LittletoN, Judge,
delivered the opinion of the court:
The plaintiff sues for payment for electricity furnished to the Government which was not recorded by the meter because of a “phase unbalance” caused by a blown fuse. The questions presented are whether the contract prohibits recovery, and if not, the time that the fuse blew and the amount which plaintiff is entitled to recover.
[277]*277On April 1, 1943, plaintiff entered into a contract with the Army to supply electricity to Fort Buchanan, an Army post located at Guaynabo, Puerto Rico, and to various other points within a 25-mile radius. The plaintiff supplied electricity to defendant under the terms of the contract at all times material to this action. The plaintiff submitted and defendant paid monthly bills for the electricity supplied as shown by the meter readings. A clause in the contract provided:
* * * The contractor shall furnish, install, maintain, calibrate and read meters to measure the electrical energy to be furnished hereunder. Meters will be read monthly in the presence of the Contracting Officer or his authorized representative. * * *
In case of error in meter readings, it is mutually agreed that the percentage of error shall apply only to the electricity used since the date of the meter reading applying to the last bill.
The contract required plaintiff to install metering devices to measure the demand in kilovolt-amperes (KYA) at specified intervals and the consumption in kilowatt-hours (KWH) continuously. The plaintiff was to be paid for the current consumed on the basis of a formula using both demand and consumption, for each of which rates were specified. The defendant agreed to furnish and maintain the transformer substations and local distribution system and to connect its load so as to give a balanced load on all phases. If a phase unbalance exceeding ten percent should be found to exist, the plaintiff could require the defendant to rearrange the connection of its load to correct the unbalanced conditions.
The plaintiff’s transmission lines carried 38,000 volts to two substations, one known as the Army Terminal substation, at which connections were made for emergency use only, without metering devices, and another substation at the South Gate of Fort Buchanan, where the meter was installed. Thereafter, when current was supplied through the Army Terminal substation, the amount used was estimated and paid for on the basis of prior use for an agreed period as recorded at the South Gate substation.
[278]*278During March, April, and May 1946, while the transmission lines within the Fort Buchanan area were being relocated, the current was supplied through the Army Terminal substation and defendant paid for the electricity based on an average of the preceding six months.
The transformers at the South Gate substation received current of the strength of 88,000 volts, carried by three transmission lines. The voltage was reduced by the transformers to the 4,000 class. On the primary, or high voltage side, each transmission line was attached to a fuse, through which the current passed into the transformer. Each fuse was between the meter and the transformer. The fuses were in the portion of the installation maintained by the Army and the plaintiff had no responsibility for their inspection, maintenance, or replacement. The meter corresponded to the requirements in the contract and was in proper working order at all times.
During the 20-month period beginning with July 1944 and ending with February 1946, the average KVA demand was 1808.73 while the average KWH consumption was 592,911. The average of the monthly bills was $6,848.91. The military strength at the fort declined between February and August 1946, but remained fairly constant thereafter. During this period of decline there were contingents of military personnel at the post from time to time using it as a way station in the shifting of manpower.
On July 6, 1948, while Government employees were trimming trees adjacent to the power lines, a branch of a tree fell across the lines and caused interruption of the electric service. Investigation disclosed that two of the three fuses were burned out. When the fuses were replaced, full service was restored and the meters accurately recorded the current furnished. The parties agree that one fuse blew at this time and that the other was blown at some earlier date. When one fuse was blown service was not disrupted but a phase unbalance exceeding ten percent was caused, and the meter was caused to record an inaccurately low consumption. When two fuses were blown service was disrupted.
