PRIVILEGED AND CONFIDENTIAL — ATTORNEY WORK PRODUCT
MEMORANDUM
TO: Theodore Langston, Partner
FROM: Simone Laroche, Senior Associate; Jeffrey Park, Associate
DATE: November 3, 2023
RE: Review and Critique of Claimant's Expert Reports — Prof. Dieter Langfeld (Damages) and Dr. Martina Kessler (Metallurgy)
MATTER: Whitmore Aerospace LLC — ICC Case No. 27914/MHM
ENGAGEMENT NO.: 1019-00001
Right-click to update Table of Contents
I. Purpose and Executive Summary
The purpose of this memorandum is to provide a detailed review and critique of the two expert reports submitted by Claimant Kestrel Dynamics GmbH ("Kestrel") on October 16, 2023, in support of Kestrel's claims totaling approximately €47.3 million against Whitmore Aerospace LLC ("WAL") in connection with ICC Case No. 27914/MHM. This memorandum is prepared for the use of Theodore Langston and the broader litigation team in connection with the preparation of rebuttal expert reports and hearing strategy, and is intended to be treated as privileged attorney work product in its entirety.
The two expert reports under review are:
(1) Expert Report of Prof. Dieter Langfeld, Partner, Rheinsberg Advisory AG (Talstrasse 62, 8001 Glenrock Assurance, Switzerland), dated October 16, 2023, addressing the quantification of damages allegedly suffered by Kestrel as a result of WAL's purported material breach of the Master Supply Agreement dated June 15, 2018 (the "MSA"); and
(2) Expert Report of Dr. Martina Kessler, Director, Albrecht Prüftechnik GmbH (Robert-Bosch-Straße 12, 71229 Leonberg, Germany), dated October 16, 2023, addressing metallurgical and technical issues relating to lot conformance, with particular focus on the porosity specifications applicable to the Ti-6Al-4V ELI titanium alloy turbine housings supplied by Kestrel under the MSA.
We have reviewed both reports in detail, together with the supporting exhibits and appendices attached to each. We have also cross-referenced the reports against the MSA (including Amendment No. 1, dated February 10, 2020, and Amendment No. 2, dated September 3, 2021), the relevant contractual specifications (WAL-SPEC-TH-401 through WAL-SPEC-TH-410), WAL's Quality Control Procedures manual (WAL-QCP-100), and the documentary record produced in these proceedings pursuant to Procedural Orders Nos. 1 through 4.
Executive-Level Conclusions. Both reports contain significant methodological weaknesses and unsupported factual assumptions that should be exploited in rebuttal reports by our retained experts, Dr. Nadia Ostrovsky of Ridgeline Economic Consulting LLC (economic damages) and Dr. Raymond Chow of Arclight Materials Science Inc. (metallurgy and materials science), respectively. While neither report is facially implausible — indeed, both authors are credentialed professionals whose reports are presented in a professional format — the substantive vulnerabilities are substantial and, in several instances, may be dispositive of entire categories of claimed damages.
Summary of Principal Vulnerabilities. The principal weaknesses we have identified in the two reports are as follows:
Prof. Langfeld's Damages Report:
• Langfeld's damages model inflates Kestrel's claimed gross margin of 22.4% on the WAL account, which is inconsistent with Kestrel's own financial disclosures and the contemporaneous cost data obtained in document production. The actual blended margin on the WAL account appears materially lower, particularly when accounting for raw material cost increases in Ti-6Al-4V ELI titanium alloy and the additional labor and quality-assurance costs Kestrel incurred during Contract Years 3 and 4.
• The single largest component of Kestrel's claim — lost profits for the two discretionary renewal periods, totaling €27.15 million (more than 57% of the total claim) — is legally and analytically unsupportable. The renewal options were exercisable solely at WAL's discretion; Langfeld assumes a 100% probability of renewal without any evidentiary basis; and the claim is barred by the consequential damages waiver in § 8.7 of the MSA.
• Langfeld claims €3.90 million in reputational damages that are entirely speculative, lack any identified lost customer contracts or third-party communications, and are expressly barred as consequential damages under § 8.7.
• Langfeld's rework-cost claim of €3.42 million is internally inconsistent with Kessler's opinion that the lots were conforming. If the lots conformed to specifications, no rework should have been necessary. Additionally, rework costs for lots that Kestrel successfully cured were contractual obligations under § 8.4, not recoverable damages.
Dr. Kessler's Metallurgical Report:
• Kessler's re-testing was conducted on only 4 of the 14 rejected lots, and the results from this limited sample are extrapolated to all 14 lots without adequate statistical justification.
• Kessler's testing methodology (computed tomography scanning at stated resolution parameters) differs from the testing methodology prescribed in WAL-QCP-100 for incoming inspection, which is the contractually governing procedure. The methodological difference may yield systematically lower porosity readings.
• Kessler's "unpublished ≤0.15% standard" theory, while superficially supported by certain internal WAL engineering emails, does not withstand scrutiny when measured against the express porosity specification of ≤0.25% set forth in WAL-SPEC-TH-401, Appendix C, Table 3, and when considered in light of the lot-by-lot inspection data showing that the majority of rejected lots (11 of 14) exceeded the ≤0.25% contractual threshold.
• Kessler's reliance on Kestrel's AS9100 Rev D certification as evidence of adequate quality management conflates system-level certification with lot-specific conformance and ignores Kestrel's own internal quality records documenting increasing porosity concerns during CY4 production.
The sections that follow provide a detailed, section-by-section analysis of each report, together with specific recommendations for the rebuttal reports of Dr. Ostrovsky and Dr. Chow. We conclude with a discussion of cross-cutting issues, including the internal inconsistency between the two Claimant expert reports, and a summary of recommended next steps in advance of the November 13, 2023 filing deadline for rebuttal reports.
II. Review of Prof. Dieter Langfeld's Damages Expert Report
A. Overview of Langfeld's Report and Qualifications
Prof. Dieter Langfeld is identified in the report as a Partner at Rheinsberg Advisory AG, a Glenrock Assurance-based consulting firm that specializes in financial analysis, valuation, and economic damages quantification in commercial and manufacturing disputes. Langfeld's curriculum vitae, attached as Exhibit 1 to his report, indicates that he holds a doctorate in economics from the University of Glenrock Assurance and has served as an expert witness in approximately thirty-five international arbitration and litigation proceedings over the past fifteen years. His prior engagements have primarily involved supply-chain disputes and manufacturing sector damages, with a particular focus on the European aerospace and automotive industries. He has published several articles in trade and academic journals on the quantification of lost profits in long-term supply agreements.
