Decision FortisAlberta Inc Phase II Distribution Tariff. January 27, 2014

Transcription

1 Decision FortisAlberta Inc Phase II Distribution Tariff January 27, 2014

2 The Alberta Utilities Commission Decision : FortisAlberta Inc Phase II Distribution Tariff Application No Proceeding ID No January 27, 2014 Published by The Alberta Utilities Commission Fifth Avenue Place, Fourth Floor, 425 First Street S.W. Calgary, Alberta T2P 3L8 Telephone: Fax: Website:

3 Contents 1 Introduction Details of the application Responses to AUC directions Distribution related Transmission related AUC directions from Decision PBR implications PBR framework and Phase II applications Allocation of K and Y factor amounts Rate design and cost allocation overview Local costs multiplier for Rate Cost allocation Distribution cost allocation AUC assessment fees Load settlement cost allocation Transmission cost allocation Load forecast (billing determinants) Revenue-to-cost ratios Transmission rate design Recovery of bulk transmission charges Distribution rate design Rate 11 - residential service Rate 21 - farm service Rate 26 - irrigation service Rates 31, 33 and 38 - lighting service Rate 41 - small general service Rate 45 - oil & gas service Rate 61 - general service Rate 63 - large general service Distribution bill (rate) impacts Distribution adjustment rider Review of Fortis Rate 44/45 charges Background of the transition of customers from Rate 44 to Rate Proceeding ID No and AUC Decision Fortis response to AUC directions from Decision Rate 44/ Interpretation of rate schedules TransAlta proceeding AUC Decision (January 27, 2014) i

4 Billing determinants Revenue impacts Relief approved Payment in lieu of notice Billing adjustment EQUS/NPP jurisdictional issue Incorporating the 2012 Phase II results in 2012, 2013 and 2014 rates Finalizing 2012 and 2013 distribution rates Establishing distribution rates for 2014 and beyond Terms and conditions of service AMI meters Order Appendix 1 Proceeding participants Appendix 2 Summary of Commission directions Appendix 3 Rate 44 and Rate 45 rate schedules List of tables Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Table 7. Table 8. Table 9. Table 10. Table 11. Table 12. Table 13. Allocators for Y factor and K factor amounts... 8 Restructured rates initially applied for Amendments to application Sample feeders compared to total distribution system Fortis response to IR AUC-FAI Comparison of Fortis 2010 and 2012 distribution cost studies Impact of the 2012 enhancements Fortis allocation of distribution costs Fortis 2013 transmission cost forecast Commission summary of transmission cost allocation method for distribution connected load Revenue-to-cost ratio changes Commission summary of 2013 forecast AESO charges to Fortis Impacts of energy as billing determinant ii AUC Decision (January 27, 2014)

5 Table 14. Table 15. Table 16. Table 17. Table 18. Table 19. Table 20. Table 21. Table 22. Impacts of demand as billing determinant Illustration of residential rate and rate structure changes Proposed farm rate structure as revised March 4, Illustration of irrigation rate and rate structure changes Illustration of small general rate and rate structure changes Illustration of oil & gas rate and rate structure changes Illustration of large general rate and rate structure changes Number of accounts moved from Rate 44 to Rate Change in 2012 revenue if the 2012 Phase II results had been reflected in the 2012 rates DRAFT AUC Decision (January 27, 2014) iii

6

7 The Alberta Utilities Commission Calgary, Alberta Decision FortisAlberta Inc. Application No Phase II Distribution Tariff Proceeding ID No Introduction 1. On January 18, 2013, FortisAlberta Inc. (Fortis or FAI) filed an application with the Alberta Utilities Commission (AUC or Commission) requesting approval of its Phase II distribution tariff application for the purposes of finalizing rates for 2012, and to establish performance-based regulation (PBR) rates for 2013 and 2014 (application), subject to any PBR proceedings in progress. 2. Notice of the application (notice) was published in the four major Alberta newspapers (the Edmonton Journal, the Edmonton Sun, the Calgary Herald, and the Calgary Sun) on Tuesday January 29, 2013, and was distributed by on January 22, 2013 to parties on the Commission s interested party distribution list. In addition, the notice was posted on the Commission s website on January 22, In response to the notice, the following parties registered to participate in this proceeding: ATCO Electric Inc. (ATCO Electric) Canada West Ski Areas Association (CWSAA) The Consumers Coalition of Alberta (CCA) Desiderata Energy Consulting Inc. (DESI) EPCOR Distribution & Transmission Inc. (EDTI) Equs Rural Electrification Association Ltd. (EQUS) The Industrial Power Consumers Association of Alberta (IPCAA) North Parkland Power Rural Electrification Association Ltd. (NPPREA) The Office of the Utilities Consumer Advocate (UCA) 3. On March 4, 2013, Fortis filed revisions to its proposed residential and farm rates and provided an updated rate design, bill impacts and rate schedules for approval. 1 A black lined version of Fortis rate schedules was later provided in attachment UCA-FAI Exhibit 30. Exhibit AUC Decision (January 27, 2014) 1

8 4. The Commission established the following process to test the application: Process step Due date Information requests (IRs) to Fortis round 1 March 7, 2013 IR responses round 1 April 5, 2013 Intervener evidence May 10, 2013 IRs on intervener evidence May 22, 2013 IR responses on intervener evidence May 31, 2013 Rebuttal evidence June 7, 2013 IRs to Fortis round 2 June 28, 2013 IR responses round 2 July 8, 2013 Argument July 17, 2013 Reply argument July 29, 2013 IR to Fortis round 3 October 15, 2013 IR responses round 3 October 22, 2013 Submissions from parities October 29, On June 11, 2013, the Commission received a letter, dated May 25, 2013, from the Potato Growers of Alberta (PGA). The letter from the PGA outlined concerns that the PGA had with respect to increases in rates to its growers. 3 For this proceeding, the Commission considered the letter as the PGA s statement of intent to participate, as well as intervener evidence. The Commission revised the schedule for the proceeding to allow parties to file evidence in reply to the PGA. Fortis submitted a letter dated June 18, 2013, indicating that it had in its rebuttal evidence already addressed rate mitigation measures for irrigation Rate 26 and therefore had no further reply to the PGA By letter dated July 30, 2013, the CCA indicated that it had inadvertently left out some material in its reply argument, and uploaded a corrected version of its reply argument. 7. The Commission issued a third round of IRs to Fortis on October 15, 2013 and received responses on October 22, For the purposes of this decision, the Commission considers that the record of this proceeding closed on October 29, In reaching the determinations contained within this decision, the Commission has considered all relevant materials comprising the record of this proceeding, including the evidence and argument provided by each party. 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. 3 4 Exhibit Exhibit AUC Decision (January 27, 2014)

