EPCOR Energy Alberta GP Inc.

Decision 20633-D01-2016 EPCOR Energy Alberta GP Inc. 2016-2017 Regulated Rate Tariff Application December 20, 2016

Alberta Utilities Commission Decision 20633-D01-2016 EPCOR Energy Alberta GP Inc. 2016-2017 Regulated Rate Tariff Application Proceeding 20633 December 20, 2016 Published by the: Alberta Utilities Commission Fifth Avenue Place, Fourth Floor, 425 First Street S.W. Calgary, Alberta T2P 3L8 Telephone: 403-592-8845 Fax: 403-592-4406 Website: www.auc.ab.ca

Contents 1 Introduction... 1 2 Forecast revenue requirement and monthly non-energy charges... 3 3 Compliance with Commission directions... 4 4 Issues... 5 4.1 Corporate vacancy rate... 5 4.2 Salary escalations... 7 4.3 Operating costs... 9 4.3.1 Customer service costs... 9 4.3.2 Late payment charges... 12 4.4 Tax and deferral accounts... 13 4.4.1 Bad debt deferral... 14 4.5 Depreciation and return... 16 4.5.1 Weighted average cost of capital... 16 4.5.2 Interest during construction and allowance for funds used during construction... 17 4.6 Risk compensation... 19 5 Other matters... 29 5.1 Price schedules and terms and conditions... 29 5.2 Errors and omissions... 29 6 Order... 30 Appendix 1 Proceeding participants... 31 Appendix 2 Summary of Commission directions... 32 List of tables Table 1. Forecast non-energy RRT revenue requirements for 2016 and 2017... 3 Table 2. Forecast monthly non-energy charges for 2016 and 2017... 3 Table 3. Historical approved and actual non-union salary escalators... 9 Table 4. Variance between forecast and actual customer service costs... 10 Decision 20633-D01-2016 (December 20, 2016) i

Alberta Utilities Commission Calgary, Alberta EPCOR Energy Alberta GP Inc. Decision 20633-D01-2016 2016-2017 Regulated Rate Tariff Application Proceeding 20633 1 Introduction 1. On July 16, 2015, EPCOR Energy Alberta GP Inc. (EEA) filed an application with the Alberta Utilities Commission requesting approval of its non-energy regulated rate tariffs (RRT) for service provided by EEA to eligible customers in the EPCOR Distribution & Transmission Inc. (EDTI) and FortisAlberta Inc. (FAI) service territories, for the period from January 1, 2016 to December 31, 2017 (the test period or the test years). EEA also requested confidential treatment of 1772387 Alberta Limited Partnership s, operating as ENCOR, customer site counts. EEA stated that the public release of this information would prejudice ENCOR s position in the competitive retail market by enabling competitors to gain insight into the effectiveness and potential future direction of ENCOR s marketing strategies. 2. On July 21, 2015, the Commission issued a notice of application. Any party who wished to intervene in the proceeding was requested to submit a statement of intent to participate (SIP) to the Commission by the participation closing deadline of August 7, 2015. The Commission received SIPs from the Consumers Coalition of Alberta (CCA) and the Office of the Utilities Consumer Advocate (UCA). 3. The CCA and the UCA requested that the Commission establish a process including written information requests (IRs), argument and reply argument. The CCA indicated that it would be able to determine whether or not it objected to the application after reviewing EEA s IR responses. The UCA indicated that it would be able to comment on the need for further process after reviewing EEA s IR responses. 4. On August 7, 2015, EEA filed a motion requesting approval to initiate a negotiated settlement process in respect of EEA s 2016-2017 RRT application. 5. The Commission issued a letter on August 12, 2015, allowing parties the opportunity to comment on EEA s confidentiality request and motion to pursue a negotiated settlement by August 17, 2015, and allowing EEA to reply to those comments by August 21, 2015. 6. Comments were received from the CCA and the UCA on August 17, 2015. EEA filed its reply comments on August 21, 2015. 7. The CCA agreed that it is reasonable for the Commission to grant confidential treatment of ENCOR s site counts. 8. The UCA stated that it had concerns about the nature and quantum of confidentiality requests received by the Commission. The UCA further stated that EEA s request for confidential treatment ought to be denied because EEA had failed to demonstrate that exceptional circumstances exist to rebut the presumption in favour of the open-court principle in Commission proceedings. In particular, the UCA argued that EEA had failed to describe the nature of the information for which confidential treatment was sought with the specificity Decision 20633-D01-2016 (December 20, 2016) 1

necessary to support its claim as to the harm that would result from public disclosure of the information. 9. EEA responded that it is obvious that ENCOR s site counts are commercially sensitive and confidential, and that the UCA s arguments to the contrary are without merit. EEA stated that the nature and quantum of confidentiality requests received by the Commission are not applicable to this proceeding as EEA has only made one confidentiality request of a limited nature. 10. On September 1, 2015, the Commission issued a ruling granting confidential treatment to certain portions of the document containing the ENCOR site counts and directing EEA to file a redacted version of the document on the public record of this proceeding. 11. The ruling also approved EEA s request to pursue a negotiated settlement process, after a round of IRs and IR responses, and directed EEA to provide an update on the status of negotiations no later than November 16, 2015. The deadline to file an update on the status of negotiations was subsequently extended to November 26, 2015, and later to December 4, 2015. EEA filed an update on December 4, 2015, indicating that parties had failed to reach a negotiated settlement. On January 6, 2016, EEA filed responses to a number of IRs asked by the UCA and the CCA during the negotiated settlement process. 12. On January 11, 2016, EEA submitted a supplementary filing outlining its proposed non-energy risk compensation for 2016 and 2017. 13. There were a number of deadline extensions and process additions throughout the proceeding. The Commission maintained a written process to assess the application, which resulted in the schedule below: Process step Date IRs to EEA September 21, 2015 Information responses from EEA October 7, 2015 Commencement of negotiated settlement process October 15, 2016 Update on status of negotiated settlement process December 4, 2016 EEA to file information responses after negotiations January 6, 2016 EEA to file supplemental filing on risk compensation January 11, 2016 IRs to EEA on supplemental filing January 19, 2016 Information responses from EEA January 28, 2016 Intervener evidence March 4, 2016 IRs to interveners March 18, 2016 Information responses from interveners April 1, 2016 Rebuttal evidence April 12, 2016 Argument April 29, 2016 Reply argument May 13, 2016 14. The Commission considers that the close of record for this proceeding was May 13, 2016. 2 Decision 20633-D01-2016 (December 20, 2016)