Langfeld was retained by Kestrel's counsel, Haber Reinhardt Schlosser Partnerschaft mbB, Frankfurt am Main (Dr. Maximilian Haber, Lead Partner; Katrin Schäfer, Senior Associate), on behalf of Kestrel Dynamics GmbH. His report is dated October 16, 2023, and was filed on the same date in accordance with the procedural timetable established by the Tribunal.
Langfeld's central conclusion is that Kestrel is entitled to approximately €47.3 million in damages resulting from WAL's alleged material breach of the MSA. Langfeld organizes his damages calculation into four principal categories:
(i) Lost profits for the remaining term of the MSA following WAL's termination on October 3, 2022 through the scheduled expiration on June 14, 2023 (approximately 8.5 months of CY5): €12.83 million;
(ii) Lost profits for the two optional one-year renewal periods that Langfeld assumes WAL would have exercised: €27.15 million;
(iii) Rework costs allegedly incurred by Kestrel in attempting to cure the 14 rejected lots: €3.42 million; and
(iv) Reputational damages allegedly suffered as a result of WAL's termination and the associated market perception: €3.90 million.
We note at the outset that Langfeld's qualifications, while genuine, are not beyond challenge. His experience lies primarily in financial modeling and economic damages rather than in aerospace manufacturing cost accounting specifically. Additionally, several of the prior matters listed on his curriculum vitae involved significantly smaller claims and less complex supply-chain arrangements than the present case. However, we do not recommend a frontal challenge to Langfeld's qualifications, as such challenges are rarely successful before international arbitration tribunals and may distract from the more productive substantive critiques outlined below. Instead, we recommend that Dr. Ostrovsky's rebuttal report address Langfeld's methodology on its merits, demonstrating through rigorous analysis where his assumptions and calculations are unsupported or incorrect.
B. Critique of Langfeld's Lost-Profit Analysis — Remaining Contract Term (€12.83 Million)
Langfeld's first damages category addresses lost profits for the remainder of the MSA term following WAL's issuance of the Notice of Termination for Cause on October 3, 2022. The MSA was originally scheduled to run through June 14, 2023 (the end of Contract Year 5). Langfeld calculates this component of damages by projecting Kestrel's lost revenue for the period from the termination date through the scheduled contract expiration, applying the following inputs:
• Volume: 3,600 units (which Langfeld describes as the "anticipated" order volume for CY5, prorated for the remaining term);
• Unit price: €5,353.49, reflecting the CY5 unit price under the MSA pricing schedule as escalated pursuant to the 2.5% annual escalator in § 5.2;
• Gross margin: 22.4%, which Langfeld asserts is Kestrel's average gross margin on the WAL account based on management representations and internal financial data.
Langfeld applies the 22.4% margin to the projected revenue of approximately €19.27 million (3,600 units × €5,353.49) to arrive at lost gross profits of approximately €4.32 million for the prorated CY5 period. When annualized and adjusted for the full remaining term, Langfeld arrives at a total remaining-term lost-profits figure of €12.83 million, which includes what Langfeld characterizes as "incremental overhead absorption" losses in addition to the direct margin impact.
Weaknesses Identified. We have identified three principal weaknesses in Langfeld's remaining-term lost-profit analysis, each of which is discussed in detail below.
(1) The volume assumption of 3,600 units is unsupported. Langfeld assumes that WAL would have ordered 3,600 units during CY5. However, the MSA at § 4.1 specifies a minimum annual purchase commitment of only 2,400 units per contract year. WAL was entitled, but not obligated, to order volumes above the minimum. Langfeld bears the burden — and, by extension, Kestrel bears the burden — of demonstrating that WAL would in fact have ordered volumes above the contractual minimum during CY5. Langfeld does not carry this burden.
To the contrary, the historical purchase order data undermines Langfeld's volume assumption. A review of the purchase orders produced in this proceeding shows that WAL's actual order volumes during CY3 and CY4 were at or near the contractual minimum of 2,400 units. Specifically, WAL ordered approximately 2,520 units in CY3 and approximately 2,440 units in CY4. These figures are far closer to the 2,400-unit minimum than to the 3,600-unit figure assumed by Langfeld. Langfeld's volume assumption appears to be derived from CY1 and CY2 volumes, when WAL was ramping up the program and ordering higher quantities; however, by CY3 and CY4, WAL had stabilized its ordering pattern near the minimum commitment level. Langfeld does not address this downward trend or explain why CY5 volumes should be projected at the elevated CY1/CY2 levels rather than at the more recent CY3/CY4 levels.
Additionally, we note that even the 3,600-unit figure does not correspond to any specific contractual threshold or benchmark in the MSA. The MSA establishes a minimum of 2,400 units (§ 4.1) and, following Amendment No. 1 (February 10, 2020), a maximum of 4,000 units per contract year. Langfeld's 3,600-unit figure falls between these contractual bookends but is not derived from any contractual provision. It appears to represent an arithmetic average of historical order volumes across all contract years, which is a methodologically questionable approach given the clear downward trend in recent years.
(2) The 22.4% margin assumption is inflated. Langfeld's application of a 22.4% gross margin to the projected revenue is critical to his damages calculation, and we believe this figure is materially overstated. Our review of Kestrel's financial disclosures, together with the production documents obtained pursuant to Procedural Order No. 2 (March 17, 2023), indicates that Kestrel's actual blended gross margin on the WAL account was materially lower than 22.4% during the relevant period.
Several factors support this conclusion. First, Kestrel experienced significant raw material cost increases for Ti-6Al-4V ELI titanium alloy during CY3 and CY4. Kestrel's own internal procurement records, produced as part of the document production, reflect titanium alloy price increases of approximately 12% between CY2 and CY4. The MSA's annual price escalator of 2.5% (§ 5.2) was not sufficient to offset these raw material cost increases, compressing Kestrel's margins over time. Langfeld does not adequately account for this margin compression in his model; instead, he applies a single blended margin figure across the entire contract term that appears to be weighted toward the more favorable early-year margins.
Second, Kestrel incurred additional labor and quality-assurance costs during CY4 as a result of the quality issues that gave rise to the lot rejections. Kestrel's own internal cost reports reflect increased inspection labor, additional NDT testing, and rework activities — all of which are direct manufacturing costs that reduce gross margins. Langfeld does not appear to have deducted these incremental costs from his margin calculation, which suggests that his margin figure reflects a normalized or idealized production environment rather than the actual cost structure during the relevant period.
