Depreciation Shell Game


Accounting reforms might force regulators to abandon their live-now, pay-later practices.

Fortnightly Magazine - April 2008

When an advisory committee of the Securities and Exchange Commission (SEC) voted recently to phase out special accounting treatment for various industries, it signaled the end may be near for power plant depreciation deferral mechanisms. Such mechanisms are a mainstay of regulatory accounting in many states, and their discontinuation could send plant owners and regulators back to the drawing board to find a new, GAAP-compliant way to recognize asset depreciation in financial reports.

Specifically, the Advisory Committee on Improvements to Financial Reporting issued a report on Valentine’s Day 2008 recommending the Financial Accounting Standards Board (FASB) avoid special treatment for various industries in its accounting rules on transactions and activities.1 The committee identified 18 industries, including “regulated operations,” as having industry-specific guidance—i.e., exceptions to generally accepted accounting principles (GAAP) —that can create problematic complexities and inconsistencies in financial reports.

For investor-owned utilities and other public companies that own power plants, an important example of such guidance is SFAS 71, Accounting for the Effects of Certain Types of Regulation (see sidebar, “GAAP and Depreciation Deferral”). SFAS 71 allows utilities to use special regulatory accounting for their income statements, with the differences from GAAP disclosed on the balance sheet as regulatory assets and liabilities.

Depreciation deferral mechanisms, imposed by regulators in many states, conflict with GAAP in the way they determine a power plant’s life span, as well as the way they deal with upgrades, operational changes and ultimately end-of-life demolition and removal costs.

Rescinding SFAS 71 and related accounting standards specific to regulated entities would, in most cases, preclude regulatory depreciation from being reflected in financial statements, except to the extent reflected in recorded revenues. While this trend will create short-term confusion for industry CFOs and regulators, it’s a good thing for the industry—especially its shareholders and ratepayers—because these mechanisms too often lead to misleading regulatory accounting, and can increase ratepayer costs over the life of the affected assets.

Mind the GAAP

Some depreciation deferral mechanisms apply to the investment portion of depreciation, and some apply to the salvage and cost of removal portion. Mechanisms related to end-of-life issues pose particularly significant uncertainties for utility regulation and financial reporting.

The reliability of any electric generating unit deteriorates with age, and steam and nuclear units typically deteriorate sufficiently by an age of about 30 years to require remedial actions that influence certain of these deferral mechanisms. The available actions include: retirement; changing the mode of operation to peak load or standby service; and reconditioning through refurbishment or repowering by boiler replacement.

Prior to 10 or 15 years ago, the most common remedial action for steam units was to change the mode of operation, and today the most common action is to recondition the unit. Initially, the most common action for nuclear units was retirement—usually well prior to the 40 years allowed by their operating licenses—and today the most common action is to recondition through operating license renewal.

Regulation frequently responds to these actions in ways that defer recording and recovery of depreciation costs, generally to minimize the immediate effect on electricity tariffs. Deferral reduces near-term revenue requirements, which makes rate-case decisions appear more palatable to electricity customers. Unfortunately, the regulatory process tends to obscure the total revenue requirements resulting from depreciation deferral over the life of the assets. In addition to increased life-span costs, depreciation deferral—and its potential eventually to become denial—creates risks that increase the utility’s cost of capital, further increasing costs imposed on ratepayers.

The accounting implication is a consequence of deferral mechanisms causing regulatory accounting to be in conflict with the financial accounting dictated by GAAP. The cost implication is a consequence of the way depreciation affects the utility’s balance sheet. Annual depreciation expenses are a positive component of revenue requirements, and annual expenses accumulated in the book reserve are a negative component. The typically long life of energy and water utility assets causes the negative component to eventually overwhelm the positive component, which means the apparently beneficial initial ratepayer impact of any depreciation change will reverse itself within a few years.

Retirement Planning

The determination of depreciation rates is essentially an effort to predict the future, commonly by analyzing past experience. But such analyses for power plants are unlikely to provide a reasonable indication of the future, unless the company has retired at least one station with unit life spans similar to those expected for the remaining units. Nevertheless, power plant depreciation rates frequently are based on predicted retirement dates.