15. On September 27, 2016, the Commission issued a letter notifying parties that the original panel for this proceeding had recused themselves upon becoming aware of institutional circumstances that they determined gave rise to reasonable apprehension of bias. A new panel was appointed to consider the record and make the determinations herein. 16. In reaching the determinations set out within this decision, the new panel has considered all relevant materials comprising the record of this proceeding, including the evidence, argument and reply argument. Accordingly, references in this decision to specific parts of the record are intended to assist the reader in understanding the Commission s reasoning relating to a particular matter and should not be taken as an indication that the Commission did not consider all relevant portions of the record with respect to that matter. 2 Forecast revenue requirement and monthly non-energy charges 17. EEA applied for approval of its non-energy RRT revenue requirement and rates for service to eligible customers in the EDTI service area and the FAI service area for the period from January 1, 2016 to December 31, 2017. The forecast non-energy RRT revenue requirements for the 2016 and 2017 test years, included in the application, are set out in Table 1 below: Table 1. Forecast non-energy RRT revenue requirements for 2016 and 2017 2014 decision 2014 actual 2015 decision 2015 updated forecast 2016 forecast 2017 forecast $ million 44.76 40.09 45.19 42.09 37.80 38.47 18. The forecast monthly non-energy charges, as set out in financial Schedule 5 of the application, are reproduced in the table below: Table 2. Forecast monthly non-energy charges for 2016 and 2017 Customer type FAI service area 2016 forecast 2017 forecast $ per site per month Residential 5.51 5.62 Farm 5.09 5.19 Irrigation 3.77 3.61 Small commercial 5.70 5.81 Oil & gas 3.53 3.52 Lighting 5.27 5.36 EDTI service area Residential 5.40 5.50 Small commercial 5.20 5.30 Lighting 4.88 4.95 Decision 20633-D01-2016 (December 20, 2016) 3

19. EEA requested approval of the following: Continuation of a hearing cost reserve and short-term incentive deferral accounts. Price schedules, including miscellaneous fees and credits. Terms and conditions of service. 20. In the sections that follow, the Commission considers the issues that were raised by parties or that arose upon the Commission s review of the application, and provides its findings on EEA s application. Subject to the findings and directions outlined in this decision, the Commission approves EEA s: Forecast customer site counts Operating costs Corporate services costs Property taxes Hearing cost deferral account Depreciation expense 3 Compliance with Commission directions 21. Appendix D to the application contains EEA s descriptions of its treatment of Direction 21 from Decision 2012-272 1 and directions 1, 7, 8, 9, 12, 13, 14 and 16 from Decision 2014-303. 2 22. The Commission has reviewed EEA s proposed treatment of Direction 21 from Decision 2012-272 and Direction 1, which deals with disallowed EPCOR Utilities Inc. (EUI) investment costs, Direction 7 (quarterly revenue updates for bad debt), Direction 9 (effect of acquisitions), Direction 13 (forward interest rate curve analysis), Direction 14 (return margin) and Direction 16 (ENCOR operating costs) from Decision 2014-303. EEA s responses to these directions are complete and the Commission considers that, for the purposes of this application, EEA has complied with these directions. 23. EEA s proposed treatment of Direction 8 from Decision 2014-303, regarding late payment charge forecasting is discussed in Section 4.3.2 of this decision. EEA s proposed treatment of Direction 12 from Decision 2014-303, regarding EEA s capital structure is discussed in Section 4.5.1 of this decision. 1 2 Decision 2012-272: EPCOR Distribution & Transmission Inc., 2012 Phase I and II Distribution Tariff, 2012 Transmission Facility Owner Tariff, Proceeding 1596, Application 1607944-1, October 5, 2012. Decision 2014-303: EPCOR Energy Alberta GP Inc., 2014-2015 Non-Energy Regulated Rate Tariffs, Proceeding 2986, Application 1610188-1, November 4, 2014. 4 Decision 20633-D01-2016 (December 20, 2016)

