20 Cl. Ct. 184 | Ct. Cl. | 1990
OPINION
This action is brought under the Contracts Disputes Act of 1978, 41 U.S.C. §§ 601-13 (1982). The case arises out of a contract awarded by the United States Corps of Engineers, New Orleans District (“Corps” or “Government”) to the A1 Johnson Construction Company and Massman Corporation as a joint venture (“Johnson-Massman”) for the construction of Phase II of the Old River Control Auxiliary Structure in Concordia Parish, Louisiana. Johnson-Massman has filed a complaint alleging a differing site condition and a constructive change to the contract. It seeks recovery of $1,913,008.00, plus interest and costs. Trial was held on the issue of liability only. The court concludes that there was no differing site condition, but that Johnson-Massman is entitled to recover on part of its claim based on a constructive change to the contract.
BACKGROUND FINDINGS
The Mississippi and the Atchafalaya Rivers are connected by what is known as the Old River. Before 1950, part of the Mississippi River flowed unhindered through the Old River, toward the Atchafalaya River to the west, while the remaining flow continued southward toward New Orleans. In the 1950’s, prompted by concern that the Atchafalaya would eventually capture all of the Mississippi River’s flow and deprive New Orleans and other cities of river flow, the Mississippi River Commission implemented a project known as “Old River Control” to regulate the flow from the Mississippi River into the Atchafalaya through the Old River.
The Old River Control project involved the construction of two water control structures at the juncture of the Mississippi and the Old River. The main structure, known as the Low Sill structure, spans the mouth of the Old River where it meets the Mississippi. It is, in essence, a dam. Directly north of Low Sill, and on higher ground, is the overbank structure, which at high river levels controls the degree to which flooding from the Mississippi River’s banks will overflow into the Old River. The Low Sill structure and the overbank structure, together with a system of smaller levies, locks and closures, constitute the Old River Control. They were completed in the late 1960’s.
In 1973 a major flood damaged the Low Sill structure, and the Corps determined that Old River Control needed to be augmented. It decided to construct an additional control structure, now called the Old River Control Auxiliary Structure (“OR-CAS”). The ORCAS project required open cut excavation of a channel, located approximately one mile below Low Sill, that would allow additional (but regulated) flow from the Mississippi River into Old River.
ORCAS was completed in two phases. Phase I, which was not performed by plaintiff, involved preliminary excavation of the future channel using a hydraulic dredge. This dredge was floated onto the site through a channel cut from the Old River. No construction of the actual control structure was begun during Phase I. After Phase I the ORCAS site resembled a wide dirt path from the Mississippi to the Old River. Excavation was deepest at the end of the path nearest Old River, where the actual control structure was to be built. There, in preparation for construction of the control structure, enough material had been removed to create a four to five acre oval cavity. This cavity had become filled with water during Phase I, and was commonly referred to as the “bathtub.”
On July 1,1982, the Corps issued solicitation No. DACW 29-82-B-0072 for Phase II, during which the excavation was to be dewatered and the control structure built. Johnson-Massman was low bidder and was awarded the contract for the construction of Phase II.
One of Johnson-Massman’s obligations under the contract was to design, install, and operate a dewatering and surface water control system to assure that the area was kept dry during construction. Specifically, Johnson-Massman was required under the contract to “[l]ower and maintain the piezometric surface
In the bid documents, the Corps provided information on the ORCAS site which contractors were to use in designing the dewa-tering system according to specifications which were to be incorporated into the contract. In particular, the Corps provided, at section 2N of the specifications, data obtained from a pump test, performed for the Corps by Woodward-Clyde Consultants in 1981. The pump test data was furnished to enable bidders to approximate transmissivity, or “T”, a geotechnical term which describes the amount of water, in gallons per day per foot (GPD/ft) of aquifer,
Johnson-Massman subcontracted with Subgrade Construction Corporation, a sub
To estimate M, the thickness of the aquifer, Hoody began with a portion of Corps Design Memorandum 16 (“DM 16”), which showed a longitudinal profile of the soils along the center line of the excavation (the bathtub). Hoody had obtained this portion of DM-16 from another contractor, Morrison-Knudson, although it was available through the Corps. Hoody examined the profile, which showed the types of soil recovered in the borings, as well as the depths at which these soils were found. Hoody’s task was to determine whether the soils described by these borings were part of the aquifer, that is, whether they would allow free flow of water into the construction site. Using this information he attempted to estimate the depth of the aquifer at each boring. By averaging the aquifer depths indicated by each boring, he arrived at an estimated average aquifer depth.
Hoody established a preliminary average aquifer depth of 116 feet using only DM 16.
In making his determination of aquifer thickness, Hoody excluded as part of the aquifer soils classified as “silty sands” because, in his judgment, they did not bear a significant amount of water. Hoody also testified that the approximately 50 foot deep abandoned channel
In estimating transmissivity, Hoody relied on the data generated for the Corps by Woodward-Clyde’s pump test. That test utilized a single screened, filter-packed pumping well placed in the bathtub area at the end facing the Mississippi. Piezome-ters to measure the level of the phreatic surface were placed on three lines radiating out from this well. These lines, referred to as piezometer lines, formed roughly equivalent 120 degree angles. Pie-zometer line PA radiated westward toward the Old River outlet channel. Line PB ran toward the Mississippi River to the north. Line PC ran to the southeast. The Corps provided a map showing the precise location of each piezometer, as well as a chart
The test well was pumped at five different rates for a total of 5.5 hours, then allowed to recover for 14 hours, then pumped continuously at a rate of 1200 gallons per minute (or “GPM”) over a 48 hour period before being shut off completely. During this 48 hour period, the system had reached steady-state, that is, the piezome-ters readings had stabilized, while the rate of pumping and rate of recharge to the aquifer were equal. Piezometer readings were taken while the system was at steady-state. After the well was shut off, recovery readings were taken at the various piezometers, over several days and at increasing time intervals, to record water levels as the aquifer “recharged.”
The Corps provided in the bid documents a graphic representation of what Hoody described as the “final result of [the pump test] effort.” It consisted of three graphs, one for each piezometer line, showing for each piezometer on that line the water level (or drawdown) at steady-state, and the distance of that piezometer from the test well.
The Aquifer Test Report also set forth values for transmissivity calculated from the steady-state and non steady-state data obtained from the pump test. Five methods were used to obtain the estimates for transmissivity, two using steady-state data, and three which used non steady-state data. The range of transmissivity values given in the Aquifer Test Report was 113,-600 to 149,500 GPD/foot. The transmissivity values calculated with methods utilizing steady-state data were consistently lower than those calculated with methods which utilized non steady-state data.
