OPINION
{1} The Public Regulation Commission (PRC) granted Southwestern Public Service Company’s (SPS) application to (1) include a prepaid pension asset in its rate base in order for SPS to earn a return on this asset, and (2) obtain a renewable energy cost rider to recover approximately $22 million of renewable energy procurement costs from those customers who do not have a legislatively imposed limit on their renewable energy costs (non-capped customers). The Attorney General appeals the PRC’s final order granting SPS’s application, arguing that the approved rates are unjust and unreasonable because the inclusion of the entire prepaid asset in the rate base is not supported by substantial evidence, and the PRC acted contrary to law in allowing SPS to recover the aforementioned renewable energy costs from non-capped customers. We affirm the PRC because (1) SPS is entitled to earn a reasonable rate of return on the investor-funded prepaid pension asset, and (2) SPS may recover its renewable energy costs in excess of the large customer cap from non-capped customers because such a recovery mechanism is the only viable method of cost recovery that is consistent with the purposes of the Renewable Energy Act, NMSA 1978, §§ 62-16-1 to -10 (2004, as amended through 2011).
I. THE INCLUSION OF SPS’S PREPAID PENSION ASSET IN THE RATE BASE
{2} SPS applied to the PRC to include a prepaid pension asset in its rate base to allow its shareholders, who funded the asset, to receive a corresponding return on their investment. By including this prepaid pension asset in the rate base, the asset is treated as a capital investment, allowing SPS to recover *716the asset as an expense, thereby increasing SPS’s revenue requirement. See Joseph P. Tomain, Symposium Article, “Steel in the Ground Greening the Grid with the iUtility, 39 Envtl. L. 931, 945-46 (2009) (providing and discussing the rate making formula, which sets the amount of money utilities may receive for their investments and expenses). Importantly, inclusion of an investment asset in the rate base does not enable investors to recover the value of their investment, but instead only allows investors to earn a return on the asset. See id. (noting that utilities generally recover the value of an investment by treating the depreciation of the asset as an operating expense).
{3} The parties agree that a prepaid pension asset is the amount by which investor contributions to a pension trust and earnings on these contributions exceed pension expenses. S. Co. Servs., Inc., 122 FERC ¶ 61,218, at *62235, 2008 WL 630079, slip copy at 5 (FERC 2008) (order on tariff filing), order clarified by 128 FERC ¶ 61,276, 2009 WL 3043950 (slip copy) (FERC 2009); In re Delmarva Power & Light Co., 2014 WL 3964914, slip copy at 18, 315 P.U.R. 4th 10 (Del. P.S.C. 2014) (“A prepaid pension asset occurs when the accumulated contributions and growth in the pension plan exceed the accumulated expenses associated with the pension obligations.”). For example, SPS’s expert stated that if the annual pension contribution over a five-year period is $100 and the annual pension expense over the same period is only $90, at the end of the five-year period, the prepaid pension asset would be $50 ($100 x 5 - $90 x 5), plus any return on the $50 prepaid pension asset. “Conversely, when [accrued expenses] exceed[] contributions to [a] fund, a prepaid pension liability accrues.” See In re Sw. Pub. Serv. Co., 2008 WL 4226018 n.256, slip copy at 114 (NMPRC) (final order partially adopting recommended decision), order clarifying final order sub nom. 2008 WL 9888273 (slip copy) (NMPRC 2008). The SPS expert also stated that the prepaid pension asset is an artifact of timing; over a long period, pension contributions and pension expenses may even out, but over short and intermediate periods there will surely be differences, which are recorded as either prepaid pension assets or pension liabilities.
{4} SPS’s expert testified that pension contributions and expenses differ because the federal Employee Retirement Income Security Actof 1974, 29 U.S.C. §§ 1001-1461 (2011), and the Internal Revenue Code, 26 U.S.C. §§ 1-59 (2012), dictate how much the utility must contribute to its employee pension program, whereas the Financial Accounting Standards Board promulgates codified accounting standards1 that govern how pension expenses are determined. The expert continued by testifying that as a result of these differing federal and industry standards, pension contribution and expense calculations utilize different assumptions, attribution methods, and periods of time over which the costs are required to be recognized. SPS’s expert stated that these dissimilarities often result in differing annual contribution and expense amounts. When mandated contributions and income earned on the contributions exceed expenses, a prepaid pension asset accrues. See S. Co. Servs., Inc., 122 FERC ¶ 61,218, at *62235, 2008 WL 630079, slip copy at 5; see also In re Delmarva Power & Light Co., 2014 WL 3964914, slip copy at 18.
{5} The expert witness also testified that utilities cannot legally withdraw any funds from pension trusts except to pay pension benefits and expenses. See S. Co. Servs., Inc., 122 FERC ¶ 61,218, at *62235, 2008 WL 630079, slip copy at 5. However, SPS *717customers benefit from a prepaid pension asset because the earnings on this asset are deemed to be income for SPS, which reduces the amount of revenue it must collect from its customers. See Ind. Office of Util. Consumer Counselor v. Ind. Mich. Power Co., 1 N.E.3d 1025, 2014 WL 934350, at *12 (Ind. Ct. App. 2014) (memorandum decision) (nonprecedential). The following hypothetical offered by SPS’s testifying expert illustrates the indirect benefit SPS customers receive.
[S]uppose that in a given year the utility had a revenue requirement of $300, and that it expected to earn a 6% return on the pension fund. The $3.00 return on [a hypothetical] $50 prepaid pension asset (0.06 x $50) . . . would be credited against the revenue requirement, so that the utility could only collect $297 from its customers through [the] rates. Thus, the revenue requirement is reduced by $3.00 as a result of the prepaid pension asset.
SPS customers therefore would benefit from rate reductions generated by the prepaid pension asset, but SPS would not earn a return on the prepaid pension asset if the asset is not included in SPS’s rate base.
