This appeal marks the second time this sad affair of a ruptured pipeline has been before this court. A bifurcated trial was ordered below, and the liability findings and conclusions were affirmed per curiam in an unpublished opinion by a panel of this court in 1976. The decision rendered by the district court in the damages portion of *1167these proceedings, recorded at 468 F.Supp. 802 (E.D.La.1979), prompts the present appeal. An abbreviated summary of pertinent facts is presented here; the factual situation and the attendant relationships among the several parties are more than a little confusing, and the district court opinion adequately sets out the facts of the accident and the course of proceedings below.
Signal Oil & Gas Company, Louisiana Land & Exploration Company, Amerada Hess Corporation, and Marathon Oil Company (“SLAM”) jointly operated an oil production rig in the Gulf of Mexico off the coast of Louisiana. The SLAM pipeline was the conduit for production from the SLAM platform to shore. Sun Oil Company (“Sun”) operated a platform nearby but did not enjoy a similar access to shore. SLAM and Sun negotiated an agreement allowing Sun to hook up to the SLAM pipeline. Sun agreed to indemnify SLAM for any losses it might suffer thereby. Sun contracted with J. Ray McDermott (“McDermott”) to handle this construction. McDermott, in turn agreed to indemnify Sun for any losses incident to the construction. McDermott contracted with WilliamsMcWilliams (“Williams”), the vessel owner, for the use of a barge and crew in aid of its construction activities. McDermott and Williams did not enter into any indemnity agreement.
While operating near the SLAM platform, the Williams barge fouled anchor on an obstruction — the SLAM pipeline, as was shortly seen. Meeting resistance in his efforts to retrieve the anchor and change positions, the barge superintendent, Southon, ordered a “dogging” technique to free it. In “dogging,” the anchor line is drawn tight and secured while the barge is in a trough; when the barge is lifted by the next swell, the power of the sea itself is used to dislodge the anchor. Unfortunately for all concerned, this retrieval also “dislodged” a chunk of SLAM pipeline. SLAM production operations were interrupted and substantial repairs required.
A roundrobin of claims were consolidated for trial. In the liability phase of this bifurcated trial, the judge arrived at the following conclusions: (1) the negligence of the barge superintendent, for which Williams bore full responsibility, was the sole proximate cause of the accident; (2) Williams was consequently liable to SLAM in tort; (3) Sun was liable to SLAM on its contract of indemnity; (4) McDermott was liable on its indemnity agreement to Sun; and (5) Williams was liable for tort indemnity to Sun and McDermott. These findings were affirmed on appeal in an opinion of this court dated November 29, 1976, and are not before us on this appeal.
The tidy circle above, in which the negligent party was to bear the loss, was broken by the judge’s more recent decision on damages that prompts this appeal. The district court found SLAM entitled to full recovery of its loss — $1,116,234.62. The tortfeasor Williams, however, pursuant to 46 U.S.C. § 181 et seq., was entitled to limit its legal liability to the value of the barge, subsequently found by the district judge to be $450,000. The insurance carried by Williams in excess of its first $500,000 of coverage was held not subject to the Louisiana direct action statute. The combined effect of Williams’ resort to the limitations statute and the inaccessibility to suit of its excess insurance coverage produced a maximum recovery, extractable from Williams and its insurers, of $500,000. Plaintiff SLAM consequently turned to indemnitor Sun for relief and Sun to McDermott. McDermott, left holding the bag with instructions to fill it, is the chief appellant here. It maintains that the district court erred (1) in allowing Williams to limit its liability for the results of this accident; (2) in allowing Williams an improper filing of this limitation; (3) in arriving at an incorrect valuation of the barge; (4) in awarding SLAM recovery of initial, unsuccessful repair costs; (5) in finding the umbrella insurance policy covering Williams unreachable under the Louisiana direct action statute; and (6) in finding McDermott liable on its contract of indemnity to Sun. Plaintiff SLAM, awarded recovery from Williams, its insurers, and *1168McDermott, complains of an insufficient award of interest on its judgment and seeks attorneys’ fees expended in securing its tort and indemnity recoveries.
We find no merit in any of McDermott’s objections or in SLAM’s complaint of an improper rate of interest; accordingly, we affirm the district court on those questions. We remand for computation and award to plaintiff SLAM of its attorneys’ fees incurred in prosecuting its tort claim against Williams.
I. Williams' Right to Limit its Liability.
McDermott, a nonnegligent party, has been cast in judgment for over half the award to the SLAM group. McDermott implores this court to reverse a result “based upon a fragile, interrelated structure of erroneous findings and conclusions, reversal of any one of which will cause the structure to topple.” We reject that characterization of the judgment below. McDermott’s plight stems from a contract willingly signed and a federal statute of ancient and venerable origin. In rejecting its claims, we turn first to its attacks on the statute’s application to this case.
McDermott does not dispute that the nature of the accident and the parties makes this incident generally the kind in which the vessel owner may move to limit its liability. Rather, the argument that this limitation defense should have been denied Williams relies on McDermott’s allegations that by its actions Williams has incurred liability that traditionally escapes the statute’s protective sweep. McDermott argues that Williams’ actual liability should exceed the limitation amount under the “Personal Contract Doctrine.”
Properly invoked, this doctrine deprives a shipowner of the benefits of limitation and exposes him to the full burden of liability his actions produced. Surveying the history and purpose of the limitations statute, the Supreme Court in Richardson v. Harmon, 222 U.S. 96, 107, 32 S.Ct. 27, 30, 56 L.Ed. 110 (1911), wrote:
Thus construed, the section harmonizes with the policy of limiting the owner’s risk to his interest in the ship in respect of all claims arising out of the conduct of the master and crew, whether the liability be strictly maritime or from a tort nonmantime, but leaves him liable for his own fault, neglect, and contracts.
