Full opinion text
MEMORANDUM OBERDORFER, District Judge. Introduction There is before the Court the ongoing dispute arising from claims by Aero Corporation, now supported by opinions of the Acting Comptroller-General released June 5, 1981, and September 9, 1981, that the Department of the Navy has violated its duties under the Armed Services Procurement Act (“ASPA”), 10 U.S.C. §§ 2301-2737, and Section 3 of the Armed Services Procurement Regulations (“DAR”), see 32 C.F.R. parts 1-39 (vol. 1). The specific provision of ASPA that plaintiff claims defendant has violated is 10 U.S.C. § 2304(g), which creates a requirement that procurements, even if they cannot be competed through formal advertising, must, if possible, be awarded on the basis of competitive negotiation with all qualified potential contractors. That section provides in relevant part: In all negotiated procurements in excess of $10,000 ... in which time of delivery will permit, proposals, including price, shall be solicited from the maximum number of qualified sources consistent with the nature and requirements of the supplies or services to be procured.... 10 U.S.C. § 2304(g). Even when a noncompetitive, or “sole source,” negotiated procurement is permissible under section 2304(g), the regulations require the contracting officer to take steps to avoid unnecessary subsequent sole-source procurements, and to position himself to compete, if possible, any subsequent contracts related to the sole-source contract. As Armed Services Procurement Regulation ¶ 3-101(d) provides in pertinent part: Negotiated procurements shall be on a competitive basis to the maximum practical extent. When a proposed procurement appears to be necessarily noncompetitive, the contracting officer is responsible not only for assuring that competitive procurement is not feasible, but also for acting whenever possible to avoid the need for subsequent noncompetitive procurements. This action should include both examination of the reasons for the procurement being noncompetitive and steps to foster competitive conditions for subsequent procurements, particularly as to the availability of complete and accurate data ... and possible breakout of components for competitive procurement. ... 32 C.F.R. parts 1-39, vol. 1, at 327 (1979) [hereinafter cited without reference to the Code of Federal Regulations as “DAR ¶ 3-101(d)”]. Without conceding that defendant’s actions were to any extent lawful under section 2304(g), plaintiff also contends that defendant’s conduct violated the separate express obligations of DAR ¶3-101(d). The dispute and the litigation derive from decisions by the Chief of Naval Materiel (“CNM”) to award to Lockheed-Georgia Corporation (“LGC”) a series of contracts for the overhaul of 49 propeller-driven C-130 airplanes manufactured for the Navy by LGC in the 1950’s and thereafter. Defendant began to award the overhaul contracts to LGC in November 1979, and LGC has now been designated as the overhaul contractor for all 49 aircraft. The overhaul program is designated as a “Service Life Extension Program” (“SLEP”), and has as its purpose to increase the number of allowable safe flight-hours for aircraft that otherwise would need to be retired from use. The determinations made by the Navy concerning the C-130 SLEP undertaking were to procure the necessary overhaul work on all 49 aircraft through a negotiated procurement, rather than by formal advertising, and further to negotiate contracts for all 49 aircraft on a “sole-source” basis from LGC. The contracting officer has asserted that, under the circumstances of this case, section 2304(g) did not require a competitive negotiation. The decision to award SLEP contracts on a sole-source basis was based upon a determination by the commander of the Naval Air Systems Command (“NAVAIR”) that no contractor other than LGC could perform the overhaul on the C-130 aircraft designated for SLEP until LGC had furnished to that contractor the “kits” needed to accomplish the roughly 40 major individual overhaul tasks NAVAIR intended for each airplane. LGC had indicated to NAVAIR that it would require a very long time to furnish such kits to another aerospace firm. NAVAIR decided, in essence, that the Navy’s C-130 aircraft were in such bad condition that if overhaul were delayed for the time LGC represented to NAVAIR would elapse before the required kits could be furnished, many of the aircraft would become unsafe and would be grounded. That result, NAVAIR concluded, would jeopardize the critical fleet missions for which the Navy and the Marine Corps operate the aircraft. Plaintiff has contended that the CNM, NAVAIR and LGC (on whom the CNM and NAVAIR heavily depend in calculation of the timetable for possible kit manufacture) failed to consider in sufficient depth the fact that plaintiff and at least one other aerospace firm, Hayes International Corporation (“Hayes”), have for many years performed elaborate maintenance service on Navy, Air Force, and Marine Corps C-130 aircraft as so-called “Standard Depot Level Maintenance” (“SDLM”) and “Programmed Depot Maintenance” (“PDM”) contractors. That experience, plaintiff claims, was ignored by the Navy to such a degree as to make defendant’s decisions to procure SLEP for the C-130 aircraft on a sole-source basis irrational and a violation of section 2304(g) of ASPA and DAR ¶ 3-101(d). In its two opinions prepared for the parties and the Court, the General Accounting Office (“GAO”) concluded that competitive SLEP procurement was possible for some of the Navy’s C-130 fleet currently scheduled for induction in the program, and found the Navy’s rejection of competitive options “premature” in light of those possibilities. GAO determined, however, that the Navy properly could insist that any SLEP contractor other than LGC use kits in performing SLEP overhaul on the C-130. GAO also found that the opportunities for competition might be increased if competitive negotiation were limited to experienced C-130 maintenance contractors like Aero, Hayes, and LGC, because the kits needed for SLEP accomplishment by firms with the experience of Hayes and Aero would be less complex and could be furnished before the kits needed to assist industry-wide competition. GAO’s determinations were based upon documentary submissions and briefs filed by plaintiff, defendant, and LGC. After careful consideration of the elaborate GAO opinions, the voluminous documentary evidence considered by GAO, the memoranda filed by counsel, and the testimony adduced at several hearings (including testimony by the present commander of NAVAIR and the expert testimony of LGC employees), the Court is persuaded that plaintiff probably will prevail on the merits of many of its most important contentions. The Court will probably conclude that the Navy decisions to employ sole-source procurement for all 49 aircraft was unlawful. The Court will probably conclude that the Navy, beginning in 1979, so “transgressed the statutory boundaries” as to render the contracts and options since awarded to LGC for SLEP of many of the aircraft invalid. Additionally, the Court is likely to find that defendant has repeatedly breached its duties under DAR ¶ 3-101(d), as well as its obligations under this Court’s Order of March 4, 1980, to take steps to reconsider and minimize noncompetitive negotiated procurements. After assessing the effects of defendant’s continued preclusion of competition on plaintiff and on the