This case involves the valuation of a barge that was sunk in a towing accident and subsequently determined to be a total loss. Finding that the district court's determination of damages and prejudgment interest was not clearly erroneous, we affirm.

Plaintiff-appellant E.I. DuPont de Nemours & Co. (DuPont) was the owner of the barge EIDC-3, which was used to carry sulfuric acid to a refinery located on the Mississippi River. On April 20, 1984, the tugboat M/V RANDY JETT was towing the EIDC-3 and another barge from Burnside, Louisiana to Pascagoula, Mississippi. While on this trip, the RANDY JETT lost power in one of its engines, causing it to lose control of the barges, which in turn caused the EIDC-3 to hit the anchor chain of an anchored bulk carrier and sink the next day.

After several attempts to reach a settlement failed, DuPont filed suit in July 1986 against Robin Hood Shifting & Fleeting Service (Robin Hood), which was under a long-term towing contract with DuPont and charterer of the RANDY JETT, and Transload & Transfer Inc. (TTI), the owner and operator of the RANDY JETT.1 The RANDY JETT and TTI were insured by Employers Insurance of Wausau (Wausau) for up to $1 million under a hull and towers policy provision. Republic Insurance Co., Arkwright-Boston Manufacturers Mutual, and American Home Assurance Co. (collectively excess insurers) provided coverage for excess liability. DuPont sought damages of $1.4 million for the loss of the barge, and $120,000 for loss of the cargo of sulfuric acid.

Prior to the bench trial, all cargo and related claims were settled. The insurance companies also stipulated that they were liable for damage to the barge, leaving the amount of damage the only undetermined issue. Following testimony from expert witnesses, the district court awarded DuPont $250,000 for the value of the barge, but denied any increased award for uniqueness and special value of the barge to DuPont. The court also awarded DuPont $44,500 for loss of use and charter hire, but denied any award for the period of time DuPont knew and/or should have known that the EIDC-3 was a total loss. Finally, the court limited DuPont's award of prejudgment interest because it found that DuPont failed to conduct settlement negotiations in good faith. After the district court issued a series of amended judgments to correct minor errors, it awarded DuPont damages of $294,450.00 and prejudgment interest of $66,821.50. DuPont contends that the court made a number of errors in determining the value of the barge, the amount of damages for lost use, and the amount of prejudgment interest.

In reviewing a district court's valuation of a vessel in a bench trial, we must accept all factual findings unless clearly erroneous. Greer v. United States, 505 F.2d 90, 93 (5th Cir.1974); Fed.R.Civ.P. 52(a). In King Fisher Marine Serv., Inc. v. NP Sunbonnet, 724 F.2d 1181 (5th Cir.), reh'g denied with opinion, 727 F.2d 315 (1984), this court set out the law in determining damages for total loss of a vessel:

It is fundamental that when a vessel is lost or damaged, "the owner is entitled to its money equivalent, and thereby to be put in as good a position pecuniarily as if his property had not been destroyed." Standard Oil Company v. Southern Pacific Company, 268 U.S. 146, 155, 45 S. Ct. 465, 466, 69 L. Ed. 890 (1924). To further this policy courts have long held that where a vessel is a total loss the measure of damages is the market value at the time of loss. Id. at 155, 45 S.Ct. at 466. Where there are insufficient sales to establish a market other evidence such as replacement cost, depreciation, expert opinion and the amount of insurance can also be considered to determine the value of the vessel.

Id. at 1185. DuPont objects to the district court's valuation of the EIDC-3 on two grounds. The first ground is that the district court failed to take into account the special value of the EIDC-3 to DuPont. The second ground is that the district court improperly calculated the replacement cost of the EIDC-3.

DuPont first argues that the award of damages should be increased to compensate it for the special value that the EIDC-3 had to DuPont. Prior to the accident, DuPont used five barges, including the EIDC-3, to service the Chevron refinery in Pascagoula, Mississippi. These barges would deliver virgin acid to the refinery and pick up spent acid, which would be transported back to DuPont plants for reprocessing into virgin acid. DuPont points out that the EIDC-3 was uniquely suited for this task.

The EIDC-3 was a 2,000 ton dual-capacity barge. Its centerline tanks were designed to carry 2,000 tons of virgin sulfuric acid to the refinery, while its wing tanks carried 2,000 tons of spent sulfuric acid from the refinery. This dual capacity enabled the EIDC-3 to carry up to 2,000 tons of acid each way, thereby avoiding the costly and environmentally hazardous need to clean the tanks before loading them with virgin acid. The problem with valuing the EIDC-3 is that there are no barges comparable to it, and thus no market value for dual-capacity, 2,000 ton barges.

The closest equivalents are 1,400 ton and 2,000 ton single-capacity barges. But even if these barges were retrofitted with dual tanks, they still would only have a one-way capacity of 700 to 1,000 tons. At trial, DuPont introduced evidence that in order for DuPont to fulfill its contract with Chevron, it would take two 1,400 ton barges 86 round trips at a yearly towing cost of $946,000. Two barges like the EIDC-3 would require only 30 round trips at a yearly towing cost of $330,000. This is strong evidence that the EIDC-3 was much more economical than other barges, and that there is no comparable market for the EIDC-3.

