Trailblazer Pipeline Company LLC Docket No. RP Exhibit No. TPC-0093

Transcription

1 Trailblazer Pipeline Company LLC Docket No. RP- -000

2 UNITED STATES OF AMERICA BEFORE THE FEDERAL ENERGY REGULATORY COMMISSION Trailblazer Pipeline Company LLC ) ) ) Docket No. RP SUMMARY OF PREPARED DIRECT TESTIMONY OF PATRICK R. CROWLEY ON BEHALF OF TRAILBLAZER PIPELINE COMPANY LLC Patrick R. Crowley is employed by Brown, Williams, Moorhead & Quinn, Inc., a nationally recognized energy consulting firm, as a depreciation expert in regulated oil and natural gas pipelines. He provides Prepared Direct Testimony in this proceeding on behalf of Trailblazer Pipeline Company LLC ( Trailblazer ) regarding the proper and adequate depreciation rates and negative salvage rates for Trailblazer s facilities based on reasonable remaining life estimates and a broad group straight-line average remaining life depreciation methodology. Witness Crowley explains the concepts behind depreciation analysis and survivor curve theory, both of which establish the average service lives of the assets of the utility, the retirement decline curve, and the interim retirements that set the average remaining life for each account. Witness Crowley incorporates the economic lifespan recommended by Witness Alexander J. Kirk, and the terminal retirement cost study performed by Witness Steven R. Fall.

3 UNITED STATES OF AMERICA BEFORE THE FEDERAL ENERGY REGULATORY COMMISSION Docket No. RP Page of Trailblazer Pipeline Company LLC ) ) ) Docket No. RP PREPARED DIRECT TESTIMONY OF PATRICK R. CROWLEY ON BEHALF OF TRAILBLAZER PIPELINE COMPANY LLC June, 0

4 Page of TABLE OF CONTENTS I. INTRODUCTION... II. DEPRECIATION THEORY... III. SURVIVOR CURVE THEORY... IV. SIMULATED PLANT RECORD ANALYSIS... V. REMAINING LIVES AND DEPRECIATION RATES... A. Transmission Plant... B. General Plant... VI. NEGATIVE SALVAGE RATE...

5 Page of GLOSSARY OF TERMS Acc. Res. Depr. Acc. Res. Neg. Salv. ARL ASL BWMQ CI C.F.R. Commission Depr. FERC L Curve Neg. Salv. NGA NS R Curve S Curve SPR Trailblazer Accumulated Reserve for depreciation Accumulated Reserve for negative salvage Average Remaining Life Average Service Life Brown Williams Moorhead & Quinn, Inc. Conformance Index Code of Federal Regulations Federal Energy Regulatory Commission Depreciation Federal Energy Regulatory Commission Left Modal Iowa Survivor Curve Negative Salvage Natural Gas Act Negative Salvage Right Modal Iowa Survivor Curve Symmetrical Modal Iowa Survivor Curve Simulated Plant Record theory Trailblazer Pipeline Company LLC

6 UNITED STATES OF AMERICA BEFORE THE FEDERAL ENERGY REGULATORY COMMISSION Docket No. RP Page of Trailblazer Pipeline Company LLC ) ) ) Docket No. RP PREPARED DIRECT TESTIMONY OF PATRICK R. CROWLEY ON BEHALF OF TRAILBLAZER PIPELINE COMPANY LLC 0 I. INTRODUCTION Q. Please state your name, occupation, and business address. A. My name is Patrick R. Crowley and my business address is th Street, NW, Suite 00 Washington, DC 000. I am Vice President of Brown Williams Moorhead & Quinn, Inc. ( BWMQ ), an energy consulting firm in Washington D.C. Q. Please briefly state your professional experience and qualifications. A. I graduated from DePaul University in Chicago, Illinois with a Bachelor of Arts degree in economics in and a Master of Arts degree in economics in, with a concentration in mathematical economics. Upon graduation from DePaul University in, I joined the Chicago, Rock Island & Pacific Railroad Company for a short time working in the general manager s office before I joined the Federal Energy Regulatory Commission ( FERC or Commission ) in. I was employed at FERC for years. For of those years, I was employed in the litigation division of the Office of Pipeline and Producer Regulation and its