4 Issues 4.1 Corporate vacancy rate 24. In response to a Commission IR, EEA updated its global vacancy rate to 4.5 per cent. 3 This vacancy rate is applied to all contact centre full-time equivalents (FTEs) that are not customer service consultants, and to all other EEA FTEs other than customer service consultants. 25. EEA stated that this vacancy rate is not applied to EUI corporate employees because EUI does not track FTEs. Instead, EUI monitors employees on a headcount basis. 4 26. In evidence submitted on behalf of the UCA, Ms. Radway stated that regulated utilities including AltaLink L.P. (AltaLink), ATCO Electric Ltd. and FAI show the FTEs for corporate staff in their Schedule 25.5 Corporate Manpower Full Time Equivalents, and that a vacancy rate is applied for corporate employees. ATCO Electric, AltaLink and FAI apply the same vacancy rate as applied to either transmission FTEs or distribution FTEs as they do for corporate employees. 27. Ms. Radway stated that there are vacancies within EUI, just as there are within EEA. Ms. Radway cited EEA s application in which two positions within information services application support were vacant for a period of time in 2014. For that period, the forecast direct assigned cost for information services application support was $1.57 million, while the actual was $1.19 million. Not accounting for the vacancies resulted in over-forecasting in the amount of $0.38 million. 28. Ms. Radway noted that, although EUI is an unregulated entity, it does allocate some of its corporate costs to EEA and EDTI, both of which are regulated entities. Because some corporate costs are allocated to regulated entities, Ms. Radway recommended that a 4.5 per cent vacancy rate should apply to all corporate equivalent headcounts for EUI and there should be a corresponding reduction in salary costs, including all benefits and incentive costs for the regulated entities. Because the vacancy rate for corporate services may differ from EDTI and EEA, Ms. Radway recommended that the vacancy rate could be tracked to ensure an accurate forecast of corporate services, union and benefits and incentives costs for future test years. 5 29. In its rebuttal evidence, EEA stated that Ms. Radway s recommendation is flawed and inaccurate, and must be rejected. Since EUI uses headcount, it cannot accurately forecast a vacancy rate. Due to the finance optimization project, which is anticipated to be complete in 2016, EUI can now accurately monitor and track its actual corporate services vacancies. With this newly available data, EEA can commit to providing the actual vacancy rate for corporate services employees for 2016 and 2017, and then applying a forecast vacancy rate for all corporate headcount equivalents for the next test period. 6 30. EEA stated that the global vacancy rate is calculated based on actual vacancy rates from previous years and is not transferrable to EUI personnel. EUI corporate services and EEA are distinctly different businesses with different operations and personnel types. Most of EEA s 3 4 5 6 Exhibit 20633-X0062, information responses, EEA-AUC-2015OCT07-004, tables 1.4.3-1 and 1.4.3-2, PDF pages 11-12. Exhibit 20633-X0073, information responses, EEA-UCA-2015OCT07-020(a) and (b), PDF page 75. Exhibit 20633-X0130, evidence of Shelley Radway, paragraphs A40 and A41, PDF page 28. Exhibit 20633-X0143, EEA rebuttal evidence, paragraph A90, PDF page 64. Decision 20633-D01-2016 (December 20, 2016) 5

personnel are unionized entry-level positions, while corporate services staff are mostly management and professional positions. Some of these corporate services positions are rarely vacant and when the departure of an incumbent is announced, an overlapping transition period is often used. 7 31. EEA also submitted that AltaLink, ATCO Electric and FAI apply vacancy rates to FTEs, not to headcounts. EUI is not a regulated entity and is, therefore, not required to make FTE calculations. EUI, therefore, does not have historical information available upon which to forecast corporate vacancies accurately. If it were to calculate a vacancy rate, it would be calculated based on a point in time during the year because EUI does not track FTEs. 8 32. The UCA argued that EUI s shared services salary and labour costs for 2016 and 2017 are $30.4 million and $31.44 million, respectively. Reducing those values by 4.5 per cent would equate to $1.37 million for 2016 and $1.41 million for 2017. This is equivalent to calculating the total salary and labour costs, and dividing them by the total headcount to determine the average cost per headcount. The same results would be obtained by applying the 4.5 per cent reduction to the headcount and multiplying that value by the average cost per headcount. 9 33. The UCA argued that it is not necessary for EEA to track its FTEs for a vacancy rate to apply to EUI for the test years. 34. In reply argument, the UCA cited EEA s statements regarding the finance optimization project and forecasting a corporate vacancy rate in future test years, and then argued that EEA seems to recognize that applying a vacancy rate to corporate employees is appropriate and necessary. The UCA argued that EEA relies heavily on the fact that EUI is not required to track corporate FTEs, and indicated that the other utilities mentioned by Ms. Radway apply corporate vacancy rates that mirror those applied to the companies transmission and distribution employees. 10 Commission findings 35. EEA argued that a corporate services vacancy rate cannot be calculated and forecast because EUI tracks headcounts rather than FTEs. EEA provided an example to support its assertion that any corporate services vacancy rate calculation would be based on a single point in time. In the example provided, a vacancy calculation is only calculated on a single day during the test year. The Commission finds that EEA does not adequately explain why such calculations could not be made throughout the year to determine an average vacancy rate for the year. There is not sufficient evidence to suggest that an alternative method could not be employed by which a vacancy rate could be calculated for corporate services. EEA has not demonstrated that a vacancy rate specific to corporate services cannot be calculated for regulatory purposes. 36. Furthermore, given the evidence presented by Ms. Radway regarding vacancies in information services application support and given EEA s proposal to track corporate vacancies during 2016 and 2017, vacancies do exist within EUI corporate services and the application of a vacancy rate to those costs is supported. 7 8 9 10 Exhibit 20633-X0143, EEA rebuttal evidence, paragraph A91, PDF page 65. Exhibit 20633-X0143, EEA rebuttal evidence, paragraph A92, PDF page 65. Exhibit 20633-X0147, UCA argument, paragraph 63, PDF page 16. Exhibit 20633-X0150, UCA reply argument, paragraph 90, PDF page 23. 6 Decision 20633-D01-2016 (December 20, 2016)

