Seton Construction, Inc. v. United States

OPINION

BRUGGINK, Judge.

This is an action brought pursuant to the Contract Disputes Act of 1978, 41 U.S.C. §§ 601-13 (1988). Plaintiff seeks to recover a portion of the contract balance for certain road construction work. Trial was held in Portland, Oregon on December 8th-10th, 1992. For the reasons that follow, the court concludes that plaintiff is only entitled to partial recovery.

FACTUAL BACKGROUND

Seton Construction, Inc., the nominal plaintiff, brings this action on behalf of its subcontractor, Puget Sound Asphalt Paving, Inc. (“Puget”). The contract at issue involved construction of approximately five miles of road in the Olympic National Forest to be used for logging and recreation. It was executed on July 29, 1987. Puget was the subcontractor for paving. All issues of contract performance relevant to the following discussion relate to dealings between Puget and the Forest Service of the Department of Agriculture, the contracting agency. This was Puget’s first venture into paving for the Forest Service. In the past, its government contracting had been as a supplier of aggregate. Puget is owned and directed by Mike Shaw. He was the primary witness for the company.

The Forest Service ultimately determined that the paving was in “reject status,” because, insofar as relevant here, it did not meet contract specifications for compaction and gradation. Rather than reject the work, however, as it would have had the right to do under the contract if in fact the paving was unsatisfactory, the Forest Service devised a pay scheme that netted Puget 64% of the contracted-for amount. Puget’s argument in this action is that it is due all or a part of the balance withheld because both compaction and gradation were within specification. The primary issue is one of contract interpretation, informed by the parties’ experts.

Only a few contract specifications bear on the outcome. They come primarily from specifications specially adapted for the contract. Section 401 was entitled, “Plant Mix Bituminous Pavement — General.” Section 401.02 required Puget to submit a job-mix formula for the Forest Service’s approval before it could begin paving. This formula reflected Puget’s proposal for the physical characteristics of the asphalt mix in such terms as gradation of aggregate, density, percent bitumen, and voids in mineral aggregate (“VMA”). The gradation of the aggregate was indicated by the amount of rock passing through sieves of various size. Although parts of § 401.02 were changed by the special project specifications, the portion dealing with the job mix design were the same as were contained in the standard project specifications. That part carried forward from the standard project specifications referred to the nine components of mix design as “target values” for the mix.

Puget submitted three mix designs, with accompanying test data, before its proposal was finally accepted. The first design had insufficient information. The second was rejected for several reasons, including a VMA value that was too low. Puget submitted a third mix design, which included a change in the percentages of aggregate passing certain sieve sizes. It also utilized a different method for measuring specific gravity, which resulted in a higher value for VMA. This design was approved by Gordon Hanek, the Forest Service Project Engineer.

Work Order “U”, in which the approval was relayed, reflected nine “target values” corresponding to the nine elements of mix design from § 401.02. Special project specification § 102 defined “target value” as “[vjalues which are established according to contract specifications and from which any deviations are measured.” ¹ The work order itself also contains the following note: “The Contractor must understand that his performance is measured against these Targets, and significant price adjustments are made if the materials are not within the allowable deviation from the Targets. Disputes regarding the Target Values must be resolved prior to paving ...” (Emphasis in original.)

Work Order U adopted as the target value for the amount of aggregate passing through various sieve sizes the mix characteristics proposed by Puget. Each of the target values had a certain allowable deviation. With respect to the amount of aggregate passing the quarter inch sieve, for example, the target value was 60%, with a deviation of ±7%. The permissible range for percentage aggregate passing the quarter inch sieve was thus 53% to 67%.

