Board of Contract Appeals General Services Administration Washington, D.C. 20405 __________________________________ GRANTED: June 30, 1998 __________________________________ GSBCA 12503, 13051 PRODUCTS ENGINEERING CORP., Appellant, v. GENERAL SERVICES ADMINISTRATION, Respondent. Jack Rephan of Hofheimer Nusbaum, P.C., Norfolk, VA, counsel for Appellant. Michael J. Noble, Office of General Counsel, General Services Administration, Washington, DC, counsel for Respondent. Before Board Judges PARKER, HYATT, and WILLIAMS. HYATT, Board Judge. Appellant, Products Engineering Corp. (PEC), has appealed the termination for default of its contract for the supply of combination squares (GSBCA 12503) and the contracting officer's subsequent decision assessing excess costs of reprocurement (GSBCA 13051). PEC's contract was terminated for default for failure to timely deliver products conforming to the specifications and for failure to maintain plant quality and process controls. For the reasons stated below, the appeals are granted. Findings of Fact Background 1. PEC is a manufacturer of precision measuring tools and the largest manufacturer of fixed combination squares in the world. PEC has been in business since 1960 and has been selling combination squares to the General Services Administration (GSA) since the early 1970s. Transcript at 461-62, 473. 2. On July 31, 1992, the General Services Administration (GSA) awarded to PEC fixed-price contract number GS-00F-50322 to provide the Government its requirements for combination squares, under National Stock Numbers (NSNs) 5210-00-540-3513 and 5210-00- 078-8949, for the period August 1, 1992 through July 31, 1994. The contract required that the combination squares comply with Federal Specification GGG-S-656D, dated September 25, 1987. The contract also provided that orders for the combination squares were subject to the sampling and testing procedures of MIL-STD- 105. Delivery was required to be made at destination within seventy-five calendar days after receipt of the order. Appeal File,[foot #] 1 Exhibit 1; Appellant's Hearing Exhibit 20. 3. A combination square is a multipurpose precision tool consisting of a ruler, or blade, and an assortment of heads used for a variety of measuring tasks and calculations, such as checking square and marking straight cuts across boards. See Transcript at 466-70. The applicable specification, GGG-S-656D, defined the basic Type VI combination blade for machinists to consist of a graduated blade and a square and miter head with blade-clamping device, level glass, and scriber. If requested by the purchaser, a center head and a protractor head will also be furnished with each unit.[foot #] 2 For the Type VI square, the center head attachment is used to determine the location of a hole in a round piece of steel. The squares were required to conform to stated dimensions set forth in a table accompanying the specification. Section 6.2 of the specification states that the squares and attachments defined therein are to be used by carpenters, diemakers, machinists, and others in layout of angles and in determining angles when fabricating parts from wood, metal, or other materials. Supplemental Appeal File, Exhibit 1 at 11; Transcript at 469. Contract Provisions 4. Included among the general specification provisions applicable to the squares delivered under the subject contract are the following: 3.5 Finish and coatings. All utile or machine-ground and polished surfaces bearing graduations . . . shall have a finish of 32 microinches roughness height rating or better. Determination of surface finish of ----------- FOOTNOTE BEGINS --------- [foot #] 1 All references are to the appeal file in GSBCA 12503, unless otherwise specified. [foot #] 2 The combination squares supplied under this contract included a center head and a protractor head. ----------- FOOTNOTE ENDS ----------- blade edges shall be made by use of standard surface finish masters for visual comparison. Graduations and figures shall have a finish which will make it possible for easy accurate reading. 3.6 Graduations. All graduation lines shall be etched or machine cut to a width and depth . . . as specified for each type and class. The edges of graduation lines shall be sharp, straight, legible, and permanent. Supplemental Appeal File, Exhibit 1. 5. Some of the specific requirements applicable to the Type VI blades are as follows: 3.18.2 Blade. The blade shall be steel, of uniform thickness . . . . Edges and faces shall be straight and true. 3.18.5 Center head. The inside arm surfaces, including their adjacent side surfaces, shall be finished straight and true and square. The arm shall include a 90 degree angle with a tolerance of + 15 minutes, the vertex of which shall center with the edge of the blade. 3.18.7.1 Scale accuracy. The scale error taken at the one-half inch graduation from each end shall not exceed . . . +0.005 inch or -0.0035 inch in 12 inches. 3.18.7.5 Center head. The blade shall bisect the included angle, as measured, of the center head within a tolerance of + 3 minutes. 3.23 Workmanship. The squares shall be free of all burrs, fins, sharp projections, excessive surface roughness, or any other imperfections which may impair their durability, serviceability, and appearance. Supplemental Appeal File, Exhibit 1. 6. Under the contract, PEC was required to maintain an inspection system in compliance with all requirements of Federal Standard 368 and acceptable to the Government for the supplies to be provided. Supplies tendered to the Government for acceptance must have been inspected under that system and found to be in conformance with contract requirements. Although the Government would normally rely upon the contractor's certification as to the quality of supplies shipped, it reserved the right under Federal Acquisition Regulation (FAR) clause 52.246-2 (Inspection of Supplies -- Fixed Price) to inspect and test before acceptance, all supplies provided under the contract. The contract inspection clauses further provided that: During the contract period, a Government representative may periodically select samples of supplies produced under this contract for Government verification inspection and testing. . . . Samples from a rejected lot will be returned to the Contractor or disposed of in a time and manner agreeable to both the Contractor and the Government. Appeal File, Exhibit 1 at 12-13. 7. Section 4.2 of the specification sets forth guidelines for sampling lots to determine conformance with specifications. For specification items to be tested, the proper sample size is thirteen. For other features to be examined, the proper sampling size is eight. Section 4.3 of the specification provides for visual and dimensional examination of the squares selected under the sampling procedures for the various defects listed in table XIII. Listed as major defects under table XIII are the following items: 101 Type, length, and width (of blade) not as specified. 122 Graduations or figures illegible, incorrect, not sharply defined, not straight, not uniform, graduations not etched, machine cut or stamped, inch graduations not numbered, or graduations not as specified (except type IX). Supplemental Appeal File, Exhibit 1. 