Granite Construction Co. v. United States , 24 Cl. Ct. 735 ( 1991 )

OPINION

BRUGGINK, Judge.

This is an action brought pursuant to the Contract Disputes Act of 1978.¹ Plaintiff, Granite Construction Co., seeks an equitable adjustment based on an asserted Type I or Type II differing site condition, on failure to disclose superior knowledge about the difficulty of doing the work, and on improper acceleration of the work. At the conclusion of plaintiff’s trial presentation, the court granted a motion pursuant to RUSCC 41(b) as to the Type I and Type II differing site condition claims, and denied the motion in other respects. The reasons for granting the motion were stated on the record. After the conclusion of trial presentation, and for the reasons set out below, the court concludes that the plaintiff is not entitled to recover on its claims of superior knowledge and acceleration.

BACKGROUND FACTS AND PROCEDURAL HISTORY

The subject of the litigation is Granite’s construction of a salmon fingerling bypass in the John Day Dam, an existing dam on the Columbia River. The river historically has supported a large migratory run of salmon. With the construction of a series of dams along the Columbia and its tributaries by the U.S. Army Corps of Engineers (“Corps”) and the Bureau of Reclamation, salmon fingerling mortality rose as they were required to pass over spillways or through turbines as they traveled downstream.

In an effort to reduce fingerling mortality, the Corps and others began to develop plans for various bypass systems to be incorporated into new dams being constructed and retrofitted into existing dams. These bypass systems were intended to divert the fingerlings from powerhouse intakes and direct them into a bypass tunnel and outfall sluice leading back to the river below the dam.

The John Day Dam is a multipurpose facility that is comprised of, among other things, a powerplant, non-overflow portions, a series of service, powerhouse and skeleton bays, a spillway, a navigation lock, and fish passage facilities. The service and powerhouse bays are composed of mass concrete. There are expansion joints between each of the bays, at intervals of approximately 90 feet.

A fingerling bypass tunnel, comprised of a steel conduit, was incorporated into the John Day Dam when it was built, but experience revealed that it was ineffective. As a replacement for the original system, the Corps determined to pursue the project at issue, which was a barrier screen system that diverted fingerlings from powerhouse intakes toward orifices leading to a transportation conduit and outfall sluice that returned them to the river below the dam. The transportation conduit portion of the project required the excavation of a tunnel through the mass concrete of John Day Dam.

The present litigation arises out of a contract between the United States, acting through the Corps, and Granite for the construction of this bypass system. The contract was awarded on January 5, 1984 in the amount of $7,898,465.00. There were four principal items of work: vertical barrier screens and guides, excavation and lining of the transportation conduit, outfall chute, and evaluation facility for a juvenile bypass.

The dispute concerns only one element of this work, the transportation conduit. Approximately 1000 feet of this tunnel had to be excavated out of the existing mass concrete that comprises, principally, the nine bays of the dam. The tunnel was not uniform in shape. The profile was rectangular, but the floor sloped down to the north end to induce water flow, and the squared sides tapered from north to south. Not including the wedge added by the sloping floor, the shape at the large end was approximately ten feet (high) by twelve feet.

The tunnel thus went through the length of the dam, perpendicular to the flow of the river. At several points along the roof of the tunnel small holes were to be punched through to the gate slots. Fish would be attracted to the holes by light, and then would be drawn to the tunnel and then through a separate, much shorter tunnel that had to be excavated through monolith 14, to the outfall chute and eventually to the river below.

In order to provide a proper chronological background for Granite’s claim of superior knowledge, the discussion has to be diverted to a prior Corps project. The Walla Walla District of the Corps in July of 1978 contracted with S.J. Groves and Sons Company (“Groves”) to construct a fingerling bypass system at the existing Little Goose Dam along the Snake River in Washington. This project, in part, called for the excavation of a tunnel through mass concrete. Groves attempted to complete at least part of the tunnel excavation at Little Goose Dam with a machine called a road-header. This is a large piece of mining equipment which uses a rotating cutting head, attached to a long arm. In bidding the Little Goose project, Groves proposed to use an Alpine Equipment Company Model F16 roadheader, weighing 40,000 pounds. However, in performing the work, Groves actually attempted to excavate the tunnel with a modified Alpine Model F6A, weighing about 26,400 pounds.

The specifications for the Little Goose work contained the following paragraph:

1.3 The existing concrete in the path of mining may be expected to range in strength from about 3000 psi to over 8000 psi. Maximum coarse aggregate size is approximately 3 inches. Additional information on the existing concrete can be found in the publication ‘Report of Concrete Operations and Tests, Little Goose Dam’, available for review from the Army Corps of Engineers District Office in Walla Walla, Washington. It is expressly understood that interpretations, conclusions and deductions made on the basis of this report are solely the responsibility of the Bidder.

Contrary to the specification, Groves in fact encountered substantial quantities of coarse aggregate or cobble ranging in size up to six inches or more. The F6A initially put on the project proved to be totally inadequate. It was modified with a larger cutting arm, cutting head, and cutting head motor. Although it was eventually able to excavate a substantial part of the tunnel, its performance was not considered to be satisfactory. Groves submitted a differing site condition claim to the Corps, asserting that its difficulties in excavating the tunnel were attributable to the presence of significant quantities of six-inch aggregate in the mass concrete when the contract specifications indicated that the maximum aggregate size would be only three inches.

By letter dated February 28, 1979 the Corps’ Contracting Officer (“CO”) issued a final decision denying Groves’ differing site conditions claim upon the ground that its difficulties were attributable to the inadequacies of the modified F6A Alpine road-header. The Corps took the position, based in part on expert reports prepared for it and attached to the decision, that the size of the aggregate was not a factor in the difficulty of the work. The CO final decision reflects that the Alpine roadheader was frequently not operating; that it milled² the concrete when it did operate; and that the machine was ultimately pulled out and the work completed by hand-mining techniques. The matter was ultimately tried before the Armed Services Board of Contract Appeals, and, based on the trial judge’s indication that Groves’ claim was valid, the parties settled.

The invitation for bids for the John Day bypass project was issued by a different Corps district than the one which had been in charge of the Little Goose project. The project at John Day was handled by the Portland District of the Corps. The Invitation for Bids (“IFB”) was issued on November 8,1983. Bid opening was originally set for December 6, 1983, but was extended to December 15. The bidding and other contract documents furnished to bidders made no reference to the Little Goose project. Internal Corps memoranda make clear, however, that the experience at Little Goose was known to the technical staff in Portland. The engineering division acknowledged “great difficulty at Little Goose.” This knowledge later was considered when the estimate was formulated.

David Gove was the area manager for Granite. Typically, he would have taken an active role in preparation of the John Day bid. Due to the press of other business, however, he turned over to Don Gillis, an estimator, the task of preparing the part of the bid related to the tunneling operation. At that time, Gillis was in Granite’s Atlanta office. When Granite later got the contract, Gillis agreed to move to Oregon for six months to be project manager for the work.

Neither Gove nor Gillis had experience with excavating concrete. Both were generally familiar with blasting concrete, but had never cut it to dimension. Since they were unaware at that time of the Little Goose project, the idea of tunneling through concrete was, to their understanding, a unique undertaking.

One of plaintiff’s experts, Terrence McCusker,³ testified that in 1983, the “presumptive” method for cutting concrete to dimension would have been with hand-mining techniques, such as drilling and blasting. Blasting was forbidden in the IFB for John Day, however. Other available hand-mining techniques would include drilling and removal through chemical or hydraulic expanders. Unknown to any of the bidders, the Corps calculated its pre-bid cost estimate for the excavation of the tunnel based upon such a hand-mining operation.

Mass concrete, which is used in massive structures such as John Day Dam, is placed in large quantities without reinforcing steel throughout most of the structure. The parties have stipulated that the primary differences between mass concrete and other concretes are the larger aggregate sizes and greater ratio of aggregate to cement. Both Gove and Gillis had experience with placement of mass concrete. Granite was a joint-venture partner for the Bonneville second powerhouse contract, which required the placing of mass concrete. Both men also acknowledged that they understood the general principle that in mixing concrete, it is generally desirable from economic considerations to use as much aggregate as possible.