McDermott argues that Williams now must answer for its “own fault, neglect, and contracts” apart from its vicarious liability for the negligence of its barge superintendent, which latter liability, has been limited under the federal statute. The elements of this alleged fault are: (1) provision to McDermott of an unseaworthy vessel; (2) breach of an implied warranty to perform services in a workmanlike manner; and (3) breach of Williams’ warranty to McDermott that it enjoyed “effectual” insurance coverage.
Before further considering its application to this case, we examine briefly the nature of the Personal Contract Doctrine. The cryptic reference in Richardson to the vessel owner’s inability to limit liability for his own fault, neglect, and contracts, rather than those of his captain arid crew, does not offer much guidance. This language could have been read to mean that limitation was available only against tort claims — that such a contractual undertaking by a vessel owner as with a shipper, for example, was not subject to the statute despite satisfaction of the stated requirements for limitation. As noted by G. Gilmore & C. Black, The Law of Admiralty § 10-26, at 899 (2d ed. 1975) (hereinafter Gilmore & Black), subsequent cases have rendered such a reading of the doctrine improper. “It is by no means clear exactly what a ‘personal contract’ is, but it is clear that not all contracts are ‘personal’, even though entered into personally by the shipowner.” Id. Analysis of the cases delineating the types of contracts that fall within the designation “personal” is of scant assistance to our inquiry. For example, bills of lading have been held not to be personal contracts, Earle & Stoddart, Inc. v. Ellerman’s Wilson Line, Ltd., 287 U.S. 420, 53 S.Ct. 200, 77 L.Ed. 403 (1932); Gilmore & Black, § 10-26 at 899, Charter parties, on the other hand, *1169have been four times treated as personal contracts by the Supreme Court in denying limitation to vessel owners. Gilmore & Black, id. The parties have not directed our attention to, nor have we found, any case discussing the proper characterization of this contract for the services of a work-over barge in offshore oil production.
The initial characterization, moreover, does not necessarily dispose of the question whether liability resulting from a vessel owner’s contractual relationships is properly subject to limitation under the statute: “To call a given type of contract ‘personal’ does not automatically lead to the conclusion that the shipowner will under all circumstances be denied the right to limit against the claims of his promisee.” Gilmore & Black, § 10-27 at 904. Learned Hand, sitting as a district judge in The Soerstad, 257 F. 130 (S.D.N.Y.1919), suggested a narrow reading of the personal contract doctrine, restricting its application to situations in which the breach, as well as the execution, of the contract could be said to be “personal” to the vessel owner. His approach in The Soerstad has gathered support in subsequent cases, including dicta considered supportive by Gilmore and Black, see § 10-27 at 904-05, in Cullen Fuel Co. v. W. E. Hedger, Inc., 290 U.S. 82, 88-89, 54 S.Ct. 10, 11, 78 L.Ed. 189 (1933).1 There is no more recent or more certain authority. This case, however, does not require us to fill that gap. Whether The Soerstad’s narrowing of the personal contract doctrine is correct or appropriate need not concern us here; McDermott, the party seeking application of the doctrine, presents its grounds for Williams’ liability in post-Soerstad terms: breaches of warranties of seaworthiness, performance in a workmanlike manner, and provision of “effectual” insurance coverage. McDermott makes no claim that Williams should be denied limitation simply because it personally agreed to provide the vessel in aid of the construction operation. We leave, then, further consideration of the scope of the personal contract doctrine to scholarly commentators and to judges directly confronting the issue.
Williams is entitled to limit its liability for the results of the rupture of the SLAM pipeline unless its culpability falls within the personal contract doctrine. We have examined McDermott’s grounds for disallowing Williams’ limitation here and find them unpersuasive. As a preliminary matter, we note that McDermott’s attempt in this damages portion of the proceeding to attribute fault to Williams beyond the negligence of the barge superintendent, found the cause of the accident in the liability hearing, may run afoul of the “law of the case” doctrine. This doctrine, “a restriction self-imposed by the courts on themselves in the interests of judicial efficiency, generally operates to preclude a reexamination of issues decided on appeal, either by the district court on remand or by the appellate court itself upon a subsequent appeal.” Conway v. Chemical Leaman Tank Lines, Inc., 644 F.2d 1059, 1061 (5th Cir. 1981). Unlike res judicata, law of the case does not operate to bar subsequent consideration of matters that could have been, but were not, raised and resolved in the earlier proceeding. Like the doctrine of collateral estoppel, it prohibits relitigation only of matters decided, expressly or by necessary implication, in the prior appellate proceeding. Id. at 1062.
At least the first two of McDermott’s asserted grounds for invocation of the personal contract doctrine (provision of an unseaworthy vessel and breach of warranty to perform services in a workmanlike manner) may be barred by the law of the case. The trial judge found in the liability phase of this bifurcated proceeding that the negligent action of the barge superintendent in “dogging” the anchor free was the sole *1170proximate cause of this injury and that the barge W-701 was a seaworthy vessel. These creative warranty arguments were either rejected by the trial judge or never presented to him in that proceeding, where they could properly have been argued. The liability findings have been affirmed by this court and cannot be disturbed. Generally, in applying the law of the case doctrine, it makes a difference whether the challenged issue or claim was actually presented to the earlier courts for their decisions. Whether or not raised in the court below, however, these two warranty claims must be considered rejected by necessary implication in the trial court’s conclusion that Southon’s negligence alone caused the accident.