7. Section 4.4 of the specification governs testing of samples for conformance to visual and dimensional requirements. Section 4.4.1 states that each combination square shall be tested to determine conformance with the specification's dimensional requirements, as follows: 4.4.1 Dimensions. Each sample square shall be tested at a temperature of 68 degrees F. Whenever practicable, to determine conformance with the applicable dimensional requirements of this specification, readings taken at other temperatures shall be adjusted to 68 degrees F. 4.4.2 Out-of-squareness. All sample squares shall be tested to determine their squareness. Squareness shall be determined by means of a surface plate and a precision cylinder square used with a comparator-square or with a master square of known accuracy. 4.4.3 Squareness test for inside. All sample squares shall be tested to determine their accuracy. A surface plate and angle plate may be used with a thickness gage to measure the error. 4.4.4 Angular trueness. All sample type VI squares with protractor heads shall be tested for angular intervals. This shall be determined by means of either a circular measuring machine, sine bar, precision angle blocks, or other methods of equivalent accuracy. 4.4.5 Miter accuracy. Samples of type I and VI squares shall be tested for miter accuracy. This shall be determined by means of either universal angle gages or, an approved master miter of known accuracy may be used. 4.4.6 Hardness. Hardness testing shall be performed in accordance with ASTM E 18. 4.4.7 Bevel Accuracy. Bevel accuracy of the types IV, V, and VII blades and the miter of the type III, style 2 beam shall be tested for bevel accuracy. This shall be determined by means of a universal protractor of known accuracy reading to 5 minutes. 4.4.8 Accuracy. Type IX square shall be checked for accuracy by the use of a coordinate measuring machine or by other means approved by the Government. Supplemental Appeal File, Exhibit 1. 8. The standard clause applicable to default termination of a fixed-price supply or service contract was included in the contract. Among other things, this clause permitted termination of the contract for default for the contractor's failure to deliver supplies within the time specified under the contract or any extension thereof. The contract provisions governing inspection of supplies also stated that the contract could be terminated for default if subsequent Government inspections disclosed that plant quality or process controls were not being maintained, that supplies not conforming to the specification requirements were being shipped, or if there was a failure to comply with any other requirement of the inspection clause. Appeal File, Exhibit 1. Contract Performance and Administration 9. PEC had a quality assurance system that was approved by GSA. PEC's quality assurance specialist had a checklist, which included numerous characteristics to be verified, to complete when product was ready for shipment. Under this system, a variety of instruments, including those referenced in section 4.4 of the specification, were used to test the combination squares for conformance with applicable tolerances. Transcript at 481- 83; Supplemental Appeal File, Exhibits 1, 30. 10. The contract required first article testing. Appeal File, Exhibit 1. GSA witnessed the first article tests on September 10, 1992, and by letter dated September 14, 1992, notified PEC that its first article and first article test reports had been approved. Appeal File, Exhibit 2; Appellant's Hearing Exhibit 20. 11. On September 2, 1992, GSA notified PEC of its intention to invoke its right to inspect orders filled under the contract. Subsequently, on September 16, 1992, GSA inspected the first such order, number FWJ-92151. Samples were selected from the lot, and shipped to the GSA National Laboratory in San Francisco for evaluation. Under the sampling plan as interpreted by GSA, thirteen squares were tested. Five of these thirteen squares were rejected for failing the length of blade requirement under the specification. Appeal File, Exhibit 3. The samples were tested with a coordinate measuring machine (CMM), which had been purchased by the GSA laboratory in 1986. Transcript at 62-63. 12. GSA's testing laboratory was certified and accredited by the American Association of Laboratory Accreditation. Transcript at 78-80. During the period relevant to this dispute, 1992-1993, the equipment used was calibrated annually in May. Respondent's Hearing Exhibits 2-3; Transcript at 105. The operators of the equipment had training in various metrology methods and in the use of the CMM. Transcript at 56-57, 77-78. 13. GSA's materials engineer, who operated the equipment and conducted the tests on PEC's squares, testified that the CMM was used instead of the usual metrology methods to achieve enhanced repeatability and to minimize the effects of operator dependency. He stated that under the older metrology methods results depended heavily on the experience of the analyst conducting the test. Transcript at 103-04. On cross- examination, he admitted that he had no familiarity with the use of the cylinder gauge to test angularity, and stated that the GSA laboratory did not have such a gauge. Transcript at 222. 14. The applicable specification does not mention testing with a CMM except with respect to the testing of Type IX squares, see Finding 7, which consist of a blade and a beam, with no center head or miter attachment. For Type IX squares, section 4.4.8 provides that accuracy shall be checked with a CMM or by other means approved by the Government. Supplemental Appeal File, Exhibit 1. 15. The laboratory checked with the specification writer and was assured that the use of the CMM to test various characteristics of the combination squares was not inconsistent with the specification's testing provisions. Respondent's Hearing Exhibit 22; Transcript at 98-100. 16. The tests were conducted pursuant to a written GSA policy, FSS P 2900.6, CHGE 9, that provided the following general guidance with respect to the extent of testing of supplies by GSA laboratories: (2) Complete testing shall be the objective of the laboratory. The laboratory head may permit reduced testing only when he has determined that a sufficient number of samples under the same specification from the same supplier have been completely tested and accepted. However, all key characteristics of salient features must be tested each time. (3) Complete testing shall be performed on bid samples, preproduction samples, quality approved manufacturer verification samples, qualification samples, and the first lot of any contract not previously subjected to preproduction or qualification inspection. (4) With the exception of the testing cited in (3), above, testing shall be performed to the point of major failure; i.e., a failure that would cause the material to be rejected regardless of the outcome of other test results. Respondent's Hearing Exhibit 1. The samples taken from PEC's shipments under this contract were quality approved manufacturer verification samples, for which the policy prescribes complete testing. Transcript at 424-25. Nonetheless, the GSA laboratory consistently stopped testing at the point at which a failure was first identified and did not test for further defects in the sample units. Supplemental Appeal File, Exhibits 16-26. The employee who conducted the tests of PEC's squares under the defaulted contract testified that this practice had been adopted to expedite sample turnaround time. Transcript at 93-94. 