37. The Commission accepts EEA s explanation of the differences between EEA personnel and EUI corporate services personnel for the purposes of determining a vacancy rate. The Commission agrees that these differences suggest that the global vacancy rate may not be directly applicable to EEA s corporate services costs. However, EEA has not provided any evidence that would allow the Commission to determine a non-zero vacancy rate for EUI corporate services personnel. 38. In the absence of evidence to support a vacancy rate that should apply to EUI s corporate services allocated to EEA, the Commission considers that the global vacancy rate of 4.5 per cent should be applied to the EUI corporate services costs allocated to EEA. Accordingly, the Commission directs EEA to include a 4.5 per cent vacancy rate for EUI s corporate services costs allocated to EEA in the compliance filing. 4.2 Salary escalations 39. Consistent with previous applications, EUI engaged Towers Watson to prepare a report on the competitiveness of EEA s, EDTI s and EUI s unionized and non-unionized positions and to recommend the level of escalation for EEA s non-union salaries for 2016 and 2017. For 2016 and 2017, Towers Watson recommended a range of escalation for non-union salaries of three per cent to four per cent. 11 40. According to EEA, the lower end of the range recommended by Towers Watson is generally consistent with factors applied prior to the decline in oil prices and the higher end reflects potential increases if oil prices return to historical norms. The recommendation was based on projections that the Canadian economy is expected to rebound by the end of 2015 and return to full productivity in 2016. 12 41. Based on Towers Watson s analysis, EEA originally forecast escalation for non-union salaries at four per cent for 2016 and 2017, which represents growth in Alberta with oil prices returning to historical levels. However, in its rebuttal evidence, EEA revised this forecast down to three per cent, which EEA stated aligns with the three per cent escalator approved for EDTI in Decision 3539-D01-2015. 13 Views of the parties 42. In evidence submitted on behalf of the UCA, Ms. Radway cited the Towers Watson memorandum, in which Towers Watson stated that the projected range is highly sensitive to the economic recovery assumptions noted above and the range may decrease if oil prices and overall economic growth do not increase as anticipated. 14 43. Ms. Radway stated that Towers Watson s recommendation was predicated on oil prices returning to historical norms, which has not happened, and that the low end of the Towers Watson range may be too high. Ms. Radway further stated that the Economic Outlook for the 11 12 13 14 Exhibit 20633-X0003, application, paragraph 128, PDF page 47. Exhibit 20633-X0003, application, paragraph 129, PDF page 47. Decision 3539-D01-2015: EPCOR Distribution & Transmission Inc., 2015-2017 Transmission Facility Owner Tariff, Proceeding 3539, Application 1611027-1, October 21, 2015. Exhibit 20633-X0130, evidence of Shelley Radway, paragraph A26, PDF page 18. Decision 20633-D01-2016 (December 20, 2016) 7

Government of Alberta s Budget 2015 15 shows that for 2016-2017, weekly earnings will increase by 2.4 per cent. 44. Ms. Radway noted that, for the last three years, EEA has awarded salary escalators 0.3 per cent lower than the lowest number of the range recommended by Towers Watson. As a result, Ms. Radway proposed a salary escalation rate of 2.7 per cent. 16 45. The UCA argued that a three-year average is commonly used to test the reliability of forecasts and this time frame is consistent with EEA s own forecasting methodology for vacancy rates. 17 46. EEA stated that only the high end of the escalator range recommended by Towers Watson was predicated on oil prices returning to historical norms. EEA also stated that it is important to note that there is not a direct relationship between oil prices and EEA s salary escalation forecasts. In its report, Towers Watson considered the effect of the decline in oil prices and observed a more moderate approach taken within the utility sector relative to the oil and gas sector. 18 47. EEA also stated that the weekly earnings forecast in the Economic Outlook for the Government of Alberta s Budget 2015 is not a reasonable comparator to EEA s non-union salary escalation. The weekly earnings indicator is the gross taxable payroll divided by the number of employees for all employees across Alberta. The weekly earnings includes overtime that is not applicable to EEA s non-union employees and part-time hours, which will drive down the weekly earnings. It also includes all industries, such as oil and gas companies, which are not included in EEA s comparator companies. 19 48. EEA examined six years of data from 2010 to 2015 to show that EEA has, on average, awarded a salary escalation 0.1 per cent above the low end of the ranges recommended by Towers Watson. 20 Commission findings 49. Table 3 below illustrates the approved non-union salary escalators and the actual escalators for the years 2004 to 2015, as provided by EEA in its application and IR responses: 21 15 16 17 18 19 20 21 Government of Alberta, Budget 2015, Fiscal Plan 2015-18, Economic Outlook, October 27, 2015. Exhibit 20633-X0130, evidence of Shelley Radway, paragraphs A27 and A28, PDF page 18. Exhibit 20633-X0150, UCA reply argument, paragraph 70, PDF page 19. Exhibit 20633-X0143, EEA rebuttal evidence, paragraph A65, PDF page 47. Exhibit 20633-X0143, EEA rebuttal evidence, paragraph A66, PDF page 48. Exhibit 20633-X0143, EEA rebuttal evidence, paragraph A67, PDF page 49. Exhibit 20663-X0003, application, Table 1.6.1.3-1, PDF page 48; Exhibit 20633-X0062, information responses, Table EEA-AUC-2015OCT07-007-1, PDF page 20. 8 Decision 20633-D01-2016 (December 20, 2016)

Table 3. Historical approved and actual non-union salary escalators AUC/EUB (1) approved (%) EEA actual 2004 3.4 5.0 2005 4.5 7.6 2006 4.0 8.0 2007 5.0 4.6 2008 5.0 4.1 2009 5.0 2.6 2010 3.5 3.5 2011 3.5 3.2 2012 4.0 3.9 2013 4.0 3.4 2014 4.0 3.3 2015 4.5 3.0 Note 1: EUB Alberta Energy and Utilities Board. 50. As shown in the table above, from 2007 onwards, with the exception of 2010, EEA has consistently implemented actual non-union salary escalators lower than those approved by the Commission and its predecessor, the Alberta Energy and Utilities Board (the board). These data also show a declining trend in actual salary escalators awarded by EEA from 2012 to 2015. 51. In its memorandum on non-union salary escalation factors, Towers Watson stated that the decline in 2015 salary budgets and the continued economic uncertainty make it difficult to provide a firm estimate of 2016 and 2017 salary escalators. Towers Watson further stated that it anticipates 2016 salary escalators will increase from depressed 2015 levels because of a positive outlook for 2016. Towers Watson asserted that the salary escalator range it projected is highly sensitive to Towers Watson s economic recovery assumptions and that the range may decrease if oil prices and overall economic growth do not increase as anticipated. 22 52. Given Towers Watson s stated uncertainty with regard to its projections, the positive outlook that those projections are predicated on and the current economic climate in Alberta, the Commission finds that the three to four per cent range recommended by Towers Watson is no longer reflective of salary escalators, given the current climate in Alberta, and does not constitute a reasonable salary escalator to be applied to the test years. 53. For the reasons discussed above, the Commission accepts the UCA s proposed 2.7 per cent non-union salary escalator for 2016 and 2017, and directs EEA to update its non-union salary escalators to 2.7 per cent for both 2016 and 2017 in the compliance filing. 4.3 Operating costs 4.3.1 Customer service costs 54. EEA applied for customer service costs of $21.69 million and $21.97 million for 2016 and 2017, respectively. Those totals were composed of client services costs of $7.78 million for 2016 and $7.72 million for 2017; technical training and communications costs of $1.11 million 22 Exhibit 20633-X0022, Appendix I, PDF page 50. Decision 20633-D01-2016 (December 20, 2016) 9