As we have noted, however, the law of the case doctrine is a restriction “imposed by the courts on themselves in the interests of judicial efficiency.” Id. at 1061. Considering the uncertain application of the doctrine to this case and the substantively weak grounds upon which McDermott bases its attempt to invoke the “personal contract” exception, “judicial efficiency” is best served here by disposition of McDermott’s arguments on their merits.
Even if the first two warranties were made and could theoretically form the basis of a recovery beyond the limitation fund, they were not here breached. McDermott’s seaworthiness and workmanlike performance arguments would impose on Williams a responsibility more direct than respondeat superior for the barge superintendent’s negligent act. In loosing on an unsuspecting world an inadequately trained and poorly schooled master of the vessel, the argument runs, Williams has incurred a liability that should penetrate the statute’s protective cover; it must answer for its “own fault, neglect, and contracts.” Whether the law might support McDermott’s contentions need not concern us, since the facts clearly do not. The trial judge found that “the on-the-job training which was provided to him [Southon, the barge superintendent] created an obviously capable man with an excellent reputation ... . ” 468 F.Supp. at 815. In ten years as a barge superintendent Southon had suffered only one similar incident. He was an experienced, able captain with a good safety record. As the district court found, “it is clear that Captain Southon was not incompetent but merely negligent.” 468 F.Supp. at 814. Furthermore, despite McDermott’s vigorous argument to the contrary, we find that Williams violated no duty in failing to promulgate formal written procedures for freeing anchors and unfouling lines. Its reliance on the skill and judgment of its superintendents does not become negligently misplaced merely because of this accident. The attempt to bootstrap the negligence of Southon into a breach of warranties of seaworthiness and workmanlike performance is insupportable in fact. From the results of its employees’ negligence, Williams is entitled to the full benefit of the limitation statute.
The third claim of breach of warranty, that Williams failed to secure its promised insurance coverage, is of a different stripe. Here at last is a claim of fault unconnected with Southon’s negligent conduct. This ground for denial of limitation arises from an alleged contractual breach related to the accident only by the fact that, absent that accident, Williams’ insurance coverage would simply not not have been an issue. As a theoretical matter, a claim of this nature might well be the source of a recovery outside the scope of the limitation of liability statute. It could perhaps have formed the basis for an independent suit against Williams; that avenue has apparently been forsaken, however, as Williams’ alleged breach is raised for the first time on this appeal as a reason for denying limitation under the statute. While differently based than the first two warranty charges, this claim is entitled to no more sympathetic a reception.
The warranty of “effectual insurance,” if such there be, was simply not breached here. Williams purchased all the coverage indicated in its notices of insurance to McDermott. That insurance was in effect and available to satisfy appropriate claims. Unfortunately for McDermott, a statute *1171embodying the congressional policy of encouraging the mid-nineteenth century American shipping industry allowed Williams and its insurers to cut the losses they otherwise would have suffered. Under these circumstances, the trial court’s decision that the umbrella insurance policy held by Williams’ parent corporation could not be reached directly, see Part IV, infra, does make this coverage in some sense “ineffecas leaving McDermott liable for an unanticipated loss. McDermott’s quarrel, however, is with Congress and the Louisiana legislature, not with Williams. In this instance, McDermott stands exactly where it would in any one of the 48 states that allow no direct action against insurers. Insurance in those states is clearly not “ineffectual” merely because tortfeasors alone, and not their carriers, may be sued.2 tual,”
*1172After review of the arguments raised to avoid application of the limitation statute here,' we arrive at the same conclusion reached by the district judge: Williams is entitled to limit its liability to the value of the barge W-701 at the time of the accident. Resolution of that question, however, does not end our inquiry into the application of the statute to the facts of this case. We turn next to McDermott’s challenge to the trial court’s decision about the value of the offending barge.
II. Value of the Barge W-701.
46 U.S.C. § 183 provides that the amount recoverable from the owner of a vessel, under the circumstances stipulated in the statute, cannot “exceed the amount or value of the interest of such owner in such vessel, and her freight then pending.” Williams’ successful limitation of its liability fixed its exposure at a level well below plaintiffs’ actual loss. Applying the statutory formula to the W-701, the trial judge found the 1969 value of the barge to be $450,000. With insurance coverage of only $500,000 subject to the Louisiana direct action statute, the liability of Williams and its insurers could consequently not exceed the half-million dollar figure. Plaintiffs were found entitled to recover over $1.1 million for their loss. McDermott, as ultimate indemnitor without recourse, was held liable for the difference. To avoid this notable loss, McDermott argues that the value placed by the trial judge on the barge was far too low; for every dollar above $500,000 added to the value of the vessel, McDermott’s liability drops proportionately.
To succeed in its effort on appeal, McDermott must demonstrate that the trial judge’s finding on the value of the barge was “clearly erroneous.” McAllister v. United States, 348 U.S. 19, 75 S.Ct. 6, 99 L.Ed. 20 (1954). “No greater scope of review is exercised by the appellate tribunals in admiralty cases than they exercise under Rule 52(a) of the Federal Rules of Civil Procedure [citations omitted].” Id. at 20, 75 S.Ct. at 8. McDermott fails in this difficult showing. There was ample evidence from which the district judge could reasonably have derived a figure of $450,000 for the value of the barge. Three expert witnesses, two for Williams and one for McDermott, testified on the value of the W-701. The Williams’ experts agreed on the $450,-000 figure; McDermott’s reckoned an amount in the neighborhood of $1.3 million. The careful consideration given this conflicting evidence by the trial court, reflected in its opinion, 468 F.Supp. at 809-11, convinces us that McDermott’s shot here is far wide of the mark. On the entire evidence, we are not “ ‘left with a definite and firm conviction that a mistake has been committed.’ ” McAllister, 348 U.S. at 20, 75 S.Ct. at 8, quoting United States v. Oregon State Medical Society, 343 U.S. 326, 339, 72 S.Ct. 690, 698, 96 L.Ed. 978 (1952), and United States v. United States Gypsum Co., 333 U.S. 364, 395, 68 S.Ct. 525, 541, 92 L.Ed. 746 (1948).
