The opinion
Case: 20-20663 Document: 00516194200 Page: 1 Date Filed: 02/07/2022
United States Court of Appeals
for the Fifth Circuit United States Court of Appeals
Fifth Circuit
FILED
February 7, 2022
No. 20-20663 Lyle W. Cayce
Clerk
Hess Corporation,
Plaintiff—Appellant/Cross-Appellee,
versus
Schlumberger Technology Corporation,
Defendant—Appellee/Cross-Appellant.
Appeal from the United States District Court
for the Southern District of Texas
USDC No. 4:16-CV-3415
Before Clement, Southwick, and Willett, Circuit Judges.
Leslie H. Southwick, Circuit Judge:
Hess Corporation contracted with Schlumberger Technology
Corporation to provide safety valves for several of Hess’s deep-sea oil wells
in the Gulf of Mexico. Hess experienced problems with the valves, and
Schlumberger recalled them. Hess then attempted to revoke acceptance of
the equipment and sued Schlumberger for breach of contract. After a bench
trial, the district court found that Hess had not proven its revocation claim
and held for Schlumberger. We find no reversible error and AFFIRM.
Case: 20-20663 Document: 00516194200 Page: 2 Date Filed: 02/07/2022
No. 20-20663
FACTUAL AND PROCEDURAL BACKGROUND
Hess Corporation drilled six wells in the Tubular Bells oilfield in the
Gulf of Mexico. Hess needed surface controlled subsurface safety valves
(“SCSSVs”) and contracted with Schlumberger Technology Corporation to
provide the valves for what it designated as Wells B, C, and D.
These valves are large, complex, and expensive pieces of equipment
that are installed in the upper wellbore. The valves serve as emergency safety
devices that are not designed for general operational activities, such as
slowing or stopping production, or for backflow control. 1 The valves operate
by way of a flapper mechanism on the downhole end of the valve. When
closed, the flapper limits hydrocarbon flow into the production tubing. To
open and close the 15-foot-long, 1,600-pound valve, operators conduct
hydraulic pressure to two piston systems connected to the flapper. Further,
each piston system has five metal spring energized (“MSE”) seal assemblies
that, when activated, open the valve and allow the flow of hydrocarbons. The
valve fails “safe,” or closed, so that hydrocarbons do not reach the surface in
the event of malfunction.
Federal regulations require certain pieces of safety equipment, like the
SCSSVs at issue in this case, to satisfy the industry standards published by
the American Petroleum Institute (“API”). The SCSSVs were to be
“manufactured and marked pursuant to ANSI/API Spec. Q1.” See 30
C.F.R. § 250.801(a)–(b). API Specification 14A provides specific guidance
for subsurface safety valves, dictating practices for design and inspection.
1
The parties contested typical usage of these valves at trial. Hess’s expert, though,
referred to the SCSSV as “the ultimate fail safe device in the well [that] stops uncontrolled
flow of the well into the environment in the event of a catastrophic failure.” The district
court found that the valves were emergency safety devices. We see no clear error in this
factual finding.
2
Case: 20-20663 Document: 00516194200 Page: 3 Date Filed: 02/07/2022
No. 20-20663
The agreement between Hess and Schlumberger required Schlumberger to
provide valves that complied with then-current API guidance.
The first Schlumberger-provided valve failed in a Hess well in 2015.
Ultimately, four different Schlumberger valves failed in three different Hess
wells. 2 Schlumberger conducted an investigation and identified the MSE
Seal Spring as the cause of the failure. Schlumberger recalled the valves at
issue in this case. It informed customers that the failure of the MSE seal was
“due to [a] non-conforming MSE Spring,” which it attributed to the
manufacturer of the seal, Greene Tweed.
Hess stated its intent to revoke acceptance of the valves in May 2016.
It filed this suit in November 2016 in the U.S. District Court, Southern
District of Texas. In its Third Amended Complaint, Hess sought breach of
contract remedies available to buyers who properly revoked acceptance
under Section 2.608 of the Texas Business and Commerce Code.
