Georgia Kaolin International appeals from a judgment awarding damages to a vessel and her owners on their counterclaim based upon the damages suffered by the vessel when she broke away from GKI’s dock in thé Savannah River. The trial court’s judgment was based upon its determination that the dolphin1 to which the vessel was made fast was inadequate to withstand the pull on the hull of the vessel caused by the incoming tide on July 4,1978.
*413The appellant, to the contrary, contends that the failure of the dolphin was caused by the fact that the four stern lines making the ship fast to the dock were not tended as the tide rose approximately nine feet and that this failure necessarily caused the lines to tighten as the vessel rose, thus creating the pull testified to the appellees’ witnesses. If we are to take the testimony of the expert witnesses for its full value, there is no doubt but that the trial court could have accepted the vessel’s explanation of the cause of the incident. However, it is the appellant’s position that the expert evidence upon which the trial court relied was based on several assumptions which were either unproved at the trial or in conflict with the proven facts. There is no dispute between the parties but that the expert opinions given must be based on record facts in order to carry probative weight with the trial court.
For convenience in later reference to the fact findings, some of them are reproduced here:
17.... The four stem lines extending to the nine pile mooring dolphin were approximately 100 feet in length .... ******
21. Mr. Alonza deF. Quinn, a consulting marine engineer, reviewed the drawings and specifications prepared by Mr. McArthur as modified by Mr. Wilkinson, together with a report on boring samples of the subsurface soil at the site prepared by Whitaker Laboratory, Inc. on August 8, 1978, after the incident in question. Mr. Quinn, author of an internationally recognized textbook on marine structure, testified that the capacity of the dolphin at failure was twelve tons. Since the dolphin should have a safety factor of two, the design was adequate for an anticipated pull of six tons. In order to safely serve as mooring for a vessel as large as the GRAND JUSTICE, the dolphin should have been designed for a lateral pull of 35-50 tons. The dolphin did not fail in tension. Instead, the force on the dolphin was sufficient to break the skin friction of the outer piling on the dolphin and cause the far piling to pull out. According to Mr. Quinn, whom the Court credits, 15,000 lbs. of force was being exerted on the stern by the tide. Mr. Quinn took into account the proximity of the vessel to the river bottom and the slant of the bank. The horizontal force was 22,000 pounds. The pull on the rear piling was 68,000 pounds which was equal to the force necessary to break the skin tension and cause the piling to pull out. Once the skin tension was broken, the dolphin lost rigidity and the pilings were broken off.
23. Nor is the fact that the M/V GRAND JUSTICE had been moored at Berth # 32-33 for three days before the breakaway without incident sufficient to draw the inference that the breakaway was caused by poor adjustment of the lines. The tides were approaching spring tide conditions. Thus, the tide was becoming progressively stronger day by day. The force on the stern exerted by the tide was stronger at the time of the breakaway than it had been during the previous few days.
The facts which the appellant claims were either not supported by anything in the record or were inconsistent with the court’s final conclusion are all contained in the quoted portion of the court’s findings of fact:
1. The court stated: “The four stern lines extending to the nine pile mooring dolphin were approximately 100 feet in length.”
Quinn’s computation by which he arrived at the “horizontal force” of 22,000 pounds was based upon his inclusion in the figure of 7,000 pounds which he found to exist by reason of the proximity of the vessel to the river bottom and the slant of the bank. However, he assumed that the angle between the stem line and the face of the dock was 45. degrees, which was based upon the assumption that the stern lines were 70 feet long rather than 100 feet long. Moreover, Quinn testified that if the angle was less than 45 degrees, the force would be reduced somewhat.
*4142. The trial court found that at 6:15 p. m. on July 4 the tide was “flooding at slightly less than two knots.”
Quinn testified that he assumed the tide was flowing at 1.7 or 1.8 knots, and the computations put in evidence by witness Stuber on behalf of the appellee showed the current at 3 feet per second, which may be translated into 1.78 knots in one of his equations and 1.8 in another, whereas the only proof in the record was from a Coast Guard employee to the effect that the current at the time of the accident was 1.5 knots.