public interest, as well as the possible impact of preliminary injunctive relief on the public interest and the allocation of power among the executive, legislative, and judicial branches of government, the Court has determined that affirmative interim relief is required in this case. Accordingly, an accompanying Order will direct defendant to begin engineering design for a kit-assisted competitive negotiation that would permit competitive award of SLEP contracts on as many aircraft as possible in the C-130 fleet scheduled by NAVAIR for SLEP accomplishment. The Order will leave undisturbed the schedule of SLEP inductions established by NAVAIR, and requires no action inconsistent with timely and uninterrupted SLEP accomplishment for the fleet. The Order does, however, require defendant to conduct the competitive procurement required by Congress unless and until the Secretary of the Navy or his delegate credibly and rationally determines that the costs of competitive negotiation or the risks attendant to such a procurement relieve the Navy from the mandate for competition under 10 U.S.C. § 2304(g) and DAR ¶3-101((1). FINDINGS OF FACT Background 1. When this action was filed in 1979, NAVAIR had proposed to contract on a sole-source basis for overhaul of a total of 49 C-130 aircraft. On cross-motions for summary judgment and plaintiff’s motion for a preliminary injunction, the Court ruled on March 4, 1980 that the Navy’s decision to award contracts for overhaul of 13 C-130 aircraft to LGC in the SLEP program and to give LGC an option for SLEP contracts for an additional seven aircraft was not unlawful, even though the contracts were negotiated without competition. Aero Corp. v. Department of the Navy, 493 F.Supp. 558, 556-67 (1980). With respect to the other 29 aircraft that the Navy planned to overhaul in the SLEP program, however, the Court emphasized that [t]he mandate of section 2304(g) of the [Armed Services Procurement] Act .... is clear[:] ... competition is required to the maximum extent feasible by solicitation of proposals “from the maximum number of qualified sources consistent with the nature and requirements of the supplies and services to be produced.” 493 F.Supp. at 568. Thus the Court found that while the decision not to compete the first 20 SLEP contracts was not unreasonable, the same could not be said for NAVAIR’s plan to contract for the remaining 29 without competition. 2. The Court further found in its March 1980 decision that the timing of the process used by the Navy in deciding to award the SLEP contracts to the original C-130 manufacturer had finessed plaintiff’s efforts to make a timely protest to the General Accounting Office, and had denied plaintiff an opportunity for timely judicial review of the sole-source decision as to the first 20 aircraft. Reviewing the Navy’s deliberate and unexplained failure to conduct design and procurement of SLEP for the C-130 in a manner permitting the effective review contemplated by statute and the law of this Circuit, the Court concluded: The Navy-Lockheed actions [in the last months before the CNM decision to procure on a sole-source basis] served no material planning or operational purpose of the Navy, and had as [their] principal apparent effect frustration of Aero’s opportunity to obtain considered review of the decision by GAO and this Court. The Navy [in the months after filing of the action] freely and strongly invoked national defense concerns in pressing the GAO and the Court to act in great haste, to review in days or weeks a decision which the Navy required over two years to make, creating the impression that the Court had a dramatic choice between doing its duty under the law and the Constitution or jeopardizing the Navy’s ability to defend the Nation. In these circumstances, the Navy’s failure to facilitate GAO consideration of Aero’s earlier protests was a breach of its duty, and seriously interfered with the ability of the GAO and of this Court to perform their duties. Neither of these failures by the Navy justifies interference with its plans for the first 20 planes. However, they do require the fashioning of a remedy that will guarantee that the Navy further considers all possibilities of competing some or all of SLEP for the remaining planes. 493 F.Supp. at 568-69. To this end, the March 1980 Order required the Navy [to] continue in good faith to consider the feasibility of competitive procurement for the remaining planes to undergo SLEP, including the use of kits tailored to depot level contractors experienced with the C-130.... 493 F.Supp. 570. In addition, the March 1980 Order required the Navy to monitor LGC’s work on the first overhaul contracts closely to determine whether kits were necessary, and declared that, as the procurement law required, all prospective contracts or options would be terminable upon determination by defendant that competitive procurement was practical and mandated by law. The Court explicitly based the requirement that the Navy monitor LGC’s SLEP experience, and explore use of kits tailored to SDLM and PDM contractors, on DAR ¶ 3-101(d): the Court reminded defendant that “[i]f the nature of the services in SLEP is such that providing limited kits is the only means of obtaining any competition at all, the Navy must do so, if it can be done practically.” 493 F.Supp. at 568. 3. Pursuant to the Navy’s master schedule, the first aircraft was inducted for SLEP accomplishment at LGC’s facilities on May 15, 1980, and other inductions soon followed. A team of Navy aerospace personnel (“the Review Team”) undertook on-site study of LGC’s progress on the first aircraft, and began to prepare a written report for use by the Navy decision-makers who would have to determine whether, and when, competition would be feasible, and for presentation to the Court in connection with the litigation. 4. In August 1980 plaintiff moved to supplement the March 1980 Order by requiring defendant inter alia to begin the procurement process for the last 29 aircraft scheduled for SLEP in time to permit competition. According to plaintiff, the Navy’s conduct in delaying solicitation and evaluation of proposals for the group of aircraft under section 2304(g) would have had the effect of making it impossible to compete contracts for the last group of aircraft in much the same way that the pressure of time made competition as to the earlier aircraft infeasible. Plaintiff also alleged that the Navy’s procurement of parts for installation in SLEP in certain configurations and combinations would make SLEP by any firm other than LGC unnecessarily difficult. In August 1980, however, the NAVAIR Review Team was still at work on its evaluation of LGC’s performance and the potential for kit-based competition. Moreover, GAO, prompted by an Aero protest filed in October 1979, had under active consideration the reasonableness of the Navy’s decision not to compete a group of the SLEP contracts. Accordingly, in light of GAO’s expertise and responsibility for reviewing contract procurement decisions, and reluctant to examine defense-related decisions of the Navy, the Court retained Aero’s prayer for an injunction requiring competition under advisement. 