9 2 Details of the application 10. Fortis stated that this application was unique from previous Phase II applications in that, rather than just setting rates by rate class for a single cost of service test year, i.e., 2012, Fortis must also make proposals for setting rates for 2013 and Fortis submitted that its Phase II cost allocation and rate design in the application incorporated elements of PBR as approved in Decision Proposals in Fortis application recognized that certain determinations on Phase I parameters in related PBR proceedings were outstanding 6 at the time of its application, but were expected to be resolved in As such, Fortis incorporated in its application placeholders equal to the amounts proposed in the PBR proceedings. Once these parameters are determined, Fortis expects to make the required adjustments and true-ups to this application for the purposes of setting rates by rate class for 2013 and With respect to cost allocation, Fortis stated that its distribution tariff is comprised of two distinct components: the transmission and distribution components. Costs are allocated to rate classes based on distinct transmission and distribution cost drivers and characteristics. Fortis submitted that the methods used to allocate costs are largely unchanged from those used in previous studies. 9 The majority of distribution costs in Fortis cost allocation study are allocated based on its component allocation method (CAM) model. The CAM model was introduced in 2002 and has since been refined. The various versions of the model were approved by the Commission in subsequent applications With respect to rate design, Fortis stated that its greatest concern in entering a PBR framework was transitioning the overall rate levels to 100 per cent revenue-to-cost ratios by rate class as soon as possible. As such, in the application, Fortis proposed to implement new rate design structures to achieve 100 per cent revenue-to-cost ratios by rate class in With respect to Fortis 2012 rates by rate class, Fortis submitted that it is not proposing to alter its rates from those that were in effect on an interim basis in Fortis requested the Decision : Rate Regulation Initiative, Distribution Performance-Based Regulation, Application No , Proceeding ID No. 566, September 12, PBR related proceedings outstanding at the time of Fortis application included: Proceeding ID No. 2130: the PBR Compliance Filing for 2013 rates in response to Decision , Proceeding ID No. 2131: the PBR Capital Tracker Filing, and Proceeding ID No. 2240: the Review and Variance (R&V) Application of Decision At the time when this decision was issued the Commission had ruled on: Proceeding ID No. 2130, the result was Decision : 2012 Performance-Based Regulation Compliance Filings, AltaGas Utilities Inc., ATCO Electric Ltd., ATCO Gas and Pipelines Ltd.,, EPCOR Distribution & Transmission Inc. and FortisAlberta Inc., Application No , Proceeding ID No. 2130, March 4, 2013, which set interim rates for 2013 and required a second compliance filing which is Proceeding ID No. 2477, Proceeding ID No. 2240, the result was Decision : Rate Regulation Initiative, Distribution Performance-Based Regulation, Decision on Preliminary Question, Requests for Review and Variance of AUC Decision , Application Nos , , , , and , Proceeding ID No. 2240, March 4, 2013 where the Commission denied the R&V of Decision , and this has now become a court matter, lastly Proceeding ID No was still on-going. Exhibit 1, paragraphs 2 and 3. Exhibit 1, paragraph 5. Exhibit 1, paragraph 8. Exhibit 1, paragraph 10. Exhibit 1, paragraph 16. AUC Decision (January 27, 2014) 3

10 Commission approve, on a final basis, the 2012 distribution tariff rates that were in effect on an interim basis from January 1, 2012 to December 21, Specifically, Fortis applied for the following approvals: 13 (a) Declare that the 2012 distribution tariff rates, that were in effect on an interim basis from January 1, 2012 to December 31, 2012, be approved on a final basis. (b) Approve Fortis approach for setting final 2013 rates, using existing rate structures, as set out in the application, after and subject to final approval of the PBR parameters in the PBR related proceedings pending before the Commission. (c) Approve the 2012 distribution cost allocation study, methods and cost allocation percentages, as set out in Section 2 of the application. (d) Approve the 2013 transmission cost allocation methods and cost allocation percentages, as set out in Section 3 of the application. (e) Approve for use, the 2013 billing determinant forecast on existing rate structures, and as filed in the PBR compliance filing in Proceeding ID No. 2130, for purposes of establishing the target 2013 PBR revenue, after and subject to final approval of the PBR parameters in the related PBR proceedings pending before the Commission. (f) Approve Fortis Phase II approach to apply the cost allocation percentages approved as part of the 2012 distribution cost study (per part (c) above) to the target 2013 PBR revenue approved in part e above, for purposes of establishing newly structured 2013 rates. These rates will not be implemented, but will be subsequently used as the basis to establish and propose newly structured 2014 rates in the next annual PBR rates filing. (g) Approve the converted 2013 billing determinant forecast, reflecting the proposed new rate structures, for purposes of establishing the newly structured 2013 rates. These rates will not be implemented, but will be subsequently used as the basis to establish and propose newly structured 2014 rates in the next annual PBR rates filing. (h) Approve the new rate structures and schedules as set out in Section 4 and Appendix 1 of the application for purposes of establishing newly structured 2014 rates. These rates, adjusted once final approval of the PBR parameters in the related PBR proceedings are issued, are proposed to be implemented in billing effective January 1, 2014, in the next annual PBR rates filing. (i) Approve Fortis proposed K and Y factor allocators and methods, as set out in Section 2.4 of the application, for allocating K factor and Y factor amounts in future calculations of the distribution adjustment rider (DAR) by rate class, after and subject to approval of the PBR parameters in the related PBR proceedings once issued by the Commission. (j) Confirm that Fortis has responded and complied with all outstanding Commission Phase II directions. (k) Approve Fortis proposal with respect to resolution of the Harvest Operations Corp. application to review Fortis Rate 44/45 charges, pursuant to Decision Exhibit 1, paragraph 21. Exhibit 1, paragraph AUC Decision (January 27, 2014)

11 (l) Approve collection of the difference between revenue on final rates and revenue on interim rates for the months that customers were on interim rates in 2013, after final approval of the PBR parameters in the related PBR proceedings is issued, and/or Bulletin is rescinded by the Commission. 16. Some of the approvals requested above have already been provided in related PBR decisions. 3 Responses to AUC directions 3.1 Distribution related 17. Decision included Direction 8, Direction 10 and Direction 11 related to the distribution cost allocation that Fortis was to address as part of its this Phase II application. The Commission has reviewed Fortis evidence and supporting documentation and subject to the findings made by the Commission in Section of this decision regarding Direction 11 is satisfied that Fortis has complied with these directives from Decision Transmission related 18. Decision included Direction 2, Direction 5 and Direction 7 related to transmission cost allocation that Fortis was to address as part of its this Phase II application. The Commission has reviewed Fortis evidence and supporting documentation and subject to the findings made by the Commission in Section 6.2 of this decision regarding Direction 2 is satisfied that Fortis has complied with these directives from Decision AUC directions from Decision Fortis compliance with AUC directions from Decision is addressed in Section 13.3 of this decision. 4 PBR implications 4.1 PBR framework and Phase II applications 20. Fortis is subject to a five-year PBR plan for the years 2013 through 2017, approved in Decision As set out in Decision , the PBR framework provides a formula mechanism for the annual adjustment of rates. In general, the companies rates are adjusted annually by means of an indexing mechanism that tracks the rate of inflation (I) relevant to the prices of inputs the companies use less an offset (X) to reflect the productivity improvements the companies can be expected to achieve during the PBR plan period. As a result, with the exception of specified adjustments, a utility s revenues are no longer linked to its costs. Companies subject to a PBR regime must manage their businesses and service obligations with the revenues derived under the Decision : Harvest Operations Corp., Application to Review FortisAlberta Inc. Rate 44/45 Charges, Application No , Proceeding ID No. 2006, January 3, Bulletin , Government of Alberta request regarding electricity rates, March 13, Decision : FortisAlberta Inc., 2010 Phase II, Application No , Proceeding ID No. 362, July 22, AUC Decision (January 27, 2014) 5