3. The court stated “The tides were approaching spring conditions. Thus, the tide was becoming progressively stronger day by day. The force on the stern exerted by the tide was stronger at the time of the breakaway than it had been during the previous few days.”
The only evidence, other than testimony of the captain and crew members of the vessel as to how the current appeared to be flowing was the testimony of Coast Guard investigating officer Yelton, who testified that the tide was flowing 1.5 knots at 1838 hours. (The breakaway occurred at 1815 hours.) With respect to the statement in this finding that “the tide was becoming progressively stronger day by day and that the force on the stern exerted by the tide was stronger at the time of the breakaway than it had been during the previous four days,” the graph of the tides during the previous four days shows that the tide had been higher on one of the previous days and had been as high on another one, during which the vessel was moored to this dock.
Finally, although it is not mentioned in the findings of fact, appellants contend that the record discloses that witness Quinn, whom the trial court credited, assumed the design of the vessel to have been one with a box-like or flat stern with a beam of some 67 feet and a depth below water of 30 feet, whereas the vessel here involved was designed with a narrowed stern or as described by one witness as having been built very “fine,” that the pressure on the stern of the vessel of a box-like shape would necessarily be much in excess of that on a vessel having a narrowed hydrodynamically designed stern.
We consider each of these points of dispute, because each of them deals with an element of the formulation used by Mr. Quinn in his computations which were accepted by the trial court.
To begin with, we should note that by Mr. Quinn’s testimony, there would not be expected to be a failure of this dolphin unless the pressure had been in excess of 12 short tons, 24,000 pounds per square foot. He testified, of course, that in designing the dolphin it would have been proper to have provided a 100 percent safety factor or more. That, of course, has nothing to do with the question of causation of this accident, which depends upon the determination whether the dolphin collapsed under pressure which it should have withstood under the existing circumstances.
1. Length of Lines
As already indicated, the trial court found that the lines to the dolphin were approximately 100 feet. There is further persuasive evidence that the lines were actually longer than this, because a witness testified that the drawing to scale made it impossible for the lines to be any less than 130 feet. However, we will assume the correctness of the trial court’s determination of 100 feet. This, in itself, is inconsistent with the acceptance by the trial court of the precise figures testified to by Mr. Quinn as to the additional 7,000 pounds of pressure caused by the slope of the bank, etc. Mr. Quinn assumed the angle of the line to the dock to be 45 degrees, based upon the schematic drawing by witness Stuber. That drawing used a line 70 feet long. Substituting a line 100 feet long would have reduced the angle to something approximately 31 degrees, by rough calculation. We do not know the formula by which Mr. Quinn computed the 7,000 pounds of pressure assuming a 45 degree angle, so we are unable to compute the reduction in this figure that would result from the reduction of the angle from 45 to 31 degrees. However, Mr. *415Quinn conceded in his testimony that a reduction in the size of the angle would reduce the amount of the pressure. Quinn stated: “In other words, the flatter the angle, the less would be the increased force on the mooring lines.”
2. Speed of the Current
The speed of the current was obviously the most important factor in determining the force against the vessel under the theory advanced by the defendant and accepted by the court. The exact effect of a slight change in the rate of the current will be demonstrated by the equations illustrated below. The trial court casually stated that the current was “slightly less than two knots.” Quinn testified that he assumed a current of 1.7 or 1.8 knots. He didn’t seem to know which. He stated: “We have assumed that the flood current was somewhere in the order of 1.7 or 1.8 knots, I believe.” Since in his testimony, his attention was called to computations made by another defense witness, Elmer R. Stuber, and he indicated that he was not in disagreement with any of the formulas used by Stuber, we consider it appropriate to take Quinn’s figure of 1.8 for the current which he used in his computations, although Stu-ber used alternatively a figure of 3 feet per second which amounts to 1.777 knots. Because of the substantial effect of using the correct figure of 1.5 knots as against 1.8, as demonstrated below, we agree with the appellant that the trial court also erred in accepting Quinn’s estimate of 22,000 pounds per square inch without modification based on this incorrect factor.