Hoody obtained the non steady-state data. He did not, however, obtain the entire Aquifer Test Report. A.W. Macomber, Marketing Director for Stang, had requested and received the Aquifer Test Report from the Corps and had told Hoody about the transmissivity values indicated there. Hoody and Macomber determined that it was not necessary to obtain the entire report because, in their view, only transmis-sivity values calculated with steady-state data were reliable, and non steady-state data should only be used when steady-state data is not available. Hoody therefore used only the steady-state graphs originally provided by the Corps in the bid documents to determine transmissivity.
The method Hoody employed to calculate transmissivity is known as Jacob’s straight line analysis.
The second step in Jacob’s Straight Line Analysis involves the formula T = (528 X Q)/dH, where 528 is a constant, Q is the pumping rate of the pump test, and dH is the differential in head calculated from the semi-log paper graph. Using a Q of 1200 GPM (the pumping rate for the Woodward-Clyde pump test), Jacob’s formula gave a value for T of 119,550 GPD/ft. When Hoody was later advised that the actual value of Q was 1150 GPM, he used the corrected value and obtained a T of 114,180 GPD/ft. Having obtained a value for M and T, Hoody solved the equation T = cKM to arrive at a K of .094 ft/minute.
As will be discussed below, the value for K which Hoody thus established is used in another formula to estimate the quantity of water, in GPM, which will be needed to dewater ORCAS under the most severe conditions contemplated by the contract— Mississippi River elevation of 70 feet and Old River outlet channel elevation 61.8 feet. See Contract Section 2N, 116.1. This gallon per minute quantity is often represented in formulae by the symbol Qt (because it is the theoretical quantity of water which will have to be pumped to dewater the site). Once a value for QT is known, the size of the dewatering system can be established. The specific analysis which Hoody employed to calculate Qt with his given value of K is described below.
The dewatering system which Hoody decided to use was a deep-well system with the fully penetrating wells being placed around the excavation, roughly in a circle. This decision impacted on how Hoody calculated Qt. Such a system, for purposes of calculating QT, is treated as behaving in theory like a single, large well, whose diameter is the diameter of the circle formed by the ring of wells. The formula for calculating Qt for such a system is known as the Equivalent Radius Approximation Theory Formula.
One of two versions of the Equivalent Radius Approximation Theory Formula is used, depending on what assumptions a designer makes about the geological characteristics of the particular site. If a designer assumes there is a line source condition, that is, if the designer assumes that the aquifer will be supplied with water (recharged) from a direct source, such as in this case the Mississippi River, the formula is:
[7.5(pi)k2Mh„] Qt = - Ln 2(L + re)/re
In this formula, pi is the mathematical symbol for the relationship between a circle’s diameter and its circumference (pi = circumference/diameter), 7.5 represents the approximate gallons of water in a cubic foot, K is permeability, L is the distance to the line source, re (or radius equivalent) is the radius of the ring of wells, and h0 is the “differential in head,” that is, the difference between the highest river elevation, + 70 feet, and the level to which the water must be drawn down to achieve the required drawdown of 5 feet below the 35.5 foot deep excavation, — 40.5 feet; the value for h0 is therefore 110.5 feet. The symbol Ln is for the natural logarithm. Hoody testified that the value of L, the distance to the Mississippi, is 3,150 feet, and that the value for re is 1,152 feet.
If a designer assumes that there is no line source, but that there is a circular recharge boundary, and that water recharges the aquifer from all directions equally, a slightly different version of the formula is used:
n [7.5(pi)k2Mh0] Wt = - Ln (R + re)/re
In this formula, R, the only new symbol, represents what is called the radius of influence of the system. The radius of influence is the radius of the circle whose center is the center of the ring of wells, and whose edge is formed by the points at which the aquifer is being recharged. Within the radius of influence, there is no
To perform the necessary calculations, Hoody used a computer program which he had developed himself. It operated with a hand-held computer. Hoody testified that where (as here) twice the distance from the center of the excavation to the line source (that is, 2L) exceeds the radius of influence, or R, the formula which assumes a circular recharge would always give the higher value for Qt. His program was therefore designed to default to the circular recharge boundary formula in such an instance. This was considered a conservative precaution. Thus Hoody used the circular recharge formula and, with the help of his program to perform the calculations, arrived at a figure for Qt of 26,701.6 GPM.
Hoody explained that the assumption of a circular recharge condition, despite the existence of the nearby Mississippi River and outlet channel acting as potential line sources, was justified because he expected the phenomenon of plating to occur.
The same computer program which derived a value for Qt also determined the number of 30 inch diameter wells, each with a maximum pumping capacity of approximately 1,150 GPM, which would be required to pump 26,701.6 GPM.
After his preliminary determination, Hoody checked the system's drawdown over the entire excavation by laying out a 25 well array on the contract drawing’s excavation plan. The wells were evenly spaced around the excavation and the drawdown effect of each well was estimated over the uneven excavation floor. Interpolating between each well, an overall drawdown contour was derived, and Hoody determined that the system would satisfy the contract requirements over the entire excavation.
Hoody’s final dewatering plan was submitted to the Corps on or about October 20, 1982. It provided for an array of 25 wells around the excavation, each with a pumping capacity of 1,150 GPM, spaced roughly 272 feet apart. The total capacity of the system, as quoted to the Corps, was 28,750 GPM.
A meeting between Corps personnel and representatives of Johnson-Massman and Stang was scheduled for November 4, 1982 to discuss the dewatering system. Hoody testified that such a meeting is standard
Internal Corps memoranda indicate that before the November 4 meeting to discuss Johnson-Massman’s dewatering system, certain Corps personnel considered Hoody’s design to be inadequate. A correspondence, dated October 29, 1982, from John Grieshaber to Larry Cave of the Corps gives a list of deficiencies, as well as a list of observations and suggestions, regarding plaintiff’s dewatering system. The first paragraph of the deficiencies states: “System (including number of wells and header size) as shown will not handle the flow resulting from the river stages given in the specifications.” A November 1,1982 correspondence also lists system capacity as the first item under the heading “Deficiencies.” Grieshaber admitted during cross examination that prior to the November 4 meeting certain Corps personnel considered plaintiff's dewatering system to be inadequate.