7 Page of 0 successor offices. I retired from FERC to form my own consulting firm, Crowley Energy Consulting, in February 00, where I provided energy litigation support for clients in the natural gas and oil pipeline industries. I joined BWMQ in 0. My work as an Industry Economist in the Depreciation Branch of the Pipeline Rates Division of the Office of Pipeline and Producer Regulation at FERC was mainly as an expert witness with the Trial Staff gas and oil litigation team from to. I prepared pipeline depreciation studies, long-term forecasts of crude oil and natural gas reserves and production, mortality studies of plant retirements, and cost behavior studies for pipeline facilities. From through, I worked on the operational aspects of Order No.. From through, I worked on the advisory side of the Commission where I prepared reports for Commission orders regarding proposals for revised tariff terms, new services, rate designs, and tariff rates, as well as a wide variety of utility reports and cost studies. In, I returned to the litigation side of the Commission where I worked on electric utility, natural gas, and oil pipeline rate cases, complaint cases, and show cause orders until I retired. See generally Pipeline Service Obligations and Revisions to Regulations Governing Self-Implementing Transportation Under Part of the Commission s Regulations; Regulation of Natural Gas Pipelines After Partial Wellhead Decontrol, Order No., FERC Stats. & Regs., Regs. Preambles 0,, order on reh g, Order No. -A, FERC Stats. & Regs., Regs. Preambles 0,0, order on reh g, Order No. -B, FERC, (), aff d in part and remanded in part sub nom. United Distrib. Cos. v. FERC, F.d 0 (D.C. Cir. ), cert. denied, 0 U.S., order on remand, Order No. -C, FERC, (), reh g denied, Order No. -D, FERC,0 ().

8 Page of 0 Q. Have you previously testified before the Commission? A. Yes, I have. I filed testimony before FERC in numerous dockets, as reflected in Exhibit No. TPC-00. Q. On whose behalf are you submitting your prepared testimony in this proceeding? A. I am submitting testimony on behalf of Trailblazer Pipeline Company LLC ( Trailblazer ). Q. What is the purpose of your direct testimony? A. I am presenting my recommendation regarding the proper and adequate depreciation rates for Trailblazer based on appropriate remaining life factors applicable to the Trailblazer natural gas pipeline system. I am also 0 recommending Negative Salvage ( NS ) cost recovery rates for Trailblazer s transmission functions based on the terminal dismantling, removal, and restoration costs estimated by Trailblazer Witness Steven R. Fall in Exhibit No. TPC-00. Q. Have you provided any exhibits with your testimony? A. Yes. I have included the following exhibits with my testimony: Prepared Direct Testimony of Patrick R. Crowley Exhibit No. TPC-00 Curriculum Vitae of Patrick R. Crowley Exhibit No. TPC-00 Depreciation Workpapers Exhibit No. TPC-00 Transmission Survivor Curve Analysis Q. What materials are included in your Exhibit No. TPC-00? A. Exhibit No. TPC-00 is made up of the workpapers supporting my depreciation and NS rate recommendations for the Trailblazer pipeline system. The schedules

9 Page of present the major components of the process of developing the proper and adequate depreciation rates given Trailblazer s current operations and the recovery of its investment in plant over the remaining useful life of those assets. The following schedules are included in Exhibit No. TPC-00: 0 Schedule No. Schedule No. Schedule No. Schedule No. Schedule No. Schedule No. Schedule No. Schedule No. Schedule No. Recommended Depreciation Rates Plant Balances Near-Term Plant Additions Compression Depreciation Model Parameters Average Remaining Lives Transmission Depreciation Rate Derivation Terminal Negative Salvage Estimate Transmission Negative Salvage Rate 0 Q. What materials are included in your Exhibit No. TPC-00? A. Exhibit No. TPC-00 is the survivor curve analysis workpapers developed for the major property accounts in the Transmission Function. Q. Please provide a brief overview of your prepared testimony. A. My prepared direct testimony will address six general areas. In this first section, I provide a summary of current and proposed depreciation rates. In the second section, I provide a short discussion about depreciation theory to place in context the causes and concepts behind depreciation rate derivation. In the third section, I discuss survivor curve theory, a cornerstone of depreciation theory which derives the average service lives, the retirement decline patterns, the average remaining lives, and the interim retirement forecasts. In the fourth section, I discuss Simulated Plant Record ( SPR ) theory, which applies survivor curve theory to

10 Page of 0 the plant data available. In the fifth section, I discuss the account-by-account specific results of the survivor curve analysis and its impact on the depreciable lives on each account. In the last section, I discuss NS theory regarding the dismantling and removing of the pipeline system upon the termination of its useful life. Q. Will you summarize your recommendation in this case? A. Yes. Provided below is a summary of Trailblazer s current depreciation rates along with the rates I support in this testimony: Current Rates Proposed Rates Depr. Neg. Salv. Depr. Neg. Salv. Transmission Existing System Plant Expansion System Plant 0.%.0% 0.% 0.%.0%.% General 0.00% 0.% II. Q. Please describe depreciation theory. DEPRECIATION THEORY 0.% 0.% A. Depreciation is a term used in accounting, economics, and finance to convey the concept of the inherent loss of value in an entity s capital assets over time and the associated allocation of that loss in capital value over a defined period. Capital costs are those costs incurred to acquire plant and equipment that will be used over several accounting periods to facilitate the provision of an entity s goods and services. When investors purchase assets they expect to get their money back and earn a profit on that investment, i.e., the return of investment as well as the return on the investment. Depreciation and amortization are the means by which capital

