delivered the Opinion of the Court.
Plaintiff Montana-Dakota Utilities Company (MDU) appeals a December 7, 1986, judgment of the First Judicial District Court, Lewis and Clark County. The District Court upheld the decision of the Montana Public Service Commission (PSC), which reduced the coal expenses used by MDU in setting rates for Montana electric service. We affirm.
MDU raises three issues for our review:
1. Did the PSC err in adopting the “rate of return” method to determine the amount of MDU’s coal expense which could be passed through to Montana ratepayers?
2. Did the PSC err in applying the “rate of return” method to MDU’s coal expense?
3. Does the record support the PSC decision?
In September of 1983, MDU filed a request with the PSC for a $8,731,439 rate increase, using 1982 as the “test year.” As part of its supporting evidence, MDU reported a $15,622,000 expense for lignite coal used by MDU to generate electricity for Montana consumers. The coal expense was paid by MDU to Knife River Coal Mining Company (KRC). KRC is a wholly-owned subsidiary of MDU Resources Group, Inc., of which MDU is a division. The coal expense represented 100 percent of MDU’s coal purchases for 1982. MDU’s purchases represented 25 percent of KRC’s total coal sales for 1982.
Montana Consumer Counsel (MCC), the agency established to re*120present ratepayers’ interests, contended that MDU’s coal expense should be reduced by $585,000. MCC’s expert witness, Dr. John Wilson, compared KRC to six natural resource companies which sold both coal and oil. Dr. Wilson testified that the six companies’ 1982 average rate of return on equity was 15 percent, while KRC’s rate of return on equity was 22 percent.
After hearing the arguments, the PSC disallowed $347,000 of MDU’s claimed coal expense. The PSC determined that KRC’s return rate of 22 percent was too high and “evaded the spirit of regulation.” However, the PSC found weaknesses in both MCC and MDU’s methods for calculating MDU’s coal expense. PSC determined MDU’s annual coal expense for ratemaking purposes by applying a hybrid rate of return of 14.565 percent to KRC on owner’s equity. To arrive at this percentage, the PSC calculated the rate of return for the six companies suggested by Dr. Wilson and one comparable company, Baukol-Noonan, Inc., suggested by MDU. The PSC then adjusted the rate of return to “normalize” KRC’s 1982 test year and also factored in KRC’s tax rate. The PSC calculated the rate of return on KRC’s total year-end equity, but limited its final adjustment to the 25 percent of KRC’s total sales to MDU. Thus, the PSC avoided regulating the price of coal charged by the subsidiary KRC, and yet fixed a fair price for coal expense to be incurred by MDU when considered for ratemaking purposes.
The PSC justified its calculation procedure in Finding 165:
“The classification of coal reserve operations as a nonutility or utility function becomes important to electric ratepayers due to the different ratemaking treatments afforded to the coal fuel expense. It is not clear to the Commission why coal reserves of Knife River Coal Company should be considered a nonutility function with its ratemaking treatment based on comparable profits and prices. Public utilities are required to provide service at the lowest reasonable rate, and the Commission is required to allow rates that reflect the lowest reasonable costs.”
Upon judicial review, the District Court found that the record supported the positions of both parties. However, the court concluded that the PSC “rate of return” analysis method was not an abuse of discretion and affirmed the PSC. The District Court noted that recent Montana Supreme Court decisions emphasize “the reasonableness of the price for coal paid by the parent to the subsidiary and not on the propriety of ratebasing the subsidiary.” (Emphasis added.) MDU appeals.
*121Issue 1. Adoption of “rate of return” method.
MDU contends that PSC’s rate of return method is flawed because four of the six companies were not truly comparable to KRC. MDU argues that these companies deep-mined bituminous coal in the East, where the high overhead costs squeezed profit margins. MDU asserts that Baukol-Noonan, which earned a higher return than KRC, was the only company truly comparable to KRC.
