WRIGHT, J., filed an opinion concurring in part and dissenting in part, in which GUNTHER, J., joined.

Argued November 17, 1954.

OPINION BY RHODES, P.J., March 28, 1955:

The Manufacturers Light Heat Company filed with the Pennsylvania Public Utility Commission new tariff supplements providing for the following increases:

(1) Industrial rates estimated to yield additional annual revenue of $1,597,907; (2) a rate applicable only to Lukens Steel Company, Coatesville, Pennsylvania, estimated to yield additional annual revenue of $137,967; (3) general service rates applicable to residential and commercial customers in Pennsylvania estimated to yield additional annual revenue of $4,128,765. The total estimated increases amounted to $5,864,639 for the twelve months ending July 31, 1953, adjusted to reflect rates in effect on that date. The first two tariff supplements were to become effective December 8, 1953; the third tariff supplement was to become effective December 9, 1953. All three tariff supplements were suspended by the commission for a total of nine months from their effective dates. Eighteen complaints were filed prior to the effective dates of the proposed supplements; one complaint was withdrawn. The commission on its own motion instituted an investigation for the purpose of determining whether the proposed rates and charges of Manufacturers were fair, reasonable, just, and lawful; such investigation was to include consideration of the lawfulness of the existing rates and the imposition of temporary rates. A rate case was presented before the commission; the record including exhibits contains 2,041 pages.

The commission, on August 23, 1954, by its order allowed operating revenues of $39,921,985 in contrast to operating revenues of $42,789,698, which were to be realized from the proposed tariff supplements. Manufacturers was directed to file tariffs designed to produce the operating revenues prescribed by the commission. The commission's order therefore granted about $2,900,000 of the proposed increases of over $5,800,000.

From the commission's order appeals to this Court were taken by the City of Pittsburgh (No. 256, April Term, 1954); Mayer China Company (No. 266, April Term, 1954); Ceramic Color Chemical Manufacturing Company (No. 267, April Term, 1954); McDanel Refractory Porcelain Company (No. 268, April Term, 1954); and Manufacturers Light Heat Company (No. 265, April Term, 1954).

The commission disposed of the issues in one order relating to all of the complaints before it. In this opinion we shall consider the questions raised by the various appellants.

The contentions of the appellants on these appeals from the commission's order may be stated as follows:

(1) Did the commission err in determining the amount of accrued depreciation and depletion; (2) did the commission err in determining the amount of annual depreciation and depletion expenses; (3) was it permissible for the commission to reject upward adjustments of test year operating revenues and expenses calculated on basis of ten-year average temperatures; (4) did the commission err in the allocation of property and operating expenses to Pennsylvania retail sales; (5) was it error for the commission to reject proposed upward adjustments of test year expenses for increases subsequent to the cut-off date; (6) was the rate structure unreasonably discriminatory; (7) was the commission's finding of 6 1/2 per cent rate of return arbitrary, unreasonable, and unsupported by the evidence.

The commission determined the fair value of the utility's property used and useful in the public service allocated to Pennsylvania retail sales to be $80,000,000, at July 31, 1953, upon consideration of original cost and reproduction cost [or trended original cost], the latter at the fair average price of materials, property, and labor. As to such measures of value the commission made the following findings: Original cost $62,845,024; original cost trended to average price level of 1952, $100,809,036, to average price level of 1951-52, $99,408,318, to average price level of 1950-52, $96,837,895, to average price level of 1948-52, $91,919,727. In arriving at these amounts, accrued depreciation and depletion premised upon the reserve requirement study were deducted. From original cost there was deducted $16,173,536 (23.60%). From original cost trended the following deductions were made: One-year price level, $43,370,350 (32.40%), two-year price level, $42,650,925 (32.37%), three-year price level, $41,451,110 (32.38%), five-year price level, $39,139,838 (32.37%).