11 Page of costs are allocated over the economic lifespan of the assets. Depreciation theory and practice encompass several concepts regarding the lifetime of an asset: 0 Physical life the time over which an asset can physically function given the ordinary forces that cause an asset to depreciate. Average Service Life ( ASL ) the average amount of time over which the group of assets in the class will provide useful service absent truncating events that bring economic usefulness to an end. Economic life the time over which an asset will provide useful economic value. Remaining economic life the time between the depreciation study date and the terminal closure and dismantling of the system. Average Remaining Life ( ARL ) the time between the study date and the terminal date as adjusted for the interim retirement of plant and equipment. 0 0 The maximum physical life of an asset generally is expected to substantially exceed the asset s ASL, but the economic life of a particular asset can be substantially shorter than either the physical life or ASL. The recovery of the capital costs must occur within the economic lifespan of an asset. Q. How does the Commission define depreciation? A. The Commission defines depreciation as: [T]he loss in service value not restored by current maintenance, incurred in connection with the consumption or prospective retirement of gas plant in the course of service from causes which are known to be in current operation and against which the utility is not protected by insurance. Among the causes to be given consideration are wear and tear, decay, action of the elements, inadequacy, obsolescence, changes in the art, changes in demand and requirements of public authorities, and in the case of natural gas companies, the exhaustion of natural resources. C.F.R. Pt. 0, Definitions.B (0).

12 Page 0 of 0 Q. What does loss in service value mean? A. Loss in service value is the diminishment of the ability of an asset to provide useful service to the utility. Loss in service value occurs broadly from two sources: first, physical causes such as wear and tear, decay, and action of the elements; and, second, what can be classified as economic causes (inadequacy, technological or economic obsolescence, changes in the art, changes in demand, requirements of public authorities, and the exhaustion of natural resources). Q. Based on the factors you have discussed, what economic life are you using for Trailblazer s facilities? A. I am using an economic life of years for Trailblazer. Trailblazer Witness Alexander J. Kirk supports a -year economic life for Trailblazer, which I rely upon for my recommendation concerning applicable depreciation rates. As 0 discussed in Witness Kirk s direct testimony, Exhibit No. TPC-00, long-term forecasting becomes more and more speculative the further out the forecast. It is reasonable to assume natural gas will be available for at least years consistent with the timeframe traditionally used by FERC to assess natural gas supply. However, Witness Kirk discusses certain factors that are likely to reduce the demand for Trailblazer s transportation services in the future and takes note of the fact that reputable projections, such as the U.S. Energy Information Administration s Annual Energy Outlook, only extend years to 0. Thus, Witness Kirk concludes that while sufficient natural gas supply may be available to Trailblazer over the next years, he does not believe there is reliable evidence

13 Page of to support an economic life for Trailblazer beyond 0. For those reasons, he supports a -year economic life for Trailblazer, which I adopt. Q. What depreciation methodology did you use for Trailblazer? A. I used the broad group, straight line, ARL method of depreciation. Under this method, which is the standard method for FERC-regulated pipelines, all of the assets within a group are considered to be homogeneous units of plant used and treated alike across the system regardless of the vintage, construction techniques, or retirement rate. In practice, there are two levels of grouping by FERC account and by function. Generally, all assets within a FERC account are 0 considered as one group and a depreciation expense is derived. Then the FERC accounts are combined into a larger functional group, such as storage or transmission, with one depreciation rate for the whole function. Where 0 operational considerations warrant, assets within a given FERC account are grouped in a different function, such as offshore transmission versus onshore transmission, to reflect the distinctive use and depreciable life expectations for those assets. I used whole life rates for general plant accounts. Q. On Schedule No. of Exhibit No. TPC-00, you report your estimated average service lives for each property account. Why have you truncated the service lives in the Average Remaining Life column? A. Most pipelines incorporate a truncation date in deriving depreciation rates to reflect the fact that the average actual useful lifespan of the assets can be significantly shorter than the physical ASL. The incorporation of a truncation date is often unrelated to the physical characteristics of the asset itself, but may

14 Page of 0 0 instead be implemented due to reasons such as the loss of reserves supporting its use, technical obsolescence bringing about replacement, or the requirements of public authorities that may lead to economic obsolescence of certain facilities. The truncation may cause the remaining life of the assets to be less than the average physical life, as is the case for Trailblazer. Q. Schedule No. of Depreciation Workpapers Exhibit No. TPC-00 includes near-term capital plant additions. Explain. A. I estimated near-term additions for the period 0 through 0 for inclusion in Schedule No. by taking the average of the plant additions net of recent remediation expenses for the years 0 through 0 for each transmission function account. Q. Is the concept of relying upon near-term plant balances consistent with current depreciation theory and practice? A. The concept of using the average plant balance is inherent in the use of survivor curves to develop the average remaining life. It is a long-accepted depreciation concept that depreciation rates are calculated using the average plant balance over the remaining life of the assets to ensure the proper rate of recovery of the plant investment. Incorporating only the estimated reductions in plant balances (via the survivor curve), but ignoring known and measurable near-term additions, would cause an under-recovery. For the period the rates are in effect, the depreciation recorded on Trailblazer s books will reflect the near-term plant additions but will not be matched by accruals through rates if the plant is not included in the rate development.