17. On October 21, 1992, a quality deficiency notice was issued to PEC with respect to the combination squares under order FWJD-92151 that had been tested by GSA's laboratory. The quality deficiency notice explained that five of the blades were below the minimum length requirement. Appeal File, Exhibit 3. As a result of these failures, PEC was placed on lot-by-lot inspection. Appeal File, Exhibit 4. 18. In response to the quality deficiency notice, PEC wrote to GSA explaining that it had reinspected its stock and determined that a small percentage of twelve inch blades were under the minimum length requirement. Although 100% of machinist blades were checked for length, the nonconforming blades were not caught because of a random error in PEC's digital caliper. To avoid the possibility of future errors in blade length, PEC switched to an electronic height gauge to measure the overall length of the blades. Appeal File, Exhibit 4; Transcript at 498, 553. 19. After the initial rejection of order FWJD-92151, GSA tested samples from two additional lots, and found that blades in the lots were similarly not in conformance with section 3.18.7.1's stated tolerances for length of blade. The laboratory reports for samples from two orders which were taken on November 2, 1993, and received in the laboratory on November 4, 1993, reflect that two of thirteen blades failed the test for scale accuracy for one lot and one of thirteen failed for the other lot. The tests also reflect that GSA did not test the samples for any other deficiencies. Supplemental Appeal File, Exhibits 16-17; Respondent's Hearing Exhibit 4. 20. GSA submitted these same samples to an outside laboratory for verification. The reports provided to GSA by this laboratory confirmed the determination of the GSA laboratory that the sample squares did not meet the specification's tolerance requirements for scale accuracy. Respondent's Hearing Exhibits 5-6; Transcript at 96, 109. 21. On January 14, 1993, GSA issued a cure letter to PEC. That letter points out that after issuance of the quality deficiency notice concerning scale accuracy, and receipt of PEC's response on October 29, two subsequent samples taken on November 2, see Finding 16, failed testing for the same characteristic. As a result, GSA stated its concern that efforts to correct the situation had been ineffective, causing the agency to doubt whether PEC's quality control system was adequate to identify and correct deficiencies. The cure letter asked PEC to respond in ten days with a written and detailed plan showing the steps implemented to remedy the "unsatisfactory condition," and warned that recurrences of the noted deficiencies, i.e., scale accuracy and blade length discrepancies, could result in termination of the contract without further notice. Appeal File, Exhibit 4. 22. By letter dated January 19, 1993, PEC responded to the cure notice. In this letter, PEC took issue with the sample size taken for purposes of inspecting for scale accuracy. PEC took the position that length of blade was a visual and dimension defect under specification section 4.3, rather than a testing deficiency under section 4.4, and under the applicable military standard, the proper sample size was eight, rather than thirteen, combination squares. As such, PEC argued, one of thirteen would not have failed. PEC also made a request, in this letter, to witness dimension inspection and testing of samples taken from three additional lots. Appeal File, Exhibit 5. 23. GSA responded to PEC's letter in a letter dated February 24, 1993. This letter addressed PEC's contention that GSA used the wrong sample size to reject its combination squares and that if the proper sample size had been used, the lot would have passed inspection. In GSA's view, to determine scale accuracy, testing, as opposed to dimensional and visual examination, is required and called for under the specification. In addition, GSA pointed out that PEC itself tested thirteen samples for scale accuracy under its own inspection program. This letter also expressed GSA's understanding and approval of PEC's efforts to correct deficiencies and stated that under the circumstances GSA did not "see the benefit of a lab visit at this time." Appeal File, Exhibit 6. 24. On March 3, 1993, the administrative contracting officer wrote to PEC addressing three orders sampled and rejected under the subject contract: P.O. No. Date Sampled Due Date Rejection Date FWJD710-1 1/6/93 1/28/93 2/17/93 SWJD706-1 1/6/93 1/28/93 2/24/93 FWJD785-1 1/13/93 2/14/93 2/24/93 The letter required PEC to respond within five days, explaining the basis for the "delinquent status"of these orders, and providing details of the conditions leading to the "delinquencies." The letter further stated that the Government did not intend to waive its rights under the contract. Appeal File, Exhibit 10. 25. The record contains at least ten test reports issued by the GSA laboratory with respect to the results of inspecting and testing PEC's combination squares for compliance with specifications. Test reports for samples received in the laboratory on November 4, 1993, from two different orders, reflect rejections based on failure to pass scale accuracy requirements. These tests were conducted on November 19 and 20; the reports were issued on December 11. Supplemental Appeal File, Exhibits 16-17. A sample from yet another order, received in the laboratory on November 12, and tested on December 4, passed all of the tests and was determined to be compliant. Supplemental Appeal File, Exhibit 18. Laboratory test reports issued on January 28, February 11, February 12, and February 17 similarly reflect rejections based on problems with scale error and other imperfections in the blades. Once these problems were determined, testing was halted. Supplemental Appeal File, Exhibits 19-22. Laboratory reports issued on March 30, April 27, and May 6 found all squares in the samples, which were received in the laboratory on March 12, March 25, and April 5, respectively, to be compliant in all respects. Supplemental Appeal File, Exhibits 23-25. 26. By letter dated March 8, 1993, PEC responded to GSA's March 3 letter . In its response, PEC pointed out that each order had been delivered approximately three to four weeks before the contract delivery date, and that the "delinquencies" were a direct result of the laboratory's delays in testing and notifying PEC of rejections. In addition, PEC explained that it was actively working with the agency to clear up problems and redeliver conforming squares. Appeal File, Exhibit 8. 27. In a letter dated March 15, 1993, PEC updated the administrative contracting officer on its progress in identifying and resolving problem areas with respect to the rejected lots. PEC requested GSA to allow resubmission of the lots by March 31, 1993. Appeal File, Exhibit 8. In response, the administrative contracting officer agreed to modify the contract to reestablish the delivery due dates on the affected orders by extending the delivery dates to March 31 in exchange for a mutually agreed upon consideration of $200. The letter cautions that the delivery clause of the contract is not automatically negotiable, however, and that failure to make timely delivery under the modification or on future orders could result in termination of the contract for default. Appeal File, Exhibit 9. 