{6} In this case, the New Mexico jurisdictional share2 of SPS’s prepaid pension asset is approximately $36.9 million. According to SPS, this asset resulted in $1.7 million in earnings that effectively reduced SPS’s pension expense by $ 1.7 million, which reduced SPS’s revenue requirement by the same amount. SPS sought “to include the net amount of its prepaid pension asset of approximately $22 million” in the rate base to earn a return on its $22 million (the $36.9 million asset minus a $14.9 million tax deferred asset).
{7} In a recommended decision, the PRC hearing examiner concluded that because the prepaid pension asset reduced the pension expense by $1.7 million, that $1.7 million should be included in the rate base for recovery. The PRC hearing examiner did not recommend that the $22 million net prepaid pension asset amount be included in the rate base. SPS disagreed with this recommendation, contending that the examiner’s proposal would enable “SPS [to] earn a return only on the amount of the reduction in the cost of service rather than on the amount of the asset that resulted in the reduction.”
{8} The PRC also disagreed with the hearing examiner. In its final order, the PRC authorized the inclusion of the net amount of the prepaid asset in SPS’s rate base because doing so “recognizes that ratepayers benefit from the prepaid pension asset and that the utility should earn a return on the prepaid pension asset in order for the utility to recover its full cost of service.” The Attorney General appeals, arguing that substantial evidence does not support the inclusion of the entire prepaid pension asset within the rate base.
A. Standard of Review
{9} In determining whether a PRC final order is supported by substantial evidence, we review the whole record, “view[ing] the evidence in the light most favorable to the decision made by the [PRC].” PNM Gas Servs. v. N.M. Pub. Util. Comm ’n (In re PNM Gas Servs.), 2000-NMSC-012, ¶ 4, 129 N.M. 1, 1 P.3d 383 (internal quotation marks and citation omitted). “ ‘Substantial evidence’ [is] such relevant evidence as a reasonable mind might accept as adequate to support a *718conclusion.” Rinker v. State Corp. Comm ’n, 1973-NMSC-021, ¶ 5, 84 N.M. 626, 506 P.2d 783. “The supreme court shall have no power to modify the action or order appealed from [(in this case, a PRC final order)], but shall either affirm or annul and vacate the same.” NMSA 1978, § 62-11-5 (1982). “The [PRC] is vested with considerable discretion in determining whether a rate to be received and charged is just and reasonable.” Hobbs Gas Co. v. N.M. Pub. Serv. Comm 'n, 1980-NMSC-005, ¶4, 94 N.M. 731, 616 P.2d 1116. A party challenging a PRC final order has the burden of establishing that the order is “arbitrary and capricious, not supported by substantial evidence, outside the scope of the agency’s authority, or otherwise inconsistent with law.” N.M. Indus. Energy Consumers v. N.M. Pub. Regulation Comm’n, 2007-NMSC-053, ¶ 13, 142 N.M. 533, 168 P.3d 105 (internal quotation marks and citation omitted).
B. The Prepaid Pension Asset
{10} The Attorney General contends that only the earnings generated by the prepaid pension asset should be included in the rate base because this is the amount by which the ratepayers have benefitted, or the amount by which the utility’s revenue requirement is reduced. The Attorney General argues that only $1.7 million should be included in the rate base, whereas the PRC’s final order enables SPS to include $22 million in the rate base, which is the net amount of its prepaid pension asset. In resolving this issue, we explain the rationale for electric utility regulation.
{11} Electric utilities are regulated because their industry has natural monopoly characteristics. Joseph P. Tomain, The Persistence of Natural Monopoly, 16 Nat. Resources & Env’t. 242, 242 (2002). In natural monopoly settings, both the benefits and the possibility of competition are limited. Omega Satellite Prods. Co. v. City of Indianapolis, 694 F.2d 119, 126 (7th Cir. 1982). If the electric industry was a competitive free-for-all, different companies would attempt to build separate electric grids and sign up customers as quickly as possible to reduce their average costs of business more rapidly than their rivals. See id. This competitive process would last until a single company was left standing “because until a company serves the whole market it will have an incentive to keep expanding in order to lower its average costs.” Id. Thus, until a single company wins, competition within the electric industry would produce wasteful duplicate grids that would needlessly raise average costs for consumers. See id.; Tomain, The Persistence of Natural Monopoly, supra, at 242 (“A specific service area needs only one set of electric . . . wires — the investment in any other set of wires is wasteful.”). To avoid wasteful duplication, a government may choose to give one firm a monopoly within a service area “in exchange [for] a commitment to provide reasonable service at reasonable rates.” Omega Satellite Prods. Co., 694 F.2d at 126.
{12} Electric utility regulation consequently reflects a compact between utilities and the public. See Jersey Cent. Power & Light Co. v. Federal Energy Regulatory Comm'n (FERC), 810 F.2d 1168, 1189 (D.C. Cir. 1987) (Starr, J., concurring). A utility is given a monopoly over a service area, and in exchange accepts government regulation of its business, including price regulation. Id. Under this arrangement, utility investors obtain a stability in earnings that would likely be unattainable in less regulated industries, while “ratepayers are afforded universal, non-discriminatory service and protection from monopolistic profits.” Id.
{13} Regulators attempt to set prices that mimic market conditions and ensure that utilities are “profitable enough to attract *719capital investment.” Tomain, "Steel in the Ground, ” supra, at 945. The following rate making formula traditionally determines utility revenues to be received from ratepayers: R = O + (V - d)r, where
R represents the utility’s revenue requirement — that is, the amount of money the utility needs to stay in business. O represents the utility’s prudently incurred expenses. In short, ratepayers reimburse the utility for its expenditures dollar for dollar. The utility’s rate base is represented by (V - d), which stands for the value of a utility’s capital investment minus depreciation, which is returned to the utility as expenses. Finally, r represents the rate of return on the rate base.
