From Casetext: Smarter Legal Research

Swift Chemical Co. v. Usamex Fertilizers, Inc.

United States District Court, E.D. Louisiana
Jun 6, 1980
490 F. Supp. 1343 (E.D. La. 1980)


Civ. A. Nos. 74-76 "G", 79-812 "G".

June 6, 1980.

Jerry A. Brown, Richard W. Bussoff, New Orleans, La., and John W. Hofeldt, Rolf O. Stadheim, and J. Calvin Langston, Chicago, Ill., for Swift Chemical Co.

Bill Durkee, Michael O. Sutton, Houston, Tex., and J. William Vaudry, Jr., New Orleans, La., for Usamex and Fertilizantes Fosfatados Mexicanos, S.A.

Swift Agricultural Chemical Corporation (Swift) filed an action against Fertilizantes Fosfatados Mexicanos, S.A. (FFM) and its wholly owned subsidiary, Usamex Fertilizers, Inc. (Usamex), for infringement of a patent on the manufacture of liquid ammonium polyphosphate fertilizer, U.S. Letters Patent No. 3,464,808 (hereinafter referred to as the '808 or Kearns patent). The matter was tried in late November and early December 1976, and I rendered an opinion on October 7, 1977 that the patent-in-suit was valid and infringed. On December 19, 1977, I entered a judgment on liability, ordered an accounting to determine infringement damages, and enjoined defendants from infringing the '808 patent; however, I suspended the injunction pending defendants' appeal. In early 1978, the parties settled past infringement damages, costs, and attorney's fees and Usamex took a license. On March 3, 1978, a consent judgment on the issue of damages was entered, the appeal was dismissed, and the order suspending the injunction was vacated.

Almost one year later, on March 2, 1977, defendants filed a motion for relief from judgment pursuant to FRCP 60(b), essentially on the ground that newly discovered evidence revealed that their process did not infringe plaintiff's patent. In my opinion on the patent's validity and defendants' infringement, I noted that a critical aspect of the patent claim was the reaction time within which ammonia and phosphoric acid are contacted in a jet reactor at specified temperatures, and that plaintiff's patent called for a reaction residence time of less than one second. Defendants now claim they have new evidence proving the residence time of their process was greater than one second. Furthermore, in a declaratory judgment action brought pursuant to 28 U.S.C. § 2201, 2202, and 1338(a), and subsequently consolidated with the original case, defendants claim that shortly before the opinion on liability was rendered, they modified their process to employ a 33-ft, rather than a 12-ft reactor, creating a new process with a greater than one second residence time that also does not infringe the '808 patent. Therefore, they seek a declaratory judgment of noninfringement, and $108,001.54 in damages, which is the amount of royalties they paid Swift through March 1979 under the license agreement. I instructed the parties to stipulate the undisputed facts pertinent to these matters, and to brief the disputed legal issues.

Rec.Doc. No. 226. The published opinion is Swift Chemical Co. v. Usamex Fertilizers, Inc., et al., 197 U.S.P.Q. 10 (E.D.La. 1977).

Throughout this opinion, Usamex and FFM will be referred to as defendants and Swift as plaintiff, although, of course, the parties are reversed in the declaratory judgment action. Additionally, for purposes of brevity, defendants will be referred to simply as Usamex, rather than Usamex and FFM.

I. Background

The invention of the '808 patent was made by Tommy Carter Kearns, who was employed in fertilizer research and development by Swift. The patent discloses a method of producing a liquid ammonium polyphosphate fertilizer product by reacting ammonia and wet phosphoric acid under conditions that simultaneously neutralize and molecularly dehydrate the acid. Because of its relatively high polyphoephate content, the resulting reaction product, when dissolved in water, has self-sequestering properties, i.e., forms a solution wherein impurities, such as iron and aluminum salts, present in the wet process phosphoric acid are sequestered or held in solution, instead of precipitating out in the form of a gelatinous sludge that would hinder subsequent storage and handling of the liquid fertilizer product. More particularly, the method involves a direct ammoniation of orthophosphoric acid in such a manner that the exothermic heat of the reaction supplies the energy requirements to molecularly dehydrate the acid, i.e., convert a substantial portion of the orthophosphates to polyphosphates without formation of appreciable quantities of highly insoluble iron tripolyphosphates and metaphosphates.

Rec.Doc. No. 226 at 2, 197 U.S.P.Q. at 13. The basic technology involved in the process is detailed in Rec.Doc. No. 226 at 2-9; 197 U.S.P.Q. at 13-16.

Kearns filed his application for the patent-in-suit on August 8, 1965, and it was issued on September 2, 1969. The process is defined in the '808 patent claims:

1. A process for preparing ammonium polyphosphates having self-sequestering properties comprising: supplying a stream of ammonia to a jet reactor; supplying a stream of phosphoric acid having P2O5 content of between about 54% and about 68% to said jet reactor; and contacting said stream of ammonia with said stream of phosphoric acid in said reactor at temperatures of between about 450° F. and about 650° F. for a period of less than one second to form molten droplets of ammonium polyphosphate.
2. The method of claim 1 wherein the droplets of molten ammonium polyphosphate are subsequently quenched.
3. The process of claim 1 wherein the P2O5 content of acid is between about 60% and about 62%.
4. The process of claim 1 wherein at least 50% of the orthophosphate in the acid is converted to nonorthophosphate.

Rec.Doc. No. 226 at 5 and n. 11; 197 U.S.P.Q. at 14 and n. 11.

A significant feature of the process is that the ammonia and phosphoric acid are contacted in a reactor at a temperature of from about 450° F. to 650° F. for a duration of less than one second, after which the reaction product is quenched or cooled. Such rapid reaction at a high temperature, referred to as "residence time," was found by Kearns to produce a molten ammonium polyphosphate (melt) in which a high percentage of the phosphate was in the non-ortho form, yet with little or no formation of undesirable insoluble iron and aluminum polyphosphate compounds.