Forty years ago, the typical generation planning horizon was about five years, which required utility planners to conduct special studies to predict the retirement dates needed for depreciation purposes. This horizon since has expanded substantially and many planners now incorporate retirement date predictions in their regular activities, thereby eliminating the need for special studies for depreciation purposes. Even when special studies were needed, station operators commonly predicted retirement dates to determine whether large capital expenditures or maintenance projects are justified. Whether predicted by planners or by operators, the resulting retirement dates provide a suitable basis for determining depreciation rates, but are sometimes rejected by regulators and replaced with longer lives based on industry data not demonstrated as being comparable—initially due to a lack of station retirement experience and more recently due to life spans being extended through reconditioning.

For example, Missouri regulators2 don’t accept depreciation rates based on predicted retirement dates for steam plants, but they do for nuclear units. Consequently, steam-plant depreciation rates are based on analyses of past experience. Lacking sufficient experience with station retirements leads to rates that presume unit life spans that are unrealistically long. For example, one Missouri case authorizes depreciation rates for turbogenerator units that assume life spans of 120 and 200 years.3

While it’s not unusual to encounter steam plant equipment more than 100 years old in museums, it is unusual to encounter 100-year old equipment in operation. Such unrealistically long life spans resulting from regulatory approaches to depreciation rates are inconsistent with GAAP, and therefore require the company to apply the provisions of SFAS 71.

In addition, unrealistic life spans may impose a need to test for asset impairment. The life spans of nuclear units are imposed by the termination dates of Nuclear Regulatory Commission (NRC) operating licenses, rather than estimated by planners or operators, and such dates typically serve as the basis for their regulatory depreciation. NRC licenses initially were issued for a term of 40 years that began at the issue date of the construction permit. This resulted in operating life spans shorter than 40 years for units that experienced delays in construction and operation, which prompted the NRC to change the 40-year term to start at the issue date of the operating license and to modify existing licenses accordingly. The NRC later began allowing 20-year extensions to operating licenses of existing units through license renewals, upon request and demonstration of adequate equipment condition.

EES North America

Whether operating licenses are renewed or not, actual operating life spans are unlikely to be as long as is allowed, because of the increasing difficulty in justifying large capital and maintenance expenditures as the license termination date approaches. No nuclear unit has yet to reach the life span allowed by its original operating license, but some are getting close.

Missouri requires nuclear depreciation rates to presume operation until the termination dates of operating licenses, and in some cases to presume a renewed license that does not exist. However, the actual operating license and the likelihood of early retirement would need to be recognized for GAAP depreciation purposes. As a result, the depreciation rates for financial accounting purposes are higher than they are for regulatory accounting, again bringing SFAS 71 into play. The potential for not receiving a license renewal and for early retirement would impose a need to periodically test for asset impairment.

Makeover Accounting

All types of generating units experience component replacements and the addition of new components during their lifetimes, as part of routine maintenance and repairs. The depreciation terminology for this activity is “interim additions and retirements.” Average service life applies to the depreciable investment, which interim additions and retirements cause to be considerably shorter than the life span of the unit. For example, Ameren UE testimony in a 2007 case shows the average service life experienced by two retired steam units was about half of their life span. 4

It’s quite common for power plant depreciation rates based on estimated generating unit retirement dates to recognize estimated future interim retirements, but to exclude recognition of future interim additions—most of which are for replacement of the estimated interim retirements—until after they have occurred. For example, Missouri does this for nuclear units. Interim additions have more influence on depreciation rates than do interim retirements, because their magnitude is typically five to 10 times the magnitude of interim retirements.

Deferring the recognition of interim additions causes the depreciation rate to increase at each recalculation, if the life span does not change. This conflicts with the typically decreasing pattern of usage for steam units and constant pattern for nuclear units, so it brings SFAS 71 into play. This increasing pattern of rates qualifies as a phase-in plan under SFAS 92. However, SFAS 92 gives Missouri a way out, because it requires the phase-in plan to be more deferred than under the ratemaking policy in place prior to 1982, and Missouri’s plan was in place prior to 1982.

Altered Operation Mode

GAAP and the FERC Electric Uniform System of Accounts specify that depreciation is to be systematic and rational, with rational meaning depreciation be recorded in a way that’s consistent with the usage of the asset. For regulated entities, such as electric utilities, which are required to practice the group concept of depreciation, this requirement is accomplished by the pattern of depreciation rates being the same as the pattern of asset usage.

Reconditioning to extend life involves substantial expenditures, and regulators commonly dictate that the extended life be recognized in depreciation rates immediately, and that the related expenditures not be recognized until after they have been recorded. This is the same treatment of interim additions addressed above and has the same implications for financial reporting.