But the district court was entirely correct in holding that DuPont was not entitled to any value greater than what it would cost to replace the EIDC-3. When no market value exists for a vessel, "other evidence such as replacement cost, depreciation, expert opinion and the amount of insurance can also be considered." King Fisher, 724 F.2d at 1185 (emphasis added). By using replacement cost, the district court took into account whatever special value the EIDC-3 had to DuPont.

The seminal case of The President Madison, 91 F.2d 835 (9th Cir.1937) provides an excellent example. In The President Madison, the court valued a boat specially designed for the local Skagit River by looking to the replacement cost of the boat. In so doing, the court refused to consider the market price at which the boat would sell on the Columbia River, because such buyers had different needs, and "could not give the owner, who needs these special features in his trade, the value of what he had lost." Id. at 844. Similarly, in King Fisher, this court found that because a barge lost at sea fulfilled the owner's unique and legitimate needs as a drydock, the district court was not clearly erroneous in awarding the owner the cost of replacement rather than the market price the owner paid for the barge only two days before the accident. King Fisher, 724 F.2d at 1182. The lesson of King Fisher and The President Madison is not that an owner gets more recovery than replacement cost if the vessel had special value. Rather, these cases merely demonstrate if there is no market for a comparable vessel, replacement cost is a reliable method of determining damages.

Here, the district court used two methods of valuing the EIDC-3, both of which estimated the actual replacement cost of the EIDC-3. By using replacement cost, the court properly accounted for any special value of the barge to DuPont, and thus there was no error.

Given the fact that replacement cost was the proper measure of damages, the next issue addresses DuPont's objection that the court failed to calculate replacement cost properly. The district court determined the value of the EIDC-3 by using two methods, the first of which figured the actual cost to replace the EIDC-3, depreciated to account for the barge's remaining useful life at the time of the sinking.

Before trial, the parties met in an attempt to determine the cost of building a new barge like the EIDC-3. Specifications were submitted to four different shipyards, and the lowest bid received was $904,250 from Nashville Bridge Company. Of course, there can be no argument that this figure does not take into account the special value of the EIDC-3, because the specifications submitted were for a 2,000 ton dual-capacity barge--a virtual carbon copy of the EIDC-3. The district court used the Nashville Bridge bid as the basis for calculation of replacement cost. Although DuPont objects that the court should have used the average of the four bids, absent any suggestion as to why Nashville Bridge would not be a proper contractor, DuPont's duty to mitigate damages suggests that the lowest bid for replacement of the EIDC-3 is the appropriate bid.

The district court then noted that different experts had different opinions as to the useful life, condition, and value of the EIDC-3. All agreed however, and the district court found, that DuPont maintained the EIDC-3 in better-than-average condition. The district court found that of the opinions of the experts testifying at trial, the most credible was that of Mr. Peter Merrill of Merrill Marine Services, Inc., an expert who testified on behalf of the excess insurers. Mr. Merrill testified that the average useful life of a barge on the Mississippi River was 20 years, but that the EIDC-3 had a useful life of 30 years due to DuPont's superior maintenance program. Having been built in 1960, the EIDC-3 would have been useful until 1990. Thus, at the time of sinking in 1984, the EIDC-3 had a remaining useful life of 6 years.

Based on these facts, the district court used the lowest bid for replacing the EIDC-3, rounded to $900,000 for ease of calculation, and multiplied it by the EIDC-3's remaining useful life of six out of thirty years, which equals $180,000. To that figure, the court added $100,000 to account for DuPont's superior maintenance program, for a total of $280,000.

DuPont levels two attacks at this method. The first is that because the EIDC-3 was in above-average condition, a useful life of 30 years was too low, and that a 40-year life was more appropriate. But the district court recognized the superior condition of the EIDC-3 in two ways. First, it adjusted its calculation up $100,000 in recognition of DuPont's superior maintenance program. Second, although the average useful life of a barge on the Mississippi River was 20 years, the court adjusted the EIDC-3's useful life to 30 years. Thus, the court did consider the EIDC-3's longer-than-average lifespan.

In addition, other evidence indicates that a 30-year life was not clearly erroneous. Portions of the barge, including the gunnel, deck, cargo tanks, bow rake, and stern transom were original equipment, and not renewed during the substantial repairs made in 1971 and 1981. The excess insurers also introduced evidence that the cargo tanks would have to be replaced, and that the EIDC-3 had an underwater crack in need of permanent repair. Considering all of the evidence and the adjustments made by the district court, a 30-year life was not clearly erroneous.