Rec.Doc. No. 226 at 6; 197 U.S.P.Q. at 15.

To provide for such extremely brief residence time at high temperature, the patent requires the use of an elongated, open-ended pipe in which to contact the ammonia and phosphoric acid, as opposed, for example, to a tank or other such reaction vessel. The patent describes the reaction pipe as a "jet" or "jet reactor," and the process is sometimes referred to as a "jet process.".


The accused process employed by Usamex at its plant in St. Rose, Louisiana by March 1971 involved a 12-ft reactor pipe and was described in U.S. Letters Patent No. 3,734,708 (hereinafter referred to as the '708 or Burns patent), issued May 22, 1973 to Tom V. Burns, an FFM employee. By July 1975, Usamex was employing a process for manufacturing liquid fertilizer generally described in U.S. Letters Patent No. 3,998,140 (hereinafter referred to as the '140 or Burns and Ortega patent).

Rec.Doc. No. 226 at 9; 197 U.S.P.Q. at 16.

Rec.Doc. No. 337, Stipulation No. 10.

II. Stipulated Facts

In the development of the Kearns process, which culminated in the issuance of Swift's '808 patent, the melt was observed as being suspended in steam and exiting the reactor at a high velocity. Although neither the Kearns '808 patent, the Burns '708 patent, nor the Burns and Ortega '140 patent contains a statement that the residence time was to be determined by that of the steam, it was so calculated by the respective patentees.

When Usamex first began developing its process for the production of liquid ammonium polyphosphate, residence time in its 12-ft reactor was calculated on the basis of steam residence time. The actual flow characteristics in the pipe were not known exactly and were not physically assessed, and Burns never observed the steam or melt exiting the reactor. It was assumed that the melt traveled through the pipe at the same speed as the steam. Burns did not know of any way of determining residence time except by calculating steam residence time, and was unaware that radioactive material tests could be performed to determine the residence time of the steam or melt. He did not spend a great deal of time trying to find another way to calculate the residence time. Calculating steam residence time was a suitable, convenient, and practical way to determine residence time; it was convenient for evaluating plant design and indicating retention time for purposes of polyphosphate conversion.

Id. No. 14.

Id. Nos. 11, 12, 15.

In the process employed by Usamex until early October 1977, a 12-ft reactor was used, but at that time, shortly before I rendered my opinion, Usamex modified the 12-ft reactor pipe by adding 21 feet of pipe and altering the configuration. The daily production rate was increased from 600 to 700 tons. Both the 12-ft and 33-ft processes employ acid and temperatures in the ranges specified in Claim 1 of the Kearns '808 patent; achieve conversion of at least 50% of the orthophosphates of the phosphoric acid to polyphosphate form, the polyphosphates having self-sequestering properties; avoid formation of highly insoluble iron tripolyphosphates and metaphosphates; and avoid reconversion of the polyphosphates back to orthophosphate form. In both processes, the calculated steam residence time is less than one second. Usamex made no efforts by test or otherwise to determine actual residence time of the liquid melt in its reactor pipes in the design and development of its 12-ft and 33-ft processes.

Id. Nos. 11, 12, 15.

Id. No. 13.

In preparation for the 1976 trial, Everett M. Mortenson, Swift's expert, calculated the residence time of the steam in the Usamex 12-ft reactor. He did not consider the scale formed in the reactor and the volume of melt because he thought they were minor relative to the volume of steam. He testified that the melt flowed through the reactor at the same velocity as the steam; that the reaction created a froth that was carried along with a large volume of steam and gushed through the pipe at a relatively high velocity, sweeping everything out of the reactor in a jet stream or gush of fluid, steam, and melt, and discharging it in the form of droplets of melt; and that the steam would represent about 98-99% of the total volume of the material going through the Usamex reactor. The residence time of melt suspended in steam in a pipe reactor is approximately that of the steam. Dr. Graham Wallis, another Swift expert, testified that the flow regime in a reactor of peculiar geometry, such as Usamex's, must be determined by actual tests, and that calculations would give clues, but are "very dangerous" in attempting to determine the actual flow regime.

Id. No. 18.

Id. No. 16.

Id. No. 19.

The possibility that the liquid (melt) and gas (steam) phases might separate in the Usamex reactor, and that melt residence time might be greater than one second, first was considered and investigated by Usamex's experts, W.R. Mustian, Jr. and A.V. Slack, in March 1976, which was eight months prior to trial and three months prior to submission of the pre-trial order. Usamex hired both men to help prepare its defense to the infringement allegation, and they were listed as "will call" witnesses. Shortly before their March 1976 tests, Mustian asked Usamex's patent counsel, Walter Gillis, whether the question of noninfringement due to a different residence time had been explored; Gillis said it had. Gillis testified that "all the information we had was that everything blew through the reactor because of the generation of the steam," and that when counsel previously asked the client whether the residence time might be increased to avoid infringement, the technical advice was that was not feasible. Gillis said that prior to spring 1976, he "never dreamed" that the steam and melt might separate in the reactor. He said the March tests and the question of residence time were under "constant discussion" up to the time of trial.

Id. No. 22.

Id. No. 23.

Usamex admits its three March 1976 tests were crude and not conducted under normal operating conditions of the Usamex process, but contends they attempted to simulate normal operating conditions. Usamex's counsel decided to stand on the opinion of Mustian and Slack "that maybe there was phase separation confirmed, and that the tests, insofar as they were relevant to that issue, confirmed it. They certainly did not preclude us from introducing evidence of the opinion they had before the tests." Usamex did not include its theory of phase separation and longer residence time in the pre-trial order, and did not inform Swift of those theories until ten days before trial. Swift first learned of the tests when Mustian testified at trial. When Swift questioned the theories and tests on cross-examination, defendants made a tactical decision to withdraw the phase separation/residence time noninfringement defense. They did not call Slack to testify about the March 1976 tests, or Ronald Fogg, FFM technical director, to testify about a melt withdrawal he observed in Usamex's reactor in 1971 and his opinions based on it.