Third, Langfeld's margin figure is based in part on what he describes as "management representations" from Kestrel's Chief Financial Officer, Dr. Tobias Fleischer. Such representations, without corroboration from audited financial data, are inherently unreliable in the context of a disputed damages calculation. We note that Langfeld does not disclose whether he independently verified the margin figure against Kestrel's audited financial statements or underlying cost accounting records.
(3) Langfeld fails to deduct mitigation income. A fundamental principle of damages law under New York law — which governs the MSA pursuant to § 14.2 — is that a claimant must mitigate its damages by making reasonable efforts to replace lost business. The record in this proceeding suggests that Kestrel redirected some of its manufacturing capacity to other customers following WAL's termination. Kestrel's own production records indicate that the manufacturing line previously dedicated to the WAL program was partially redeployed to fill orders for other aerospace customers within approximately three months of the termination. Any revenue earned, or reasonably earnable, from these alternative customers should be deducted from Kestrel's lost-profit claim.
Langfeld makes no deduction for mitigation income in his damages model. His report does not address the mitigation issue at all, which is a significant omission. Under New York law, the burden of proving failure to mitigate rests on the party asserting it (here, WAL), but the complete absence of any mitigation analysis in Langfeld's report is nevertheless a vulnerability, particularly given the documentary evidence of Kestrel's post-termination production activity.
Recommendation for Dr. Ostrovsky's Rebuttal Report. We recommend that Dr. Ostrovsky's rebuttal report on the remaining-term lost-profits component include the following analyses:
(a) Re-calculate lost profits using the contractual minimum volume of 2,400 units per contract year, prorated for the remaining CY5 term, rather than Langfeld's unsupported 3,600-unit assumption;
(b) Apply a corrected gross margin based on an analysis of Kestrel's actual historical margins on the WAL account, drawing on the production documents and financial data in the record, coordinating as necessary with Sandra Gillespie, Engagement Partner at Canfield & Thorpe LLP, for forensic accounting inputs regarding Kestrel's cost structure;
(c) Quantify the mitigation income that Kestrel earned or could reasonably have earned from the redeployment of manufacturing capacity following the termination; and
(d) Present a revised remaining-term lost-profits figure that reflects these corrections, which we anticipate will be substantially lower than Langfeld's €12.83 million.
C. Critique of Langfeld's Renewal-Period Lost Profits (€27.15 Million)
Langfeld's second damages category — and by far the largest single component of Kestrel's claim — addresses lost profits for the two optional one-year renewal periods set forth in § 3.2 of the MSA. Langfeld calculates this component by assuming that WAL would have exercised both renewal options and that the MSA would have continued for an additional two years beyond the original June 14, 2023 expiration date. He projects revenue for each renewal year at escalated pricing (reflecting the 2.5% annual escalator in § 5.2) and assumes order volumes of 3,600 units per year. Applying his 22.4% margin assumption, Langfeld arrives at a total renewal-period lost-profits figure of €27.15 million.
This component represents more than 57% of Kestrel's total claimed damages of €47.3 million. It is, accordingly, the single most consequential element of Langfeld's report from a financial perspective. It is also, in our view, the most vulnerable to challenge on both legal and analytical grounds.
Weaknesses Identified. We have identified five principal weaknesses in Langfeld's renewal-period analysis.
(1) The renewal options were exercisable solely at WAL's discretion. The MSA at § 3.2 provides that the agreement "may be renewed for up to two (2) additional one-year periods upon written notice by [WAL] delivered not less than ninety (90) days prior to the expiration of the then-current term." The language is unambiguous: the renewal options are WAL's to exercise, not Kestrel's. Kestrel had no contractual right to compel renewal. In the absence of an obligation to renew, Kestrel has no legally cognizable expectation interest in renewal-period profits under New York law. A claim for lost profits must be grounded in a reasonably certain expectation of future business; discretionary renewal options, by their nature, do not create such certainty.
The New York Court of Appeals has consistently held that lost-profit damages must be established with "reasonable certainty" and cannot be based on speculative assumptions about the exercise of discretionary contractual provisions. See Kenford Co. v. County of Erie, 67 N.Y.2d 257, 261 (1986); Ashland Mgmt. Inc. v. Janien, 82 N.Y.2d 395, 403 (1993). Langfeld's assumption that WAL would have exercised both renewal options is precisely the type of speculative projection that New York courts have repeatedly rejected.
(2) Langfeld assumes a 100% probability of renewal. Even if lost profits for discretionary renewal periods were legally permissible in principle, Langfeld's analysis assumes with absolute certainty that WAL would have exercised both options. This assumption is analytically indefensible. Langfeld provides no probability-weighted analysis of the likelihood of renewal. He cites no evidence of WAL's stated intent to renew, no internal WAL communications indicating a plan to exercise the renewal options, and no market analysis of alternative suppliers that might bear on WAL's renewal decision. The only support Langfeld offers for his renewal assumption is the general assertion that "long-term supply relationships in the aerospace sector are typically renewed absent material intervening events." This assertion is vague, unsupported by empirical data, and circular — the "material intervening event" here is precisely the quality dispute that precipitated the termination.
A responsible damages analysis would, at minimum, apply a probability weighting to the renewal-period profits to reflect the inherent uncertainty of an optional renewal. No such weighting appears anywhere in Langfeld's report.
(3) The consequential damages waiver in § 8.7 bars recovery of forward-looking lost profits beyond the contract term. Section 8.7 of the MSA provides, in relevant part:
"IN NO EVENT SHALL EITHER PARTY BE LIABLE TO THE OTHER FOR ANY INDIRECT, INCIDENTAL, CONSEQUENTIAL, SPECIAL, OR PUNITIVE DAMAGES, INCLUDING BUT NOT LIMITED TO LOST PROFITS BEYOND THE THEN-CURRENT CONTRACT TERM, ARISING OUT OF OR IN CONNECTION WITH THIS AGREEMENT OR ANY BREACH HEREOF, REGARDLESS OF WHETHER SUCH DAMAGES ARE BASED ON CONTRACT, TORT, STRICT LIABILITY, OR ANY OTHER THEORY."