After the November 4 meeting, Stang began installation of the dewatering system as designed. Hoody testified that sometime in January he was notified by A.W. Macomber, Marketing Director for Stang, that the Corps planned to run a capacity test at system startup in February. According to contract section 2N-7, the contractor itself was required to “test and evaluate each component of the system to demonstrate ... that the system is capable of satisfying the [contract] requirements.” Hoody testified, however, that a test run in February, at system startup, would be premature because the excavation would not have been unwatered at that time, and because steady-state would not have been reached. According to Hoody, if the test were run before unwatering, water in the excavation would be supplied to the pumps and the test could not accurately evaluate whether the system could control water supplied only by the aquifer. Hoody stated that he had never experienced anyone evaluating a construction dewatering system in that manner.
Joseph McGowan was Johnson-Mass-man’s Project Manager at ORCAS in January of 1983 when the subject of the Corps’ proposed startup test was raised. Like Hoody, McGowan expressed concern about doing a system check at startup, because the excavation would not be unwatered at that time.
The Corps’ test was run at system start-up on February 3, 1983. Half the wells were started on the first day by turning on every other well, beginning with number one. A total of thirteen wells were made operational on the first day.
In a letter addressed to Johnson-Mass-man dated February 22, 1983, the Corps stated that “based on data obtained during start-up of the dewatering system, the capacity of the system presently installed is not adequate____(This data was corrected for the effect of the existing surface water in the excavation).” No supporting data or other explanation was provided. The letter also requested that Johnson-Massman's “proposed method(s) for increasing the capacity of the system be submitted by COB March 4, 1983.”
Brian Kaub was Project Engineer for Johnson-Massman during this time. He testified that Johnson-Massman was “very confused” by the February 22 letter. According to Kaub, Johnson-Massman could not understand how the Corps could make such a decision at that time, and believed that the decision was premature. Kaub said he felt the Corps was “pushing to get some action.” Douglas Johnson, President of the Heavy Construction Division of A1 Johnson Construction Company, stated, “[b]ased on this letter of February 22, we knew we had to do something.”
Johnson-Massman attempted to accommodate the Corps by proposing the addition of two wells. The proposal involved the drilling of a new well, No. 26, between wells number 25 and 1, and in addition the installation of a pumping assembly in a pre-existing government well located approximately 100 feet east of well number 13. This new well was designated number 27. A new piezometer was also to be installed on the center line of the excavation. Brian Kaub discussed this proposal with George Grate of the Corps by telephone. According to Grate's handwritten notes, dated March 24, 1983, regarding the conversation, “after those changes are made, the system would be re-evaluated. Should the system not be considered adequate (by the contractor), deep well(s) would be installed in the bottom of the excavation.” Brian Kaub testified that George Grate’s initial reaction to the proposal was “good,” although Grate stated that he planned to have discussions with people “higher up.”
On March 26, 1983, Johnson-Massman received a letter from Donald Hull, Chief of the Construction Division of the Corps. That letter stated in part: “We concur with your conclusion that the system capacity needs to be increased and adding the well amd [sic] pump is a step in the right direction but, delaying the analysis of the system until after their installation is not necessary and therefore not acceptable.” (Emphasis in original). In addition, the letter stated that “the completion schedule on this project is very tight and it is imperative that we promptly conclude our deliberations concerning the dewatering sys-tem____” The letter closed by urging Johnson-Massman to analyze the system “immediately upon pool unwatering.”
Johnson telephoned Hull to discuss the situation. According to Johnson, Hull expressed concern over flooding of the Mississippi because of the approaching springtime, and stated that it was the “focal point” of the Corps to get the project completed. Johnson told Hull that Johnson did not have enough information to evaluate the system, and that they wanted to unwa-ter the excavation and install a piezometer in the center for proper evaluation. Hull emphasized that waiting was unacceptable. Johnson felt his only choices were to risk default on the contract or cooperate with the Corps. Johnson and Hull scheduled a meeting between representatives of the Corps, Johnson-Massman, and Stang for April. The purpose of this meeting, in Johnson’s view, “was to find out what they [the Corps] were going to accept.”
On April 7, in preparation for the following day’s meeting, Howard Hoody of Stang took readings from piezometer 13 to determine the water level below the excavation.
To calculate transmissivity based on his April 7 readings, Hoody used the following formula:
T = (528 X 42,210)/110.5
where 42,210 is the amount of water which must be pumped at maximum drawdown, 110.5 is the differential in head at maximum drawdown, and 528 is a constant. This formula is identical to that used to calculate transmissivity earlier, except that the figure in the denominator does not represent a comparison of two piezometric points over a log cycle, but rather is the differential in head at high river stage. Hoody testified that this formula is an acceptable modification of Jacob’s formula. Although he testified that this formula could be found in dewatering texts, he could not name one specifically. With the
At the April 8 meeting, Douglas Johnson explained Hoody’s calculations to the Corps. Johnson-Massman told the Corps that they believed they had encountered a differing site condition, and that an additional 20 wells (not including the 2 already proposed but not installed) would be needed. According to McGowan, someone from the Corps said at that meeting, “you finally got to our number of wells.” Douglas Johnson recalled that Grieshaber agreed with 20 wells as the correct number. According to Douglas Johnson, Grate said at the end of the meeting, “you consider that 20 wells is firm and proceed to do it.”
Stang ordered materials and began installing the additional wells immediately following the April 8 meeting. Including the two wells agreed to earlier, Johnson interpreted the Corps employee’s comments as an order to install, in total, 22 additional wells. On April 8, it filed a notice of intent to file a claim.
McGowan testified that shortly after the April 8 meeting, the water level of piezome-ter 13 began to drop, even as the river rose. Johnson-Massman and Stang personnel began discussing the possibility that only 16 additional wells would be necessary. Hoody explained the reaction of pie-zometer 13 as evidence of the effects of plating having begun. On May 12, Johnson-Massman sent the Corps a letter requesting a reduction from 20 to 16 additional wells. This request was thereafter approved, subject to performance.
Of the 16 additional wells ultimately installed, 9 were placed, by Johnson-Mass-man’s design, on the side of the excavation closest to the Mississippi River. The maximum number of wells operated at any one time during the life of the ORCAS project was 34. All of the wells, however, were used at one time or another.