Specifically, Hess sought recovery of the cost of retrieving and replacing the
non-conforming valves in Wells B, C, and D, as well as incidental damages,
consequential damages, attorneys’ fees, and other relief.
The district court held a bench trial on a single breach of contract
claim. The court found that Hess failed to prove the required elements for
revocation and denied the claim. Nonetheless, because substantial trial time
had been devoted to the matter of damages, the court made a contingent
2
Over the time-period relevant to the contract, Schlumberger had 137 valves with
the seals Hess alleged to be faulty in service in the Gulf. Of those, 114 were of a different
model which uses the same seals but is rated for lower pressures. Other valves made by
Schlumberger failed to various degrees before and after the failures in these Hess wells.
The parties dispute the relevant comparison set and definitions of “failure,” but Hess’s
failure rate was significantly higher than that of other producers.
3
Case: 20-20663 Document: 00516194200 Page: 4 Date Filed: 02/07/2022
No. 20-20663
ruling on damages to be applied in the event of reversal. Hess timely
appealed. Schlumberger cross-appealed the contingent award of damages.
DISCUSSION
When we review a district court’s decision following a bench trial, we
apply a standard of clear error to the court’s findings of fact and review legal
issues de novo. Guzman v. Hacienda Recs. & Recording Studio, Inc., 808 F.3d
1031, 1036 (5th Cir. 2015). Factual findings made during a bench trial deserve
“great deference.” Id. A district court’s finding of fact is clear error only if
it is “implausible in the light of the record considered as a whole.”
Brumfield v. Cain, 808 F.3d 1041, 1057 (5th Cir. 2015) (quotation marks and
citation omitted). “Where there are two permissible views of the evidence,
the factfinder’s choice between them cannot be clearly erroneous.”
Anderson v. City of Bessemer City, 470 U.S. 564, 574 (1985). This court
“grant[s] even greater deference to the trial court’s findings when they are
based on determinations of credibility.” Deloach Marine Servs., L.L.C. v.
Marquette Transp. Co., 974 F.3d 601, 607 (5th Cir. 2020) (quotation marks
and citation omitted); see also Fed. R. Civ. P. 52 (a)(6).
Hess’s claim is that it was entitled to revoke its acceptance of the
valves that Schlumberger had provided. The parties agree that the law
applicable to the claim is found in Section 2.608 of the Texas Business and
Commerce Code. Of particular relevance is this: a “buyer may revoke his
acceptance of a lot or commercial unit whose non-conformity substantially
impairs its value to him if he has accepted it . . . without discovery of such
non-conformity if his acceptance was reasonably induced either by the
4
Case: 20-20663 Document: 00516194200 Page: 5 Date Filed: 02/07/2022
No. 20-20663
difficulty of discovery before acceptance or by the seller’s assurances.”
Tex. Bus. & Com. Code § 2.608(a). 3
Caselaw has broken down the elements of revocation this way:
(1) initial acceptance (with a reasonable assumption that the
non-conforming item would be cured and it is not cured, or
without discovery of the non-conforming item if acceptance
was induced by difficulty of discovery or by seller’s assurance);
(2) of [a] non-conforming item; (3) such non-conformity
substantially impairs the value to the buyer; (4) and revocation
occurs within a reasonable time; (5) in any event, the
revocation must occur before a substantial change in the
condition of the goods occurs (which change is not caused by
defect of the goods).
Neily v. Arron, 724 S.W.2d 908, 913–14 (Tex. App. — Fort Worth 1987, no
writ). The district court, following the trial, found that Hess failed to prove
that the valves were non-conforming; that even if they were non-conforming,
there was no evidence any deviation had substantially impaired the value of
the valves; and, finally, that Hess had not proven it had revoked acceptance
before Hess itself damaged the valves.
In our review of the district court’s decision, we start with that court’s
interpretation of the contract and the applicable API standards.
I. Legal error in contract interpretation
Hess claims legal error in the district court’s interpretation of two
sections of API 14A standards that were incorporated into the sales contract.
