia courts would impose an implied duty to market on lessees under oil and gas leases.” Legard, 2011 WL 86598, at *10, adopted in Legard v. EQT Prod. Co., No. 1:10CV00041, 2011 WL 4527784, at *1 (W.D. Va. Sept. 28, 2011). The first question is therefore whether this “duty to market” obligates EQT to pay royalties based on the highest price obtainable, as the plaintiff claims, or whether this duty merely obligates EQT to act as a reasonably prudent operator. The second is whether its practice for calculating royalty payments—using sales prices—meets this obligation. It is clear that both of these common questions bear on the ultimate issue of whether EQT underpaid royalties. The answer to the first question defines EQT’s obligation to the putative class members; the answer to the second determines whether EQT did, in fact, underpay royalties.

I must next decide whether these common issues predominate and, again, must consider the elements of the plaintiffs claims. I conclude that the common questions predominate over individual issues as to the claim for breach of contract, but only for the modified class I described above, see supra at III. B.l.b: the holders of the 3,388 leases that contain identical language regarding royalty payments. I also conclude that the common questions predominate as to the claim for conversion as to the entire proposed class.

As noted above, the elements of a claim for breach of contract are (1) a legal obligation by the defendant to the plaintiff, (2) a breach of that duty, and (3) a consequential injury to the plaintiff. Brown, 467 S.E.2d at 807. I find that common questions predominate as to this claim because all members of the modified class are subject to the same entitlements: the entitlements arising from EQT’s implied duty to market.⁴⁴ Because EQT uses the same methodology to calculate royalty payments for every member of the modified class, the common questions described above—whether EQT’s “duty to market” requires it to calculate royalty payments based on the highest price obtainable, and whether its practice of calculating royalty payments based on sales prices fulfills that duty—can be answered for every member of the modified class in one fell swoop. There will be no need to consider individualized evidence as to the breach because every class member’s lease is the same. Moreover, although the actual royalties to which each gas owner is entitled will vary, I believe the bifurcated approach I adopt here allows common questions to predominate over these individual damage calculations. See supra at III.A.3.f.ii. I therefore conclude that the Adkins plaintiff has satisfied the commonality and predominance requirements as to her breach of contract claim on the question of whether EQT calculated its royalty payments based on improperly-low prices.

As to the conversion claim, for the reasons described supra at III.B.3.a, I find that the Adkins plaintiff has satisfied the commonality and predominance requirements on the question of whether EQT calculated its royalty payments based on improperly low prices.⁴⁵

c. The First Marketable Product Rule and “Marketability.”

Under the First Marketable Product Rule, oil and gas lessees—like EQT—have an “implied duty to bear the cost of putting the oil and gas in a marketable condition after it is removed from the well, including common postproduction expenses for gathering, compressing, and dehydrating oil and gas.” EQT Prod. Co., 764 F.3d at 364 (citing sources). Whether the Virginia courts follow the First Marketable Product rule is currently an open question, but I have concluded that they are likely to do so.⁴⁶ See Legard, 2011 WL 86598, at *10-11 (magistrate judge’s Report and Recommendation, holding that Virginia courts would follow the First Marketable Product Rule), adopted by 2011 WL 4527784, at *1; Adkins v. EQT Prod. Co., No. 1:10CV00031, 2011 WL 6178438, at *5 (W.D. Va. Dec. 13, 2011) (noting that “Virginia courts would follow the ‘first marketable product’ rule”).

In light of my conclusion, the plaintiff now seeks certification on a third question: when, exactly, CBM first becomes “marketable” within the meaning of the Rule. She contends that EQT’s gas is marketable when two conditions are met: quality standards and location, According to the plaintiff, gas is in “marketable condition” when it “meet[s] the natural gas quality standards of the interstate pipelines” and has been “delivered into the transmission pipelines.” Adkins v. EQT Prod. Co., 1:10CV00041, Graham Rep. 3, Ex. H, ECF No. 858-8; see also id. Pl.’s Mem. Supp. Class Certification 13, ECF No. 357. By contrast, the defendant contends that its gas is marketable once it has met the required quality standards, without regard to its location.