On July 16,1984, Johnson-Massman submitted its dewatering claim to the Corps seeking an equitable adjustment. Johnson-Massman alleged that the well pump data as contained in the contract specifications was inadequate and defective, and did not accurately set forth the conditions ultimately to be encountered. Johnson-Mass-man also alleged a differing site condition under General Provision (“GP”) 4 of the contract, based on Hoody’s April 7 calculation of transmissivity and the resulting value for permeability. Finally, Johnson-Massman alleged that the Corps had ordered a change to the contract within the meaning of GP-3. Regarding the changes claim, Johnson-Massman urged that according to the contract, the system, once approved, would be altered only if it “proves inadequate.” See Contract § 2N-9. Johnson-Massman argued that the system never proved inadequate, and that the events leading up to and including the April 8 meeting amounted to a change in the procedure for altering the system, a deprivation of Johnson-Massman’s rights to determine the adequacy of the system, and an order for additional work and materials on the project.
Johnson-Massman’s claim was denied on April 29, 1986. On April 27, 1987, Johnson-Massman filed suit in this court requesting a de novo review of its claim.'
DISCUSSION
A. Differing Site Condition.
The Differing Site Condition clause of the Johnson-Massman Contract is found at GP-4. That provision describes two types of differing site conditions, commonly referred to as Types I and II. JohnsonMassman is alleging a Type I differing site condition. It must therefore demonstrate that it encountered at ORCAS “subsurface or latent physical conditions at the site differing materially from those indicated in th[e] contract.” See generally P.J. Maffei Bldg. Wrecking Corp. v. United States, 732 F.2d 913, 916 (Fed.Cir.1984). In this instance, the contract did not make any affirmative representation as to the conditions which would be encountered. Rather,
1. What conditions were indicated in the contract?
According to plaintiff, reasonable manipulation of the data accompanying the bid package indicated that the conditions at ORCAS would be such that at contract high river stage of 70 feet 26,701 GPM would have to be pumped to dewater the excavation. Johnson-Massman anticipated that a 25 well system with a maximum capacity of 27,596 GPM would be adequate to deliver this flow. As discussed earlier, plaintiff’s dewatering design was based in large part on Hoody’s assumption that the phenomenon of plating would occur at OR-CAS. In particular, and most significant, Hoody believed that because of plating neither the Mississippi River nor the Old River outlet channel would be a line source of water recharging the aquifer, and that he could therefore design his system as if there would be a circular recharge boundary at the ORCAS site.
At the outset, the court questions the validity of the plating theory in this context.
In addition to testifying that they had never heard of plating, Grieshaber, Napoli-tano, and Kemp also testified that they were unable to understand how plating could work in theory. Napolitano thought it very unlikely that the relatively insignificant draw of water from the dewatering system could have the dramatic effect of sealing off the entire bottom of the Mississippi River. The Mississippi River is over one-half mile wide at ORCAS. Napolitano testified that as much as 2,100,000 cubic feet of water pass by ORCAS each second. The dewatering system, as designed by Hoody, was to pump at its maximum 27,596 GPM, which translates to 459 gallons per second. According to Napolitano, the likelihood that such a relatively insignificant flow could seal off the Mississippi River “stretches the technical aspects of the problem.”
Kemp questioned the plating theory on different grounds. His difficulty with the plating theory related to the necessary implication that plating only occurs during pumping: “I don’t understand how a period of pumping induces any more seepage into an area than a high water hydrostatic head____” Grieshaber expressed a similar criticism. He stated that the amount of water that is fed into the aquifer by the Mississippi River is so much larger than the 20 to 30 thousand GPM pumped by plaintiff’s dewatering system that “I would have to believe that if in fact plating was going to occur, it would have occurred 7 to 8 hundred years ago____”
The court is persuaded by these general criticisms of Hoody’s plating theory. Moreover, from a strictly empirical point of view there is no clear evidence that plating actually occurred at ORCAS. Hoody testified that plating was demonstrated by the behavior of piezometer 13 after April 7, because even though the river was rising then the water level in piezometer 13 was dropping. As pointed out by Grieshaber at trial, however, the noticeable drop in pie-zometer 13 in April was concurrent with the activation of wells 26 and 27 on April 12. The court also notes that between April 9 and 11 the total number of wells operating increased by a total of 4. The graph at Tab 13 of Plaintiff’s Exhibit 59a appears to show a general correlation between river stage and pumping rates. As river stage increases there is reflected an increase in pumping, while a subsequent decrease in the pumping rates corresponds with a drop in river levels. It is also worth noting that plaintiff decided to add 9 of the 16 wells it ultimately installed along the edge of the excavation facing the Mississippi River. This suggests that even plaintiff ultimately believed there was a line source condition.
Based on the above discussion, the court concludes that plating did not exist at OR-CAS, and that it was not reasonable to
It still remains to be determined what a reasonable interpretation of the information accompanying the bid package indicates about the conditions at ORCAS. The Corps introduced into evidence an analytical tool known as the flow net analysis which is used to predict the amount of water which will be experienced at a particular location when there is a line source condition. The flow net introduced by the Government as Exhibit 26F was performed by the Corps in 1982, before it issued the solicitation for bids, as part of its cost estimate for the ORCAS project. The motivating premise behind a flow net analysis is that the aquifer at the site in question will be recharged by a line source. One of the benefits of such an analysis is that it predicts not only the quantity of water which will be experienced, but also the direction from which it will come. This is particularly important where, as here, there are two potential line sources, the Mississippi River and the Old River outlet channel. Hoody admitted that the flow net is a valid method for estimating the quantity of water which will be needed to dewa-ter a particular area under specified conditions, but stated that the flow net was not applicable here because, due to plating, there was no line source at ORCAS.
A flow net analysis is performed as follows. Using information obtained from a pump test such as that conducted by Woodward-Clyde, a contractor plots on a map of the excavation a series of concentric rings, known as equipotential lines, along which, in the contractor’s estimation, the phreatic surface is of equal depth.
After such a map is constructed, the following formula is used to calculate the number of gallons per minute which would have to be pumped under given conditions to keep the excavation dry:
Q = KDH(Nf/Ne)
In this formula, K is permeability, D is aquifer depth (abbreviated M in Hoody’s formulae) and H is the differential in head. The term Nf/Ne is called the “shape factor.” It is simply the number of flow lines divided by the number of equipoten-tial lines.
The Corps used a value for K of .0984, which it obtained from the Woodward-Clyde Aquifer Test Report, and a value for M of 130 feet, which was determined through its own analysis of the boring data. The Corps determined a shape factor of 3.4. Using these figures, the Government’s flow net analysis indicates that 36,-111 GPM would have to be pumped at maximum river stage to dewater the excavation. A 32 well system would be needed to deliver this much flow. According to the Corps, therefore, reasonable manipulation of the bid information suggested that the conditions encountered at ORCAS would be such that a system generating 36,111 GPM would be necessary to dewater the excavation at high river stage.