Compaction is a measure of the density of the mix as laid on the road. It is expressed as a percentage. The higher the percentage, the more compacted the road. With respect to compaction, Work Order U reflected the following: “(i) Maximum Specific Gravity [measured] by AASHTO T 209.... Density: 154.2 pcf (100%) 141.9 pcf (92%).” AASHTO refers to the American Association of State Highway and Transportation Officials. That organization has established various test procedures that have become standards in the road and bridge construction industry. The figure generated by the procedure in T 209 reflects the specific gravity of the sample tested by using a method to measure only the solid material, eliminating any voids. It thus reflects an absolute density with 0% voids. This method yields a figure known as Rice Specific Gravity. As a practical matter, normal commercial road construction methods would not achieve 100% of this absolute density. In fact, a road compacted to that extent would be of poor quality because it would tend to rut. Hence the possibility of achieving a 1.0 or even 1.05 pay factor by only achieving a portion of this density.

Special section 105.033 set up an elaborate matrix .by which to measure the amount of pay Puget would earn in terms of a set of pay factors. Each pay factor reflected Puget’s performance against the target values for such things as compaction and gradation. This use of the term target value was peculiar’ to the special project specifications. It was possible to achieve a pay factor as low as 75% and still be in “accept” status with respect to a particular target value. It was also possible for a contractor to be paid as much as 105% of the contract amount. Puget’s ultimate pay would be based on the lowest acceptable pay factor. The risk to Puget, however, was that it had to be minimally acceptable on all criteria. Thus if it failed any parameter, theoretically the work was unacceptable. Since gradation across each of the eight sieve screen sizes was a separate basis for rejecting the paving, Puget was faced with multiple possibilities for rejection.

During the first season, which began for Puget in October, 1987, it paved approximately 1000 feet. After repair of certain isolated defects, the Service made a temporary partial payment based on a 79% pay factor. In reducing the interim payment, the Service determined the percentage compaction by comparing the density of the road as laid to the AASHTO T 209 figure for maximum density which Puget supplied as part of its mix design. When converted into pounds per cubic foot (“pcf”), the figure used by the government as a target was 154.2 pcf. Using this figure, the Forest Service calculated a pay factor of 79% with respect to compaction.

Throughout the period of paving, Puget was responsible for its own quality testing. It hired for that purpose Polaris Testing, which, among other things, tested the density of pavement through use of a nuclear gauge. Under § 401.22 the Forest Service could have elected to employ this same method for acceptance testing. The gauge requires a target value to be programmed in, however, and when Polaris asked the Forest Service for a number, it was furnished the higher Rice Specific Gravity figure. Although it is not entirely clear from the testimony, apparently Polaris’ test results reflected that Puget was not consistently achieving 92% compaction when measured against Rice Specific Gravity. Shaw referred to figures being both above and below the 92% lower control limit. He was not unduly concerned, however, because of his belief that acceptance testing would be done in accordance with a different procedure, set out in AASHTO T 230, and because of his understanding “from the people there” that the values obtained using AASHTO T 230 were normally higher than nuclear gauge figures.

Mike Shaw complained at the time to Don Markham, the Contracting Officer’s Representative, that the figure for density used by the Forest Service to measure the first season’s work was too high. He believed that use of the Marshall Method of measuring density, set out in AASHTO 166, would have yielded a lower target of approximately 147.8 pcf.

At the beginning of the second construction, season Shaw spoke with Service representatives and asked if he could utilize a different mix design. He was concerned with whether he could achieve the target figure for gradation with respect to lk inch aggregate. He wanted to raise from 60% to 65% the target value for the amount of aggregate passing the lk inch sieve. This would have yielded an acceptance range of 58% to 72%. He was told to submit the request in writing, as required by section § 401.22 of the specifications, but never did submit an adequate change request.

The Forest Service tested the road for final acceptance in September, 1988. It concluded that the work was in reject status because it did not fall within acceptable levels with respect to either compaction or gradation. Use of the pay chart was thus inappropriate from the Service's perspective. Rather than reject the work altogether, however, the Service, in conjunction with an outside consultant, attempted to determine what the projected life expectancy of the road would be compared with its intended useful life of twenty years. A pay factor of 64% was ultimately derived.