28. PEC redelivered the lots forming the basis for the modification of the contract. On April 5, 1993, samples from order FWJD-7101 were received in the GSA laboratory. Tests were conducted on May 10, 1993. The report issued by the laboratory on May 19 reflects that various features were tested, and that the lot was rejected because one of the thirteen squares tested failed the tolerance for center head accuracy of the bisect angle. The tolerance for the bisect angle was + 3 minutes; the CMM found a deviation of more than thirteen minutes. The angle of the center head arms was not tested. Appeal File, Exhibit 11. 29. PEC's vice president and general manager, Paul Wien, testified that PEC's inspector used a cylinder gauge to verify the included angle of the center heads. This gauge had been used successfully for many years and PEC had no reason to believe that the angle of the center heads would not pass GSA's inspection at the time of delivery. Transcript at 484-85. Mr. Wien further attested that a variance of the center head's bisect angle for specified tolerances is minor in nature and would not render the product unusable for its intended purpose. Additionally, he confirmed that the defects could have readily and quickly been corrected by replacement of any center heads that did not meet the tolerance. Transcript at 506, 542-43. 30. The CMM used by GSA to test for center head accuracy is a large granite plate with two bars across and an I-beam across the top. Attached to the machine is a probe with a little ball on the end. The probe is operated to touch the part to be tested in various places to determine an angle. Transcript at 536-37. Mr. Wien also stated that the customary use of the CMM is to check machined products, to measure holes, such as valve or piston holes, and to measure the angles of a block -- e.g., the block of a car. Transcript at 513. 31. Specifically with respect to the center head measurement, Mr. Wien testified that the CMM can repeat a single reading, but to test properly the reading should be of a plane under the specification which refers to the edge of the blade to one arm. According to Mr. Wien, the variance within the CMM itself was 2.8 minutes, while the tolerance for the bisect angle was only three minutes. Thus, if the angle was within tolerance, but slightly off, the variance within the CMM itself could cause a finding that the center head bisect angle error exceeds three minutes. Transcript at 537-38. Mr. Wien further testified that the use of the CMM can also result in inaccurate and variable readings due to operator practices. For example, with respect to the bisect angle of the center head, he explained that the reading will vary depending on how tightly the center head is affixed to the blade. The clamping motion required to tighten the blade and center head can change the angle and thus the reading. Transcript at 565-66. 32. On May 25, 1993, the GSA quality assurance specialist (QAS) assigned to PEC issued a quality deficiency notice informing appellant of the rejection of order FWJD-710-1 for failure to meet the tolerance specified for the bisect angle of the center head. In a letter dated June 10, 1993, PEC wrote to GSA's Assistant Commissioner for Quality and Contract Administration to express its concerns about the practices of the GSA laboratory. The letter raised a number of issues, including the rejection of the resubmitted order FWJD-710-1. PEC took exception to GSA's insistence that it request copies of the laboratory reports under the Freedom of Information Act in order to obtain access to these documents, and also objected to the practice of discontinuing testing after the rejection of a unit based on one characteristic. PEC explained that these practices frustrated the company's desire to investigate and take expeditious action to correct any problems with its products. With respect to the rejection of this specific order, PEC acknowledged that the QAS had shown it a copy of the laboratory test, and stated that in its view the test had been conducted improperly because the angle of the center head arms had not been measured. Supplemental Appeal File, Exhibit 28. 33. The Assistant Commissioner responded to PEC's letter on June 29, 1993. In addressing PEC's concerns, he stated that (1) complete testing is preferred, (2) in the future it would not be necessary to invoke the Freedom of Information Act to obtain copies of laboratory test reports, and (3) that the laboratory had been advised to permit the contractor to observe testing and to discuss testing techniques and procedures. He disagreed, however, with PEC's contention that the test to determine the bisect angle had been conducted improperly. Supplemental Appeal File, Exhibit 29. 34. On June 18, 1993, GSA terminated contract number 50322 based on PEC's failure to deliver conforming Type VI combination squares under order number FWJD-710-1 by the extended delivery due date. The letter noted that the rejection of this reoffered order, based on the failure of the bisect angle to meet the specified tolerance, constituted the second rejection of this lot and that eight out of twelve laboratory tests conducted on PEC shipments had been rejected for various defects (scale accuracy and length of blade), leading the Government to conclude that PEC's quality control system did not conform to contract requirements. Appeal File, Exhibit 12. The administrative contracting officer, in a memorandum of position addressing the reasons for terminating the contract for default, made similar observations concerning PEC 's delivery, on eight of twelve occasions, see Finding 25, of nonconforming combination squares, which caused him to conclude that PEC had failed to practice effective quality control. Appellant's Hearing Exhibit 20. 35. The letter terminating PEC's contract for default was signed by contracting officer William Lacey, on behalf of the assigned contracting officer, Paul Swanson, because of Mr. Swanson's unavailability. Transcript at 420. Mr. Lacey was authorized to sign the letter on behalf of Mr. Swanson. In addition, Mr. Lacey was himself informed about the decision to terminate for default and had independent knowledge of the GSA laboratory tests finding that PEC's combination squares were not meeting specification tolerances. Transcript at 420-23. 36. Subsequent to the termination for default of PEC's contract, GSA conducted a reprocurement for combination squares to replace those to have been supplied under contract number 50322. In February and April of 1994, GSA awarded contracts to PEC's wholly owned subsidiary, Sigma West Corporation, for replacement combination squares. Appeal File, GSBCA 13051, Exhibits 4-5. The squares purchased from Sigma West were of substantially higher quality than those to have been supplied under contract number 50322 and, as a consequence, were considerably more expensive. Transcript at 544-45. On October 12, 1994, the contracting officer issued a decision assessing reprocurement costs in the total amount of $114,414.74. These costs were comprised of the added costs of purchasing the combination squares ($96,733.41) and administrative costs associated with the conduct of the reprocurement ($17,681.33). Appeal File, GSBCA 13051, Exhibits 6-7. 