for 2016 and $1.13 million for 2017; and information services costs of $4.86 million for 2016 and $4.82 million for 2017. 23 Views of the parties 55. In her evidence, filed on behalf of the UCA, Ms. Radway stated that EEA has a history of over-forecasting client services, technical training and communications, and information services for 2012, 2013 and 2014. 56. Ms. Radway used the data provided by EEA to show the over-forecast percentage in 2012, 2013 and 2014 for each of the three customer service cost areas and provided an average over these three years. The variances are reproduced in Table 4, below: Table 4. Variance between forecast and actual customer service costs Year 2012 2013 2014 Average Client services 11.2 14.6 9.7 11.8 Technical training and communications 11.5 13.4 11.4 12.1 Information services 9.5 11.0 8.6 9.7 Source: Exhibit 20633-X0130, UCA evidence of Ms. Radway, Table 1, PDF page 6. 57. Ms. Radway commented that the pattern continues for 2015, but that 2015 was not included in the calculations because updated forecast data were only available for the first quarter of 2015. 58. Ms. Radway noted that while the individual subcategories that comprise each of the cost categories within customer services costs may not show evidence of over-forecasting, this is not the case when looking at the total categories where annual over-forecasting becomes readily apparent. If the sum of the individual forecasts consistently demonstrates over-forecasting, then Ms. Radway stated that the total must be adjusted for the test years. 59. For client services, Ms. Radway acknowledged that site count variance needed to be accounted for in an analysis of the historical variances. Ms. Radway adjusted for the site count variance by dividing the forecast and actual client services costs by the forecast and actual site counts, respectively. After adjusting for site counts, Ms. Radway s result showed an average variance of 10.0 per cent over the years 2012 through 2014. Ms. Radway did not propose a site count adjustment specifically for technical training and communications and information services. 24 60. Based on her analysis, Ms. Radway recommended reducing EEA s 2016 and 2017 forecasts for client services by 10.0 per cent, for technical training and communications by 12.1 per cent, and for information services by 9.7 per cent. 25 (%) 23 24 25 Exhibit 20663-X0003, application, Table 3.1-1, PDF page 60. Exhibit 20633-X0130, evidence of Shelley Radway, paragraphs A10 and A11, PDF page 7. Exhibit 20633-X0130, evidence of Shelley Radway, paragraph A15, PDF page 10. 10 Decision 20633-D01-2016 (December 20, 2016)

61. Ms. Radway referenced three instances in which the Commission reduced forecast costs for test years based on a history of under- or over-forecasting. In Decision 2013-407, 26 the Commission reduced AltaLink s general operating and maintenance expense; in Decision 2014-303, the Commission reduced EEA s utility affiliates site counts; and in Decision 3577- D01-2016, 27 the Commission reduced ATCO Pipelines costs for capital projects. 28 62. In rebuttal evidence, EEA submitted that Ms. Radway s analysis is flawed and misleading, and it does not address the detailed reasons provided by EEA for historical variances from forecast. As such, EEA stated that Ms. Radway s recommendations should be rejected for the following reasons: 29 (a) For 2012 and 2013, Ms. Radway uses EEA s forecast amounts rather than decision amounts, which included reductions to EEA s forecasts. For example, in Decision 2013-110, 30 the Commission made several determinations reducing EEA s customer service costs. 31 (b) Ms. Radway ignores organizational transfers of positions within EEA that reflect a shifting of costs between categories. For example, the transfer of webmail costs and the customer relations manager from client services to billing services resulted in decreased client services costs of $0.29 million in 2013 and $0.10 million in 2014, with corresponding increases in billing services. 32 (c) Ms. Radway s approach double counts the impact of historical variances because EEA s forecasts incorporate historical results. EEA provided its global vacancy rate as an example of how historical variances are incorporated into forecasts. (d) Ms. Radway ignores the underlying causes of variances and does not account for unusual or one-time events, such as EEA s contact centre consolidation project and higher than anticipated attrition in 2014, that will not affect the current test period. 33 (e) Ms. Radway does not provide any reasons for the expectation that historical variances will occur again. The occurrence of a variance does not mean that the variance will occur again. (f) Ms. Radway has not demonstrated any flaws in EEA s forecasting methodology, or that her recommendation will result in a more accurate forecast. 26 27 28 29 30 31 32 33 Decision 2013-407: AltaLink Management Ltd., 2013-2014 General Tariff Application, Proceeding 2044, Application 1608711-1, November 12, 2013. Decision 3577-D01-2016: ATCO Pipelines, 2015-2016 General Rate Application, Proceeding 3577, Application 1611077-1, February 29, 2016. Exhibit 20633-X0130, evidence of Shelley Radway, paragraph A14, PDF page 9. Exhibit 20633-X0143, EEA rebuttal evidence, paragraph A49, PDF page 31. Decision 2013-110: EPCOR Energy Alberta Inc., 2012-2013 Regulated Rate Tariffs, Proceeding 1872, Application 1608427-1, March 21, 2013. Exhibit 20633-X0143, EEA rebuttal evidence, paragraph A52, PDF page 34. Exhibit 20633-X0143, EEA rebuttal evidence, paragraph A54, PDF page 36. Exhibit 20633-X0143, EEA rebuttal evidence, paragraph A49, PDF page 32. Decision 20633-D01-2016 (December 20, 2016) 11