With respect to gradation, the Forest Service took thirteen core samples of freshly laid pavement. After use of solvents to remove the bitumen, the remaining aggregate was sized. The 69% figure derived with respect to aggregate passing the lk sieve was beyond the permissible range of 53% to 67%.² Puget challenges the accuracy of the test results for gradation, and the enforceability of the target values.

As to density, the Service test results found that compaction achieved, 141.6 pcf, was less than 92% of the target value used by the Service, 154.2 pcf. As to this figure, Puget contends that the Service’s interpretation of the contract is improper; that the 92% figure should have been applied against a lower target value obtained using the Marshall Method.

DISCUSSION

Gradation

Final acceptance testing by and for the Forest Service on core samples from the road showed that Puget’s work was out of specification with respect to the amount of lk inch aggregate. The maximum allowable under the approved mix was 67%. The Forest Service measured an average of 68.7% of the aggregate passing the lk inch sieve. A subsequent test by the Washington State Department of Transportation Labs showed 68.4% passing the same sieve. Puget has two arguments with respect to why this element of acceptance testing should either be ignored or adjusted. The first is that Gordon Hanek should have permitted a change in target values during contract performance. Shaw wanted to raise the figure with respect to lk inch sieve gradation from 60% to 65%. This would have generated an acceptable range of 58% to 72%. Hanek was under no obligation to approve such a change, however. Specification 401.02 called for the contractor to submit proposed job mix-formulas in writing at least five days prior to the start of mixing operations. When the subcontractor orally proposed a change in the formula, he was told a few days later that the proposal had to be in writing. At that point Shaw pursued the subject no further with the Forest Service.³

Puget’s second argument is more substantial. It contends that the gradation achieved with respect to lk inch aggregate was in fact less than 67%. Puget put on evidence indicating that the process of testing the core samples had the effect of breaking down the average size of the aggregate and thus gave the false impression that it was smaller than it actually was. Puget knew that its cold feed (before introduction of heat and bitumen) was within specification, albeit at the high end. Cold feed averages from July 17 to August 18, 1988, in the midst of construction, were 66.2% for lk inch aggregate.⁴ It also knew that the process of mixing aggregate with bitumen eliminated through air pollution equipment approximately one and a half percent of # 40 and smaller aggregate. Puget thus calculated the actual amount of Vi or less going on the road to be 65.69%.

Puget put on evidence of tests run by General Testing Lab and the Washington State Department of Transportation reflecting a two to three percent difference in the percentages of xk inch aggregate passing the sieve before and after mixing, compacting, coring and testing. Puget speculated that the difference could be due to a breakdown in the aggregate as the core is drilled and solvents are used to separate out the bitumen. Gordon Hanek conceded that there is a breakdown in the aggregate during testing. He suggested that the fifteen minutes of shaking over a sieve might account for the degradation, as well as the alternate heating and drying that is required in testing.

The Forest Service offers the argument that the subcontractor has only itself to blame for not discovering that final acceptance testing would be problematic, since it only used cold feed sampling to assess the character of the aggregate. It does not challenge Puget’s evidence that there is in fact a certain amount of degradation that takes place. What the Forest Service argument reduces to, is that Puget did not discover that the acceptance tests were inaccurate until after performance. If it had discovered that earlier, it could have compensated by making the mix coarser. This may be correct,⁵ but it begs the question of whether Puget was obligated to make such an adjustment, if in fact the pavement was within specification.

The Forest Service test results reflect only the aggregate remaining after breaking apart and dissolving the sample core. The government has offered no justification for applying the target values to anything other than that which is on the road. Nor would the evidence support a finding that the parties had a tacit understanding when the target values were developed that the actual characteristics of the pavement in place would be different. The court concludes therefore that the relevant question is whether the pavement in place met the target specifications. In that regard, the cold feed test results provide the most accurate information available with respect to the character of the aggregate actually incorporated into the paving. They reflect for a relevant period that the amount passing the xk inch sieve was no higher than 66.2%.⁶ This was within the upper range of the specifications,⁷ and would have generated a pay factor of 81%.