5. Thereafter, on December 16,1980, the Navy reported to the Court that it had decided to award five of the remaining 29 SLEP contracts to Lockheed. The Navy decision was based on a determination that any contractor other than LGC could not be allowed to attempt SLEP performance unless it was required to use so-called “mil-spec” kits. Mil-spec kits would contain, in addition to the parts needed for installation on the aircraft in the overhaul process, all the tools that any firm having the minimum technical qualification for maintenance work on Navy aircraft would possess. The language in the technical directives that would accompany such a mil-spec kit would employ the universal “military specification” terminology comprehensible to any such qualified firm, even if that firm had never worked on the C-130 aircraft before. LGC had previously informed NAVAIR that 36 months could elapse after a decision to compete SLEP contracts before those mil-spec kits would be available to a contract competitor. That time lapse, the Navy had found, would involve a militarily unacceptable delay in the SLEP program and thus make competition impracticable. 6. Plaintiff reacted with a motion seeking the sanction of contempt for alleged Navy noncompliance with the Court’s March 1980 Order. In doing so, it renewed its earlier prayer for an order requiring the Navy to conduct a competitive negotiated procurement involving, at the least, the three firms that were experienced C-130 maintenance contractors as to the remaining 29 aircraft, as well as any of the 20 aircraft not yet inducted into SLEP. Again, in deference to the Navy and to GAO, the Court retained Aero’s motion under advisement and requested GAO to respond to a series of questions framed in a February 26, 1981 stipulation entered into between plaintiff and the Navy. The Court thus requested GAO’s opinion whether there was a rational basis for the following decisions by defendant: (1) The Navy’s decision to award installation of SLEP on a sole-source basis; (2) The Navy’s determination that certain kits would be required for experienced C-130 maintenance contractors; and (3) The Navy’s determination that it would require approximately 36 months to prepare the required kits from the time of a decision to procure SLEP on a competitive basis. See Stipulation & Order of February 26, 1981. In addition, the Court identified for possible consideration by GAO the question whether [t]he Navy, in order to maximize competition for future SLEP procurement, has a duty ... to procure immediately the parts, data and tooling that the Navy deems necessary for a competitive procurement and to take the administrative steps that the Navy deems necessary for a competitive procurement. Id. at 2. In response to this and other Court orders, the Navy and Aero made elaborate submissions to GAO concerning the Navy’s December 16, 1980 decision. When, on May 11, 1981, the Navy determined to award all of the remaining SLEP contracts to LGC, plaintiff, defendant and LGC made further submissions to GAO. The GAO Opinions 7. On June 5, 1981, the Acting Comptroller-General submitted his opinion on the issues framed by the February 26, 1981 Stipulation and Order. In brief, the Acting Comptroller-General drew the following conclusions: a. There is a rational basis for the Navy’s determination that “certain” kits are required for performance of C-130 SLEP overhaul by experienced C-130 maintenance contractors other than LGC. b. Preparation of the kits necessary for competitive negotiated procurement could be accomplished in “considerably less time than the Navy has allocated,” and could be completed within 12-18 months’ time rather than three years. c. Because it was not possible, given those two facts and in light of all the other pertinent considerations, to conclude that competitive negotiation was impossible, the Navy erred in prematurely determining that competition under 10 U.S.C. 2304(g) was infeasible and in electing to make a final and complete sole-source procurement from LGC. d. The Navy had and has a duty to procure the parts, data and tooling the Navy deems necessary for competitive procurement under DAR ¶ 3-101(d). 8. In response, on June 15, 1981, Aero again moved for a preliminary injunction. While contesting the Acting Comptroller-General’s approval of the Navy’s finding that certain kits were required, Aero prayed again for an order enjoining the Navy from taking any action or failing to take any action that would delay competition for SLEP contracts among experienced C-130 SDLM contractors. When, however, the Navy sought GAO reconsideration of the Acting Comptroller General’s opinion of June 5, 1981, the Court requested GAO to entertain the Navy’s reconsideration petition, and retained the latest Aero motion under advisement with all the rest. 9. On August 13, 1981, after argument on the pending motion for preliminary injunction, and pending a GAO response to the Navy’s request for reconsideration, the Navy consented to an Order pursuant to which it would take the first administrative step toward competitive procurement: it undertook to contract by September 1, 1981 with LGC for a so-called “Engineering Change Proposal” (“ECP”). The ECP, according to defendant, represented a first step in the competitive process. If the Navy elected to compete the SLEP contract, the ECP would be used to prepare the technical directives or working-papers, also called the air frame change (“AFC”), needed for SLEP by a firm other than LGC. The ECP ordered by defendant pursuant to the Order of August 13, 1981, however, contemplated competition among all qualified aerospace firms rather than a competition limited at the outset to experienced C-130 SDLM contractors. Such an ECP would thus not fully reflect existing tooling and expertise possessed by firms like Aero and Hayes, who were not strangers to C-130 overhaul. The decision to order such a mil-spec-based ECP made it more difficult for defendant to explore fully the possibility of timely competition within the existing C-130 engineering community of LGC, Hayes and Aero. 10. On September 9,1981, GAO responded at length to the Navy’s request for reconsideration of its June 5 opinion. It rejected each element of the Navy’s continued opposition to any further attempt at competition. The Acting Comptroller-General found that the Navy had not justified placing additional noncompetitive C-130 SLEP contracts with LGC. On the question whether the Navy erred in projecting a three-year lead time for kit preparation, the Acting Comptroller-General emphasized that, in himself projecting an 18-month period, he had considered the evidence before him “in light of the court’s directions to the Navy to examine the extent to which ‘tailored kits’ might be used to foster competition.” C.R. 84 at 4. The reconsideration petitions by the Navy and LGC, however, apparently had made clear to the Acting Comptroller-General that [the Navy and LGC] have no intention of tailoring kits to reflect differences in difficulty and risk. ... Rather they now indicate they plan to describe every task in equal detail, without regard for the skill of experienced C-130 maintenance contractors, notwithstanding the impact this would have by increasing the time required to prepare the technical materials included in the kits. C.R. 84 at 4-5. According to the Acting Comptroller-General, however, [the] use of tailored kits reflecting only the minimum necessary technical detail is the lynchpin of our conclusion ... that it would be proper for the Navy to insist that contractors other than Lockheed use kits. C.R. 84 at 5. GAO indicated that when it endorsed the Navy decision to require “certain kits” it did not envision the Navy as writing detailed technical documentation without regard for its needs. Insofar as the record shows, preparation of kit documentation for use by experienced C-130 maintenance contractors should involve little more than documenting tasks which Lockheed has performed on the initial quantity of C-130 series aircraft to undergo SLEP. C.R. 84 at 5. Finding the Navy estimates of the time and difficulty involved in the preparation of kits intelligently tailored to the needs of experienced C-130 maintenance contractors therefore to be greatly exaggerated, the Acting Comptroller-General concluded: there is no impediment of which we [GAO] are aware preventing [the Navy] from conducting a negotiated procurement within the next few months, in connection with which it could properly weigh the relative risks and costs it would incur using proposed alternative approaches. C.R. 84 at 11-12. 