12 PBR indexing mechanism and adjustments provided for in the formula. The PBR framework is intended to create efficiency incentives similar to those in competitive markets. 22. In accordance with the provisions of the PBR plan approved for Fortis in Decision , in addition to the I-X mechanism, the company s distribution rates for each year may include an adjustment to fund necessary capital expenditures (K factor), an adjustment for certain flow-through costs that should be directly recovered from customers or refunded to them (Y-factor), and an adjustment to account for the impact of material exogenous events for which the company has no other reasonable cost recovery or refund mechanism within the PBR plan (Z-factor). 23. In the application, Fortis observed that under PBR, Phase II methodologies will be needed to allocate any approved K, Y and Z factor amounts. With the advent of PBR in 2013, certain distribution costs will be recovered outside of the I X component of base rates, including any applicable K, Y and Z Factor rate adjustment amounts, which are proposed to be recovered through a separate Distribution Adjustment Rider (DAR). In accordance with Decision , these amounts are to be allocated consistent with the Phase II methods and allocators. Accordingly, FortisAlberta has proposed the basis for these allocations in Section 2.4 to permit their timely use in the establishment of final rates and riders for Fortis further commented on the interaction between the PBR framework and the Phase II matters discussed in this proceeding as follows: [ ] of greatest concern in entering a PBR framework is transitioning the overall rate levels at 100% Revenue-to-Cost (R/C) ratios by rate class as soon as possible, while still ensuring this transition is accomplished in a reasonable, orderly and acceptable manner However, in response to AUC-FAI-019 regarding why the objective of 100 per cent revenue-to-cost ratios is more important under PBR, Fortis acknowledged that revenue-to-cost ratios can be adjusted during a PBR, and do not need to be adjusted to 100 per cent in entering a PBR framework. 26. Under traditional cost of service regulation, in years when there is no new comprehensive Phase II cost allocation proceeding, rates are usually updated by increasing the existing rates by a proportionate amount to recover the approved Phase I revenue requirement. PBR breaks the link between revenues and costs that is the basis of cost of service regulation, where rates are set to recover approved forecast costs. 27. In Decision , the Commission noted that PBR is unrelated to the requirement to periodically update rates through a Phase II process. Nevertheless, the Commission stated that: 996. [ ] during the PBR term the companies may file applications for Phase II adjustments to their rate design and cost allocation methodologies and the Commission will make a determination at that time as to whether the adjustments are warranted. For purposes of a cost of service study, the companies shall use the revenue requirement Exhibit 1, paragraph 11. Exhibit 1, paragraph AUC Decision (January 27, 2014)

13 resulting from going-in rates adjusted by the PBR formula (including the I-X mechanism, K factors, Y factors and Z factors) and the latest updated billing determinants Furthermore, although the intent of the PBR plan is to break the link between revenues and costs, any approved K, Y and Z factor amounts would be recovered on a cost of service basis, thereby linking revenues and costs for these expenditures. Therefore, Phase II methodologies are still required under PBR to convert the approved K, Y and Z factor dollar amounts into customer rates. This issue is further discussed in Section Allocation of K and Y factor amounts 29. In Decision , the Commission directed the companies to allocate items outside of the I-X mechanism, including K, Y and Z factors (except for items subject to flow-through treatment and collected by way of a separate rider) to rate classes based on the most recent approved forecast of billing determinants along with the Phase II methodologies currently in place This matter was further discussed in Decision , where the Commission determined that both the simplified factor allocation (i.e., using the projected base revenue per rate class as the allocator) and classifying, functionalizing and then allocating the K, Y, and Z factor amounts by rate class using the last approved Phase II methodologies are acceptable for allocating K, Y, and Z factor amounts that apply to all rate classes. 21 In addition, the Commission noted that, in the event that any of the applied-for K, Y or Z factors do not apply to all customer classes, they should be allocated to rate classes using the approved Phase II methodologies which involve classifying, functionalizing, and then allocating any rate-class specific amounts In response to Information Request AUC-FAI-007 in the PBR Compliance Filings, Proceeding ID No. 2130, Fortis confirmed that the underlying costs in any final 2013 rates and riders would be allocated on a basis consistent with its approved Phase II cost allocation methods and allocators. Fortis added: While some Y Factor and K Factor amounts could be allocated based on the existing approved allocators, others may not fit squarely into one of the above categories. In these cases, additional allocators may need to be developed to ensure cost causation is maintained by rate class. FortisAlberta plans to discuss such allocations in its upcoming Phase II Application, and if such allocations are accepted in early 2013, this would allow timely use in the establishment of final rates and riders for Decision , paragraph 996. Decision , paragraph 993. Decision , paragraph 65. Decision , paragraph 66. Proceeding ID No. 2130, Exhibit 94.02, AUC-FAI-7(b). AUC Decision (January 27, 2014) 7

14 32. On this basis, Fortis proposed the following allocators for each Y factor and K factor amount applied for in the related PBR proceedings: Table 1. Allocators for Y factor and K factor amounts Y factors AUC assessment fees Hearing costs for interveners Load settlement costs Property and business taxes ROE deferral AESO contribution deferral Metering capital deferral Metering operating costs deferral Settlement System Code costs Farm transmission credit Carrying costs Proposed allocator Cost subtotal Cost subtotal Site count Cost subtotal Rate base subtotal Transmission DTS POD charges AMR metering property Cumulative meter sites AMR metering property Farm transmission allocator Cost subtotal K factors Customer growth AESO contributions Substation upgrades Line moves/ipp Distribution control centre/scada Proposed allocator As per Capital Tracker application Transmission DTS POD charges CAM RCN shared CAM RCN shared Cost subtotal 33. No party objected to the proposed allocators. Commission findings 34. The Commission has considered the proposed allocators for each Y factor and K factor, as shown in the table above. The Commission is satisfied that the approach adopted by Fortis to allocate the approved K and Y factors is consistent with the Commission s directions in decisions and Accordingly, the Commission approves the allocation of K, and Y factors as proposed by Fortis. Fortis is directed to use these allocators throughout the PBR term for any approved K and Y factors. 35. However, the Commission approved a simplified allocation methodology, which uses base revenue per rate class as the allocator, for the K factor placeholder amounts in Decision Accordingly, the Commission finds that Fortis may use this simplified allocation methodology for any K factor placeholders approved by the Commission until such time as the Commission approves a final K factor. 36. The Commission expects that Fortis will address the issue of Z factor allocation in any application requesting approval of a Z factor amount. 5 Rate design and cost allocation overview 37. Rate design is the process used by utilities to establish rates which direct how customers will be billed for their electricity usage. It includes the selection of rate classes, a cost of service study to establish the cost of providing electrical service by customer class; the basis for charges, 8 AUC Decision (January 27, 2014)