Compaction

The parties’ competing positions turn on the meaning of special provision 401.22. It is quoted in relevant part below:

401.22 Acceptance Sampling and Testing of Bituminous Mixtures (Compaction).

After the bituminous mixture has been placed and compacted, the lot will be accepted with respect to compaction using statistical evaluation procedures in accordance with subsection 105.03. The pay factor for the lot will be determined in accordance with subsection 105.03. The following tolerance, also referred to as the Lower Control Limit, will be allowed from the target value.

(a) For ... dense graded hot mix at least 92 percent compaction.

Samples, when required, will be taken by the Contractor at randomly selected locations. Tests will be performed by the Engineer. Testing will be in accordance with AASHTO T 230 or ASTM D 2950 or other approved methods. The specific gravity of standard specimens shall be determined in accordance with AASHTO T 166 METHOD A.

(b) If directed by the Engineer, a control strip may be used to establish a Target Value density.

(Emphasis added by the court.)

Puget contends that the reference to a Target Value in the opening paragraph refers to a figure generated as part of a testing process under AASHTO T 230. This latter provision, in turn, explains that standard specimens of the mix are prepared and tested in accordance with AASHTO T 167 and T 166. AASHTO 167 lays out a procedure whereby loose mix is compacted in a cylinder using a certain number of blows of a piston of specified weight. This simulates a process of ideal compaction comparable to what might be achieved using road equipment. The specific gravity of this sample is then measured using a process of immersion and weighing set out in T 166. This technique is known as the Marshall Method. The density obtained for the standard specimen made from the mix becomes a standard against which to measure a core sample drawn from the pavement. The specific gravity of pavement specimens is determined, per AASHTO T 230, according to AASHTO 166. The results of using T 230 is a figure, expressed in percentage terms, reflecting the degree of compaction of the pavement specimen as against the standard specimen.

From Puget’s perspective, therefore, § 401.22 sets out a self-contained method for deriving a target value for compaction.⁸ The target value is determined at the time of acceptance testing by use of the Marshall Method. The significance of this is that the standard specimen tested under this method has approximately four percent air voids. It is thus easier for a paving contractor to achieve 92% compaction of a Marshall Method figure than it is to achieve the same degree of compaction with respect to the Rice Specific Gravity, which assumes 0% air voids.

Since there were no standard specimens made from the hot mix pursuant to AASHTO 167 at the time of acceptance testing, Puget used as a target value the Marshall Method figure it submitted as part of the original mix design, 147.8 pcf. Under Puget’s interpretation, it achieved a pay factor for compaction of 105%.

Puget’s construction of the contract specifications rests solely on § 401.22. It contends that the language of that section strongly suggests that the target value for compaction is determined by the Marshall Method, in. accordance with AASHTO T 230. When § 401.22 is read without considering § 401.02, Puget is plainly correct. “Testing ... in accordance with AASHTO T 230” is unambiguous. The government’s suggestion that the only part of the procedure set out in T 230 which should be utilized is that which controls obtaining test samples is indefensible, at least when only § 401.22 is considered. Obviously more is involved in testing than obtaining a specimen of the material to be tested. T 230 itself draws a distinction between sampling, to which the government wants to limit the utility of the provision, and testing. Testing according to T 230 would direct the contractor to the Marshall Method, set out in AASHTO T 166. As Puget’s witnesses testified without contradiction, the Marshall Method generates a value for density from a standard specimen made from the loose mix against which the core sample is judged. No standard specimen was developed in the Forest Service’s acceptance testing.

