"THIS OPINION WAS INITIALLY ISSUED UNDER PROTECTIVE ORDER AND IS BEING RELEASED TO THE PUBLIC IN ITS ENTIRETY ON JUNE 26, 1995" DENIED: March 13, 1995 GSBCA 13103-P TITAN CORPORATION, Protester, and EER SYSTEMS CORPORATION, Protester/Intervenor, and MICROLOG CORPORATION, Intervenor, v. DEPARTMENT OF COMMERCE, Respondent. John R. Tolle and William T. Welch of Barton, Mountain & Tolle, McLean, VA, counsel for Protester EER Systems Corporation. Richard A. Thompson, President of Microlog Corporation, Germantown, MD, appearing for Intervenor. Jerry A. Walz, Mark Langstein, and Lisa J. Obayashi, Office of General Counsel, Department of Commerce, Washington, DC, counsel for Respondent. Before Board Judges DANIELS (Chairman), DEVINE, and HYATT. DEVINE, Board Judge. This protest was originally filed by Titan Corporation. EER Systems Corporation, referred to as protester herein, intervened as of right, as did Microlog Corporation. Thereafter Titan Corporation was dismissed as a party by agreement. EER Systems continued as a protester. Microlog Corporation took no active part. Findings of Fact On June 9, l993, the National Oceanic and Atmospheric Administration (NOAA), a part of the United States Department of Commerce (DOC), issued a solicitation that sought to acquire a new weather radio system for its National Weather Service. Protest File, Exhibit 3. The system sought is known as the Weather Radio Console Replacement System or CRS. Id. The principal mission of NOAA's National Weather Service is to forecast weather patterns nationally and to disseminate its forecasts to interested media and the general public, particularly those forecasts dealing with expected weather extremes, with a view to saving lives and preserving property. Transcript at 233-36. Weather forecasts are transmitted by NOAA Weather Radio, a broadcast service which operates a network of transmitters throughout the United States and parts of the Carribean Sea and the western Pacific Ocean. Protest File, Exhibit 3. This is the system that warns of tornadoes, hurricanes, blizzards and flash floods, and such non-weather perils as earthquakes and volcanic activity. Suspension Hearing Transcript (hereafter: S. Transcript) at 8. The existing radio system consists of 386 transmitters driven by over 300 broadcast consoles, located in more than 200 weather service offices nationwide, reaching about 75% of the population. S. Transcript at 14. It is based on outmoded 8-track technology which requires manual manipulation by system operators, thus slowing down response times, which is especially undesirable in extreme weather emergencies. Id. at 12-13. The new system to be acquired under the procurement at issue is highly automated. It will convert digital data coming from various forecast sources into synthesized speech, which will be automatically stored, sequenced, scheduled and transmitted, thus greatly speeding critical warnings. S. Transcript at 15, 41. In addition, the broadcasts will be available to 95% of the population. Id. at 14. The Solicitation Requirements. This was a two phase procurement. Phase I contemplated that offerors would propose hardware/software systems looking to the development of system prototypes. Offerors were told that either one or two Phase I contracts would be awarded on a cost-plus- fixed- fee basis. They were also told that a single award might be made where one of the proposals was considered substantially superior to the others, or where the award of two contracts would exceed Government budget estimates. Protest File, Exhibit 3. In the event, only one Phase I contract was awarded because of a money shortage. Phase II, available to NOAA as an option under the Phase I contract, was to be a single firm-fixed-price contract to produce and deploy the new system. Id. Both Phase I and Phase II were to be evaluated together. The cost figures submitted for both phases by all offerors were necessarily educated estimates, especially those proposed for Phase II. Until a successful Phase I prototype was developed and accepted, no offeror could know exactly what Phase II would involve. The detailed requirements of the system that NOAA sought in this procurement were necessarily of the performance variety. NOAA set out the end results that it expected the system to deliver and asked the offerors to propose a mix of hardware and software that could deliver the results required. The proposals were to be in enough detail that NOAA's computer people and evaluators could make informed judgments as to the merits and demerits of each system proposed, and intelligently assess both the performance risks and the cost risks presented by each proposal. Protest File, Exhibit 3. To achieve reliability at reasonable cost, the solicitation required non-developmental or commercial off-the-shelf (COTS) hardware and software wherever practicable and stressed functional modularity of both hardware and software. The idea of this last requirement was to acquire modular units that could be replaced when a failure occurred without requiring undue changes in other system components. Protest File, Exhibit 3. In accordance with the solicitation, those offerors in the competitive range were advised during discussions of deficiencies in their offers and were twice given an opportunity to make corrections and submit revisions. Protest File, Exhibit 3 at M.5.2. Following discussions, offerors were to submit their best and final offers (BAFOs). Id. Proposal Evaluation. Three major factors were considered in the