One argument against acceptance of the $450,000 figure merits brief notice. McDermott points to evidence introduced by Williams at trial that on the date of the accident the barge was insured for $500,000. It argues that under the statute the insured *1173value of a vessel should be the starting point for its valuation and that the correct figure should perforce range upward from there. We are offered neither authority nor persuasive argument why any presumption based on insurance coverage should lie, much less a presumption of undervaluation as opposed to overvaluation in securing this insurance. It is common knowledge that, wittingly or unwittingly, items of property are frequently over- or under-insured. Thus, while the amount of coverage on an item may be relevant to establishing its value, it clearly does not lay a floor beneath it. The value of the barge W-701 as found by the trial judge will remain undisturbed on this appeal.
III. Alleged Defects in Seeking Limitation of Liability.
McDermott’s final challenge to the limitation statute’s application to this suit is a claim that Williams untimely and inadequately sought the statute’s protection. The statute offers two avenues of limitation. A vessel owner may petition the district court, pursuant to 46 U.S.C. § 185, for limitation of liability within six months of written notification to it of a possible claim; under section 185, the owner’s petition in a particular case may well be the first filing in court. Section 183, which established the substantive right of a vessel owner to limitation, also allows limitation to be pled as a defense in answer to an earlier filed damage suit. Williams took this route.
McDermott’s first procedural objection is that Williams raised the limitation defense in an untimely manner — in its March 1971 answer to McDermott’s cross-claim rather than within six months of notice of the claim, as stipulated in section 185. Acceptance of this argument would require a finding that the time requirement for filing a section 185 petition carries over into an assertion of limitation in answer to a complaint. McDermott’s approach does not lack some support in the cases. See, e. g., Cantwell v. Meade, 120 F.Supp. 406, 407 (E.D.N.Y.1954) (“I do not believe that it was the intention of Congress to limit the application of section 185 of Title 46 U.S. C.A. to cases in which the shipowner petitioned for limitation. ... It is my opinion that sections 183(a) and 185 should be read together .... ”). The better answer, however, is that adopted in The Chickie, 141 F.2d 80, 85 (3d Cir. 1944):
We think it is abundantly clear, in view of the historical development of the statute, that the time limitation put in § 185 by the 1936 amendment and made expressly applicable to a proceeding by petition at the suit of the owner, is not to be read into § 183(a) by implication .... To read the time limitation into § 183(a) would militate against the well settled rule that the courts are to construe the statute liberally. It would tend, without an express declaration by Congress, to increase the economic burden of ship owners, thus placing them at the competitive disadvantage which the statute sought to avoid, [citations omitted]. It would also require that once the sole injured party gave notice of claim, but delayed filing suit, the ship owner must of necessity bring a proceeding within six months to save his rights, since a libel may not be filed against him by the injured party for more than six months after notice.
See also Deep Sea Tankers v. The Long Branch, 258 F.2d 757 (2d Cir. 1958), cert. denied, 358 U.S. 933, 79 S.Ct. 316, 3 L.Ed.2d 305 (1959); Odegard v. E. Quist, Inc., 199 F.Supp. 449, 451 (E.D.N.Y.1961) (pointing-out that Cantwell v. Meade, supra, must be considered overruled by Deep Sea Tankers ). At least since The Scotland, 105 U.S. 24, 26 L.Ed. 1001 (1881), the right of a vessel owner to raise limitation for the first time in his answer to a complaint has been recognized. It would often frustrate that right to allow it only to that defendant who is sued within six months of receiving written notice of a claim. We find Williams’ invocation of the limitation protection timely, coming as it did within the time limit allowed for answers in the Federal Rules of Civil Procedure.
The core of McDermott’s second procedural objection is that, when limitation is *1174raised in the answer under section 183, the defendant vessel owner must create multiple limitation funds for multiple claims — or, at a minimum, for multiple claims that, as here, are of more than one character, where Williams was sued for tort damages by SLAM and for tort indemnity by codefendants Sun and McDermott. Accepting this argument would require, at minimum, doubling the amount for which Williams would be liable and would go far toward erasing the sum owed by McDermott as indemnitor.
McDermott offers neither substantial authority nor much in the language of the statute for this approach. On the other hand, Williams locates no analogous cases (that is, ones involving a mixture of tort and indemnity claims) directly rejecting McDermott’s formulation. After examination of the few existing cases and of the structure of the statute, we conclude that a vessel owner’s decision to plead limitation as a defense in a single action with multiple claimants under section 183, rather than to petition under section 185, does not open it to greater total liability.
McDermott accurately cites The West Point, 83 F.Supp. 680 (E.D.Va.1949), as supporting its position:
[I]f there is more than one claim, then the defendant owner can confine his aggregate liability on all the claims to the single value of the vessel, or his interest therein, only by the petition prescribed by the statute. R.S. 4285, 46 U.S.C.A. § 185. If he does not follow the procedural steps of the statute, but uses the answer to raise such defense, the owner may be held personally liable on each claim to the extent of the value of his ship, or his interest in her.
Id. at 681. The cursory resolution of the issue in The West Point is less satisfactory, however, than that supplied by another district court in a multiple claim situation. In Blunk v. Wilson Line of Washington, Inc., 341 F.Supp. 1345 (N.D.Ohio 1972), a defendant vessel owner pled limitation in a single answer to forty consolidated claims arising from physical and psychic injuries claimed to have been suffered by school children on a vessel that ran aground. Recognizing the scant comment on this issue since The West Point, the court in Blunk conducted a searching analysis of the development of the limitation statutes, concluding that the single fund limited to the vessel’s value was sufficient to satisfy all the claims. Id. at 1349-50. We are likewise convinced and find it insignificant that all forty claims in Blunk were of an identical nature, while here differing substantive claims are raised.