The error of the Forest Service’s interpretation of § 401.22 is highlighted by the meaning Hanek assigned to the term “standard specimen” as used in that provision. Section 401.22 states, “The specific gravity of standard specimens shall be determined in accordance with AASHTO T 166 METHOD A.” Hanek suggested that the reference here is to core samples drawn from the completed pavement. It is clear from T 230 lift 3.4 and 3.5, however, that “standard specimens” refers only to laboratory compacted samples, not to core samples. T 230 consistently and unambiguously refers to the core samples drawn from the pavement as “pavement specimens.” Defendant’s reading of § 401.22 therefore assigns a meaning to the term “standard specimens” which is completely at odds with AASHTO T 230. Defendant’s construction requires that the court treat the phrase “testing” in § 401.22 to mean sampling, and the phrase “standard specimen” to mean the core sample. AASHTO T 230 is titled “Determining Degree of Pavement Compaction of Bituminous Aggregate Mixtures.” Gordon Hanek conceded that in no sense did the Service’s use of that testing method yield a determination of pavement compaction.

The government offers two challenges to Puget’s interpretation. The first is that § 401.02 and Work Order U, as set out above, establish Rice Specific Gravity target value for compaction. Although the work order was admittedly a post-execution contract document, the figures it generated were called for within the advertised specifications. Specifically, § 401.02(i) requires the contractor to propose, inter alia, as a “definite single value[ ] ... [f]or dense graded hot mix, the maximum specific gravity of bituminous paving mixtures as determined by AASHTO T 209. It shall be the maximum density for the target value bituminous content.” By overlaying the § 401.22 statement that a lower control limit of 92% will allowed “from the target value,” what is produced for purposes of § 401.22, according to the government, is the higher, absolute density target value of 154.2 pcf.

The Rice Specific Gravity target value was called for in the specifications with respect to the project mix proposal. Puget submitted this figure, and it became a target value identified in the work order. Use of that higher number, however, is inconsistent with the literal phrasing of § 401.-22. That section calls for acceptance testing by use of AASHTO T 230, which in turn refers to AASHTO T 166 and T 167. The Marshall Method incorporated in those provisions assumes the creation and testing of a standard specimen against which the road work is measured. Testing of that standard specimen creates a new target value against which to assess the simultaneously-measured pavement cores. There is no question that if that method had been used to create the target value, the road work would have met or exceeded a 92% lower limit.

Was the ambiguity such that it should have been apparent to a typical road contractor or subcontractor? The government expert, R.G. Hicks, believes so. In his opinion, the use of a 92% lower limit is incompatible with a target value generated by use of the Marshall Method. The resulting pavement would have upwards of 10% air voids, and in Hicks’ view, this is sufficiently out of the ordinary that a reasonable contractor would know immediately that the lower limit would apply against the Rice Specific Gravity figure. Applying it to the Marshall Method target value would, in his opinion, produce an inferior road.⁹

According to Shaw, there was no ambiguity in the specifications. In his view, the target value for density generated pursuant to § 401.02 would be used by the government to “monitor the weight of the rock and the oil to see that it didn’t vary.” In other words, it had a use independent of acceptance testing. This statement, which was unelaborated on direct examination and unchallenged on cross-examination, leaves much to be desired as an explanation. Somewhat more satisfactory was the explanation of Robert Dunning, Puget’s expert witness. In his view, the figure generated in response to § 401.02(i) can only be a target value for maximum specific gravity. It could not meaningfully serve as a target value, in the sense of being an optimum, for compaction, because pavement which achieved that level of compaction would be “mush.”¹⁰ It would be too dense. The utility of the Rice Specific Gravity figure, therefore, is in doing other calculations in designing the job mix formula. In order to know whether the field compacted specimen actually achieves a desired four percent voids, it is necessary to know the specific gravity of a sample with no voids.

Thus, while the only use of maximum specific gravity as a target value during the contract was for acceptance testing, it was established that it is a figure that must be known in order to calculate VMA. It thus had some utility in evaluating the mix design.

More importantly, whatever ambiguity is required by Puget’s construction was present even before the special specifications were drafted. Gordon Hanek testified that the only purpose for the information called for by special § 401.02(i) was to establish the target value for compaction. Although the figure would have other uses in connection with creating the design formula, “you would not need to list it as a target value in item (i) in order to accomplish that. It would be redundant, no need.” Transcript at 375. What this testimony fails to account for is that § 401.02(i) of the standard specifications already in existence also referred to the Rice Specific Gravity as a “target value.” That figure was not used as a target value for acceptance testing, however, since testing in the default version of § 401.22 was by reference to AASHTO T 230 without mention of “target values.” This is the same figure which the Service now insists is the only possible target value referred to in § 401.22. Hanek’s criticism of Puget’s interpretation would have applied equally to the standard specifications.