I do not believe this question is itself an appropriate basis for class certification. Because it amounts to a disagreement regarding Virginia law, it is more properly raised at the merits stage of the litigation. At the class certification stage, I may consider “[m]erits questions ... only to the extent ... they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied.” Amgen Inc., 133 S.Ct. at 1195, Addressing this particular merits question is tantamount to addressing a motion for summary judgment, and, importantly, it is not necessary to address this question in order to make my Rule 23 determinations. As stated above, I will certify the proposed version of the Adkins class as to the conversion claim on the other two questions raised by the plaintiff, and conditional on the appointment of an appropriate class representative, I will certify the modified version of the Adkins class as to the breach of contract claim on the second question. The parties are free to renew their arguments regarding the applicable law at the appropriate time.

Rule 23(b)(3) also requires that plaintiffs demonstrate that “proceeding as a class ⅛ superior to other available methods for fairly and efficiently adjudicating the controversy.’” EQT Prod. Co., 764 F.3d at 371 (quoting Fed. R. Civ. P. 23(b)(3)). I previously concluded, and the Fourth Circuit agreed, that multiple factors weigh in favor of class action litigation in this case. See EQT Prod. Co., 764 F.3d at 371 (citing Adair, 2013 WL 5429882, at *40). The Fourth Circuit has now instructed me to additionally “consider how the dominance of state law issues may affect the suitability of this litigation in a federal forum and what state-law mechanisms may be available to resolve the underpayment claims as an alternative to a class action.” Id. The plaintiff maintains that “a class action is the superior (and indeed, the only practical) method for adjudicating these claims.” Adkins v. EQT Prod. Co., 1:10CV00041, Pl.’s Mem. Supp. Class Certification 19, ECF No. 357. The defendant argues, however, that “[bjecause of these substantial state interests that are involved and the multiple Erie guesses that would be required, this case should not be litigated in a federal forum,” Id., Def.’s Mem. Opp’n Class Certification 33-34, ECF No. 361.

It is true, as the defendant states, that “[t]his case involves ... questions of Virginia law that have never been addressed by any court in Virginia.” Id. at 33. The applicability and intricacies of the First Marketable Product Rule in Virginia pose questions of state law that are highly relevant to the merits of the plaintiffs’ claims. However, the fact that federal courts must occasionally make an educated guess as to how a state court might rule on a given issue is the reality of the American judicial system under Erie R.R. Co. v. Tompkins, 304 U.S. 64, 58 S.Ct. 817, 82 (1938). Certainly, the fact that a case necessitates such a federal prediction is not, by itself, grounds for refusing to certify a class.

Furthermore, I do not believe the prevalence of state law issues in this ease affects the suitability of this litigation in a federal forum. There are other factors weighing heavily in favor of litigating this action in a federal court, which I discuss supra at III. A.4.a. This is a diversity case, and as such, state law issues are obviously dominant—but a federal class action is nevertheless the superior means of resolving the dispute. For the reasons described above and supra at III.A.4.a, I find that the Adkins plaintiff has satisfied the superiority requirement of Rule 23(b)(3).⁴⁷

5. Fail-Safe Class.

I must also consider whether it would be possible to define the Adkins class without creating a fail-safe class, A fail-safe class exists where “[e]ither the class members win or, by virtue of losing, they are not in the class and, therefore, not bound by the judgment.” Randleman, 646 F.3d at 362; see also supra at III.A.5. The defendant argues that the class proposed by the Adkins plaintiff is a fail-safe class. I disagree.

