Matter of: AT&T File: B-250516.3 Date: March 30, 1993

B-250516.3: Mar 30, 1993

Additional Materials:

Contact:

Shirley Jones
(202) 512-8156
jonessa@gao.gov

 

Office of Public Affairs
(202) 512-4800
youngc1@gao.gov

PROCUREMENT Competitive Negotiation Discussion Adequacy Criteria PROCUREMENT Competitive Negotiation Requests for proposals Terms Compliance Protest is sustained where: (1) proposal selected for award fails to comply with material solicitation requirement. The solicitation was issued by the University of California as the management and operations (M&O) contractor for DOE's Lawrence Livermore National Laboratory (UCLLNL). Which is responsible for contracting for the operation of the ESNET. That it was prejudiced because discussions were conducted only with Sprint. BACKGROUND The procurement is part of an overall program to encourage the development of very high speed data communications capabilities using commercial cell-relay and state-of-the-art technologies.[1] The solicitation contemplated the award of a predominantly fixed-price contract.

Matter of: AT&T File: B-250516.3 Date: March 30, 1993

PROCUREMENT Competitive Negotiation Discussion Adequacy Criteria PROCUREMENT Competitive Negotiation Requests for proposals Terms Compliance Protest is sustained where: (1) proposal selected for award fails to comply with material solicitation requirement, offer of communications switch not yet built could not reasonably be viewed as meeting solicitation requirement that proposal conclusively demonstrate current availability of proposed technology; and (2) contracting officials conducted discussions with proposed awardee, by visiting its proposed subcontractor's facilities to determine technical acceptability of proposed communications switch, but did not conduct discussions with the other offerors in competitive range.

Attorneys

DECISION AT&T protests the proposed award of a contract to Sprint Communications Company under request for proposals (RFP) No. 160506, issued for commercial data communications services for the Energy Sciences Network (ESNET), a nationwide computer data communications network funded by the Department of Energy (DOE). The solicitation was issued by the University of California as the management and operations (M&O) contractor for DOE's Lawrence Livermore National Laboratory (UCLLNL), which is responsible for contracting for the operation of the ESNET. AT&T contends that Sprint's proposal failed to conform to the material requirements of the solicitation, and that it was prejudiced because discussions were conducted only with Sprint.

We sustain the protest.

BACKGROUND

The procurement is part of an overall program to encourage the development of very high speed data communications capabilities using commercial cell-relay and state-of-the-art technologies.[1] The solicitation contemplated the award of a predominantly fixed-price contract, with a 3-year base period and two 1-year options, to furnish commercial data communications services between research facilities that are part of the ESNET based upon "fast-packet technology as it evolves on an early availability and cost effective and on-going basis."[2] Although the solicitation encouraged the use of state-of-the-art technology, and specifically provided for additional capabilities to be incorporated into the network as they become available, it cautioned that "[i]t is considered important to begin high-performance cell-relay based services by late 1992." Traffic load on the ESNET doubles approximately every 6 months and, according to testimony given at the hearing our Office conducted regarding this protest, initial implementation by an early date was considered by UCLLNL to be "important." Transcript (Tr.) at 14, 232. Consistent with the necessity to assure timely, successful initial implementation, the solicitation generally provided that the technology to be utilized by the contractor "to implement the required service is expected to be targeted as a commercial offering"; when "initially provisioned . . . the technology is expected to be in a commercial product/service early release state, comparable to `alpha' or `beta' release status."[3] According to the solicitation, it was: "neither intended nor desired that the services be provided on a private test-bid or non-commercial implementation basis. It is expected that the facilities are, or will be, part of a publicly available shared infrastructure."

The current communications network consists of a number of dedicated, point-to-point circuits which interconnect various sites and are capable of transmitting data at a rate of 1.544 megabits per second. Tr. at 15. The solicitation provided for the replacement of this system by one in which data was transmitted to a contractor-owned-and-managed switching network using cell-relay technologies and capable at initial implementation of providing network service between user sites at a rate of 44.736 megabits per second and communication from a user site to the network (ingress) and from the network to a user site (egress) at a rate of 34 megabits.