The views of the personnel of each party as to the level of performance required of the subcontractor are obviously sincerely held. The court is not left with the impression after listening to the witnesses either that Forest Service personnel were setting a trap for the unwary, or that Shaw was knowingly gambling. If Hanek is representative of Forest Service personnel, the agency expected to get pavement that had no more than eight percent voids in it. On the other hand, Shaw’s testimony that he assumed the target value referred to in § 401.22 would be developed by the Marshall Method is credible.

Whether Shaw’s was a reasonable belief depends in part on the weight to be given the testimony of Dr. Hicks that 92% of a Marshall Method figure would produce an unusable road. Ninety two percent of a figure that already has three or four percent air voids would yield a road with eleven or twelve percent voids. There are a number of reasons that the court finds Dr. Hick’s testimony unpersuasive. The first arises out of a serious inconsistency between the standard specifications and his testimony. The standard language of § 401.22 used contemporaneously with the contract period did not contain the term “target value.” Acceptance testing for compaction was simply “92 percent for dense graded hot mix;” testing to be “in accordance with AASHTO T 230.” The default version of § 401.22 therefore pointed to the use of a “target value” to be generated by the Marshall Method. As Dr. Hicks pointed out, this value for density assumes a standard sample with approximately four percent voids. Ninety two percent of a Marshall density figure is therefore plainly contemplated by the standard language in use at the time.¹¹

In a written report he prepared for the Forest Service concerning Puget’s claim, even Dr. Hicks was apparently confused about how to piece together the standard and special specifications.¹² He also conceded that “AASHTO T 166 is the bulk density of a compacted mix and comparison to either T 209 or a mix design is needed. [I.e. to accomplish the Forest Service’s interpretation.] This is in conflict with section 401.02 of the contract specifications and could, in my opinion, cause problems in court.” In explaining this during trial, Dr. Hicks testified that the confusion was limited to special § 401.22. He found that the confusion lifted, however, when Work Order U was considered. The obvious difficulty with this explanation is that the subcontractor did not have the benefit of Work Order U at the time of the bid. It was a post contract formation document.

Dr. Hicks further added to the difficulty with the Forest Service’s reading of § 401.-22, moreover, by stating that its preferred interpretation of the term “target value” only applies with respect to dense graded hot mix. Section 401.22 refers to the Marshall Method with respect to cold mixes and open graded hot mixes: “[Section 401.22] says for cold mixes and open graded hot mix, at least 95 percent compaction, and that would ... be appropriate for the Marshall method that we described, and the 92 percent is from the target value that was specified in the job mix formula, which is 92 percent of T 209.” In other words, the phrase has a fluctuating meaning even within § 401.22. The contractor is supposed to know simply by the percentage picked whether the reference is to T 209 or T 230. Presumably if the percentage with respect to cold mix had been 92%, the contractor should have known it was directed to T 209.

A contractor cannot be expected to engage in that degree of mind reading. Puget’s interpretation merely creates latent ambiguities. The Forest Service’s interpretation creates multiple inconsistencies. While the court is left with impression that the exercise of a larger degree of imagination either by Puget personnel or Forest Service personnel would have avoided the present lawsuit, the government was responsible for drafting the specifications, and has to assume greater responsibility when they do not literally read as intended.

The court will use Puget’s interpretation of §§ 401.02 and 401.22. The former section did not create target values with respect to density acceptance testing, and the latter section directed the Service to use all relevant aspects of AASHTO T 230. Puget was therefore within specifications on compaction.