The waiver expressly addresses "lost profits beyond the then-current contract term." Renewal-period lost profits are, by definition, profits that would have been earned beyond the then-current term. The waiver therefore bars Kestrel's recovery of the entire €27.15 million renewal-period component as a matter of contract. Langfeld's report acknowledges the existence of § 8.7 but dismisses it in a single footnote, asserting that the waiver applies only to "tort-based consequential damages" and not to "contract damages flowing naturally from a material breach." This interpretation is inconsistent with the plain language of § 8.7, which explicitly references lost profits "arising out of or in connection with this Agreement or any breach hereof" and applies "regardless of whether such damages are based on contract, tort, strict liability, or any other theory."
(4) Langfeld uses the 3,600-unit volume assumption for renewal years without acknowledging Amendment No. 1. As discussed in Section II.B above, Langfeld's volume assumption of 3,600 units per year is unsupported even for the remaining contract term. The same criticism applies with even greater force to the renewal periods, where the uncertainty of WAL's ordering behavior would be compounded. Moreover, Langfeld's use of 3,600 units for the renewal periods creates an internal inconsistency in his own report: Amendment No. 1 (February 10, 2020) increased the maximum annual order volume from 3,600 to 4,000 units. If Langfeld's volume assumption is based on the contractual maximum, it should be 4,000 units, not 3,600. If it is based on historical averages, the more recent CY3 and CY4 data would suggest volumes closer to 2,400 units. The 3,600-unit figure appears to be selected neither from the contractual terms nor from the historical data, but from a methodology that yields the most favorable result for Kestrel. This selective use of volume figures undermines the credibility of Langfeld's overall approach.
(5) No discounting to present value is applied. Langfeld's renewal-period damages are stated in nominal terms without any discounting to present value. The renewal periods would not have commenced until June 15, 2023 at the earliest and would not have concluded until June 14, 2025. Any damages award for future lost profits should be discounted to present value to reflect the time value of money. The failure to apply any discount rate to the renewal-period amounts results in an overstatement of the present value of the claimed damages. While the appropriate discount rate is a matter of expert judgment, the complete absence of discounting is a basic methodological error that should be highlighted in Dr. Ostrovsky's rebuttal.
Recommendation for Dr. Ostrovsky's Rebuttal Report. We recommend that Dr. Ostrovsky's rebuttal report on the renewal-period component take the following approach:
(a) The primary argument should be that renewal-period lost profits must be excluded entirely as speculative and legally barred under the consequential damages waiver in § 8.7. This position is well-supported under New York law, and the express language of § 8.7 is unambiguous;
(b) In the alternative, if the Tribunal were to consider renewal-period damages notwithstanding the contractual bar, Dr. Ostrovsky should demonstrate that Langfeld's 100% renewal probability assumption is unsupportable and should apply a probability weighting of no more than 10–20%, reflecting the uncertain market conditions, WAL's documented supplier diversification efforts, the unresolved quality issues, and the absence of any evidence of WAL's stated intent to renew;
(c) Any renewal-period amounts, even if probability-weighted, should be discounted to present value using an appropriate discount rate; and
(d) The corrected volume and margin inputs discussed in Section II.B should be applied to any alternative renewal-period calculation.
We note that if the Tribunal accepts the primary argument that renewal-period damages should be excluded entirely, this would reduce Kestrel's total claimed damages from €47.3 million to approximately €20.15 million — a reduction of more than 57%.
D. Critique of Langfeld's Rework-Cost Claim (€3.42 Million)
Langfeld's third damages category addresses the costs allegedly incurred by Kestrel in attempting to cure the 14 rejected lots (totaling 1,680 units) within the 45-calendar-day amended cure period established by Amendment No. 2, dated September 3, 2021. Langfeld calculates the rework costs based on Kestrel's internal cost estimates for labor, materials, retesting, and associated overhead allocable to the cure efforts.
It is undisputed that WAL rejected 14 lots (Lot KDG-2022-0307 through Lot KDG-2022-0919) between March 7, 2022 and September 19, 2022 for failing to meet the porosity specification of ≤0.25% set forth in WAL-SPEC-TH-401, Appendix C, Table 3. It is also undisputed that Kestrel undertook cure efforts on all 14 lots and that Kestrel successfully cured 7 lots within the 45-day cure period, resulting in acceptance of those lots by WAL. The remaining 7 lots were not successfully cured within the cure period.
Weaknesses Identified. We have identified four principal weaknesses in Langfeld's rework-cost analysis.
(1) Rework costs for the 7 successfully cured lots are not recoverable as damages. Of the 14 rejected lots, Kestrel successfully cured 7 within the 45-day cure period. For these lots, the rework costs were incurred in the performance of Kestrel's contractual obligation to cure non-conforming goods under § 8.4 of the MSA. Section 8.4 provides that upon rejection of non-conforming goods, "Supplier shall, at its sole cost and expense, cure the non-conformance or replace the defective goods within the cure period." The rework costs associated with successful cure are therefore Kestrel's ordinary cost of performance — they are an expense Kestrel was contractually required to bear — and are not recoverable as damages. Including these costs in a damages claim is fundamentally inconsistent with the contractual allocation of risk under § 8.4.
(2) Rework costs for the 7 uncured lots are internally inconsistent with Kestrel's position. For the 7 lots that Kestrel failed to cure within the 45-day period, Langfeld includes the costs of partial cure attempts in his damages model. This inclusion creates a significant logical inconsistency with Kestrel's overall case theory. Kestrel's primary position in this arbitration — as reflected in Kessler's expert report — is that the rejected lots were conforming and should not have been rejected. If the lots met specifications as delivered, then no rework should have been necessary, and the rework costs should not exist. Yet Langfeld claims rework costs for cure attempts on these lots, which implicitly concedes that the lots required correction. This internal contradiction between Kestrel's two expert reports is a significant vulnerability that we discuss further in Section IV.A below.
(3) The per-unit rework cost figures are unsubstantiated. Langfeld cites per-unit rework cost figures that, on review, do not appear to be supported by contemporaneous cost records. Langfeld's figures are derived primarily from summary schedules prepared by Kestrel's finance department, which were created after the commencement of this arbitration. No contemporaneous time-and-materials records, labor logs, or material purchase orders have been produced to substantiate the claimed rework costs on a per-unit basis. In the absence of such documentation, the per-unit figures must be regarded as unreliable estimates rather than actual incurred costs.
We have requested that the litigation team confirm whether any contemporaneous rework cost documentation exists in the production record. If no such documentation was produced, this omission should be highlighted in Dr. Ostrovsky's rebuttal and may warrant a request for additional production if the Tribunal will entertain such a request at this stage.