Id. Nos. 24, 25, 26, 27.

In October 1977, I found in favor of Swift, holding the '808 patent valid, enforceable, and infringed by defendants. I determined that Usamex's phase separation/residence time defense was "obviously groundless," and its tardy assertion and subsequent withdrawal was one of the bases upon which I awarded Swift attorney's fees. After entry of my judgment order of December 19, 1977, permanently enjoining defendants from further infringement, and ordering an accounting to determine damages, defendants moved to suspend the injunction pending appeal. They did not inform Swift or the Court that they had been employing the 33-ft reactor for several months. I granted the motion and suspended the injunction after defendants posted a $900,000 bond. By that time, defendants had spent $600,000 for attorney's and consultant's fees and disbursements in preparation for the 1976 trial.

Rec.Doc. No. 226.

Id. at 35; 197 U.S.P.Q. at 30-31.

Rec.Doc. No. 337, Stipulation Nos. 29, 30.

Id. No. 28.

In late 1977, Mustian, then Usamex's manager of production, prepared a 36-page report entitled "A Brief in Support of Usamex's Assumption of a Position of Non-Infringement," recommending that competent experts be consulted to confirm his phase separation/residence time arguments, and that Usamex explore with counsel ways in which to return to court to prove that neither the 12-ft nor the 33-ft process infringed Swift's patent. Before the report was completed, defendants had decided to "settle friendly" in accordance with my findings, and approached Swift to obtain a license under its patent and to settle past infringement damages, costs, and attorney's fees. At the same time, Usamex intended to search for another way to contest infringement, but did not so inform Swift or the Court. Usamex took a license, and on March 3, 1978, a final judgment was entered by consent. The judgment awarded Swift damages based on Usamex's sales of the fertilizer prior to January 1, 1978. Part of the sales included fertilizer produced since October 1977 using the 33-ft reactor, although Swift did not know it, nor did it know that Usamex intended to continue using the 33-ft reactor under the license.

Id. Nos. 31, 32, 33.

Id. Nos. 34, 35, 36, 37.

In April and May 1978, Mustian wrote a second report entitled "A Review In Re Swift v. Usamex." One of its stated purposes was "to provide bases for determining the merits of actions to be considered leading to termination and reversal of the penalties assessed against Usamex." In July 1978, defendants consulted new patent counsel for advice concerning the payment of royalties to Swift, and the attorneys advised that at least one impeccable technical expert should be consulted to prove their position of noninfringement.

Id. No. 39.

Id. No. 40.

Drs. Dukler, Parker and Houze were contacted in October 1978 and tests were planned. In December 1978, after Dukler had made calculations of residence time, but before he had run any tests, Usamex told Swift it had a "new process" that did not infringe, and invited Swift to observe the tests; Swift declined. Tests were run in December 1978 and January 1979 to determine residence time in the 33-ft reactor. Another test was made February 24, 1979 employing a reconstructed 12-ft reactor. These were the first tests run since March 1976 for the purpose of determining residence time.

Id. Nos. 41, 42, 44.

In January 1979, Usamex made its quarterly royalty payment covering the use of its process (in which Usamex was employing the 33-ft reactor) "under protest." It filed its motion for relief from judgment under Rule 60(b) on March 2, 1979, and its declaratory judgment action the same day. Usamex did not make its royalty payment in April 1979, contending its process using the 33-ft reactor did not infringe. After Swift threatened to terminate the license if the payment was not made, Usamex moved to enjoin termination; however, I denied the motion, and Usamex immediately made the royalty payment.

Id. Nos. 43, 45, 47, 48.

III. Contested Facts

The following is Usamex's description of the new test results and expert opinions drawn from them; Swift denies the validity of the tests, the conclusions drawn, and their relevancy to these matters:

(a) The melt and steam produced in the Usamex reactors separate and do not flow through the reactors at the same speed;

(b) Conversion to non-ortho polyphosphates is higher at the end of the 33-ft reactor than at any preceding test location;

(c) A reaction to achieve a conversion to higher polyphosphates proceeds along the entire length of the 33-ft reactor;

(d) The melt produced in the 33-ft reactor has a range of residence times, the most probable residence time as determined by the radioactive tracer test is about 10 seconds, and as determined by the chemical tracer test, about 30 seconds.

(e) The temperature along the entire length of the 33-ft reactor during the tests performed on December 16, 1978 and January 2, 1979 was within the range of from 487° F. to 523° F.

(f) The melt flows along the bottom of both the 12-ft and 33-ft reactors in a stratified flow pattern separate from the steam, rather than being carried through the reactors as "molten droplets suspended in the steam and exiting at high rates of speed."

(h) Prior to the tests of December 16, 1978, Dr. Dukler calculated that the melt flows through the 12-ft and 33-ft reactors in a stratified flow pattern separated from the steam, that the residence time of the melt in the 33-ft reactor is about 11.7 seconds, and that the residence time of the melt at 12 feet in the 33-ft reactor is about 4.3 seconds.

Red.Doc. No. 337, Appendix.

IV. Rule 60(b) Motion

FRCP 60(b) provides, in pertinent part: On motion and upon such terms as are just, the court may relieve a party . . from a final judgment, order, or proceeding for the following reasons: (1) mistake, inadvertence, surprise, or excusable neglect; (2) newly discovered evidence which by due diligence could not have been discovered in time to move for a new trial under Rule 59(b);. . . (5) . . . it is no longer equitable that the judgment should have prospective application; or (6) any other reason justifying relief from the operation of the judgment. The motion shall be made within a reasonable time, and for reasons (1), (2), . . . not more than one year after the judgment, order, or proceeding was entered and taken. A motion under this subdivision (b) does not affect the finality of a judgment or suspend its operation.