15 such as energy, or capacity (typically on a cost causation basis) and the determination of the amount of each charge. 38. Fortis indicated that it has used standard rate design (Bonbright) principles, restated as they pertain to an owner of an electric distribution system operating under a PBR framework. 24 As customers have different preferences regarding rates, the principles are weighted and balanced by rate class. Recover FortisAlberta s expected target revenue associated with providing electric distribution service when the rates are applied to the forecast load and billing determinants; Recognize cost causation, by reflecting the results of its cost allocation study and analysis with respect to transmission and distribution costs; Avoid undue discrimination between rate classes and individual customers within each class; Allow for simplicity of understanding and acceptance by end-use customers and their retailers and provide for ease of administration and economy of practical billing; Consider the existing rates, including trends, changes and stability of the rate levels; Recognize the value of service and support optimum use of electric distribution service; and Consider the rates and practices of other utilities providing comparable service. 39. Noting its proposal to retain the existing rate classes, Fortis submitted that an end-use basis is the best measure of homogeneity respecting costs and preferences for smaller loads and that load size remains the best measure to distinguish larger general service customers. Further, maintaining the existing rate categories allows stability and continuity for customers ease-of-understanding and stable billing implementation, and is naturally aligned with the data that is established and used in the cost allocation study. 40. Fortis proposed a number of changes to its distribution rate design, including elimination of the use of estimated billing demand based on breaker size for farm rates, inclusion of a service charge (Block 1) on either a $/site or a minimum demand charge, and changing the large general service (Rate 63) $/kilometer (km) charge to a $/kilowatt (kw) (kw x km) charge. Fortis stated that it had simplified rates and reduced variability. 24 Exhibit 1, application, paragraphs 144 and 145, PDF page 46. AUC Decision (January 27, 2014) 9

16 41. The following table 25 presents the restructured rates initially applied for: Table 2. Restructured rates initially applied for Restructured 2013 Rates Summary (Rounded) kw $/site/day R11 - Residential 0.79 $/site R Farm (kwh 1.32 Metered) $/site R Farm (kva 1.32 Metered) $/site R REA 0.02 $/site R26 - Irrigation 0.19 $/kw R31 - Light: Investment 0.59 $/fixture R33 - Light: No-Investment 0.21 $/fixture R38 - Light: Yard Lighting 0.37 $/fixture R41 - Small General Service 0.39 $/kw R44 - Oil &Gas (Capacity) 0.69 (Closed) $/kw R45 - Oil &Gas 0.69 $/kw R61 - General Service R63 - Large General Service R65 - Transmission Connected Service R66 - Opportunity Transmission Rider E - Customer Specific Facilities Charge Option A - Primary Service Credit Option I - Interval Metering Option 30 $/site 35 $/site 30 $/site 0.82 $/meter Blk 1 Blk 2 Blk 3 Blk 1 Blk 2 Blk 3 Rate Demand (Daily) (Daily) (Daily) Size Size Size Min & Max Ratchet 7.89 >30 /kwh $/site none kwh 7.81 /kwh All kwh 0 $/site none $/kw 0.02 $/kw 2.20 /kwh 0.01 $/kw 0.53 $/kw 0.25 $/kw 0.50 $/kw 0.50 $/kw $/kw $/kw.km 0.37 $/kw 0.47 $/kw 0.47 $/kw $/kw >= 0 kw 0-3 kw 0-3 kw 0-3 kw 0-50 kw > 0 kw >3-15 kw >3 kw >3-15 kw >3-15 kw > kw >15 kw >15 kw >15 kw >500 kw $/site $/site Contract kw $/fixture $/fixture $/fixture 3 75 kw none n/a 95% of Expected kw n/a n/a n/a 85% 50 kw 3 75 kw 100% 3 75 kw 85% 50 2,000 kw 2,000 kw Min $/site 85% 90% n/a 42. Fortis s 2013 interim rates were approved in Decision , 26 On March 4, 2013, Fortis amended its application for Rate 11 (residential) and rates (farm). 27 In the letter, Fortis compared the restructured rates originally applied for with the revised restructured rates, as set out in Table Exhibit 1, application, paragraph 161, PDF page 52. Decision : 2012 Performance-Based Regulation Compliance Filings, January 1, 2013 Interim Rates for each of ATCO Electric Ltd., EPCOR Distribution & Transmission Inc. and FortisAlberta Inc., Application No , Proceeding ID No. 2130, December 17, Exhibit AUC Decision (January 27, 2014)

17 Table 3. Amendments to application Restructured 2013 Rates Summary (Rounded) ($/Day) Filed January 18, 2013 Blk 1 Blk 2 Blk 3 $/site Blk 1 Rate Blk 2 Rate Blk 3 Rate Size Size Size R11 - Residential 0.79 $/site 7.89 /kwh <=30 kwh >30 kwh R Farm (kwh Metered) 1.32 $/site 7.81 /kwh All kwh R Farm (kva Metered) 1.32 $/site 0.53 $/kw 0.37 $/kw >3-15 kw >15 kw Restructured 2013 Rates Summary (Rounded) ($/Day) Revised Blk 1 Blk 2 $/site Blk 1 Rate Blk 2 Rate Size Size R11 - Residential 0.79 $/site 0.94 /kwh 1.86 /kwh <=30 kwh >30 kwh R Farm (kwh Metered) 2.79 $/site 3.56 /kwh All kwh R Farm (kva Metered) 2.79 $/site 0.42 $/kw >3 kw 43. Fortis explained that the originally proposed rates for residential and farm customers would result in bills that increased significantly in months with higher consumption. Fortis recognized that these customers prefer bill stability. 28 The proposed restructured rates were designed to keep the overall revenue in each rate class unchanged. Further these changes would also minimize the variability in bill impacts within these rate classes. 44. The majority of costs in the cost allocation study are allocated based on the CAM model. In response to AUC-FAI-007, Fortis provided a history of the CAM model. The model was developed in 2000/2001 with nine feeders, selected to represent all rate classes. In the distribution tariff application (DTA), an additional 40 feeders were selected on a random basis. In response to a direction by the board, the sample was increased in the DTA (based on 2005 data) with the addition of a further 51 feeders, selected on a random basis for a total of 100 feeders. A statistical analysis of the reliability of the CAM model at that time found that the impact of adding additional feeders would typically vary the results by less than two per cent. 45. Since 2005, three feeders have been replaced as they were removed from the system. During the intervening period the number of distribution feeders has increased from 480 to 560 and for some rate classes the change in customers on the sample does not reflect the corresponding change in customers in the system as a whole. In response to IPCAA-FAI Fortis provided the following information comparing the sample feeders to the total distribution system Exhibit Exhibit AUC Decision (January 27, 2014) 11

18 Table 4. Sample feeders compared to total distribution system A B C D E F G Installed Capacity [2.2-R Col G] kva (1) Sample Feeders Number of Customers (1) Allocator [2.2-R Col H] [ A B ] # KVA/customer Total Distribution System Installed Capacity (1) Number of [2.2-Q2 Col D] Customers kva/customer kva # [ D E ] Relative Difference [ (C-F)/F ] Line No. Rate Class Description Rate Code 1 Residential ,933 61, , , % 2 FortisAlberta Farm ,691 11, , % 3 REA Farm ,788 7, , % 4 FortisAlberta Irrigation 26 95,721 1, , % 5 Exterior Lighting 3X 21 10,570 6 Small General Service ,274 10, ,247 50, % 7 Oil and Gas ,393 3, , % 8 General Service ,067 1, ,407 8, % 9 Large General Service , ,816 1, , % 10 Total FortisAlberta 2,000,015 96, , , % Note: (1) From FAI Phase II DTA 46. In response to IRs from IPCAA, Fortis identified some of the assumptions in the CAM model and explained the reasons for some of the numbers in Schedule 2-2 Q-R 2012 CAM feeder analysis. For example, Fortis stated that in the CAM model all residential customers are assumed to have the same on-peak utilization factor regardless of whether they have a dedicated or shared transformer, such that the on-peak load for a residential customer with its own 10kVa transformer is two times the on-peak load for a residential customer who shares a 10 kva transformer. 47. Fortis stated in the application that 90 per cent of property is distribution grid property and 10 per cent facilities and fleet property. The return on the former is based on original cost, while the return on the latter is based on replacement cost new. In response to AUC-FAI-013, Fortis provided the following tables indicating the impact of the different bases on property allocation and allocated distribution costs. 12 AUC Decision (January 27, 2014)