The court concludes that the flow net analysis is the most reasonable one here because a line source condition existed at ORCAS. The required pumping rates indicated by the flow nets are therefore a valid measure of what the bid documents indicated about conditions at ORCAS. The court concludes that reasonable manipulation of the contract data should have indicated a required pumping capacity in the low to middle thirty thousand range to dewater ORCAS.
2. What were the conditions at OR-CAS?
Plaintiff bases its differing site condition claim on the value for T of 201,880 GPM which it measured on April 7, 1983. It also points to the corresponding permeability (at plaintiff’s calculated aquifer thickness of 113 feet) of .166 feet per minute. The 201,880 GPM figure for transmissivity is 77% larger than Johnson-Massman’s prebid value for transmissivity of 114,180 GPM. The court is unable to understand why plaintiff bases its claim on a calculation of transmissivity. As Hoody himself admitted, the contractor seeks transmissivity ultimately to determine the quantity of water which must be pumped to dewater a location. Actual pumping volumes and pie-zometric levels would therefore seem the relevant “conditions” to consider. The court also finds it preferable to consider pumping rates as indicative of the actual conditions encountered at ORCAS because the contractor kept daily records of river stage, piezometer levels, and pumping rates throughout construction at ORCAS, and because these records are not disputed. In any event, as discussed below, the court rejects plaintiff’s 201,880 GPD/ft figure for transmissivity.
According to Hoody, the modification of Jacob’s formula which he used to arrive at the April 7 value for T is an acceptable modification of Jacob's formula, and one which has been used by Stang in the past. He therefore believed it accurately suggested the conditions encountered at OR-CAS. Hoody could not, however, identify any external source supporting his formula. When asked if the formula can be found in any recognized text, Hoody responded that it could be, but when asked for a specific source stated, “I don’t know — I can’t name you the text book but if — it’s a development of the single well formula based on K and M.”
John Grieshaber of the Corps, who testified as an expert geotechnical engineer, stated that he did not understand Hoody’s
Testifying as a rebuttal witness for plaintiff, Hoody said that he did use two points, the radius of influence and the equivalent radius. Hoody was asked to elaborate during cross examination. The relevant portions of his testimony are set forth:
Q Where is the radius of influence shown on that [formula]?
A It does not — I stated it does not show on there but it was used.
Q How was it used?
A It — it was used as the log of R + RE — the radius of influence plus the equivalent radius divided by the equivalent radius.
******
Q Okay, but this isn’t ... taken over a log cycle?
A This is taken over the — over the comparison of the radius of influence plus the equivalent radius divided by the equivalent radius. And that’s from the — from the one point to the other for the full distance of — of the 110 and half feet of drawdown. ***** it’s a distance relationship of — of the radius — of the radius between two points. ******
Q But that isn’t a log cycle?
A It’s a distance relationship of — of the radius — of the radius between two points.
Hoody was asked a few moments later if his measurement was taken over a log cycle. He replied, “it was not taken over a log cycle.”
Hoody, explaining the formula further, testified that he determined the radius of influence used in his April 7 calculation by employing Siekart’s formula.
The court does not accept Hoody’s April 7 figure for transmissivity. The formula used to calculate that figure had only three variables, two of which are simple measurements. Hoody’s attempt to support the formula involved a seemingly random sampling of other concepts relating to the various steps he undertook in developing his system. The court does understand how these concepts relate to the April 7 formula,
The court turns to actual pumping experience as the best indication of the conditions encountered at ORCAS. A graph derived from the contractor’s own daily records, submitted as part of the original claim, indicates that after the erratic and unstable behavior experienced during start-up in February, pumping rates at ORCAS during the period from March 1 to the end of July 1983 ranged from a low of 20,500 GPM, at river stage 37 feet, to a high near 32,000 GPM at a river stage of slightly under 63 feet. (The highest river stage reached was approximately 64 feet according to this graph.)
The relevant question becomes, were the conditions encountered at ORCAS, as revealed by the above data, materially different from those indicated in the contract? Reasonable manipulation of the contract data using a flow net analysis should have indicated that a pumping capacity in the low to middle thirty thousand GPM range would be required. During the life of the ORCAS project, 32,202 GPM was the highest pumping volume ever required. This amount was needed when the river was at 47.06 feet. A system capable of pumping in the low to middle thirty thousand GPM range would have been sufficient to handle the highest pumping volume actually encountered. At the highest river stage encountered, 63.09 feet, just under 32,000 GPM was needed. A system capable of pumping in the low to middle thirty thousand GPM range would therefore also have been sufficient to handle the pumping required at the highest river stage experienced. The above analysis suggests that the conditions indicated in the contract did not differ materially from those actually encountered.
The court recognizes, however, that the contract required the dewatering system to keep the excavation dry at river stage 70 feet, and that the river never reached that level. At 63.09 feet, the highest river stage experienced, 30,615 GPM were pumped to dewater the excavation. It might be argued that at river stage 70 feet a system generating in the low to mid thirty thousand GPM range would have been inadequate, and that the conditions at ORCAS therefore did in fact differ materially from those indicated by the contract. The court, admittedly, cannot predict with precision what capacity would have been required to dewater the excavation at river stage 70 feet. It notes, however, that only an additional seven feet of drawdown from the experienced high of 63.09 feet would have been required had the river reached 70 feet. The evidence does not suggest that this additional seven feet of drawdown would have required so much more pumping as to warrant the conclusion that conditions were materially different from those
B. Withheld Information — The Low Sill Report.
Though not argued in their original claim, Johnson-Massman now alleges that the Government deliberately withheld a report associated with the Low Sill Project (“the Low Sill Report”), and that the report contained relevant information which they should have been provided. To recover on a superior knowledge claim, plaintiff must demonstrate that the allegedly withheld information was pertinent. See, e.g., Wm. A. Smith Contracting Co. v. United States, 188 Ct.Cl. 1062, 1086-87, 412 F.2d 1325, 1338 (1969); Leal v. United States, 149 Ct.Cl. 451, 460, 276 F.2d 378, 383 (1960).