(4) Rework costs are not recoverable under the exclusive remedy provision of § 8.7. Section 8.7 of the MSA establishes an exclusive remedy framework for quality-related disputes. Under this provision, Kestrel's available remedies in the event of a quality dispute are limited to replacement of non-conforming goods, credit against future orders, or termination of the affected purchase order. Rework costs incurred voluntarily by Kestrel in attempting to cure non-conforming goods are not among the enumerated remedies and are therefore not recoverable under the contractual exclusive-remedy regime. Langfeld's report does not address the impact of § 8.7 on the recoverability of rework costs.
Recommendation for Dr. Ostrovsky's Rebuttal Report. We recommend the following approach in Dr. Ostrovsky's rebuttal:
(a) Challenge the internal inconsistency between Langfeld's rework-cost claim and Kessler's opinion that the lots were conforming — Kestrel cannot simultaneously contend that the lots met specifications and claim damages for the costs of fixing them;
(b) Demand substantiation of per-unit rework costs through contemporaneous cost records, and highlight the absence of such documentation in the current record;
(c) Argue that rework costs for the 7 successfully cured lots are ordinary costs of contractual performance under § 8.4, not recoverable damages; and
(d) Argue that the exclusive remedy provision of § 8.7 bars recovery of voluntarily incurred rework costs.
E. Critique of Langfeld's Reputational Damages Claim (€3.90 Million)
Langfeld's fourth and final damages category is a claim for reputational damages in the amount of €3.90 million. Langfeld's methodology for this component is, to put it charitably, impressionistic. Langfeld asserts that WAL's termination of the MSA and "the associated market perception" caused Kestrel to lose business opportunities with unnamed third-party aerospace customers. He estimates the reputational harm by reference to a "market impact assessment" that consists primarily of general statements about the aerospace industry's emphasis on supplier reliability and quality reputation.
Weaknesses Identified. We have identified four fundamental weaknesses in this claim.
(1) The claim is speculative and lacks objective evidentiary support. Langfeld identifies no specific customer contracts that Kestrel lost as a result of the WAL termination. He names no specific customers that declined to do business with Kestrel as a result of the dispute. He cites no declined bids attributable to the WAL termination. He produces no communications from third parties referencing the dispute or expressing concerns about Kestrel's quality reputation. He provides no before-and-after analysis of Kestrel's customer pipeline, win rate, or revenue trajectory. In short, there is no evidence whatsoever of a causal link between WAL's termination and any specific lost business opportunity. The €3.90 million figure appears to be derived from a generalized assumption that some unquantified percentage of Kestrel's business was adversely affected, which does not meet the "reasonable certainty" standard for damages under New York law.
(2) Reputational damages are expressly barred by the consequential damages waiver in § 8.7. Reputational damages are consequential in nature under New York law. They do not flow directly from a breach of contract but rather from the indirect effects of the breach on the injured party's relationships with third parties. As such, they fall squarely within the scope of the consequential damages waiver in § 8.7, which bars "indirect, incidental, consequential, special, or punitive damages." Langfeld's report does not address the waiver in the context of the reputational damages claim.
(3) No recognized methodology for quantifying reputational harm in a B2B manufacturing context is applied. Langfeld does not apply any recognized or peer-reviewed methodology for quantifying reputational damages. There is no event study, no market analysis, no regression analysis, and no comparable-transaction analysis. The €3.90 million figure appears to be a rough estimate based on Langfeld's judgment rather than a rigorous quantification.
(4) The claimed amount is disproportionate and unsupported by Kestrel's actual financial performance. Kestrel's total revenue for FY2022 was approximately €312 million. The claimed reputational damages of €3.90 million represent approximately 1.25% of annual revenue. Yet Langfeld provides no evidence that Kestrel's revenue was adversely affected at all. In fact, preliminary indications from Kestrel's publicly available financial disclosures suggest that Kestrel's overall revenue was relatively stable in the period following the WAL termination. If Kestrel suffered a 1.25% revenue impact, one would expect to see some corroboration of that decline in the financial data. No such corroboration is provided.
Recommendation for Dr. Ostrovsky's Rebuttal Report. We recommend that Dr. Ostrovsky argue for the complete exclusion of reputational damages on the grounds that (a) the claim is speculative and fails to meet the reasonable-certainty standard; (b) the claim is expressly barred by the consequential damages waiver in § 8.7; and (c) there is no causal nexus between the termination and any demonstrated revenue loss. If the Tribunal nevertheless considers the claim, Dr. Ostrovsky should challenge the absence of any recognized quantification methodology and the lack of any supporting financial data.
F. Langfeld's Treatment of WAL's Counterclaim
In the final section of his report, Langfeld briefly addresses WAL's counterclaim of €5.17 million for replacement-component costs incurred by WAL in procuring Ti-6Al-4V ELI titanium alloy turbine housings from Valkyr Precision Components Inc. (Tucson, Arizona) to replace the rejected Kestrel lots. WAL's counterclaim is based on the per-unit cost premium that WAL paid to Valkyr above the CY4 contract price with Kestrel.
Langfeld challenges WAL's counterclaim on three grounds: (1) WAL allegedly failed to mitigate by not seeking competitive bids from alternative suppliers earlier in the process; (2) the Valkyr unit price of €8,432.92 is allegedly above market for comparable titanium alloy turbine housings; and (3) the per-unit premium of approximately €3,210 (€8,432.92 minus Kestrel's CY4 unit price of €5,222.92) is allegedly inflated because WAL could have obtained better pricing had it acted more promptly.
Weaknesses in Langfeld's Critique of the Counterclaim. We have identified three principal weaknesses in Langfeld's analysis of WAL's counterclaim.
(1) Langfeld does not account for the Valkyr introductory discount. Langfeld's critique of the Valkyr pricing fails to account for the 4.2% introductory discount that Valkyr offered WAL on the initial order tranche and that WAL already applied in its counterclaim calculation. WAL's counterclaim calculation of €5.17 million reflects the net premium after the introductory discount (i.e., 1,680 units × €3,210 per-unit premium × 0.958 discount factor = approximately €5,166,763, rounded to €5.17 million). Langfeld's failure to recognize the discount in his critique overstates the degree to which WAL's counterclaim figure is inflated.
