delivered the opinion of the court:
This is an appeal from the judgment of the circuit court of De Kalb County in which the court awarded to the plaintiff the sum of *9244.81 representing the balance of the purchase price of certain insulated pipe and awarded to the defendant (counterplaintiff) the sum of *61,628 as damages arising out of a construction contract.
*585In January, 1964, the Teachers College Board (College Board) let a contract for a certain steam heating system to be installed at Northern Illinois University. The contractor was L. J. Keefe, one of the defendants, and later counterplaintiff (Keefe). The original plaintiff (and later, counterdefendant) was Durant Insulated Pipe Company (formerly known as Portage Insulated Pipe Co.) (herein called Durant). Durant was the supplier of the pipe which Keefe ordered as contractor and which Keefe proceeded to install as per drawings prepared by Durant and approved by the College Board’s consulting engineer and architect. Some difficulty was encountered in laying the pipe due to moisture in the ground and in one place, quicksand, and the job was not completed until October, 1964. On October 23,1964, Durant acknowledged by letter that Keefe had installed the piping system “in accordance with our current installation instructions except that all portions of the system that became wet during installation must be dried out when steam is turned on.”
The system was not activated for another 6 months because of the delay in completing another and completely different heating system at Northern Illinois University, which was to be connected with the system being installed by Keefe. Steam was first introduced into the system in April, 1965, and very shortly problems developed — the system did not operate satisfactorily and the contractor was not able to dry it out so as to make it work efficiently. In June, 1965, it was decided by everyone involved that there were probably leaks in the system. Excavations were begun and a crack in the piping was discovered at a point known as Point 5. This crack was at a connection which had been welded at the Durant plant. Keefe repaired the crack by rewelding it in accordance with Durant’s specifications but the system still did not operate properly and after further excavations another cracked weld was discovered at Point 6 in the system. This also was repaired by Keefe according to Durant’s specifications.
The point which had cracked in both cases was a weld at a place where the pipe made a 90-degree turn. The system consisted of two pipes, a steam pipe about 10 inches in diameter and a return pipe about 5 inches in diameter (to draw off the water which had not turned to steam), both housed in a large 23-inch conduit made of corrugated sheet metal coated with asphalt and protected by asphalt felt. The steam pipe had a tendency to expand considerably and in order to give it plenty of room at the bends or turns in the system, the straight section and the right-angle section of the conduit where the turns were made were not designed to meet head-on or flush but were off-set 1% inches at such bends, giving the maximum room for expansion toward the direction the steam was flowing when it made its right-angle turn. This, of course, created a space between the nonaligned ends of the conduit, which were filled in by M-*586inch steel plates. These steel plates were welded to the ends of the conduit so as to close the gap. It was at the point of these welds where the cracks developed.
After the cracks were repaired the system apparently passed a satisfactory test. However, the system still could not be dried out properly and it became the opinion of all concerned that water was still getting into the system, although the source could not be discovered. Keefe hired the Heath Engineering firm to determine the cause of the problem but after some investigation they reported they could not determine the cause of the wet condition of the conduit. Thereafter, in August, 1966, Keefe employed another engineering firm, Adams Company, but they also were unable to come up with a satisfactory answer. Next, the firm of Consoer-Townsend was employed and they tried to find the source of the system’s problems but were unable to do so and they also gave up. The total fees of these three engineering firms was *11,856, but in spite of the time and effort spent in testing, analyzing and excavating, the source of the water in the conduit was not determined and water continued to get into the conduit. In addition to all the tests made by the three engineering firms, by Keefe himself and by university personnel, a test was made by a State chemist to determine whether the water in question was steam water or ground water. This, however, was also inconclusive. Finally, in 1968, having been informed by the engineers that the life of the system had been considerably shortened by water damage to the conduit, the College Board decided to scrap that part of the system which was thought to be the “wet” part. New piping was laid and a new manhole was constructed in the 515 feet of piping which was thought to be the defective part and after this was done the system operated satisfactorily.
The cost of installing the new piping was *48,723, the cost of the new manhole was *11,063 and Keefe spent *8,821 in repairing the cracked welds. By 1967, Keefe found himself financially unable to continue with the contract and his bonding company, Aetna Casualty and Surety Company, took over and was subrogated to any rights Keefe might have against anyone in connection with the contract. Aetna paid the engineering fees and the cost of laying the new pipe and for construction of the additional manhole.