The commission did not use the utility's book reserve for depreciation and depletion and the corresponding trended amounts, but concluded that the reserve requirement study which was made at the commission's direction was a reasonable reflection of the utility's retirement experience with respect to both annual and accrued depreciation and of the extent of depletion.

It is generally recognized that depreciation is the loss not restored by current maintenance which is due to all the factors causing the ultimate retirement of a utility's property. These factors embrace wear and tear, decay, inadequacy, and obsolescence. Lindheimer v. Illinois Bell Telephone Company, 292 U.S. 151, 167, 54 S. Ct. 658, 78 L. Ed. 1182, 1192. Naturally, a utility's depreciation reserve could represent the actual accrued depreciation in the utility's properties. This is seldom, if ever, the case, as the book reserve representing a historical accumulation may be accumulated by various methods which may or may not reflect the mortality which actually has occurred in the utility's properties. To determine fair value, accrued depreciation must be deducted from the various measures of value which have been properly presented, while annual depreciation is an allowance to be made each year as an operating expense.

A properly determined reserve requirement study may be the best practical measure of depreciation. Such study is an analysis of the utility's recent experience in retiring from service its units of property for various causes; it is based on present-day knowledge and judgment concerning lives of property. A reserve requirement study should present a true current annual depreciation expense chargeable to present cost of service. Using the reserve requirement study to calculate accrued depreciation should disclose the consumption of property to date. The same is true as to depletion which is the exhaustion of natural resources; the term "depreciation" is usually limited to exhaustion of artificial resources.

ACCRUED DEPRECIATION AND DEPLETION: The book reserve offered in evidence by the utility showed higher depreciation and depletion than that shown by the reserve requirement study. Thus the book reserve for depreciation and depletion amounted to 32.2 per cent of depreciable and depletable property allocated to Pennsylvania retail sales at the original cost measure of value, or $22,182,836, prior to necessary adjustments. The city argues that the commission was bound to accept the higher book reserve as a measure of accrued depreciation and depletion particularly since the city and the utility relied upon such book reserve.

The commission concluded that the utility's book reserve was not reliable and that it did not afford a proper basis for the calculation of accrued depreciation and depletion. With reference to the record, the commission pointed out its reasons for rejecting the book reserve in determining accrued depreciation and depletion as of July 31, 1953. The reserve for depreciation and depletion was first established by the utility in 1910. In the intervening forty-three years there were twenty-three revisions made in the annual accrual rates. Since 1948 depreciation has been accrued on a service-life basis by plant accounts, and depletion has been determined on a unit of production basis. The commission further pointed out that the estimates of service-lives were not based on actuarial studies of the utility's retirement experience but rather on the judgment of the management and of the operating personnel, and that, no matter how accurate the estimates may have been from time to time in previous years, only by coincidence would the book reserve for depreciation and depletion be a reliable measure of accrued depreciation and depletion at a later date.

It further appears from the record that the utility and its predecessor companies made depreciation accruals regardless of earnings. There is no countervailing evidence in the record that past rate payers were overcharged or that management abused its discretion in determining annual charges for depreciation and depletion. Certainly, the commission was not obliged to find that the excess of book reserve over the reserve requirement was genuine. See Board of Public Utility Commissioners v. New York Telephone Company, 271 U.S. 23, 46 S. Ct. 363, 70 L. Ed. 808. If the commission had found that the utility had always earned at least a fair return on fair value, the excess appearing in the book reserve could have been considered genuine; but if the utility had not earned at least a fair return the excess to that extent would be false. But it must be recognized that, if the amounts credited to the reserve account were excessive and were charged to operating expenses, to that extent consumers would have provided, in effect, capital contributions to the utility. Lindheimer v. Illinois Bell Telephone Company, supra, 292 U.S. 151, 54 S. Ct. 658, 78 L.Ed. 1182.