(2) Langfeld provides no evidence of alternative suppliers. Langfeld asserts that WAL could have obtained Ti-6Al-4V ELI titanium alloy turbine housings at lower cost from suppliers other than Valkyr, but he identifies no specific alternative supplier that could have met WAL's stringent specifications (WAL-SPEC-TH-401 through WAL-SPEC-TH-410) at a lower price and within the required timeframe. The manufacture of aerospace-grade titanium alloy turbine housings is a specialized field with a limited number of qualified suppliers. WAL's specifications include demanding porosity, dimensional tolerance, and surface-finish requirements that not all titanium alloy fabricators can meet. Langfeld's assertion that WAL should have obtained "competitive bids" is facile without evidence that competitive alternatives actually existed.
(3) Langfeld's timing criticism ignores the cure-period obligations. Langfeld criticizes WAL for not seeking alternative supplier bids sooner, but this criticism ignores the contractual cure-period timeline. Under Amendment No. 2 (September 3, 2021), WAL was required to provide Kestrel with a 45-calendar-day cure period following each lot rejection before taking further remedial action. WAL complied with this requirement, providing Kestrel with the full cure period for each of the 14 rejected lots. Only after Kestrel failed to cure 7 of the 14 lots did WAL begin the process of engaging an alternative supplier. Langfeld's suggestion that WAL should have sought bids earlier would have required WAL to breach its own contractual obligations to Kestrel by preemptively replacing Kestrel before the cure period expired.
Recommendation for Dr. Ostrovsky's Rebuttal Report. We recommend that Dr. Ostrovsky's rebuttal report defend the reasonableness of WAL's counterclaim as follows:
(a) Demonstrate the reasonableness of Valkyr's pricing given the specialized nature of Ti-6Al-4V ELI titanium alloy turbine housings and the limited number of qualified suppliers capable of meeting WAL's specifications;
(b) Show that WAL's transition timeline was driven by Kestrel's repeated failures to cure the rejected lots and by WAL's compliance with its cure-period obligations;
(c) Confirm that the 4.2% introductory discount is already reflected in WAL's counterclaim calculation of €5.17 million; and
(d) Quantify the costs WAL would have incurred had it delayed the Valkyr engagement further, including potential production-line downtime and program delays.
III. Review of Dr. Martina Kessler's Technical/Metallurgical Expert Report
A. Overview of Kessler's Report and Qualifications
Dr. Martina Kessler is identified in her report as the Director of Albrecht Prüftechnik GmbH, a testing and quality-certification firm based in Leonberg, Germany (Robert-Bosch-Straße 12, 71229 Leonberg). Kessler holds a doctorate in materials science from the University of Stuttgart and has over twenty years of experience in non-destructive testing (NDT) and quality certification for aerospace components. Her curriculum vitae lists numerous engagements involving the testing and certification of titanium and nickel-based superalloy components for European aerospace OEMs and Tier 1 suppliers. She is a certified Level III NDT inspector under EN ISO 9712 and is listed as a recognized testing authority by several European aviation regulators.
Kessler's report was filed on October 16, 2023, concurrently with Langfeld's damages report. She was retained by the same counsel, Haber Reinhardt Schlosser Partnerschaft mbB.
Kessler's central conclusion is that the 14 lots rejected by WAL (Lot KDG-2022-0307 through Lot KDG-2022-0919, totaling 1,680 units) met the contractual porosity specification and should not have been rejected. Kessler further contends that WAL applied an unpublished, stricter internal porosity tolerance of ≤0.15% rather than the contractual standard of ≤0.25% as set forth in WAL-SPEC-TH-401, Appendix C, Table 3. This "unpublished standard" theory is central to Kestrel's broader case theory that WAL's rejections were pretextual and motivated by WAL's desire to replace Kestrel with a lower-cost supplier (Valkyr Precision Components Inc.).
We note at the outset that Kessler's qualifications are more directly relevant to the subject matter of her opinions than Langfeld's are to his. Kessler has substantial experience in NDT for aerospace components, and her expertise in porosity testing is genuine. However, as discussed below, her testing methodology, sample selection, and interpretive framework are subject to serious challenge. Additionally, her "unpublished standard" theory, while superficially supported by certain internal WAL documents, does not withstand scrutiny when examined against the full documentary record and the lot-by-lot testing data.
B. Critique of Kessler's Porosity Testing Methodology
Kessler's approach to the porosity issue involved conducting independent re-testing of representative samples from a subset of the rejected lots. She employed two primary testing methods: computed tomography (CT) scanning and optical microscopy. Based on this re-testing, Kessler concludes that the tested samples met the contractual porosity specification of ≤0.25% and that the lots should not have been rejected.
Weaknesses Identified. We have identified five principal weaknesses in Kessler's re-testing methodology and the conclusions she draws from it.
(1) Kessler's re-testing was conducted on only 4 of 14 rejected lots. Of the 14 rejected lots, Kessler obtained and tested samples from only 4 lots. She does not explain why she was unable to obtain samples from the remaining 10 lots. She acknowledges this limitation in a footnote but proceeds to extrapolate her findings from the 4 tested lots to all 14 lots. This extrapolation is not statistically justified. Each lot consists of 120 units manufactured in a specific production run, and there is no basis for assuming that porosity levels are uniform across lots produced at different times during CY4. Indeed, the very nature of Kestrel's porosity issues — which appear to have been influenced by variations in raw material quality from Kestrel's titanium supplier — suggests that porosity levels may have varied significantly from lot to lot. Kessler's extrapolation from 4 lots to 14 is a fundamental methodological flaw that undermines the reliability of her overall conclusion.
(2) The re-testing was performed more than 12 months after original rejection. Kessler's re-testing was conducted in approximately September 2023, more than 12 months after the most recent lot rejections. While titanium alloy is not inherently susceptible to degradation under normal storage conditions, the chain of custody and storage conditions for the retained samples are not documented in Kessler's report. She does not describe where the samples were stored, at what temperature and humidity conditions, or whether the samples were sealed against contamination. In the aerospace industry, testing protocols typically require that samples be tested within a defined period after manufacture and under controlled storage conditions. The lapse of more than 12 months raises legitimate questions about whether the tested samples remained representative of the lots as delivered and inspected.
(3) Kessler's CT scanning methodology differs from WAL-QCP-100. The contractual incoming inspection procedures are defined in WAL-QCP-100, which prescribes specific testing methods, equipment parameters, and acceptance criteria for porosity evaluation. WAL-QCP-100 specifies the use of metallographic cross-sectioning with optical microscopy at defined magnification levels, supplemented by ultrasonic testing, as the primary methods for porosity assessment during incoming inspection.