The proposed class includes lessors from the Nora Field who have received at least one royalty payment from EQT since June 8, 2006 and whose leases neither expressly authorize or expressly prohibit the taking of deductions. EQT argues that this is a fail-safe class because

In order to determine if any particular lessor is in the class, the Court will have to decide whether the lease allows deductions. If the lease allows deductions, the lessor is not in the class and no judgment can be entered against him. If the lease does not allow deductions, the lessor is in the class and—by virtue of being in the class—has won on the issue of deductions.

Adkins v. EQT Prod. Co., 1:10CV00041, Def.’s Mem. Opp’n Class Certification 33, ECF No. 361. However, I believe this is a miseharacterization of the class. The class includes leases that do not expressly authorize or expressly prohibit deductions, Whether a lease contains express language permitting or prohibiting deductions and whether a lease actually permits or prohibits deductions are two different issues that the defendant seeks to conflate.

A lease is included in the class because of its lack of express language regarding deductions, not because it actually permits or prohibits deductions. The finder of fact could find that a lease that does not expressly authorize (or prohibit) deductions nevertheless allows EQT to take certain deductions. In that event, contrary to EQT’s assertion, the lease would remain in the class—because its express language would remain unchanged—EQT would prevail on that issue, and the leaseholder would not be entitled to damages for underpayment of royalties. The lease is included in the class due to its express language; its membership is not contingent on the fact-finder’s conclusion. Likewise, the finder of fact could find that such a lease does not allow EQT to take certain deductions. In that event, the leaseholder would, indeed, be in the class and prevail on that issue. However, again contrary to EQT’s assertion, he would not prevail on the issue merely by virtue of being in the class—he would prevail on the issue because the fact-finder happened to make a finding in his favor.

IV. Conclusion.

For the reasons stated, I hereby find as follows:

The putative Hale class is certified for all claims as to the following issues: (1) allegedly excessive deductions; (2) royalties based on allegedly improperly low prices; (3) deduction of severance taxes; and (4) request for an accounting. The class is not certified as to any issues not listed above. Damages ■will be addressed in later proceedings as necessary.

The putative Addison class is not certified as to any issues.

The putative Adair class is certified for all claims as to the following issues: (1) royalties based on allegedly improperly low prices; (2) deduction of severance taxes; and (3) request for an accounting. The class is not certified as to any issues not listed above. Damages will be addressed in later proceedings as necessary.

The putative Kiser class is not certified as to any issues.

The putative Adkins class is certified for the conversion claims as to the following issues: (1) allegedly improper deduction of marketability costs and (2) royalties based on allegedly improperly low prices. A modified version of the class is conditionally certified, pending the appointment of an appropriate representative, for the breach of contract claims as to the following issue: royalties based on allegedly improperly low prices. The class is not certified as to any issues not listed above. Damages will be addressed in later proceedings as necessary.

Separate orders will be entered in each case in accord with the foregoing.

1 . While I anticipated the parties might wish to present live witness testimony at the joint hearing on the renewed motions for class certification, see Adair v. EQT Prod. Co., 1:10CV00037, Scheduling Order ¶ 8, ECF No. 560, the parties relied exclusively upon deposition testimony and various declarations, along with exhibits, all of which I have considered.

2 . During oral argument, plaintiffs' counsel also asserted that the issue of whether the plaintiffs are entitled to an accounting is a sixth issue that should be included with this list, because the request for an accounting is predicated on separate causes of action such as conversion, unjust enrichment, breach of contract, and trespass. This position constitutes a departure from the plaintiffs’ brief, which lists the five legal questions above as common issues that are subordinate to the accounting question. See Hale v. CNX Gas Co., 1:10CV00059, Pl.'s Mem. Supp. Class Certification 10-21, ECF No. 459; Kiser v. EQT Prod. Co., 1:11CV00031, Pl.’s Mem. Supp. Class Certification 10-14, ECF No. 350.

3 . Because the briefs in this matter were written before January 1, 2016, the briefs contemplated that as a future date.