15 Page of Q. How is the development of depreciation rates different from NS rates? A. The rate development of depreciation rates and NS rates are quite similar. The difference is that depreciation rates recover the past cost of building the system and NS rates recover the future cost of removing the system. In both cases, the amount already accrued is subtracted from the base amount to determine the amount remaining to be accrued over the remaining life. That net amount is divided by the remaining life and then divided by the plant in service to reveal the depreciation or NS rate, as illustrated below. Depreciation Rates Negative Salvage Rates 0 0 Plant in Service Terminal Removal Costs - Acc. Res. Depr. - Acc. Res. Neg. Salv. = Net Plant = Net Neg. Salv. to Recover ARL ARL = Depr. Expense = Neg. Salv. Expense Plant in Service Plant in Service = Depr. Rate = Neg. Salv. Rate III. SURVIVOR CURVE THEORY Q. What is a survivor curve? A. A survivor curve is the pictorial end result of an actuarial analysis of hundreds of thousands of transactions that make up the life story of industrial property accounts. With each passing year, the retirements of property, if any, leave a smaller percentage of the original installation in place. If retirements were uniform in size and regularity, a simple straight-line projection would provide an adequate forecast of future retirements, and, in turn, allow for the calculation of the ARL of the assets. But the retirement patterns of industrial property do not follow a straight line. The retirement patterns of industrial property are

16 Page of characterized by a complex life trajectory, which includes a transition point where survivorship takes a dramatic downward turn. The retirement rate and survivorship rate are inversely related phenomena. The upside-down bell curve shape of retirement frequency distribution creates the ski-slope shape survivorship curves. After a period of substantial retirements, the retirement pattern passes through another transition point where retirements fall off, leaving a long tail of lingering survivorship. The overall lifespan survivorship trajectory for most 0 industrial property follows this ski slope pattern seen in Graph No. that, despite an appearance of simplicity, requires complex mathematical formulae to replicate. Graph No. Retirement Distribution & Survivorship Rate Adding to the complexity, additions to plant, transfers between accounts, and various adjustments to plant accounts over time, can obscure the patterns of retirements, making it difficult to discern the physical life expectancy of plant and equipment. Survivor curve analysis translates the hundreds of thousands of data

17 Page of points into recognizable patterns, enabling an analysis and forecast of future life expectancies. Q. Is survivor curve theory accepted by the Commission? A. Yes, the Commission has approved depreciation rates based on survivor curve theory for over 0 years. Q. How does survivor curve theory work? A. The survivor curve analysis primarily deals with two survivor curves: one being a curve that traces the actual surviving dollars from each vintage of plant addition and the other a prototypical Iowa Curve selected to carry the trend of the actual 0 data out into the future for forecasting purposes. Once the original data is synthesized into an original experience survival curve (see Graph No. ), the curve is compared to hundreds of prototypical curves (see, e.g., Graph No. ) to find one that will best forecast the most likely experience of future interim retirements. With the retirement forecast in hand, the ARL can be calculated.

18 Page of Graph No. Survivor Curve of Original Data

19 Page of Graph No. R Family of Iowa Curves Q. What are Iowa Curves? A. Iowa Curves represent standardized retirement patterns of industrial property developed from actuarial studies conducted in the 0s. The Iowa Curves 0 consist of families of curves that reflect left-modal, symmetrical-modal, and rightmodal frequency distributions, called simply L, S, and R curves. Each family of curves includes four to five curve sets within the family, labeled R, R, R, and so on, each with slightly different slope configurations (see Graph No. above). The modality of the curves simply reflects whether the most frequently occurring retirement age is younger or older than the average retirement age. For example, an L Curve represents retirement data in which the most frequently occurring (i.e., the mode) retirement age is younger, or to the left of the average on the graph.

20 Page of 0 An R Curve represents retirement data in which the most frequently occurring retirement age is older, or to the right of the average on the graph. Graph No. also illustrates the wide variety of retirement patterns that can occur within each family of curves, from plant that experiences retirements almost immediately after installation (as in the R type curve) to plant that may go a very long time before any significant retirements take place (as in the R type curve). Q. How does a difference in curve selection affect the depreciation rate? A. The curves estimate the future retirement patterns. A steeper decline rate will reduce the economic life and average remaining life, leading to higher depreciation rate as illustrated in Graph No. below. The L Curve has a gentle slope and, over the course of the next years, predicts a lower level of retirements than the R Curve. Between the study date and the terminal date, the R Curve suggests that almost all the plant will be retired so that the ARL will be significantly lower than the remaining economic life of years. With the L Curve, most of the plant will survive all the way to the th year so that its ARL will be much closer to years than would the R Curve plant.