evaluation of offers: Technical; Technical Management; and Cost. The first two were to receive a point score rating and a narrative description. The third was to be evaluated for magnitude and realism. Protest File, Exhibit 3 at M.6. The Technical factor contained five subfactors, listed in descending order of importance: System Design and Effectiveness; Technical Approach; System Reliability, Maintainability, and Availability; Test Methodology; and Implementation and Integration. Id. The Technical Management factor contained three subfactors, also listed in descending order of importance: Program Management Plan; Corporate Experience and Capabilities; and Personnel. Id. The Technical factor (60) and the Technical Management factor (20), taken together, when compared to the Cost factor (20), represented a technical-to-cost ratio of 80/20. Put another way, the technical factors were four times as important as the cost factor. Protest File, Exhibit 3 at M.6.2. The offerors' proposals were evaluated by a Technical Advisory Committee (TAC) of ten members who evaluated selected elements for all proposals. Evaluators assigned an adjective rating and a numerical score to the various elements, supported by a narrative description of strengths and weaknesses. Protest File, Exhibit 12. The range was from 0 to 1.0 for each selected evaluation element. Id. Once all the individual evaluators had completed their work, a series of meetings of the TAC members was held to arrive at a consensus for each evaluation element. These consensus scores were then multiplied by the appropriate weighting factor to arrive at final scores. Id. The maximum score for the Technical factor was 1200. The TAC's final evaluation report gave CommPower a technical score of 980 points (out of 1200), while protester received 805. Protest File, Exhibit 39 at 2-3. CommPower was better than protester in four of the five subcategories, lagging behind only in Technical Approach. Id. CommPower's score for the category Techical Management was 286 (out of 400) while protester's was 307. The Business Evaluation Committee evaluated all of the offerors' cost proposals for magnitude and realism. Its final report showed a total evaluated price for CommPower of $12.6 million and for protester, $9.3 million. Protest File, Exhibit 41. The Protest Grounds. Protester asserted a number of protest grounds in its various pleadings. Some were not further pursued. Most were subsumed into the three grounds dealt with in protester's post- hearing brief. The first of these was the issue of whether or not NOAA's delegation of procurement authority (DPA) was violated in the course of this procurement. The second was the issue of whether or not NOAA should have told the offerors that only one Phase I contract would be awarded, and the third dealt with the correctness of the scoring of both CommPower's and protester's technical proposals, especially with respect to the use of subcontractors, and also with respect to protester's ten concurrent voice channel capability. The DPA History. At the commencement of the procurement in dispute, the DOC estimated that the system needed would cost $13.1 million. It held a specific DPA from the General Services Administration (GSA) to acquire automated data processing equipment (ADPE) in the amount of $17.5 million, of which $13.1 million was redelegated to NOAA. Protester's Exhibit 9. On November 23, 1994, GSA increased DOC's specific DPA to $20 million, of which $19.552 million was redelegated to NOAA. Protester's Exhibit 5. The contract at issue was awarded on December 7, l994, and totalled $19,551,649. The Proposal Histories. Six proposals were submitted in the fall of 1993. The first round of evaluations took place in February and March of 1994. Protest File, Exhibits 7-12. Five were found to be in the competitive range, that is, they were rated acceptable or capable of being made so after discussions, which began in March 1994. Id., Exhibit 13. The initial offers ranged from a low of $13.5 million to a high of $31.78 million. Id., Exhibit 12. First revised proposals were submitted in April 1994. Id., Exhibits 15-19. These were evaluated in May 1994, with both oral discussions during face-to-face meetings and written questions and answers during June 1994. Id., Exhibits 26-30. NOAA directed 119 questions to protester with respect to this submission and 51 to CommPower. Id., Exhibits 14-16. Following this evaluation by the TAC, protester's proposal was rated good with one technical deficiency and six weaknesses. The deficiency questioned protester's ability to meet the emergency override interruption described in the solicitation. This was subsequently corrected. The six weaknesses were: (1) the validity of EER's calculations with respect to reliability and availability, (2) whether the proposed speech synthesis system could handle the required ten voice channels to service ten transmitters, (3) EER's lack of a detailed explanation as to how the proposed stand-alone hard disk drive could be shared between two processors, one running Windows NT and the other SCO UNIX, (4) the ability of the offered VME Bus chassis to accommodate sufficient boards, (5) the use of the Developer's Workbench (DWB) software design, and the lack of experience with this software of those personnel proposed to develop it, and (6) the lack of detail in testing and reporting schemes. A technical deficiency in the Technical Management area dealt with the TAC's concern that the proposed hardware and software