37. After termination of its contract, PEC requested return of the samples of squares that had been selected for testing. GSA did not return the samples because it had decided to hold them at the laboratory pending litigation. During discovery conducted in the course of this appeal, the Government informed appellant that, except for nine squares, it no longer had any of the samples and could not return them to PEC. GSA has not been able to provide an explanation as to why the remaining squares are missing. Transcript at 165, 168, 377-78, 433-45. Prior Sales of Type VI Combination Squares 38. PEC has been selling Type VI combination squares to GSA since the late 1960's. Since 1984, and exclusive of the units shipped under the subject contract, PEC had sold to the Government more than 43,000 of the same combination squares as those purchased under contract number GS-00F-50322. Since 1987, these sales were subject to the same specification and sampling and testing procedures as those imposed under contract number GS- 00F-50322. Transcript at 473-78. Copies of prior contracts under which Type VI combination squares were sold to the Government under the same terms and conditions as those applicable to contract 50322 have been provided by appellant. Appellant's Hearing Exhibits 4-17. 39. Prior to the rejections experienced by PEC under this contract, PEC was not aware of any GSA test results showing its product to be nonconforming under the specification.[foot #] 3 There is testimony and evidence in the record that on two occasions under prior contracts the laboratory had determined that PEC's combination squares had failed the bisect angle tolerance requirement. These two occasions reflect testing performed in response to customer complaints. Respondent's Hearing Exhibits 30, 31. The laboratory reports generated show that the dates of manufacture and contract numbers are unknown. There are no reports or letters showing that this information was ever transmitted to PEC. Although the laboratory employee who performed the tests testified that he believed Mr. Wien had been notified of these test results during a visit to the laboratory that took place in the late 1980s, he had very little recollection of the purpose of or detailed issues discussed in that visit and could only speculate that PEC would have been informed of these test results. Transcript at 59-61. 40. Mr. Wien testified that the visit referred to by the laboratory technician was in connection with test results relevant to another tool altogether and that he was not at that time informed of any testing taking place on Type VI combination squares. He unequivocally stated that he had no knowledge of the use of the CMM to test Type VI squares prior to the 1992 contract. Transcript at 473-74. 41. Since at least 1984, PEC has used the same type of cylinder gauge and quality assurance system to ensure that the combination squares shipped to GSA met the tolerance requirements of the specifications. Throughout this period, PEC's quality assurance systems and testing methods have been observed and approved by GSA. The instruments used by PEC to test the squares for compliance were regularly calibrated. During that period PEC shipped and GSA accepted more than 43,000 squares with the same blade and center head that were the subject of rejections under contract number GS-00F-50322. PEC asserts that until GSA began testing with a CMM under contract number GS-00F-50322, there were no rejections of combination squares for failure to meet tolerance requirements. Transcript at 473-84; Appellant's Hearing Exhibit 25. Discussion It is well-settled that a termination for default is a drastic sanction, to be imposed upon a contractor only for good cause and on the basis of solid evidence. DeVito v. United States, 413 F.2d 1147 (Ct. Cl. 1969); Foremost Mechanical ----------- FOOTNOTE BEGINS --------- [foot #] 3 Mr. Wien testified that to his knowledge the only rejection of a Type VI combination square supplied by PEC involved a single item shipped to a military base in the early 1980s. Transcript at 473. This was an out-of-warranty product and the incident has no relevance to the matter before us. Transcript at 478. ----------- FOOTNOTE ENDS ----------- Systems, Inc. v. General Services Administration, GSBCA 12335, et al., 95-1 BCA 27,382 (1994); A.J.C.A. Construction v. General Services Administration, GSBCA 11541, et al., 94-2 BCA 26,949, at 134,203. Default terminations are strictly construed. DeVito, 413 F.2d at 1153. It is, moreover, the Government's burden to prove, by a preponderance of the evidence, that the termination for default was justified. Lisbon Contractors, Inc. v. United States, 828 F.2d 759 (Fed. Cir. 1987). In evaluating whether the Government is justified in exercising its right to terminate a contract for default, we look to the facts and circumstances giving rise to the termination. In this case, the Government has met its initial burden to produce evidence in support of its decision, but has not sustained the overall burden of persuasion that the default termination of PEC's contract was warranted. The termination for default of this contract was initiated after the second rejection of order number FWJD-710-1. The original due date for this order was January 28, 1993. PEC tendered the shipment on January 6, 1993. Samples were taken promptly, but were not tested by the GSA laboratory until mid- February 1993. Following the initial rejection of the lot for nonconformance with scale accuracy requirements, and issuance of a cure notice, the delivery due date was extended by agreement of the parties until March 31. PEC corrected the problems giving rise to the original rejection and redelivered this order by March 31. Samples from the order arrived in the GSA laboratory on April 5; tests were not conducted until May 5. This time, the squares were rejected for an entirely different reason -- failure of the center head bisect angle to conform to the tolerances set forth in the specification. The Government argues that the foregoing evidence, supported by the written records of laboratory testing and the testimony of the GSA employees conducting the tests, satisfies its burden to show the termination was proper based upon failure to make timely delivery of conforming products and failure to maintain an adequate quality assurance system. In support of this position, GSA points to the test results showing the previous rejections of squares for failure to meet specifications governing length of blade and to the final rejection of order FWJD-710-1 based on failure to meet the tolerances for the bisect angle. GSA contends that the resultant failure to timely deliver conforming squares under order FWJD-710-1, together with the history of rejections under this contract, which demonstrated unresolved problems with PEC's quality assurance system, justified the termination for default of the contract. Although this suffices to satisfy GSA's initial burden to come forward with evidence supporting the termination decision, PEC has raised a number of defenses to the default termination of its contract. These