3
In response, Schlumberger argues for a “resulting from” standard found in the
statute’s damages provision. See Tex. Bus. & Com. Code § 2.715 (entitling a non-
breaching buyer to damages “resulting from the seller’s breach”). Schlumberger, though,
presents no authority to show that Section 2.715 is relevant to the preliminary inquiry into
whether a non-conformance substantially impaired the value of goods.
5
Case: 20-20663 Document: 00516194200 Page: 6 Date Filed: 02/07/2022
No. 20-20663
Alternatively, Hess claims that the district court’s findings of fact relating to
Schlumberger’s compliance with one of these two sections were clearly
erroneous. If Hess prevails on either of its API 14A interpretation claims,
Schlumberger would, as a matter of law, have delivered non-conforming
goods.
a. API 14A Section 6.3.2.2.
API 14A Section 6.3.2.2, which pertains to “design criteria” for
SCSSVs, provides that equipment “shall be manufactured to drawings and
specifications that are substantially the same as those of the size, type, and
model [SCSSV] equipment that has passed the validation test.” The trial
court record shows that the drawings kept on file by the seal assembly
manufacturer, Greene Tweed, did not substantially change from 1998 to
2015. Hess is correct that the seals that had been used for the initial validation
and testing process in 2004 to qualify the valve did not exactly match the
dimensions specified in the drawings themselves. A partial explanation is
that their shapes had been changed by testing. The validation package did
not contain any untested seals. 4
As a simple factual matter, this means that in API 14A 6.3.2.2 terms,
there were no “drawings and specifications” of the valves that “passed the
validation test”; consequently, Hess claims that Section 6.3.2.2 was not —
indeed, could not — be satisfied. To use Hess’s words, “Section 6.3.2.2
requires the drawings and specifications of the validated valve to precisely
4
The district court found that the valves sold to Hess complied with Section
6.3.2.2: “Because drawings of MSE Assembly . . . remained the same from April of 1998
until September of 2015, the MSE Assemblies . . . in the SCSSVs that Schlumberger
manufactured for Hess in 2013 . . . were manufactured to drawings and specifications that
are the same as the drawings and specifications of the MSE Assembly that passed the
validation test in 2004.”
6
Case: 20-20663 Document: 00516194200 Page: 7 Date Filed: 02/07/2022
No. 20-20663
match the validated valve itself,” but no drawings of the validated valve even
exist.
Hess contends it would be inappropriate to “read in” a “substantially
the same” standard for the variation between the devices and the drawings
for three reasons. First, Hess claims that when Section 6.3.2.2 says “of the
size, type and model equipment that has passed the validation test” the “of”
denotes an “exact match” between the drawing and the equipment.
Allowing anything else, Hess says, gives “no assurance that the new valve
has any relation to the validated valve.” Second, Hess believes that because
“substantially the same” was used to qualify the relationship between the
two sets of drawings (of the original qualified equipment and the new
produced equipment), its absence in the provision comparing the drawings
to the products suggests a required exact match. Finally, Hess argues that
accepting the district court’s reading of the contract “would frustrate the
entire purpose of [the] standard,” because the standard would be
meaningless if the valves were not required to be manufactured exactly to the
drawings and specifications.
The district court, though, determined that “Hess’s contention . . . is
not supported by the evidence that Section 6.3.2.2 . . . is a design
requirement, not a quality control provision.” According to Schlumberger,
this reading “is supported by the text of the provision, industry custom and
practice, and the structure of API 14A.” At trial, Schlumberger relied on
witness testimony from experts who helped draft the API 14A standards, who
stated that nothing else beyond a comparison between “drawings and
specifications” was required to comply with 6.3.2.2.
Under Texas law, “[e]ven if a contract is unambiguous as a matter of
law, a court may still consider the surrounding facts and circumstances as an
aid in the construction of the contract’s language.” Barrow-Shaver Res. Co. v.