On review of this issue, we note that the PSC is vested by statute with the duty to supervise and regulate the operations of public utilities and to see that their rates are “just and reasonable.” Section 69-3-330, MCA.
In determining what is just and reasonable, the PSC is not restricted to any single formula, if the method followed and the order entered “when applied to the facts and viewed as a whole do not produce an unjust or arbitrary result.” Montana-Dakota Utilities Co. v. Bollinger (Mont. 1981), [_ Mont. _,] 632 P.2d 1086, 1091, 38 St.Rep. 1221, 1227 (where the PSC reduced MDU’s coal expenses from KRC using Dr. Wilson’s rate of return method, but this Court remanded the case to the PSC to determine if a factual basis existed for the rate of return method).
The PSC has the power to adopt any non-arbitrary method it chooses. Bollinger, 632 P.2d at 1091. In the instant case, the PSC adopted Dr. Wilson’s method to compare KRC’s rate of return to the coal industry generally. Perfect comparisons are impossible because coal is not a fungible commodity. Coal has many variables such as density, latent BTU’s, sulphur content, debris, extraction costs, and transportation costs. At the delivery site, the coal consumer’s boilers must be custom-matched to the coal producer’s particular fuel. The comparable which could precisely correlate with the MDU-KRC rate of return does not exist. The PSC used similar comparables, then extrapolated the variables. In doing so, the PSC was simply trying to determine whether the coal expense paid by MDU was reasonable.
The PSC decision conforms to our past holdings on this issue. The PSC adopted a method which was far from perfect, but it was also far from arbitrary. In the balance, we find it was reasonable. We hold that the method was valid on its face and now move to the validity of its application.
Issue 2. Application of the method.
MDU contends that the PSC exceeded its jurisdiction when it ap*122plied the rate of return method to KRC’s sales, thereby improperly treating KRC as part of MDU. MDU asserts that the PSC must limit its scrutiny to KRC’s sales to MDU and may not examine the other 75 percent of KRC sales. MDU further asserts that KRC’s most profitable sales were not to MDU, and therefore these other sales distorted the profit margins on sales to MDU.
In reviewing this issue, we note that the “interest of the PSC is to see that MDU does not reap an unfair profit on its investment in its subsidiary by allowing the subsidiary to overcharge the parent for coal when the coal expense will be passed on to the ratepayers.” Bollinger, 632 P.2d at 1089.
Contrary to MDU’s assertions, MDU’s coal purchases are at issue, not KRC’s coal sales. The key factor in the PSC’s examination was the extent to which KRC functioned as an organ of MDU. As the PSC noted: “The vertical integration by MDU into the coal mining business may provide the Company with an opportunity to circumvent effective return regulation by capturing monopoly profits in its affiliated upstream coal operations.” Throughout the proceedings, PSC maintained its regulatory focus on MDU, not KRC.
In determining a reasonable rate of return, the PSC made several economic corrections in MDU’s favor. The final PSC decision came close to splitting the difference between the $585,000 adjustment advocated by MCC and the zero dollar adjustment advocated by MDU. The final $347,000 adjustment amounted to approximately a 2 percent reduction in MDU’s allowable coal expense. The final adjustment had no effect on KRC, but only on MDU.
As applied, the PSC method scrutinized the captive coal transaction to ensure a reasonable rate. “When one of the expenses submitted by MDU is caused by transactions with a subsidiary company, the scrutiny applied by the PSC must be all the more intense.” Bollinger, 632 P.2d at 1089. We hold that the PSC’s “rate of return on equity” method was properly composed and applied.
Issue 3. Support in the record.
MDU contends that the PSC’s decision is not supported in view of the whole record. MDU argues that its evidence was more probative. MDU asserts that the price KRC charged MDU for its coal was competitive and that MDU could not acquire cheaper coal from any other source. Furthermore, MDU asserts that the contracts to sell coal to MDU were negotiated in an arm’s length transaction by a third-party independent utility. MDU concludes that the PSC *123actions reflected a predetermined decision that PSC was going to reduce MDU’s coal expense regardless of the evidence in the record.