On the other hand, the commission found that the reserve requirement study which it directed the utility to have prepared resulted in a reliable estimate of accrued depreciation and depletion of the utility's plant allocated to Pennsylvania retail sales at July 31, 1953, with certain exceptions. The commission then proceeded to base its findings of accrued depreciation and depletion on such reserve requirement study. The city's contentions that the reserve requirement study was unreliable related to matters which the commission could accept or reject. From our examination of the record, we are convinced that the reliability of the reserve requirement study was for the commission to determine. The utility's book reserve was not credited over the years so that it necessarily reflected present-day accrued depreciation and depletion. For some periods book reserve had been computed on the basis of gas delivered, fixed yearly percentages, and other unrealistic methods. There were transfers from surplus, and accruals were not always charged to operating expenses. Neither from a historical or engineering point of view can we say that the reserve requirement study was unreliable and the book reserve reliable.

Possibly in this rate proceeding the commission could have used either the reserve requirement or the book reserve as a proper deduction for depreciation and depletion; its conclusion depended upon all the pertinent considerations — historical, equitable, economic, reliability, and accuracy. As we said in Harrisburg Steel Corporation v. Pennsylvania Public Utility Commission, 176 Pa. Super. 550, 559, 109 A.2d 719, the commission, within the sphere of its authority, could have accepted the utility's book reserve if convinced of its reliability and accuracy. The commission is not bound to accept any particular method in estimating accrued depreciation and depletion which are essentially judgment figures, and if based on substantial evidence are binding on appellate review. Philadelphia v. Pennsylvania Public Utility Commission, 174 Pa. Super. 641, 649, 102 A.2d 428; Pittsburgh v. Pennsylvania Public Utility Commission, 174 Pa. Super. 363, 370, 101 A.2d 761. But in so determining accrued depreciation and depletion the book reserve is by no means conclusive and may be inaccurate. Orlosky v. Pennsylvania Public Utility Commission, 171 Pa. Super. 409, 416, 89 A.2d 903. The fact that the book reserve showed a higher or lower amount than the reserve requirement study has no effect on the principles applicable to such administrative finding.

The result of the commission's acceptance of the reserve requirement study and the rejection of the utility's book reserve was to increase the various measures of value. On the other hand, the calculation of annual depreciation and depletion on the basis of the reserve requirement study representing the annual cost of property consumed in rendering service was to reduce the allowable operating expenses of the utility in this rate case.