Cell-relay technology includes at least two methods of transferring data across a network at very high speeds: switched multi-megabit data service (SMDS) and direct asynchronous transfer mode (ATM). Tr. at 26-27, 36. SMDS is an older, more mature technology, with better defined specifications than ATM. In contrast, ATM offers greater capabilities and was viewed by contracting officials both as the likely future switching structure for major communications carriers and as the long-term goal for the ESNET. Tr. at 36-42, 46-47, 162, 241-243.

As issued on February 24, 1992, the solicitation provided in paragraph 4.2.1.3.1 of the Requirements Specification Document that:

"The user access interface shall initially conform to the SMDS Subscriber Network Interface . . . Allowable exceptions are noted below."

"Alternate subscriber cell-relay based interfaces, such as direct ATM access, may be proposed for longer term alternatives, if significant advantages will result. Any alternative scheme proposed must be sufficiently comprehensive to provide end-to-end (router to router) operational capability."

As noted above, the solicitation also emphasized the incorporation of new capabilities into the network; it specifically listed direct ATM access among the capabilities "expected to be included" in the network. Subsequently, however, UCLLNL advised offerors that direct ATM access could be substituted for SMDS for initial implementation if certain conditions were met. Specifically, by letter of March 30, prior to the submission of proposals, UCLLNL responded to a question concerning initial implementation of direct ATM access as follows:

"Q. Paragraph 4.2.1.3.1 of the Requirements Specification Document specifies that user access shall be provided as SMDS SNI [subscriber network interface]. However, Figure 1 [of the Requirements Specification Document] shows several access methods, one of which is a direct ATM access. Will [UCLLNL] consider direct ATM user access instead of SMDS SNI?

"A. [UCLLNL] believes Paragraph 4.2.1.3.1 is quite clear on this point. However, to emphasize: Any alternative scheme proposed must be sufficiently comprehensive to provide end-to-end (router to router) operation capability. Therefore, the proposal must conclusively demonstrate current availability of the required end-to-end operational capability."

The solicitation generally provided for award to be made to the offeror whose proposal was determined to be most advantageous to UCLLNL, cost and other factors considered. The RFP specifically stated that proposals would be evaluated on the basis of technical approach and management, and that technical approach would be approximately five times more important than management. The solicitation described cost as important and indicated that it would be considered in the selection of the contractor; it provided that cost or pricing data would be evaluated to determine the probable cost, cost realism and cost reasonableness of proposals.

Proposals were received from four offerors, including Sprint and AT&T, by closing time on May 5. Based upon its evaluation of initial proposals and specific supplemental information requested by contracting officials, UCLLNL's Source Evaluation Board (SEB) awarded Sprint the highest overall technical/management score (80 points), 11 points higher than AT&T's second place score (69 points). Sprint was evaluated as the "technical leader among the offerors"; all but one point of the overall difference in scores relative to AT&T was accounted for by Sprint's higher technical score. In particular, the SEB noted that only Sprint had proposed to bypass SMDS and instead offer direct ATM access for initial implementation. (While Sprint proposed to complete laboratory testing of ATM by August 1992 and commence initial service implementation by late 1992, AT&T proposed to upgrade service to ATM only in 1994.) Further, 9 of the 17 evaluated strengths found in Sprint's proposal under the evaluation factor for implementation strategy, which was the most important factor under technical approach and was twice as important as the next ranked factor, concerned Sprint's proposal of direct ATM access and its "benefits" and "advantages."

Although the SEB determined that "AT&T is within the competitive range," AT&T's proposal prices were found to be "expensive and unreasonable." In view of AT&T's prices and Sprint's position as the "technical leader," the SEB report stated that no best and final offers were necessary as Sprint had been "preselected without further discussions or negotiations with the other offerors."