Id. at 945-46.
{14} The utility’s rate base — the total amount of investment made by a utility to provide its service — is determined by adding the utility’s investment in physical properties to its working capital. Cent. La. Elec. Co. v. La. Pub. Serv. Comm’n, 373 So. 2d 123, 129 (La. 1979). Thus, a utility can include physical properties such as a power plant, see, e.g., Hobbs Gas Co., 1980-NMSC-005, ¶ 6, and working capital — operating funds essential to pay for current obligations — in its rate base. Gov't of Guam v. Fed. Mar. Comm 'n, 329 F.2d 251, 256 (D.C. Cir. 1964); Ariz. Pub. Serv. Co., 5 FERC ¶ 63,038, at *65179, 1978 WL 16416, slip copy at 2 (FERC 1978) (defining working capital as “the money which a utility puts up to finance the services provided until it is compensated by customers”).
{15} In the context of utility regulation, working capital
does not include the total liquid funds with which the business is conducted. It is not the property which the business has; that is, it is not the excess of current assets over current liabilities. Working capital, rather, is an allowance for the sum which the company needs to supply from its own funds for the purpose of enabling it to meet its current obligations as they arise and to operate economically and efficiently.
Gov’t of Guam, 329 F.2d at 256 (internal quotation marks and citation omitted). As a result, only utility contributions, not ratepayer contributions, can be properly included in the rate base as working capital. For example, if a utility were to prepay for natural gas with investor funds, the utility should expect to receive a reasonable return on its investment. Zia Nat. Gas Co. v. N.M. Pub. Util. Comm 'n (In re Zia Nat. Gas Co.), 2000-NMSC-011, ¶ 22, 128 N.M. 728, 998 P.2d 564. Conversely, if ratepayers have paid in advance for the natural gas, the utility would have no expectation of a return because its capital was not used to buy the natural gas. Id.
{16} A utility can include prep ayments for pension expenses in its rate base “because the utility is out-of-pocket for such costs until they are recovered from ratepayers and is therefore entitled to recover its cost of financing such prepaid expenses.” S. Co. Servs., Inc., 122 FERC ¶61,218, at *62235, 2008 WL 630079, slip copy at 5. For example, in the context of prepaid pension assets, income earned on the pension fund is reported under generally accepted accounting principles as a reduction to the utility’s pension expense. Id. “If that reduction in pension expense is used in determining a utility’s rates, there will be a corresponding reduction in the amounts collected from ratepayers.” Id. Under these circumstances, the utility must finance the reduction because it cannot use the income from the pension trust to pay other current *720obligations; as a result, the utility is allowed to recover the costs of financing the reduction by including the pension income in the rate base. See id. The Attorney General’s position is that the utility can only recover the costs of financing the reduction of the utility’s revenue requirement, i.e., the utility can only earn a return from the pension income generated by the prepaid pension asset.
{17} However, a utility may not only be out-of-pocket for reductions in its revenue requirement resulting from pension fund earnings. A utility may also be out-of-pocket for investor-funded contributions that are in excess of pension expenses. Basically, when a utility supplies working capital to fund contributions in excess of pension expenses to create an income-producing prepaid pension asset, the utility finances the cost of the entire prepaid pension asset. See, e.g., In re Rocky Mountain Power, 2014 WL 7526282, at *14 ¶¶ 52, 53, *36 (Wyo. P.S.C. 2014) (noting that a “prep aid pension asset represents [a utility ’ s] contributions to its pension ... plans in excess of what is expensed to that time” and the utility “finances the asset with a combination of debt and equity financing”).
{18} Other jurisdictions have allowed utilities to recover the financing costs of the net prepaid pension asset by including the asset in the rate base as a component of working capital. See, e.g., Ind. Office of Util. Consumer Counselor, 7 N.E.3d 1025, 2014 WL 934350, at *12 (upholding a regulatory determination that a prepaid pension asset be included in the rate base because the “asset amounted to working capital that benefited the ratepayers by reducing the total pension costs needed in [the utility’s] revenue requirement”); In re Rocky Mountain Power, 2014 WL 7526282, at *14, *36 (finding persuasive a utility’s argument that it should recover the financing costs of its prepaid pension asset by including the asset in the rate base to enable the utility to earn a return on that asset). But see In re Pub. Util. Comm ’n of Or., 2015 WL 4710466, at *7 (Or. P.U.C. 2015) (affirming a “long-standing policy of allowing a utility to recover its pension contributions [only as an] expense and rejecting] the . . . Utilities’ proposal to include their current prepaid pension assets in rate base”).
{19} On appeal, the Attorney General does not argue as a matter of law that the prepaid pension asset cannot be included in the rate base. The Attorney General’s only evidentiary challenge is that inclusion of the net prepaid pension asset will result in ratepayers paying more to SPS than the benefit ratepayers have enjoyed from the pension fund earnings. We interpret the Attorney General’s argument to be that SPS did not prove how much of the net prepaid pension asset resulted in consumers paying $1.7 million less to SPS. We disagree. We hold that some or all of a prepaid pension asset should be included in the rate base to the extent that the evidence evinces that the asset was investor-funded, as opposed to ratepayer-funded.3 See In re Potomac Elec. Power Co., 2008 WL 516553, slip copy at 29, 263 P.U.R. *7214 th 1 (D.C. P.S.C.) (finding that “investor-supplied cash contributions have resulted in an asset from which [utility] customers receive a tangible benefit in the form of reduced pension expenses” and including the prepaid pension asset in the rate base), adhered to on denial of reconsideration sub nom. 2008 WL 483 1456 (slip copy) (D.C. P.S.C. 2008); In re N. Ill. Gas Co., 2005 WL 2445944, slip copy at 23 (Ill. C.C. 2005) (noting that a prepaid pension asset “was created by ratepayer-supplied funds, not by shareholder-supplied funds,” and finding that the “prepaid pension asset should be eliminated from rate base”); In re Zia Nat. Gas Co., 2000-NMSC-011, ¶ 22 (noting that only investor-supplied working capital may be included in the rate base); In re Cent. Tel. Co. of Tex., 19 Tex. P.U.C. Bull. 929, 1993 WL 595464, slip copy at 13 (Tex. P.U.C. 1993) (concluding that conversely, when ratepayer-supplied money overfunds a pension plan, investors are not entitled to “earn a return on the prepaid pension asset because [this] . . . would have the effect of charging ratepayers again for amounts they have already paid”). Similarly, while a prepaid pension asset may be included in the rate base, prepaid pension liability must be subtracted from the rate base. See, e.g., In re Ky.-Am. Water Co., 1997 WL 34863470, slip copy at 10 (Ky. P.S.C. 1997) (noting that although pension liabilities can be utilized to reduce the rate base, if “a pension asset is created, then the asset should be included as a rate base addition”), opinion modified on denial of reh 'g sub nom. In re Adjustment of the Rates of Ky.-Am. Water Co., 1997 WL 34863471 (slip copy) (Ky. P.S.C. 1997).