Gillis and Gove never seriously considered handmining, because they felt it would be too labor intensive and costly. Both were familiar with the use of road-headers to do tunneling. Gove had been involved with a tunneling job in Montreal in the 1970’s in which a roadheader was used to excavate in shale, a relatively low-strength material. Gove was “dissatisfied” with the performance of the machine, however, because it was unable to cut through intermittent lenses of much harder dolomitic rock. Granite was also using a roadheader at a job cutting a tunnel through mudstone and sandstone at Ridgeway, Colorado. Gillis called Steve Elmore, project manager for Granite at Ridgeway. Elmore told Gillis he was satisfied with a Dosco Mark II-A that Granite was using there.⁴ Gillis told him the John Day work would involve excavating concrete, to which Elmore responded that he did not know how well it would perform. Gillis was also familiar with use of road-headers in softer materials from previous jobs.

Both Gillis and Gove understood that roadheaders were used primarily in soft material applications, such as coal mining. They also understood that the IFB’s indications of compressive strength (expressed in terms of pounds per square inch or “psi”), though low, related to the amalgam as a whole, and not to the aggregate. They understood that the aggregate would be extremely hard, and probably well beyond the normal range of a roadheader. Gove testified that in his view, using a roadheader in concrete would be an “experimental” application. Gillis stated that at that time he would not have considered using a road-header in material with a compressive strength in excess of 15,000 psi.

In view of these initial contra-indications, Gove and Gillis’ willingness to believe that the cutting head of a roadheader would be effective against mass concrete is surprising. One factor militating in favor of the roadheader, however, was Gove’s belief that the cutting head of the roadheader would be in very little contact with the aggregate. He judged, in part from pictures of a core sample of John Day concrete, that the percentage of large aggregate was relatively low, and thus the teeth or “picks” of the cutting head would be able to push through the cement matrix and break off chunks with aggregate imbedded. Gillis’ understanding was similar. He knew that the roadheader could not cut through aggregate, but felt that the matrix was sufficiently weak that, even if the teeth of the cutting head were to hit the cobble instead of cement, the stone would break loose from the cement.

To some extent this understanding was based on contacts with the Dosco Corporation (“Dosco”), the third party defendant. Although there were a few other companies of which he was aware that manufactured roadheaders, Gove felt the most promising supplier was Dosco. Based on previous contacts, he had no interest in pursuing the others. Gillis similarly was convinced that the other products either would not fit in the tunnel or were underpowered.

Dosco engages primarily in two types of business. It manufactures and markets tunnel support systems, and it also markets roadheaders manufactured in England by a related company, Dosco Overseas Engineering Ltd. (“Dosco Overseas”).⁵ Since at least 1980, as part of the development of roadheader equipment, Dosco Overseas has sponsored extensive research and testing of roadheader equipment, including its capacity to cut concrete, at the University of Newcastle-Upon-Tyne, in England.

Gillis’ initial contacts at Dosco were with J.E. Marianski, at that time Vice President of Dosco. In his initial contact, Gillis did not tell Marianski the project for which the machine was needed. He told him that the machine would have to cut through shale. Gillis did this because he considered the John Day project to be relatively small. He hoped it would not be generally known and thus did not want to alert potential rivals by disclosing the precise project.⁶ After the initial contact, Marianski designated A.D. Smith, western sales representative, to follow up, and tasked Ian McKinlay, a Dosco sales representative to do some background research on the project.

On November 18, Dosco sent Gillis a brief cover letter to some literature describing two types of roadheader. This literature contains certain standard estimated performances. The larger of the two machines, and the one ultimately used, an SL120, was quoted at a production rate of 80 tons per hour. In the interim Gillis told Smith the true nature of the work. Smith, only a three week employee at Dos-co, called Marianski to ask how to find out if a roadheader could cut concrete. Marianski told him to contact Dosco Overseas.⁷ Based on information from England, and on Marianski’s general knowledge about roadheaders, Dosco passed along new figures, lowering estimated production. The letter put in evidence contains Gillis’ handwritten notes of these subsequent telephone conversations with Dos-co personnel. They indicate that “per Marianski” Dosco would conduct a site visit on December 5, and would report back. The next notes were from conversations in early December, with Ian McKinlay, and on December 12, with Doug Smith, about which more later.

During this same period, Dosco contacted or was contacted by other prospective bidders on the John Day project, including A1 Johnson Co., Dillingham Construction Co., and Pacific Ventures, Inc. Marianski testified that the differences in the contractors’ assessments of the difficulty of the project gave him some elevated concern. Roughly at that time he asked McKinlay to research the project. It is this upcoming mission which is reflected in the notes on the November 18 letter to Granite.

It is noteworthy that on the same date (November 18) Dosco sent two letters to other prospective bidders that were very different from the one initially sent to Granite. The one sent to Pacific Ventures, for example, warned that

The performance of the machine would be directly linked to the size, hardness and density of the aggregate. In the worst situation where the aggregate would be closely packed and of a quartz nature, it would be impossible for the machine to sump in under its own power. If the material is going to be of this nature we would recommend either drill and shoot or impact hammers. However, where there is sufficient cement to attack with the cutting head teeth, the machine would be able to successfully rip out the aggregate.

“Sumping in” is a reference to the need for a machine to make the initial breach in an otherwise solid face of concrete. Once a hole is cut, the cutting head can arch around the inner edges of the hole more easily. Pacific Ventures was quoted a production rate of 40 tons per hour. This is the rate ultimately quoted to Granite by telephone. A similar letter was sent to Dillingham Construction Co.

Dosco had had some earlier unsuccessful experience with use of an SL120 roadheader in hard material. With reference to a tunneling project at Glenn Canyon Dam, which involved a small amount of tunneling through concrete at the end of a sandstone tunnel, Marianski wrote a contractor on August 26, 1983, that “the aggregate is extremely dense and consists of 3" and 4" diameter quartz boulders. There is no way can any roadheader type bit make any impression on this material.” To some extent this description of the concrete part of the work at Glenn Canyon is very different than that given by Ken Payne, a Dosco service representative present at that work site. He testified that the roadheader tried for “five to ten minutes” to penetrate the concrete, but that it encountered boulders 18 to 24 inches in diameter, as well as two to three inch reinforcing steel bars.

It is difficult to fully accept both Payne’s and Marianski’s statements about Glenn Canyon. First, there is an inconsistency in the size of the aggregate. Also, it is difficult to credit both that the machine made no impression on the material, and that, despite working only five or ten minutes, it uncovered rocks two feet across which, according to Payne, were not visible on the surface of the concrete. Perhaps Marian-ski had reports from others describing conditions differently. Resolving the discrepancies is not important, however. It is sufficient, for reasons discussed below, to note that Dosco personnel had reason to be very skeptical of the ability of an SL120 to sump into concrete.

The rates quoted to Granite were “instantaneous” rates, i.e., assuming maximum efficiency and constant contact of the cutter head with the concrete surface. Both Marianski and Granite personnel understood that actual effective production rates would be much lower due to inherent loss of contact time of the head with material.

Notes on Granite's copy of the November 18 letter reflect Dosco’s modification of production rates. McKinlay called back to modify the estimated output and cost figures shown in the standard literature. For the SL120, the stronger of the two machines, the output figure dropped from “80 to 100 tons per hour” to 25 tons per hour (approximately 10-15 cubic yards). The cost per ton of replacing the bits or teeth on the cutting head went up from $.50 per ton to “$1.00—$1.50” per ton. The maintenance cost went from $.50 per ton to $1.50 per ton. When Doug Smith called back, the production figures went up, to 40 tons per hour, but the cost of bits also went up dramatically, to $15.00 per ton. Maintenance cost went down by half, to $.75 per ton.

By this time, Smith had heard from a friend that a similar project had been done at Little Goose. He called the Walla Walla office of the Corps and asked about that previous work. He does not remember with whom he spoke, but whoever it was told him that the Alpine had had a difficult time with the concrete tunneling. The Corps employee was not willing to discuss the details of the work, saying that there was a claim in litigation. Smith knew Groves had done the work, but did not get in touch with the contractor.

Because of his misgivings, around December 1, Marianski told McKinlay to conduct a site visit at John Day, and told him to find out how the machine at Little Goose operated. McKinlay was shown all around the John Day dam. He walked the galleries that ran through the same mass concrete in which the tunnel was to be cut. He was shown core samples from mass concrete said to be from the navigation lock on the north end of the dam. McKinlay was also directed three miles downstream to a staging area where surplus aggregate from the dam was still stored. He took several photographs.