Kessler's approach relies primarily on CT scanning at stated resolution parameters. While CT scanning is a legitimate and widely used NDT technique, it differs from the methodology prescribed in WAL-QCP-100 in several important respects. CT scanning evaluates the volume of a component non-destructively and generates a three-dimensional map of internal features, including pores. However, the resolution of CT scanning is limited by voxel size, and small sub-surface pores may not be resolved at the resolution levels Kessler employed. This means that Kessler's methodology may systematically underestimate porosity compared to the metallographic cross-sectioning method prescribed in WAL-QCP-100, which directly reveals pore morphology and distribution at the microscopic level. Kessler does not address this methodological discrepancy or explain why CT scanning should be preferred over the contractually prescribed method for assessing conformance to the ≤0.25% porosity specification.
(4) Kessler does not disclose full raw data. Kessler's report presents only summary statistics for her porosity measurements — specifically, mean porosity values and ranges for each tested lot. The full raw data underlying these summary statistics (i.e., individual sample measurements, CT scan images, and optical microscopy images with measurement annotations) are not attached to the report or otherwise disclosed. This lack of transparency makes it impossible to independently verify Kessler's calculations, assess the distribution of porosity readings within each lot, or identify outlier measurements that may affect the summary statistics. In scientific and technical expert reports, the disclosure of raw data is a baseline expectation that allows the opposing party's expert to conduct a meaningful review. Kessler's failure to provide this data is a significant vulnerability.
(5) Even Kessler's most favorable readings approach the contractual limit. Perhaps most telling is the fact that even on Kessler's own favorable testing — conducted on only 4 of 14 lots, using a methodology that may systematically underestimate porosity — several tested samples showed porosity levels between 0.20% and 0.24%. These readings are very close to the ≤0.25% contractual limit. They do not conclusively demonstrate conformance even for the lots Kessler tested, let alone for the 10 lots she did not test. A reading of 0.24% porosity, for example, leaves only a 0.01 percentage-point margin below the specification limit — a margin well within the measurement uncertainty of even the most precise testing methods. Kessler does not address measurement uncertainty in her report, which is a further omission that weakens her conclusions.
Comparison with Dr. Chow's Findings. Dr. Raymond Chow of Arclight Materials Science Inc. (McLean, Virginia), our retained metallurgical expert, tested samples from all 14 rejected lots using the methodology prescribed in WAL-QCP-100 for incoming inspection. Dr. Chow's results show porosity levels ranging from 0.19% to 0.38%, with the majority of lots (11 of 14) exceeding the ≤0.25% contractual limit. This is a fundamentally different picture from the one presented by Kessler's limited and methodologically divergent testing. The disparity between the two experts' findings underscores the importance of the testing methodology: when the contractually prescribed WAL-QCP-100 method is used, the majority of lots fail; when Kessler's alternative CT scanning method is used on a limited sample, the results are more favorable to Kestrel but still close to the specification limit.
Recommendation for Dr. Chow's Rebuttal Report. We recommend that Dr. Chow's rebuttal report on Kessler's porosity testing methodology include the following:
(a) Highlight the limited sample size of Kessler's testing (4 of 14 lots) and the lack of statistical justification for extrapolating results to all 14 lots;
(b) Challenge the validity of re-testing conducted more than 12 months after the original lot rejections, with specific attention to sample storage conditions and chain of custody;
(c) Contrast Kessler's CT scanning methodology with the contractually prescribed WAL-QCP-100 incoming inspection procedures, explaining the technical reasons why CT scanning may yield systematically lower porosity readings;
(d) Request and analyze Kessler's full raw data (which should be sought through counsel if not voluntarily provided);
(e) Provide a full lot-by-lot data comparison between Kessler's results (for the 4 lots she tested) and Dr. Chow's results (for all 14 lots), as set forth in Appendix B to this memorandum;
(f) Address measurement uncertainty and demonstrate that Kessler's readings of 0.20%–0.24% do not conclusively establish conformance given the inherent uncertainty in porosity measurement; and
(g) Emphasize that even Kessler's most favorable readings show results approaching the contractual limit, casting doubt on the conformance of the untested lots.
C. Critique of Kessler's "Unpublished Standard" Theory
Kessler's most prominent contention is that WAL applied an internal, unpublished porosity standard of ≤0.15% to reject Kestrel's lots, rather than the contractual standard of ≤0.25% set forth in WAL-SPEC-TH-401, Appendix C, Table 3. This theory is central to Kestrel's broader narrative that WAL's lot rejections were pretextual — motivated not by genuine quality concerns but by WAL's desire to replace Kestrel with Valkyr Precision Components Inc. as a lower-cost alternative.
Kessler bases her "unpublished standard" theory on two principal pieces of evidence: (i) certain internal WAL emails produced in document production, in which WAL engineers reference a "target" porosity of ≤0.15% in the context of internal engineering guidelines for turbine housing design optimization; and (ii) the observation that some of the rejected lots had measured porosity levels above 0.15% but below 0.25% — in other words, lots that would have been accepted under the contractual ≤0.25% standard but rejected under the alleged ≤0.15% internal standard.
Weaknesses Identified. We have identified five principal weaknesses in Kessler's "unpublished standard" theory.
(1) The contractual specification is unambiguous and governs. The porosity specification applicable to Kestrel's deliverables is expressly stated in WAL-SPEC-TH-401, Appendix C, Table 3 as ≤0.25%. This specification is incorporated into the MSA by reference and constitutes the governing acceptance criterion for incoming inspection of Kestrel's lots. Internal WAL engineering "targets" or aspirational guidelines, whatever their purpose, do not modify the contractual specification unless formally incorporated into the MSA through an amendment. No such amendment was executed. The contractual specification is the ≤0.25% limit in WAL-SPEC-TH-401, Appendix C, Table 3 — and it is the only standard that matters for purposes of determining whether Kestrel's lots were conforming.