Rule 60(b) "must be equitably and liberally applied to achieve substantial justice." Blois v. Friday, 612 F.2d 938 (5th Cir. 1980). Accord, Laguna Royalty Co. v. Marsh, 350 F.2d 817, 823 (5th Cir. 1965). This rule, which allows the trial court to reopen a case, is:

most liberally applied to default judgments; its main application is to those cases in which the true merits of a case might never be considered because of technical error, or fraud or concealment by the opposing party, or the court's inability to consider fresh evidence. [Citations omitted.] The purpose of the motion is to permit the trial judge to reconsider such matters so that he can correct obvious errors or injustices and so perhaps obviate the laborious process of appeal. Weighing against the grant of a 60(b) motion is the desirability of finality in judgments. This is particularly true where the reopening of a judgment could unfairly prejudice the opposing party. [Citation omitted]. But even without such prejudice, the desirability of orderliness and predictability in the judicial process speaks for caution in the reopening of judgments. These are matters that are addressed to the sound discretion of the trial court, and its ruling . . . will be reversed on appeal only upon a showing of abuse of discretion. [Citations omitted].
Fackelman v. Bell, 564 F.2d 734, 735-36 (5th Cir. 1977).

Defendants originally brought their Rule 60(b) motion under subsections (1), (2), (5), and (6); however, in my June 6, 1979 minute entry denying Swift's motion to dismiss the Rule 60(b) motion, I concluded, and I now reaffirm, that "the thrust of the defendants' argument is that newly discovered evidence [subsection (2)] shows their process did not infringe that of the plaintiff."

Rec.Doc. No. 280 at 2.

Defendants contend that the case should be reopened pursuant to FRCP 60(b)(1) because the testimony of Swift's expert, Mortenson, that the melt and steam flow through the 12-ft Usamex reactor at the same velocity, and that droplets of melt are discharged, was "mistaken"; however, they do not cite any case involving similar facts where the court allowed the Rule 60(b)(1) movant to characterize testimony at trial as "mistaken" where that testimony was an expert opinion based upon evidence available at the time of trial. Mortenson testified the residence time of the steam in the Usamex reactor was less than one second. Since defendants do not now contradict that measurement, but only seek to offer evidence of the residence time of the melt in the Usamex reactor, it is improper to label Mortenson's testimony as "mistaken."

Subsections (5) and (6) of Rule 60(b) also do not provide suitable avenues for reopening this case. Rule 60(b)(5) covers situations where it is no longer equitable for a judgment to have prospective application. It is a proper remedy where the court finds that due to changed conditions, the continuous effect of an injunction imposes an undue hardship on the moving party. C. Wright A. Miller, Federal Practice and Procedure § 2863 (1973). As explained subsequently, no material operative facts have changed; therefore, the equities lie with Swift and the continuous effect of the injunction. Rule 60(b)(6) is the catchall section that is available where "any other reason justifying relief from the operation of the judgment" is present. The "other reasons" are mutually exclusive from the five more specific grounds listed in Rule 60(b); when a reason for relief is set forth in another subsection, Rule 60(b)(6) cannot be the basis for relief. Transit Cas. Co. v. Security Trust Co., 441 F.2d 788 (5th Cir. 1971). In this case, the ground for which relief is sought is contained in Rule 60(b)(2).

A motion for a new trial under Rule 60(b)(2) is an extraordinary motion, and the requirements of the rule must be strictly met. Ag Pro, Inc. v. Sakraida, 512 F.2d 141, 143 (5th Cir. 1975). The motion may not be granted unless: (1) the new evidence was discovered following the trial; (2) the evidence is not merely cumulative or impeaching; (3) due diligence on the part of the movant to discover the new evidence is shown or may be inferred; (4) the evidence is material; and (5) the evidence is such that a new trial would probably produce a new result. Id. These requirements are apropos to patent litigation. Id. The existence of requirements (1) and (2) does not appear to be disputed; however, I must determine whether Usamex satisfies (3), (4), and (5).

A. Due Diligence

The cases cited by defendants where courts found due diligence to discover new evidence simply are not very helpful, as this is the type of determination that necessarily turns on the facts in a particular case. For example, defendants rely on Laguna Royalty Co. v. Marsh, 350 F.2d 817 (5th Cir. 1965), where the court granted a Rule 60(b) motion to vacate on the ground of newly discovered evidence. Laguna was an action for damages to a gas well brought by owners of working interests and leaseholders against the landowner/lessor, Marsh. One of the defense theories was the sands in the well area were "bentonitic," which would decrease the porosity of the sand and prevent gas from flowing in the well, thus ending the well's productivity. Both sides offered expert testimony on the issue, but all of the experts agreed that the only way to determine conclusively whether the sands were bentonitic was by analysis of core samples. Although a method for drilling and analyzing core samples was available prior to trial, no one had done so. The facts here, however, are not analogous. In Laguna, the defendant, among other things, threatened a working interest owner with trespass charges if any action was taken on the leasehold other than the removal of salvage equipment. The necessary tests were of cores obtainable during well drilling operations. The court held:

"[T]he Trial Court could not seriously consider the contention made by the defendant to us that the plaintiffs were not diligent since they didn't drill a well to get cores upon which to refute Marsh's theory of water damage. The lawsuit was about the cost of a replacement well. It was complex enough without getting in a second one."
Laguna, supra, at 825. The facts in the other cases cited by defendants are also inapposite, and do not merit discussion.

In this case, Usamex's experts, as early as March 1976, attempted to validate their phase separation/melt residence time theory by conducting some rather crude tests. Other than the fact that the theory and these tests were under "constant discussion" up until trial, no efforts were made to discover a test that would validate the theory until Mustian prepared reports in December 1977 and April-May 1978 promoting the theory. Experts were not contacted until October 1978 to conduct further tests. Once contact was made, tests allegedly favorable to defendants' position were completed in less than five months. Yet Usamex concedes it was "aware of the criticality of the residence time of the phosphate melt in the reactor to a finding of infringement," although "no one was aware of a way to accurately measure this time." Usamex has failed to meet its Rule 60(b)(2) burden of proving it exercised due diligence to discover its new test evidence.