11. On September 11, 1981, the Court held a conference to consider the effect of the Acting Comptroller-General’s opinions on the motions pending before the Court. At that hearing a representative of the Acting Comptroller-General advised the Court that executive branch agencies almost invariably comply with decisions of the Comptroller-General concerning the legitimacy of procurement decisions. Counsel for the Navy noted that the responsible Navy officials had not then had an opportunity to study the Acting Comptroller-General’s September 9 decision and decide whether to acquiesce in it. In view of the reported practice of agency acquiescence in GAO decisions and the Navy’s recent agreement to the Order respecting the ECP, the Court agreed to the request of Navy counsel for a continuance. The Court wished to give high-level Navy executive officers a reasonable time to review and give considered judgment to the immediate implications of the strong GAO opinions of June 5 and September 9, as well as their broader implications, particularly the practice of acquiescence in GAO determinations. The Sole-Source Rationale 12. The Court then held a supplemental hearing on the pending motions on September 25, 1981. Plaintiff, except for brief testimony from one of its consultants, submitted on the record. Defendant filed affidavits and adduced several hours of testimony from Vice-Admiral Ernest R. Seymour, commander of NAVAIR. Admiral Seymour testified that as commander of NAVAIR he reported to the CNM; that the CNM reported to the Chief of Naval Operations; the Chief of Naval Operations reported to the Secretary of Navy. NAVAIR’s commander, Admiral Seymour explained, is responsible for procurement of all Navy aircraft and for their support, and this responsibility involves the annual expenditure of about $5.5 billion for procurement and $2.5 billion for maintenance. NAVAIR’s “warrant” or responsibility, its commander testified, is to contract legally and “prudently.” C.R. 94 at 12-16, 26. No civilian official short of the Secretary of the Navy was identified as having a role in the NAVAIR procurement function, and Admiral Seymour identified neither the Secretary, nor any other civilian, as participating in the decision of NAVAIR and the CNM not to follow the Acting Comptroller-General’s opinions. C.R. 94 at 122-23. 13. At the September 1981 hearing, the NAVAIR commander also testified concerning the rationale for sole-source procurement of SLEP for the remaining fleet of C-130 aircraft. That testimony, the admiral’s affidavits, and the entire record in this case define the rationale for defendant’s repeated decisions to procure SLEP on a sole-source basis as follows: a. The procurement officers within NAVAIR have had, throughout the SLEP undertaking, three principal concerns regarding SLEP contract procurement and management: adherence to the schedule for induction of aircraft into the overhaul program and for their return to the fleet, maintenance by the SLEP contractor of the technical standards critical to the SLEP undertaking, and prudent use of the public’s money to accomplish SLEP. Management of those three objectives — -which, as Admiral Seymour stated, may involve trading achievement of one objective for another — often must place considerations of cost after those of scheduling and performance. C.R. 94 at 102-03. The critical mission of the C-130 has made adherence to the SLEP schedule NAVAIR’s paramount concern. C.R. 94 at 53. b. As to the first two concerns, adherence to schedule and assurance of competent performance, LGC has always presented the least risk, among all possible SLEP contractors, of the types of failures that would jeopardize the SLEP undertaking. C.R. 94 at 53-54. This low risk has resulted from LGC’s experience with the C-130 as NAVAIR’s prime manufacturing contractor for the aircraft, LGC’s more recent work as the SLEP contractor, and the generally close relationship of LGC with NAVAIR and NAVAIR’s dependence on LGC for technical guidance respecting the C-130. c. Since 1979, when SLEP procurement began, NAVAIR has believed that competition, even through a negotiated procurement limited to LGC and the few firms experienced in C-130 service and mechanical support, would jeopardize either NAVAIR’s adherence to its schedule, or the maintenance of a sufficiently high level of technical performance by the SLEP contractor. (1) The risk of technical failure in a competitive procurement could be made acceptably low only if NAVAIR procured from LGC, once solicitation and evaluation of proposals were nearly complete, kits for SLEP accomplishment that had been tested and validated as technically adequate by LGC. The production and release to a successful competitor of such kits would, however, have required postponement of SLEP accomplishment for aircraft critically in need of overhaul, and thus have violated NAVAIR’s first criterion: that the SLEP work stay on schedule. According to NAVAIR’s 1981 timetable, kits would not be available until October 1985, approximately eight months after the last airplane is scheduled to undergo SLEP. C.R. 92 at 2-3. (2) Adherence to the schedule through a competitive procurement from a firm other than LGC would be possible only if the SLEP contractor began overhaul before kits became available. SLEP accomplishment without the kits required would, however, have entailed unacceptable technical risks. C.R. 92 at 2. An alternative would have been for NAVAIR to have ordered the production of kits by LGC (estimated to take 17 months, excluding AFC development) and require on-line, task-by-task validation (at the same time the kits are in production), so that kits might be delivered to a competitor in time for SLEP accomplishment on some aircraft. That alternative, however, would be “imprudent,” for at least two reasons: (1) Because LGC deems on-line, task-by-task kit evaluation to be inferior to complete validation once a kit is fully prepared, such “piecemeal” validation should not be employed. C.R. 84 at 6. (2) LGC itself requires no kits to perform SLEP. If, following commencement of competition or as a result of competition, LGC were awarded the remaining SLEP contracts, the sums spent on kits would be wasted. C.R. 94 at 26, 62. 