{20} The evidence indicates that SPS has a net prepaid pension asset of approximately $22 million. The evidence also indicates that including $22 million of the net prepaid pension asset in the rate base would generate approximately $2.5 million in revenue for SPS, which exceeds the $ 1.7 million by which SPS asserts the pension expense was reduced. SPS maintains that its actual annual pension expense is $5.36 million, but the $1.7 million return on the prepaid pension asset reduced the pension expense to $3.66 million.
{21} Although the Attorney General is correct to make an evidentiary contention, the premise of its argument is incorrect. Utilities are able to recover the costs of financing their business operations through the inclusion of investor-supplied working capital in the rate base. See In re Zia Nat. Gas Co., 2000-NMSC-011, ¶ 22. In his written testimony, Gene H. Wickes stated that “[t]he portion of the prepaid pension asset due to these contributions has therefore come exclusively from shareholder capital and should be included in rate base.” It is uncontested that SPS investors made contributions to the pension fund that are required by law. These contributions exceeded expenses and generated earnings that effectively reduced SPS’s — and consequently the ratepayers’ — pension expense. Had the ratepayers advanced the contributions to the pension fund, their contributions would not have been included in the rate base. See In re N. Ill. Gas Co., 2005 WL 2445944, slip copy at 14. However, because the ratepayers did not make the contributions, the investors, notthe ratepayers, absorbed the cost of funding the pension program, and therefore the net prepaid pension asset was properly included in the rate base. See, e.g., In re Pub. Serv. Co. of Colo., 1993 WL 494141, slip copy at 17, 148 P.U.R. 4th 1 (Colo. P.U.C. 1993) (“In order to compensate investors for the additional funds they supply to meet the higher contribution levels, the resulting prepaid assets are an appropriate addition to rate base.”); In re Potomac Elec. Power Co., 2008 WL 4831456, slip copy at 3 (concluding that inclusion of an investor-supplied prepaid pension asset in the rate base is supported by substantial evidence because “the earnings on *722the prepaid pension asset will reduce the annual [utility] expense, thus benefiting customers by reducing the revenue requirement”); Ind. Office of Util. Consumer Counselor, 7 N.E.3d 1025, 2014 WL 934350, at *12 (upholding a regulatory determination that a prepaid pension asset amounted to working capital that should be included in the rate base).
{22} We note, however, that contributions to pension funds should be scrutinized to ensure that utility investments are “used and useful” so as to inure to the benefit of consumers. See N.M. Indus. Energy Consumers v. N.M. Pub. Serv. Comm ’n, 1986-NMSC-059, ¶ 29, 104 N.M. 565, 725 P.2d 244 (internal quotation marks omitted) (noting that the “ ‘used and useful’ concept is but one factor among many to be considered by the [PRC] in its rate base analysis”). Utilities should not voluntarily overfund their pension funds simply to earn a favored rate of return. In re Appalachian Power Co., 2011 WL 2150661, slip op. at 27, 288 P.U.R. 4th 185 (W. Va. P.S.C. 2011) (“Prepayments should be subject to the same review as any other investment or expense of a utility. Inclusion of prepayments in rate base should not be used for a utility to find a convenient place to deposit funds and then expect to earn a return on those funds.”). On the other hand, mandatory contributions to pension funds are useful. Such contributions may benefit customers by generating an income-earning prepaid pension asset to reduce pension expenses, see, e.g., In re Potomac Elec. Power Co., 2008 WL 516553, slip copy at 29, and also fund the pension programs that make it possible for the utility to attract and retain highly-skilled workers. See, e.g., In re Advice Letter No. 830 - Gas of Pub. Serv. Co. of Colo., 2013 WL 5799983, at *46-47 (Colo. P.U.C.).
{23} We conclude that the Attorney General has failed to meet its burden of showing that the PRC’s inclusion of the entire prepaid pension asset was unreasonable or unlawful for lack of substantial evidence. See In re PNM Gas Servs., 2000-NMSC-012, ¶ 4.
II. THE LAWFULNESS OF THE RENEWABLE PORTFOLIO STANDARDS RIDER
A. SPS’s Recovery of Renewable Energy Procurement Costs from Non-Capped Customers
{24} The Attorney General also contends that the PRC acted contrary to law when it approved SPS’s renewable energy cost rider because the rider sought to recover renewable energy costs from non-capped customers, customers who are not subject to a legislatively imposed limit on their renewable energy costs. We review issues of law de novo. N.M. Attorney Gen. v. N.M. Pub. Regulation Comm’n, 2013-NMSC-042, ¶ 10, 309 P.3d 89. However, “[w]hen an agency that is governed by a particular statute construes or applies that statute, [we] will [accord] some deference to the agency’s interpretation.” Id. ¶ 12 (internal quotation marks and citations omitted). We will reverse the agency’s interpretation of a statute if it is unreasonable or unlawful. Id.