7
Case: 20-20663 Document: 00516194200 Page: 8 Date Filed: 02/07/2022
No. 20-20663
Carrizo Oil & Gas, Inc., 590 S.W.3d 471, 483 (Tex. 2019) (quotation marks
and citation omitted). 5 When “construing a specific contractual term, we
must give consideration to the meaning attributed to that term in the
industry.” Kona Tech. Corp. v. Southern Pac. Transp. Co., 225 F.3d 595, 611
(5th Cir. 2000) (quoting Personal Preference Video, Inc. v. HBO, 986 F.2d 110,
114 (5th Cir. 1993)) (applying Texas substantive law). In Kona, the court
relied on expert testimony and treatises to determine what “same or related
origins to same or related destinations” meant in the context of the shipping
industry. Id. at 611–12. The district court did not err in doing the same
regarding the interpretation of API 14A Section 6.3.2.2.
Schlumberger’s interpretation, adopted by the district court, is the
better one under Texas law. By understanding “of” to allow for some
insubstantial variation, the district court does not destroy any link between
the drawing and the qualified equipment. The interpretation requires that
any difference between physical product and drawing be insubstantial. This
interpretation accords with the expert testimony that the district court heard
from some of those who assisted in drafting the API 14A standards. The
district court did not err in considering such testimony. See Kona, 225 F.3d
at 611. The district court did not err in interpreting API 14A Section 6.3.2.2
to require only that the drawings remain substantially the same and that the
valves be manufactured using those drawings.
b. API 14A Section 7.6.2
In addition to its assertion that the district court erred in interpreting
Section 6.3.2.2, Hess argues that the district court also erred in interpreting
API 14A Section 7.6.2. Hess asserts that the MSE seal spring, known also as
5
Hess concedes that the contract is “unambiguous” and “that all undefined terms
in API 14A are given their ‘plain, ordinary meaning.’”
8
Case: 20-20663 Document: 00516194200 Page: 9 Date Filed: 02/07/2022
No. 20-20663
the “rosette spring,” was a “traceable component” under API 14A Section
7.6.2. That section requires “[a]ll traceable components, except non-
metallic seals, shall be dimensionally inspected . . . during or after the
manufacture of the components but prior to assembly.” Hess argues that,
under the “plain, ordinary meaning” of Section 7.6.2, because the rosette
spring could be traced, it is a traceable component requiring inspection.
Identifying the springs, not the entire seal, as traceable components would
mean the springs needed a dimensional inspection. That inspection did not
occur.
“Traceable component” is not defined in the contract. The district
court determined that whether the spring was a traceable component was a
question of fact rather than a question of law. After trial, the district court
found that “[t]he traceable components for purposes of [Section] 7.6.2 are
the MSE Seal assemblies [containing the rosette springs], not the rosette
springs” themselves. The district court primarily relied on the fact that the
seal assemblies were the “lowest level of traceable component” identified in
the Schlumberger Bill of Materials and the Hess Inspection Matrix — both
included in the contract — and expert testimony suggesting that it was
industry practice to define the lowest level of traceable component.
We consider the following to be dispositive. First, the parties set out
a quality control plan in which the parties stipulated to inspection of the seal
(but not its subcomponents). Second, the district court found that
Schlumberger put on credible testimony that it is industry practice to
enumerate components to be inspected and accounted for in the inspection
plan in order to be qualified as “traceable parts.” Finally, inspecting the
rosette spring inside the seal would require the destruction of the seal,
something that Schlumberger could not feasibly do as the purchaser of the
completed seal assembly. To inspect would destroy the assembly, and a new
one would be needed — only also to be destroyed during inspection. That
9
Case: 20-20663 Document: 00516194200 Page: 10 Date Filed: 02/07/2022
No. 20-20663
cannot be. We conclude that the contract contemplated dimensional
inspection of the seal assemblies rather than inspection of the rosette springs
within that assembly.
c. The district court’s API 14A factual determinations
Hess argues that the district court also made clearly erroneous factual
findings relative to Schlumberger’s compliance with 6.3.2.2. Namely, Hess
asserts that “the district court clearly erred in finding that the difference
between the Greene Tweed drawings and the 2004 validated valve was
‘insubstantial.’”