14. Within NAVAIR’s analysis, evaluation of the competitive option has been linked to an assumption that competition, if attempted at all, should not initially be limited to C-130 SDLM contractors. C.R. 94 at 70. According to the procurement officer, the AFC that NAVAIR would order, for example, would not be addressed only to experienced SDLM contractors, but would be adapted for use by any qualified military aircraft overhaul contractor. C.R. 94 at 75. Moreover, the time NAVAIR would allow for the “competitive cycle,” which is the period needed for solicitation and evaluation of contract proposals, contemplated solicitation and evaluation of proposals from firms other than LGC and the C-130 SDLM contractors. C.R. 94 at 74. Indeed, according to its commander, NAVAIR has never undertaken a “formal” or systematic effort to evaluate the practical and legal considerations in competition limited to SDLM contractors and LGC, as against a sole-source procurement, or a competition involving a larger group of aircraft overhaul firms. Id. As Admiral Seymour stated, it has been NAVAIR’s position that competition limited to LGC and SDLM contractors would be “competition for competition’s sake,” C.R. 94 at 73, and would be inadvisable when compared to wider competition within the aerospace industry. C.R. 94 at 73-79. The record thus shows that NAVAIR did not consider with any specificity the costs and potential benefits of such limited competition. 15. At the September 25 hearing, the Court explored with the procurement officer the basis for his preference for industry-wide competition over a more limited competition. NAVAIR evidently considered valuable the “certification” LGC could be required to provide with the fully detailed and tooled mil-spec kits that would be furnished for industry-wide competition. See C.R. 94 at 33-35, 75-77, Final Report §§ 8.1, 8.2.1, 8.2.3, 8.4. That certification, as Admiral Seymour indicated, would act as a remote warranty that the ultimate SLEP accomplishment by the contractor actually using LGC’s kits would be adequate. See C.R. 94 at 75-77. Such a remote warranty might not be available from LGC if less extensive kits were purchased, the Admiral implied. C.R. 94 at 77. The procurement officer also indicated, however, that NAVAIR had not considered whether requiring a direct warranty of fitness as to SLEP accomplishment on a given aircraft by the SLEP contractor that performed SLEP overhaul on the aircraft would be obtainable and, if it were, whether that warranty would be adequate or would be enhanced by an overlapping remote warranty. C.R. 94 at 75-78. If, of course, NAVAIR assumed that it would expect to purchase full mil-spec kits even if the SLEP contract went to an experienced SDLM contractor, there would be no advantage — in terms of kit requirements — in limiting competition to SDLM contractors and LGC. Apart from that preference for a remote warranty by LGC that SLEP would be properly accomplished using its kits, NAVAIR’s other basis for avoidance of limited competition stemmed from acceptance, as an axiom, of a principle that competition should be as broad-based as possible. C.R. 94 at 70, 73-74. 16. No study commissioned by NAVAIR has ever assessed the costs and time needed to prepare kits specifically designed for use by experienced C-130 SDLM contractors. No study commissioned by NAVAIR has ever assessed the costs and time needed to prepare the AFC required by experienced C-130 SDLM contractors. No study commissioned by NAVAIR, therefore, has ever determined whether SLEP accomplishment on any number of aircraft under the established NAVAIR timetable for SLEP of the C-130 would be possible through competitive negotiation limited to LGC and other experienced SDLM contractors. NAVAIR never weighed the costs of such a limited competition against the savings that might be achieved through such a competitive procurement. 17. NAVAIR received the mil-spec-based ECP it had ordered from LGC on November 30, 1981. While the ECP was being studied by members of Admiral Seymour’s APRB, the APRB also received a report from the Naval Air Rework Facility (“NARF”) at Cherry Hill, North Carolina, on prospects for competitive SLEP procurement. Both the ECP and the NARF advice were considered at an APRB meeting held January 14, 1982. Evaluation of the ECP confirmed, in NAVAIR’s view, the procurement officer’s testimony that the Navy’s critical schedule needs for SLEP could not be satisfied through a kit-based competition. The NARF recommendation took a different approach to the same result: officials at NARF, upon reading the ECP prepared by LGC, reported to the APRB that the “complexity” of SLEP was “beyond the scope of a retrofit kit competitive procurement approach.” AIR-01 Memorandum dated Jan. 18, 1982 (minutes of APRB meeting) (filed January 19, 1982). The APRB, after examining NARF’s position, rejected it, and concluded that competitive kit-based procurement was not technically infeasible, but (as the procurement officer had advised the Court in September 1981) that it was impossible under the current SLEP schedule. As the minutes of the APRB state: ... [T]he Board concluded that the risks associated with the ECP were not technical risks inherent in the ECP itself, but rather in the impact of the associated risks on the program schedule. Id. at 1. The APRB consequently determined to advise the NAVAIR commander not to pursue further competitive procurement for substantially the reasons offered by the NAVAIR commander in his testimony in these proceedings. On the same day that the APRB met, the procurement officer accepted the APRB’s advice that “the established SLEP program should continue to completion.” Id. at 3. The minutes of the APRB offer no indication that the competitive options discussed in the GAO opinions or in proceedings before the Court — including use of “tailored” kits — were considered at the APRB meeting. Factors Affecting the Procurement Decision 18. In NAVAIR’s own estimation, LGC is not the only source capable of satisfying the Navy’s technical requirements for SLEP accomplishment if parts and kits are available. NAVAIR’s judgment on the technical-risk factor, most recently embodied in the minutes of the January 14, 1982, APRB, appears to be based principally on the Final Report of the Review Team assigned to monitor SLEP and to evaluate Hayes and Aero. From the Review Team’s Final Report and other parts of the record, these facts have been established: a. NAVAIR, in evaluating technical capabilities of LGC and other possible SLEP contractors, requires “a high probability of technical success” from any firm awarded a SLEP contract. The Team commissioned to evaluate LGC’s early SLEP performance and the capabilities of Hayes and Aero focussed on four broad considerations: the engineering staffs of each of the three firms it examined, the three firms’ physical facilities for overhaul work, the availability of parts needed in SLEP accomplishment, and the availability of technical data. Final Report §§ 9.4 — 9.5. b. Taking account of all the relevant technical considerations, including parts and technical-data availability, the Review Team prepared at the conclusion of its study a “Technical Risk Assessment” for each of the firms it had evaluated. See Final Report §§ 5.1-5.6.5. LGC received the highest score of the three and obtained an “excellent” overall rating. Final Report § 9.5.1. Hayes received an “adequate” rating if furnished kits, and Aero a “questionably satisfactory” rating if furnished kits. Final Report §§ 9.5.2, 9.5.3. Hayes’ rating varied from that of LGC principally because, in the Team’s opinion, “unplanned parts” — parts discovered to be needed only after overhaul of a given aircraft had begun — were relatively