Rec.Doc. No. 339 at 9.

Even if Usamex exercised due diligence, its motion for relief from judgment must be denied if it fails to prove its newly discovered evidence is material and is such that a new trial would probably produce a new result.

B. Materiality of evidence and question whether new evidence is such that a new trial would probably produce a new result

Since Usamex does not now reurge the invalidity of Swift's patent, the only question here is, assuming the new scientific evidence of Usamex on its 12-ft process is valid, does the process infringe the Swift patent?

In my opinion following the trial, I noted that defendants did not contest the fact that the Usamex 12-ft process is the same as that described in the Swift patent. Moreover, Usamex has since stipulated that its process achieves conversion of at least 50% of the orthophosphates of the phosphoric acid to polyphosphate form, the polyphosphates having self-sequestering properties; avoids formation of highly insoluble iron tripolyphosphates and metaphosphates; and avoids reconversion of the polyphosphates back to orthophosphate form. The process employs acid and temperatures within the ranges specified in Claim 1 of the Swift patent. Furthermore, the calculated steam residence time in the Usamex reactor is less than one second.

Id. No. 226 at 9-10; 197 U.S.P.Q. at 17.

Usamex now urges, however, that even though I correctly found Swift's patent to be valid, a consideration of Usamex's new evidence would result in a determination that its process does not infringe the '808 patent, because its "residence time" is greater than one second. The new evidence is that the melt and steam produced in the Usamex reactor separate and do not flow through at the same speed, i.e., the melt flows along the bottom of the reactor in a stratified flow pattern separated from the steam, rather than being carried through the reactor as "molten droplets suspended in the steam and exiting at high rates of speed," as in the '808 process. Although the residence time of the steam is less than one second in the Usamex process, the residence time of the melt is about four seconds in the 12-ft reactor, as determined by the radioactive tracer test.

Usamex contends that the scope of the '808 patent is limited to processes in which the melt residence time is less than one second, and that the steam residence time is irrelevant. Usamex bases its construction of the patent's scope on my opinion, the patent itself, the prior art, and statements of Swift's counsel in presenting its claim to the Patent Office and in prosecuting its infringement action. It argues that the law of the case is that residence time relates to the time "the fertilizer product" or "the reaction product," i.e., the molten ammonium polyphosphate, is in the reactor, not the time the steam is in the reactor.

This construction is not supported by the record. As I wrote:

"A significant feature of the process is that the ammonia and phosphoric acid are contacted in a reactor at a temperature of from about 450° F. to 650° F. for duration of less than one second, after which the reaction product is quenched or cooled. Such rapid reaction at a high temperature referred to as "residence time" was found by Kearns to produce an ammonium polyphosphate melt in which a high percentage of the phosphate was in the non-ortho form, yet with little or no formation of undesirable insoluble iron and aluminum polyphosphate compounds.

Rec.Doc. 226 at 6; 197 USPQ at 15 (emphasis added).

I also found that prior art reveals that the phrase in Claim 1 of the '808 patent, "contacting said stream of ammonia with said stream of phosphoric acid . . . for a period of less than one second," refers "not to the speed of the molecular reaction, but to the time required to adequately mix the reactants."

Rec.Doc. 226 at 30; 197 USPQ at 28.

The real issue, then, is whether Usamex can avoid a finding of infringement by determining the residence time required to adequately mix the reactants by calculating the melt residence time rather than the steam residence time.

Although the '808 patent specifies that "[t]he reaction product as it emerges from the pipe consists of molten ammonium polyphosphate suspended in a high velocity stream of superheated steam," and it is stipulated that "[t]he residence time of melt suspended in a steam in a pipe reactor is approximately that of steam," there is nothing in the patent or elsewhere in the record that supports a finding that the less than one second residence time is determined by calculating the melt, rather than the steam, residence time.

Swift '808 Patent following Rec.Doc. No. 341, Col. 5, 11. 25-27.

On the contrary, it is stipulated that the residence time stated in the '808 patent is "the calculated residence time of the steam in the reactor." Additionally, although the patents describing the processes employed by Usamex — the Burns '708 patent and the Burns and Ortega '140 patent — do not explicitly state that their residence times are the calculated residence times of the steam, it is so stipulated. Usamex never physically assessed the flow characteristics of the melt and steam in the reactors used in its processes, but simply assumed that the melt traveled at the same speed as the steam. Having been taught by the '808 patent that an extremely short residence time was a factor, Usamex perfected the accused process. Calculating steam residence time was a suitable, convenient, and practical way for Usamex to determine residence time in order to evaluate its plant design.

It is particularly incongruous for Usamex to assert that the residence times specified in the Swift patent and the prior art are melt residence times, when it alleges in its Rule 60(b) motion that no one knew of a way to accurately measure melt residence time. Indeed, Usamex did not develop evidence of phase separation and a longer melt residence time until late 1978 and early 1979, when it was attempting to avoid this court's previous finding of infringement. The new evidence has had no practical use, as Usamex's processes were perfected based on calculations of steam residence time.

In deciding whether an accused device or composition infringes a valid patent, resort is had in the first instance to the words of the claim, and "[i]f accused matter falls clearly within the claim, infringement is made out and that is the end of it." Graver Tank Mfg. Co. v. Linde Air Products Co., 339 U.S. 605, 607, 70 S.Ct. 854, 855, 94 L.Ed. 1097 (1950); Studiengesellschaft Kohle v. Eastman Kodak Co., 616 F.2d 1315 (5th Cir. 1980). "In considering literal infringement, the patent's claims must be read in connection with [the] patent's specification and its file history, and the claims of [the] patent cannot be given a construction broader than the teachings expressed in the patent." Studiengesellsehaft Kohle, supra, at 1324. Accord, Marvin Glass Assoc. v. Sears, Roebuck Co., 448 F.2d 60 (5th Cir. 1971). The '808 patent teaches that favorable results can be obtained in the manufacture of liquid ammonium polyphosphate when a stream of ammonia and a stream of phosphoric acid having P2O5 content of between about 54% and 68% are supplied to a jet reactor and contacted at temperatures of 450°-650° F. for less than one second. Because this is the process employed by Usamex, it literally infringes the '808 patent.