(g) Ms. Radway s approach is inconsistent with forecast cost-of-service ratemaking, which allows the applicant to keep the benefits of improvements during the test years and then pass those benefits on to customers in future test periods. 34 63. The UCA argued that the cost transfers from client services to billing services were already accounted for in the costs provided in EEA s application, that EEA continues to overforecast even after accounting for historical variances in forecasts and that the onus is not on Ms. Radway to provide an explanation for why historical variances will reoccur. 35 The UCA argued further that Ms. Radway s approach is consistent with cost-of-service regulation since it ensures that the resulting savings are passed on to customers in subsequent test periods. 36 Commission findings 64. The Commission acknowledges the conflicting positions of EEA and the UCA with regard to the reasonableness of EEA s customer service cost forecasts. The Commission has taken these viewpoints into account and considers that the issue of over-forecasting is adequately addressed in Section 4.6 below and additional findings are not required in this section of the decision. 4.3.2 Late payment charges 65. In Decision 2014-303, the Commission approved EEA s proposed trend methodology for forecasting late payment charges, but provided the following direction: 193. While the Commission accepts EEA s proposed method to forecast late payment charges, the Commission does have concerns regarding the applicability of this method for forecasting and whether it is an adequate measure from which to forecast late payment charges. The Commission, therefore, directs EEA to include in its next nonenergy application, in addition to the method approved in this decision, an analysis of alternative methods of forecasting late payment charges. The methods examined should include, but not necessarily be limited to, a trend line to historical actual late payment charges methodology and a two-step methodology based on: (i) the percentage of bills, out of the total number of bills issued, to which late payment fees have been charged; and (ii) the amount or portion the late payment charges represent on those bills. 66. For the current application, EEA undertook a review of its late payment charges forecasting method and evaluated 20 different alternatives using historical data from 2006 to 2014 in order to evaluate the percentage error for each alternative. 67. EEA found that the two-step methodology suggested by the Commission was not reliable because there is not a one-to-one ratio between sites and customer accounts. The same was found in EEA s analysis of a methodology based on the percentage of billed sites by rate class. 68. EEA also reviewed a trend methodology, as approved in Decision 2014-303, three different historical average methods (three-year average, four-year average, and historical simple average using data from 2006 to 2014), and a weighted average of the three most recent years, which was previously used by EEA for bad debt forecasting. Each of these five methodologies 34 35 36 Exhibit 20633-X0143, EEA rebuttal evidence, paragraph A49, PDF page 31. Exhibit 20633-X0147, UCA argument, paragraphs 4-9, PDF page 4. Exhibit 20633-X0150, UCA reply argument, paragraph 20, PDF page 7. 12 Decision 20633-D01-2016 (December 20, 2016)

was investigated using monthly and annual data, and using billed revenues and settled revenues for a total of 20 alternatives. 69. EEA s analysis showed that the alternatives using monthly data and billed revenues produced lower average percentage errors than the alternatives using annual data and settled revenues. For the EDTI service area, the four-year average produced the lowest percentage error, while for the FAI service area, the historical simple average produced the lowest average percentage error. 70. EEA then performed the same analysis with the data for the EDTI and FAI service areas combined. In this analysis, the historical simple average using monthly data produced the lowest average percentage error. 71. Based on the results of the analyses, EEA proposed to change its late payment charge forecasting methodology to a historical simple average method calculated using monthly data. 37 As a check on the reasonableness of this proposal, EEA compared the 2016 and 2017 forecasts using this methodology to forecasts prepared using the most accurate methodology for each service area. 72. For the EDTI service area, 2016 and 2017 forecasts using the proposed historical simple average were $3.50 million and $3.73 million, respectively. Forecasts using a four-year average were $3.48 million and $3.71 million, respectively, demonstrating a variance of only $0.02 million for each test year. 38 Commission findings 73. The Commission considers that EEA has provided a sufficient review of alternative late payment charge forecasting methodologies and accepts the results of EEA s analyses. 74. Accordingly, the Commission approves EEA s late payment charge forecasting methodology using a historical simple average method calculated using monthly data, as proposed in the application. The Commission considers that in providing the various alternatives for calculating forecasts of late payment penalties, EEA has complied with the Commission s direction at paragraph 193 of Decision 2014-303. 4.4 Tax and deferral accounts 75. EEA applied for the continuation of a hearing cost reserve account, and the continuation of a deferral account for short-term incentive costs. 39 Following the guidance of the Commission in Decision 2013-110, EEA applied to include, as part of the hearing cost reserve account, any incremental depreciation and debt costs related to regulatory projects driven by regulatory initiatives that occur during 2016 and 2017. EEA indicated that it has not forecast any costs of this type for 2016 and 2017. 40 76. No comments were received on these two accounts from the CCA or the UCA. 37 38 39 40 Exhibit 20633-X0003, application, paragraph 544, PDF page 169. Exhibit 20633-X0003, application, paragraph 546, PDF page 170. Exhibit 20633-X0003, application, paragraph 1191, PDF page 413. Exhibit 20633-X0003, application, paragraph 1193, PDF pages 413-414. Decision 20633-D01-2016 (December 20, 2016) 13