The following day, December 7, McKinlay went to the Division Headquarters at Walla Walla to inquire about the problems at Little Goose. He spent about three hours there. He was told that an Alpine F6A had been used. He was not told that the machine had been modified. That information was contained in the reports attached to the CO’s decision on Little Goose, however, and McKinlay was offered access to the decision, but declined. He was told that the machine did not complete the work, and he understood that it had serious difficulties. He was shown closeup photographs of the surface of the tunnel and could see that some aggregate up to six inches had been dislodged and other pieces had been sheared. McKinlay assumed that since the F6A was a “soft rock machine,” the heavier SL120 would be able to do the work. Based on his visit to Walla Walla, McKinlay deduced that the John Day tunnel would be very hard to excavate. There would be very high bit costs, and it would be a “severe application.”

To some extent this is reflected in a report McKinlay sent to Marianski concerning his two visits. The report was accompanied by the photographs. As to John Day, McKinlay reported that his tour gave him little understanding of the nature of the concrete, but that he was shown a single core sample, which, in his judgment, contained a very low percentage of six inch materials. He mailed these photographs to Marianski and some were introduced at trial. As to the surplus aggregate pile he was directed to, it contained “no trace of any 6" material.”

Marianski testified that he was not concerned by the core sample photographs, since in his judgment they reflected concrete that was not that dense and did not appear to have six inch aggregate. The court’s examination of the same pictures causes it to question his sanguine reaction. A four-inch cigarette pack in the picture to show scale suggests that cross-sections of many of the stones are at least four inches, and that the aggregate is sufficiently dense that the chance of the parallel teeth of the cutting head hitting matrix would be slim indeed.

McKinlay also noted, moreover, that he was told that the aggregate at John Day might be harder than that used at Little Goose. He also reflected that Groves’ operating records were somewhat contradictory. It had both “high operational costs” and a “low advance rate” in six inch aggregate, as well as “low operational costs and high advance rate in 6" aggregate.” He attached figures on costs and production at Little Goose. They bear out the former observation. They show that the production rate per operating hour was only 5.6 tons per hour.

On December 13, Smith wrote a letter to Granite covering a proposal to furnish the SL120 for the excavation work. As to performance and cost, it is the same as the information reflected in the notes of the telephone call to Gillis of the previous day. It specifically “estimated performance in concrete with 4"—6" aggregate up to 40 [tons per hour].” Marianski, from whom Smith got this data, testified that he would expect an effective utilization rate of no less than 15 tons per hour. The letter goes on to make the following disclaimer, however:

The nature of the project is such that we cannot guarantee production performance or cutter costs, however the following estimates are offered as a guide only to enable you to estimate progress in preparation of your proposal.

The cutter costs could be extremely high due to the quartz and basalt aggregates and these estimates are a result of experiences at the factory in England during machine testing in preformed concrete blocks and also from observations by our Engineering Service Manager during a visit to the job site and also a visit to the Walla Walla District Office of the Corps of Engineers which supervised a similar project at the Little Goose Dam on the Snake River.

Gillis testified that he received this letter after bid opening on December 15, 1984, but that in his view the substance of it was conveyed in earlier conversations. He had already heard Dosco’s revised estimate of production, and based his own calculations for Granite’s bid on it. He discounted the 40 ton per hour figure (which Dosco informed him was conservative) by 35 percent, to 25 tons. Gove discounted it even further, to 23.5 tons per hour. These reductions were based in part on Gillis’ lack of familiarity with the machine.

There was no scheduled pre-bid tour on the John Day project. Nor was there a pre-bid conference. Instead, the bidding documents provided that prospective bidders could arrange for their own tours of the site, or could contact David Illias, a designated Corps employee to ask about “administrative” matters relating to plans and specifications. Gillis never contacted Illias, although he took a tour of the dam facility. He spent four or five hours at the site. He made no specific inquiries about the concrete and was furnished no information about it. Prior to the submittal of Granite’s bid, neither Granite nor Dosco advised the Corps that Granite intended to utilize a roadheader on the John Day project.

Seventeen bids were submitted to the Corps for the John Day project. Granite’s bid of $7,898,465 was the lowest submitted; the Corps’ estimate for the project work, not including profit, was $11,818,899.

At the request of the Corps, Granite, by letter of December 21, 1983, verified that its bid contained no error. Gove interpreted the request for verification as inviting a confirmation of the mathematics of the bid. He did not notify Gillis that the request was made. Thereafter, on January 5,1984, the Corps awarded the contract for the John Day project work to Granite.

Gillis testified that at no point prior to the bid was he aware that “a similar project,” such as referred to in Dosco’s December 13 letter, had taken place. He was not told about Little Goose by Dosco personnel. The reduction in production figures made on December 12 was attributed merely to “additional data.” It was not until the day of the bid that, after opening, someone from Groves told Gillis that it had unsuccessfully attempted to use a road-header.

Shortly afterwards, both Granite and Dosco found out that the roadheader used by Groves was modified with a larger boom and more powerful motor. This caused some concern. As a result, Dosco arranged a test of the SL120 on concrete with large aggregate. It found an abandoned bridge abutment near Grand Junction, Colorado and succeeded in demolishing it with the SL120.

In a report prepared for Granite, Dosco indicates that, although the compressive strength of the test concrete was less than that at John Day (maximum measured at 3,700 psi) the matrix materials were comparable in strength. The report states that the strength of the aggregates was unknown, and while acknowledging “the extreme hardness of the aggregates used in the construction of the John Day Dam,” nevertheless, “other than increasing bit consumption and probably higher shock loads the extraction rate should be fairly similar.” This gave some assurance to Gillis and Elmore, who also attended the test, and Granite made no changes in its plans to use the roadheader.

Granite proposed to excavate the concrete tunnel with the Dosco roadheader in two passes from a caisson attached to the south face of the dam. Granite proposed first to cut a rectangular tunnel in the first pass through the service and powerhouse bays and then, in a second pass, cut a lower, wedge-shaped bench that would decline in depth as Granite proceeded away from the south face. Once the tunnel through the bays was completed, the road-header would be used to excavate a portion of monolith 14 leading to the outfall chute.

On January 24, 1984 Granite received from the Corps a notice to proceed. The original date for contract completion was set at October 20, 1985. Because of a labor dispute, the Corps extended the subject contract performance time by one day. No further time extensions were granted. The contract provided that if work was not completed by October 21, 1985, delay damages could be assessed.

The contract also provided at paragraph SP-1.5: “The vertical barrier screens and guides, outfall chute, evaluation facility, and transportation conduit shall be complete and operational by 85 APR 01.” This interim milestone date had no provision for delay damage. Granite’s revised schedule showed that tunneling work, which depended on completion of the caisson, was to begin on June 1,1984 and end in the middle of August that year. Construction of the caisson, which had been anticipated to commence in March, did not commence until April of 1984 and was not completed until June. An additional delay of several weeks was caused by the capsizing of the completed caisson as it was being towed into position. Excavation of the tunnel with the Dosco roadheader did not commence until July 25, 1984.

On May 23, 24 and 25, 1984, as reflected in a June 8, 1984 letter to Gillis, Concrete Technology Laboratories performed compressive strength tests for Granite on concrete core samples obtained from the mass concrete at John Day. The four samples tested reported an average compressive strength of 5,178 psi. In October 1984, the Corps performed similar compressive strength tests on three “four-inch” and seven “eight-inch” cores of concrete removed from the John Day Dam mass concrete. The compressive strength of the three “four-inch” cores averaged 5,010 psi. The compressive strength of the seven “eight-inch” cores averaged 4,587 psi.

From the first instant Granite attempted to use the roadheader to excavate, it was apparent that the machine would work poorly at best. The teeth of the cutterhead were unable to sump into the face of the concrete. Even after a breach was made with drilling, the roadheader was only able to make extremely slow progress. What little progress it made took a tremendous toll on the machine. Granite went through its entire inventory of replaceable teeth for the cutterhead in the first week. The constant pounding and vibration weakened joints, gears, and arms. The machine was almost continually in a state of breakdown and repair.

Through perseverance, ingenuity, and intense labor, the project was eventually completed, but at a cost far in excess of that anticipated. The work was submitted on time, and accepted by the Corps on October 25, 1985. The tunnel and outfall effectively function to pass salmon fingerlings safely through the dam.