ANNUAL DEPRECIATION AND DEPLETION — REMAINDER LIFE THEORY: The utility claimed $1,710,478 for annual depreciation and depletion allocated to retail sales in Pennsylvania related to plant at July 31, 1953. The commission allowed a total of $1,305,308 for annual depreciation and depletion made up of $1,231,012 for depreciation and $74,296 representing depletion. The commission computed such annual depreciation and depletion as it had accrued depreciation and depletion, primarily on the basis of the reserve requirement study. We have repeatedly held that accrued and annual depreciation must be calculated on a reasonably consistent basis. Blue Mountain Telephone Telegraph Co. v. Pennsylvania Public Utility Commission, 165 Pa. Super. 320, 67 A.2d 441; Schuylkill Valley Lines, Inc., v. Pennsylvania Public Utility Commission, 165 Pa. Super. 393, 68 A.2d 448; City of Pittsburgh v. Pennsylvania Public Utility Commission, 171 Pa. Super. 187, 90 A.2d 607. As previously stated, the utility's book reserve was higher than the reserve requirement study indicated to be necessary. Thus the reserve requirement study showed $16,173,536, or 23.6 per cent of original cost representing accrued depreciation and depletion, whereas the book reserve unadjusted was $22,182,836, or 32.2 per cent of the original cost of property allocated to Pennsylvania retail sales. It follows that there was an apparent "excess" in the book reserve of approximately $6,000,000. The city urgently contends the commission was bound to apply the remainder life theory and to reduce the utility's annual allowance by an amount sufficient to amortize this excess over the remaining life of the property. The commission, however, gave full weight to the reserve requirement study in the matter of annual expense, consistent with its finding on accrued depreciation and depletion, and refused to give any weight to book reserve or apply the remainder life theory. Certain general principles concerning application of the remainder life theory may be deduced from our recent cases. Whether the remainder life theory shall be applied in any given case, and the extent of its application are primarily matters for the commission. Expressed in another way, the relative weight to be given any particular evidence on depreciation is, under ordinary circumstances, for the commission. Thus the relative weight to be accorded book reserve and a reserve requirement study would be for the commission as the trier of fact. The question, whether an apparent "excess" or "deficiency" in the book reserve is genuine, is subsidiary to the main problem of credibility of evidence. Where the evidence warrants it, we have approved application of the remainder life theory. Harrisburg Steel Corporation v. Pennsylvania Public Utility Commission, supra, 176 Pa. Super. 550, 560, 109 A.2d 719. Here the commission, acting within its proper sphere, refused to give any weight to book reserve, but consistently based both annual and accrued depreciation and depletion on the reserve requirement study. A depreciation reserve set-up on a utility's books is not necessarily conclusive in the eventual recovery of full original cost of a utility's property; and the commission is not always obliged to apply the remainder life theory because of an excess or deficiency in the book reserve as disclosed by a reserve requirement study. In Duquesne Light Company v. Pennsylvania Public Utility Commission, 176 Pa. Super. 568, 107 A.2d 745, we held that the commission, having applied the remainder life theory in a previous proceeding involving the utility, could not subsequently make a complete change of its policy to the utility's prejudice. Under the evidence in the present case the commission was in no way bound to apply the remainder life theory, and its finding as to annual depreciation and depletion will be affirmed.

ADJUSTMENT FOR TEMPERATURE DEFICIENCIES: The city presented evidence showing the difference between the temperatures during the test or base year ending July 31, 1953, and ten-year average temperatures. The city contends the commission erred in not making upward adjustments of operating revenues and expenses for the test year by taking into consideration ten-year average temperatures as well as test year average temperatures. Based on such averages the ten-year average degree-day deficiency was greater than the degree-day deficiency for the test year. It is the view of the city that such adjustments would more accurately reflect for rate purpose the utility's space-heating sales. Such adjustment of gas sales would seem to rest entirely on temperature deviations. The city argues that such adjustments are necessary because those sales are influenced almost solely by temperature. Obviously there are other factors as the commission points out, and the record indicates that the use of gas for space-heating is not directly proportional to degree-day deficiency. The commission rejected any adjustment of test year operating revenues for deviations of test year temperatures from the ten-year average temperatures, as well as any concomitant adjustment to test year operating expenses. There may be merit in the city's proposed adjustments to revenues, expenses, and allocation of cost to reflect ten-year average temperatures. But the evidence used by the commission showing temperatures for the test year was proper and sufficient to support its findings as to revenues and expenses in so far as related to temperature. We cannot say that the test year average temperature readings did not constitute evidence reasonably calculated to reflect average conditions. Neither can we say, as a matter of law, that the commission failed to give proper weight to the evidence before it. See section 1005, Art. X, Act of May 28, 1937, P.L. 1053, 66 P. S. § 1395. The commission is granted a wide area of discretion as to the extent and the type of adjustments to be made to test or base year figures, where such adjustments are warranted by substantial evidence in the record. Duquesne Light Company v. Pennsylvania Public Utility Commission, 174 Pa. Super. 62, 69, 70, 99 A.2d 61. But the commission is not bound to accept any particular evidence or to make any specific adjustments. On the present record the commission, to which is delegated the duty of weighing the evidence, acted within the sphere of its authority as a fact-finding body in refusing to make the adjustments for ten-year average temperatures for which the city has argued.