Commission findings 77. As noted by EEA in the application, 41 the hearing cost reserve account and the short-term incentive costs deferral account were established in previous years and have been in existence for a number of years. The addition of the incremental depreciation and debt costs related to regulatory projects was reasonably the result of a Commission consideration in Decision 2013-110. There is no reason provided on the record to change the treatment of these two accounts during 2016 and 2017 and, therefore, the Commission approves their continued use in 2016 and 2017. 4.4.1 Bad debt deferral 78. Referring to the bad debt risk, site counts risk and cost risk underlying EEA s request for non-energy risk compensation, as described further in Section 4.5 of this decision, the Commission asked EEA if deferral accounts related to any of these risks would mitigate risk to customers as well as to EEA. The Commission also asked EEA to explain why it had not requested deferral account treatment for site counts, bad debts and all remaining costs, but instead had requested non-energy risk compensation. 42 79. In response, EEA indicated that the use of a deferral account would be a way to mitigate these three risks to customers and EEA. EEA added that these three risks are not an exhaustive list of all the business risks that EEA faces. Using the criteria identified by the Commission in Decision 2010-189 43 for the establishment of deferral accounts, EEA provided an evaluation of whether site counts risk, bad debt risk and cost risk met these criteria. EEA indicated that only bad debt risk satisfies all of the Commission s criteria for deferral account treatment. 44 EEA stated that based on the Commission s denial of a bad debt deferral account in Decision 2014-303, it did not request deferral account treatment for bad debts as part of this application. 45 80. In the same response, EEA proposed the same bad debt deferral account mechanism as it proposed in its 2014-2015 non-energy RRT application. 46 This deferral account mechanism subjects 50 per cent of the variance between the actual and forecast bad debt expenses to deferral, with EEA retaining the remaining 50 per cent of the variance. EEA added that if this deferral account is provided, then its non-energy risk compensation would be reduced from 0.50 per cent to 0.42 per cent. 47 81. The CCA submitted that, based on a historical analysis of actual and forecast cost levels, the approval of a bad debt deferral account would decrease the risk of negative outcomes for EEA and increase the positive risk to EEA. The CCA added that the negative outcomes of bad debt would no longer be dragging down the net positive effect of variances in site counts and other costs. 48 41 42 43 44 45 46 47 48 Exhibit 20633-X0003, application, paragraph 1191, PDF page 413. Exhibit 20633-X0108, EEA-AUC-2016JAN19-001(b) and (c), PDF pages 1-2. Decision 2010-189: ATCO Utilities Pension Common Matters, Proceeding 226, Application 1605254-1, April 30, 2010. Exhibit 20633-X0114, information responses, EEA-AUC-2016JAN28-001(b), PDF pages 4-7. Exhibit 20633-X0114, information responses, EEA-AUC-2016JAN28-001(c), PDF page 9. Proceeding 2986. Exhibit 20633-X0114, information responses, EEA-AUC-2016JAN28-001(c), PDF page 8. Exhibit 20633-X0129, evidence of Jan Thygesen, paragraph 24, PDF page 10. 14 Decision 20633-D01-2016 (December 20, 2016)

82. The UCA described two concerns with the introduction of a bad debt deferral account. The first concern is the reduction of the incentive to control bad debt levels, and the second is that EEA will have less risk compared to other regulated rate option (RRO) providers. 49 The UCA submitted that EEA has updated its methodology for forecasting bad debt expense in each of the last three RRT applications, and these refinements have reduced the forecast errors. The UCA stated that EEA is able to forecast its bad debt expense. 50 83. The UCA indicated that over the last five years, EEA has the least amount of forecast error for bad debt expense compared to the other RRO providers; namely, Direct Energy Regulated Services (DERS) and ENMAX Energy Corporation (EEC). 51 The UCA stated that DERS and EEC do not have bad debt deferral accounts. It added that if EEA has a bad debt deferral account, which reduces EEA s forecast risk compared to DERS and EEC, and EEA continues to obtain a return compensation figure that is equivalent to the return compensation figures for DERS and EEC, then EEA will be overcompensated. 52 84. The UCA recommended that the request for a bad debt deferral account be denied. 53 85. In its rebuttal evidence, 54 as well as its written argument, 55 EEA stated that it was not amending its application to apply for a bad debt deferral account. EEA included the following clarification about its response to the Commission s IR regarding a deferral account for bad debt expense: To clarify, in EEA-AUC-2016JAN28-001(b), EEA was asked whether a deferral account would mitigate bad debt risk (amongst other risks) and, if so, how the deferral account would function. In response to this question, EEA explained how it would propose that such a deferral account would function. EEA was not, however, amending its Application to apply for a bad debt deferral account. 56 Commission findings 86. Before EEA submitted its clarification to the response to the Commission s IR EEA- AUC-2016JAN28-001(b), the Commission s understanding was the same as that of both the UCA and the CCA, in that, EEA was amending its application and proposing a bad debt deferral account. However, based on the clarification in EEA s rebuttal evidence and written argument, it is clear to the Commission that EEA is not applying for a bad debt deferral account. The Commission considers that no finding is required because EEA is not proposing a bad debt deferral account. 49 50 51 52 53 54 55 56 Exhibit 20633-X0130, evidence of Shelley Radway, paragraph A37, PDF page 25. Exhibit 20633-X0130, evidence of Shelley Radway, paragraph A37, PDF page 26. Exhibit 20633-X0130, evidence of Shelley Radway, paragraph A37, PDF page 26. Exhibit 20633-X0130, evidence of Shelley Radway, paragraph A37, PDF page 27. Exhibit 20633-X0130, evidence of Shelley Radway, paragraph A38, PDF page 27. Exhibit 20633-X0143, EEA rebuttal evidence, PDF page 59. Exhibit 20633-X0149, EEA written argument, paragraph 133, PDF page 69. Exhibit 20633-X0143, EEA rebuttal evidence, PDF page 59. Decision 20633-D01-2016 (December 20, 2016) 15