Suit was originally filed by the College Board for the use and benefit of Durant against Keefe and Aetna for the remaining balance of *6,121 due for the piping material, Keefe having refused to pay this balance after the trouble developed. Keefe, by Aetna, then filed a counterclaim against Durant for all the sums expended by him and his insurer, Aetna, in investigating, repairing and rebuilding the system, amounting in total to some *108,000. Upon trial of the cause in November, 1974, before the *587court without a jury, the court, after hearing voluminous testimony and after considering the arguments and briefs of all the parties, being Durant, the College Board, Keefe and Aetna, issued its judgment order finding that:
(1) Durant was entitled to recover from Keefe (to be paid by Aetna) the sum of *6,121.28 plus some *3,000 interest for the remaining balance due on the bill for piping;
(2) Aetna, as subrogee of Keefe, was entitled to recover (including interest)
(a) *12,822.04 for welding repairs done by Keefe,
(b) *2,436.57 (including interest) for engineering fees of Heath;
(3) Durant and Keefe were jointly responsible (including interest) for:
(a) *1,164 for extra work done by Keefe;
(b) *12,699 for Adams Engineers;
(c) *2,458 for work of Consoer-Townsend Engineering;
(d) *14,381.90 for manhole construction, and
(e) *62,037.75 for replacement and installation of piping.
The net judgment in favor of Aetna and Keefe, therefore, was *52,384.66 of which part, as agreed, was awarded to the College Board.
In this appeal, Durant contends it was not proved responsible for the trouble in the system because it was not established by the evidence either that (1) the cracks were the cause of the water getting into the conduit or (2) that the cracks in the weld were the fault of Durant. They point out with regard to the cracks in the weld that (a) the ground in which the piping was laid was very wet to begin with and water could have gotten into the system from other causes since the system lay in the ground, inactive, from October, 1964, to April, 1965; (b) there was testimony that rough handling by Keefe had damaged the protective coating of the conduit pipes; (c) there was considerable testimony indicating that Keefe may have dug too deep a ditch in which to lay the piping, which would cause both extra water to gather around the piping and more weight from fill on top of the conduit, which extra weight could have caused the cracks at the welds or done other damage not related to the cracked welds; (d) Keefe used timbers to lay the conduit on in order to get the proper downward drainage — the theory of Durant being that this caused the conduit to sag in certain places when the timbers moved, thus causing undue stress on the conduit; (e) moreover, Durant pointed out that the system continued to be wet after the cracks at Points 5 and 6 were repaired, therefore the trouble could not be attributed to these cracks.
In response, Aetna argued-that Keefe took all possible precautions against water getting into the conduit when it was being laid but some water is bound to get in which is not unusual and would normally dry out if there were no leaks allowing water to come in from the outside. As to *588the question of rough handling damaging the protective cover, they maintain this was repaired or was minor and did not cause the leaks; as to digging too deep a trench and placing the conduit on timbers, they point to Durant’s letter of October 23, 1964, stating Keefe had installed the pipes in accordance with Durant’s specifications. As to the fact the system continued to give trouble after the weld and cracks were repaired, it was the contention of Keefe that this was due to further cracks or the opening of the same cracks due to a defective design which did not allow sufficient expansion for the steam pipes. As neither side saw fit to do any further excavating this point was never resolved.
Much of both briefs is taken up with a technical discussion as to whether or not the proper technique was used in the welding and whether the off-set design of the conduit at the turns was a correct design. It was contended by Keefe that the design was inherently susceptible to cracking at the welds where the conduit turned due to lack of expansion room as well as to the method of welding being wrong. The method of welding was criticized by Aetna’s expert as being a “wash” method whereas he thought the “fillet” method would have been stronger. There was some dispute, however, as to whether the wash method was used and this was not resolved by the expert testimony. The fillet method uses additional metal which is fused with the parts to be welded together whereas the wash method depends on heat, that is, the contact between the metals being heated to a point where the fusion is made. Also, Durant contended the model from which Aetna’s expert testified was not an accurate representation of the actual conduit, but Keefe’s testimony disputed this. The point was not resolved.
Since three engineering firms, a contractor and a consulting engineer could not pin down the source of the trouble, even though they were on the spot and used every test at their disposal, it is evident that there are a great many uncertainties in this case, but five facts are, we believe, established, and the trial court apparently bottomed its decision on them: (1) that Keefe accepted the piping material from Durant and installed it without making any objection; (2) that Durant gave its approval to Keefe (in the letter of October 23,1964) as to the manner of installing the pipe; (3) that cracks were discovered in the welds of the conduit at two places and these are logical places of water penetration; (4) that the life of the system was hopelessly impaired by the continued presence of water inside the conduit; (5) that the Teachers College Board suffered considerable damage and expense due to the failure of the heating system.