4 , At the joint oral argument on all of the class certification motions, counsel for the plaintiffs acknowledged that the proposed accounting would occur only if the factfinder found, in Phase 1 of the litigation, that the defendants’ "specific challenged practices" were "improper.” Adair v. EQT Prod. Co., 1:10CV00037, Hr'g Tr. 109:20-24, Sept. 18, 2015, ECF No. 581.

5 . CNX also alleges that the Hale and Addison plaintiffs "cannot satisfy the adequacy and typicality requirements” of Rule 23(a). Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp'n Class Certification 19 n.19, ECF No. 463.

6 . For example, EQT estimates that the severance taxes at issue amounted to only $191,941 between Februaiy 1995 and October 2011. Kiser v. EQT Prod. Co., 1:11CV00031, Def.'s Mem. Opp’n Class Certification 8, ECF No. 352.

7 . Where a plaintiff seeks class certification under Rule 23(b)(3), the predominance requirement supersedes Rule 23(a)’s commonality requirement. Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 609, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997). I discuss commonality and the predominance requirement more fully infra at III.A.3.

8 . CNX argues that the Hale and Addison plaintiffs cannot satisfy the numerosity requirement because their proposed classes are not ascertainable. Hale v. CNX Gas Co., 1:10CV00059, Def.'s Mem. Opp’n Class Certification 16 n,13, ECF No. 463.1 disagree, and I address the ascertainability requirement infra at III.A.2. Moreover, CNX has not disputed the plaintiffs’ assertion that the classes as currently defined would comprise hundreds of individual gas owners (multiple claimants for each of 529 units), EQT does not contest the numerosity requirement.

9 . These two requirements typically merge with Rule 23(a)’s commonality requirement as well. Broussard, 155 F.3d at 337. However, since the plaintiffs seek certification under Rule 23(b)(3), I address commonality separately as part of my predominance inquiry. See infra at III.A.3.

The Supreme Court has also noted that the adequacy of representation requirement "raises concerns about the competency of class counsel.” Gen. Tel Co. of Sw., 457 U.S. at 157 n.13, 102 S.Ct. 2364. None of the parties have raised such concerns here.

10 . The Hale plaintiff also raises a claim for trespass regarding CNX’s premature production of CBM gas. However, because I deny certification of the Hale class as to this issue, see infra at III.A.3.e.i, I need not address the question of whether Hale fulfills the typicality and adequacy requirements as to this claim.

11 . CNX asserts that it is not a fiduciary and thus owes no fiduciary duty to Hale or any putative Hale class members. See Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp'n Class Certification 22, ECF No. 463. I express no opinion here as to whether CNX and the force-pooled gas owners do or do not have a fiduciary relationship.

12 . CNX also argues that Hale’s claims are not typical of the class because the class ”include[s] working interest owners who are not similarly situated” to him. Hale v. CNX Gas Co., 1:10CV00059 Def.'s Mem. Opp’n Class Certification 19 n.19, ECF No. 463. However, in Hale’s Proposed Revised Class Definition, he expressly excludes from the class “any person who holds a working interest in a well operated by CNX in Virginia.” Id. Pl.’s Revised Class Definitions 2, ECF No. 481. This objection is therefore moot.

13 . EQT makes no assertions regarding Adair’s satisfaction of the adequacy and typicality requirements.

14 . CNX argued that the plaintiffs’ proposed classes were not ascertainable at the time the plaintiffs initially moved for certification of the classes. At this time, however, the deadline for identifying claimants imposed by the legislature has passed, and for the reasons described above, the members of the proposed classes are readily identifiable.