The claims which are the subject matter of this litigation were received on August 12, 1985. They were denied by a final decision issued on March 10, 1986. This action was filed on September 26,1986, and was assigned to this judge on April 17, 1991.

DISCUSSION

A. The Superior Knowledge Claim

The legal basis for Granite’s superior knowledge claim is succinctly expressed in J.F. Shea Co. v. United States, 4 Cl.Ct. 46, 53 (1983):

If the government possesses special knowledge which is vital to the performance of a contract but which is unknown and not reasonably available to a bidder who is thereby misled, the government must disclose its superior knowledge or be held liable for breach of contract. Hardeman-Monier-Hutckerson v. United States, 198 Ct.Cl. 472, 487, 458 F.2d 1364, 1370 (1972); Helene Curtis Indus., Inc. v. United States, 160 Ct.Cl. 437, 442, 312 F.2d 774, 777 (1963); Ragonese v. United States, 128 Ct.Cl. 156, 120 F.Supp. 768 (1954). The government is under no duty to disclose information, such as an opinion or conclusion of its geologist, where the knowledge of both the government and the bidder is based on data equally available to both parties and where more conclusive data is available, but neither party chooses to obtain it.

See also Petrochem Servs., Inc. v. United States, 837 F.2d 1076, 1079 (Fed.Cir.1988); McCormick Constr. Co. v. United States, 18 Cl.Ct. 259, 265 (1989), aff'd, 907 F.2d 159 (Fed.Cir.1990).

In the present case, the precise nature of the Government’s knowledge that Granite views as superior is difficult to isolate. Plainly, the relevant knowledge centers on the Little Goose project, but the Government’s knowledge about it was extensive, and Granite does not contend that everything about the project should have been disclosed. During closing argument, counsel suggested that at a minimum, it would have been adequate if the Corps had disclosed that similar work had been attempted at Little Goose, and that a roadheader had difficulties. This argument suggests the following inquiries. What facts did the Government have in its possession after Little Goose? Was comparable information readily available to Granite from a source other than the Corps? What information did Dosco obtain that it did not pass along? Would it have made any difference to Granite if it had in fact had that information prior to bid?

1. What did the Government know?

The Government plainly knew that Little Goose involved a tunnel through concrete. While there is some question as to whether similar work had ever been performed, there was reason for the Corps to realize that the work at Little Goose was one of the very few instances in which such work had been attempted. Granite does not argue, however, that the mere uniqueness of Little Goose is sufficient to meet the legal threshold. The court agrees. Absent more, such information is merely relevant, not necessarily vital, and in any event was probably known or accessible to Granite.

The Corps also knew that a roadheader had been used at Little Goose. Again, the court declines to find, and plaintiff does not contend, that failure to notify John Day bidders of this additional fact is sufficient for liability.

It is only the knowledge gained about the experience of the roadheader in concrete that activates, in Granite’s view, an obligation to disclose. In that regard, the Corps had information that the contractor had serious problems with productivity on Little Goose, and that the Corps attributed those problems, at least in part, to using a smaller machine than the one listed in the bid. One of the Government-hired experts, Terrence McCusker, reported that virtually all material was being removed by milling, rather than by being plucked out or fractured in large pieces.

The Government also had some conflicting data, however. Groves contended (and submitted its own reports) that its problems were due, not to the inadequacies of the roadheader, but to the presence of large quantities of aggregate in excess of three inches in size. The Government also knew that the Board of Contract Appeals judge who heard Groves’ claim was prepared to rule that the problem at Little Goose was the quantity of oversize aggregate, and therefore, by necessary implication, not the roadheader.

Critical pieces of relevant data about Little Goose were thus the subject of controversy. What was uncontroverted was that Groves had used a roadheader, that the roadheader had problems, that the road-header was smaller than other models available, and that the board found that the problems were not due to the fault of the roadheader so much as to the presence of oversize materials. What is missing, however, from plaintiff’s proof is a basis for finding that the Little Goose experience uniquely informed the Corps that roadheaders are inherently unfit for tunneling in concrete. It would have been not at all unreasonable for Corps personnel at Portland, knowledgeable about Little Goose, to conclude that the salient lesson to learn about the first project was that bidders should be warned about the size and quantity of aggregate. It also would not have been unreasonable to conclude that a larger roadheader would have worked better.

2. What did Granite know?'

The next relevant issue is, was information comparable to that allegedly omitted available to plaintiff? Or, to put it differently, what did plaintiff know or have reason to know, at the time of the bid? An important question here concerns whether plaintiff can be charged with the knowledge of, or knowledge imputed to, Dosco. The Government contends that knowledge in Dosco’s possession, or available to it, should be imputed to Granite. Recognizing the significance of that issue, the court first considers what Granite personnel knew or should have known without resort to knowledge not passed along by Dosco.

As to what Granite personnel did not know, the court finds that they did not know or have reason to know about the work at Little Goose. A fortiori they did not know S.J. Groves had difficulty with a roadheader on that project. As to what they did know, Gove and Gillis knew that Granite had never excavated concrete to dimension before. To the extent they were familiar with removing concrete, or cutting very hard material to shape, they knew the standard methodology would be blasting, or drilling and shooting, or use of chemical or hydraulic expanders. They had very little experience with roadheaders, and had certainly never used one to excavate concrete. From their standpoint, therefore, use of a roadheader to excavate concrete would be an experimental application in a novel task. Similarly, the concern of other bidders about the possible difficulty of doing the work with a roadheader was reflected in their calls to Marianski.

Both men knew that roadheaders were used primarily in soft materials. Gove knew from his experience in Montreal that the roadheader used on that job had difficulty cutting through intermittent lenses of harder material. Gillis would not use a roadheader in material with a compressive strength of over 15,000 psi. Both knew that the aggregate would have a compressive strength potentially far higher than that number. Both men also knew that the psi of concrete is not a reflection of the strength of the imbedded aggregate. They knew further that mass concrete uses aggregate of larger size, and uses it in greater proportions than other concrete. Even without knowing about Little Goose, therefore, Granite had every reason to be extremely skeptical of the ability of the roadheader to cut mass concrete.

Granite elected not to do any serious investigation of the capabilities of a road-header in concrete. It chose instead to rely on representations of Dosco. In getting information from Dosco, however, Granite did not tell Dosco that the material to be excavated was concrete. Instead it told Dosco personnel that the machine would be used for cutting in shale, a relatively soft material. The difficulty with that tactic in this case, however, is that it probably prevented Granite from getting the alternative November 18 letter sent to Pacific Ventures and Dillingham Construction. If it had gotten that letter, it would have known that, if the aggregate was of a quartz nature, and densely packed, the roadheader could not sump in under its own power, and that drilling and shooting would be recommended. The aggregate actually encountered was both dense and extremely hard.

The court makes three observations about the information in the November 18 letter. First, Gillis and Gove probably had enough information at hand to draw a similar conclusion. Second, the court is persuaded that the only reason Granite did not get such a letter is that it misinformed Dosco at the outset about the nature of the material. Third, the only reasonable reaction to this information would be to either rule out a roadheader, or make absolutely certain that the disqualifying conditions were not present. In sum, the information in the alternative November 18 letters should be treated as information constructively in Granite’s possession.

3. What did Dosco know?

Marianski had been getting different assessments from other potential bidders about the difficulty of the job. Dosco also had ready access to data from Dosco Overseas to the effect that compressive strength is not a reliable indicator of the difficulty of cutting concrete with a road-header. In addition, Marianski had reports about performance of a roadheader in concrete at Glenn Canyon that should have caused skepticism about the practicality of use of a roadheader in concrete. Marianski was persuaded that the roadheader would only work if the proportion of matrix to aggregate were sufficiently high to permit the cutterhead teeth to function. The court finds that the photographs of John Day core samples should not have allayed those concerns.

These generalized types of concerns about excavating concrete, and the limitations of roadheaders in hard material were not unique to Dosco, however. They were certainly not unique to the Government. They are the result of knowledge about materials and equipment that the Government can assume any competent bidder possesses.

As to Little Goose, however, Dosco personnel clearly knew more than Granite. They were well-acquainted with the fact that Groves had had serious problems at Little Goose using the Alpine roadheader. From this information, and even with an incorrect understanding of the precise roadheader used, McKinlay deduced that the tunnel would be very hard to excavate—an extreme application. He was told that the aggregate at John Day might be harder than that at Little Goose. He knew the production rate achieved was dismal. Although McKinlay was not told the Alpine machine had been modified, he was given access to highly relevant documents that would have revealed that fact, and he declined to look at them.