more difficult for Hayes to obtain than for LGC to obtain. Aero’s less favorable rating than Hayes and LGC appears to have resulted from that firm’s lack of any substantial cadre of experienced and fully-trained engineers; both Hayes and LGC scored well above Aero in that respect. Nevertheless, the Team was not able — perhaps owing to its knowledge of Aero’s work as a SDLM contractor, see Finding of Fact 18(d) infra — to find that Aero was an unsatisfactory candidate for SLEP of the C-130. Moreover, the Final Report does not consider the effect of possible changes in the firms’ engineering staffs on their relative technical competence or the possibility of NAVAIR’s requiring such changes as condition for a SLEP contract award. c. In sum, it appears from the evidence that while LGC is the firm whose performance of SLEP presents the least risk of technical failure, it is not, according to NAVAIR’s criteria and assessment, the only firm technically qualified for the work if kits are available. Final Report § 11.1. The principal technical consideration in which LGC compares favorably to Hayes is parts availability; the principal technical criteria in which LGC exceeds Aero are parts availability and present engineering staff. d. This assessment by NAVAIR that LGC is superior but has not been proven to be the only firm capable of SLEP accomplishment is not a reversal of a prior position that only LGC could perform the work. It was evident before the Team had prepared its final report that other firms would probably be found to have the technical capability to accomplish SLEP, if provided parts and kits for their installation. Team members noted in 1980 that firms other than LGC at that time did “more replacement of SLEP type items” than did LGC itself. They also indicated, in 1980, that the Team “has seen no technical data so complicated during the SLEP on [the first two aircraft inducted for SLEP] that it would create any risks of a technical nature.” They commented that other firms had much of the tooling needed for SLEP already on hand. Thus at the time NAVAIR was developing a procurement program for SLEP of the C-130 aircraft scheduled for induction in the later years of the program, NAVAIR presumably was guided by an assumption confirmed by the Team’s final report: LGC might be technically superior, but it was not proven to be the only known source capable of SLEP accomplishment. The SLEP monitoring study by the Review Team, commenced in the summer of 1980, was NAV AIR’s most extended and systematic attempt to evaluate possible SLEP procurement sources. There is nothing in the record indicating that defendant earlier had assumed that technical considerations forbade competition; defendant’s position since the beginning has simply been that technical factors required kits. 19. As already noted, NAV AIR has never determined the cost and period of time required for preparation and delivery of kits specifically designed to permit SLEP accomplishment, in a technically satisfactory manner and with acceptable results, by experienced SDLM contractors. See Finding of Fact 16, supra. However, the following facts, relevant to the calculation of time and cost for kits, are established by the record: a. LGC, the probable manufacturer of the kits, has completed SLEP on a number of aircraft and has thus performed work on each of the three principal model classes on which later SLEP overhaul has been planned. b. With regard to technical directives and warrantable written commitments by SLEP contractors that work will be performed in particular ways and with certain results, Aero, as a Navy SDLM contractor, has experience in generating apparently complete and valid descriptions of the tasks required in the SDLM process. There has been no determination and apparently no attempt to determine whether Hayes, as an Air Force C-130 SDLM contractor, has equivalent experience in generating working-papers that also would be useful in stating overhaul warranty terms. 20. GAO undertook what it described as long and careful review of every timetable submitted to it by the Navy regarding the time needed to prepare kits needed for a limited competition among the SDLM contractors and LGC. It also considered Aero’s claim that Aero did not need kits to perform SLEP on the C-130. GAO made the following findings: a. As to the dispute on whether kits were required for SLEP accomplishment by Aero, GAO found that Aero had not demonstrated that the Navy lacked a rational basis for the kit requirement. See Finding of Fact 7, supra. b. As to the time required to prepare the appropriate kits, GAO found that close examination of the record before it indicated that the Navy had overstated various periods needed for kit design and fabrication. In the June 5, 1981 opinion, the Acting Comptroller-General summarized his conclusions in the following statement: We do not agree with the Navy that the record shows that it is impossible to complete kit preparation in less than three to four years. Our analysis of the work required to perform SLEP as disclosed by the Navy’s reports indicates that, although preparatory work remains to be done, the Navy as a result of its SLEP monitoring effort now has a fairly broad understanding of the technical problems which may be encountered, generally knows what equipment and locating tooling (tooling used to align parts during assembly) is needed, and has identified much of the technical data which may be necessary to compete its requirements. The record indicates that there is some information which the Navy does not yet have, particularly so-called “fit-up/clamp-up” criteria describing acceptable tolerances, torque to be applied to bolts and similar information concerning the assembly of the parts to be installed in SLEP. Lockheed, however, does have that data, and the Navy intends to turn to Lockheed to design, fabricate and test any kit which is used. Charts submitted by the Navy with its reports indicate that planned additional engineering can be completed within six months. A Lockheed proposal which that firm submitted to the Navy with regard to KC- and C-130 SLEP kits (which present more difficult work than does an EC-130 kit requirement) is consistent with that estimate and indicates that kit fabrication (assuming parts are available) requires a similar length of time. The cost of the Lockheed design effort in terms of time (manhours) and money (excluding parts) is nominal when compared with the total cost of SLEP. Nevertheless, the Navy explains that unless it validates kit design by having Lockheed perform trial installation, it rather than Lockheed must assume responsibility for the completeness and accuracy of the kit. The Navy allows approximately a year to perform concurrent validations of one KC- and one C-130 kit. While we view the Navy’s concern in this regard as appropriate, the record does not indicate why the work which Lockheed would perform to prepare kits cannot be overlapped. Assuming parts availability at Lockheed, assembly of the test kits might overlap in part work related to kit design, which largely involves documenting procedures Lockheed is now using. For the same reason, validation through Lockheed test installation might be accomplished in conjunction with SLEP on some of the aircraft currently scheduled for SLEP at Lockheed (which include a KC- and C-130 aircraft scheduled to undergo SLEP early next year). Instead, by assuming a discontinuous