4.5 Depreciation and return 4.5.1 Weighted average cost of capital 87. EEA did not request any return on equity with respect to its rate base through its nonenergy rates for 2016 and 2017. 57 However, EEA indicated that it still requires an approved forecast for its weighted average cost of capital (WACC). This WACC figure is used for calculating the return on capital component for purposes of determining corporate asset usage fees. 58 EEA calculated a WACC for 2016 and 2017 of 6.01 per cent, based on a debt to equity ratio of 61/39, a return on equity figure of 8.30 per cent, and a cost of debt figure of 4.55 per cent. 59 The 61/39 debt to equity ratio is a change from previous years, for which EEA used a capital structure of 100 per cent debt to calculate WACC. 60 88. EEA referred to Decision 2014-303 in which the Commission accepted that EEA operates as a stand-alone entity, 61 and a subsequent direction from that decision for EEA to explore its capital structure as part of the current application, and to explain the proposed change in the debt to equity ratio. 89. EEA proposed the 61/39 debt to equity ratio based on Decision 2009-216, 62 as this was the last time the Commission approved EEA s capital structure prior to the 100 per cent deemed debt for EEA. 63 90. Referring to Decision 2191-D01-2015, 64 the 2013 Generic Cost of Capital (GCOC) decision, the Commission asked EEA in an IR if the one per cent reduction to the equity ratios approved for all the utilities involved in that proceeding should be reflected in EEA s proposed debt to equity ratio. In response, EEA considered that the one per cent reduction should be applied, and EEA proposed to incorporate such a change in the compliance filing if the Commission agreed to the change. EEA advised that the effect of such a change would be an increase in the revenue requirement for 2016 and 2017 of approximately $10,000 in each year. 65 Commission findings 91. The Commission issued Decision 20622-D01-2016 66 on the 2016 GCOC proceeding on October 7, 2016. The awarded return on equity and capital structures from that proceeding are applicable for the years 2016 and 2017. The Commission considers that the figures from the decision on the 2016 GCOC proceeding are the most up to date and should be incorporated into EEA s 2016 and 2017 RRT, because these figures will apply to the same test years. Therefore, the Commission directs EEA to incorporate any applicable results from Decision 20622-D01-2016, into its compliance filing. 57 58 59 60 61 62 63 64 65 66 Exhibit 20633-X0003, application, paragraph 1325, PDF page 452. Exhibit 20633-X0003, application, paragraph 1135, PDF page 370. (0.39*.083) + (0.61*.0455). Exhibit 20633-X0003, application, paragraph 1325, PDF page 452. Decision 2014-303, paragraph 230, PDF page 55. Decision 2009-216: 2009 Generic Cost of Capital, Proceeding 85, Application 1578571-1, November 12, 2009. Exhibit 20633-X0007, Appendix D, paragraph 17, PDF page 7. Decision 2191-D01-2015: 2013 Generic Cost of Capital, Proceeding 2191, Application 1608918-1, March 23, 2015. Exhibit 20633-X0062, information responses, EEA-AUC-2015OCT07-035(b), PDF page 100. Decision 20622-D01-2016: 2016 Generic Cost of Capital, Proceeding 20622, October 7, 2016. 16 Decision 20633-D01-2016 (December 20, 2016)

4.5.2 Interest during construction and allowance for funds used during construction 92. EEA currently maintains two concurrent calculations of capitalized interest. 67 93. For regulatory purposes, EEA currently capitalizes interest charges for capital projects using the allowance for funds used during construction (AFUDC) methodology. The AFUDC methodology involves calculating the capitalized interest by applying EEA s WACC to the midyear balance of capital work-in-progress. 68 94. For external financial reporting purposes, EEA calculates capitalized interest using the interest during construction (IDC) methodology, in accordance with the requirements of the International Financial Reporting Standards. The IDC methodology uses EEA s cost of debt rate multiplied by capital work-in-progress for any capital project exceeding both a minimum capital threshold of $0.5 million and a duration of six months. 69 95. As part of its application, EEA proposed to adopt IDC for regulatory purposes. EEA stated this would be administratively more efficient because it would result in EEA maintaining a single set of accounting records. EEA submitted that this change would not have a material effect on its revenue requirement, because it has minimal capital assets. 70 It provided information showing the effect on revenue requirement resulting from the use of IDC compared to AFUDC. For the years 2010 to 2015, the combined revenue requirements would have decreased by approximately $10,000 if IDC had been used instead of AFUDC. For the 2016-2017 forecast years included in the application, the combined revenue requirements are approximately $10,000 greater as a result of using IDC instead of AFUDC. 71 96. The UCA recommended that there be no change for the accounting treatment for AFUDC and that only the interest on the debt portion of the capital structure be included in costs. The UCA recommended further that the AFUDC for corporate asset usage fees only include the interest costs. 72 97. The UCA indicated that EEA has adopted a 61/39 debt to equity ratio as its capital structure and as such, EEA collects interest costs for debt for its working capital and reserve activity based on this structure. However, as the UCA added, EEA recovers its WACC of 6.01 per cent through its return for corporate asset usage fees. 73 98. The UCA submitted that since the return compensation determined for EEA in Decision 2941-D01-2015 74 is considered all inclusive, there should be no return on equity for either AFUDC or the return portion of the corporate asset usage fees. 75 The UCA submitted that, in 67 68 69 70 71 72 73 74 75 Exhibit 20633-X0003, application, paragraph 1302, PDF page 445. Exhibit 20633-X0003, application, paragraph 1301, PDF page 445. Exhibit 20633-X0003, application, paragraph 1301, PDF page 445. Exhibit 20633-X0003, application, paragraph 1303, PDF page 445. Exhibit 20633-X0003, application, tables 7.1-1 and 7.1-2, PDF page 446. Exhibit 20633-X0130, evidence of Shelley Radway, PDF page 24. Exhibit 20633-X0130, evidence of Shelley Radway, PDF page 21. Decision 2941-D01-2015: Direct Energy Regulated Services, ENMAX Energy Corporation and EPCOR Energy Alberta GP Inc., Regulated Rate Tariff and Energy Price Setting Plans Generic Proceeding: Part B Final Decision, Proceeding 2941, Application 1610120-1, March 10, 2015. Exhibit 20633-X0130, evidence of Shelley Radway, paragraph A33, PDF pages 22-23. Decision 20633-D01-2016 (December 20, 2016) 17