An explanation was demanded by defendant when the July bill, which was $9,704.57 (after the two fuses were replaced), [279]*279was compared with the June bill of $5,840.38. It was noticed that both EWH consumption and amounts paid by defendant had been substantially reduced from earlier patterns since August 1946, and that there had been a rather sharp reduction of KVA demand in July 1946. Neither demand nor consumption had been as high, since midsummer 1946, as they were before that time or as they were shown to be in July 1948. When the figures were plotted on a graph, valleys in both demand and consumption were reflected, extending over a 2-year period, approximately, in what otherwise appeared as a slowly ascending line for KWH and a slowly descending line for KVA. The bakery was closed in March 1948, for conversion to oil, which reduced the KVA load. It was reopened the following June and the KVA load was slightly increased. From July 1948, to the early part of 1949, some electrical equipment was added and the KVA was increased by about 800.
The plaintiff contended that the first, fuse was blown in July 1946; that the failure of this fuse affected the meter in such a way that it failed to record all of either consumption or demand; that defendant had been underbilled for current during the 24-month period from July 1946 to July 1948; that the application of an experience pattern drawn from five months immediately before and after the period in controversy affords a reasonable basis of estimating such current; and that on such basis, defendant had and used power for which it has not paid of the valu* of $80,449.78,
The defendant contends that the above-quoted clause in the contract relative to meter readings precludes recovery and in the alternative, that plaintiff’s claim is speculative and, in any event, it is just as reasonable to say that the first fuse was blown in December 1947 as it is to say that it was blown at an earlier date. The Armed Services Board of Contract Appeals found that plaintiff had established a claim for electric energy furnished the defendant, but was of the opinion that the meter reading clause prohibited recovery, except for the period following the last bill before the restoration of service on July 6,1948.
We think defendant’s contention that this was a “meter reading” error is without merit. The evidence in the record [280]*280fails to establish a trade usage meaning for this term. There was no error on the part of personnel in reading or recording figures from the meter. The blown fuse, for which the Government was responsible, caused the phase unbalance which caused the meter to record inaccurately. The Government drew the contract. If defendant intended to limit or define its liability resulting from mechanical failures caused by itself or for which it was responsible, less ambiguous language should have been employed. The plaintiff agreed to furnish and the Government agreed to pay for the electricity demanded and consumed, at specified rates. The plaintiff has fulfilled its part of the contract and now seeks to be paid for it. To give the meter reading clause the construction contended for by defendant would be unjust and contrary to the obvious intent of the parties. United States v. Southern Gulf Lumber Co., 106 F. Supp. 815 (S. D. Ala.). In that case, where the clause was identical to the one in the instant case, the Government employee had actually misread the meter by adding a zero to the correct figures. The court allowed the Government to recover the overpayment.
The two remaining questions are when did the fuse blow, and how much electricity did defendant use? The commissioner has found from the evidence of record that the fuse blew about September 1,1946. Although plaintiff contends it blew in July of that year, it acquiesces, however, in the commissioner’s finding. The defendant contends that a determination of when the fuse blew would be speculative, and at any rate the fuse did not blow earlier than December 1947. The commissioner has found that there is nothing in the evidence other than the blown fuse to explain the precipitate drop in energy consumption in September 1946, as compared with consumption of previous months. We agree with this finding. There was also a large unexplained drop in the recorded (KVA) demand at about that time. The actual amount of electricity used by defendant at Fort Buchanan was fairly constant from August 1946 to July 1948. We agree with and adopt the commissioner’s finding that the evidence reasonably indicates that the fuse blew in September 1946.
[281]*281The plaintiff contends that it is entitled to recover $72,456.50 for the amount of current furnished defendant for the period from September 1,1946 to July 6,1948. The plaintiff bases this amount on an estimate predicated on a 5-month average before and a 5-month average after the period in controversy, computed at the rates then in effect. The commissioner of the court has found that the amount of recovery should be $66,500. After considering the slight variations which tended to increase the use in the 5-month period after June 1948, and the 2’0-month period preceding the period in controversy, we concur in the commissioner’s finding that $66,500 should be the amount of recovery.
The plaintiff is entitled to recover and judgment will be entered in the sum of $66,500.
It is so ordered.
Laramoke, Judge/ MaddeN, Judge; Whitaker, Judge; and JoNes, Chief Judge, concur.