ALLOCATION: The commission's allocations of the utility's property and operating expenses to Pennsylvania retail sales are questioned by the city. These allocations were based on the combined average of one-day peak and annual MCF sales. In allocating the proportion of total Production and Transmission Plant to Pennsylvania retail sales (the utility operating in other states), the commission used the factor of 50.41 per cent, that is, the commission found that 50.41 per cent of this type of property was devoted to service of Pennsylvania retail customers. This percentage was obtained by taking an equally weighted composite of two ratios, that is, the ratio of peak-day retail sales in Pennsylvania to total peak-day sales by the utility to all of its customers (52.65%) and the ratio of annual MCF sales in Pennsylvania at retail to total MCF sales by the utility to all of its customers (48.17%). Therefore, the peak-day demand factor was the first ratio, 52.65 per cent, and the commodity factor, the second ratio, 48.17 per cent.

It is not questioned that Distribution Plant allocated to Pennsylvania represents the original cost of such property located in Pennsylvania. The commission points out that the portion of Intangible Plant and General Plant allocated to Distribution Plant was allocated between Pennsylvania and the other states on the basis of the ratio of the Distribution Plant in Pennsylvania to the total Distribution Plant.

It is the contention of the city that a three-day peak average, with adjustments for temperature, should have been used as a basis for calculating the demand factor in place of the one-day peak for allocation purposes. The commission gave consideration to this question and rejected the city's contention that the three-day peak should be used in determining allocations to Pennsylvania retail sales, or that the combination of the three-day peak and the ten-year average temperatures should be used for the same purpose.

We find no substantial basis for the city's argument that the classification of operating expenses between demand costs and commodity costs was fatally inconsistent. Under the circumstances, allocations of plant, revenues, and expenses are approximations based upon the evidence. The results reached by the commission seem to be substantially correct. For example, the commission, in its brief, has calculated the allocation of transmission operating expenses to Pennsylvania retail sales would be $1,335,889 under the city's theory, while the commission's allowance was $1,369,483. The other complaints of the city relative to the classification of items as demand costs or commodity costs are unsound or would, if accepted, produce negligible results.

The city's use of three-day peak sales to calculate a demand factor results in a percentage of 50.95 per cent without temperature adjustments, and 52.29 per cent with temperature adjustments; the commodity factor as calculated by the city would be 47.98 per cent with temperature adjustments. The commission's demand factor was 52.65 per cent and commodity factor 48.17 per cent. We have previously discussed the question of temperature adjustments. As to the use of the three-day peak advocated by the city, the commission in its order stated: "While City of Pittsburgh would have us use the three-day peak (January 5, 6, 7, 1953), rather than the single-day peak (February 17, 1953), we are unable to see any justification for such a step. The unalterable fact is that respondent did meet the demands imposed on its system on the single day of February 17, 1953. Plant should be allocated on the basis of single-day peak demand, because the plant was used on such basis to provide the demands imposed by the respective groups of customers."

There was no reversible error in the commission's method of allocation. The day of greatest system-wide demand in the test year to meet the needs of Pennsylvania retail customers was February 17, 1953. The one-day peak represented the day of greatest system-wide usage by Pennsylvania retail customers. The utility's system was obliged to be prepared to meet this demand. The city's proposed method of allocation on the basis of an average of three-day peak sales would not meet the maximum plant requirement. The city contends, however, that, if there was any deficiency of plant to meet a single peak-day demand, such deficiency could be met by the use of gas in transmission lines, that is, "line pack," but such removal of gas would naturally reduce the pressure to some extent and possibly result in inadequacy of service. Allocation of utility property among several states in which a utility serves consumers is necessarily an approximation; and where the commission's findings are supported by proper evidence and the method used is reasonably accurate, we cannot say that an alternative method must be accepted by the commission. These same observations apply to the commission's classification of operating expenses between demand cost and commodity cost. Demand costs were defined as those cost elements incurred in connection with meeting peak-day requirements of the utility's customers; commodity costs were defined as those cost elements incurred in connection with supplying the full volume requirements of the utility's customers. Demand operating expenses were allocated to Pennsylvania retail sales on the basis of the demand ratio of 52.65 per cent; commodity operating expenses were allocated to Pennsylvania retail sales on the basis of the commodity ratio of 48.16 per cent.