21 Page of Graph No. Comparison of Curves 0 Q. How is a best fit curve selected? A. The selection of a curve is done by a combination of a statistical comparison and informed knowledge of the nature of the assets. The statistical assessment is a simple calculation of the differences between the original data and the selected Iowa Curve. The differences are squared to eliminate positive and negative differences from cancelling each other out as well as to accentuate deviations. The curve with the least sum of squared difference between the actual book value of the account and the predicted value of the account is generally the best fitting curve and, unless some other factor weighs heavily in the analysis, that curve will be used to forecast future retirements. This concept is illustrated in Graph No..

22 Page 0 of Graph No. Iowa Curves Represent the Original Data 0 Q. Do you always use the curve identified by the model as the best fit curve? A. No, I do not. The Iowa Curve with the least sum of squared differences may fit the overall pattern of the original survivor curve, but may not fit the portion of the original life curve relevant to the timely recovery of the utility s investments. For depreciation purposes, the interim period between the study date and the termination date defines the period over which the remaining undepreciated plant investment must be recovered. The sum-of-least-squares calculation measures the differences between the actual and forecasted curves along the entire span of the curve from 0 to 00 percent of the ASL. Tracking the retirement pattern over the interim period is more important for estimating the ARL relevant to recovery of these assets than tracking a long-term pattern that will not come to pass due to the truncation of the life of the assets.

23 Page of Q. Describe the concept of the average service life. A. The physical plant of large industrial entities like pipelines is made up of thousands of units of property. At some point assets begin to drop out of service so that some plant units have shorter lives than others. Since we don t know, a priori, what the ASL is, we use the survivor curve methodology to estimate that figure. By matching the original survivor curve (see Graph No. ) with 0 representative curves for each possible surviving age for each curve type, we can find a pair of curve types and ASLs that best mimic the original pattern. My model tests age groups from years to 0 years ASL. Q. Describe the concept of the average remaining life. A. The ARL calculation is restricted to the time between the study date and the termination date. The ARL, commonly known as the area under the curve as illustrated in Graph No. below, represents the time period over which the company s remaining net plant will be depreciated. Dividing the sum of the surviving balances as calculated by the survivor curve by the starting balance provides the ARL, which is used in the depreciation calculations.

24 Page of Graph No. - Average Remaining Life & Area Under the Curve 0 Q. In instances where a pipeline has property accounts that have not experienced significant retirements, can survivor curves still be useful? A. Yes, they can. Remember that survivor curve theory rests on the massing of huge amounts of data about industrial property lifespans. The work done by Robley Winfrey in the 0s demonstrated that there were clear patterns to the retirement of industrial property. The massing of data allows the highs and lows to even out so that the underlying patterns can be more clearly seen. Where any given year (or operating entity, or lateral grouping, or vintage era) may experience a lull in retirement activity, the overarching data patterns alert us to the fact that the retirements are coming. Survivor curves allow us to estimate a normal retirement pattern for planning and investment recovery purposes. Where the data

25 Page of is insufficient, or retirements are late to arrive, we can still forecast with confidence that retirements will happen and from those forecasts the ARL can be derived for depreciation purposes from the survivor curves. IV. SIMULATED PLANT RECORD ANALYSIS 0 Q. The actuarial survivor curve model you described requires vintaged retirement, i.e., retirement data tied to the year of installation of the assets. What happens if the vintaged data is not available? A. In the absence of sufficient data tying the annual retirements back to the specific year of installation, as is the case here, the SPR model can provide a similar forecast of annual retirements for depreciation rate purposes. In fact, it is more often the case that sufficient vintaged data is not available to perform the full actuarial analysis of retirement patterns. Fortunately, we have the SPR model as a reliable, alternative tool for estimating survivorship rates. An SPR model relies on the survivor curve theories for its foundation but does not generate a leastsquares analysis for each data point. Instead, the SPR model applies a prototype Iowa Curve to each annual plant addition and calculates a final balance for the account, assuming the plant has retired in a pattern similar to that of the Iowa Curves. The selected curve is used to forecast future retirements, which provides the ARL and ultimately the depreciation rate. The SPR model I use was 0 developed by BWMQ and is proprietary intellectual property. Q. How does the SPR model represent the actual plant activity? A. As plant ages, the surviving plant ratio falls as it moves along and down the survivor curve. The average age of the plant in each account determines where