include challenges to the validity of the testing methods and procedures used by the Government, a defense based on prior course of dealing, and the doctrine of substantial compliance. PEC also argues that GSA's failure to return the samples tested constituted a breach of the contract, placing PEC at a disadvantage in defending itself, and should nullify the test results forming the basis for the termination. The Testing Method PEC challenges the propriety of the testing method used by the GSA laboratory to determine conformance with specification requirements. In particular, PEC contends that the CMM was not at the time commonly used to evaluate small precision instruments such as the Type VI square, had a sufficient margin of error due to its own tolerances to produce skewed results in testing the various characteristics of PEC's squares, and was not the type of instrument contemplated to be used to test Type VI squares under the specification. In particular, PEC takes exception to the use of the CMM to test the center head bisect angle of the samples and argues that the Government's long-standing approval of its inspection system entitled it to rely on the use of the gages and instruments under that system for testing conformance to specifications. For its part, respondent maintains that the testing provisions of specification GGG-S-656D do not identify a particular means of determining compliance with the specification, and that use of the CMM to determine whether the combination squares conformed to the specification was thus within the prerogative of the Government. Respondent further asserts that it has shown through the testimony of its laboratory employees that this method of testing was reasonably calculated to produce accurate test results. We address first the differing interpretations of section 4.4 of the specification, which addresses testing of particular characteristics . GSA urges that the specification does not delineate or restrict the Government to any particular means of testing squares for compliance with specification tolerances. PEC disagrees, pointing out that most of the sections that address testing identify a particular means of doing so and at most allow for use of an equivalent method. See Finding 7. PEC also points out that the only mention of a CMM is for the testing of the accuracy of a Type IX square -- a tool that has no center head. Finding 14. PEC urges that the omission of the CMM as a suitable instrument for testing other squares and their individual characteristics supports the conclusion that the specification writers recognized that the CMM was not suitable for testing other combination squares. Neither position is entirely correct with respect to the respective proposed interpretations. Although we agree with PEC that the testing provisions overall are rather specific as to the types of gages and instruments to be used to test certain characteristics, see Finding 7, the specification is silent with respect to testing of the center head bisect angle tolerances. This does not permit the conclusion, however, that this characteristic should not be tested or that the Government is limited to the test method used by PEC to determine compliance with this characteristic. The specification sets forth tolerances applicable to the bisect angle -- as such it is reasonable to conclude that the squares may be inspected and tested for compliance with this specification. Since no particular test method is specified or recommended for this characteristic, the issue becomes whether the test used is reasonable for evaluating conformance to specifications. See Technics EMS, Inc., GSBCA 6679-COM, 84-1 BCA 17,060. In Technics EMS, the Board determined that the Government's test was reasonable because it consisted of operation of the machine in the manner in which it was designed to be used. The individuals operating the machine were scientists who were familiar with the underlying theory, function and operation of the machine. Unlike Technics EMS, the testing procedures used by the parties here did not consist of using the tools to determine if they functioned properly. Rather, various precision instruments, subject to their own tolerances and deviations from specifications, were used to evaluate the properties of the Type VI combination squares. As a general proposition, the Government is entitled to use whatever test it prefers to determine contract compliance, so long as it does not impose a higher standard than that included in the contract. Nash Metalware Co. v. General Services Administration, GSBCA 11951, 94-2 BCA 26,780, (citing Air, Inc., GSBCA 9847, 91-1 BCA 23,352); cf. Pinay Flooring Products, Inc, GSBCA 9286, 91-2 BCA 23,682 (termination based on appellant's inability to test whole roll of flooring material rather than a limited portion of it as previously agreed to by the parties was improper). The evidence of record does not establish that the CMM would necessarily result in the imposition of a higher standard of testing than that provided for in the specification or established by prior conduct of the parties. There is evidence suggesting that the results of the CMM may be subject to error, particularly with respect to measurement of the bisect angle. At the same time, at least with respect to the earlier rejections, it appears that there is similarly a possibility that the rejections could be justified based on potential errors in the gages used by PEC. In these circumstances, we do not hold that the Government could not use the CMM to test the squares for compliance or that it was required to notify PEC in advance of any decision to use this testing method. As discussed in more detail below, however, the conflicting evidence concerning the validity of the CMM results does cast doubt on the propriety of the decision to terminate this contract for default.[foot #] 4 ----------- FOOTNOTE BEGINS --------- [foot #] 4 We view the Government's inability to locate or return the samples in a similar light. Although PEC makes much of the fact that it has been deprived of the ability to test the (continued...) ----------- FOOTNOTE ENDS ----------- Prior Course of Dealing PEC has also challenged the default termination of its contract on the ground that even if the tolerance deviations for the bisect angle of the center head were not met, the facts of record establish a prior course of dealing barring the termination for default of its contract for failure to conform to the specifications. In particular, PEC contends that the Government's previous approval of PEC's quality control system and methods of testing for compliance with the specification bars the use of a completely different method of testing, particularly where that method has not been shown to be equivalent to the approved method and produces varying results because of built-in error in the testing apparatus. Specifically, PEC argues that the evidence of record shows an established course of dealing under which GSA had, since 1984, under previous contracts repeatedly accepted the same equipment, subject to identical specifications, based on the same quality control procedures implemented by PEC without subjecting the equipment to testing with a CMM. This long history led PEC to believe that combination squares which were within tolerance