An adoption of McDermott’s approach would not follow the statutory scheme; it would make the section 183 alternative a most risky and unattractive one. Requiring multiple funds would defy the logic in statutorily encouraging the shipping industry by placing a relatively low cap on potential liability. “Statutory provisions ... for limitation of liability should be construed liberally in order to effectuate their beneficent purposes.” Larsen v. Northland Transportation Co., 292 U.S. 20, 24, 54 S.Ct. 584, 585, 78 L.Ed. 1096 (1934). Individually manageable figures, if multiplied, could aggregate a liability mammoth in relation to the value of the vessel involved. In declining so to construe the statute, we recognize a contrary schematic argument — that limiting the satisfaction of all claims to a single fund when limitation is raised as a defense under section 183 arguably removes incentives for vessel owners to avail themselves of section 185 petition procedures, a petitioning vessel owner being liable for the costs of that limitation proceeding itself.
We note that the statute and cases indicate no preference for either limitation alternative; each is to be, within the confine of statutory requirements, equally available and equally attractive to the vessel owner. That no incentive is given to choose one over the other is not troubling. For whatever reasons, an owner may prefer to strike first in establishing the outside limits of its liability by petition under section 185. Requiring multiple funds for multiple claims would limit the practical availability of the section 183 alternative for the prudent owner to the situation where there is a single claimant only. A factual situation not un*1175like this one suggests the unsatisfactory nature of such an interpretation. For the results of its employee’s negligence, a vessel owner is confronted with a single tort claim. Unaware, perhaps, of the existence of indemnity agreements or uncertain of their application to the particular incident, the owner fails to foresee possible tort indemnity claims and delays assertion of limitation until its answer to plaintiff’s complaint, filed after the six-month period in section 185. Now, unable to petition under section 185, the defendant vessel owner is served with cross-claims by plaintiff’s indemnitors. Acceptance of McDermott’s multiple claim-multiple fund argument would subject the defendant in this hypothetical case, as it would Williams, to a liability not contemplated by the statute. Without more and without implying any particular limitation, we find that in such a case, where the multiple claims are properly advanced in a single proceeding and some are in a sense derivative of others, the defendant vessel owner’s liability is limited to the single value of its vessel.
IV. Inaccessibility to Direct Action of Williams’ Umbrella Insurance Coverage.
That Williams was able to limit its liability to the value of the barge W-701 did not alone remit McDermott to its unenviable position here. The value assigned to the barge was $450,000, but Williams enjoyed over $5 million of insurance coverage. The Louisiana direct action statute, La.Rev.Stat. Ann. § 22:655, allows a plaintiff to sue the tortfeasor’s insurer directly. In a Louisiana maritime tort case, that the vessel owner may limit its liability is a concern to the plaintiff only insofar as plaintiff’s injuries exceed the greater of the value of the vessel or the amount of insurance that can be reached under the direct action statute.
A direct action against the insurer is available when the insurance contract at issue was written or delivered in the State of Louisiana or if the injury occurred within the state. Webb v. Zurich Insurance Co., 251 La. 558, 205 So.2d 398 (1967). The first two policies held by Williams on its operations, totalling $500,000, were clearly subject to direct action; Williams and its insurers readily admit that these policies were delivered to Williams in Louisiana. At the time of the accident Williams enjoyed additional coverage under an umbrella policy secured by its parent corporation, Zapata Norness, from a British insurer; this umbrella policy was to cover the operations of a score of Zapata divisions and subsidiaries in various locations around the globe. It is this uppermost layer of coverage, totalling $5 million, that McDermott would like held subject to direct action, thus removing McDermott from its costly position.
McDermott’s efforts to bring this third policy within the Louisiana statute are unpersuasive. It is undisputed that the accident occurred in the Gulf of Mexico, well beyond the boundaries of the state. The policy was written in London and delivered to Zapata’s corporate headquarters in Houston, Texas. McDermott attempts an argument on “constructive delivery” but is unable, as pointed out by the district judge, 468 F.Supp. at 816, to find any legal or factual support for that position. McDermott argues that it contravenes public policy to allow a company with significant and ongoing business contacts with the State of Louisiana to “evade” the direct action statute by purchase and delivery of insurance by an out-of-state affiliate; what merits, if any, that argument possesses are plainly addressed to the wrong audience. We take the statute as written by the legislature and reject McDermott’s theories.
V. McDermott’s Liability to Sun on Their Indemnity Agreement.
[II] In the liability phase of the trial, the district court found McDermott liable on its contract of indemnity to Sun, subject to an exception clause, paragraph 15 of that agreement. Through this agreement McDermott is now made to assume Sun’s burden of indemnifying SLAM for its unrecovered loss. McDermott’s objections to this obligation on appeal come substantively too little and procedurally too late. The liability findings in the court below have *1176been affirmed previously by this court and cannot now be disturbed. McDermott makes no argument that paragraph 15, the only question related to its indemnity liability to Sun left open in the earlier proceeding, was misapplied or misunderstood below. That exception related to certain losses, chiefly “remote or consequential damages,” that McDermott was expressly not agreeing to satisfy. Rather than complaining, McDermott applauds the district judge’s finding that it escapes liability, pursuant to paragraph 15, for over $71,000 in certain consequential damages suffered by SLAM. The paragraph 15 exclusion has no effect on McDermott’s general indemnity liability, and thus arguments against imposition of such liability are laid to rest by the law of the case.