Lockheed kit development schedule punctuated by internal Navy review, the Navy has developed an overall kit schedule which produces a finished kit design only after work presently under order at Lockheed will have been completed. As indicated, the suggested kit preparation schedules include substantial lengths of time for Navy administrative, GAO and judicial review. Some of these time frames are based on average time spent by involved activities on past reviews and do not reflect the time which would be required if available time is critically short. Others can be disregarded. The eight month period allowed for GAO and judicial review in two of the proposed schedules is not relevant to a determination as to whether competition is possible if the Navy acted now to permit it, since if the Navy did so there would be no need for such review. In some of the schedules, time appears to have been allocated twice to correct kit deficiencies by scheduling time after validation is completed even though Lockheed’s schedule (on which all of the estimates are based) assumes changes would be made at the time they are discovered. Nor is it apparent why some of the proposed schedules allocate more time to review the Lockheed kit design than Lockheed needs to prepare it in the first place. The so-called “optimistic” Navy schedule, correcting some of these difficulties, allows considerably less time to Navy review than do the other possible schedules. It, however, remains keyed to parts availability and allows approximately nine months for administrative review. We believe the optimistic schedule still overstates the minimum time required to design, assemble, and validate an initial kit package, provided the Navy permits Lockheed to perform aspects of the required work concurrently, i.e., to assemble and validate the kits as documentation is prepared and provided the Navy reviews that work while it is in progress. Bearing in mind that Lockheed principally would be documenting procedures which are now being used to perform SLEP, with which the Navy is familiar as a result of the Navy SLEP monitoring study, we believe kit preparation can be accomplished in considerably less time than the Navy has allocated and that it certainly should be possible to complete it within no more than one to one and one half years time. C.R. 63 at 6-8. c. At the heart of GAO’s difference with the Navy on the time needed to prepare kits was GAO’s assumption that the kits would be tailored to the requirements of experienced SDLM contractors performing SLEP. The Acting Comptroller-General stated the difference between his approach and that of the procurement officer in the following terms: In reaching our decision with respect to whether the kits the Navy says would be required can be prepared by the time they are needed, we recognized that the time needed to prepare such kits would depend upon a number of factors: (1) how detailed the technical instructions included in the kits would be; (2) how many and what types of parts would be furnished in the kits; and (3) whether and if so how scheduling constraints would be interrelated. We also considered the Navy’s legal obligations with respect to kit design in light of the court’s directions to the Navy to examine the extent to which “tailored kits” might be used to foster competition. The Navy’s monitoring study reports state that any kits which are used would be tailored to the skill and experience level of the contractor believed qualified to install them. Taking the Navy at its word that tailored kits would be used, we stated in our June 5 letter that (as we understood what the Navy intended to do) the Navy would prepare kits varying in detail from task-to-task, depending on the Navy’s assessment of the degree of difficulty and risk involved in performing each planned SLEP task. The Navy and Lockheed now dispute this conclusion, contending in effect that they have no intention of tailoring kits to reflect the differences in difficulty and risk associated with particular tasks. Rather, they now indicate they plan to describe every task in equal detail, without regard for the skill of experienced C-130 maintenance contractors, notwithstanding the impact this would have by increasing the time required to prepare the technical materials included in the kits. C.R. 84 at 4-5. The Acting Comptroller-General explained GAO’s assumption that such tailored kits were more appropriate than fully detailed and tooled mil-spec kits in the following way: Commonly, competition is enhanced if an agency can describe the work to be done under a contract in great detail because a larger class of offerors may be able to compete, or those which do may be able to do so more effectively. The more detail an agency seeks to provide, however, the greater the time which may be necessary to prepare for the procurement. Where, as here, time is of the essence and there are two potential classes of offerors, one which the agency believes can perform satisfactorily if given less technical data than the other, the level of detail insisted upon by the agency, if unrelated to its actual needs, can become an unreasonable impediment to competition. We did not envision the Navy as writing detailed technical documentation without regard for its needs. Insofar as the record shows, preparation of kit documentation for use by experienced C-130 maintenance contractors should involve little more than a process of documenting tasks which Lockheed has performed on the initial quantity of C-130 series aircraft to undergo SLEP. Although there are a large number of separately identified SLEP tasks, many of the tasks are relatively simple and straightforward. Only a handful are identified in the SLEP monitoring study as complex enough to require special skill, tooling or care in any particular aspect of the required work. While the few tasks which can be considered to be generally complex (i.e., which potentially involve a number of complex or interrelated functions) involve a significant amount of effort and cost for parts, the sequence of sub-tasks, inspection criteria, tooling, and fit-up/clamp-up criteria which Lockheed has used is known, at least to Lockheed. In fact, were this not so the Navy would have no assurance that Lockheed was itself performing the work in a uniform, standardized manner. C.R. 84 at 5. 