Second, DuPont contends that the court erred in using straight-line depreciation as opposed to "progressive depreciation." See Ozanic v. United States, 165 F.2d 738, 741 n. 2 (2d Cir.1948). First, DuPont waited until after trial to argue that progressive depreciation should be used. The district court rejected its use, stating that it "heard no testimony to support the use of progressive depreciation," and that "all the experts that testified about depreciation used straight line depreciation." This alone justifies the preclusion of progressive depreciation. Second, given its widespread and accepted use, only in rare cases will we find the use of straight line depreciation clearly erroneous. DuPont cites only one case where this circuit rejected the district court's use of straight-line depreciation. See Freeport Sulphur Co. v. S/S Hermosa, 526 F.2d 300 (5th Cir.1976). But Freeport Sulphur only held that straight-line depreciation was inappropriate where repairs to the damaged object extended its pre-accident useful life. In Freeport Sulphur, postcollision repairs to plaintiff's pier extended its precollision useful life by 10 years. This court held that a deduction should be made to the extent that the useful life of the pier was extended, otherwise the plaintiff would be put in a better position than before the accident. Id. at 305-06. Such a situation is not before us.

The problem is that the district court averaged the result of this method with the result of a second method. This second method of calculating replacement cost was based on the market value of 14 single-capacity barges, most of which had a total capacity of only 1,400 tons. Considering their remaining useful life, the average value of these barges was $120,000. To account for the dual-capacity of the EIDC-3, the court added $100,000 as the cost of retrofitting one of these barges with dual tanks. Alternatively, the court found that the extra costs associated with using single-capacity barges and cleaning the tanks to enable the shipment of virgin acid was also $100,000. Thus, total replacement cost of the EIDC-3 using this method was $220,000. When averaged with the result of the first method, the value of the EIDC-3 was $250,000.

The fault with this second method is that even if the single-capacity barges were retrofitted with dual tanks, they still could only have a one-way capacity of 700 tons as opposed to the EIDC-3's 2,000 tons. Similarly, even if one makes adjustments to account for extra tank cleaning costs, the one-way capacity is only 1,400 tons. As explained above, neither of these adjustments account for the increased towing costs that would result from the increased number of trips required to fulfill the Chevron contract. Thus, because this method relied on the value of barges that were not truly comparable to the EIDC-3, this method of calculating replacement cost was erroneous.

Nevertheless, because we can find no fault with the district court's first method of calculating replacement cost, the judgment is modified to recognize the result of that method, and thus we find that the value of the EIDC-3 at $280,000 was not clearly erroneous.

DuPont also objects to the district court's calculation of damages for loss of use and charter hire. The established rule is that in a case of a total loss, the owner is not compensated for the loss of use of the boat. King Fisher, 724 F.2d at 1187. The district court found that DuPont knew or should have known that the EIDC-3 was a total constructive loss by August 1984. Thus, it limited loss-of-use damages to the period from the time of sinking on April 21, 1984, to August, 1984. DuPont counters that it did not know that the EIDC-3 was a total loss until November 5, 1985, when the barge was actually raised.

The district court relied on evidence that in August 1984, marine surveyors hired by the insurers began a second attempt at a salvage operation, but learned that damage was more severe than earlier reported. These surveyors told a DuPont representative that the EIDC-3 was not salvageable and was a constructive total loss. Although DuPont argues that its representatives disagreed and thought the barge could be salvaged, the court found that DuPont should have realized from the August survey that the EIDC-3 was a total loss. The court also noted that DuPont itself never hired surveyors to determine if the EIDC-3 was a total loss. We find that this evidence provides ample support for the district court's conclusion, and thus there was no clear error.

Finally, DuPont challenges the district court's limit on the amount of prejudgment interest. The court limited the award of prejudgment interest to the period from June 22, 1986, forward. DuPont argues that prejudgment interest should have been awarded beginning on April 21, 1984, the date of the accident. In Ryan Walsh Stevedoring Co. v. James Marine Servs., Inc., 792 F.2d 489, 492 (5th Cir.1986), this court noted that an award for prejudgment interest under general maritime law is the rule rather than the exception; prejudgment interest must be awarded unless unusual circumstances make an award inequitable.

One of the grounds relied upon by the court for limiting prejudgment interest was that DuPont made an unreasonably high valuation of the EIDC-3 until June 22, 1986. In support of this finding, the court noted that DuPont's valuation of the EIDC-3 at $1 million was too high because it failed to make any deduction for depreciation. Although DuPont points out that the $1 million dollars was to settle the whole case, including loss of cargo, DuPont's valuation of the EIDC-3 at $659,000 was still unreasonably high. It not only fails to take into account depreciation, but it is over twice as high as the district court's valuation of $280,000. Because we have concluded that this valuation was not clearly erroneous, we cannot say that the district court abused its discretion in denying prejudgment interest on the grounds that DuPont placed a value on the EIDC-3 that was unreasonably high. Darling v. Scheimer, 444 F.2d 514, 515 (9th Cir.1971) (per curiam).

In summary, we affirm the district court's calculation of loss of use damages and prejudgment interest. Further, although the court's second method for calculating replacement cost failed to account for the reduced towing cost associated with a vessel like the EIDC-3, we find that the first method was not clearly erroneous, and thus we modify the judgment to reflect the value of the EIDC-3 as calculated under the first method. Therefore, the judgment of the district court is

AFFIRMED AS MODIFIED.