{25} The resolution of this issue necessitates a discussion of the Renewable Energy Act, §§ 62-16-1 to -10. As a preface to our discussion of this issue, some background on renewable energy promotion is warranted.
{26} Under the traditional rate formula, utilities receive a reasonable rate of return for capital project investments such as power plants. See Tomain, "Steel in the Ground," supra, at 946. Utilities consequently have an incentive to invest in capital projects, see id., and “prefer low-risk, conventional technologies that can be built quickly instead *723of long-term, innovative technologies that would be riskier.” Virginia R. Hildreth, Comment, Renewable Energy Subsidies and the GATT, 14 Chi. J. Int'l L. 702, 707 (2014). As a result, “government assistance is often key to encourage investment in industries like renewable energy.” Id. Such encouragement is desirable because there are numerous benefits to renewable energy such as “lessened dependence on foreign fossil fuel supplies, heightened national security, overall cleaner air, and local and rural job creation.” Shelley Welton, From the States Up: Building a National Renewable Energy Policy, 17 N.Y.U. Envtl. L.J. 987, 995 (2008); see also Brent M. Haddad & Paul Jefferiss, Forging Consensus on National Renewables Policy: The Renewables Portfolio Standard and the National Public Benefits Trust Fund, 12 The Elec. J. 68, 69 (Mar. 1999) (listing benefits of renewable energy).
{27} Renewable portfolio standards are among the most popular methods of encouraging renewable energy development. See Lincoln L. Davies, State Renewable Portfolio Standards: Is There A “Race ” and Is It “To the Top ’’?, 3 San Diego J. Climate & Energy L.3, 10(2011-2012). These standards mandate that utilities incorporate renewable energy sources into their electric generation portfolios, id. at 13, and frequently enable utilities to purchase renewable energy credits4 to satisfy the mandates of renewable portfolio standards. Id. at 11. A renewable portfolio standard therefore combines “both a potentially inflexible regulatory directive and the malleable tool of economic trading,” id. at 10, to “inject an element of economic efficiency into [renewable portfolio standard] schemes.” Id. at 11.
{28} Under the Renewable Energy Act, New Mexico has a renewable portfolio standard that both mandates the incorporation of renewable energy sources into electric generation portfolios and allows for the purchase of renewable energy certificates (credits) to satisfy the mandates. Sections 62-16-4 to -5. Pursuant to this renewable portfolio standard, before the proceedings in this case, the evidence indicates that SPS received PRC approval to (1) purchase the outputs of two New Mexico wind farms, (2) pay incentives encouraging customers to install solar and biomass generation systems, (3) obtain renewable energy credits from various sources, and (4) purchase solar photovoltaic systems.
{29} The controversy over the permissibility of SPS’s proposed rider arises from a disagreement as to how renewable energy costs are allocated between different rate classes. In utility regulation, customers are often divided into different classes that are charged different rates. II Leonard Saul Goodman, The Process ofRatemaking 964-65 (1998). The creation of rate classes involves the consideration of various factors such as alternate fuel capability and types of customer, which can be classified as residential, commercial, or industrial. Id. at 965. Differential rates can be utilized to implement various policies. Tomain, “Steel in the Ground," supra, at 946-47.
{30} In this case, cost allocation has been utilized to address a problem that is incidental to the promotion of renewable energy generation. The use of renewable energy tends to raise energy costs relative to the consumption of fossil fuels. See Hildreth, supra, at 716 (“The technology needed for renewable energy tends to be more expensive than traditional fuel sources.”); Trevor D. Stiles, Renewable Resources and the Dormant Commerce Clause, 4 Envtl. & Energy L. & Pol’y J. 34, 43-44, 45 (2009) (numerically *724illustrating how there is a “vast price discrepancy between renewable energy sources and fossil fuel sources for energy generation”). The prospect of “overly high renewable implementation costs” has prompted concern that commercial and large industrial customers may leave the utility system or exit states that implement prohibitively expensive renewable energy promotion plans. California Commissioner Seeks Consideration of Shale Gas, 4054 PUR Util. Reg. News 1, 1 (Jan. 20, 2012). These large customers may have the capacity to self-generate their energy needs or simply close their plants in areas where energy costs are high. See Charles G. Stalon & Reinier H.J.H. Lock, State-Federal Relations in the Economic Regulation of Energy, 7 Yale J. on Reg. 427, 449 (1990). When these large customers are driven from the utility system, utility rates have to be raised even further for remaining customers, which exacerbates the potential for other customer exits. See id. In light of the potential for large customers to exit the grid, the Legislature enacted Section 62-16-4(A)(2), which limits the annual amount large customers can be charged for renewable energy procurement. The PRC calls this limit the “large customer cap.” Accordingly, costs that exceed the large customer cap may be called “large customer cap costs.”
{31} In earlier proceedings, the evidence indicates that the PRC had already approved SPS’s “requested procurements without any reduction to SPS’s overall [renewable portfolio standards] to account for large customer cap costs.” When the PRC learned that SPS’s costs exceeded the large customer cap, the PRC specifically approved treatment of that amount as a deferred cost. “A ‘deferred cost’ is one that has been paid by the [utility] but is postponed for inclusion in rates until a future period.” I Leonard Saul Goodman, The Process of Ratemaking 321 (1998). This may occur, for example, when a utility “has a major future liability, and before collecting anything through rates, its management decides that the books should reflect the liability.” Id. Under these circumstances, a utility “may apply for [regulator] approval to fund an account, and to reflect on its books a deferred debit or ‘regulatory asset,’ . . . which later can be charged to ratepayers and amortized over a future period.” Id. Regulatory assets are often created to spread out the recovery of nonrecurring costs over a period of years so as to avoid substantial rate increases, which may occur if full recovery was allowed as soon as the utility made an expenditure. City of Corpus Christi v. Pub. Util. Comm’n of Tex., 51 S.W.3d 231, 244-45 (Tex. 2001).