The Supreme Court has explained how to apply a clear-error standard
to a district court’s credibility findings at a bench trial. See Anderson v. City
of Bessemer City, 470 U.S. 564 (1985). The Anderson Court cautioned that a
trial court could not “insulate [its] findings from review by denominating
them credibility determinations” and outlined certain “factors” for
consideration that could show error. Id. at 575. Namely, “[d]ocuments or
objective evidence may contradict the witness’ story; or the story itself may
be so internally inconsistent or implausible on its face that a reasonable
factfinder would not credit it.” Id. If “such factors are present, the court of
appeals may well find clear error even in a finding purportedly based on a
credibility determination.” Id.
We applied Anderson in an appeal involving a fatal maritime collision
between a tug and a shrimper; the district court had considered physical
evidence, expert testimony analyzing the physical evidence, and independent
witness testimony. In re Luhr Bros., Inc., 157 F.3d 333, 339–40 (5th Cir. 1998).
The district court determined that the tug was at fault. Id. at 339. We
considered the “plausibility and internal consistency” of the shrimper’s
account, in addition to the actual evidence. Id. We found that “physical
evidence strongly support[ed]” the tug’s case; the tug’s expert witness was
10
Case: 20-20663 Document: 00516194200 Page: 11 Date Filed: 02/07/2022
No. 20-20663
far more qualified than the shrimper’s expert and considered more
information in making his assessment; the independent witness testimony
was “inconsistent with the [shrimper’s] account of the collision”; and the
shrimper’s account smacked of “sheer implausibility.” Id. at 340–42.
Accordingly, we were left with the “definite and firm conviction” that the
evidence showed clear error by the district court. Id. at 342.
We are not left with that conviction in this case. The drawings for the
seal did not change from 2003 to 2014, and Schlumberger presented some
evidence showing a series of springs from 2005 to 2015 that were
manufactured within the tolerances specified in the drawings. Although it is
clear that Greene Tweed produced springs that were outside the tolerances
dictated by the drawings and thus did not conform, it is certainly not
“implausible” that Greene Tweed manufactured its valves “to the qualified
drawings” under the design-requirement-only interpretation of Section
6.3.2.2 adopted by the district court. We accept the district court’s
interpretation of Section 6.3.2.2, meaning the evidence supports the district
court’s factual findings that the degree of difference was insubstantial.
II. Impairment of goods
We just analyzed and rejected the claim that the manufactured valves
failed to conform to the relevant drawings. We also will consider whether
Hess provided evidence that any alleged non-conformity “substantially
impair[ed] the value” of the goods to Hess. Such impairment is an element
of a revocation claim under Section 2.608(a) of the Texas Business and
Commerce Code. Hess’s theory at trial was that the non-conformities
caused the valve failures, which substantially impaired their value. The
district court, though, found that “the alleged violations of API 14A . . .
neither caused the SCSSVs to fail nor substantially impaired their value to
Hess.” Hess argues that the district court erred as a matter of law, first by
11
Case: 20-20663 Document: 00516194200 Page: 12 Date Filed: 02/07/2022
No. 20-20663
applying an incorrect standard of causation in determining the source of the
valve failure, and second in applying that standard and determining that
Schlumberger was not at fault for the substantial impairment of the value of
the valves. We separately consider those two arguments.
a. The district court’s applied standard of causation
Section 2.608(a) does not identify a standard of causation to be
applied when considering whether nonconformity impaired the value of
goods. The district court did not identify a standard either, stating simply
that the non-conformity of the SCSSVs “neither caused the SCSSVs to fail
nor substantially impaired their value to Hess.” Hess argues that the district
court should have relied on a line of Texas cases that inquire whether a
contract breach was a “producing cause” of an injury. See, e.g., Hunt v.
Ellisor & Tanner, Inc., 739 S.W.2d 933, 937–38 (Tex. App. — Dallas 1987,
writ denied). 6 Hess infers error from the district court’s ruling by arguing
that the court “treated causation as an ‘either/or’ proposition, such that if
Hess’s operating practices contributed to the valve failures, then the non-
conforming springs could not have been a legal cause of the failures.”