when tested pursuant to PEC's own quality assurance system would be acceptable to GSA. PEC competed for this contract on this basis. Evidence of a prior course of dealing may establish the intent of the parties with respect to the proper interpretation of contract language. It may also establish that a contract requirement has effectively been waived: "A contract requirement for the benefit of a party becomes dead if that party knowingly fails to exact its performance, over such an extended period, that the other side reasonably believes the requirement to be dead." Gresham & Co. v. United States, 470 F.2d 542, 554 (Ct. Cl. 1972); accord General Security Services Corp. v. General Services Administration, GSBCA 11381, 92-2 BCA 24,897, at 124,169-70. PEC relies heavily upon the Board's decision in Unlimited Supply Co. v. General Services Administration, GSBCA 12371, 94-3 BCA 27,170, to support its claim that the default termination was improper. In Unlimited Supply, the Government terminated for ----------- FOOTNOTE BEGINS --------- [foot #] 4 (...continued) samples for itself, and that the failure to return these articles constitutes a breach of the contract, we do not regard the loss of the sample units as an event invalidating the test results altogether. The test results were recorded in writing by the laboratory at the time the tests were conducted and supported by the testimony of the GSA employee who conducted them. Loss of the samples and any possible prejudice to PEC is, however, another factor to be weighed in assessing the propriety of the default termination. ----------- FOOTNOTE ENDS ----------- default a contract for the supply of stainless steel mixing bowls on the ground that the bowls did not meet the specification for capacity. The Board overturned the default termination principally because of the existence of conflicting evidence in the record as to whether the bowls actually met the capacity requirement; the conflict was resolved against the party with the burden of proof. Additionally, however, the Board observed that a prior course of dealing had been shown under which the Government had, under nineteen prior purchase orders subject to the same provisions as the one that had been terminated, accepted the same bowls, made with the same molds. The record contained no evidence that the Government had ever previously advised Unlimited Supply of its discovery that the bowls did not conform to specifications, although apparently some testing performed after acceptance on earlier contracts had alerted the Government to this fact. Because appellant relied on the prior unqualified acceptances in pricing the terminated orders, the Government would not be permitted, without notice, to exact strict conformance with the specifications. Id. at 135,391, 135,393-94. PEC's situation is not identical to the facts of Unlimited Supply. The PEC contract was terminated because GSA, after testing various lots, as it was entitled to do under the instant contract, determined that the combination squares did not meet the specifications. Although PEC has provided testimony that its combination squares were accepted by GSA on numerous prior occasions under the same contractual terms and conditions, in contrast to Unlimited Supply the record does not reflect that GSA knowingly accepted nonconforming products. There is virtually no evidence concerning prior inspections by GSA prior to acceptance of the squares under prior contracts. At most, it appears that the agency accepted products based on PEC's own inspections.[foot #] 5 Nor do we believe that the ----------- FOOTNOTE BEGINS --------- [foot #] 5 The record does not enable us to conclude, however, as the Government contends, that PEC had reasons to know of prior tests conducted with a CMM that identified discrepancies in its combination squares. The testimony of GSA's witnesses concerning the two laboratory reports that predated the testing under the subject contract was inconclusive. These witnesses had no particular recollection of Mr. Wien's visit to the laboratory other than that it occurred. Although they state that they believe they would have disclosed the use of the CMM to Mr. Wien based on their typical practice, they did not affirmatively recall making such a disclosure. Findings 37-38. There is no written evidence to corroborate any of this testimony. The reports were generated not in connection with acceptance testing, moreover, but in response to customer complaints. The record does not show the age of the squares or identify the prior contracts under which they were sold to GSA. As such, these reports establish neither prior knowledge of defects on PEC's part, nor a waiver of tolerances on GSA's part. ----------- FOOTNOTE ENDS ----------- circumstances warrant extending the prior course of dealing doctrine to bar the Government from using different test instruments from those used by PEC without prior notification. As we stated above, GSA is entitled to use a different test method, if not precluded by the specification, so long as it does not impose a more rigorous standard than that provided in the specification. GSA's knowing failure to enforce the tolerance requirements is a prerequisite to the application of the prior course of dealing doctrine to these circumstances. The evidence thus is not sufficient to invalidate the default termination under the prior course of dealing doctrine. Nothing in the record suggests that GSA knowingly accepted squares that were out of tolerance or that GSA gave PEC reason to believe that the Government was willing to waive the tolerance requirements. Substantial Compliance Doctrine Finally, PEC argues that the shipments made were timely because delivered in the good faith belief that the products conformed to specifications and would be accepted. PEC points out that delays on the part of GSA's laboratory in testing samples were contributing factors in the alleged failure to deliver conforming goods on time. PEC delivered its initial lots well in advance of the delivery date and the reoffered lot was also delivered by the reestablished due date. Findings 24, 28. The shipments were offered in the good faith belief that the products complied with the specification. Finding 29. Under Radiation Technology, Inc. v. United States, 366 F.2d 1003 (Ct. Cl. 1966), PEC argues, it should not have been summarily default terminated, but rather should have been granted an extension of time to correct minor defects. In Radiation Technology, the Court addressed the circumstances where timely delivery of supplies might give rise to an entitlement to a reasonable time to cure where an unexpected rejection occurs. This decision delineates the doctrine of substantial compliance, which is predicated on the contractor's ability to meet certain prerequisites: (1) the supplies must have been delivered or tendered by the due date; (2) the contractor must have had a reasonable belief that the supplies conformed to the contract requirements; and (3) the defects must have been minor in nature and readily correctable. 