{32} SPS filed an application seeking to obtain a rider to recover approximately $22 million of renewable energy procurement costs. Riders are surcharges applied to directly recover specific costs. See Chesapeake Utils. Corp. v. Del. Pub. Serv. Comm’n, 705 A.2d 1059, 1063 (Del. Super. Ct. 1997) (referring to a rider as a surcharge); Citizens Util. Bd. v. Ill. Commerce Comm ’n, 651 N.E.2d 1089, 1102 (Ill. 1995) (“[A] rider mechanism . . . facilitates direct recovery of a particular cost.”). These surcharges give regulators more flexibility in spreading out costs charged to ratepayers over a period of time. See Chesapeake Utils. Corp., 705 A.2d at 1063 n.3.
{33} Section 62-16-4(A)(l)(a)-(d) mandates that a certain percentage of a “public utility’s total retail sales to New Mexico customers” be comprised ofrenewable energy. The required percentage escalates over time. See id. However, under the large customer cap provision of Section 62-16-4(A)(2), the renewable portfolio standards mandated in Section 62-16-4(A)(l)
shall be reduced, as necessary, to provide for the following specific procurement requirements for *725nongovernmental customers at a single location or facility, regardless of the number of meters at that location or facility, with consumption exceeding ten million kilowatt-hours per year [(capped customers)]. On and after January 1, 2006, the kilowatt-hours of renewable energy procured for these customers shall be limited so that the additional cost of the renewable portfolio standard to each customer does not exceed the lower of one percent of that customer’s annual electric charges or forty-nine thousand dollars ($49,000) [(large customer cap)].
Section 62-16-4(A)(2). The large customer cap in Section 62-16-4(A)(2) also escalates over time such that capped customers can continue to be charged increasing amounts for renewable energy.
{34} The evidence indicates that SPS sought to recover the renewable energy procurement costs that exceeded the large customer cap from non-capped customers. The Attorney General opposed SPS’s application, arguing that SPS could only recover its costs from large customers. The Attorney General argues that recovery of large customer cap costs from non-capped customers is contrary to both Section 62-16-4(A)(2) and 17.9.572.15 NMAC, a regulation concerning renewable energy cost recovery. The Attorney General contends that Section 62-16-4(A)(2) mandates the reduction of renewable energy procurements if such procurements would generate costs in excess of the large customer cap. Under the Attorney General’s reasoning, because large customer cap costs should not have arisen as a matter of law, they cannot be allocated to non-capped customers.
{35} In a supplemental recommended decision, the hearing examiner recommended that SPS be allowed to recover large customer cap costs from non-capped customers because given the cost limits on large customers, SPS’s ability to collect excess costs from large customers in the future would be speculative and uncertain. In a final order partially adopting the recommended decision (the final order), the PRC agreed with the hearing examiner. We affirm the PRC on this issue because its actions are consistent with Section 62-16-4(A)(2) and the Renewable Energy Act as a whole.
B. Discretion to Reduce Renewable Energy Procurements
{36} Section 62-16-4(A)(2) states that the New Mexico renewable portfolio standards mandate “shall be reduced, as necessary” to accommodate the large customer cap. According to the Attorney General, the word “shall” indicates that renewable energy procurement reductions are mandated whenever renewable energy procurement costs would otherwise exceed the large customer cap. One opposing interpretation of Section 62-16-4(A)(2) is that the phrase “as necessary” modifies the phrase “shall be reduced” to indicate that the PRC has discretion to reduce renewable energy procurements, even if large customer cap costs would result from such procurements. The Attorney General argues that when large customer cap costs arise, the PRC has discretion regarding the amount by which renewable energy procurement should be reduced, but not whether the renewable portfolio standards should be reduced. We hold that (1) Section 62-16-4(A)(2) does not mandate’ a reduction in renewable energy procurement whenever large customer cap costs arise, and (2) the PRC has discretion to reduce renewable energy procurement when large customer cap costs arise.
{37} Our analysis begins with the plain text of the statute. Garcia v. Gutierrez, *7262009-NMSC-044, ¶ 53, 147 N.M. 105, 217 P.3d 591. In analyzing the phrase “shall be reduced, as necessary,” we knowledge that “shall” is a word of mandate. See Black’s Law Dictionary 1375 (6th ed. 1990). However, the phrase “as necessary” indicates discretion. Norris v. Emanuel Cty., 561 S.E.2d 240, 244 (Ga. Ct. App. 2002). Because “as necessary” modifies the word “shall” in Section 62-16-4(A)(2) (internal quotation marks omitted), the statute’s plain text indicates that when large customer cap costs arise, the PRC has discretion to determine whether renewable energy procurement reductions are necessary.
{38} The next sentence in Section 62-16-4(A)(2) provides that “the kilowatt-hours of renewable energy procured for these customers shall be limited so that the additional cost of the renewable portfolio standard to each customer does not exceed” the large customer cap. The word “shall” in this sentence is not modified by any words of discretion. The Attorney General apparently relies upon this lack of discretionary language to argue that Section 62-16-4(A)(2) mandates renewable energy procurement reductions whenever large customer cap costs arise. We disagree. This sentence merely precludes capped customers from being charged costs in excess of the statutory cap. Logically, should large customer cap costs arise, the PRC can ensure compliance with the statutory cap in two ways: the PRC can either reduce renewable energy procurement or adjust what is actually charged to capped customers. In adjusting what is actually charged to capped customers, the PRC “may authorize deferred recovery of the costs of complying with the renewable portfolio standard.” Section 62-16-4(A)(2). In other words, the PRC can choose not to reduce procurements, even when large customer cap costs arise, by deferring the excess costs for later recovery, so as not to charge capped customers with statutorily prohibited costs. See id. {39} A plain reading of Section 62-16-4(A)(2) indicates that the PRC has the authority not to reduce renewable energy procurements, even when large customer cap costs increase. This interpretation is strongly supported by a reading of the Renewable Energy Act in its entirety, and it should therefore be adopted. See Arnold v. State, 1980-NMSC-030, ¶ 10, 94 N.M. 381, 610 P.2d 1210 (“Legislative intent is to be determined primarily from the language used in the Act or statute as a whole.”). Moreover, this reading acknowledges the difficulty of avoiding large customer cap costs.