This latter information was not passed on to Granite. Well after it found out what project Granite was about to bid on, Dosco personnel elected to be less than entirely forthcoming. The time to disclose the problems at Little Goose would have been when the production rate figures were changed.

4. Would the additional information have mattered?

Granite received two additional inputs shortly after bid opening. It was informed, apparently by a Groves representative, about the latter’s difficulties at Little Goose. This prompted the additional inquiries to Dosco, and the test at Grand Junction. Granite also received the letter of December 13 from Dosco. That letter, written after Dosco was informed of the problems at Little Goose, persists in the “up to 40 tons per hour” estimate, but is different from earlier written communications, however, in that it contains caveats. It states that Dosco will not guarantee performance, and that cutter costs could be “extremely high” due to the hardness of the aggregate. The latter conclusion, it points out, was based on tests by Dosco Overseas and also reflects the experience at Little Goose.

Gillis testified that, even if he had received the letter of December 13 prior to bid opening, it would not have affected his thinking, because the substance of the letter had been previously conveyed. Presumably by substance he was not referring to Little Goose, but merely that he knew that the extreme hardness of the aggregate could result in very high cutter costs.

Granite personnel were sufficiently concerned after finding out about Groves’ use of a modified, but lighter weight roadheader, however, that they went back to Dosco for assurances. This resulted in the test on the bridge abutment with a Dosco SL120. This was as close a comparison as Dosco personnel could arrange. The test results gave some assurance to Granite personnel, and they proceeded with the roadheader.

Gillis’ testimony and Granite and Dosco’s actions after bid opening cast unfavorable light on plaintiff’s case. First, it becomes obvious from Gillis’ reaction to the letter that merely knowing the aggregates were extremely hard, that bit consumption could be unusual, and that there had been a “similar project at the Little Goose Dam” would not have daunted Granite. Even if those latter words had been entered into the equation by the Corps prior to bid, therefore, more would have been required to alarm Gillis.

Based on the testing by Dosco on the bridge abutment, it is also difficult to follow Granite’s causation analysis. Granite has to be comfortable with the answer to the question, “What would the company have done if the Corps had stated that a modified F6A roadheader had been employed at Little Goose and encountered problems?” Gillis testified that if he had known the F6A was a modified machine he would have pressed Dosco harder for assurances that the SL120 could do the job. In fact, there is no reason to think Granite would have done anything other than what it did, namely, rely on Dosco. Even after both companies were informed that there had been problems at Little Goose, and that a modified lightweight machine had been used, personnel were satisfied with performance of the SL120 in concrete. A logical inference is that, short of instructing bidders they could not use a roadheader, merely pointing to Groves’ problems at Little Goose would not have deterred Granite.

5. Analysis

Before weighing these preceding findings in light of the legal test to be applied, the issue of imputation of Dosco’s knowledge to Granite must be resolved. The court ruled at trial that it would not find the relation between these two parties to be that of agent and principal. There is no evidence in the record to support such a conclusion. The only contract between the plaintiff and third party defendant was executed after the bid.

The Government instead points to the fact that Granite elected to rely on Dosco to furnish information to it, rather than conducting its own investigation. There is also no question that Dosco had at least some of the information plaintiff says was critical. Defendant offers nothing else. The support it cites for its proposition that “[ujnder the circumstances, where Granite relied upon Dosco in formulating its bid, Granite necessarily must be charged with all of the information and knowledge that Dosco had at the time of its bid and before beginning performance,” is totally inapposite. Fleming v. United States, 648 F.2d 1122, 1127 (7th Cir.1981), states the unremarkable proposition that a client is bound by his attorney’s authorized representations. Plainly there was no attorney-client relationship here. The reference to the Restatement (2d) of Agency, §§ 272, 275 (1958), is to the general proposition that an agent speaks for his principal. This obviously presumes that an agent-principal relationship already exists. What is conspicuously missing, however, is support for the apparently novel proposition that a bidder’s consultations with a potential supplier create such a legal relationship.

Acceptance of the Government’s argument would be unfair to contractors. Clearly, designation of an agent must be done in a conscious manner and in such a way that it binds the putative principal and can be relied on by third parties. The present facts do not permit the finding of such a designation. Instead, the court concludes that the only possible relevance of Dosco’s knowledge would be to the issue of whether the information it had was generally available, and thus not exclusive to the Government. Other than the testimony by Marianski that his conversations with other bidders elevated his concern about the project, the Government made no effort to support such a finding.

It remains then to consider what Granite knew in light of the Government’s duty, under certain circumstances, to share its special knowledge. For Granite, the conclusion is straight forward. It did not know, and had no reason to know, of the prior difficulties encountered by Groves in tunneling through concrete with a road-header; the Government did. This approach, however, proves too simplistic. It does not account for some of the peculiarities in the present facts, which make them a poor fit to the hornbook paradigm.

The implied duty to share special information is discussed by the Supreme Court in United States v. Atlantic Dredging Co., 253 U.S. 1, 11, 40 S.Ct. 423, 425, 64 L.Ed. 735 (1920). There the Government had issued maps of the river bottom to be dredged, reflecting that the material to be excavated was sand and mud. This was accurate as to some, but not all of the test borings. Field notes accompanying test borings were not affirmatively furnished to bidders. They were simply made available. Those notes reflected that test borings encountered cobble and other impenetrable material. Plaintiff was unable to complete the dredging with the equipment it had furnished. The Court found that, although bidders could have gotten access to the field notes, the recitations on the maps lulled them into a reasonable belief that only sand and mud would be encountered. In finding for the contractor, the Court relied on the principle stated in United States v. Spearin, 248 U.S. 132, 136, 39 S.Ct. 59, 61, 63 L.Ed. 166 (1918), and Christie v. United States, 237 U.S. 234, 242, 35 S.Ct. 565, 567, 59 L.Ed. 933 (1915), that a contractor is justified in relying on representations made by the Government in bid solicitations. These representations constitute a warranty, the breach of which is actionable, or they can constitute a defense to failure to complete.

Atlantic Dredging was first cited in the Court of Claims in Ruff v. United States, 96 Ct.Cl. 148, 163 (1942). The facts there also involved the furnishing by the Government of deceptive drawings. The court permitted recovery by the plaintiff, where, in reasonable reliance on a supplier’s investigation, excavation turned out to be more expensive than inadvertently suggested by the Government.

In Potashnick v. United States, 123 Ct.Cl. 197, 105 F.Supp. 837 (1952), the court found that the “untruthful” representations by the Corps of the results of boring tests constituted a misrepresentation of subsurface conditions. Information which was provided was warranted as correct, and was inconsistent with information withheld. The court cites both Atlantic Dredging and Christie.

The Court of Claims in Arcole Midwest Corp. v. United States, 125 Ct.Cl. 818, 113 F.Supp. 278 (1953), extracted from this line of cases the proposition that “where the Government makes positive statements in the specifications or drawings for the guidance of bidders, that a contractor has a right to rely on them regardless of contractual provisions requiring the contractor to make investigations.” Arcole, 125 Ct.Cl. at 822. Similarly, in Railroad Waterproofing Corp. v. United States, 133 Ct.Cl. 911, 915, 137 F.Supp. 713 (1956), the court concluded that the Government must answer in damages whenever it has “misled” one of its contractors, also citing Atlantic Dredging.

A decision frequently cited as more clearly stating the theory of liability in terms of withholding information is the Court of Claims decision in Ragonese v. United States, 128 Ct.Cl. 156, 120 F.Supp. 768 (1954). The introduction to that opinion states that plaintiff was pursuing liability on the alternate theories of a Type I and II differing site condition claim, or “because defendant concealed from it the existence of [excess ground] water and thus prevented plaintiff from including in its bid an amount to take care of it.” Ragonese, 128 Ct.Cl. at 158. In articulating the legal pedigree for a ruling favorable to plaintiff on the latter claim, the court cites Atlantic Dredging to the following effect: “The withholding of this information, especially in view of its representation that it had given bidders the best information available, was a breach of contract.” Id. at 163.