Changing to a peak load or standby mode of operation late in life results in little or no usage from then on, causing a rather distinctive lifetime pattern of usage. This pattern is what the units-of-production (UOP) depreciation procedure was devised to address, but UOP rarely is applied to power plants. While UOP is valid for this situation, there are several ways to determine depreciation rates based on life defined by time (rather than by production) that will emulate this usage pattern. Regulators, however, do not allow their use. Texas has gone so far as to interpret its Substantive Rules for regulatory depreciation as allowing only life defined by time. However, the Texas restructuring included generation, so power plant owners in Texas now can use UOP depreciation methods.

A pattern of depreciation rates different from the pattern of usage is inconsistent with GAAP, so it would trigger SFAS 71 provisions.

Terminal Demolition Costs

To satisfy laws, regulations and public safety considerations, regulated entities remove utility assets or safely abandon them in place at the end of their useful lives. Steam stations eventually are removed, and their removal expenditures substantially exceed their salvage proceeds. Missouri does not allow estimated terminal net removal costs to be recognized in depreciation for steam plants, but it does for nuclear plants through decommissioning fund contributions and for other types of energy utility assets through depreciation.

The Uniform Systems of Accounts that jurisdictional entities must comply with specify that removal costs must be a component of depreciation, but the FASB and SEC preclude removal costs from being recorded as depreciation for financial reporting purposes. This FASB and SEC exclusion is the result of misinterpreting the GAAP definition of depreciation accounting, the reasons for which are beyond the scope of this discussion. However, as long as FASB and SEC hold this position, depreciation rates excluding removal costs can be utilized for income statements without recording a regulatory asset or liability.

Unlike Missouri, most states allow terminal demolition costs to be reflected in the depreciation rates for non-nuclear generating units. (Such costs are not reflected in depreciation rates for nuclear units, because the NRC requires that decommissioning costs be funded externally.) The more common deferral mechanism for the demolition costs of non-nuclear power plants is to require recognition of site-specific cost estimates at the current price level, rather than at the price level expected at the time of demolition. However, regulators have used other mechanisms, such as placing a cap on the extent of demolition costs allowed to be reflected in power plant depreciation rates, and amortizing demolition costs over some specified period of time, after the expenses have been recorded in the book reserve.

Rational Depreciation

Missouri has not restructured its utility industry, so it makes a convenient example for analysis. But Missouri is by no means unique. Many states impose these deferral mechanisms, which is unfortunate because they lead to financial reporting that doesn’t accurately disclose the results of operations and the current financial position—and ultimately create higher costs for ratepayers. Energy cost increases imposed by regulatory practices are particularly interesting, because they make it more difficult for those involved with area development to attract new businesses and to expand existing businesses. 5

Michigan provides an interesting illustration. Early last year the state’s Public Service Commission sent a report to Gov. Jennifer Granholm, asserting the “current market structure does not provide sufficient certainty to finance a major generating plant on reasonable terms.” The report makes some recommendations for actions needed to address this situation. 6

Adequate capital recovery is an attribute of a regulatory environment consistent with addressing this situation. Michigan’s restructuring has not deregulated power plants, so the state continues to practice some of the power plant deferral mechanisms that increase risks and costs. Conspicuous by its absence from the commission’s recommended actions is the abandonment of its power plant depreciation deferral mechanisms.

Eliminating power plant regulatory depreciation deferral mechanisms—as might happen if FASB follows the SEC committee’s recommendations—would benefit ratepayers and the economic viability of utility service territories. It would reduce power costs in the long-term, and also would benefit users of financial statements by more accurately disclosing the results of utility operations and their current financial positions.



1. Progress Report of the Advisory Committee on Improvements to Financial Reporting to the U.S. Securities and Exchange Commission, Feb. 14, 2008.

2. The author’s letter to the editor in the December 2007 issue of Fortnightly asserts certain power-plant depreciation issues merit attention by the Missouri Public Service Commission. This article identifies the Missouri practices mentioned in that letter.

3 . Missouri Case No. ER-2004-0570 (Empire District Electric Co.)

4 . Missouri Case No. ER-2007-0002.

5 . Large energy users consider the quality of regulation when evaluating new sites or expansion of existing sites. The author has assisted some to do so.

6. Michigan’s 21st Century Electric Energy Plan, Michigan Public Service Commission, January 2007.