26 Page of 0 0 the account is, vis-à-vis the survivor curve, at the study date. The SPR method calculates a theoretical retirement trajectory that it applies to each vintage of additions. The curve that best forecasts a plant balance closest to the actual plant balance is deemed, generally, to be the best representative pattern for all vintages. That declining survival ratio determines the interim retirements expected to take place between the study date and the terminal date. These retirements, in turn, are the foundation for determining the ARL for depreciation purposes. Q. Is there a goodness-of-fit measurement to gauge the accuracy of the model? A. There is a goodness-of-fit measurement that traditionally has been used in depreciation studies called the Conformance Index ( CI ). The CI is derived by dividing the actual ending balance by the absolute value of the difference between the actual ending and the predicted ending balance. Actual surviving balance Predicted surviving balance Actual surviving balance The predicted ending value is squared to eliminate negative numbers and then the square root is taken to hold the predicted value as close to the actual value as possible. If the difference between the predicted and actual ending balances is high, then the CI ratio will be low. Conversely, if the difference between the predicted and actual ending balances is low, then the CI ratio will be high. The rule of thumb for ranking CIs is:

27 Page of Over excellent fit 0 to good fit to 0 fair fit Under poor fit 0 0 The rationale for the CI valuation is that in order for the CI to reach a value of, the difference between the actual and predicted balances must be within. percent of the actual balance. A CI value of 0 indicates a differential of only percent. This ranking system thus requires the forecasted values to fall close to the actual values to be considered even a fair fitting of a hypothetical Iowa Survivor curve to the actual data. A CI value above 00 indicates a forecast fit that is within percent of the actual data; larger values for the CI over 00 do not indicate a significantly better fitting curve. If more than one curve has a CI beyond 00, the analyst incorporates other factors to select an appropriate curve. As the difference between the predicted ending balance and the actual ending balance gets smaller, the CI value increases. As the difference approaches zero, the CI approaches infinity. Q. Is the CI a reliable basis for determining a best fit curve? A. Not always. In fact, the CI often can calculate a fit for an Iowa Curve that significantly misrepresents the likely survivor pattern of a category of property. The CI calculates the closeness of fit that each prototype Iowa Curve achieves in forecasting the actual surviving plant balance, i.e., a specific dollar value at a point in time. However, for depreciation purposes we need more than a forecast of the surviving balance at one point in time; it is also important to glean the

28 Page of trajectory of the decline curve and the amount of annual retirements. A good forecast should reflect actual experience as much as possible, but it is often the case that the best fit curve and service life pair come from a survivor curve pattern that predicts near-term retirements that are wildly divergent from the pipeline s actual recent experience. For example, in Graph No. above, both survivor curves accurately predict the current surviving balance and would thus have high CIs but take very different trajectories to get there. The L Curve has a shallower curvature and forecasts modest retirements over the remaining life of the asset. The R Curve has a steep declining curvature and forecasts the 0 0 retirement of almost all the plant over the remaining life. In such cases, I try to select an ASL/Iowa curve that forecasts near-term retirements as close to the actual experience of retirements so that the resulting depreciation rate reflects the actual ARL of the plant. Q. Do you incorporate any other measures to test the goodness-of-fit for the survivor curves? A. Yes, I do. As noted above, there can be many ASL/Iowa Curve pairs that are plausible representatives for the long-term retirement pattern. There can be a significant difference in the forecasted interim retirements among the contending curve and ASL pairs, and thus a significant difference in the derived depreciation rate (see Graph No. above). To help select among the contending curves, I compare the average five-year near-term retirement forecast to the average of the most recent five-years of actual retirements.

29 Page of 0 0 V. REMAINING LIVES AND DEPRECIATION RATES Q. How did you determine the average remaining lives for each property account? A. The BWMQ SPR model estimates the average remaining lives for each property account by simulating the retirement pattern and declining survivorship of the plant. The model assumes that all plant investment within each property account follows the same retirement pattern, calculating a surviving balance for every year s additions as of the study date. The model makes this calculation for curve types over a range of ASLs from to 0 years. The combination that best matches the actual experience of the pipeline is then used as the underlying ASL for forecasting surviving balances through to the truncation date, revealing the average remaining life. The selection of the specific curve is made on the basis of the CI to the extent the trajectory of the decline curve reasonably reflects actual experience, the extent to which near-term retirements match up to recent actual retirements, the long-term pattern of plant retirements as seen on the graphs in the associated exhibits, and my experience with retirement data resulting from over 0 years of conducting depreciation analyses. Q. Describe the schedules included in Exhibit No. TPC-00, Transmission Survivor Curve Analysis. A. Exhibit No. TPC-00 includes the schedules that reflect the data, survivor curves, and retirement rates for each of the major accounts in the Transmission function. Each schedule within the exhibits has five pages. Each schedule in Exhibit No. TPC-00 represents a property account for which a survivor curve