Even if the actual flow pattern and melt residence time in the Usamex reactor were considered in determining infringement, and it was found that there is phase separation and a more than one second melt residence time, under the judicially created doctrine of equivalents, infringement still would exist. Because minor modifications in a patented invention are sufficient to put the item beyond the scope of literal infringement, courts have developed the doctrine of equivalents to protect patentees from inventions that perform "substantially the same function in substantially the same way to obtain the same result." Graver Tank, supra, 339 U.S. at 43, 70 S.Ct. at 13; Studiengesellschaft Kohle, supra, at 1324. What constitutes equivalency "must be determined against the context of the patent, the prior art, and the particular circumstances of the case." Graver Tank, supra, 339 U.S. at 609, 70 S.Ct. at 856. Usamex's new evidence does not change the fact that its process operates in the same manner under the same physical laws to produce the same result as the Swift patented process. At most, the new evidence would show that Swift was ignorant of the physical laws at work inside the Usamex reactor, and perhaps even inside its own reactor, and could not accurately determine the melt residence time by calculating the steam residence time. Yet the processes accomplish their objectives when the critical less than one second residence time is determined according to the calculated steam residence time. In perfecting both its 12-ft and 33-ft processes, Usamex determined residence time according to the calculated steam residence time.

Furthermore, a party cannot avoid a finding of infringement by relying on tests not known to the art at the time of the application for the patent. In an infringement action, Raybestos-Manhattan, Inc. v. Texon, Inc., 268 F.2d 839 (1st Cir. 1959), the court held that a test not recognized at the date of the patent application cannot be used either to limit or extend the scope of the patent, as the patent must be interpreted in light of the skill of the art to which it pertains at the date of application. Id. at 842. The patent-in-suit was for a method of producing thermo-setting resin impregnated cellulose fibre web sheet material. The claim element at issue called for a web moisture content of "about 4% to 8%," which the district court construed as an absolute limitation. The patent was silent as to any test for determining the claimed range of water content; however, the court found that the Karl Fischer test was the most accurate test, and it revealed a water content of .5 to 2.9% in the accused method. The plaintiff did not offer any evidence that the Karl Fischer test was inaccurate, and the court granted summary judgment for the defendant. On appeal, the inventor of the plaintiff's patent stated in an affidavit that in determining the amount of moisture required for his process, he relied on the "standard" desiccation (dehydration) method, and at that time, did not know of the Karl Fischer method. Reversing, the First Circuit held:

"[T]he district court, under the circumstances presented here, was required to make a finding as to what test would be employed at the time of the patent application to determine the amount of moisture resulting from the process in question, by a person skilled in the art to which the application pertained. This is an objective standard and does not depend on the subjective intent of the inventor. Moreover, it does not cause the patent to mean one thing at the time of its issuance and another at some later date upon the discovery of a more accurate test."
Raybestos-Manhattan, supra, at 842.

Usamex cites Helene Curtis Industries v. Sales Affiliates, 233 F.2d 148 (2d Cir. 1956), in support of its position that post-application tests may be considered. There, plaintiffs sought a declaratory judgment that defendant's patent for a chemical composition suitable for permanent hair waving was invalid. The defendant's patent specified mercaptan as a waving agent, and the Second Circuit construed the patent as teaching that a mercaptan pH ceiling of 10 was a critical limitation, because with a higher pH, hair destruction or injury might occur before a satisfactory wave was obtained. At trial, plaintiff offered tests using the "direct application" method of hair waving to show that successful waving without hair damage could be accomplished by the use of mercaptan solutions having an alkalinity in excess of pH 10; therefore, that pH level was not critical. On appeal, defendant asserted that the tests were not admissible on the issue of criticality, because at the date of his patent application, the commercial practice was to wave by the "circulation" method. The court said this contention was "fallacious in law," and continued,

"It is, of course, true that patentable invention must be measured by the advance which it teaches over the prior art. But in determining the existence or nonexistence of such physical phenomena as demonstrate a critical advance over the prior art, we think the court is entitled to such aid as derives from any relevant technique, — even one not developed until after the invention at issue."
Helene Curtis, supra, at 154-55. Its holding, however, was that defendant's argument was factually unsound, because at the date of the patent application, the "direct application" method "had substantial commercial vogue." Id. at 155. The language in Helene Curtis relied on by Usamex not only is dicta, but also addresses an issue not disputed here. In this case, the validity of the Swift patent, and hence the criticality of the less than one second residence time, is not challenged. Only the manner in which the residence time is determined is disputed.

The record is clear that the test for determining residence time that would have been employed in August 1965, the time of Kearns' patent application, by a person skilled in the art, was calculation of steam residence time. It was not until March 1976 that Usamex's technical consultants for the infringement action thought of another way to calculate residence time, but did not produce a convincing test. Their idea was not revived until late 1977, when Usamex sought a way to reverse the penalties assessed against it for infringement, and the chemical and radioactive tracer tests were not perfected until late 1978 and early 1979. In the interim, Usamex designed, built, and utilized its 33-ft reactor, although a test for calculating melt residence time still had not been performed. Usamex concedes that such tests were unknown to those working in the art: "[T]he [new] evidence . . . relates to two sampling techniques which were previously unheard of in the field of fertilizer manufacturing and which took months to develop, successfully run, and obtain complete results from." Consequently, Usamex's new evidence of melt residence time is legally irrelevant, as reliance on it would "cause the patent to mean one thing at the time of its issuance and another at some later date." Raybestos-Manhattan, supra, at 842.

Rec.Doc. No. 269 at 7.