In Ragonese, the Government had contracted with an engineering firm to do soil borings. The work involved laying of sewer lines. The borings made no reference to the water table. The court thus found that the water actually encountered could not be inconsistent with the drawings, and a Type I differing site condition claim could thus not be sustained. As to the Type II claim, the court, without holding that the claim failed, implied that subsurface water cannot be that uncommon. As to the claim of withheld information, however, a different result obtained. The Government knew from the engineering report that water was encountered at levels much above the level of the to-be-placed sewer. After finding that this information would have been highly relevant to plaintiff, the court went on to find:

Defendant not only did not furnish plaintiff with this information, but it said, in effect, that it did not have knowledge of the existence of this water, because it said that the information which it had furnished on subsurface conditions “represented the best information available.”

Id. at 161-62.

It was not enough, however, for plaintiff to show that the information withheld was relevant. It also had to “show that the withholding of this information in fact misled it. If plaintiff should have known that it would encounter large quantities of water, although this information was withheld, then it cannot be said that it was misled.” Id.

What these cases have in common, and what perhaps distinguishes them somewhat from the typical Type I differing site condition claim, is that, while the information in the specifications or other contract documents might be literally accurate, nevertheless, in light of information not furnished, the bidder is left with an inaccurate picture. The information furnished creates a false impression because of information withheld by the Government and not reasonably available to the contractor.

A somewhat different underpinning of liability for failure to disclose arises in subsequent decisions. In Bateson-Stolte, Inc. v. United States, 145 Ct.Cl. 387, 172 F.Supp. 454 (1959), the Court of Claims held that the contractor stated a cause of action based on the Government’s failure to disclose that a concurrent project would have a dramatic effect on minimum wage rates. Although no direct authority was cited there, the court suggests, by referring to United States v. Beuttas, 324 U.S. 768, 772, 65 S.Ct. 1000, 1002, 89 L.Ed. 1354 (1945), that the obligation to share data was in part premised on the implied duty of a contracting party not to interfere with the other party’s ability to perform.

The next significant step in the articulation of a cause of action for breach of duty to share information came in Helene Curtis Industries, Inc. v. United States, 160 CtCl. 437, 312 F.2d 774 (1963). There the plaintiff had contracted to provide disinfectant powder to the Army, based on a formula provided by the latter. It turned out to be very difficult to mass produce the desired product without knowledge of the fact, known to the Government but not shared with plaintiff, that a key ingredient had to be ground to certain dimensions. The Government had obtained its information from its extensive contacts with researchers working on a formula for the disinfectant at the Government’s initiative. The court found that the disinfectant was novel and had never been mass produced; that the Government had sponsored research and knew much more about the formula than its bidders did; and it also knew that bidders would likely be ignorant of the special processing necessary to achieve desirable results. As the court held, “[although it is not a fiduciary toward its contractors, the Government— where the balance of knowledge is so clearly on its side—can no more betray a contractor into a ruinous course of action by silence than by the written or spoken word.” Helene Curtis, 160 Ct.Cl. at 444, 312 F.2d at 778.

Primarily in reliance on Bateson-Stolte and Helene Curtis, the Court of Claims subsequently issued a string of decisions,⁸ the lessons of which are captured in Hardeman-Monier-Hutcherson v. United States, 198 Ct.Cl. 472, 487, 458 F.2d 1364, 1371-72 (1972). Three elements were distilled there to find liability: the Government had information that would be material to performance of the contract; the contractor lacked that knowledge and had no reason to be on notice of it; and the Government either knew or should have known that the plaintiff was in ignorance of the relevant information. The element of misrepresentation, inadvertent or deliberate, is not directly relied upon, but may be implicit in the requirement that the Government know or have reason to know that the contractor is, in effect, stepping into a trap.

Hardeman-Monier-Hutcherson involved construction of a radar tracking station on a remote coast in western Australia. The Navy had in its possession two reports on the severity of the weather and sea conditions at the construction site. The only information included in the bid documents was a comment that the area was subject to periodic hurricanes. In fact, the Navy knew that the site was virtually unusable during certain predictable periods of the year. The reports were not shared with bidders. The court had no difficulty in concluding that the information was relevant, indeed vital, and was unique to the Navy.

It is obvious that the present facts are not within these lines of development in the case law. To the extent that an indicia of liability would be a misrepresentation, arising out of a combination of what was stated and what was not stated, there was, in the specifications, no affirmative indications that the work could be performed with a roadheader. Unlike Helene Curtis, for example, the Government did not indicate how the work should be performed. Or, to make the circumstances here analogous to those in Ragonese, there was no representation that the Government had divulged everything it knew about what techniques not to employ. Unlike the circumstances in Ruff and Potashnick, there were no representations about the concrete or tunneling that would be erroneous in the absence of withheld information.

Granite argues at this point, however, the same facts it advanced in connection with the rejected differing site condition claim, namely that the bid documents suggested softer concrete than that encountered, and created a false impression through the requirement of ceiling bolts in the tunnel. As to the former, tests run after the contract commenced confirmed the accuracy of the strength figures provided. Moreover, Granite’s employees recognized the differences between strength of matrix and strength of aggregate. As to the latter, the technical provisions of the contract merely call for rows of concrete bolts to be placed in the ceiling after tunneling. As the court held during trial in connection with the Type I differing site condition claim, the Corps’ caution cannot be construed as an assurance that the concrete would be unstable and excavation therefore easy. In addition, Gillis visited John Day Dam and walked the galleries. According to Ian McKinlay, who did the same thing, there was no indication the concrete was deteriorated or sloughing. In short, nothing in the Government’s statements created a false sense of security for Granite, and certainly not in the face of its own reasons to be skeptical of a roadheader’s ability to excavate in concrete, an application it knew would be experimental.

Nor is the present set of facts similar to Helene Curtis, or other cases in which the missing information was truly unique to the Government. Unlike Helene Curtis, the Corps had not been involved in research into the methodologies of doing the work for which it contracted, or at least not the one Granite employed. For the same reason, Hardeman-Monier-Hutcherson is distinguishable. There, the Government had in its hands a highly relevant and unique weather report, and, despite a request from the contractor, refused to reveal its contents. Unlike knowledge about the performance of a piece of generally available equipment (as in the present case), the weather information in Hardeman-Monier-Hutcherson was prepared by the Australian government for the United States and concerned a remote stretch of ocean. The very character of the information in that case was such that it would be difficult for a private party to duplicate.

Similarly, in J.A. Jones Constr. Co. v. United States, 182 Ct.Cl. 615, 390 F.2d 886 (1968), and Bateson-Stolte, the information withheld (knowledge of upcoming contracts) was generated by the Government. Here, the Corps had no proprietary interest in and no authorship of the information which Granite says was essential. In no sense did the Government sponsor Groves’ work as an experiment in use of a road-header.

The factor in the present case that creates a colorable argument for Granite is that that precise information the contractor says it needed—knowledge that a road-header selected at Little Goose proved inefficient in mass concrete—was indeed not generally known. Nor is it information that the Corps could expect to be apparent from a site visit at John Day. But it is the very precision with which Granite defines the information it lacked that differentiates this case. So far as the court is aware, in no prior decision has the Government been held liable for failure to disclose the uniqueness of a similar earlier project, or a prior contractor’s lack of success with a particular (and, in effect, experimental) methodology, or for lack of disclosure of relevant, albeit controverted reports. Instead, the clear import of the cases most nearly on point is to the contrary.

American Ship Building Co. v. United States, 228 Ct.Cl. 220, 654 F.2d 75 (1981),⁹ for example, involved construction of a research vessel for the Department of Commerce. The ship was the fourth of its kind, and problem had arisen in connection with construction of previous vessels. There had been cost overruns and substantial delays. In developing its own cost estimate, the Commerce Department had utilized cost data derived from these earlier contracts. It did not take into account cost overrun data, however, because it was thought that overruns were the result of contractor inefficiencies.

The contractor in American Ship Building argued, and the trial judge found, that, under the doctrine of superior knowledge, this information about previous delays, costs and risks of performance should have been disclosed. The Court of Claims reversed, however:

In the instant case the specifications were sufficient to show the plaintiff what was required under the contract. There is no issue of defective or misleading specifications. Also, there was no withholding of a vital fact affecting performance which the government knew plaintiff was unaware of. Since the government was not asked about earlier contracts, it might have assumed plaintiff obtained information elsewhere. Even more important, it is difficult to determine what fact the government should have provided to plaintiff. It would be irresponsible procurement to seek to deter bidders from bidding, or even seem to do so. The evidence shows the government believed reasonably or unreasonably, but in good faith, that the problems in the earlier contracts were largely attributable to contractor inefficiency and inexperience.