The defendants correctly point out that, because the revised statute allows the Board to extend the deadline at the request of the gas well operators, not all gas claimants would have necessarily been identified by the statutory deadline. However, because the plaintiffs' proposed class definitions include only the claimants identified as of a particular date—rather than all possible claimants—the class members are readily identifiable regardless of any extensions granted by the Board. Furthermore, at this point, the defendants have had more than a full extra year to comply with the statute's requirement. Allowing the defendants to block certification of these classes by delaying their compliance with the law—no matter how necessary and understandable the delay—would undermine the objectives of fairness and judicial economy that underlie the very concept of a class action suit. See Advisory Committee Notes on 1966 Amendment, Subdivision (b)(3), Fed. R. Civ. P. 23.

15 . The decision of the Virginia Supreme Court in Swords Creek Land Partnership, 762 S.E.2d 570, also simplifies the ascertainability analysis by eliminating those who own coal rights, but not gas rights, as potential class members.

16 . The court of appeals previously instructed this court to "determine the number of potential class members who have obtained their interest in the gas estate.” EQT Prod. Co., 764 F.3d at 360. Because the classes have been defined in terms of actions that the defendants will have actually taken prior to or within a specific date range, determining the number of potential class members will be a relatively simple endeavor.

17 . The plaintiffs for the Hale and Addison classes filed identical renewed motions and memoranda in support of class certification. For purposes of simplicity, I generally cite to the filings in the Hale case here. The identical Memorandum in Support of Class Certification for the Addison case can be found at Addison v. CNX Gas Co., 1:10CV00065, ECF No. 377.

18 . The plaintiffs for the Adair and Kiser classes likewise filed identical renewed motions and memoranda. I generally cite to the filings in the Kiser case here; the identical memorandum for the Adair case can be found at Adair v. EQT Prod. Co., 1:10CV00037, ECF No. 566.

19 . Discussion of this issue is included alongside discussion of issue (2): calculation of royalties based on improperly low prices.

20 . CNX states that because its deductions differ "depending on where it sells the CBM,” its "deduction calculations sometimes vary between and even within wells during different time periods." Hale v. CNX Gas Co., 1:10CV00059, Def.’s Mem. Opp’n Class Certification 24, ECF No. 463 (quoting EQT Prod. Co., 764 F.3d at 366-67). However, because these deduction rates are applied equally to all members of the class, I do not believe they preclude a finding of commonality. The common practice in question here is CNX’s application of a standardized deduction scheme.

21 . The Adair plaintiff asserts that EQT "applies uniform deduction rates to its deemed lessors.” Kiser v. EQT Prod. Co., 1:11CV00031, Pl.'s Mem. Supp. Class Certification 13, ECF No. 350. However, the only support cited for this assertion is a letter from EQT to the Virginia Division of Gas and Oil, in which an EQT Vice President relays the company’s monthly post-production expense deductions for the year 2008. Id., West Letter 1, Ex, M, ECF No. 351-17. There is no context indicating that the monthly deduction rates stated in the letter applied uniformly to all deemed lessors, as the plaintiff suggests.

22 , If the Adair plaintiff had shown that his putative class members own gas interests only in the Nora Field, certification on this basis might be appropriate for the same reasons given for the Hale class. If all class members are subjected to the same deduction scheme, EQT’s application of that scheme is a common issue. Moreover, if all class members are subject to the same deduction scheme—as deemed lessors are—determining whether EQT’s application of that scheme produces excessive deductions might well predominate over any individual analyses. However, where the deduction scheme to which the class members are subject varies by gas field, there is no longer a common issue, and individual analy-ses predominate.

23 . In addition, like the Addison plaintiff, the Kiser plaintiff would be unable to fulfill the superiority requirement as to her claim for conversion.

24 . The expert noted that the "overwhelming majority” of the leases contained identical language, but this does not change the fact that the putative class members’ leases can be sorted into twenty variations, Kiser v. EQT Prod. Co., 1:11CV00031, Hoge Decl. 3, Ex. O, ECF No. 351-19.Further-more, the expert sorted more than 3,500 leases into three macro-categories that impose completely different royalty payment obligations. Id.