Assuming without deciding that Hess is correct that the proper
standard is “producing cause,” the district court’s order is consistent with
the application of such a rule. The district court first determined that
Schlumberger’s actions did not cause the failure of the SCSSVs. It then
determined that Hess’s operation of the well caused the valves to fail. From
there, the district court determined that Hess had “failed to establish by a
preponderance of the credible evidence that the alleged [API 14A]
violations . . . substantially impaired the value of the SCSSVs to Hess.”
6
Hess has cited no authority applying the “producing cause” standard in a
revocation case.
12
Case: 20-20663 Document: 00516194200 Page: 13 Date Filed: 02/07/2022
No. 20-20663
We do not find in the district court’s analysis an “either/or”
understanding of causation. Rather, the district court considered whether
the valves would have failed regardless of the supposed non-conforming
manufacture, and the court found that they would have. The only question,
then, is whether the district court’s factual findings were clearly erroneous.
b. The district court’s causation findings
At trial, Schlumberger introduced evidence that inadequate pressure
could “allow debris to flow with the pressure up to the MSE seals” and
scratch the seals. Schlumberger’s claim was that low pressure, combined
with the stress of temperature and pressure shifts from opening and closing
the valves a large number of times, would create stress on the valves to the
point of failure. The district court, in large part, adopted Schlumberger’s
view. Hess claims that this “ignored a mountain of compelling evidence from
internal, contemporaneous Schlumberger documents” and that “[t]here is
no plausible view of the evidence taken as a whole” that supports the district
court’s conclusion that the springs “did not at least substantially contribute
to the valves’ failures.”
Hess is correct that Schlumberger’s initial investigations and
causation analyses identified the rosette spring as the source of the valve
failure, consistent with what was found in an investigation into a failure of a
BP valve that pre-dated the Hess failures. Further, Schlumberger employees
communicated with each other internally, expressing confidence that the
rosette spring had changed over time and that the spring was the cause of at
least some valve failures. Hess is also correct that, in early investigations,
Schlumberger disfavored alternative explanations for the failures such as low
operating pressure, debris, and temperature swings.
The record reveals explicit assertions by Schlumberger from early in
this dispute that are strongly contrary to Schlumberger’s trial evidence.
13
Case: 20-20663 Document: 00516194200 Page: 14 Date Filed: 02/07/2022
No. 20-20663
Even with that change in position, though, there remains the fact question on
causation to be answered: “If the district court’s account of the evidence is
plausible in light of the record viewed in its entirety, the court of appeals may
not reverse it even though convinced that had it been sitting as the trier of
fact, it would have weighed the evidence differently.” Anderson, 470 U.S. at
573–74. The district court’s apparent crediting of Schlumberger’s
explanation that it had first relied on what it learned from an above-ground
test-failure of the valve and thus reached an incorrect conclusion is a
reasonable interpretation of why the first explanation of the problem can be
rejected. Such an account was supported by Schlumberger’s expert
testimony. Hess also had no explanation the district court had to accept as
to why Hess’s own failure rate is markedly higher than that of other
Schlumberger customers. Such a finding requires a weighing of the evidence,
and such weighing was the district court’s role, not ours.
Schlumberger’s account was not “so internally inconsistent or
implausible on its face that a reasonable factfinder would not credit it.” Id.
at 575. This is not a case in which neither physical evidence nor credible
expert evidence supported the district court’s determination. Luhr Bros., 157
F.3d at 342. Instead, “[e]ach [side] has support in inferences that may be
drawn from the facts in the record.” See Anderson, 470 U.S. at 577.
Accordingly, the district court did not clearly err in finding that any alleged
non-conformity did not cause the valves’ failure which in turn would have
impaired their value.
* * *
We AFFIRM the denial of Hess’s revocation claim. We need not
review the damage award, which would become relevant only upon reversal
of the finding of no liability. The cross-appeal is DISMISSED as moot.
14