There is no reversible error in the commission's allocation of Production and Transmission Plant to Pennsylvania retail sales by use of the composite factor of 50.41 per cent. The city argues that a portion of the cost and expenses of a 14-inch interstate transmission pipeline running from south to north in eastern Pennsylvania was improperly allocated to Pennsylvania retail sales. Allocation of mass property by means of factors reflecting peak-day sales and annual consumption by Pennsylvania retail customers rather than by individual allocations of each transmission line appears to have been reasonably accurate.

ADJUSTMENTS TO TEST YEAR EXPENSES AND REVENUES AFTER CUT-OFF DATE: Subsequent to the cut-off date, July 31, 1953, there were several increases in the price of gas to the utility, totaling $1,779,057. Of these the commission adjusted test year operating expenses upward, and allowed $683,484, being increases occurring September 1, 1953, and October 1, 1953. Later increases in the price of purchased gas, occurring in January, March, April, and May, 1954, were disallowed by the commission. The utility claims the commission erred in not making adjustments for all gas increases and also for $310,628, representing additional wage and related pension expenses. In refusing to make these further adjustments the commission reasoned they were too remote, especially since a substantial increase in revenues and changes in the volume of operations may have occurred; and that it would not be feasible to determine plant and operating revenues in 1954. It has been said that the purpose of a cut-off date is to provide a time when plant, revenues, and expenses can all be synchronized. Although findings are based primarily on test year figures, adjustments may be made by the commission as we have indicated in Duquesne Light Company v. Pennsylvania Public Utility Commission, 174 Pa. Super. 62, 69, 70, 99 A.2d 61. Under the evidence in the present record, it cannot be said that the commission exceeded its authority in refusing to make the adjustments sought by the utility. Cf. City of Pittsburgh v. Pennsylvania Public Utility Commission, supra, 171 Pa. Super. 187, 209, 210, 90 A.2d 607. The objective of test year figures is to reflect typical conditions. Adjustments are made in furtherance of such objective; and where evidence relates to remote changes and would, if accepted, distort test year figures, it may be rejected.

RATE STRUCTURE: The industrial consumer appellants complain of the rate structure of the tariffs which were filed in compliance with the commission's order of August 23, 1954. It is the contention of these appellants that the rate structure unduly differentiates between the smaller industrial consumers and the larger industrial consumers. The rates effective September 7, 1954, under the tariffs filed by the utility to comply with the commission's order, contain seven consumption blocks, while under the previously existing rates there were five consumption blocks. It is obvious that every consumer would pay exactly the same rates for the gas used in each of the consumption blocks, and the percentage increase contained in each of the consumption block rates would be borne by every consumer using gas in that consumption block. There was no uniform percentage increase in the rates effective September 7, 1954. The use of block type rates in rate structures whereby a consumer pays at a lower rate as his consumption increases, that is, as he passes from one block to another, is generally accepted as fair and reasonable. It should allocate equitably basic fixed charges to the cost of delivery of gas at a given point. It is the contention of the industrial appellants, however, that in increasing the rates the same percentage increases should have been applied to the respective blocks. The proposed rates under supplement 5, filed October 7, 1953, and which were the subject of investigation, never became effective. The industrial appellants further argue that the schedule of rates filed to produce the allowable annual operating revenues is not supported by any evidence in the record to show the reasonableness of the classification made by the new tariffs as approved by the commission.