VI. Recovery by SLAM of its Initial Repair Costs.
McDermott’s final complaint on appeal is of the inclusion in the $1.1 million award to SLAM of $77,000 for expenses incurred in an initial, unsuccessful attempt by SLAM to repair the ruptured pipeline. McDermott argues, apparently for the first time in this court, that in attempting this repair SLAM was negligent in failing adequately to test a valve. Once in place, the valve began leaking, necessitating further repairs. McDermott asserts that in allowing this amount the judgment permits plaintiffs to profit from their own negligence. Aside from the small matter that contributory negligence was not alleged by McDermott below, the facts simply fail to support its claim that plaintiffs were negligent in this regard. The repair effort was a tricky one, necessitating some innovative engineering techniques. SLAM understandably wanted to move as fast as possible to hold down its losses and thus those of the potential defendants. Tests done on the valve failed to reveal a broken fitting; other tests that might have revealed the problem arguably were available. Substantial evidence supports the trial court’s findings on the amount of damages properly recoverable by SLAM; the award is fully affirmed.
VII. Proper Rate of Interest on SLAM’s Recovery.
Having considered and rejected McDermott’s various points on appeal, we turn now to those raised by SLAM as cross-appellant.
The trial court awarded SLAM over $1.1 million, with seven percent interest to run on that amount from December 20, 1969, the date of the accident. No one argues that the date from which interest runs was improperly set; all accept this court’s statement in Complaint of M/V Vulcan, 553 F.2d 489, 490 (5th Cir. 1977), that “in admiralty cases the award of prejudgment interest from the date of loss is the rule rather than the exception.” That some interest was due is likewise fully accepted. “Discretion to deny interest must be based on the existence of peculiar circumstances, none of which are present in this case.” Id. at 490-91. The dispute, such as it is, is over the proper rate of interest.
SLAM argues here, as it did unsuccessfully in the court below, that its alleged cost of borrowing money during the period since the accident should have been awarded, it as interest — 13.4 percent rather than 7 percent. Williams now asserts in a reply brief that seven percent is itself too high and that, under a proper construction of the applicable statutes, five percent is the appropriate rate. McDermott says that seven percent is just right. McDermott’s original and reply briefs were filed with this court before Williams raised its argument that the seven percent figure erred on the side of generosity; McDermott, as the one shouldering most of this burden, would undoubtedly accept a reduction from seven to five percent.
Neither argument — that the trial judge set the figure too low or too high— prevails. Williams’ request that the rate be lowered is quickly disposed of. This argument is made for the first time in Williams’ second appellate brief. There is no indication that the trial judge’s alleged mistake in relying on particular statutory interest pro*1177visions rather than others was ever brought to his attention. There is no reason to consider this argument now.
SLAM relies on the above-cited Vulcan decision for its claim to 13.4 percent interest. In Vulcan this court found an award of interest at 12 percent (that particular plaintiff’s cost of borrowing) as opposed to the state statutory rate not to be an abuse of that trial judge’s discretion. There exist two problems with SLAM’s sole reliance on Vulcan here: (1) we see no reason to turn the holding of that case inside out to make an award at a rate other than plaintiff’s actual cost of borrowing an abuse of discretion; and (2) SLAM, unlike the Vulcan plaintiff, offered no proof that it borrowed any money and incurred these higher interest costs. The decision of the trial court to award prejudgment interest at a rate of seven percent is affirmed.
VIII. Indemnification of SLAM’s Attorneys' Fees.
In filing suit against Williams for maritime tort recovery and in seeking indemnification from Sun, SLAM incurred substantial attorneys’ fees. SLAM sought to recover these fees from Sun on Sun’s promise to “indemnify and hold [SLAM] harmless from and against any and all loss, cost, damage, expense, claims, actions and liability on account of any and all damage to and loss of or destruction of any property belonging to [SLAM] . . . arising out of or in connection with the installation, repair and maintenance of said riser on platform A.” Because the litany of anticipated liabilities (“loss, cost, damage, expense, claims, actions and liability”) did not include the words “attorneys’ fees,” the district court denied SLAM that item of recovery.
Initially, SLAM did not claim for these attorneys’ fees in the district court; that omission was cured in a post-trial submission, and by leave of that court the record on appeal includes an affidavit from SLAM’s attorneys of time invested in this matter and a suggested hourly rate. No procedural irregularity was the rationale for the district court’s action, as that court clearly denied the award on the merits. The issue was not waived and is properly before us on appeal. SLAM asks this court to find the unwillingness of the district court to order indemnification of SLAM’s fees to be error. Sun and McDermott oppose any reversal of the district court’s judgment in this regard.
The indemnity agreement between > Sun and McDermott expressly obligated McDermott to cover legal fees and expenses incurred in a situation of this kind, and the district judge awarded Sun this reimbursement. The inclusion of that express obligation in the Sun-McDermott agreement emphasized for the district court the apparent inadequacy of the SLAM-Sun indemnity in relying on more general language.
The parties skirmish lightly about the applicable law on this question of indemnity contract interpretation within the context of a maritime tort suit. Mercifully, the parties expend little time or energy on that discussion. They apparently recognize that the proper result in the case does not hinge on resolution of this choice-of-law question. The Louisiana cases cited to us lead to a conclusion no different than those applying the law of another jurisdiction. Indeed, the cases point to no very clear conclusion at all; on at least one of the two questions presented by this general request for attorneys’ fees, we are directed to no clear authority from any jurisdiction. On the other, three analogous cases are available from different circuits and thus are not controlling on our disposition, whether as constructions of state or of federal law. Finding no close cases that settle the issues or bind us in our construction of the contract, we have considered cases from within and without Louisiana for whatever persuasive authority they may carry.