19 Table 5. Fortis response to IR AUC-FAI-013 FortisAlberta Response to Information Request AUC-FAI-013 Sensitivity of Original Cost and RCN Cost Allocators A B C D E Impact on Property Allocation Line No. Rate Class Description Rate Code Total Property [Sch 2.1-N11 Col E] $ 000 Total Property [Original Cost Allocator Only] $ 000 Total Property [RCN Cost Allocator Only] $ 000 Original Cost Allocator Only vs Filing [(B -A)/A] % RCN Cost Allocator Only vs Filing [(C- A)/A] % 1 Residential 11 $ 970,641 $ 970,729 $ 969, % -0.1% 2 FortisAlberta Farm , , , % 10.4% 3 REA Farm , , , % 0.7% 4 FortisAlberta Irrigation , , , % -8.6% 5 Exterior Lighting 3X 265, , , % -31.3% 6 Small General Service , , , % -0.1% 7 Oil and Gas , , , % 9.3% 8 General Service , , , % 3.7% 9 Large General Service 63 92,406 94,273 76, % -17.2% 10 Transmission Connected Total FortisAlberta $ 3,449,111 $ 3,449,111 $ 3,449, % 0.0% Impact on Allocated Distribution Costs Rate Class Description Rate Code Total Distribution Costs [REA allocation = LS Costs] [Sch 2.1-F Col F] $ 000 Total Distribution Costs [REA allocation = LS Costs] [Original Cost Allocator Only] $ 000 Total Distribution Costs [REA allocation = LS Costs] [RCN Cost Allocator Only] $ 000 Original Cost Allocator Only vs Filing [(B -A)/A] % RCN Cost Allocator Only vs Filing [(C- A)/A] % 12 Residential 11 $ 134,066 $ 134,068 $ 134, % 0.0% 13 FortisAlberta Farm ,469 61,970 66, % 6.8% 14 REA Farm % 0.0% 15 FortisAlberta Irrigation 26 14,173 14,279 13, % -6.4% 16 Exterior Lighting 3X 19,495 20,212 13, % -31.3% 17 Small General Service 41 50,666 50,666 50, % 0.0% 18 Oil and Gas ,832 42,535 45, % 5.9% 19 General Service 61 62,606 62,439 64, % 2.3% 20 Large General Service 63 9,737 9,877 8, % -12.3% 21 Transmission Connected 65 1,019 1,019 1, % 0.2% 22 Total FortisAlberta $ 397,291 $ 397,291 $ 397, % 0.0% 48. The practice of using original cost was adopted in the DTA which directed Fortis to take into account the effective age of assets serving each customer class Fortis suggested that for the next cost allocation, the number of feeders in the CAM model should be increased by 20 per cent or 20 feeders to be selected randomly from feeders added to the system since The benefits of updating the sample size in this manner include recognizing the growth that is being experienced on the system and keeping the sample representative of the total distribution system Two enhancements to the CAM model were introduced in this application: the allocation of poles and conductors separately and the determination of the average on-peak kw/installed kva for irrigation sites based on 12 months rather than seven months. The impact of these enhancements are reflected in the 2012 distribution cost study. The direct allocation of poles enables pole depreciation and return to be calculated and allocated separately from conductor depreciation and return to more accurately reflect the age of assets serving individual customers As quoted in AUC-FAI-013. AUC-FAI-007. AUC Decision (January 27, 2014) 13

20 since pole replacements frequently occur without replacing the associated conductor. For irrigation customers, the use of seven months, rather than 12, in the on-peak kw/installed kva allocator, resulted in more distribution costs being allocated to irrigation customers A comparison of Fortis 2010 and 2012 distribution cost studies is set out in Table 6: Table 6. Comparison of Fortis 2010 and 2012 distribution cost studies A B C D E 2010 Distribution Cost Study 2012 Distribution Cost Study Change Percent of Percent of Line Cost Category $ Millions total $ Millions total [(D-B)/B] 1 Vegetation Management % % -12% 2 Other Operations % % 8% 3 Depreciation % % -6% 4 Return % % 8% 5 Distribution Property Subtotal % % 1% 6 Metering % % 34% 7 Wholesale Billing % % 8% 8 Load Settlement % % -60% 9 Other Customer Service % % -16% 10 Information Technology % % -3% 11 General and Admin % % -25% 12 Miscellaneous Revenues (12.0) -3.5% (11.7) -2.9% -16% 13 Total % % 0% 52. The impacts of the 2012 enhancements are reflected in Table 7. The separate allocation of poles and conductors resulted in a 4.9 per cent increase in costs to residential customers with decreases primarily to Fortis farm (4.8 per cent), large general service (3.3 per cent) and general service (3.2 per cent). The costs allocated to irrigation customers decreased by 11.9 per cent, 33 but other rate classes were impacted by less than one per cent UCA-FAI-004. As noted in IPCAA-FAI , The impact of the proposed change to a 12-month resulted in an 11.1 per cent decrease in the allocated Property retirement Units (PRU) costs for all sample feeders. 14 AUC Decision (January 27, 2014)

21 Table 7. Impact of the 2012 enhancements FortisAlberta Response to Information Request CCA-FAI-005 Summary of the Impact on the Affected Rate Classes as a Result of the Two Proposed Enhancements A B C D E Line No. Rate Class Description Rate Code Total Distribution Costs [Sch 2.1-F Col F] $ 000 Total Distribution Costs [W.O. Pole and Conductor Allocation Enhancements] $ 000 Total Distribution Costs [On-Peak kw/installed kva Based On 7 Month] $ 000 Relative Difference [ (B-A)/A] % Relative Difference [ (C-A)/A] % 1 Residential 11 $ 134,066 $127,827 $ 133, % -0.2% 2 FortisAlberta Farm ,469 65,615 61, % -0.8% 3 REA Farm % 0.0% 4 FortisAlberta Irrigation 26 14,173 14,281 15, % 11.9% 5 Exterior Lighting 3X 19,495 19,077 19, % -0.1% 6 Small General Service 41 50,666 50,731 50, % -0.5% 7 Oil and Gas ,832 43,777 42, % -0.5% 8 General Service 61 62,606 64,676 62, % -0.7% 9 Large General Service 63 9,737 10,067 9, % -0.1% 10 Transmission Connected 65 1,019 1,014 1, % 0.1% 11 Total FortisAlberta $ 397,291 $ 397,291 $ 397, % 0.0% 53. EQUS and NPP reviewed the CAM model and the results of the cost allocation study. They concluded that the CAM model is not a reliable and reasonable predictor of the costs associated with integrated operations with REAs or the identification of those costs for allocation. 54. EQUS and NPP argued that, based on a random sample of five feeders, actual line ownership within their respective service areas was significantly different (and in favour of EQUS and NPP) than the 61.4 per cent Fortis/38.6 per cent REA ownership ratio used by Fortis to derive costs to serve REAs. Further, they claimed the CAM model allocation appeared not to take into account and deduct costs for previously owned lines of REAs which Fortis has a perpetual obligation to maintain and repair. 55. EQUS and NPP confirmed that no regular exchange of mapping information between EQUS and NPP and Fortis has taken place since Without regular exchanges of facilities and mapping information, EQUS and NPP argued that the information in the CAM model would likely be inaccurate and not reflective of the real-time ownership share of the integrated distribution system as between REAs and Fortis. 56. EQUS and NPP compared the data from the CAM model with their data for a sample of five feeder lines and concluded that Fortis allocation of $12.1 million of net distribution costs to REAs was not supportable. 57. EQUS and NPP clarified that it was not their intention to challenge the CAM model with respect to its overall allocation of costs to Fortis own customers for the purposes of Fortis distribution tariff. Further, EQUS and NPP confirmed that they are not proposing any other model or approach for Fortis allocation of costs to its customers. EQUS and NPP argued that Fortis has no authority to impose, nor does the Commission have the jurisdiction to approve a distribution tariff for Fortis which imposes a distribution/usage charge upon REAs and their members. Only load settlement and flow-through Alberta Electric System Operator (AESO) related charges components of Fortis costs should be recovered from REAs under Fortis distribution tariff. AUC Decision (January 27, 2014) 15