25 . In addition, there is a question as to whether this claim will prevail at the merits stage of the litigation. A claim for unjust enrichment is based on a theory of implied contract. However, Addison and the members of her putative class are all parties to leases—express contracts—with CNX. "The law will not impose an implied contractual relationship upon parties in contravention of an express contract.” Nedrich v. Jones, 245 Va. 465, 429 S.E.2d 201, 207 (1993) (citing Royer v. Bd. of Cty. Supervisors, 176 Va. 268, 10 S.E.2d 876, 881 (1940)). Thus, where the rights of the parties are defined by an express contract, recovery for a breach of those rights is not available under a theory of implied contract. See id. Here, the rights and obligations of the class members and CNX are defined by the terms of express contracts, namely, their leases.

However, a district court may consider merits questions at the certification stage “only to the extent ... that they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied.” Amgen Inc. v. Conn. Ret. Plans & Tr. Funds, 568 U.S. 455, 133 S.Ct. 1184, 1195, 185 L.Ed.2d 308 (2013). Because the Addison plaintiff has not satisfied the predominance requirement on this claim as to this issue, this particular merits question does not bear on the propriety of Rule 23 certification. I have therefore ruled on the certification question without regard for the merits of this claim.

26 . See supra note 25.

27 . Although it remains unclear whether these putative class members own gas interests only in the Nora Field or in all three of EQT’s Virginia fields, see supra at III.A.3.a.ii, on the issue of prices used to calculate royalty payments, this factor is not determinative. Unlike its deductions practices, EQT has not asserted, and there is no evidence, that its practice of calculating royalty payments based on sales prices differs by field. Moreover, there does not appear to be any evidence that the pooling orders governing EQT’s obligations to the members of the Adair class impose disparate obligations on EQT with regard to the applicable price standard.

28 . There is also a question as to whether the claim for unjust enrichment would prevail on the merits. See supra note 25.

29 . If a common practice existed, it is possible that the Kiser class could fulfill the predominance requirement as to the claim for conversion. Even if this were the case, however, it would be unable to meet the superiority requirement. See supra note 23 and accompanying text.

30 . The Addison plaintiffs do not seek certification on this issue.

31 . The Kiser plaintiffs do not seek certification on this issue.

32 . It is unclear whether the pooling orders governing each of the three gas fields contain identical terms regarding the deduction of severance taxes. However, EQT does contend that the pooling orders differ in this regard. Furthermore, even if the pooling orders do differ slightly on this specific issue, the additional analysis required to answer the question as to all class members would not change my conclusion that common issues predominate, given that this practice was applied equally to all members of the class and that answering the question would require examination of, at most, three sets of pooling order terms.

33 . The Addison, Adair, and Kiser plaintiffs do not seek certification on this issue.

34 . The Hale and Addison plaintiffs move for a hearing under Rule 23(d) regarding CNX’s "efforts, or lack thereof, to comply with the 2015 amendments to the Virginia Gas and Oil Act.” Hale v. CNX Gas Co., 1:10CV00059, Pl.'s Mot. for Hearing 1, ECF No. 505; Addison v. CNX Gas Co., 1:10CV00065, Pl.'s Mot. for Hearing 1, ECF No. 409. I will deny these motions as unnecessary under the circumstances.

35 , In so concluding, I do not decide these conversion claims on the merits. Addison, Kiser, and their putative class members are not estopped from raising these claims in any individual lawsuits.

36 . Because I find that the Addison and Kiser plaintiffs have not satisfied the Rule 23 requirements for class certification, I need not consider fail-safe class arguments as they apply to these classes,

37 . EQT does not argue that the class proposed by the Adair plaintiff is a fail-safe class. However, because the definition proposed for the Adair class is virtually identical to the definition proposed for the Hale class, my conclusion as to the Hale class applies equally to the Adair class.