3. The Height of the Tide
Quinn testified that the deck of the vessel was above the level of the dock both at low water and at high water, conceding that as the tide flowed it raised the deck of the ship accordingly, so that at high tide it would be eight or nine feet more above the level of the dock than at low tide. Based on this factor, the appellant contended that the fact that the ship had been moored alongside the dock for three days previously during which time it was testified that the crew lengthened and shortened the lines as the tide rose and fell was strong evidence that the error here resulted from the failure of the ship’s crew to adjust the lines from the low water mark. The failure was admitted by the ship’s master. The trial court discounted this evidence, because it stated: “The tides are approaching spring tide conditions. Thus, the tide was becoming progressively stronger day by day. The force on the stern exerted by the tide was stronger at the time of the breakaway than it had been during the previous few days.” As already indicated above in discussing the current, these statements of fact were incorrect. GKI introduced at trial a Corps of Engineers’ tidal graph showing the level of the tides in the Savannah River from July 1 through July 5. It shows that on July 4, the day of the breakaway the afternoon high tide, which occurred two hours after the breakaway, was 9.1 feet. At the time of the breakaway, the tide was only 8.1 feet. On July 3, the high tide was also 9.1. On July 2, the afternoon high tide was 9.2 feet and on July 1, the day the Grand Justice was tied up originally, the afternoon high tide was 9.3 feet. Thus, there were several tides higher than that at the time of the incident.
4. The Shape of the Huil
Finally, we come to the contention of the appellant that Quinn’s estimate of the force exerted against the hull of the vessel was improperly based upon a “box-like” stern, rather than a vessel with a fine design. We are unable to agree with the appellant as to this argument, because Quinn stated that his computations were based upon an equation that included what he called a “hull factor.”2 For want of further development *416of what this hull factor comprises it is impossible for us to say that this variance in the equation would not adequately solve the problem as to the shape of the vessel.
We now come to a consideration of the effect on Quinn’s computations of the incorrect fact findings by the trial court. We deal with two of them, taking them in inverse order to their discussion above.
Witness Stuber testified to the formula which he obtained from a book written by Quinn and which, so far as apparent here, was the basis of Quinn’s own estimates. For some reason, the parties did not include Quinn’s own computations in the evidence in this case, although they were present in court at the time of Quinn’s testimony. This formula is comprised of three parts. The first was “force current against stern end of ship parallel to current.” It was expressed ’
W FC = 2G V2.
W represents the weight of salt water, G represents the force of gravity and V is the current stated in feet per second. As Stu-ber resolved this equation, he determined that FC = 9 pounds per square foot. He then applied this figure to the ship’s beam, 67.35, times the ship’s draft of 30 and arrived at a figure of 18,185 pounds.
If, instead of using a current of three feet per second (1.777 knots) we substitute 1.5 knots, the result is 6.2 pounds per square foot. This figure multiplied by the beam and draft give us a final figure of 12,527 pounds of pressure instead of 18,185.
In addition to the item of force current against stern, Stuber added two small forces, one against what he called the “hull wetted surface” and one against “the propeller”. By the use of a current of 1.8 knots in both instances he added 656 pounds and 821 pounds respectively for these two items so that his grand total was 19,662 pounds. Using the current of 1.5 knots, these two figures would be 455 and 570 respectively for a total of 13,552 pounds of pressure.
Mr. Stuber also made a computation of the final force when considering the slope of the river bank. He expressed this in the equation F =-X 19662. If we substitute the figure 13552 for the latter figure and multiply it by the square root of two, we arrive at a final computation of 19,121 feet of pressure. This would be the total force according to Stuber’s computation, with corrections based on the speed of the current. It is to be compared with the strength of the dolphin estimated by Mr. Quinn to be 24,000 pounds. This, of course, is on the assumption that there is to be no reduction based on the demonstrated fact that the final figure would be somewhat less because the angle is less than 45 degrees.
We now turn to Quinn’s computation of 15,000 pounds of pressure. This must be converted by using 1.5 knots instead of the 1.8 which he used. In arriving at his figure Quinn used only the first of the equations outlined above from Stuber’s report. He ignored the latter two. We have already determined that the substitution of 1.5 for 3 f. p. s. gives us 6.2 pounds per square foot or 12,527 pounds of pressure against the stern.
