Widespread concern over nuclear plant decommissioning has triggered similar interest in the decommissioning of fossil-fired steam generating stations. This rising interest stems in part from the emergence of a competitive market in electric generation, which, among other things, threatens impairment of assets.
Fossil decommissioning issues are not nearly as contentious as those that attend nuclear plants. Nevertheless, the magnitude of cost estimates for fossil decommissioning, when expressed as a percentage of station investment, is high enough to demand attention from accountants and regulators.
Past Experience
Over the 30 years during which I have been involved with setting fossil station depreciation rates, I have seen a steady progression in the need for and attention paid to plant decommissioning. Where fossil stations once carried a zero decommissioning cost (em with the contractor keeping the salvage (em fossil decommissioning today incurs significant costs. This change derives partly from strict rules for handling hazardous materials, and partly from the design of the stations involved.
Past experience is unlikely to provide a meaningful indication of decommissioning costs for most existing fossil stations. Most prior experience has largely come from removal of units with self-supporting boilers. Modern stations, however, come equipped with very large top-hung boilers that require more extensive (and costly) removal procedures. Yes, modern stations will employ less insulation containing asbestos, which will reduce decommissioning costs. But salvage value will be limited for modern stations, because large top-hung boilers will hold only scrap value and the other equipment, being quite old, is unlikely to have any reuse value.
In particular, I am familiar with two specific and unusual instances of recent decommissioning at modern stations. Both cases involved quite young turbogenerator units with high reuse values. The salvage for one greatly exceeded the demolition and site restoration costs, because part of the deal to buy the turbogenerator was that the purchaser demolish the station at his expense. For the other, the reused turbogenerator produced about 5 percent salvage and the actual demolition cost was about 40 percent, which did not include complete site restoration.
Unfortunately, this experience is not indicative of what can be expected for most modern fossil stations.
Magnitude of Costs
Although high, estimates of fossil plant decommissioning costs will be influenced by site conditions, station design, estimate assumptions, and the experience of the estimator. Cost estimates that are a matter of public record show quite a range of costs and cost components, but do not disclose whether the variations derive from one of these factors specifically or from a combination thereof. Station design is particularly important, however, because it is generally recognized that modern top-hung boilers are more expensive to remove than self-supporting boilers. Just how much more expensive depends on the assumed removal process, and opinions differ on whether certain processes can be implemented safely.
One can express decommissioning costs in at least two ways: 1) unit costs, or 2) net salvage factor (that is, salvage less cost of removal, expressed as a percentage of depreciable investment). I collect site-specific fossil station decommissioning cost estimates that are a matter of public record, and periodically summarize them in terms of unit costs and net salvage factors:
Gas & Coal
Oil Units Units
Unit Cost:
1993 Cost Level $30/Kw $ 40/Kw
Removal Date Cost Level $62/Kw $174/Kw
Net Salvage Factor (37)% (49)%
While I find net salvage factors less meaningful than unit costs, the two measures do tend to correlate for the steam units of several of my clients. Thus, I find it reasonable to assume fossil station decommissioning cost levels of about 30 to 50 percent of the current depreciable investment. However, these costs are somewhat conservative, because escalation is applied to salvage amounts and removal is assumed to occur in the year in which the last unit at each station is retired.
Charts 1 and 2 show data derived from my site-specific estimate collection. Chart 1 covers coal units and Chart 2 gas and oil units; both show cost estimates expressed as amounts per Kw of capacity in relation to the average size of the generating unit or units. Charts 1A and 2A show amounts at the 1993 price level and Charts lB and 2B show amounts at the price level at the estimated date of retirement. The correlation to size for the 1993 amounts gets a bit fuzzy when cost escalation is introduced, especially for coal units.
The estimates on Charts 1 and 2 vary widely from station to station, as do the net salvage factors on Chart 3. Several reasons undoubtedly explain these variations. For instance, about a dozen different estimators are involved. And estimators are understandably reluctant to share their secrets, especially demolition contractors.
Regulatory Cost Deferral
Regulators generally do not like the negative net salvage produced by decommissioning, because it leads to higher depreciation rates and, in the short term, to higher tariffs. Therefore, they commonly defer recognition of decommissioning costs in utility depreciation rates. This task is easier at fossil stations because not much historical experience exists to be explained away.
Query: Can we determine whether regulators are in fact deferring fossil plant decommissioning costs?
If one excludes Florida and Pennsylvania (these two states do not include fossil station decommissioning costs in depreciation rates), industry data shows that state public utility commissions (PUCs) generally are approving decommissioning cost (negative net salvages) factors for fossil-fired steam generating plants in a range of 5 to 10 percent. That range clearly demonstrates the degree of deferral that exists, since, as noted earlier, fossil station decommissioning cost estimates range between 30 to 50 percent of the current depreciable investment.