22 58. The CCA submitted that the CAM method is a tried and tested method used to assign or allocate costs to rate classes. Based on this approach, the REAs have been allocated certain distribution costs caused by them, including the cost of facilities required to serve REAs. Exempting the REAs from the costs incurred by Fortis in order to serve them would clearly result in higher costs to other customer classes. The costs properly allocated to the REAs under the CAM method are based on cost causation and are properly payable by the REAs. Accordingly, the CCA argued that the REA proposal that they be exempt from paying a major portion of Fortis investment in distribution facilities and costs should be rejected. 59. Fortis stated that the majority of distribution costs are allocated based on data from the CAM model, which directly allocates property for 100 distribution feeders to the individual sites being served by these feeders. The proposed distribution cost allocation remains essentially unchanged from the last comprehensive Phase II application (2010 Phase II Application No ). 34 Fortis indicated that the CAM model included updated mapping data for all 100 sample feeders, updated load settlement data, and aligns with the 2012 Phase I NSA approved in Decision Fortis explained that, mapping discrepancies would be limited to the relatively few self-operated REA distribution lines built solely to serve REA customers (of which Fortis was aware of none) after the 2009 distribution loss study and before mid-2011 (the test year used for the CAM model samples). With respect to the maps provided by EQUS and NPP, Fortis was only able to identify nine instances of missing REA services. All of the differences were identified as additions to EQUS and NPP owned distribution lines. These lines are used solely to serve EQUS and NPP sites. Fortis observed that there was no missing mapping data with regards to joint-use distribution facilities and no differences in line ownership were noted. 61. Fortis concluded that the impact of the CAM model not including some REA use only data would be to understate the total REA load on the joint-use portions of distribution lines. This would result in an under-allocation of distribution costs to REAs which, while very small, would be to the benefit of the REA rate class. In any event, there would be no cost allocation impact for any such missing REA-owned distribution line as it only serves a REA member. Commission findings 62. The evidence of EQUS and NPP was based on a sample of five feeders. The Commission considers that this sample size is not sufficiently representative and therefore conclusions made or drawn based on this evidence are not adequately supported. 63. The Commission considered Fortis description of rate classes and its explanation that maintaining the existing rate categories maintains stability and continuity of customers ease-of-understanding and stable billing implementation, and is naturally aligned with the data that is established and used in its cost allocation study. The Commission accepts Fortis proposal for no changes to its rate classes. 64. Cost allocation studies are subject to the assumptions made and the data available and relied upon, and are therefore only approximations. 34 Exhibit 1, paragraph AUC Decision (January 27, 2014)

23 65. Given that the feeder representation in the CAM model has not been updated since 2005, despite the growth in the system, and given that Fortis indicated the model is not representative of the system for some rate classes, the Commission directs the following. Fortis shall add a further 20 feeders (or more if required to make the sample representative) to the CAM model prior to the next cost of service study and undertake a statistical analysis to ensure that the updated CAM model is representative. 66. The Commission has reviewed the comparison of the 2010 distribution study to the 2012 distribution study. Distribution property costs account for 78.5 per cent of the total in 2010 and 79.3 per cent of the total in The large amounts involved indicate the importance of assumptions regarding property in the CAM model, such as property retirement units (PRU) and the basis on which assets are valued, such as by way of replacement cost new (RCN). With respect to non-property costs, metering costs increased from $25 million (7.2 per cent) to $38.5 million (9.7 per cent), load settlement costs decreased from $7.7 million (2.2 per cent) to $3.5 million (0.9 per cent) and general and administration costs decreased from $32.6 million (9.4 per cent) to $28.1 million (7.0 per cent). The change in load settlement costs is partly related to the treatment of a portion of these costs as a Y factor adjustment under PBR. 67. The Commission has reviewed the record with respect to assumptions. For example, Fortis has assumed that all residential customers have the same on-peak utilization factor regardless of whether they have a dedicated or shared transformer. However, for practical reasons of data availability and cost benefit trade-offs, despite the short-comings of this assumption, it may be the best option. As the CAM model has been in use for many years, without a detailed review, the Commission directs that in its next Phase II application, Fortis shall provide a list of all assumptions and explain the rationale for each assumption that impacts the allocation of costs to one or more rate classes by a material amount. 68. Notwithstanding, the Commission approves the use of the CAM model and the resulting cost allocations as proposed by Fortis for the purposes of this decision. The Commission has also examined the local cost multiplier, in the following section. 5.1 Local costs multiplier for Rate In its application, Fortis calculated the local costs multiplier for Rate 63, to adjust for multi-transformer installations with additional switchgear not captured in Fortis simple line-transformer model, consistent with previous methodology. 70. Fortis submitted that without a local costs multiplier adjustment, local costs would be understated by the CAM model for Rate 63 services. In response to AUC-FAI-010, Fortis indicated that the local costs multiplier adjustment factor had not been reviewed since and was not reviewed or updated for this application. 71. IPCAA stated that in information requests it had asked Fortis to identify reasons why two customers in the sample had such high costs, but Fortis declined to respond. In its rebuttal evidence, Fortis did not respond to the IPCAA questions related to the Rate 63 local costs multiplier IPCAA proposed removal of the two large customer data points and argued that the onus is on Fortis to provide evidence that these two large customers are representative of the Rate Exhibit , IPCAA reply argument, page 2. AUC Decision (January 27, 2014) 17