It is true that the prior rate schedule is not res judicata on the question of discrimination or reasonableness; and that in determining whether rates are unreasonably discriminatory the administrative agency must be granted an area of discretion. Pittsburgh v. Pennsylvania Public Utility Commission, 168 Pa. Super. 95, 78 A.2d 35. However, we are unable to find anything in the record which shows, directly or inferentially, upon what basis or for what reason the difference in the various blocks was determined. We agree that the present tariffs are not automatically discriminatory merely because they differ in pattern from the previously existing tariffs. On May 1, 1953, the supplement No. 4 made the rates effective under five consumption blocks. The change in rate classification by supplement No. 5, making proposed increase in the rates, and by the supplement No. 11 making new rates effective September 7, 1954, by reducing the amount of income under the proposed rates, is not supported by any evidence which permits this Court to determine on review the controverted question presented and whether proper weight was given to the evidence. Pennsylvania Power Light Company v. Public Service Commission, 128 Pa. Super. 195, 218, 193 A. 427; Act of May 28, 1937, P. L. 1053, Art. X, § 1005, 66 P. S. § 1395.

A difference in rates between classes of customers upon the basis of the quantity of gas used is permissible and does not necessarily establish unreasonable discrimination; and such difference is ordinarily within "`the flexible limit of judgment which belongs to the power to fix rates.'" Philadelphia v. Pennsylvania Public Utility Commission, 162 Pa. Super. 425, 432, 57 A.2d 613. Philadelphia Suburban, Water Co. v. Pennsylvania Public Utility Commission, 164 Pa. Super. 320, 64 A.2d 500; Pittsburgh v. Pennsylvania Public Utility Commission, supra, 168 Pa. Super. 95, 78 A.2d 35; City of Pittsburgh v. Pennsylvania Public Utility Commission, supra, 171 Pa. Super. 187, 215, 90 A.2d 607.

The changes from the old to the new rates were not made with any ascertainable consistency, and we are unable to determine on the present record whether such discrimination as exists in the tariff supplement effective September 7, 1954, is reasonable. See Pittsburgh v. Pennsylvania Public Utility Commission, 165 Pa. Super. 519, 530, 69 A.2d 844; American Lime Stone Co. v. Public Service Commission, 100 Pa. Super. 158, 162; Act of May 28, 1937, P.L. 1053, Art. III, § 304, 66 P. S. § 1144.

RATE OF RETURN: The utility sought a rate of return of 7 per cent of fair value. The commission allowed a return of 6 1/2 per cent on $80,000,000 which the commission found to be the fair value of the utility's property used and useful in the public service allocated to Pennsylvania retail sales. The industrial appellants contend that on the record the rate of return should be no more than 6 per cent. It is well established that the rate of return necessarily varies with the circumstances of each case, and must be determined from the evidence presented. Solar Electric Co. v. Pennsylvania Public Utility Commission, 137 Pa. Super. 325, 388, 9 A.2d 447; Pittsburgh v. Pennsylvania Public Utility Commission, supra, 165 Pa. Super. 519, 531, 69 A.2d 844.

In arriving at a fair rate of return the commission assumed, and it was not disputed, that the cost of capital to Columbia Gas System, Inc., was identical with the cost of capital to the utility (Manufacturers). On this point the commission stated: "To summarize, reasonable capital cost rates applicable to Columbia and respondent are for debt capital, 3.09 per cent historically and 3.35 per cent on a current basis, and 9.25 per cent for common equity capital. Weighting these cost rates with the average capital structure of Columbia System for the 1949-1953 period, consisting of 52 per cent debt and 48 per cent common equity, results in a composite cost of 6.05-6.18 per cent. If a balanced capital structure were applied, the composite cost range would be 6.18-6.31 per cent. We, therefore, have for consideration in determining fair rate of return a cost of capital within the range of 6.05-6.31 per cent." Such cost of capital is based on the demands of the investors and fully takes into account the hazards and risks incurred in the business, including the fact that "wasting assets" are involved. There is no doubt that the utility obtains its necessary capital through Columbia. At the same time it is evident that the cost of capital to the parent company, consisting of many subsidiaries in various states and jurisdictions, is by no means an accurate determination of cost of capital in this proceeding to this utility operating an intrastate business in Pennsylvania.