Similarly, in the present circumstances, while it would not have been improper to refer to Little Goose, and while imagination and courtesy may have dictated some reference, the failure to do so cannot be considered a breach of contract.¹⁰ To hold otherwise would put the Government in an untenable situation, incompatible with the need to maintain independence between itself and contractors. Here, it is obvious that the only thing that might have deterred plaintiff is knowledge that its road-header would have a very difficult time performing. But based on Little Goose, the Corps could well have concluded that a larger roadheader might succeed where the Alpine failed. It was certainly not the responsibility of the Corps to make the judgment in this instance that roadheaders should be discouraged, particularly in light of its lack of success before the board.

To similar effect is Intercontinental Manufacturing Co. v. United States, 4 Cl.Ct. 591 (1984). The contract in that case involved the manufacture of containers for sea mines. There had been previous contracts for the containers, and prior contractors had encountered difficulties and filed claims. The facts are somewhat distinguishable from the case at bar in one particular—the Government had discussed at a prebid conference some of the prior disputes. Plaintiff contended that the information was insufficient and misleading, however. While finding that the disclosure had not been misleading, the court nonetheless questioned the premise underlying the claim—namely that the Government owed the duty of disclosure in the first instance. It concluded there was no duty:

The present case does not fit with [the] precedents____ [T]he case for imposing upon the Government a duty of disclosure proceeds entirely upon the assumption that since manufacturing difficulties of varying degrees had been experienced by all prior contractors, it could reasonably be anticipated that [plaintiff] too would come to share this plight and that it was the Government’s responsibility to forestall that occurrence.

... [I]t is simply too large an assumption to rest upon. It is important to recognize that with an end-product specification such as is here involved, an imposition upon the Government of a duty of disclosure regarding manufacturing processes and techniques, accomplishes, in practical terms, a reallocation of the performance risks normally shouldered by the fixed-price contractor.

____ [I]t is incumbent upon the aggrieved contractor to explain why ... it would have been beyond its properly expected skills and abilities to have foreseen the manufacturing problems that were encountered and the solutions they demanded. To recognize a duty of disclosure on any lesser basis—to require it, for example, simply because claims arose in past procurements—would carry with it the real possibility of obliging the Government to assume the duty of informing the contractor about what it ought to know.

4 Cl.Ct. at 599.

The lesson to be drawn from Intercontinental and American Ship Building is that, in the absence of affirmative misrepresentations, the Government is not a guarantor against poor judgment with respect to methodologies selected by the contractor. In this case, Gillis’ belief that the aggregate would not play any role was unreasonable. His and Gove’s assumption that the teeth of the cutter head would make constant contact with the low-strength matrix is inexplicable. It was at best a triumph of hope over data. Also, Gillis’ reliance on the requirement to place roof support bolts in the ceiling as an indication of deteriorating concrete, while more reasonable in the abstract, was inconsistent with- a tour of the dam, and was in any event, too slim a reed on which to place the great weight of contra-indications. As this court has held in other circumstances, any “misleading that occurred was due to plaintiff’s own unreasonable assumptions.” McCormick Constr. Co. v. United States, 18 Cl.Ct. 259, 267 (1989), quoting Mojave Enters, v. United States, 3 Cl.Ct. 353, 358 (1983).

B. The Acceleration Claim

The contract completion date, as amended, was October 21, 1985. Liquidated damages of $1750 per day could be assessed if completion went beyond that date. The contract also contained an interim milestone for completion of a functioning outfall tunnel by April 2, 1985. No liquidated damages were spelled out in the contract for failure to meet that date. Nevertheless, the Corps let Granite know in no uncertain terms that failure to complete the substantive parts of the project by that interim date would lead to dire consequences.

At a meeting in October 1984, when it was apparent that the tunnel excavation was well behind schedule, Col. Katin of the Corps told Steven Elmore that no time extensions would be allowed, and that Granite might face damages of $2.1 million per day for failure to meet the interim date. This threat was assertedly justified by costs associated with shutting down John Day Dam. Katin contended that the spring salmon fingerling run would be jeopardized if the turbines were functioning and bypass not completed. Therefore, the Corps anticipated closing down the turbines and paying for alternate energy and associated costs or claims, which it would pass on to Granite.

When Granite completed the tunneling portion of the work, it submitted a revised schedule to the Corps reflecting the time in which tunnel lining and other work would be performed. The schedule reflected lining work well into April, past the interim deadline. The Corps responded by letter dated February 6, stating that the schedule had to be amended to reflect completion by April 2.

Plaintiff responded, stating that it would treat the Corps letter as a demand for it to accelerate and adopt a full-time schedule. In that letter of February 21, plaintiff suggested that a time extension was appropriate in view of two asserted changed conditions, the difficulty of the excavation and excess grouting.

Granite commenced the lining portion of the contract on January 25, 1985. The Government ultimately extended the interim deadline to April 15, the date on which Granite completed the work necessary to meet the interim deadline. In the meantime, however, Granite had gone to a seven-day, 24-hour schedule, with three shifts, which Elmore explained was necessary given the compressed period of time between January 25 and April 2. There was evidence at trial that this compression of scheduling as to the lining and grouting work resulted in loss of efficiency and increased costs. Based on these facts, Granite advances its acceleration claim.

The Court of Claims articulated three elements of an acceleration claim. Granite needs to establish that it was ordered to accelerate, that any delays prompting the order were excusable, and that it in fact accelerated performance and incurred additional costs. Norair Eng’g Corp. v. United States, 229 Ct.Cl. 160,164, 666 F.2d 546, 548 (1981). The difficulty with plaintiffs claim under the present circumstances is that facts that would warrant a finding of an order to accelerate are thoroughly intertwined with delays prior to those on which this count is based. Granite was requesting extensions of time for the excavation work well before it began doing grouting work. It had been denied those requests and told that the April 2 deadline was firm. The court has found that the delays up until the end of January cannot be attributed to the Government. Therefore, the Government’s instructions in October were well founded. Literally applying the Norair analysis would thus be fatal to plaintiff’s claim. Even if the first and third elements are present, the second is not.

That analysis is arguably unfair to Granite under the present peculiar circumstances, however. First, it is plain that the Corps reaffirmed its refusal to move the April 2 date after the completion of the tunnel and before the grouting was completed. It forced Granite to develop a new schedule showing it would finish by that date and virtually ordered it to go on an all-week schedule. Moreover, Granite’s present contention is that there was an independent basis for extending the interim date, namely, a differing site condition as to water flowage. The court is therefore satisfied that there was an order to accelerate. Nor did the defendant seriously contest Granite’s evidence that the compressed schedule for grouting lead to inefficiencies. Accordingly, the only remaining issue is whether Granite is entitled to an additional period of performance based exclusively on problems associated with excess water so that it could have completed performance on time, absent the changed condition?

While the threats regarding dire financial consequences may have been legally without support, Granite cannot argue with the fact that the contract called for interim completion of the tasks at issue by April 2. Even though the Government might not be able to collect liquidated damages under that clause, it is not without import. It plainly permits the Government to insist on performance by that date, absent grounds for an extension of time under the contract. Plaintiff therefore concedes, as it has to, that it must demonstrate that a differing site condition warranted an extension of the interim deadline.

Granite asked for a time extension of 30 days in connection with the additional grouting. The precise amount of the days, as counsel for Granite points out, is not directly relevant, however, so long as plaintiff was entitled to enough of an extension to give it time, after January 25, to complete without the added costs of accelerating.

Granite points to the bid documents to support the argument that the Corps represented that a minor amount of grouting work would be required. The bid workup documents, furnished by the Corps, reflect that 36 gallons of grouting compound would be needed. In fact, plaintiff used nearly twenty times that amount, and the Corps paid it for the additional quantities.

There is no question that there were substantial quantities of water encountered during the tunneling and lining work that added to the effort involved in grouting. Elmore estimated that at its peak, the water discharge out of the tunnel and into the caisson reached 3000 gallons per minute. He estimated that, when he walked the galleries prior to construction, there was a natural runoff of water from between the bays of between 200 and 500 gallons per minute.¹¹ Even assuming that the volumes Granite had to deal with were excessive however, the court concludes that it has not proven a differing site condition claim, and consequently, was not entitled to a relaxation of the interim deadline.