To this figure, we then add the figure arrived at by using the last equation in Stu-ber’s computation to obtain the total force of the current: F =-X 12,527. This produces the figure 17,675 pounds of pressure instead of the 22,000 pounds arrived at by Quinn. It is to be noted also that in using the same formula used by Stuber to calculate the current caused by the slope of the river bank, etc., we have assumed that the angle was correctly figured at 45 degrees which, as we have indicated above, produces a larger figure than would be correct.
*417 What is the effect of all this?
In his testimony, Quinn made a computation that converted the 22,000 pounds of pressure against the lines of the vessel into pressure against the rear piling in the dolphin of 68,000 pounds. Quinn did not explain the formula by which he arrived at this figure, but it is not challenged by appellants, because Quinn stated in his testimony: “The force that I computed on the stern line — this pile in tension — in tension from that force — put a tension in that pile of about 67 'to 68 kips [a kip is 1,000 pounds] practically the same value as the ultimate friction value holding the pile from being pulled out. So, it was right on the verge, or almost identical, to pull out.” Since the “67 to 68 kips” figure was based on the multiplication of his 22,000 pound figure, it is obvious that a reduction of the 22,000 to 17,675 pounds would reduce the pressure on the “pile in tension” substantially below the 67,000 or 68,000 pounds as to which he testified. This figure would, of course, be further reduced when the effect of the cross current against the vessel resulting from the slope of the bank is recomputed because the angle of the line to the dock was 31 degrees instead of 45 degrees as assumed by Quinn.
There is no principle of law in dispute between the parties here. It is clear that if the opinion evidence could be accepted at its face value, the trial court was within its power to decide that the dolphin was pulled out because it had insufficient strength to withstand the pull of a current that should normally have been expected under the circumstances. If, on the other hand, the ' opinion evidence was based on assumptions as to the rate of the current and as to the angle, which played a part in computing the pressure on the vessel, which were incorrect as not being supported in the record or as contrary to the proof in the record and which incorrect assumptions would substantially affect the validity of the opinion evidence, then such evidence should not have been accepted by the trial court as disposi-tive of the case. Here, the trial court expressly credited Quinn’s testimony. But a careful analysis of the facts disclosed in the record indicate that the assumptions he made both with respect to the speed of the current and the angle between the line and the dock were contrary to the proven facts. It is further evident that under the proven facts, the final opinion as to the pressure against the vessel by reason of the current was excessive by at least five thousand pounds. To the extent that this figure should be deducted from the 22,000 pound figure given by Quinn, the total pressure against the vessel fell below what Quinn himself testified would have been within the capacity of the dolphin to withstand.
We conclude, therefore, that the judgment based, as it was on this testimony, cannot stand.
Appellant claims that without this evidence, the only cause for the breakaway of the vessel must have been the failure of the crew to adjust the lines as the ship rose approximately 8.1 feet above low tide on the afternoon of July 4 without any adjustment of its line. The record contains a letter from appellee’s counsel to the trial court in support of its proposed findings of fact, in which it concedes that the deck of the vessel to which the line was attached was 20 feet above the level of the dock. This, of course, would mean that as the vessel had risen from low tide the lines would necessarily become tighter. Nevertheless, we do not think the record would justify our determining the cause of the breakaway. This must be for the district court in the first instance.
The judgment is REVERSED and the case is REMANDED for further proceedings not inconsistent with this opinion.
. The dolphin here consisted of nine treated timber pilings, 65 feet in length, set at a one to twelve batter, the center piling of the cluster being driven vertically. On top of these pilings was a concrete cap in which there was a single steel bollard. It was set in the Savannah River adjacent to the GKI pier and was reached by a catwalk from the pier.
. This testimony was developed as follows:
Q: Well, let me ask you this. Doesn’t it make a tremendous difference as to how fine the vessel is?
A: How what?
Q: How fine the shape — the hull shape of the vessel is?
*416A: It makes some difference, but let me say that authority that I quoted, and that is Abbott, takes the same factors and that is the hull factors and taking the whole dynamic pressure against the hull. That velocity squared over 2 G times a hull factor, and the variation in the hull factor is a low .7 or .75 — I don’t recall — .7 or .75. I have used the lowest hull factor that would give the least resistance.