The most common deferral practice sets depreciation rates on the basis of decommissioning costs at the current price level, which causes the depreciation rates to increase when the utility periodically recalculates future decommissioning costs as generating units grow older. But under my interpretation of the AICPA definition of depreciation accounting, these increasing depreciation rates will not comply with GAAP. I believe that GAAP will permit such increases in depreciation rates only if the usage of the underlying assets also increases. However, usage of fossil-fired generating units will likely stay relatively constant or decrease over time. This conclusion is based on my interpretation of the word "salvage," in the AICPA definition, to mean net salvage. Some contend otherwise (that "salvage" means "salvage"), thereby allowing cost of removal to be excluded from depreciation and recorded as a liability on a discounted basis.
State PUCs might also defer fossil decommissioning costs by 1) sinking fund depreciation, 2) purposely underestimating costs or future cost escalation, or 3) using a cash basis instead of accrual. Other deferral practices include:
Missouri. No recovery for fossil stations.
Pennsylvania. Recovery over 5 years following expenditure (all facilities but nuclear stations).
Texas. No current recovery of fossil plant net removal costs above 5 percent of current depreciable investment.
The Coming Competition
Deferrals of decommissioning costs for fossil-fired steam generating plants carry important implications for electric utilities during the transition from a regulated to a competitive environment.
Power-supply activities in a competitive environment may involve any one or all of a number of corrective actions, including 1) a writedown of "impaired assets," 2) a write-off of "regulatory assets" (paper assets representing regulatory promises of future cost recovery), or 3) increases in the rate of depreciation made mandatory under a switch from regulatory to nonregulatory GAAP accounting to compensate for past depreciation deferral and to preclude future deferrals.
Collectively, these corrective actions fall under the rubric of stranded investment. An active debate exists about whether investors or customers should bear costs associated with stranded investment. Since each of these possible actions arises from regulatory promises of future cost recovery, logic suggests that customers should pay. Nevertheless, this conclusion may not hold for impaired assets or depreciation rate increases if regulators have never before been apprised of the extent of the decommissioning obligations.
Impaired Assets. In March 1995, the Financial Accounting Standards Board (FASB) published new rules in Financial Accounting Standard (FAS) 121. Under those rules, asset impairment occurs when undiscounted future cash flows from asset use and eventual disposition fall below the asset carrying amount (plant in service less future net salvage and depreciation reserve).
In such a case, decommissioning costs incurred at the end of plant life will create significant differences between discounted and undiscounted cash flows. Thus, under the FASB definition, the deferred depreciation methods in common use for fossil-fired plants will increase both the chance and the magnitude of asset impairment.
Switching to GAAP. If competitive electric prices emerge, set by the market rather than by cost regulation, then electric utility accounting must move out from under FAS 71 (Accounting for the Effects of Certain Types of Regulation), and switch to GAAP, as directed by FAS 101 (Regulated Enterprises (em Accounting for the Discontinuation of FAS 71).
In the switch to GAAP, the depreciation rate increases that would be needed to recover from past depreciation deferrals can be determined from the periodic studies typically conducted to test the continued validity of depreciation rates. If my sample is any indication, the process of eliminating past deferrals and precluding future deferrals would force up fossil station depreciation rates by about 50 percent.
Remedies. Many electric utilities are trying to limit their exposure to a competitive marketplace by reorganizing to segregate generation from their other operations. I consider this strategy wise, since it appears that the past depreciation deferrals for transmission and distribution facilities are larger relative to recorded asset values than are the deferrals for power plants. Nevertheless, reduced exposure to competition may not affect asset impairment, because FAS 121 applies to both regulated and nonregulated businesses.
Increasing depreciation rates to eliminate past deferrals and preclude future deferrals may not be helpful for operating in a competitive environment, because depreciation expenses will rise. Future depreciation expenses can be decreased only if the other side of the depreciation equation is also addressed, such as by writing down asset values (em in other words, recognizing asset impairment.
Moreover, a competitive environment may effectively reduce the useful depreciation lives of utility assets. This life influence is not as easy to identify as the influence of deferral of decommissioning obligations, because decommissioning costs expressed as a ratio of depreciable values are quite sensitive to the age of the assets. Therefore, decreases in useful life that, by themselves, would cause depreciation rate increases will also bring about decommissioning cost changes that, by themselves, would cause rate decreases. The net effect of these opposing forces will depend on their relative rate of change. At some rate of change these two forces will exactly offset each other. Until proven otherwise by specific circumstances, an assumption of exact offset may be reasonable under depreciation accounting. However, this offset assumption will not be reasonable under liability accounting, because net salvage would no longer form a component of depreciation. Further, a shortened life means that revenues would cease, thereby affecting future cash flow and enhancing the potential for asset impairment.
Clearly, fossil station decommissioning raises significant accounting and regulatory issues. These issues are not new, but the rapid movement of power supply into a competitive environment makes them urgent. t
John S. Ferguson will retire this month as a principal of Deloitte & Touche LLP. He is a frequent contributor to PUBLIC UTILITIES FORTNIGHTLY.
Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.