FINDINGS OF FACT
The court, having considered the evidence, the briefs and argument of counsel, and the report of Commissioner W. Ney Evans, makes the following findings of fact:
1. On April 1, 1943, plaintiff1 entered into a contract2 with the Army to supply electricity to Fort Buchanan, an Army post located at Guaynabo, Puerto Rico, and to various other points within a 25-mile radius.
The contract continued in effect and plaintiff supplied electricity to defendant under its terms at all times material to this action. During all such times plaintiff submitted and defendant paid monthly bills for the electricity supplied as shown by the meter readings. Each of such bills was certified by plaintiff as being true and correct.
2. The contract contained the following provision:3
[282]*282* * * The contractor shall furnish, install, maintain, calibrate and read meters to measure the electrical energy to be furnished hereunder. Meters will be read monthly in the presence of the Contracting Officer or his authorized representative. * * *
In case of error in meter readings, it is mutually agreed that the percentage of error shall apply only to the electricity used since the date of the meter reading applying to the last MU. [Italics supplied.] 4
This case involves no error in meter readings in the sense of failure on the part of personnel to read or record accurately.
3. The contract required plaintiff to install metering devices (some of the features of which were specified) to measure (a) the demand in kilovolt-amperes (KVA) at specified intervals and (b) the consumption in kilowatt-hours (KWH) continuously. Plaintiff was to be paid for the current consumed on the basis of a formula using both demand and consumption, for each of which rates were specified. Defendant agreed to furnish and maintain the transformer substations and local distribution system and “to connect all his load so as to give a balanced load on all phases.” If a phase unbalance exceeding ten percent should exist, the company could require the consumer to rearrange the connection of his load to correct the unbalanced conditions.
4. Plaintiff’s transmission lines carried 38,000 volts to two substations, one known as the Army Terminal substation, at which connections were made for emergency use only, without metering devices, and another substation at the South Gate of Port Buchanan, where the meter was installed. Thereafter, when current was supplied through the Army Terminal substation, the amount used was estimated (and paid for) on the basis of prior use for an agreed period as recorded at the South Gate substation.
5. (a) During March, April, and May 1946, transmission lines within the Fort Buchanan area were being relocated at defendant’s request. While this work was being done, current was supplied through the Army Terminal station. The meter at the South Gate substation was not recording [283]*283at this time. Defendant was billed and paid for current based on an average of the preceding six months, computed as 1867.8 KVA and 615,780 KWH for each of the three months. The relocation work extended into part of June, but the meter was recording during at least ten days of that month. It reflected a maximum demand of 1920.0 KVA. Consumption for the month (estimated and recorded) was billed as for 551,343 KWH.5
(b) During the 20-month period beginning with July 1944 and ending with February 1946, the average KVA demand was 1808.73,6 while the average KWH consumption was 592,911.7 The average of the monthly bills was $6,848.91.8
6. The military strength at Fort Buchanan declined between February and August 1946, so that during the latter month it was only 23 percent of what it had been in the former. After August 1946, the military strength at the post remained fairly constant.
At the hearing before the Armed Sendees Board of Contract Appeals, defendant declined on grounds of security to reveal figures or to show in detail the relation between the decrease of military strength and the consumption of electrical energy. At the trial of this case, the figures and data were no longer available.
The term “military strength” is not defined in the evidence. It was used by witnesses in the sense of military personnel stationed at Fort Buchanan, as distinguished from military personnel sent into or through the post for short periods. During the period of decline in military strength at Fort Buchanan there were contingents of military personnel at the post from time to time, using it as a way station in the shifting of manpower.
[284]*2847. Beginning in March 1948 and extending through March 1949, several changes were made in the installations at Fort Buchanan which tended to change the KVA load.
The bakery was closed in March 1948, for conversion from electric to oil burning ovens, thereby reducing the KVA load. When the bakery was reopened the following June, there was a slight increase in the KVA load.