personnel (employees of PRC, protester's software developer) had limited committment and limited involvement in the project and might be insufficient in number to complete the project. The TAC concluded that although protester's proposal remained technically unacceptable, it could be made acceptable. Id., Exhibit 26. With respect to CommPower's proposal, the TAC rated it excellent technically, noting many strengths, but found three technical weaknesses: (1) CommPower's Audio Switching Assembly had no redundancy and was a point of failure, (2) CommPower's design was hardware intensive and might be overdone, and (3) its procedures for resolving discrepancies and reporting software problems were not discussed. In addition, three weaknesses remained in the Technical Management area: (1) CommPower's program management system lacked automation and contained little information for tracking program progress against expenditures, (2) the structure and format of the monthly status report was insufficiently described, and (3) CommPower showed limited experience in synthesized voice systems. However, the TAC found the proposal technically acceptable. Protest File, Exhibit 26. This evaluation generated eighty-six questions to be answered by protester and fifteen to be answered by CommPower. Id. at 27. A dozen of the questions directed to EER reflected the TAC's concern about what EER's software tool, DWB, consisted of; how the CRS program would benefit from the DWB experience of PRC, the software subcontractor; whether DWB was developmental or a COTS item; the amount of DWB code that would be used in the CRS project; and the accuracy of EER's software development effort. A number of questions dealt with how the two different operating systems (UNIX and Windows NT) proposed by EER could share the same hard disk drive disk. A number of questions dealt with EER's proposed DECTalk solution to the problem of ten channel through-put capacity: how it could be synchronized, controlled, and timed. Protest File, Exhibit 27. Finally, the TAC questioned the availability and continuity of EER's software development personnel who were PRC employees. Id. Protester satisfied NOAA on a number of these issues, but the TAC remained concerned with DWB, protester's software tool, because the personnel of protester's software subcontractor, PRC, had no experience with it in an operational system and protester had no in-house expertise at all. Although the TAC liked the idea of a software development tool, it was concerned that DWB was still in a development stage and thus did not satisfy the requirement that both hardware and software be commercial, off- the-shelf products to the extent possible. There was also concern about the lack of documentation and inadequate quality assurance with the version proposed (5.0) although EER promised corrections. Protest File, Exhibit 34. The TAC also remained concerned that the two operating systems proposed by protester could not be successfully set up to share the same hard disk drive. EER's answers to questions in this area elicited only a list of five possible ways, set out in its revised proposal, in which a local area network (LAN) manager package, or other software, such as a network file system (NFS), might be utilized to coordinate the functions of the two processors. However, it gave little detail how any of the five solutions might work and left the final decision on which of the possible solutions to employ to be made during the design work following award. Protest File, Exhibit 34. The TAC's final serious concern was the ability of the DECTalk software operating in the Windows NT environment, which protester has proposed, to handle a stream of ten text-to-speech conversions simultaneously. Protester maintained that it could be done but did not discuss how control and synchronization were to be accomplished. Protest File, Exhibit 34. In its BAFO, protester opted for number 4 of its original five possibilities to solve the shared disk problem stated earlier. This called for a NFS using TCP/IP as the transport layer protocol for interprocessor communication. Little further detail was provided. With respect to the DWB issue little further information was added. With respect to the ten channel communication issue, the BAFO summarized information given earlier in respone to NOAA's questions but added no further detail. Protest File, Exhibit 38. Following the submission of BAFOs, the TAC's overall score for CommPower's proposal was 1266 points (out of a possible l600) with protester in second place with 1112. With respect to technical scores, CommPower had 980 while protester had 805. Protest File, Exhibit 39. The TAC report to the Source Evaluation Board (SEB) found that CommPower offered the "best system design and technical approach with minimal risk," especially when considered in the light of the procurement's "aggressive development schedule." Protest File, Exhibit 39 at 2-3. CommPower had the "best scores in four of the five Technical rating subfactors." Id. With respect to protester, it found "risk associated with the DecTalk software algorithm supporting 10 voice channels, the disk sharing between Windows NT and SCO UNIX [two operating systems], and the use of the PRC Developer's Workbench." Id. It recommended award to CommPower. The SEB met to review the Business Evaluation Report, the TAC report, and the BAFOs. Its conclusions were similar to the TAC's. It made the following recommendation to Dr. Friday, the Source Selection Official (SSO): The CommPower proposal offers the Government the best system design and technical approach with the least technical, schedule, and cost risk. It has the highest combined technical and technical management score, and is believed to be the only proposal that can meet the Government's aggressive development schedule. Based on the above, the Source Evaluation Board recommends that CommPower Incorporated be awarded the contract for the Console Replacement Solicitation. Protest File, Exhibit 41 at 27. Dr. Friday testified at hearing that his decision to make the award to CommPower, the more expensive of the two proposals in issue here, was his judgment that, being the better proposal and the one most liable to be most quickly operational, it might well assist NOAA to save lives that might otherwise be lost. Transcript at 245, 238-39. Discussion Although protester attacks the procurement in question on several grounds in its pleadings, it argues only three grounds in its posthearing brief: (1) that NOAA violated its DPA in conducting this procurement, (2) that NOAA should have advised offerors that, despite the statements in the RFP that two contract awards were possible, budgetary limitations would restrict NOAA to a single Phase I contract award, and (3) that the Technical and Technical Management evaluations were improperly done, allegedly resulting in too high a score for CommPower and too low a score for protester, thereby affecting the Cost/Technical Trade-off decison. We consider these grounds in order. In CACI, Inc. v. Stone, 990 F.2d 1233 (Fed. Cir. 1993), the Federal Circuit held that the award of a contract for ADPE where the agency confessedly had not first obtained a DPA was void. In Science Applications International Corporation v. National Aeronautics and Space Administration, GSBCA 12600-P, 94-1 BCA 26,553, 1993 BPD 328, this Board held a contract award in an ADPE procurement void where the value of the contract "substantially exceeded" the DPA's monetary limitation. Both of these cases display the "plain illegality" that the Federal Circuit talks about in CACI. The case before us, however is different. The DOC began the procurement at issue with a specific DPA for $17.5 million. Protester's Exhibit 9. On November 18, 1994, this was raised to $20 million. Protester's Exhibit 8. The contract was awarded on December 7, 1994, at a price of $19,551,649. Protest File, Exhibit 11. Based on the foregoing figures, protester asserts that NOAA exceeded its DPA authority. It reaches this result by averaging all of the offerors' initial proposals of September 1993, which protester says results in an average proposal figure of $21,186,452.33. It then averages the BAFO proposals of August 1994, which protester says results in an average BAFO of $17,717,485.60. Next, protester compares these averages (with Phase I amounts doubled to cover the possibility of a second award) with the DPA then in effect, finds the DPA exceeded, and the procurement void from the beginning. CACI tells us that a procurement without any DPA at all is illegal and therefore void. In the procurement before us, NOAA began it with a specific DPA that matched its estimate of what the procurement would cost ($13.1 million) and sought and obtained a higher DPA ($19.552 million) when it became evident that the system it was authorized to buy might cost more ($19.55 million) than its original estimate. Although the award price was $19,551,649, CommPower's evaluated price was $12.6 million and protester's was $9.5 million. By June 30, 1994, NOAA had given up the idea of awarding two Phase I contracts because it didn't have the money. It could do no more. The offers it was considering which were most likely to be accepted never exceeded in amount (especially in their evaluated amounts) the DPAs under which NOAA was proceeding. In addition, the figures on which NOAA relied in making its procurement decisions were necessarily all estimates, the best available at that time. They are principally based on a mix of labor rates multiplied by the estimated number of manhours in the various labor categories required to complete the project. Some figures included in the total will undoubtedly be higher than estimated and some lower. Such estimates are by their nature approximations. A relatively small percentage of the total is attributable to equipment or software that can be readily priced, such as those found in SAIC. Some of the figures will undoubtedly turn out to be higher than estimated and some lower. They are all approximations. This procurement is scheduled to continue over a period of some five years. If protester's views on this point were carried to a logical conclusion, any cost increase during performance over and above the DPA figure would retroactively void the contract, a result which would destroy all certainty of contract where DPAs are concerned and produce unforeseeable results. A similar situation would result if the man hour estimates are incorrect or the labor mixes are in error because the nature of the tasks finally performed were not anticipated. CACI requires that where a "plain illegality" exists, the contract must be held void. Conversely, however, where the illegality is not plain (or is non-existent as it is here) the contract should not be held void. This is so because holding a contract void makes it almost impossible to fashion a remedy for the situation in which the parties find themselves. The status of the DPA is not the responsibility of any offeror, nor