However, the SEB report was withheld from the source selection official (SSO), and any decision to make award to Sprint was postponed, pending visits to Sprint and to TRW Corporation, Sprint's proposed subcontractor for the ATM switch. Tr. at 470. According to the SEB report:

"a trip to TRW . . . is required to verify their ATM switch. A trip to Sprint is also necessary for Sprint to provide a clear explanation of their pricing methodology utilized in their proposal. Their local DCAA [Defense Contract Audit Agency] office was unable to perform the audit as requested and a visit with Sprint . . . is necessary for [UCLLNL] to develop a plan to negotiate appropriate pricing for this procurement."

According to UCLLNL's project manager for the ESNET, "the state of technology [offered by Sprint] was less far along" in development than AT&T's proposed technology and the trip to TRW "was partially [needed] to assure ourselves that they did have the capability . . . and to find out . . . [on] what kind of schedule they could produce that kind of equipment." Tr. at 149.

Although TRW's proposed ATM switch could not be demonstrated during the subsequent July 8 visit to TRW because a prototype had not yet been built and the switch was still in the design phase, another high-speed switch was demonstrated and the Technical Evaluation Committee of the SEB was "favorably impressed with [TRW's] capabilities." Tr. at 119, 201, 222-223. As a result, the SEB report recommending award to Sprint was furnished to the SSO, and on July 16 he approved the selection of Sprint. Upon being advised that its proposal had not been selected, AT&T immediately requested a debriefing; the debriefing, however, discussed AT&T's proposal and not Sprint's. Subsequently, on September 8, based upon an August 24 article in the trade press stating that Sprint had been selected on the basis of its proposal to bypass SMDS and instead initially implement ATM, AT&T filed an agency-level protest. It then protested to our Office.

TECHNICAL COMPLIANCE

AT&T recognizes that paragraph 4.2.1.3.1 of the Requirements Specification Document, as clarified in UCLLNL's March 30 question and answer, permitted offerors to propose direct ATM access, instead of SMDS, for initial implementation. AT&T contends, however, that Sprint's proposal of an ATM switch for which no prototype had yet been built at the time of evaluation did not comply with the requirement in the clarification to paragraph 4.2.1.3.1 that where an alternative scheme such as ATM is proposed, "the proposal must conclusively demonstrate current availability of the required end-to-end operational capability." Further, AT&T argues that the July 8 visit to TRW by the Technical Evaluation Committee constituted discussions and therefore required UCLLNL to conduct discussions with, and request BAFOs from, all firms in the competitive range, including AT&T.

As an initial matter, DOE maintains that "the technical requirements" of paragraph 4.2.1.3.1 of the Requirements Specification Document were intended for relative evaluation only, and did not constitute mandatory pass/fail requirements. It bases its position on the fact that the requirements of paragraph 4.2.1.3.1 were not included among the requirements listed under paragraph 4.0, "Qualification Criteria," of the Requirements Specification Document, which provided that "[t]he following are minimum criteria that must be met before further consideration of the proposal will be made. . . ."[4]

We do not find DOE's position persuasive. The language of paragraph 4.2.1.3.1 and of the subsequent clarification are phrased in the imperative: "Any alternative scheme proposed must be sufficiently comprehensive to provide end-to-end . . . operation capability. Therefore, the proposal must conclusively demonstrate current availability of the required end-to-end operational capability." (Emphasis added.) Nowhere does the solicitation expressly provide for only the relative evaluation of this requirement. Moreover, we note that the qualification criteria listed in the Requirements Specification Document appear to be primarily directed at the status and fundamental capability of the offeror. The criteria include a minimum experience requirement for the offeror and a requirement that the communications services use cell-relay facilities and capabilities provided by a single communications vendor on a nationwide basis and supported 24 hours-a-day. Inclusion in the solicitation of factors related to the status and capability of the offeror, that is, inclusion of responsibility-type factors, is not inconsistent with the presence elsewhere in the specification of mandatory material specification requirements. Given the mandatory language of paragraph 4.2.1.3.1 and the absence of any provision for only a relative evaluation, we find that the paragraph established a material, mandatory solicitation requirement.