In Schuylkill Valley Lines, Inc. v. Pennsylvania Public Utility Commission, supra, 165 Pa. Super. 393, 404, 68 A.2d 448, 454, we said: "The rate of return of utilities should not entirely ignore prevailing low interest rates or low cost of money. High values on the basis of present costs with a very high return — out of line with economic realities — on such reproduction costs would obviously produce an excessive and an unreasonable return." In fixing the fair rate of return, the commission made an allowance above the cost of capital and concluded that 6 1/2 per cent was fair and reasonable. Thus the commission, having once considered current capital cost, made an additional allowance notwithstanding the indication in the record that the current cost of money was downward. We are of the opinion that, if the cost of capital for the utility, that is, Manufacturers, is the same as the cost for Columbia, then on this record the rate of return for the utility cannot exceed the cost of capital as found by the commission for Columbia, that is, 6.05-6.31 per cent. In the first place, the inherent risks involved and the hazards of the industry are necessarily reflected in the cost of capital to Columbia. Moreover, Columbia owns fifteen subsidiaries, including the utility (Manufacturers), and the credit of the utility on this record is obviously higher rather than lower than the composite credit. For instance, in computing savings resulting from the filing of a consolidated return for federal income tax, the commission considered, equalized and made adjustments for various factors affecting individual operating companies in the Columbia system.

Rate of return is not always synonymous with the cost of capital ( Pittsburgh v. Pennsylvania Public Utility Commission, supra, 174 Pa. Super. 363, 375, 101 A.2d 761); but the reasons given by the commission for an allowance above the cost of capital, that is, normal risks and uncertainties associated with financing in the capital markets, had entered into the cost of capital as found by the commission. Having once considered these factors affecting capital costs and having made an allowance therefor, the commission could not gratuitously make a further allowance for good measure. In any event, further allowance above the cost of capital must be based on the evidence and supported thereby.

The commission's ultimate finding as to the rate of return is inconsistent with the commission's basic finding. We recognize that the commission is granted an area of discretion in making estimates and adjustments, and that it is not restricted to unreasonably narrow limits. At the same time it cannot make double allowance for the same factors such as fluctuating capital costs, and it cannot make adjustments, additions, or subtractions as a matter of policy without support in the evidence. The commission's findings, especially in a rate case, must be definite, consistent, supported by the evidence, and sufficient to enable the reviewing court to determine the legal questions involved. Aizen v. Pennsylvania Public Utility Commission, 163 Pa. Super. 305, 60 A.2d 443; Pittsburgh v. Pennsylvania Public Utility Commission, 158 Pa. Super. 229, 44 A.2d 614. While a finding as to the rate of return may be an estimate or approximation, a fair rate of return is capable of being ascertained with a higher degree of accuracy than many other factual questions. The record will be returned to the commission to make proper findings on the rate of return in conformity with this opinion and to make such other adjustments as may necessarily follow.

ORDER

The record is returned to the commission to determine the reasonableness and lawfulness of the rate structure in supplement No. 11 to Tariff Gas — Pa. P. U. C. No. 37, and to determine a proper rate of return. The commission may receive such additional evidence as the circumstances require and make further findings of fact. After such findings the commission shall revise and modify its previous order of August 23, 1954, as may necessarily follow and require proper tariff or tariffs to be filed producing the annual allowable operating revenues.

To the extent stated, the commission's order of August 23, 1954, is set aside, but the rates under the tariffs of September 7, 1954, pursuant to said order of the commission, shall remain in effect subject to the stipulation of October 7, 1954, of the Manufacturers Light Heat Company that it will promptly refund the difference between the amounts collected under said rates and the amounts which it would have collected under rates finally fixed by order of this Court or the Pennsylvania Public Utility Commission.