In July 1948, the installation of electric water heaters increased the load. Further increases occurred in August and September, upon the opening of new barracks and officers’ quarters. The load was further increased in November and December, by the opening of an induction center and the electrification of a mess hall. Other increases occurred during the early part of 1949. Defendant’s post engineer estimated the total increases in KVA load as a result of these installations to exceed 800, nearly all of which occurred after June 1948.
8. In connection with the work of relocating lines in the spring of 1946, the transformers for the meter installation were changed. The replacement transformers were placed in service on June 7,1946.
On December 16,1947, plaintiff replaced the meter at the South Gate substation.
The transformers at the South Gate substation received current of the strength of 38,000 volts, carried by three transmission lines. The voltage was reduced by the transformers to the 4,000 class. On the primary, or high voltage side, each transmission line was attached to a fuse, through which the current passed into the transformer. Each fuse was between the meter and the transformer.
9. On July 6,1948, Government employees were trimming trees adjacent to the power lines at Fort Buchanan. A branch of a tree fell across the lines, and a sharp interruption of electrical service in the Army post ensued. A single-phase service replaced the normal 3-phase service. Investigation disclosed that two of the three fuses described in the preceding finding were burned out. When the fuses were replaced, full service was restored.
10. The next bill submitted to defendant, based on the meter reading made at the end of July, was in the amount of [285]*285$9,704.57.9 Defendant compared it with the bill for the previous month, in the amount of $5,340.38,10 and demanded of plaintiff an explanation of the increase.
11. Plaintiff investigated and found that both KWH consumption and amounts paid by defendant had been substantially reduced from earlier patterns since August 1946,11 and that there had been a rather sharp reduction of KVA demand in July 1946.12 Neither demand nor consumption had been as high, since midsummer 1946, as they were before that time or as they were shown to be in July 1948. When the figures were plotted on a graph, valleys in both demand and consumption were revealed, extending over a two-year period, approximately, in what otherwise appeared as a slowly ascending line for KWH and a slowly descending line for KVA.
12. Following are the pertinent portions of paragraph 2 of the General Conditions of the contract, relating to disputes:
Except as otherwise specifically provided in this contract, all disputes^ concerning questions of fact which may arise under this contract, and which are not disposed of by mutual agreement, shall be decided by the Contracting Officer, who shall reduce his decision to writing and mail a copy thereof to the contractor at his address shown herein. Within 30 days from said mailing the contractor may appeal in writing to the Secretary of War, whose written decision or that of his designated representative or representatives thereon shall be final and conclusive upon the parties hereto. The Secretary of War may, in his discretion, designate an individual, or individuals, other than the Contracting Officer, or a board as his authorized representative to determine appeals under this Article. The contractor shall be afforded an opportunity to be heard and offer evidence in support of his appeal. * * *
13. (a) On January 17, 1949, plaintiff submitted to the Fort Buchanan headquarters command its “claim for un-[286]*286billed services supplied * * in the amount of $72,456.50. Tables and data were attached, showing how plaintiff had arrived at the amount of the current supplied and the value thereof.13
(b) On April 4,1949, the contracting officer advised plaintiff that “* * * this office is in a position to discuss with you any additional compensation which you may feel is due to the ,!! * * Authority for the period between the last meter reading prior to 6 July 1948 and that date.” Otherwise, plaintiff’s claim was denied, for the reasons that—
* * * Under the wording of the contract * * * it is apparent that the Government would in no way be responsible for failure of any type of metering equipment installed by the Contractor * * * other than to the extent detailed in * * * paragraph 5 of the Special Conditions.
* * * also * * all invoices submitted for payment during the period in question were certified as true and correct and * * * payment therefor has been made and accepted.
The contracting officer made no findings of fact.