DOE further maintains that Sprint's proposal in fact complied with a reasonable interpretation of the requirement in paragraph 4.2.1.3.1 that an offeror proposing an alternative to SMDS "conclusively demonstrate current availability of the required end-to-end operational capability." DOE claims that the requirement of paragraph 4.2.1.3.1 was phrased in terms of demonstrating a "capability" specifically in order to allow for proposals offering initial user access based on emerging technology, that is, on features not yet developed at the time of proposal submission. DOE maintains that this interpretation is consistent with the overall solicitation emphasis on encouraging the development of, and incorporating into the ESNET, emerging technology. DOE also cites in support of its position Article VI of the solicitation schedule, which allows 3 months after award for "[i]nstallation of the Initial Implementation," states that "Initial Implementation will be installed in pre-production status," and provides for implementation of a fall-back approach in the event the contractor fails to "fully demonstrate adequate progress by the due date."

Again, we find DOE's position to be unpersuasive. Although the solicitation clearly provided for and encouraged the incorporation of emerging technology such as direct ATM access into the ESNET as it becomes available, the Requirements Specification Document cautioned that "[i]t is considered important to begin high-performance cell-relay based services by late 1992." In other words, in our view, the solicitation established a two-phased approach to upgrading the ESNET: encouraging the ultimate use of emerging technology while at the same time assuring the early, initial implementation of urgently needed improvements to the network. (The stated necessity for early improvements to network capacity apparently arises from the increase in traffic load, which is doubling every 6 months.) This interpretation of the solicitation is consistent with permitting initial implementation using equipment in pre-production status but requiring offerors proposing alternatives to SMDS such as ATM to establish the "current availability" of their alternative (and providing for use of a fall-back approach if it becomes necessary).

In any case, the solicitation requirement was clear on its face: where an offeror proposed an alternative to the initial implementation of SMDS, "the proposal must conclusively demonstrate current availability. . . ." Notwithstanding DOE's views to the contrary, the requirement to conclusively demonstrate in the proposal the current availability of the proposed alternative to SMDS cannot reasonably be viewed as being met by the proposal of a switch that had not yet been built, however confident the evaluators may have been after their visit to TRW that Sprint's proposed subcontractor possessed the capability to build the switch in the future. Likewise, the fact that TRW is reported to have ultimately demonstrated the proposed switch in December 1992 does not alter the fact that at the time of proposal submission, evaluation, and source selection, its proposal failed to conform to a material solicitation requirement. In negotiated procurements, it is fundamental that any proposal that fails to conform to the material terms and conditions of the solicitation should be considered unacceptable and may not form the basis for award. See Martin Marietta Corp., 69 Comp.Gen. 214 (1990), 90-1 CPD Para. 132; Consulting and Program Mgmt., 66 Comp.Gen. 289 (1987), 87-1 CPD Para. 229.

PREJUDICE

DOE and Sprint question whether AT&T was prejudiced by its acceptance of Sprint's proposal of initial implementation of direct ATM access. First, UCLLNL questions whether AT&T could have proposed ATM even if it had known that less than "current availability" would have sufficed. In addition, Sprint points out that, while AT&T offered network service between user sites at the required rate of 44.736 megabits per second for the initial implementation, AT&T did not offer the required egress rate, from the network to a user site, of 34 megabits for initial implementation. Instead, AT&T offered an egress rate of 10 megabits for initial implementation, with a transition to 34 megabits occurring only in 1993. Finally, DOE notes that AT&T's prices were found to be unreasonable.