(2) The internal emails are engineering preference documents, not incoming-inspection criteria. The internal WAL emails on which Kessler relies are communications among WAL's design and process engineering teams discussing aspirational porosity targets for next-generation turbine housing designs. These communications reflect the engineering team's desire to achieve lower porosity levels as a design-optimization goal — not a directive to the incoming-inspection team to reject lots that meet the contractual ≤0.25% specification. Kessler conflates engineering aspiration with contractual obligation. There is no evidence in the record that WAL's incoming-inspection team applied a ≤0.15% standard. To the contrary, the incoming-inspection rejection reports for all 14 lots consistently cite the ≤0.25% standard from WAL-SPEC-TH-401, Appendix C, Table 3 as the basis for rejection. Each rejection notice references the specific specification provision and identifies the measured porosity level as exceeding 0.25%.
(3) The majority of rejected lots exceeded the ≤0.25% contractual standard. Dr. Chow's analysis shows that 11 of the 14 rejected lots had porosity levels exceeding 0.25% — meaning they would have been rejected under the contractual standard regardless of any internal engineering target. The "unpublished standard" theory is relevant, at most, to the 3 lots with porosity levels between 0.15% and 0.25% — and even those readings are contested, as the exact porosity levels measured by WAL's incoming-inspection team differ from Kessler's re-testing results. For the 11 lots exceeding 0.25%, the "unpublished standard" theory is simply irrelevant, because those lots failed the contractual specification.
(4) Kessler's own limited testing does not conclusively support the theory. As discussed in Section III.B above, Kessler tested only 4 of the 14 rejected lots and found porosity levels ranging from 0.19% to 0.24%. The 0.19% reading (which comes from a single lot) is the only data point in Kessler's testing that clearly falls below 0.25%. The other readings of 0.20%–0.24% are very close to the ≤0.25% limit and are within the measurement uncertainty of Kessler's own testing method. At most, Kessler's data suggests that one lot among the four she tested may have had porosity below 0.25%. This is an exceedingly thin evidentiary basis on which to build a theory that WAL systematically applied an unpublished standard to reject 14 lots.
(5) Kessler does not address WAL's prior acceptance of hundreds of lots during CY1 through CY3. If WAL were truly applying a covert ≤0.15% porosity standard, one would expect to see evidence of this standard being applied consistently throughout the contract term. Yet WAL accepted hundreds of lots from Kestrel during CY1 through CY3 without incident. Kessler does not explain why WAL would have begun applying a stricter internal standard only in CY4, coinciding precisely with the period in which Kestrel's quality performance deteriorated due to raw-material supply issues. The simpler and more parsimonious explanation is that WAL applied the same ≤0.25% standard throughout the contract term, and that the CY4 lots began failing that standard due to issues in Kestrel's manufacturing process and supply chain — not due to any change in WAL's inspection criteria.
Recommendation for Dr. Chow's Rebuttal Report. We recommend that Dr. Chow's rebuttal report on the "unpublished standard" theory include the following:
(a) Demonstrate that the ≤0.25% standard in WAL-SPEC-TH-401, Appendix C, Table 3 was the operative rejection criterion cited in every rejection notice for all 14 lots;
(b) Address the context of the internal WAL engineering emails, explaining the distinction between engineering aspiration (design-optimization targets for future products) and contractual obligation (incoming-inspection acceptance criteria for current deliverables);
(c) Provide complete lot-by-lot data showing that the majority of rejected lots (11 of 14) exceeded the ≤0.25% contractual limit, rendering the "unpublished standard" theory irrelevant to those lots;
(d) Rebut the theory that WAL's true motive was to replace Kestrel with Valkyr by showing that the rejection timeline and quality data are independently justified — i.e., that the rejections coincided with a documented deterioration in Kestrel's production quality, driven by raw-material supply issues, rather than with any change in WAL's inspection criteria; and
(e) Present the chronology of WAL's engagement of Valkyr to show that it postdated the first lot rejection and was a mitigation response rather than a premeditated replacement strategy.
D. Critique of Kessler's Opinion on Kestrel's AS9100 Rev D Compliance
In the final section of her report, Kessler opines that Kestrel maintained full AS9100 Rev D certification throughout the contract term and that Kestrel's quality management system was "adequate for the production of aerospace-grade titanium alloy components meeting the requirements of WAL-SPEC-TH-401 and related specifications." Kessler cites Kestrel's AS9100 Rev D certificate, issued by Aoki Reyes CPAs LLP, and notes that Kestrel's most recent surveillance audit (conducted approximately six months before the first lot rejection) resulted in no major non-conformances.
Weaknesses Identified. We have identified three principal weaknesses in Kessler's opinion regarding Kestrel's quality management system.
(1) AS9100 Rev D certification does not establish lot-specific conformance. AS9100 Rev D is a quality management system standard applicable to the aviation, space, and defense industries. Certification under AS9100 Rev D means that an organization's quality management system has been audited and found to conform to the requirements of the standard. However, AS9100 certification addresses the adequacy of the quality management system — i.e., the organizational processes, procedures, and controls — not the conformance of individual products or lots to specific dimensional and material specifications. A company can maintain a certified quality management system and still produce individual lots that fail to meet particular specifications, just as a hospital can maintain Joint Commission accreditation and still have individual adverse events. Kessler's conflation of system-level certification with lot-specific conformance is a logical error that should be clearly identified in Dr. Chow's rebuttal.
(2) Kessler ignores Kestrel's own internal quality records documenting porosity concerns. Kessler's opinion that Kestrel's quality management was adequate is undermined by Kestrel's own internal quality records, which were obtained in document production. Several internal Kestrel memoranda authored by Dr. Stefan Voss, Kestrel's Head of Quality Assurance, flagged concerns during CY4 about increasing porosity levels in Kestrel's production runs. Specifically, Dr. Voss noted in a memorandum dated May 22, 2022 that incoming raw material from Kestrel's titanium supplier exhibited higher-than-expected oxygen content, which Dr. Voss identified as a potential contributor to increased porosity in the finished turbine housings. A subsequent Voss memorandum dated July 8, 2022 recommended that Kestrel conduct enhanced incoming inspection of raw titanium alloy billets and consider sourcing from an alternative supplier.
These internal documents are significant because they demonstrate that Kestrel itself was aware of a systemic quality issue affecting its CY4 production — an awareness that is inconsistent with Kessler's opinion that Kestrel's quality management was adequate. Kessler does not reference or address these documents in her report, which is a notable omission given that the documents were available in the production record at the time of her engagement.
(3) Kessler does not examine the adequacy of Kestrel's cure attempts. Of the 14 rejected lots, Kestrel attempted to cure all 14 but succeeded in curing only 7. Kessler does not address the cure attempts at all — neither the 7 successful cures nor the 7 failures. She does not opine on whether the cure methods employed by Ke