30 Page of was modeled. At the top of each schedule is a set of salient statistics reflecting selected results of the modeling. In the first line of the first column, the circled Curve, indicates the selected ASL/Iowa Curve pair used to forecast the surviving balances and retirements by which the ARL is calculated. The three subsequent lines of the first column are presented for comparison purposes to show other approaches to selecting Iowa Curves. Q. Please explain how the schedules are structured. A. Page of each schedule contains four ASL/Curve pairs that represent different selection criteria. 0 The first is the curve I selected for this depreciation study. The second shows the pairing that most closely matches forecasted nearterm retirements to recent past retirements. The third shows the industry standard curve -R, which is often used where data does not support a strong alternative. The fourth is the pairing with the highest CI value, which as I previously discussed is sometimes characterized as the best fitting curve. 0 The first page also shows the referenced curves in graph form. Page of each schedule has graphs that present the growth in plant over time and the history of retirements. Page of each schedule show the historical data of plant additions, retirements, adjustments, and transfers. Page of each schedule presents various curve selection options with forecasted retirements, CIs, and average remaining lives for comparison. Page of each schedule presents the forecast of survivorship and interim retirements based on the curve and calculates the near- term average interim retirement and the concluding ARL figure.

31 Page of 0 0 A. Transmission Plant Q. Describe your assessment of Account. Rights of Way. A. Account. Rights of Way, Exhibit No. TPC-00 at -, includes the cost of acquiring the rights to use property for pipeline operations. The plant history, shown on page of Exhibit No. TPC-00, reflects quite low levels of retirement activity. As shown on page of Exhibit No. TPC-00, the best fit ASL/Iowa Curve pairings, as determined by the CI alone, indicate service lives of to 0 years with modest near-term retirements, yet the predicted retirements are still considerably above the recent actuals. I selected the -S curve to represent the survivorship pattern based on the near-term retirements forecast. As shown on page of Exhibit No. TPC-00, this selection results in an ARL of. years. Q. Describe your assessment of Account Structures & Improvements. A. Account Structures & Improvements, Exhibit No. TPC-00 at -, is the account that holds the costs associated with the buildings, garages, and landscape improvements that host the equipment needed to operate the system. The plant history reflects almost no retirement activity throughout its history. Exhibit No. TPC-00 at. The best fit ASL/Iowa Curve pairings, as determined by CI alone, indicate -year service lives, which is clearly untenable inasmuch as such structures last far longer than five years. Id. at 0. The available data for this account opens with a balance larger than the surviving balance which skews the comparisons of prototype curves and generates unreliable forecasts. Instead, I selected the 0-R curve to represent the survivorship pattern based on the five-

32 Page 0 of year average near-term retirements forecast. As shown on page of Exhibit No. TPC-00, this selection results in an ARL of. years. Q. Describe your assessment of Account Mains. A. Account Mains, Exhibit No. TPC-00 at -, holds the costs associated with the line pipe itself, including crossings, line pack, valves, hauling, and inspections approximately $ million. Exhibit No. TPC-00 at (showing the Form end of year balance). The plant history reflects very little retirement activity over an extended period until just recently, as indicated by the retirement graph on page of Exhibit No. TPC-00. I selected the 0-S curve to 0 0 represent the survivorship pattern based on the near-term retirements forecast. As shown on page of Exhibit No. TPC-00, this selection results in an ARL of. years. Q. Describe your assessment of Account Compressor Station Equipment. A. Account Compressor Station Equipment, Exhibit No. TPC-00 at -, is made up of costs associated with the compressor engines that pressurize natural gas to push it through the pipeline system, and the related pipes, meters, control valves, platforms, and exhaust handlers approximately $ million. Compressor station equipment is made up primarily of compressors that are powered by either electric reciprocating engines or gas turbines. The Existing System transmission includes two gas turbine compressor engines for which the average service life is based on the average annual operating hours between overhauls, as shown on Exhibit No. TPC No. 00, Schedule. The gas turbines are significant investments that are very efficient but relatively short-lived. At a

33 Page of 0 0 relatively early age in their lifespan, the engines must be removed for a complete overhaul, and replaced by a newly overhauled engine. For Expansion Plant Account assets, the plant history reflects modest routine retirement activity from 00 through 0, as shown on pages and of Exhibit No. TPC No. 00. The best fit ASL/Iowa Curve pairings indicate service lives of 0 to years with near-term retirement forecasts somewhat above the actual experience of late as shown on the table at the top of page 0. I selected the 0-R curve to represent the survivorship pattern based on the nearterm retirements forecast. As shown on page of Exhibit No. TPC-00, this selection results in an ARL of. years. Q. Describe your assessment of Account Measuring & Regulating Equipment. A. Account Measuring & Regulating Equipment, Exhibit No. TPC-00 at -, holds the costs of the meters, gauges, and minor piping needed to monitor and control the system. The plant history reflects very few retirements throughout its history. The best fit ASL-curve pairings, as indicated by the CIs, indicate service lives of 0 to years with forecasted near-term retirements within range of the actual experience of late. I selected the -S curve to represent the survivorship pattern based on the near-term retirements forecast. As shown on page of Exhibit No. TPC-00, this selection results in an ARL of. years.