Based upon our review of the record, we find that there was a reasonable possibility that AT&T was prejudiced by UCLLNL's actions in waiving the current availability requirement for initial implementation of direct ATM access. See M.C. Dean Elec. Contracting, Inc., B-248835.2, Nov. 16, 1992, 92-2 CPD Para. 346; George A. Fuller Co., B-247171.2, May 11, 1992, 92-1 CPD Para. 433. The record indicates that AT&T has had an operational, but non-production status ATM network running since approximately early 1992. The record also supports the conclusion that AT&T could have offered a preproduction ATM approach for initial implementation in 1992, and that had it known UCLLNL would accept ATM technology for initial implementation which was not currently available in production status, it would have preferred to offer direct ATM access. (Indeed, AT&T viewed an SMDS network as representing a higher risk than an ATM network because of its uncertainty as to whether there would be sufficient public customers for SMDS.) Tr. at 360, 389-390. Given UCLLNL's favorable evaluation of Sprint's proposal of early implementation of direct ATM access, which was based upon an ATM switch which had not even been built, had AT&T proposed early implementation of direct ATM access it appears that it could have substantially increased its technical score.

AT&T also states that it could have implemented the required 34-megabit egress rate on an SMDS network by late 1992 if UCLLNL had expressly questioned its proposal in this regard. (According to AT&T, it offered a lesser rate based upon its best estimate of how rapidly UCLLNL's data communication needs, which are now met by 1.544 megabit circuits, were likely to increase.) Such discussions could not properly have been avoided since Sprint's proposal failed to conform to the requirements of the solicitation.

Although AT&T's prices appear to have been substantially higher than Sprint's, a precise comparison of the proposals is difficult. There is no evidence in the record that UCLLNL calculated an overall most probable cost for either Sprint or AT&T when making the source selection.[5] As indicated above, the SEB concluded that it lacked "a clear explanation of [Sprint's] pricing methodology." Some of the pricing information supplied by Sprint in its proposal was based on 3 years of service rather than the 5 years the contract could extend if the options were exercised. As for AT&T, its proposal indicated that in general terms, the service charges to be billed to UCLLNL for SMDS service would essentially depend in part upon the percentage of network capacity benefiting ESNET users; apparently, the greater the use of AT&T's SMDS network by non-ESNET users, the lower the rates charged UCLLNL. AT&T, however, capped the service charges to UCLLNL based upon a maximum of 50 percent network usage.

AT&T has furnished calculations showing that it could have offered a substantial price reduction for an ATM network based upon its expectation of significantly greater future public demand for ATM than for SMDS. (AT&T's expectation in this regard is consistent with UCLLNL's expectation that the major communications carriers are moving toward use of the more capable ATM switching structure.) In addition, an offeror must be advised during discussions that its price is more or less than what the agency believes is reasonable. Steinhof & Sadler, Inc. d/b/a SSI, B-246604; B-246604.3, Mar. 20, 1992, 92-1 CPD Para. 299. Had UCLLNL advised AT&T that its prices were considered unreasonable, this might have resulted in additional reductions beyond those attendant upon AT&T's proposal of an ATM approach.

In summary, Sprint based its proposal, both technical and price, on an ATM switch that had not yet been built. Although we cannot know with absolute certainty what AT&T would have done, it is clear from the testimony at the hearing that, because of its expectation of a greater public demand for an ATM network than for a SMDS network, AT&T likely would have proposed an ATM approach for initial implementation had the solicitation accurately reflected UCLLNL's intentions with respect to initial implementation. It is also clear that had AT&T proposed an ATM approach for initial implementation, it would have received a substantially improved technical score. (Again, while Sprint's proposal of a nonexistent ATM switch resulted in a substantially enhanced score, AT&T was running an operational, but preproduction ATM network.) In addition, it appears that AT&T would have substantially reduced its price. We therefore conclude that AT&T was prejudiced by UCLLNL's improper actions in conducting discussions with Sprint but not with other offerors in the competitive range, and then selecting Sprint for award notwithstanding the fact that it had failed to satisfy a mandatory solicitation requirement.[6]

The protest is sustained.