Usamex not only has failed to prove it exercised due diligence in discovering its new evidence, but the evidence, even if admitted at a new trial and found to be credible, would not produce a different result, because it is legally immaterial. Accordingly, Usamex's Rule 60(b)(2) motion for relief from judgment is DENIED.

V. Declaratory Judgment Action

Just prior to the issuance of the October 1977 opinion on the validity of the '808 patent and Usamex's infringement through use of its 12-ft reactor, Usamex constructed and began operating a 33-ft reactor. Subsequently, the parties settled the amount of damage and attorney's fees Usamex was to pay Swift under the judgment on liability, and a final consent judgment was entered March 3, 1978. The settlement included payments for periods when Usamex produced fertilizer using not only the 12-ft reactor, but also the 33-ft reactor, although Usamex did not inform Swift or the court of the existence and operation of the 33-ft reactor until many months later. The parties also entered into a license agreement, which provided that Usamex agreed to pay Swift a specified royalty for ammonium polyphosphate Usamex manufactured and sold under Swift's "Patent Rights," and that Swift granted Usamex a nonexclusive license to manufacture, use, and sell ammonium polyphosphate manufactured under Swift's "Patent Rights." The agreement defines "Patent Rights" as

any process claimed and any process which by application of the Dostrine of Equivalents comes within the claims of U.S. Patent No. 3,464,808.

License Agreement following Rec.Doc. No. 341.

After Usamex obtained its new evidence of phase separation and a longer melt residence time, it sought a declaratory judgment that its process employing the 33-ft reactor does not infringe the admittedly valid '808 patent. Swift now asserts that the action should be dismissed, because it is barred by the doctrines of compromise and settlement, res judicata, and collateral estoppel.

A. The Settlement of Retrospective Damages

Ordinarily, compromises of disputed claims are favored by the courts. Williams v. First Nat'l Bank, 216 U.S. 582, 30 S.Ct. 441, 54 L.Ed. 625 (1910). Where the parties, acting in good faith, settle a controversy, the courts will enforce the compromise without regard to what the result might, or would, have been, had the parties chosen to litigate rather than settle. Hennessy v. Bacon, 137 U.S. 78, 11 S.Ct. 17, 34 L.Ed. 605 (1890). Accord, J. Kahn Co. v. Clark, 178 F.2d 111 (5th Cir. 1949). Settlements generally are viewed as binding, final, and as conclusive of rights as a judgment. Thomas v. State of Louisiana, 534 F.2d 613 (5th Cir. 1976).

Nothing in the record suggests that Swift acted in any manner except in good faith in settling the amount of damages and attorney's fees owed by Usamex following the finding of infringement. If either party acted in less than good faith, it was Usamex, which did not reveal that it had been using the 33-ft reactor during part of the period for which damages were settled. Usamex's protest that Swift failed to inquire whether it had modified its infringing process is unworthy of serious consideration. Likewise, any suggestion by Usamex that the settlement agreement is not binding due to the unequal bargaining positions of the parties is meritless. This litigation was lengthy, costly, and hard-fought. Usamex had every opportunity to acquire knowledge of the facts bearing upon the question of the validity of its defenses. See Hennessy, supra, 137 U.S. at 85, 11 S.Ct. at 19. That Usamex has discovered new evidence bears only on its Rule 60(b) motion, and cannot alter the fact that the parties ended their controversy by agreement. Finally, the compromise was not based on any mutual mistake of fact. During settlement negotiations, Usamex believed there was phase separation in its reactors; Swift believed there was not. Whatever the truth is, at best only one of the parties could have been mistaken about the issue. A unilateral mistake about a particular fact is insufficient to reform a contract otherwise properly entered into. See Stevens v. Illinois Central R.R. Co., 234 F.2d 562, 564 n. 4 (5th Cir. 1956). Accordingly, the settlement of damages and attorney's fees will not be disturbed.

B. The License

Swift contends that the judicial determination of validity and infringement and the entry of a final judgment precludes relitigation of the question whether Usamex's process using its 33-ft reactor infringes the '808 patent. In analyzing this contention, it is important to distinguish the two doctrines concerning the preclusive effect of a prior adjudication that are embodied in the rules of res judicata. The first doctrine, "claim preclusion," or true res judicata, treats a judgment, once rendered, as the full measure of relief to be accorded between the same parties on the same "claim" or "cause of action." Kaspar Wire Works, Inc. v. Leco, Engineering Mach., 575 F.2d 530, 535 (5th Cir. 1978).

"When the plaintiff obtains a judgment in his favor, his claim `merges' in the judgment; he may seek no further relief on that claim in a separate action. Conversely, when a judgment is rendered for a defendant, the plaintiff's claim is extinguished; the judgment then acts as a `bar' [Citations omitted]. Under these rules of claim preclusion, the effect of a judgment extends to the litigation of all issues relevant to the same claim between the same parties, whether or not raised at trial."
Id. The second doctrine, "issue preclusion," or the collateral estoppel arm of res judicata, bars the relitigation of issues actually adjudicated and essential to the judgment in the prior litigation between the same parties. Kaspar, supra. Furthermore, the issue to be precluded must be identical to that involved in the prior action. James Talcott, Inc. v. Allahatad Bank, Ltd., 444 F.2d 451 (5th Cir. 1971). The collateral estoppel doctrine recognizes that suits addressed to particular claims may present issues relevant to suits on other claims. Kaspar, supra, at 535.

Swift makes a clever argument that because the final judgment awarded damages for the Usamex process employing the 33-ft reactor, it necessarily includes a finding that the 33-ft process infringed the '808 patent; therefore, the judgment should have res judicata effect. I need not decide whether this is a sound argument, because collateral estoppel is an appropriate basis upon which to rule.