is it under any offeror's control. It is strictly an internal inter-agency governmental matter. The waste of time and resouces inherent in a holding that a procurement is void should not be imposed on an awardee that had no say in the matter, except in cases of "plain illegality." We are unable to determine what it is that protester would have had NOAA do with respect to DPAs. NOAA began this procurement with a DPA of $13.1 million, and at award had a DPA of $19.55 million. NOAA never awarded, nor contemplated awarding, a contract for an amount in excess of its DPAs. We find no DPA exceeded, in spirit or in fact, and no impropriety of any sort with respect to the DPAs here involved. Protester's next ground is that since NOAA knew on June 30, 1994, that there was no possibility that it would award more than one Phase I contract, although the RFP suggested the possibility of two, it should have advised the offerors of that fact. Protester says that had it been so advised before the submission of its second revised proposal and final BAFO it would have changed its proposal to eliminate the objectionable dual operating systems, proposing a single system instead and thereby eliminating one of the items that NOAA regarded as a weakness in its proposal. In addition, protester tells us that it would also have made a "design selection for achieving the ten simultaneous outputs, rather than promising multiple solutions" which would have eliminated (or mitigated) another weakness. Protester's Posthearing Brief at 38. What protester does not tell us is what the elimination of the possibility of a second award has to do with the proposal changes protester says it would have made. Protester gives not one word of logic, argument, or analysis to connect the cause with the stated effects. We can only conclude that none exists. Protester knew from the outset that NOAA would award one contract, but might or might not award two. It was not misled and cannot now successfully complain. Protester's final ground is that its own scores and those of the awardee, CommPower, in the Technical and Technical Management areas were improperly done because its scores were too low and CommPower's were too high. Protester's method of attack is to examine each subfactor in the technical area and adjust a number of its own scores upward based on its judgment that they were too low as given, and some of CommPower's downward based on its judgment that the scores were too high. None of protester's changes were radical, and the sum of them would not have changed the technical standings. Such scores are necessarily somewhat subjective, and opinions with respect to them are bound to differ. Protester has produced no evidence that any of the challenged scores were based on false premises or wrong analyses, or any other impropriety that would vitiate the TAC's findings. With respect to the Cost/Technical Tradeoff, it was clear from the solicitation, which made the technical aspects four times as important as cost, that the competition to be found in this procurement was a competition in technical excellence and not in price. The evidence before the SSO, contained in the TAC and SEB findings, convinced him that CommPower had a better plan, with respect to both software and hardware, had thought it out more completely, had backed it up better when challenged by the Government's questions, and appeared, by reason of its thoroughness in preparation, to be the best bet to be the quicker off the mark to put it into effect, in a procurement where time was important. For similar reasons the SSO had much greater confidence that CommPower's solutions would work and could be put into a prototype sooner than protester's. In other words there was a low risk of serious delay or costly failure. The higher estimated cost was balanced, in the SSO's judgment, by the fact that protester's next best proposal had several serious risks associated with it. The first was that protester was not going to handle its own software development but would farm it out to a subcontractor who would use an unproved development tool that protester itself had no experience with. There was also the fact that the subcontractor's personnel most responsible for core software development changed at least twice during the procurement. Taken together these facts constituted a risk that prototype software development might not proceed smoothly and expeditiously. There was also the fact that protester's proposal required two different operating systems to share the same disk drive, which might well prove very difficult to achieve, and protester's sketchy explanations of how it could be done did nothing to allay the SSO's doubts in this area. Finally protester's hazy description of its multiple designs for satisfying the solicitation's requirement for ten simultaneous outputs also indicated that protester might have difficulty in producing a good result in this area with the design it finally settled upon. The SSO concluded that despite the higher cost, CommPower's less risky, more thoroughly developed, technically better proposal would result in an earlier deployment, which, in turn, might well save lives and property that might otherwise be lost. We cannot say that the protester has shown his judgment to be wrong. Decision For the reasons stated the protest is DENIED. ______________________ DONALD W. DEVINE Board Judge We concur: ________________________ _______________________ STEPHEN M. DANIELS CATHERINE B. HYATT Board Judge Board Judge