That Durant should recover from Keefe for the balance owed on the pipe it ordered and used is clear. (See Commercial Code, Ill. Rev. Stat. 1963, ch. 26, pars. 2 — 606 and 2 — 607.) Also, that the College Board *589should be made whole is not open to question. The difficulty here is in apportioning the responsibility for the cost of the failure of the system as between the contractor, Keefe, and the pipe supplier, Durant. The cases cited by the parties are not much aid in making this determination since the facts here, if not unique, are too dissimilar from the cited cases to be controlling. Perhaps the closest analogy to the case before us is Wilke Metal Products v. David Architectural Metals, 92 Ill. App. 2d 265. There the buyer ordered metal window casings which he accepted from the supplier and installed and then refused to pay for, claiming they were defective. In discussing and determining whether the cause of the glass breakage in these windows was a manufacturing defect or something to do with the method of installation or the weather, the court quoted at page 275 from Condon v. Schoenfeld, 214 Ill. 226:
‘It cannot be said that the existence of a certain fact may reasonably be inferred from the evidence when the existence of another fact inconsistent with the first can be, from the same evidence, inferred with equal certainty.’ ”
Durant cites this language, contending that Aetna only showed a possible explanation of the leakage of the conduit and then had inferred a manufacturing defect from that, but that other facts cited by Durant, such as improper installation, too great a depth for the ditch, lack of shoring, and rough handling, were equally open to the inference that Keefe’s negligence caused the cracks in the welds. Therefore, they argue that these cracks should not have been taken as an inference against Durant as the other facts were open to the same inference against Keefe. In regard to the inferences against Keefe, however, we must observe that as to the method of installation, involving the question of the depth of the ditch, the laying of the pipe on proper supports, the shoring of the ditch and the handling of the conduit, such inferences are inconsistent with Durant’s letter of October 23, 1964, which, in effect, accepted the installation as being in accordance with their specifications. These issues were raised after Keefe’s counterclaim was filed and the letter of October 23 suggests that either these conditions were not considered as of critical importance at the time or were remedied before the work was completed and the letter issued. This is at least a logical inference from Durant’s letter approving the installation. The quotation from the Wilke Metal Products case is, therefore, inappropriate, as the existence of the break in the welding was uncontroverted whereas the fact of Keefe’s negligent or unsatisfactory installation was controverted by Durant’s own letter. The inferences in this case were not of equal strength and the observation in Wilke Metal Products is not apropos to the case before us.
It was alleged also by Durant that the soil was too wet when the pipe was laid. However, it was acknowledged by Durant that this was not an *590unusual condition in the construction of a large, underground steam pipe system. Since Durant’s superintendent was present during the installation and did not try to stop the operation we must assume that Durant accepted the wetness as a condition inevitable to the job. Their letter accepting the installation in noting the necessity for drying out the system acknowledges the effect of moisture but only requires that the system be dried out when the steam is turned on. It did not indicate that the system was thereby ruined.
Durant was thoroughly experienced in constructing systems for all kinds of soil conditions and (as was brought out in the testimony questioning his experience) Keefe was less so, therefore the inference that Keefe was entirely responsible for the affect of wet soil — even conceding that it was a factor — seems unwarranted. The only definite source of leakage which was established by the testimony was the cracked welds. Since Durant manufactured the pipe and did the welding at their own plant the trial judge held them responsible for the cost to repair these welds, but as to the cost of the labor and materials for the new pipe and the manhole necessary to lay it, Durant was only held jointly liable with Keefe. They were likewise held jointly liable for thé engineering work which proved so fruitless
While the evidence, therefore, was conflicting, the more definite evidence favored the resolution determined by the trial court. In view of Durant’s involvement with Keefe in the installation of the pipe, through the presence of their superintendent on the job and their letter of October 23, and the definite finding of cracked welds — whether this was due to improper design or inadequate welding — there is adequate evidence to sustain the trial court’s findings as to joint liability. Teachers College Board was an innocent party which suffered damages and must be reimbursed. The case, therefore, comes down to a question of the proper assessment of damages for failure to perform a contract as agreed. There was room for blame on both sides and to a large degree this was recognized in the trial court’s judgment assessing joint liability for the cost of the work which had to be done over. The trial court’s assessment of the evidence led him to believe that Durant was at least partly responsible for the lamentable failure of the system. In our view the evidence amply sustains a finding of at least partial responsibility on their part. The judgment, therefore, is not contrary to the weight of the evidence and should be affirmed.
The judgment of the circuit court of De Kalb County is affirmed.
Judgment affirmed.
T. J. MORAN, P. J., and DIXON, J., concur.