Plaintiff makes several arguments that the Low Sill Report is pertinent. Initially, plaintiff characterizes Low Sill as an “adjacent structure,”
Burton Kemp, District Geologist and Chief of the Geology Section in the Corps’ Foundations Materials Branch, Engineering Division, has been involved with the Low Sill project since 1972. Kemp stated that proximity has no inevitable significance regarding the geological similarities between two sites. He testified that there was a dramatic difference in the stratigra-
Plaintiff’s argument that the Low Sill Report is relevant because it would have warned of a higher permeability at ORCAS also fails simply because plaintiff failed to prove that the permeability at ORCAS was in fact higher than that which reasonably could have been calculated from the contract documents. The purported .166 ft./min. figure for permeability at ORCAS which plaintiff, in its pretrial brief, alleges was encountered follows from plaintiffs transmissivity calculation of 201,880 GPD/ft on April 7 (by virtue of the relationship T = cKM). As discussed above, the court does not accept the April 7 tran-smissivity figure as accurate. The court must conclude that the corresponding permeability figure is also invalid, and, accordingly, rejects the argument that non-inclusion of the Low Sill Report prevented contractors from anticipating accurately the permeability encountered at ORCAS.
The court also rejects the contention that the Low Sill Report was relevant because the high permeability of Low Sill would have alerted plaintiff that the abandoned channel was in fact permeable. The permeability of the abandoned channel was to be determined by examining the boring data. Both Woodward-Clyde and the Corps determined aquifer thickness that way (and both were able, using a flow net analysis, to accurately predict the conditions at ORCAS). Admittedly, Woodward-Clyde and the Corps did determine a greater aquifer thickness than did plaintiffs. The Corps, however, attributed its deeper aquifer thickness calculation to the inclusion of silty sands as part of the aquifer. Hoody admitted that the decision to exclude them in his evaluation was a matter of judgment. The difference, therefore, is not attributable to the Corps’ knowledge of permeability, or any other condition, at Low Sill. While the abandoned channel may have in fact been permeable, and while the Low Sill Report may in fact have alerted plaintiff to this possibility, this does not make the Low Sill Report “vital knowledge.” Petrochem Services, Inc. v. United States, 837 F.2d 1076, 1079 (Fed.Cir.1988) (citing American Ship Building Co. v. United States, 228 Ct.Cl. 220, 654 F.2d 75 (1981)).
Plaintiff’s attempt to establish that the Low Sill Report was relevant because Low Sill was within the dewatering system’s 7,250 foot radius of influence also fails. Initially, the 7,250 foot figure for the system’s radius of influence is itself questionable because the accuracy of that figure, according to Hoody, depended on the occurrence of plating. Plaintiff, in any event, did not demonstrate exactly what difference it would have made had it known of Low Sill’s greater permeability. Hoody did not explain in any way how he would have applied the Low Sill data in the formulae he utilized.
In conclusion, plaintiff has not demonstrated that the Low Sill Report was relo-
C. Constructive Change.
Johnson-Massman claims that under the contract’s General Provision 3 (“GP-3”), it is entitled to the cost of purchasing and installing 15 additional wells.
In this case there was no explicit government directive to install a specific number of wells. A direct order, however, is not necessary. See, e.g., J.B. Williams Co. v. United States, 196 Ct.Cl. 491, 450 F.2d 1379 (1971) (constructive change found where Government refused to approve contractor’s work); Space Services of Georgia, Inc., ASBCA No. 25793, 81-2 B.C.A. (CCH) ¶ 15,250 (1981) (government’s course of conduct found to amount to an order to put additional men on project); Triangle Elec. Mfg. Co., ASBCA No. 15995, 74-2 B.C.A. (CCH) ¶ 10,783 (1974) (constructive change found where government misinterpretation of specifications resulted in work beyond the contract). Space Services of Georgia is particularly apposite because the Government was found to have ordered extra manhours through its course of conduct. It issued memoranda marked in red ink to the plaintiff. The Chief of Services also said the supposed undermanning “made him sick” and that he expected the problem to be corrected immediately.
The court concludes that the Corps’ conduct up to and including the April 8 meeting amounted to an order to install at least an additional 20 wells.
The Corps urges that, even if there was an order, the additional wells were
The Corps cited Johnson-Massman to the February test for its decision that the dewatering system was not adequate. Plaintiff argues that the accuracy of the test is questionable because it was performed during a less than smooth system startup, when the system was not at steady-state, and when the bathtub area was still filled with water. The Corps, further, did not provide Johnson-Massman with any concrete data supporting its conclusion. Notwithstanding whether the test was an accurate basis for concluding that the system was inadequate, plaintiff also contends that the Corps’ conducting the February test was a breach of contract. The Corps relied on Section 2N-7 of the contract to justify its performance of the February test. That test requires that "upon installation,” the contractor “shall test and evaluate each component of the system to demonstrate to the satisfaction of the Contracting Officer that the system is capable of satisfying the requirements specified in 2N-6.” (Section 2N-6 gives all the performance specifications.) Johnson-Massman argues that the Corps test amounted to a test for capacity, that Section 2N-7 does not require a test for capacity, and that, in any case, 2N-7 provides that the contractor perform the test described there.
The court agrees with plaintiffs that a test for capacity like the one performed by the Corps is not contemplated by section 2N-7. While that section does require each component of the system be tested to demonstrate that the system can meet performance specifications, the test is to be performed “upon installation,” and therefore is to be performed with respect to a given number of wells, each with a given capacity (in this instance 25 wells capable of generating 1,150 GPM). In terms of capacity, a 2N-7 test will only reveal whether the system is generating the capacity for which it is designed. The question of whether this quantity of water is adequate should have already been evaluated at the time of testing, and in this case it was, by the Corps approval of the system on November 4. The court concludes that the test described in 2N-7 is a component test which requires the contractor to demonstrate that each component is operating properly. The purpose of the test is to determine whether the system can generate the capacity for which it is designed.
Because the February test was not within the Corps’ authority under the contract, and because the test itself was of questionable value, the court concludes that the Corps cannot point to the February test to demonstrate that Johnson-Massman’s de-watering system had proven inadequate within the meaning of the contract.
The court concludes that the order for 20 additional wells was not required under the contract’s “proves inadequate” clause. This, however, does not relieve plaintiff of its contractual duty to dewater the excavation, and plaintiff is not entitled to an equitable adjustment to the extent that any of the 20 wells ordered were needed to adequately dewater the excavation. With respect to such needed wells, the order did not amount to a change in the contract.