In reviewing this issue, we note:
“. . . The Commission is the judge of the facts and the court only decides questions of law. [Citation omitted.] In deciding questions of law, this Court may determine whether the PSC acted arbitrarily and unreasonably without sufficient evidence to support its findings, or exercised its authority unreasonably, or set the utility rates so low that they are confiscatory and deprive the utility of its property without due process of law. [Citation omitted.] This Court on appeal cannot substitute its judgment for that of the PSC. [Citation omitted.]
Mountain States Telephone & Telegraph Company v. Department of Public Service Regulation, Montana Public Service Commission, et al. (Mont. 1981), [_ Mont __,] 624 P.2d 481, 485, 38 St.Rep. 165.
“Our inquiry is limited to determining whether substantial evidence was presented to support the PSC’s decision or if the decision was unjust or arbitrary.” The Montana Power Company v. Dept. of Public Service Regulation, Mont. Public Service Commission, et al. (1983), 204 Mont. 224, 229, 665 P.2d 1121, 1123 (where the PSC reduced MPC’s coal expenses from its wholly-owned subsidiary using Dr. Wilson’s rate of return method, and this Court affirmed on the basis that substantial evidence supported the PSC’s decision).
Essentially, MDU argues that the decision of the PSC is not sustainable because some of the evidence in the record supports MDU’s position. MDU has misstated the burden of proof. “If a party contends that the order of the PSC is unreasonable, it is creating a fact issue, and under Section 69-3-402(4), MCA, that party has the burden of proof before the reviewing court on the issue. If it is contending that the order or decision of the PSC is unlawful, the appealing party has the burden to show PSC abused its discretion.” Montana Dakota Co. v. Montana Department of Public Service Regulation, et al. (1986), [223 Mont. 191,] 725 P.2d 548, 553, 43 St. Rep. 1648, 1653-1654.
MDU has failed to show how the PSC decision was unreasonable or arbitrary. MDU presented no evidence which tracked its own coal expenses by market price. We note that MDU’s coal purchases from KRC were an affiliated transaction, where coal is bought and sold in a vertically-integrated market. MDU had no market incentive to pay *124KRC the lowest possible price. When MDU paid KRC, MDU was paying itself.
Similarly, MDU did not present any evidence of fair market prices from comparable coal companies. The market price method may actually be less quantifiable than the rate of return method because coal is a commodity with fluctuating prices. In either case, the selection of methods lies squarely within the discretion of the PSC:
“We did not hold [in Bollinger] that the PSC must use a ‘market price’ method if a competitive marketplace can be established; the choice of methods is left to the PSC. In the instant case, the PSC chose to apply a ‘rate of return’ method in an effort to determine the reasonableness of the price paid by MPC for Western Energy coal.”
MPC, 665 P.2d at 1123.
Alternatively, MDU argues that the rate of return on equity must be based in part on the fair market value of MDU coal assets and not the depreciated original cost of the coal. This argument is flawed due to the great difference between the current price of coal and the original price of the coal reserves. In exercising its authority to determine a rate of return for MDU, the PSC essentially concluded that MDU’s proposed return on assets method, valuing coal reserves at current market prices, would result in a rate of return in excess of what the PSC determined to be reasonable.
In conclusion, we hold that the PSC’s method of reducing MDU’s coal expense is supported by substantial, credible evidence. The PSC used both Dr. Wilson’s and MDU’s comparables. The PSC’s final calculation represented a compromise. In both Bollinger and MPC, the issue turned on the sufficiency of evidence, and not on the superiority of method. PSC’s summary of evidence tracks in its conclusion. “So long as the record supports the [PSC’s] decision, that judgment must stand.” MPC, 665 P.2d at 1124.
Affirmed.
MR. JUSTICES HARRISON, SHEEHY, McDONOUGH and HUNT and HON. C.B. McNEIL, District Judge, sitting for MR. JUSTICE GULBRANDSON, concur.