(c) On April 29, 1949, plaintiff appealed the decision of the contracting officer to the Secretary of the Army, who referred the matter to the Armed Services Board of Contract Appeals. On August 25 and 26,1949, a member of the Board, representing the Army Contract Appeal Panel, held hearings in San Juan at which testimony was presented on behalf of both parties.14
(d) On April 30,1951, the Armed Services Board of Contract Appeals denied the appeal “* * * because of the limitation contained in Special Condition No. 5 * * and [287]*287directed the contracting officer “* * * to enter into negotiations fixing the amount to which appellant is entitled, for the period following the last bill before the restoration of normal service on 6 July 1948.” In the course of its opinion, the Board made several findings of fact, some of which are hereinafter noted. Plaintiff contends here, as it has alleged in its petition, that “* * * the decision of the Board is based upon an erroneous interpretation of the Special Conditions of the contract * * Plaintiff has not challenged the findings of the Board as being arbitrary, capricious, grossly erroneous, or as not being supported by substantial evidence.
14. (a) Plaintiff now contends: that it must be inferred from all the evidence (and therefore should be found as a fact) that one of the two fuses was blown out in July 1946; that the failure of this fuse affected the metering devices in such way that they failed to record all of either consumption or demand; that defendant has been underbilled for current during the 24-month period from July 1946 to July 1948; that, in the absence of precision methods for measuring the current used and not billed for, the application of an experience pattern drawn from five months immediately before and after the period in controversy affords a reasonable basis of estimating such current; and that on such basis, defendant has had and used power for which it has not paid of the value of $80,449.78.15
(b) Defendant contends: (1) that any recovery by plaintiff is precluded by the provision of the contract relating to errors in meter readings;16 (2) that (aside from the foregoing contention) the elements of plaintiff’s claim are speculative, wherefore there is no measurable quantum of damages; and (3) that (in the light most favorable to plaintiff) it cannot reasonably be inferred that the fuse was blown out earlier than December 1947.
15. (a) Following are the findings of fact made by the Armed Services Board of Appeals which are pertinent here:
[288]*288* * * The fuses were in the portion of the installation maintained by the Army and appellant had no responsibility whatever for their inspection or replacement.
* * * When they had been replaced full service was restored. * * * When the meters were next read it was discovered that the recorded consumption of energy had increased about 100 percent over the average of the 22 preceding months. * * *
* * * The meters correspond to the requirements of the contract * * *. * * * The metering devices were in proper working order at all times under discussion. * * *
* * * An experiment was made * * * on 26 August 1948 by pulling one of the high voltage fuses * * * and it was found that the current recorded while the fuse was out fell to a quantity comparable to the recorded amounts preceding 6 July 1948. * * *
* * * The record * * * fairly establishes a claim for electrical energy consumed over and above the quantity billed and paid for. * * *
(b) The foregoing findings are adopted here.17
(c) Immediately following the finding last quoted in subsection (a) above, the Board raised the question as to “ * * * the extent to which the contract permits the reopening of the account between the parties for the purpose of allowing a proper claim.” The Board decided that the account could not be so reopened, because of the “error in meter readings” clause. In the course of its opinion the Board stated the following conclusions:
* * * there are elements of uncertainty and speculation in attempting to reconstruct the facts from the only available data — the recorded registrations of the meters.
Study of the graphs submitted at the hearing poses problems to which neither group of-experts has nor can give unqualified answers.
It is the Board’s belief that the clause under consideration was intended to cut short subsequent disputes over the possible solutions to such speculative problems because of the realization of the inherent difficulty in reconstructing from the recorded data the reasons for a subsequently discovered malfunction in the electrical system.
[289]*289(d) The sentence comprising the paragraph last quoted above reflects a conclusion of law. It is not and does not purport to be a finding of fact relating to the state of mind (subjective intent) of the parties or either of them at the time the contract was made.
(e) It is not established by the evidence that all of the graphs that were submitted to the Board are in evidence here. Of those that are in evidence here, it is true, as found by the Board in relation to the graphs before it, that a study of them poses problems to which neither group of experts has nor can give unqualified answers. It is also true here, as in the evidence before the Board, that there are elements of uncertainty and speculation in attempting to reconstruct the facts from the recorded registrations of the meters.18 The Board’s conclusion that the recorded registrations of the meters represent the only available data is not applicable here.