38 . Because I find that the Addison and Kiser plaintiffs have not satisfied the Rule 23 requirements for class certification, I need not consider statutes of limitations arguments as they apply to these classes.

39 . Because EQT does not make any allegations regarding the Class Action Fairness Act, and because the Addison plaintiff does not meet the Rule 23 requirements for class certification, I consider CNX's CAFA argument only as it applies to the putative Hale class.

40 . I discuss commonality and the predominance requirement more fully infra at III.B.3.

41 . See supra note 9 and accompanying text. The defendant here has not objected to, and I find no problem with, the adequacy of representation of the Adkins plaintiff. There has been no evidence or argument either that Adkins cannot "fairly and adequately protect the interests of the class,” Fed. R. Civ. P. 23(a)(4), or that there are any doubts about the "competency of class counsel and conflicts of interest.” Gen. Tel. Co. of Sw., 457 U.S. at 157 n.13, 102 S.Ct. 2364.

42 . The plaintiff also seeks class certification on the purportedly common question of which version of the First Marketable Product Rule applies to this case. I discuss this matter infra at III. B.3.c.

43 . There is a question as to whether this claim will prevail at the merits stage of the litigation. In order to recover on a claim for conversion, "the duty ... breached must be a common law duty, not one existing between the parties solely by virtue of [a] contract.” Condo. Servs., Inc., 709 S.E.2d at 171 (citation omitted). Thus, a plaintiff cannot maintain a claim for conversion if the act of dominion exerted by the defendant was wrongful only because it breached the defendant’s duty under a contract. See also supra at III.AAb.

Here, the plaintiff alleges that EQT's actions "constitute[ ] an unlawful conversion of [her] ... Royalties and Royalty payments.” Adkins v. EQT Prod. Co., 1:10CV00041, Compl. ¶44, ECF No. 1. EQT's duty to pay royalties exists only under its lease with the plaintiff; it has no such duty under the common law. Thus, if EQT has, indeed, breached this contractual duty to pay royalties, the plaintiff’s proper recourse is in an action for breach of contract, not an action for conversion.

However, a district court may consider merits questions at the certification stage "only to the extent ... that they are relevant to determining whether the Rule 23 prerequisites for class certification are satisfied.” Amgen Inc., 133 S.Ct. at 1195. Because I grant certification to the Adkins class on at least one other claim, this particular merits question does not bear on the propriety of Rule 23 certification. I have therefore ruled on the certification question without regard for the merits of this claim.

44 .I do not believe course of performance evidence is admissible or, indeed, relevant to the certification decision on this issue. The parties’ disagreement centers on the obligations contained in EQT’s implied duty to market. This implied duty arises from the mere fact that the gas owners and EQT have entered into gas leases; it does not arise from the language of the leases themselves. Because the meaning of the leases themselves is not at issue, the parties’ underlying intent in ratifying those leases is not relevant, and course of performance evidence is not admissible.

45 . See supra note 43.

46 . The defendant now asks me to revisit this conclusion. I decline to do so. “Where state law is unclear, the [federal] court must predict how the highest court of that state would rule if presented with the issue.” Legard v. EQT Prod. Co., 771 F.Supp.2d 607, 609 (W.D. Va. 2011) (citing Wells v. Liddy, 186 F.3d 505, 527-28 (4th Cir. 1999)). I have made my prediction. EQT has not pointed to, and my research has not revealed, any changes in Virginia state law during the past six years that might render my previous ruling invalid or misguided, Absent such a change in the law, I see no reason to reconsider the question of whether Virginia courts would apply the First Marketable Product Rule.

47 . The Fourth Circuit also suggested I "assess the extent of the defendants' efforts to resolve and pay undisputed claims,” noting that "[a] finding that the defendants have not acted in good faith toward that end may weigh strongly in favor of a finding of superiority of a class action.” EQT Prod. Co., 764 F.3d at 371. However, because the Adkins plaintiff has made no such allegation, I do not consider this possibility here.