24 population. Absent some evidence that the extraordinary costs associated with the two outlier customers are representative of the Rate 63 population, IPCAA argued that these two data points are outliers in the overall customer data sample, which result in a material cost impact to Rate 63 customers and thus should be excluded from the Rate 63 local costs multiplier calculation. IPCAA argued that when applied to the total Rate 63 population, the inclusion of these two outliers increased the allocation of local property by more than $40 million, contributing roughly 25% of the total $160 million of property allocated to Rate IPCAA proposed that, such a significant cost impact is unwarranted absent some evidence these two large customers are representative of the Rate 63 population The UCA submitted that because the CAM model and multiplier methodologies have not changed since , and that the evidence is inconclusive as to whether there is anything wrong, the removal of the two outlier customers from the multiplier is not warranted at this time. 38 Further, the UCA submitted that, if the Commission has a concern with the local multiplier data, then it should direct Fortis to provide an analysis in all future cost of service studies that demonstrates that the sample data is representative of the population. This should be expanded to cover all data in the CAM model The CCA expressed concerns similar to those as expressed by the UCA and IPCAA regarding the two outlier customers. 40 The CCA opposed changing the multiplier proposed by Fortis in this proceeding but submitted that at the time of its next Phase II application, that Fortis should provide evidence showing how the outliers referred to by IPCAA, as well as how the contributions, were recognized in the multiplier calculations In reply argument, Fortis suggested that the presence of two customers with higher multipliers is to be expected due to the non-homogeneous nature of the cost to provide service to each unique large customer. For IPCAA to now consider two customers as outliers and suggest their exclusion from the data set used to determine the approved Rate 63 multiplier over recent years is inappropriate and should be rejected by the Commission. Fortis agreed that it may be appropriate for it to undertake to update its Rate 63 multiplier study for purposes of its next Phase II application. Commission findings 76. The Commission recognizes that in there were approximately 127 Rate 63 customers; this number has now increased to During the same time period the sample size used to calculate the Rate 63 multiplier decreased by three from 28 to Given these changes in customer number and sample size the Commission shares IPCAA s concern that the sample used to calculate the Rate 63 multiplier may no longer be representative. 77. IPCAA proposed that two outliers be removed from the sample size. Rate 63 is comprised of large commercial customers and the cost to provide service to each customer is unique. As the two outliers have always been part of the sample the Commission is not Exhibit , IPCAA argument, Section 3.1, paragraph 3. Exhibit , IPCAA reply argument, page 3. Exhibit , UCA argument, paragraph 8. Exhibit , UCA argument, paragraph 12. Exhibit , CCA argument, paragraph 21. Exhibit , CCA argument, paragraph 26. Exhibit AUC-IPCAA AUC Decision (January 27, 2014)

25 convinced that the two large customers should be removed from the sample used to calculate the local costs multiplier for the purposes of this decision. 78. The Commission directs Fortis to update its Rate 63 multiplier study for the purposes of its next Phase II application. Specifically, the Commission directs Fortis to undertake a statistical analysis to assess whether the sample of 25 customers is representative of the total rate class. If not, Fortis is directed to increase the sample size to make it representative of the total rate class prior to its next cost of service study. 6 Cost allocation 79. The distribution tariff is comprised of two distinct components: a transmission component and a distribution component. Each of these cost components is allocated to rate classes separately based on distinct transmission and distribution cost drivers and characteristics. The distribution cost allocations are discussed first, followed by the transmission cost allocations. 6.1 Distribution cost allocation 80. Fortis applied for approval of its 2012 distribution cost allocation study, methodology and resultant cost allocation percentages. 81. Distribution costs comprise the costs of the physical distribution facilities associated with providing regulated distribution services, and include fixed and variable operating costs (including property taxes), depreciation, administrative or overhead costs, income taxes and return. The distribution revenue requirement of $398.2 million approved in Decision , 44 adjusted for the PBR Y factor flow through ($398.2 million minus $0.9 million), which resulted in $397.3 million being allocated to the various rate classes, was used as the basis for this 2012 cost allocation study. 44 Decision : FortisAlberta Inc., Application for Approval of a Negotiated Settlement Agreement in respect of 2012 Phase I Distribution Tariff Application, Application No , Proceeding ID No. 1147, April 18, AUC Decision (January 27, 2014) 19

26 82. Fortis provided the following table setting out the allocation of distribution costs: 45 Table 8. Fortis allocation of distribution costs Cost category $ million % Allocation description total Vegetation management Overhead line property from CAM adjusted to exclude irrigation customers Other operations Distribution property allocator from CAM Depreciation Depreciation property allocator largely related to CAM Return Rate base allocator largely from CAM Distribution property subtotal Metering Direct assigned Billing Per customer except with 8% to large customers Load settlement Per site Other customer service % based on customer count, 20% allocated to interval metered customers only by revenue Information technology Costs subtotal allocator General and admin Costs subtotal allocator Miscellaneous revenues (11.7) -2.9 Costs subtotal allocator excluding transmission connected customers Cost excluding flow-through items Flow-through items 0.9 Total AUC assessment fees 83. On March 15, 2013, the Commission issued Bulletin , Stakeholder consultation on AUC Rule 025: Administration Fee (AUC Rule 025), and a draft version of AUC Rule 025 with proposed revisions. Noting that AUC assessment fees are to be based on site counts, Fortis recommended changing the allocator for AUC assessment fees to site counts. 84. The CCA considered that the AUC s net expenditures stem from a myriad of activities undertaken by the AUC. The CCA argued that the allocation of almost 70 per cent of the AUC s net expenditures to the residential class is unreasonable. The CCA argued that AUC assessment fees are more in the nature of overhead costs, similar to hearing costs for interveners, and as such, should be allocated to the rate classes using a broad-based allocator such as the costs subtotal allocator. 85. The CCA acknowledged Fortis proposed allocation basis of site counts for AUC assessment fees in its approved Y factor, in response to the Commission s recent directions in AUC Rule 025, but then proposed an entirely different allocation basis of using the costs subtotal allocator. 86. Fortis submitted that it can see no reason to deviate from the allocation method approved by the Commission in AUC Rule 025, and as such, submitted that the CCA s misaligned alternative should be dismissed and Fortis AUC-aligned allocation basis for AUC assessment costs should be approved. 45 Exhibit 1, application, paragraph AUC Decision (January 27, 2014)

27 Commission findings 87. AUC Rule 25 sets out the method for determining administration fees for electric and gas distribution utilities, as follows: (b) for Categories 2 and Category 3, for each year, the billable site count taken from the February 28 site cycle catalogue file posted on each utility s website on that date is used in determining the administration fee for each utility; 88. Given that site count is used in determining the administration fee, the Commission considers that it is reasonable to use site count to allocate the fee to the various customer classes. The Commission approves Fortis s recommended methodology of allocating AUC assessment fees on the basis of site counts Load settlement cost allocation 89. Load settlement costs which include expenses and software related to load settlement were 3.5 million or 0.9 per cent of the allocated distribution costs. In response to Direction 11 in Decision , Fortis provided an analysis of the alternatives for allocating load settlement costs, including the use of a demand allocator (Exhibit 5, Appendix 4 to the application). As a result of this analysis, and the highly automated nature of the load settlement process, Fortis recommended that load settlement costs be 100 per cent allocated based on the number of load settlement sites by rate class. The number of sites used to allocate load settlement costs includes the total number of REA sites for which Fortis performs load settlement. 90. Fortis stated that customer or rate class consumption, demand, and revenue do not impact load settlement costs in any manner. The use of billing capacity, non-coincident peak (NCP), or revenue would allocate more load settlement costs to rate classes with higher average demands, while the use of metered energy would allocate more costs to rate classes with higher average consumption. No basis could be found for the allocation of costs to customers based on demand or energy consumption. 91. The CCA pointed out that Fortis previous practice was to allocate five per cent of load settlement costs to interval metered sites and 95 per cent to non-interval metered sites. The CCA considered that using the number of sites as the only allocator was incorrect, as it would suggest that load settlement costs were entirely variable i.e., as one more site is added, there should be an increase in load settlement costs. However, all except $90,000 of costs related to the data storage requirements are fixed. This $90,000 is a variable cost and should be allocated on the basis of energy consumption. Further as Fortis submitted in response to AUC-FAI-14(d), $90,000 of the total load settlement costs should be allocated to interval metered sites. 92. To recognize that some portion of the load settlement costs are variable in the short term and costs in total are variable in the long term (due to costs exhibiting a step change nature), the CCA submitted that the AUC should direct that these costs be allocated on the historically approved basis of five per cent to interval metered sites and 95 per cent to non-interval metered sites. Alternatively, the CCA submitted that load settlement costs could be allocated based on a one-third weighting to each of demand, customer and energy, as these cost drivers all contribute to the incurrence of load settlement costs. AUC Decision (January 27, 2014) 21