34 Page of 0 0 Q. Describe your assessment of Account 0 Communications Equipment. A. Account 0 Communications Equipment, Exhibit No. TPC-00 at -, holds the costs associated with the telephones, microwave equipment, towers, and receivers used to manage the system approximately $,000. The average age of the assets in this account is. years. The plant history reflects almost no retirements. The best fit ASL-curve pairings indicate service lives of to years with forecast near-term retirements somewhat above the actual experience of late. I selected the 0-L curve to represent the survivorship pattern based on the near-term retirements forecast. As shown on page of Exhibit No. TPC- 00, this selection results in an ARL of. years. Q. Describe your assessment of Account Other Equipment. A. Account Other Equipment holds the costs associated with plant that does not fit other categories. Most of the assets in this account were retired in the mid- 0s, the remaining plant is not sufficient to run a survivor curve on its own. I substituted the R curve to model the remaining assets service life, resulting in a 0.-year average remaining life. Q. What is the composite depreciation rate for the whole Transmission function? A. I have calculated the composite depreciation rate for the Existing System Transmission plant to be.0 percent. I have also calculated a. percent rate for the Expansion System compression Ex. No. TPC-00 at Sch..

35 Page of 0 B. General Plant Q. What are your conclusions regarding General Plant depreciation rates? A. The general plant accounts are commonly used for plant and equipment that turns Acct over on a regular basis and thus the depreciation rate is designed for plant that replenishes when it reaches the end of its service life. This approach is known as a whole life method. The rate is derived by dividing 00 percent by standard ASLs as listed below. I have composited the General Plant accounts into one function-wide depreciation rate of 0. percent.. Acct Name Office Furniture & Equipment Furniture & Non-IT Equipment Average Service Life Whole Life Rate.%. Office Furniture & Equipment IT Eq. 0.00%. Transmission Plant under one ton.0%. Transmission Plant over one ton.%. Transmission Plant Heavy Equipment % Tools, Shop, & Garage Equipment..% Power Operated Equipment % Communications Equipment % Miscellaneous Equipment unknown 0.0% Although some accounts do not currently have plant in them, a whole life rate is recommended so that should plant be added in the future. Trailblazer can begin booking appropriate depreciation expense as soon as the plant is in service. I would note, however, that where there is no plant at this time, no depreciation expense associated with that account is included in the tariff rate derivations.

36 Page of Further, should plant be added, any booked depreciation expense will not be recovered in tariff rates until such time as the new plant is included in a rate case. VI. NEGATIVE SALVAGE RATE 0 0 Q. Does Trailblazer currently have NS rates? A. Yes. Trailblazer currently has a 0. percent NS rate for its Transmission plant. Q. Do you have a recommendation for establishing new NS rates on Trailblazer? A. Yes, I do, because the current NS rate is insufficient to recover the costs that will be incurred upon the closure and dismantling of the system upon the termination of its useful life. Q. How were the terminal NS cost estimates derived? A. Upon reaching the end of the useful life of the pipeline assets, the facilities are either physically removed from the service area for disposal or made safe for abandonment in place. The cost of removal and remediation for the storage and transmission function plant was estimated by Trailblazer Witness Fall. Witness Fall s exhibits reflect the cost of removal and dismantling on a project-by-project basis such as a river crossing or dismantling of a compressor station, rather than by FERC account numbers. I have reformatted those costs by FERC property accounts as shown on Schedule No. of Exhibit No. TPC-00 and calculated the NS rate for the recovery of these costs in Schedule No. of Exhibit No. TPC- 00. Because the Existing Transmission and Expansion Transmission Plant will be dismantled at the same time I developed one rate for both systems.

37 Page of 0 Q. How did you estimate interim retirements? A. Interim retirements are those plant and equipment facilities that will be withdrawn from service prior to the terminal shut-down date assumed in the depreciation model. The costs of these facilities were estimated by the survivor curve model and carried over into the NS model. Q. How did you derive the composite NS rate? A. The composite NS rate is developed by weighting the plant subject to NS by the estimated remaining life of the facilities. While the cost of removal for some facilities can be accrued over the remaining years assumed in the depreciation model, other facilities will be retired in the interim period. The cost of removal for the interim retirements should be recovered over their average remaining life. The workpapers in Schedule No., Exhibit No. TPC-00, calculate the weighted ARL for all the facilities with anticipated costs of removal. Q. What is the composite NS rate? A. The composite rate for the terminal NS is 0. percent. Q. Does this conclude your Prepared Direct Testimony? A. Yes, it does.

38