In argument before this court on SLAM’s request for attorneys’ fees in the action below, the parties do not distinguish between fees incurred in SLAM’s tort claim against Williams and those resulting from its efforts to secure payment on Sun’s (and through Sun, McDermott’s) indemnity. We have not been directed to a single case that *1178dealt with a plaintiff indemnitee’s recovery from its indemnitor of attorneys’ fees incurred in prosecuting plaintiff’s claim against the tortfeasor. The cases relied on by SLAM in arguing for a recovery are concerned with the more usual situation— the indemnitee as defendant in a tort suit is allowed, as part of its indemnity recovery from its indemnitor, attorneys’ fees incurred in defending against the tort claim. This situation, in which the indemnitee plaintiff turns to its indemnitor for recovery of attorneys’ fees, is not, judging by the absence of cases, a common occurrence. In those cases allowing recovery by the defendant indemnitee, this court and others have found attorneys’ fees properly recoverable in reliance on language in indemnity agreements no more specific than that present here. See, e. g., Olsen v. Shell Oil Co., 595 F.2d 1099 (5th Cir. 1979); Loffland Brothers Co. v. Roberts, 386 F.2d 540 (5th Cir. 1967), cert. denied, 389 U.S. 1040, 88 S.Ct. 778, 19 L.Ed.2d 830 (1968); B & G Electric Co. v. G. E. Bass & Co., 252 F.2d 698 (5th Cir. 1958), cert. denied, 357 U.S. 931, 78 S.Ct. 1372, 2 L.Ed.2d 1371 (1958). See also 42 C.J.S. Indemnity § 13d.
The parties do not directly address the question whether recovery by a plaintiff indemnitee should be as readily available as that of a defendant indemnitee. SLAM offers eases in which defendants have recovered attorneys’ fees on general indemnity language and stops; Sun and McDermott are satisfied to point out the factual distinction between this case and those without pointing out a legally significant difference.
We conclude that recovery of attorneys’ fees by the indemnitee in this situation, the reverse of the more typical one, is reasonable and within the probable intent of the parties. While not the case here, designation as plaintiff or defendant in a particular lawsuit in which fault is unclear or shared is often little more than the result of a race to the courthouse. In either case, as plaintiff or defendant, the indemnitee’s legal fees in the tort action are equally related to the accident. For its attorneys’ efforts in prosecuting the tort claim against Williams, SLAM is entitled to recover attorneys’ fees on its indemnity agreement.3
SLAM’s theory for recovery of indemnity for fees incurred in efforts to secure payment on that indemnity agreement has received previous unfavorable judicial scrutiny. The Eighth, Ninth, and Tenth Circuits have remarked that, under a general indemnity agreement like that here, the indemnitee enjoys no right to recover its legal fees incurred in establishing its right to indemnification. Bagby v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 491 F.2d 192, 198 n.9 (8th Cir. 1974) (“there should be no recovery for fees and expenses incurred in establishing the right to indemnity”); Flunker v. United States, 528 F.2d 239, 246 (9th Cir. 1975) (“The award of fees and expenses does not encompass costs necessary to establish the right against the alleged indemnitor”); Vallejos v. C. E. Glass Co., 583 F.2d 507, 510 (10th Cir. 1978) (“It is true that in connection with indemnity claims recovery may generally be had for attorneys’ fees and expenses incurred in defense against the principal claim, but not for those incurred in establishing the right of indemnity”). District court opinions within this circuit have likewise so held. See, e. g., Williams v. California Co., 289 F.Supp. 376, 380 (E.D.La.1968) (“There can be no recovery for attorneys’ services and expenses incurred in establishing the right to indemnification”). We find no Fifth Circuit opinion directly so holding. In a roughly analogous *1179case, however, Strachan Shipping Co. v. Koninklyke Nederlandsche S.M., N.V., 324 F.2d 746, 746-47 (5th Cir. 1963) (not a contractual indemnity case), the court “fully concur[red] in the clear and logical opinion” of the district judge, who “awarded to the shipowner [the defendant in the tort suit] that part of his attorneys’ fees that were incurred in defending against the longshoreman’s [plaintiff’s] claim, but refused to award any attorneys’ fees incurred in ‘getting the stevedore into the case and establishing that the stevedore, rather than the shipowner, was responsible.’ ” A decision of the Louisiana Supreme Court is in accord:
But the fee of $250, claimed by plaintiff company for services of attorneys in the present suit on the indemnity contract, is not a fee incurred “by reason of or in consequence of the bond,” since the suit is not against the surety, or the contractor, but is brought solely by the surety in its own interest against the indemnitor.
This fee was properly rejected by the lower court.
Fidelity & Casualty Co. of New York v. Bisso, 179 La. 56, 153 So. 19, 21 (La.1934). We conclude that any recovery of attorneys’ fees for services expended in establishing SLAM’s right to recover on its indemnity agreement was likewise “properly rejected by the lower court.” We decline to hold it within the contemplation of the contracting parties that the indemnitor would have to be taken to court to insure satisfaction of his agreed upon payment.
IX. Conclusion.
For the reasons discussed above, we affirm the district court’s findings and conclusions on Williams’ right here to limit its liability; the value of the barge W-701; the inaccessibility to direct action of Williams’ excess insurance policy; McDermott’s liability to Sun on its indemnity agreement; an allowance to SLAM of a recovery for its unsuccessful repair efforts; and an award of interest on the judgment at seven percent. We affirm the district court’s denial of attorneys’ fees to SLAM insofar as those fees were incurred in establishing SLAM’s right to indemnification. We reverse the court’s denial of such fees for efforts expended in prosecuting the tort claim against Williams. We remand this issue to the district court for computation of the amount of recovery to which SLAM is entitled in light of this opinion.