Although plaintiff is entitled to an equitable adjustment, it can only recover to the extent that it was damaged by the change. See Design and Production, Inc. v. U.S., 18 Cl.Ct 168, 196 (1989) (court, in discussing equitable adjustment, notes that the work ordered must generate additional expense) (citing United States v. Turner Constr. Co., 819 F.2d 283, 286 (Fed.Cir.1987)). In this regard, the court notes that every one of the wells installed at ORCAS was operated at one time or another during the life of the project. Furthermore, Brian Kaub testified at trial that due to well degradation the efficiency of the wells decreased dramatically over the operation of the system. According to Kaub, rather
CONCLUSION
The conditions encountered at ORCAS did not differ materially from those indicated in the contract. Plaintiff therefore cannot recover on its differing site condition claim. Plaintiff has also failed to demonstrate that the Government withheld information relevant to the design of its de-watering system. The Corps’ actions amounted to an order for 20 wells. The court concludes that 8 of these wells were necessary and would have been installed by plaintiff as part of its original design absent the order. Plaintiff is therefore not entitled to recover an equitable adjustment as to 8 wells. As to the 8 wells for which plaintiff is entitled to an equitable adjustment, there is some evidence that these wells benefited plaintiff. In resolving the issue of damages, such benefit, if any, to plaintiff will be addressed. The parties are directed to prepare a joint status report on or before May 7, 1990, proposing pretrial procedures on the issue of damages.
. Johnson’s interest in the venture was 60%, Massman’s was 40%. Massman, however, was strictly an investor, and Johnson had complete authority over the actual performance of the work.
. The piezometric surface is the upper face of the water table. It is also often referred to as the "phreatic surface."
. The aquifer is the layer or layers of soil through which water will move, either horizontally or vertically.
. Permeability is the ease with which water can travel through a particular soil. Permeability is expressed in feet per minute.
. The relationship between transmissivity, permeability, and aquifer thickness is expressed in the formula T = cKM, where K is the value for permeability, M is the depth of the aquifer, and c is a constant which converts the value KM into the proper units for transmissivity, gallons per day per foot (GPD/ft.). Once values for T and M are obtained, the value for permeability, K, can be obtained by solving the equation T = cKM.
. Hoody testified that he used only DM 16 initially because Stang needed to quickly compile a materials list so that it could get quotations for various components of the system.
. Hoody re-evaluated each boring a second time when he was informed by the Corps that the pumping rate used for the pump test was mistakenly given as 1200 gallons per minute, when it should have been 1150. Hoody did not indicate why a change in the pumping rate used for the pump test would prompt him to double-check boring data.
. The channel was formed when an earlier alignment of the Mississippi River or one of its tributaries filled in. The channel is revealed by uneven stratification of soils, caused by the Mississippi’s continued erosion and deposition of sediments.
. The graphs do not indicate when these readings were taken. However, on each graph is written "Q = 1200 GPM/' and from Hoody's testimony it is clear that these readings were taken during steady state.
. Hoody also referred to his method as straight line distance drawdown analysis. The Aquifer Test Report used straight line distance draw-down as one of the steady state methods for calculating transmissivity. That method yielded transmissivity values ranging from 121,800 to 126,700 GPD/foot.
. The horizontal axis of semi-log paper is not divided into evenly spaced units. Rather, it is divided into "log cycles.” According to Hoody’s testimony, when this semi-log paper is used, a plot of points which on normal paper would reveal a curve reveals a straight line.
. When Hoody plotted this information, it revealed three parallel lines very close together. Hoody testified that this was an indication that there were very uniform aquifer characteristics at the ORCAS site.
. The R for a circular recharging boundary can be calculated with Siekart's formula, R = 214ho X K1/2, where K1/2 indicates the square root of K. This formula gives a value for R here of 7,250 feet.
. The court, using Hoody’s values for K and M, arrived at a figure for Qt of 27,518.91 GPM using the Line Source formula and 27,847.943 GPM using the circular recharge formula.
. Hoody also testified that because of plating he decided not to construct a flow net, an analytical tool which is used to consider dewater-ing characteristics of a site when a line source condition is present.
. The program involved the input of other data not discussed above.
. This figure is higher than the maximum pumping capacity indicated by Hoody’s program.
. According to Corps internal correspondences, John Grieshaber recommended that the test be run after steady state had been achieved. At trial, Grieshaber admitted that his recommendations were not followed.
. Well number 3 could not be turned on. In its stead, well number 2 was activated.
. Piezometer 13 had been installed at elevation 0 on the center line of the excavation, at the end facing toward the Mississippi.
. According to Hoody, the excavation was un-watered at this time. McGowan testified similarly. Grieshaber testified that he remembers some remaining puddles. He also testified that shortly before the April 8 meeting he remembers the water level rising so much as to flood the excavation area. The court rejects Hoody’s calculation on other grounds.
. Hoody arrived at 20 wells as follows. On April 7, the wells were pumping at a rate of 28,813 GPM. Piezometer 13 on April 7 indicated that the water at that location was at —29.85 feet. The river was at elevation 45.5. The differential in head, dH, on this date was therefore 29.85 + 45.5, or 75.35 feet. Hoody divided the 28,813 GPM by 75.35 to determine what he called specific capacity, which is the amount of water, in gallons per minute, which the system must pump to achieve one foot of drawdown. The specific capacity he determined with this method was 382 GPM per foot of drawdown. Using the 382 GPM figure for specific capacity, Hoody determined that the amount of water which would have to be pumped at the maximum differential in head required by the contract, 110.5 feet, was 382 X 110.5, or 42,210 GPM. In Hoody*s judgment, an additional 20 wells would be required to pump this much water.
The court is not clear how Hoody arrives at a figure of 20 additional wells. If 42,210 GPM must be pumped, and if each well pumps approximately 1,150 GPM, it appears that a total of 36 wells would be necessary.
. Information which is referred to in the contract is considered part of the contract documents. See, e.g., Hunt & Willett, Inc. v. United States, 168 Ct.Cl. 256, 351 F.2d 980 (1964); Fel-ton Construction Co., AGBCA No. 406-9, 81-1 B.C.A. (CCH) ¶ 14932 (1981).
. Another aspect of Hoody’s pre-bid analysis which defendant has suggested was flawed is the calculation of aquifer thickness. The Government contends that the aquifer thickness should have been 130 feet, not the 113 feet Hoody used, and that Hoody’s error resulted because he excluded the abandoned channel. Grieshaber also testified that silty sands will yield water, and should have been included when calculating the depth of aquifer. With an aquifer thickness of 130 feet substituted for 113 feet in Hoody’s Equivalent Radius Approximation Theory Formula as applied to a circular recharge boundary, the value for Qt is 30,814.8 GPM. The court recognizes that there is discretion when designing a dewatering system. As will be seen below, however, the court need not consider the contention of an improperly calculated aquifer thickness in resolving plaintiffs differing site condition claim.