{40} First, the renewable portfolio standard promulgated by the Renewable Energy Act provides a minimum standard. See § 62-16-4(A)(l) (“[Rjenewable energy shall comprise no less than [a given] percent of each public utility’s total retail sales to New Mexico customers.” (emphasis added)). The Attorney General’s reading of Section 62-16-4(A)(2) would have us treat the large customer cap as providing a maximum standard. This is problematic because, mandating renewable energy procurement reductions whenever large customer cap costs arise would be inconsistent with Section 62-16-2(A)(5), which plainly states that “a public utility should have incentives to go beyond the minimum requirements of the renewable portfolio standard.”
{41} Second, Section 62-16-4(A) clearly evinces a legislative intent to systematically increase renewable energy use in New Mexico. Section 62-16-4(A)(l) escalates renewable energy procurement requirements over time, while Section 62-16-4(A)(2) increases the large customer cap over time. Mandating renewable energy procurement reductions whenever large customer cap costs arise would undermine New Mexico’s ability to systematically increase renewable energy usage.
*727{42} Third, the Attorney General’s argument is erroneously premised on the idea that Section 62-16-4(A)(2) was meant to protect non-capped customers from high renewable energy costs by banning costs in excess of the large customer cap to prevent large customer cap costs from being allocated to non-capped customers. We need not adopt the Attorney General’s interpretation of Section 62-16-4(A)(2) to protect non-capped customers from high renewable energy costs because another statutory provision already performs this function: Section 62-16-4(B) mandates setting an overall reasonable cost threshold for renewable energy procurement.
{43} Fourth, the Attorney General’s argument appears to assume that large customer cap costs can be forecast on an accurate and consistent basis so that in approving renewab le energy procurements, the PRC can systematically avoid large customer cap costs. The record proper indicates otherwise. PRC approvals of renewable energy procurement are “based on the best information available at the time the resources were being reviewed.” SPS notes that how much large customer cap costs will increase depends on future occurrences such as the fluctuation of natural gas prices. Accordingly, the evidence indicates that we cannot reasonably expect that large customer cap costs can be predictably eliminated.
{44} Reading the language of the Renewable Energy Act as a whole, we conclude that the PRC has discretion to decline to reduce renewable energy procurement, even when large customer cap costs arise. This authority is congruent with the statutory policy of increasing renewable energy use in New Mexico. Moreover, we cannot reasonably expect that either the PRC or utilities will be able to avoid large customer cap costs. Thus, adopting the Attorney General’s position that large customer cap costs have to be avoided as a matter of law also would be contrary to practical experience.
C. Section 62-16-4(A)(2) Does Not Preclude the Recovery of Large Customer Cap Costs from Non-Capped Customers
{45} The PRC exercised its discretion not to reduce renewable energy procurementwhen large customer cap costs arose, which is consistent with our interpretation of Section 62-16-4(A)(2). It then authorized the deferral of large customer cap costs for future recovery. The PRC’s final order provides for the collection of large customer cap costs from non-capped customers. We therefore determine the permissibility of collecting large customer cap costs from non-capped customers.
{46} The Attorney General does not oppose SPS’s ability to recover deferred large customer cap costs. It merely contends that such costs should not be recovered from non-capped customers, asserting that (1) enabling recovery of large customer cap costs from non-capped customers “violatefs] the basic ratemaking principle of cost[ ] causation,” and (2) Section 62-16-4(A)(2) protects non-capped customers from paying for large customer cap costs. We reject the Attorney General’s position as contrary to law and hold that large customer cap costs can be allocated to non-capped customers.
{47} The Attorney General’s contention that allocating large customer cap costs to non-capped customers violates the principle of cost causation is without merit. The plain language of Section 62-16-4(A)(2) does not mandate that renewable energy procurement costs be recovered only against those customers who caused large customer cap costs. Moreover, renewable energy procurement costs arise as a result of statutory mandate, see § 62-16-4(A)(l), such that *728neither capped nor non-capped customers can be said to cause any specific procurement costs.
{48} Similarly, the Attorney General’s assertion that Section 62-16-4(A)(2) protects non-capped customers from large customer cap costs is unsupported by law. Its argument assumes that Section 62-16-4(A)(2) mandates reductions in renewable energy procurement whenever large customer cap costs arise to protect both capped and non-capped customers from high renewable energy costs. Under the Attorney General’s reasoning, rates for non-capped customers cannot be increased by large customer cap costs because such increases would deprive non-capped customers of the protections provided by Section 62-16-4(A)(2). We reject this reasoning because Section 62-16-4(A)(2) does not evince a legislative intent to protect non-capped customers. We have already held that Section 62-16-4(A)(2) does not mandate renewable energy procurement reductions when large customer cap costs arise. The Attorney General’s contention that Section 62-16-4(A)(2) precludes large customer cap costs to protect non-capped customers is also incorrect because nothing in Section 62-16-4(A)(2) addresses the interests of non-capped customers. See State v. Diamond, 1921-NMSC-099, ¶ 5, 27 N.M. 477, 202 P. 988 (We will not insert words that are absent in a statute.). We conclude that Section 62-16-4(A)(2) does not preclude the allocation of large customer cap costs to non-capped customers.