AFFIRMED in part; REVERSED and REMANDED.
. The petitioner urges that the denial of limitation in cases like this will sweep away much of the protections afforded to ship owners by the acts of Congress. But this view disregards the nature of the warranty. The fitness of the ship at the moment of breaking ground is the matter warranted, and not her suitability under conditions thereafter arising which are beyond the owner’s control.
(citing, among other cases, The Soerstad, supra).
. In reaching this conclusion, we have considered the position advanced in Judge Tate’s partially dissenting opinion but are, for several reasons, unable to espouse it. As noted above, McDermott apparently raised this alleged “warranty of effectual insurance” for the first time on this appeal. In its original answer and cross-claim against Williams filed in February 1971, McDermott alleged generally “that it is entitled to recover over and against WilliamsMcWilliams Company and its insurers for breach of contract .... ” No more specific allegation was made of a breach of any warranty of effectual insurance.
In the liability phase of these proceedings, the district judge determined that Williams’ liability to McDermott lay in tort (specifically, maritime tort indemnity) and not contract; this finding alone would possibly conclude the matter on principles of the law of the case. In the damages phase of the trial, when the issue of limitation of liability was directly addressed and all counter-arguments were to be raised, the record is silent on any allegations of such breach. No mention was made of this warranty or its breach in the pretrial order or at the trial on damages. McDermott raised the “personal contract” exception to limitation of liability in the lower court proceeding but did so in only a very general manner; unlike its listing of three specific breaches of warranties on appeal, in the district court McDermott appears to have rested on the mere existence of a contract with Williams for its invocation of the personal contract doctrine. McDermott’s counsel argued:
On behalf of McDermott, however, we’re in a little bit different position in that we are urging that since the W-701 was operating under a contract with McDermott, and I think we sited [sic] some law to the court on that proposition, that they’re not entitled, they were not entitled to limit as to McDermott, regardless as to what they might otherwise be able to do with regard to the other parties. Because we’re in a contract relationship, whereas they’re in direct relationship with the other parties.
It may be argued in possible mitigation of any procedural irregularity on McDermott’s part that this alleged breach of warranty as grounds for invoking the personal contract doctrine could not have been recognized until after the district court’s decision on damages and limitations; that explanation highlights the substantive difficulty we have with the dissent’s position.
The dissent’s conclusion is concisely stated in its final paragraph: “Williams’ assertion of this [limitation of liability] defense as against McDermott is a breach of its ‘personal’ warranty to secure and maintain insurance coverage to protect McDermott against Williams-caused liability.” At 1181. We are unable to discern from that opinion, from McDermott’s brief, or from a review of the record where this warranty was made by Williams to McDermott. The offer into evidence of the underlying 1965 agreement between Williams and McDermott was withdrawn by McDermott’s counsel. Any light it might have shed was thus lost. The 1969 rental agreement between Williams and McDermott contains no express warranties about insurance. Attached and made a portion of that agreement was a schedule listing the insurance coverage on Williams’ operations. During the course of their business dealings, Williams provided McDermott with several Advices of Insurance, updating the coverage information. Undoubtedly, prudent businessmen consider such insurance information in decisions about their operations; acknowledging this, however, we cannot go along with the dissent’s conclusions:
[Nevertheless, McDermott entered into its contract with Williams upon the latter’s warranty that it had excess insurance to protect against all claims worldwide up to a total of $3 million. Obviously, this would be a consequential factor in McDermott’s decision to accept Williams’ rate offer and to enter into its contractual relationship with Williams.
At 1180. We are unwilling so to speculate about the parties’ intent; we do not find this warranty in the executed contract.
This claimed warranty would operate as a prospective waiver of the limitation statute; the warranty is breached solely by defendant’s resort to that statute. If this be breach, it is of a peculiar sort. Williams is to be found in breach of contract because all seven factors listed below coincided in this case; its insurance coverage is rendered “ineffectual” because:
(1) the damages suffered by SLAM exceeded the value of the barge W-701;
*1172(2) SLAM’s damages likewise exceeded the amount of Louisiana-based insurance coverage;
(3) McDermott was found liable on its indemnity to Sun;
(4) Sun was held liable on its indemnity to SLAM;
(5) the umbrella insurance policy was not subject to the Louisiana direct action statute because it was delivered in Texas, not Louisiana;
(6) the accident occurred offshore, not in the inland waters of Louisiana; and
(7) McDermott secured no indemnity agreement from Williams.
It is indeed unfortunate that a nonnegligent party is made to assume a portion of this loss; but that does not result from any contractual breach. McDermott bears this liability because of its voluntary execution with Sun of an indemnity agreement. As acknowledged by the dissent (at 1179, 1180). Williams and McDermott could have agreed to a waiver of Williams’ rights under the limitation statute by executing an indemnity agreement. This they did not do. We are unwilling to reach by indirect means what the parties chose not to reach directly.
. Sun argues that, should it be held liable to indemnify SLAM’s attorneys’ fees, this liability should pass through to McDermott on their indemnity agreement. McDermott has not challenged this assertion but has limited its argument to the initial question of SLAM’s entitlement to fees on its indemnity with Sun. Sun’s position is correct but for a reason other than the one it advances. Sun argues that the specific inclusion of the words “attorneys’ fees” in its indemnity with McDermott requires McDermott’s payment of these fees. But the responsibility specifically assumed by McDermott for attorneys’ fees was for those incurred by Sun, not SLAM. McDermott’s liability for SLAM’s attorneys’ fees is not based on the specific “attorneys’ fees” language in its agreement with Sun but on its general obligation to make Sun whole for loss or expense arising out of the construction activity.