366 F.2d at 1005-06. Looking at the record here, we find that the doctrine of substantial compliance is applicable. PEC did in fact deliver by the extended due date. Although the parties disagree as to the reasonableness of PEC's belief that the product would be found to be conforming, we find that the preponderance of the evidence supports PEC's position on this point. Mr. Wien convincingly testified that he had no prior reason to anticipate that a CMM would be used to test the features of the combination squares supplied under this contract, or, if it was, that the center heads would not meet the specified tolerances. The shipment was inspected, with the instruments customarily used by PEC to test for center head tolerances, and passed PEC's inspection. PEC had focused on resolving discrepancies in the length and accuracy of the blades, which had been raised by the laboratory tests under prior shipments. These problems were resolved. PEC, which had not been given copies of the lab reports, had no reason to expect that a new basis for rejection would arise when testing focused on another characteristic of the squares. Finally, Mr. Wien's unrebutted testimony establishes that the discrepancy identified in the center head angular tolerances was minor in nature, even in the face of what appears to be a substantial deviation from the allowable tolerance. Mr. Wien stated that the discrepancy, if it existed, would not affect the functionality of the tool, which would still locate the center of a round piece of steel for drilling, and could easily and readily be fixed by replacing any nonconforming center head attachments. Although the Government contends that these considerations are irrelevant because time was of the essence, this position is weakened by the many delays imposed by the Government itself in the testing and is further vitiated by the lengthy time that passed in the reprocurement process. There is no basis in the record to conclude that the problem identified was major or could not have been rectified in a reasonable time frame. GSA disagrees with PEC's argument that the substantial compliance doctrine should apply, urging that the Board should consider various other factors. For example, GSA contends that while it did not formally issue a cure notice as to the center head bisect angle in the squares under the reoffered shipment, it did not have to. Citing Technics EMS Inc., GSBCA 6679-COM, 84-1 BCA 17,060, GSA argues that the substantial compliance doctrine is inapplicable where an opportunity to cure has already been afforded or the Government needs the supplies. Here, in GSA's view, the contractor had been previously told in a cure notice that its quality control system appeared to be inadequate and that it would be on lot-by-lot inspection until the problem was resolved. In GSA's view, the termination is not covered by the substantial compliance doctrine because PEC was already given an opportunity to cure its quality control system and did not do so. In addition, GSA argues that the substantial compliance doctrine should not apply where the contractor shows an unwillingness to cure the defects. See Kain Cattle, ASBCA 17,124, 73-1 BCA 9999. PEC knew about the center head rejection in advance of receiving the termination notice and did not of its own volition offer to replace defective center heads. Finding 21. GSA construes PEC's failure to offer explicitly to cure or replace any defective center heads as an unwillingness to cure the defects. The difficulty with respondent's arguments arises from the overall circumstances in which the termination for default took place. Although the record does not establish clearly one way or the other whether the CMM produced testing results that were comparable to the results achieved under PEC's quality assurance system, as PEC points out, its products had been accepted without incident for many years until GSA undertook testing with a CMM. Although GSA informed PEC of its view that the quality control system was inadequate, as a result of GSA's practice of stopping testing at the point of first major failure PEC was aware only of the prior bases for rejection.[foot #] 6 PEC had no reason to anticipate that other characteristics tested under its own quality assurance system would similarly fail inspection when tested using the CMM.[foot #] 7 Finally, although GSA suggests that PEC did not voluntarily offer to replace the center heads, the notice of rejection given to PEC was initially informal and came from the quality assurance specialist assigned to PEC's plant. PEC was concerned that the testing of the bisect angle had been performed improperly and that the rejection of the squares was thus improper. In addition, PEC apparently did not realize that GSA was considering terminating for default. Until PEC's concerns about the performance of the bisect angle test were resolved, it would not necessarily have occurred to PEC to volunteer to replace the center heads. Nothing in the record suggests that PEC was unwilling to resolve these deficiencies. In sum, although it is troubling that, as testing continued with the CMM, another discrepancy was discovered, leading GSA to terminate this contract, the record before us does not support respondent's action in terminating the contract summarily. The single rejection of a square for failure to comply with the bisect angle tolerance, without giving PEC an opportunity to cure this minor deficiency, does not justify a summary termination. The delays in testing samples for compliance, and the failures to permit PEC to witness testing and to give PEC prompt access to laboratory test reports, as well as the conflicting evidence ----------- FOOTNOTE BEGINS --------- [foot #] 6 We do not find it necessary to resolve the issues raised by the parties with respect to the propriety of the test results and sampling sizes applicable to the rejections based on length of blade and scale accuracy. PEC conceded some of the deficiencies identified and proceeded to correct them. The deficiencies did not directly give rise to the termination for default action. Although GSA argues that these rejections show that the contractor's quality control system was deficient and GSA needed prompt deliveries of conforming products, these arguments are offset to a large degree by the Government's own delays in testing the samples taken from PEC's shipments and in conducting tests on a piecemeal basis, which hindered PEC's ability to identify and correct deficiencies quickly and resume deliver on a timely basis. [foot #] 7 This would seem to be further exacerbated by PEC's inability to obtain prompt copies of test reports issued by the laboratory, which would have reflected the fact that only one characteristic had been tested, and by GSA's initial refusal to permit PEC to witness the testing of its squares by the laboratory . ----------- FOOTNOTE ENDS ----------- concerning the suitability of the coordinate measuring machine for testing combination squares, are extenuating factors. Because we grant the appeal in GSBCA 12503, overturning the default termination of PEC's contract, the Government is not entitled to recover excess reprocurement costs. Decision The appeals are GRANTED. The termination for default of contract number 50322 is converted to a termination for the convenience of the Government. _______________________________ CATHERINE B. HYATT Board Judge We concur: ___________________________ _______________________________ ROBERT W. PARKER MARY ELLEN COSTER WILLIAMS Board Judge Board Judge