21. In the course of SLEP accomplishment on a particular aircraft, the contractor may discover need for “unplanned parts” in configurations that need replacement in the overhaul but that are not on the general SLEP schedule for replacement; only upon disassembly of the aircraft, for example, can the contractor see the deterioration or fatigue in some configurations. See Final Report §§ 7.1.5-7.2.7. Since the same parts can be made available to any firm performing SLEP, the length of time required for their procurement does not work for or against SLEP accomplishment by a particular firm. LGC can manufacture some of these parts, and, like other firms, can also purchase the parts directly from suppliers or through the government. For unscheduled parts most easily fabricated or purchased by LGC, of course, LGC is the most convenient SLEP contractor. There is, however, no evidence tending to show that LGC could not supply the unscheduled parts it manufactures to another SLEP contractor. There is no legal impediment to the Navy’s requiring LGC to furnish those parts, at a fair contract price, for SLEP accomplishment by another firm. C.R. 84 at 9. The Navy does not dispute this, but relies upon the evident administrative convenience of LGC’s furnishing unplanned parts to itself, rather than to another firm, as a partial justification for the sole-source procurement. However, the Acting Comptroller-General described the Navy’s resistance to competition premised on parts availability as follows: While the Navy might be able to identify some nonscheduled parts as particularly likely to be needed, and thus, include them in kits, use of such parts is defined as outside the planned scope of SLEP, and the Navy argues, most such parts must be provided from contractor inventory or through the Navy C-130 parts supply system. Lockheed can meet many unexpected parts problems quickly from its own parts inventory, by borrowing parts designated for new production aircraft, or by manufacturing or remanufacturing parts as needed. The Navy, moreover, maintains an extensive C-130 parts supply system, and indeed, furnished approximately 30 percent of the nonscheduled parts used by Lockheed for the first aircraft which underwent SLEP. Maintenance contractors such as Aero also maintain stocks of some C-130 parts. Notwithstanding the Navy’s view that only Lockheed can avoid schedule delays altogether, the Navy has not treated the non-schedule parts problem as insurmountable in its reports, but rather, has assumed throughout this litigation that competition is possible with kits which as indicated would not include most nonscheduled parts. At worst, the Navy’s reports indicate that performance of SLEP by a firm other than Lockheed might require that the Navy or its contractor maintain increased stocks of incidental parts. How much time would be required to establish a sufficient incidental parts inventory is not clear, because the Navy has not addressed this question specifically. However, since long lead time parts are apparently not involved, availability of such parts does not appear to be a factor preventing competition with respect to most if not all of the remaining aircraft. C.R. 63 at 9. And as the Acting Comptroller-General observed in his second opinion on the SLEP procurement, the Navy has a number of means of insuring that LGC, or any other parts fabricator, will deliver parts needed in connection with another ongoing defense procurement. See C.R. 84 at 7-9. There is currently no evidence in the record that the circumstances of this procurement are exceptional in that respect. Summary of Findings These, then, are the salient facts about defendant’s SLEP procurement: 1. Without determining that LGC was the only source capable of SLEP accomplishment, defendant has decided to award all SLEP contracts to LGC on a sole-source basis. 2. Notwithstanding the Court’s Order of March 1980 and the advice of GAO in June and September 1981, defendant has taken no steps to develop a kit-assisted competitive negotiation limited to experienced C-130 SDLM contractors and LGC. 3. In reliance upon the possibility of remote warranty by LGC as the kit manufacturer, defendant ordered a mil-spec-based ECP for kit development on or about September 1, 1981. It has never ordered any ECP contemplating kits specifically tailored for use by experienced C-130 SDLM contractors. 4. The Navy failed to estimate the time required for a kit-assisted competitive negotiated procurement from LGC or an experienced C-130 SDLM contractor, or to weigh the cost of such a procurement, or to undertake a comparison of the predictable cost of such a procurement with the costs of the present sole-source contract. 5. According to the Final Report of NAVAIR’s SLEP Review Team, kit-assisted procurement of SLEP by a firm other than LGC would be possible, if the necessary parts and other kit components could be furnished the SLEP contractor. 6. In the opinion of the Acting Comptroller-General, prudent kit procurement in connection with competitive negotiation limited to experienced C-130 SDLM contractors would not require so long a time as to make kit-assisted SLEP accomplishment impossible under the current NAYAIR SLEP induction schedule. CONCLUSIONS OF LAW Jurisdiction 1. As in earlier stages of the litigation, see 493 F.Supp. 558 (1980), this Court has limited subject-matter jurisdiction to review defendant’s action under the Administrative Procedure Act. Kentron Hawaii v. Warner, 156 U.S.App.D.C. 274, 277, 480 F.2d 1166, 1169 (1973); M. Steinthal & Co. v. Seamans, 147 U.S.App.D.C. 221, 233, 455 F.2d 1289, 1301 (1971); Scanwell Laboratories v. Shaffer, 137 U.S.App.D.C. 371, 386, 424 F.2d 859, 874 (1970). See also Wheelabrator Corp. v. Chafee, 147 U.S.App.D.C. 238, 244, 455 F.2d 1306,1312 (1971). Jurisdiction rests inter alia on 28 U.S.C. §§ 1331, 2201. The limits on the district court’s competence were clearly stated by the Steinthal court. As Steinthal held, a procurement officer’s discretion extends not only to the evaluation of bids submitted in response to a solicitation but also to determination by the agency with respect to the application of technical, and often esoteric, regulations to the complicated circumstances of individual procurements.... If the court finds a reasonable basis for the agency’s action, the court should stay its hand even though it might, as an original proposition, have reached a different conclusion. ... M. Steinthal & Co. v. Seamans, supra, 147 U.S.App.D.C. at 233, 455 F.2d at 1301 (footnote omitted). Plaintiff bears at this stage of the lawsuit the “heavy burden” of demonstrating that defendant’s exercise of discretion had no rational basis, or that defendant’s conduct “involved a clear and prejudicial violation of applicable statutes or regulations.” Kentron Hawaii v. Warner, supra, 156 U.S.App.D.C. at 277, 480 F.2d at 1169. Our Court of Appeals has recognized that the sensitivity required in judicial review of any aspect of government procurement, see Perkins v. Lukens Steel Co., 310 U.S. 113, 130, 60 S.Ct. 869, 878, 84 L.Ed. 1108 (1940), must be heightened when a procurement officer has acted under authority of 10 U.S.C. § 2304. Wheelabrator Corp. v. Chafee, supra; see also Cessna Aircraft v. Brown, 452 F.Supp. 1245, 1253 (D.D.C.1978). 2. The central issue in this case is whether defendant in 1979 and 1980 complied with the requirement of 10 U.S.C. § 2304(g) that proposals be solicited from the maximum number of qualified possible contractors in a negotiated procurement. Plaintiff does not contest defendant’s factually-related determination, under 10 U.S.C. § 2304(a)(10), that the circumstances precluded procurement by formal advertising. The procurement officer’s decision not to compete the contracts in 1979 and 1980 (as distinguished from his decision not to employ formal advertising) is plainly reviewable. Cf. Wheelabrator Corp. v. Chafee, supra, 147 U.S.App.D.C. at 243 n.3, 455 F.2d at 1311 n.3. Actions complaining of erroneous decisions under section 2304(g) are not barred by 10 U.S.C. § 2310(a). Compare Serv-Air, Inc. v. Seamans, 154 U.S.App.D.C. 28, 473 F.2d 158 (1972) with Wheelabrator Corp. v. Chafee, supra. Since the Navy decisions in this case under sections 2304(a)(10) and 2304(g) are factually related, however, it is noteworthy that section 2310(a), which limits review of procurement decisions by either GAO or a court, makes unreviewable only decisions “required to be made ... by the head of an agency.” By contrast, the determination made here under section 2304(a)(1