CONCLUSION

The road was not in reject status. Because the lowest pay factor was 81%, however, that is the limit of Puget’s recovery. It is entitled to the difference between what it recovered, 64%, and 81%. This amounts to $44,652.73. The Clerk is directed to enter judgment for plaintiff in that amount, plus interest pursuant to 41 U.S.C. § 611 from January 6, 1989.

1 . Under the standard project specifications in use at the time, the use of the term "target value" was limited to the nine elements of the job mix design: the recipe for the pavement. The term was not used with respect to acceptance testing for compaction or gradation.

2 . Separate testing by the Washington Department of Transportation on 14 samples also showed the V* sieve aggregate in reject status, but determined that #200 aggregate was also slightly above the limit.

3 . Donald Markham, the Contracting Officer’s Representative, received on August 1 or 2 a copy of a letter Shaw had written to the prime contractor requesting a change in the target value as to 'A sieve. The letter did not satisfy the requirements of § 401.02 with respect to proposing a new mix design. It came into Markham’s possession after half the paving was complete.

4 . Puget was responsible for its own quality control testing. In terms of aggregate size, it was permitted to use either hot extraction, or cold feed sampling. It elected to use the latter because of what it viewed as inaccurate and inconsistent results it obtained as to separate but related tests it ran on the hot extraction samples with respect to the percentage of retained asphalt.

5 . Puget inaccurately failed to give itself the full benefit of the range of tolerance at the lower end of the aggregate sieves, because of its incorrect understanding as to the application of ranges for the middle target figure. Puget thought that the results had to fall within the range of tolerances permitted by the Washington State Department of Transportation. In point of fact, Puget had to select target values from within this range but was then allowed a deviation from target values which could carry them well outside the range acceptable to Washington State. This misunderstanding may have led to Puget’s difficulty in developing a project design mix.

6 . The court declines to use Puget’s figure of 65.69%. While some fines may have been lost due to pollution control, a certain amount would have been generated by mixing and drying, which took place after the cold feed sampling.

7 . Applying a comparable reduction to other sieve sizes would not put the subcontractor out of specification.

8 . Puget's expert, Robert Dunning, testified that the purpose of T 230 is not to create target values. In the process of determining the degree of compaction of the core sample, however, it uses a figure for the degree of compaction of the standard specimen, against which the core sample is measured.

9 . Dr. Hick's testified that use of a lower control limit with a Marshall-generated target figure "[was] done, but normally you would use a higher percentage in evaluating the mix.” Transcript at 478. A compaction of 95% to 96% of the Marshall figure would be comparable to 92% of the Rice Specific Gravity figure.

10 . Even government witnesses explained that one would not actually want to achieve a road density equal to the maximum specific gravity; a 100% pay factor would be well below 100% specific gravity. In that sense, 154.2 pcf was not a target in the same sense that targets were used gradation and asphalt content, the only other aspects of the contract for which pay factors were calculated. As Hanek testified, it would be desirable to achieve 100% of the target value for those factors, but not for compaction.

11 . Puget’s expert, Robert L. Dunning, testified that both the Marshall Method and Rice Specific Gravity are used in the road building industry. The Federal Highway Administration ("FHA") uses a road compaction figure of 90% of Rice Specific Gravity as a lower control limit. The State of Washington uses 91%. He was involved with a different Forest Service road project in Idaho in which the lower limit was 90% of Rice Specific Gravity. The Federal Aviation Authority, on the other hand, measures against a field compacted standard specimen and sets a lower control limit of 96.2%. He testified that use of a lower limit of 92% with a Rice Specific Gravity on a logging road would be over-engineering. It would be a higher criteria for compaction than the FHA and the State of Washington require on roads that should be of higher quality than the one in question. Based on the results Puget achieved here, it would have been earned a pay factor of 105% on an FHA contract, and approximately 96% on a State of Washington contract. Under the Forest Service’s construction of the contract at bar, Puget might not have been paid, even if it achieved 92% of maximum density, depending on the number of samples taken.

12 . The report is dated April 2, 1990. A letter from Gordon Hanek of March 2, 1990 recites that Dr. Hicks was sent a copy of both the standard and special specifications.