The key issue Usamex might be collaterally estopped from litigating is whether melt residence time of more than one second is a valid defense to an anticipated claim that the 33-ft process infringes Swift's "Patent Rights." The three elements necessary for the application of collateral estoppel are present. First, the defense was actually litigated in the first trial, and I found it "obviously groundless." Its tardy assertion and subsequent withdrawal were part of the basis for awarding Swift attorney's fees. That Usamex now has new evidence to support the defense is irrelevant. All that is required for preclusion is that the issue was litigated, not that it was litigated well. Second, resolution of the issue was essential to the prior judgment. Had the defense been successful, Usamex would not have been found to be an infringer. Third, the issue in the declaratory judgment action is identical to that involved in the prior action. Aside from the availability of new evidence of melt residence time, the only difference between the 12-ft and 33-ft processes is 21 feet of pipe. The melt residence time in the 12-ft reactor allegedly was more than one second, just as it is in the 33-ft reactor, but the processes still infringe the '808 patent.

Usamex's allegation that conversion to non-ortho phosphates is higher at the end of the 33-ft reactor than at any preceding test location, if true, would not lead to a finding of non-infringement, and therefore is immaterial. Only Claim 4 of the '808 patent refers to the amount of conversion, and it calls for "at least 50%." Any greater conversion in the Usamex process falls within this claim. Even if the greater conversion is considered an improvement of the patented process, the process remains substantially the same as the '808 process, and infringement would not be avoided. See Tilghman v. Proctor, 102 U.S. 707, 26 L.Ed. 279 (1881); Harrington Mfg. Co., Inc. v. White, 475 F.2d 788 (5th Cir. 1973).

Nevertheless, Usamex contends that public policy prevents the application of the two res judicata doctrines in the context of patent litigation, relying on Lear, Inc. v. Adkins, 395 U.S. 653, 89 S.Ct. 1902, 23 L.Ed.2d 610 (1969). In Lear, the Supreme Court abolished the doctrine of patent licensee estoppel, a doctrine rooted in contract principles, which estopped the licensee operating under a license agreement from asserting the invalidity of the patent as a defense to the licensor's contract action for nonpayment of royalties. The rationale underlying the decision was that the public has an interest in obtaining a judicial determination of the invalidity of worthless patents. The Court noted that licensees may often be the only parties with enough economic incentive to challenge a patent's validity.

Although Lear was concerned only with the binding force of a license agreement, its rationale has been extended by analogy to consent decrees to preclude the defense of res judicata. Crane Co. v. Aeroquip Corp., 504 F.2d 1086 (7th Cir. 1974); Kraly v. Nat'l Distillers and Chem. Corp., 502 F.2d 1366 (7th Cir. 1974); Business Forms Finishing Serv., Inc. v. Carson, 452 F.2d 70 (7th Cir. 1971). Nevertheless, other courts have continued to apply the rule of Addressograph-Multigraph Corp. v. Cooper, 156 F.2d 483 (2d Cir. 1946), to give res judicata effect to consent decrees that specify both patent validity and infringement, Broadview Chem. Corp. v. Loctite Corp., 474 F.2d 1391 (2d Cir. 1973); Wallace Clark Co., Inc. v. Acheson Industries, Inc., 532 F.2d 846 (2d Cir. 1976); see also Warner-Jenkinson Co. v. Allied Chem. Corp., 567 F.2d 184 (2d Cir. 1977); and to refuse to give it such effect where the parties' intention in entering a consent decree is not clear, Brose v. Sears, Roebuck Co., 455 F.2d 763 (5th Cir. 1972).

For the remainder of this opinion, the term "res judicata" will be used in its broad sense, which includes collateral estoppel.

Although the Fifth Circuit, as other courts, has recognized that "[p]ublic policy, explicated in Lear, commands us not to accord monopolies by contract," Kaspar, supra, at 542, the concerns of Lear are not present in every patent dispute culminating in a consent decree. For example, in Schlegel Mfg. Co. v. USM Corp., 525 F.2d 775 (6th Cir. 1975), the district court entered a consent decree declaring the patent valid and defendant to have infringed it, and permanently enjoined defendant from the unauthorized use of the patent. In a subsequent contempt action for the violation of the terms of the consent decree, the Sixth Circuit gave the decree res judicata effect. The court saw a significant difference between the effect of a consent decree and the doctrine of licensee estoppel abolished in Lear. Lear was not concerned with prior litigation resulting in a consent decree, but with a licensing agreement in which the licensee never was permitted to challenge the validity of the licensor's patent.

In this case, Usamex has had a full opportunity to litigate the issues. In fact, validity of the '808 patent and Usamex's infringement was not "consented to" after discovery, but were judicial determinations. The only thing consented to was the amount of damages and attorney's fees. Usamex's argument that it has not had an opportunity to litigate whether the 33-ft process infringes the patent is without merit, because there are no material differences between the 12-ft and 33-ft processes. Accordingly, Usamex's declaratory judgment action must be dismissed.

In so doing, I have considered materials outside the pleadings; therefore, I must treat Swift's briefs urging dismissal of the declaratory judgment action as a motion for summary judgment. Pursuant to FRCP 56(c), 12(b), and 43(e), I granted Usamex ten days in which to submit additional materials in opposition to summary judgment, and they were timely submitted. After carefully reviewing the record and all materials and briefs submitted by counsel, I find that there are no genuine issues of material fact, and Swift is entitled to judgment as a matter of law.

Summaries of

Swift Chemical Co. v. Usamex Fertilizers, Inc.

United States District Court, E.D. Louisiana
Jun 6, 1980
490 F. Supp. 1343 (E.D. La. 1980)
Case details for

Swift Chemical Co. v. Usamex Fertilizers, Inc.

Case Details


Court:United States District Court, E.D. Louisiana

Date published: Jun 6, 1980


490 F. Supp. 1343 (E.D. La. 1980)

Citing Cases

Mississippi Chemical v. Swift Agr. Chemicals

In a lengthy opinion, the district court denied the motion. Swift Chemical Co. v. Usamex Fertilizers, Inc.,…

Tristar Marine, Inc. v. Port Ship Service, Inc.

These are matters that are addressed to the sound discretion of the trial court, and its ruling . . . will be…