16. At the time of the trial of this case the “error in meter readings” clause was generally used in contracts of the Army, Navy, and Atomic Energy Commission. The extent to which it had been used prior to its insertion in the contract in suit is not established by the evidence. Neither does the evidence establish a trade usage meaning or interpretation of the clause in 1943 or later.
17. As to the following findings of fact, the experts in electrical engineering who testified for the parties either agreed or made no challenge:
(1) When the tree limb fell on the transmission lines on July 6, 1948, a short circuit occurred which burned out one of the three fuses on the primary (high voltage) side of the transformers.
(2) The second of the two such fuses that were replaced on July 6,1948, had burned out at an earlier date.
(3) Electrical service on the secondary side of the transformers (inside the Army post) was drastically curtailed while two of the three fuses on the primary side were out. The curtailed service did not represent the full strength of a single-phase circuit, and users noticed the difference immediately.
[290]*290' (4) Three-phase service on the secondary side continued undiminished with only one of the three fuses on the primary side out. Users did not know the difference.
(5) With one fuse out, the meter no longer recorded accurately either the demand or the consumption. This fact was determined after July 6, 1948, by theoretical analysis confirmed by laboratory tests. No pattern of experience with such a situation was available for comparison and guidance.
(6) It is not known which of the three fuses on the primary side was out for a time prior to July 6,1948. Experiments showed that the meter would be affected and would cease to record accurately when any one of the three fuses was out; that the effect on the meter varied according to which fuse was out; and that the opening of one of the fuses produced results in the meter closely paralleling the discrepancy which plaintiff claims existed for several months prior to July 1948.
(7) Defendant arranged its connections so as to give a balanced load on all phases, insofar as concerned the proportion and use of outlets and units consuming electrical energy. The opening of one of the fuses on the primary side nevertheless caused to “exist a phase unbalance exceeding ten percent * * 19
(8) The meter would not register accurately either demand or consumption when the phase unbalance exceeded ten percent.
(9) An element known as the power factor of the load entered into the inaccuracy of the meter recordings while one of the fuses was out. This power factor varied from time to time. There is no way of knowing what it was at any given time or times. Consequently, there is no way to transform the incorrect meter registrations into correct registrations.
(10) On July 6, 1948, after the two fuses were replaced, the meter recorded accurately both demand and consumption. Its recordings had not been accurate on that day before [291]*291the two fuses were replaced, or for sometime prior to the short circuit of the transmission lines by the falling tree limb.
(11) During the period of inaccurate meter recordings, defendant received and used more electrical energy than was billed and paid for.
18. The following findings are based on inferences warranted by the record as a whole.
(1) The amount of electrical energy used by defendant at Fort Buchanan declined slightly, but never precipitately, between February and September 1946.
(2) After August 1946, and until July 1948, the amount of energy used was fairly constant.
(3) Nothing in the evidence (other than failure of the meter to record accurately) explains the precipitate drop in energy consumption in September 1946 as compared with the consumption of previous months.20
(4) Because of inaccurate meter registrations resulting from a burned out fuse for which it was responsible, defendant received and consumed more current than it paid for during the period from September 1,1946, to July 6,1948.
(5) The increases in load which began in July 1948 were slight during July and August. These two months fairly represent energy consumption comparable to that used just before the inaccurate meter registrations began.21
(6) The current received and used by defendant, between September 1, 1946, and July 6, 1948, in excess of that for which it has paid, was worth, at the rates in effect during the period, the sum of $66,500.00.
CONCLUSION OP LAW
Upon the foregoing findings of fact, which are made a part of the judgment herein, the court concludes that as a matter of law the plaintiff is entitled to recover, and it is therefore adjudged and ordered that it recover of and from the United States sixty-six thousand five hundred dollars ($66,500).