RECOMMENDATION

We recommend that UCLLNL revise the solicitation to accurately describe the state of development it considers acceptable for equipment proposed for the initial implementation, open negotiations with all offerors, and then request best and final offers. Further, we conclude that AT&T should be reimbursed its protest costs, including reasonable attorneys' fees. 4 C.F.R. Sec. 21.6(d) (1992).

1. Cell-relay is a switching technique in which data packets of fixed length are used, resulting in reduced processing overhead and higher speeds.

2. Fast-packet is a generic term for various stream-lined packet technologies which permit processing at higher speeds with minimal processing overhead.

3. The solicitation elsewhere equated alpha release status with "Test bed: off-line testing," and beta release status with "Pre-Production: initial release, limited number of sites."

4. Sprint, but not DOE, also argues that our Office lacks jurisdiction under the Competition in Contracting Act of 1984, 31 U.S.C. Sec. 3551 (1988), which governs our bid protest process, to consider bid protests by potential subcontractors. In support of its position, Sprint cites the decision in U.S. West Comms. Servs., Inc. v. United States, 940 F.2d 622 (Fed. Cir. 1991), wherein the court held that the General Services Board of Contract Appeals was precluded from considering subcontractor bid protests. However, we need not consider this argument, since DOE's regulations specifically provide for our Office to consider protests involving acquisitions by M&O contractors such as UCLLNL. Department of Energy Acquisition Regulation, 48 C.F.R. Sec. 970.7107 (1992). Thus, we would consider AT&T's protest even if Sprint's argument were correct. See 4 C.F.R. Sec. 21.11 (1992); cf. United Telephone Co. of the Northwest, B-246977, Apr. 20, 1992, 92-1 CPD Para. 374, aff'd Department of Energy, Recon. et al., B-246977.2 et al., July 14, 1992, 92-2 CPD Para. 20.

5. As a general rule, an agency is required to include cost or price as a significant factor in the evaluation of proposals, 41 U.S.C. Secs. 253a(b)(1)(A) and 253b (1988); Federal Information Resources Management Regulation Sec. 201-39.1501-1(a), and an evaluation and source selection which fails to give significant consideration to cost cannot serve as the basis for a reasonable source selection. See Lockheed, IMS, B-248686, Sept. 15, 1992, 92-2 CPD Para. 180. In our view, UCLLNL was required to calculate and take into account in the evaluation the overall most probable cost of proposals before making its source selection decision.

6. While it is unnecessary to fully address the issue because of our conclusion that Sprint's proposal was not acceptable, we note that AT&T has also raised a substantial question whether UCLLNL conducted discussions only with Sprint. Discussions occur when an offeror is given the opportunity to revise or modify its proposal or when information provided is necessary for determining the acceptability of its proposal, and, when held, they must be conducted with all offerors whose offers are within the competitive range. National Medical Staffing, Inc., B-242585.3, July 1, 1991, 91-2 CPD Para. 1; Adak Communications Sys., Inc., B-226952, June 1, 1987, 87-1 CPD Para. 556.

The SEB report stated that a visit to TRW was necessary "to verify their ATM switch," which was still in an early stage of development and had not yet been built. Tr. at 149, 201. The SEB report conditioned its findings on the outcome of visit to TRW, and the report was forwarded to the SSO only after the Technical Evaluation Committee had been briefed by TRW and reached a favorable conclusion concerning TRW and the proposed ATM switch. While an agency may conduct a preaward survey to evaluate the responsibility of a prospective awardee, the SEB made the validity of its findings contingent upon its Technical Evaluation Committee obtaining further information concerning the proposed product. Therefore, the visit to TRW appears to have constituted discussions because it was used to obtain information concerning the acceptability of Sprint's proposal. Since, according to the SEB report, AT&T was also considered to be within the competitive range, UCLLNL could not properly conduct discussions only with Sprint.