. The court notes that Napolitano’s testimony assumes, favorably to plaintiff, that all of the water pumped is supplied by the Mississippi. This is unlikely with the outflow channel’s equal proximity.
. The court understands that it is perhaps the increased draw of water that causes the plating. It is puzzling, therefore, that plating was not reflected in any of Hoody’s mathematical models. If indeed plating is related to pumping volume, it seems only prudent to at least consider the timing and extent of plating caused by the system’s expected draw.
. The designer must extrapolate between pie-zometer points to create the equipotential lines.
. The figure given by the flow net must be multiplied by 7.48 to translate the figure it generates into GPM.
. The flow nets constructed by Woodward-Clyde are contained in a report prepared for the Corps by Woodward-Clyde in 1981 as part of the Corps’ preliminary investigations of the OR-CAS site. See Joint Exhibit 2.
. The Report also suggested that after the first 21 months of the project, the inflow channel which was to be constructed would begin to recharge the aquifer. Based on this assumption, the Report calculated a shape factor of 5.1, and determined a required pumping capacity of 44,800 GPM for the project after the first 21 months.
. As discussed during recitation of the background facts, Hoody’s computer did calculate what the required pumping rate would be if the Mississippi River behaved as single line source. Hoody’s computer indicated that less than 26,-701 GPM would be needed (the exact figure is 26,627.8 GPM), and the computer therefore defaulted to the more conservative 26,701 GPM figure generated when a circular recharge boundary is assumed. No attempt whatsoever was made to consider the effect on dewatering requirements if the outflow channel, which was nearly the same distance from the excavation as the Mississippi River, also acted as a line source. In addition, Hoody never performed a flow net analysis. The court cannot accept plaintiffs cursory, incidental analysis as a reasonable manipulation of contract data, and therefore rejects the 26,627.8 GPM figure.
. This formula was described in the discussion of Hoody’s calculation of pumping capacity using the Equivalent Radius Approximation Formula.
. The court is particularly puzzled by Hoody’s discussion of the radius of influence. Not only is the court unable to understand how it relates to Hoody’s April 7 formula, but Hoody’s description of the radius of influence used for the April 7 calculation appears circular. Hoody testified that he used Siekart’s formula to determine the radius of influence (which in turn is reflected somehow in the April 7 formula), but that the value which he used for K in Siekart’s formula was derived by solving the transmissivity equation, in which he used the value for T obtained from the April 7 formula.
. John Raleigh of Raleigh Services, Inc. testified for plaintiffs as an independent consultant. According to Raleigh, the transmissivity encountered at ORCAS on April 7 was 195,734 GPD/ft. He used the very same modification of Jacob’s formula described by Hoody for his calculation but did not better explain or amplify it. Raleigh is a former Vice President and director of Stang.
. Approximation is necessary because the graph from which this data is drawn is difficult to read with precision.
. Defendant’s Exhibit 28, which shows daily pumping rates, gives an almost identical indication about the range in pumping rates at associated river stages.
. The evidence, in fact, would suggest that a system generating in the low to middle thirty thousand GPM range would have been adequate. On April 7, when attempting to determine the number of wells necessary to dewater ORCAS, Hoody introduced the court to the term "specific capacity." Specific capacity is the amount of water needed to draw the phreatic surface down 1 foot. On April 7, Hoody obtained a figure of 382 GPM for specific capacity by dividing the differential in head present on that day, in feet, by the quantity of water being pumped to maintain it, in gallons per minute. Performing the same calculation here, at river stage 63 feet the differential in head 103.5 feet. (The river is at 63 feet above elevation zero, while the phreatic is, as required by contract, 5 feet below the 35.5 foot deep excavation.) The pumping volume required to achieve that draw-down was 30,615 GPM. This number divided by 103.5 yields a specific capacity of 295 GPM. For an additional 7 feet of drawdown, therefore, an additional capacity of 7 X 295, or 2,065 GPM above the 30,615 GPM, would be necessary. This translates to a total required capacity of 32,680 GPM, which is within the low to mid thirty thousand GPM range indicated by the contract materials.
. The Low Sill structure was approximately four to five thousand feet from ORCAS.
. The relationship between permeability and transmissivity is expressed by the formula T = cKM. Discovering a higher permeability necessarily means a higher transmissivity. The opposite is also true.
. Hoody did not consider the abandoned channel water bearing, and excluded it from the aquifer. If he had considered it part of the aquifer, he would have arrived at a higher average aquifer thickness, M.
. It appears that the effect on ORCAS, if any, of a higher permeability at Low Sill would have appeared in the ORCAS pump test, as that test was designed to determine the conditions at ORCAS. In this regard, Napolitano testified for the Corps that the high transmissivities indicated by the Woodward-Clyde Aquifer Test Report, data which Hoody dismissed as unreliable, should have been a "red flag” that there may be an anomaly in the stratigraphy. This also further justifies the use of a flow net.
. Plaintiff states that it is not entitled to recover for all of the 16 additional wells installed because it admits that it needed 26 wells (at highest river stage of 63.09 feet on May 31, 1983), and that one of the 16 wells was therefore a necessary addition to the system.
. As discussed in the background facts, plaintiff before the April 7 meeting suggested the installation of two wells. It is not necessary to decide whether these were also ordered by the Corps. Only 16 wells were ultimately installed, and, of these, plaintiff asserts that it is only entitled to an equitable adjustment for 15.
. There is evidence that the Corps would have accepted less than the 20 wells proposed at the April 8 meeting. An internal Corps memorandum dated February 18, 1983 states that "[i]t appears the addition of 10-15 wells will be required for the system to meet specifications.” John Grieshaber of the Corps testified that the Corps would have accepted less than 20 wells.
. The court understands that if the February test was a section 2N-7 test, the contractor may have in fact failed it because, as discussed earlier, the system had numerous problems at start-up. The proper response from the Corps under this interpretation, however, should not have been to require more wells, but to require that the contractor correct the malfunctioning.
. This conclusion does not bear on the question of whether the system was in fact adequate. Rather, the Corps cannot rely on the “proves
. The court rejects the argument that the Government warranted plaintiff’s dewatering system when the Contracting Officer (“CO”) approved it. Contract Section 2N-9, which discusses approval, gives no such indication. The CO’s review is "for the purpose of determining the acceptability of the general design concept and layout [and] ... the system’s gross capaci-ty____” Furthermore, acceptance is "subject to performance.”