Granite had already lost six weeks in mid summer due to the problems with the caisson. It also knew, from the first or second day of the tunneling work, as Elmore testified, that the roadheader would be a continuous problem. The delay due to the caisson and the roadheader were, in part as Granite would have to concede, and, in part as the court has found, of Granite’s own making. Their net effect, however, was to remove any flexibility Granite had in dealing with water problems.

When Granite punctured the first seam between the bays, the very significant water flow resulted in a shutdown of work, and caused a great deal of consternation. Working with the Corps, and using a grouting expert promptly hired, Granite succeeded in sealing off the seam with about 100 gallons of grout.¹² Corps employees Dennis Hopman and Thomas Johnson testified for the Government. Hopman, who is Chief of the Concrete and Dam Safety Section of the Portland Division, explained that the method used on that first seam is the most efficient way to proceed. Before the seam is encountered (which can be anticipated), a ring of small holes is drilled in a circle in the wall, ceiling, and floor at the furthest point of excavation. Plugs are put into each hole after they breach the seam. Then grout is injected into each hole under conditions in which there is no water flow. He estimated that 24 gallons of grout would be adequate to seal off each seam, if it were applied in this fashion.¹³

Under conditions of water flow, large quantities of grout are wasted. After the effort to seal off the initial seam with drilling and injection, however, Granite elected not to try to use that method in advance of encountering other seams. Instead, the court finds, based on Elmore’s and Johnson’s testimony, that Granite elected to operate the roadheader every minute that it could, even if that meant creating leakage as it progressed. As Elmore explained, it was so difficult to keep the roadheader operating that everything else became secondary. While this was an eminently reasonable approach to avoid further tunneling delays, it meant that grouting was done after the fact of breach. Since grouting was done in flow conditions, enormous quantities of grout were wasted.

Another tunneling choice of Granite’s added to the volume of grout used. As Hopman explained, there were ten seams encountered. If the excavation was done in one pass, there would only be ten occasions for grouting at the seams. Granite elected to do the wedge, or bench, in a second pass, however, thereby disturbing the same ten seams a second time.

In short, the selection of a roadheader, as well as its problems with the caisson, put Granite in an untenable position. It would not have encountered the larger flows of water if it had not, through its own delays, eliminated the option of grouting in no-flow conditions. Although Granite’s efforts after that point were nearly superhuman, the need to do the additional grouting cannot be laid to the Government. Accordingly, the Government’s insistence on the April 2 deadline was not improper.

CONCLUSION

In the final analysis, the fact that there was such a wide gap at the bidding stage between Granite’s expectations and the actual capabilities of the roadheader cannot be legally attributed to the Government. It was this bidding miscalculation that saddled plaintiff with a methodology that made performance nearly impossible. That Granite was able to complete performance in a timely fashion under those circumstances is remarkable, and a tribute to the competence and persistence of its crews, but not compensable. The Clerk is directed to dismiss the complaint. No costs.

1 . 41 U.S.C. §§ 601-613 (1988).

2 . Although never given a precise definition during trial, milling is a description of the effect of the roadheader on concrete. Milling has the effect of grinding both cement matrix and aggregate into powder or small pieces. It is in distinction to breaking out chunks of combined matrix and aggregate when the tensile strength of the material is exceeded, or by cutting more readily through either the matrix or aggregate. The net effect of milling is that the cutting head has to come in contact with virtually every square inch of the concrete.

3 . McCusker prepared one of the expert reports on which the Walla Walla District relied in issuing the final decision on the Little Goose claim.

4 . The roadheader at Ridgeway was not actually used to cut sandstone. The sandstone was in a layer surrounded by much softer mudstone. Once the roadheader gouged out the mudstone, the sandstone would collapse or could be knocked out of the way.

5 . Dosco is a subsidiary of Cancall Co., which in turn is a wholly owned subsidiary of Hawkins Sydney, Canada, Ltd. Dosco Overseas is also a subsidiary of Hawkins Sydney.

6 . Marianski did not consider this ruse to be unusual. Gove testified somewhat to the contrary, saying that it is his practice to furnish the supplier as much information as possible.

7 . Sometime shortly after the bid submission, Dosco received a copy of a report prepared for Dosco Overseas by the University of Newcastle-Upon-Tyne titled "Report To Dosco Overseas Engineering Ltd., On Cuttability Of Concrete Cubes." Although the report was apparently not in Dosco’s hands until then, the court finds that it could have been accessed prior to that time. In any event, there had been telephone conversations with Dosco Overseas prior to the bid, and Dosco could or did find out what its overseas affiliate knew, namely that in assessing the ability of a roadheader to cut concrete, compressive strengths are misleading.

8 . See, e.g., J.A. Jones Constr. Co. v. United States, 182 Ct.Cl. 615, 623, 390 F.2d 886, 890 (1968); Natus Corp. v. United States, 178 Ct.Cl. 1, 13 371 F.2d 450, 458 (1967); Robertson Elec. Co. v. United States, 176 Ct.Cl. 1287, 1295-96 (1966); T.F. Scholes, Inc. v. United States, 174 Ct.Cl. 1215, 1226, 357 F.2d 963, 970 (1966); Benjamin v. United States, 172 Ct.Cl. 118, 134 348 F.2d 502, 514 (1965); Hunt & Willett, Inc. v. United States, 168 Ct.Cl. 256, 265, 351 F.2d 980, 985 (1964).

9 . See also Max Jordan Bauunternehmung v. United States, 10 Cl.Ct. 672, 678-79 (1986), aff'd, 820 F.2d 1208 (Fed.Cir.1987).

10 . Subsequent to closing of the record, Granite filed a motion to reopen trial proceedings to insert a new exhibit. It is an excerpt from a bid solicitation issued on September 30, 1991, by the Seattle District, Corps of Engineers. It calls for replacement of the intake works at Mud Mountain Dam. One of the requirements is apparently the excavation of a tunnel through bedrock. The motion focuses on the following language in the general description of work:

... The tunnel may be excavated at the option of the Contractor by the conventional drill and blast method or by use of a tunnel machine or by both methods. Conditions at the excavation site where present floor level rock is observed shows competent, hard, unweathered closely jambed rock [so] that [it] may not be practical to accomplish tunnel excavation with mechanical means such as [a] roadheader. No claim for additional compensation will be considered if the roadheader and associated equipment fails to economically do the job at the bid price.

Granite argues that the fact that the Corps inserted this language rebuts the suggestion by the Government at trial that it is the policy of the Corps not to make such judgments, and it further demonstrates that it is possible to isolate a lesson from Little Goose in a couple of brief sentences.

This solicitation was issued during trial in the present proceeding. Given the litigation arising out of Little Goose and John Day, it is not surprising that Corps personnel tried to preempt a similar suit from arising there. And the court infers no admission from it.

Even if it were considered on the points plaintiff attempts to make, however, it would not alter the outcome, and for that reason the court denies the motion to reopen. The circumstances probably make the departure from past policy sui generis. More importantly, the fact that the Corps was willing to identify the lesson from Little Goose and John Day does not answer the question of whether it had any legal obligation to do so. As the court discusses in the text, it did not.

11 . Dennis Hopman testified that he saw as much as 5,000 gallons of water leaking from the seams between the bays on some occasions. He explained that there are seasonal variations. During cold weather, the concrete contracts slightly, opening larger gaps.

12 . The ratio of grout compound to grout (which consists primarily of water) was about 1:10.

13 . Granite chose to view the relatively small amount of grout compound listed in the bid documents as an indication of low water flows. Given Dennis Hopman’s explanation for the methodology assumed by that figure (which is described generally at specification 02341 ¶ 4.1.1), 36 gallons of grout compound would not be a clear indication of low flows. In addition, there were contrary indications in the specifications. Section 02314, ¶ 1.5 warns bidders that "[ljarge amounts of leakage should be expected in and around the monolith joints as evidenced by the existing exposed drainage systems in the various galleries.” Even assuming that 36 gallons of grout compound was inconsistent with large volumes of water, it was unreasonable for Granite to rely on that indication in the face of the more explicit warning in the contract. Even if the “200-500 gallons" Elmore saw on his tour of the galleries was more normal than the volumes Hopman referred to, it should have been apparent there would be seasonal fluctuations. In any event, Elmore testified that the term “large amounts” was not consistent with his understanding of the implications of the 36 gallons of grout compound.