{49} The PRC had previously approved of SPS’s procurement plans. Under Section 62-16-6(A), “[cjosts that are consistent with commission approval of procurement plans . . . shall be deemed to be reasonable.” Thus, the renewable procurement costs in this case are reasonable as a matter of law. Because these procurement costs are reasonable, SPS is entitled under Section 62-16-6(A) to recover large customer cap costs. Id. (“A public utility that procures or generates renewable energy shall recover, through the rate-making process, the reasonable costs of complying with the renewable portfolio standard.”). The evidence indicates that if large customer cap costs only can be collected from capped customers, 20 years or more could elapse “before SPS even has the opportunity to collect” these procurement costs. Thus, as the PRC determined, the Attorney General’s proposed cost recovery mechanism “is speculative and uncertain, and would not provide a reasonable opportunity for SPS to recover [large customer cap] costs.” Forcing SPS to collect large customer cap costs only from capped customers would effectively disallow recovery of these procurement costs, contrary to the Renewable Energy Act’s guarantee that utilities can recover the reasonable costs of renewable energy procurement. Sections 62-16-4(A)(2) & -6(A).5
D. 17.9.572.15 NMAC Does Not Preclude the Recovery of Large Customer Cap Costs from Non-Capped Customers
{50} The Attorney General contends that the “plain language” of 17.9.572.15 NMAC “make[s] clear that costs associated with large [capped] customers should be borne by large customers alone.” 17.9.572.15 NMAC is a regulatory provision concerning renewable energy cost recovery that references Section 62-16-4(A)(2). Our interpretation of Section 6 2-16-4(A)(2) therefore informs our *729constructionofl7.9.572.15NMAC. Wehave previously held in this opinion that Section 62-16-4(A)(2) provides the PRC with discretion, not a mandate, to reduce renewable energy procurement when large customer cap costs arise, and does not bar the allocation of large customer cap costs to non-capped customers. Consistent with our interpretation of Section 62-16-4(A)(2), we hold that 17.9.572.15 NMAC also does not bar the allocation of large customer cap costs to non-capped customers.
{51} 17.9.572.15 NMAC provides that:
A. A public utility shall recover the reasonable costs of complying with this rule through the rate making process, including its reasonable interconnection and transmission costs and other costs attributable to acquisition and delivery of renewable energy to retail New Mexico customers.
B. Costs that are consistent with commission-approved annual Renewable Energy Act plans are deemed to be reasonable.
C. A public utility that is permitted to defer the recovery of renewable energy costs pursuant to commission order may, through the ratemaking process, recover from customers that are not subject to the rate impact limitations of Sections 62-16-4A(2) and 62-16-4A(3) NMSA 1978 the cumulative sum of those deferred amounts, plus a carrying charge on those amounts.
D. For customers that are subject to the rate impact limitations of Section 62-16-4A(2) NMSA 1978, a public utility may, through the ratemaking process, recover from those customers the cumulative sum of those S ection 62-16-4 A(2) NMSA 1978 limited deferred amounts, plus carrying charges on those amounts.
E.Any renewable energy procurement costs recovered through the utility’s fuel clause shall be separately identified in its monthly and annual fuel and purchased power clause adjustment filings and its continuation filings.
{52} The Attorney General’s argument relies on 17.9.572.15(D)NMAC to support its contention that large customer cap costs can only be recovered from large customers. The Attorney General seems to share our understanding that large customer cap costs, which arise pursuant to Section 62-16-4A(2), may be deferred. Based on this understanding, the Attorney General assumes that the term “Section 62-16-4A(2) NMSA 1978 limited deferred amounts” in Subsection D is a synonym for deferred large customer cap costs. Armed with this assumption, the Attorney General contends that because Subsection D concerns recovery of costs from capped customers and only Subsection D expressly refers to recovery of “Section 62-16-4A(2) NMSA 1978 limited deferred amounts,” large customer cap costs can only be recovered from capped customers.
{53} We reject the Attorney General’s contention. 17.9.572.15(C) NMAC states that whenever “[a] public utility ... is permitted to defer the recovery of renewable energy costs pursuant to commission order,” the utility may recover the deferred amounts from non-capped customers. Subsection C therefore authorizes public utilities to recover deferred costs, in general, from non-capped customers. By contrast, 17.9.572.15(D) NMAC explicitly authorizes only the recovery of “Section 62-16-4A(2) NMSA 1978 limited deferred amounts” from capped customers. Thus, *730although 17.9.572.15(D) NMAC arguably provides that only “Section 62-16-4A(2) NMSA 1978 limited deferred amounts” may be recovered from capped customers, 17.9.572.15(C) NMAC provides that any deferred amounts may be recovered from non-capped customers.
{54} We conclude that a plain reading of 17.9.572.15 NMAC indicates that deferred costs arising from Section 62-16-4A(2) can be recovered from both capped and non-capped customers. There is simply no language explicitly banning the collection of deferred large customer cap costs from non-capped customers. The Attorney General errs in conflating the issue of whether capped customers may be charged only for Section 62-16-4A(2) deferred amounts with whether only capped customers may be charged the aforesaid deferred amounts.
III. CONCLUSION
{55} We affirm the PRC’s final order. We will not disturb the PRC’s finding that SPS’s entire prepaid pension asset was properly included in the rate base. We also hold that the PRC properly allocated large customer cap costs to non-capped customers to enable SPS to recover its reasonable renewable energy procurement costs.
{56} IT IS SO ORDERED.
EDWARD L. CHÁVEZ, Justice
WE CONCUR:
BARBARA J. VIGIL, Chief Justice
PETRA JIMENEZ MAES, Justice
CHARLES W. DANIELS, Justice
TIMOTHY L. GARCIA, Judge Sitting by designation