28 93. Fortis argued that the CCA provided no evidence to support its suggested method of allocating load settlement costs. (i.e., using a one-third weighting of each of demand, customers and energy). 94. Based on these findings, Fortis recommended that load settlement cost be allocated 100 per cent based on the number of load settlement sites. Commission findings 95. Past practice was to allocate five per cent of load settlement costs to interval metered sites and 95 per cent to non-interval metered sites, with the percentages reflecting the different effort required for the more complex settlement of interval metered sites. The remaining 95 per cent of load settlement costs were allocated on a per site basis. 96. Since 2005, the load settlement processes for interval metered sites have become more automated to a point where there are no material differences in the costs to process interval metered sites and non-interval metered sites The Commission accepts the analysis provided by Fortis indicating that processing costs are impacted only by site count. Accordingly, the Commission finds it reasonable to allocate load settlement costs based on the number of customers. The Commission approves the change to the allocation methodology for settlement costs proposed by Fortis. 98. No party raised issues with respect to any of the remaining cost allocations for distribution as set out in Table 8 above. The Commission has reviewed the remaining cost allocations and considers them to be reasonable, given the Commission s finding that it approved the use of the CAM model and the resulting cost allocations as proposed by Fortis for the purposes of this decision. 6.2 Transmission cost allocation 99. Fortis requested approval of its 2013 transmission cost allocation methods and cost allocation percentages. Fortis provided the following summary of its 2013 transmission cost forecast: Table 9. Fortis 2013 transmission cost forecast ($ million) Distribution-connected load (AESO including Option M) Distribution-connected load (interchange) 4.7 Total distribution-connected load Transmission-connected load (Rate 65 flow-through) Total transmission costs Fortis submitted that the transmission cost allocation method for distribution connected load, which uses three years of load settlement data for the development of cost allocators, is the same methodology that was used in Fortis last Phase II proceeding (Application No ). 46 Exhibit 5, Appendix 4, 2012 load settlement. 22 AUC Decision (January 27, 2014)

29 Table 10. Commission summary of transmission cost allocation method for distribution connected load 47 AESO charge to distribution-connected load ($ million) Percentage of cost Basis of charge to Fortis Bulk system demand charges Coincident Alberta system peak demand Allocation method to rate Classes Rate class forecast monthly CP demand -DTS tariff costs charged based on capacity Per POD Rate class forecast monthly NCP demand -DTS tariff costs charged based on per POD customer charge 15 4 Per POD Rate class forecast monthly NCP demand Metered energy charges Metered energy Rate class forecast Other transmission charges 4 1 Interchange and UFLS credit metered energy Rate class percentage of total DTS charges 101. Fortis indicated that the coincident peak (CP) and non-coincident peak (NCP) demands were derived by escalating the average 2009 to 2011 load settlement data to The metered energy allocator is based on the 2013 energy forecast In Decision , the Commission issued the following direction to Fortis with respect to the allocation of transmission bulk system demand charges: 38. FAI is directed to make use of hourly data differentiated by rate class to the extent possible in its next cost of service study. For its next Phase II application, FAI is directed to explore the use of hourly load data to directly assign Bulk System Demand Charges to rate classes on the basis of each rate class s forecast load at the coincident peak hour In Decision respecting Fortis transmission and distribution tariff application, the Commission did not approve Fortis proposal to modify its metering systems to accommodate hourly-read collection capability in higher density urban areas. As a result, hourly load data for non-interval metered sites is not available but is estimated using the net system load shape Fortis indicated that by using this data, it is already using the best available hourly data to allocate bulk system demand charges. Allocating bulk system demand charges based on each rate class forecast monthly CP demand, derived from a three-year average of load settlement data, best reflects cost causation and provides rate stability by dampening the effect of any one-year anomaly. Fortis recommended the continuation of its current allocation method for bulk system demand charges Fortis also explored the use of hourly-load data to allocate billing capacity and point of delivery (POD) charges, using 2011 hourly load data in conjunction with the substations hourly load data to find the customer s peak at each individual POD. Fortis used data from Rate Exhibit 13, Schedule 3.1-A, AESO distribution connected load costs. Decision : FortisAlberta Inc., Distribution Tariff Phase I, Application No , Proceeding ID. 212, July 6, AUC Decision (January 27, 2014) 23

30 customers, as this was the only class in which metered hourly-load data by customer is available. Fortis compared the results of this analysis to the results from its current allocation methodology. Using the current methodology the amount to be allocated to Rate 63 was 18.0 per cent and under the explored method the amount would be 18.9 per cent Fortis stated that the disadvantage of the method the AUC directed it to explore is that it (i) assumed the POD is dedicated to serving that specific customer all the time (ii) does not take into consideration any load transfers or changes in customers and (iii) is data intensive and takes too much time Considering the advantages and disadvantages of each allocator, Fortis recommended allocating billing capacity and fixed POD charges using its current method as approved in Decision The CCA noted: 17. the impact of the explored method i.e. use of 18.9% vs. 18.0%, is some $2.3M to Rate 63; 12 notwithstanding Fortis remark that the resulting allocator is remarkably similar, this difference is clearly a material amount. 18. The CCA submits the use of hourly load data for POD specific allocation is more precise and better reflects cost causation as customers are charged for their contribution to system peak at that specific POD. In addition, it results in a reduction in interclass subsidy The CCA recommended that the Commission direct Fortis in its refiling to use the result of the explored method for purposes of allocating billing capacity and POD charges In its analysis of transmission costs, CWSAA noted that bulk transmission costs are allocated to rate classes based on each rate class average megawatt (MW)/month of billing capacity. With respect to this allocation, CWSAA considered that it was consistent with the AESO tariff and changes to the allocation were not necessary. Commission findings 111. The Commission is of the view that a POD-specific allocation based on hourly load data is more precise and better reflects cost causation because customers are charged for their contribution to the system peak at a specific POD. In addition, this approach results in a reduction in inter-class subsidies. However, as noted by Fortis in its disadvantages regarding this method it implicitly assumes that the POD is dedicated to serving a specific customer all the time, does not take into consideration load transfers or changes in customers and is data intensive and time consuming. Given that the resulting difference in POD charges between the two methods is not significant at this time, and recognizing that, the Commission only recently approved the allocation methodology for these charges in Decision , the Commission accepts the Fortis proposal to retain the current allocation method for the purposes of this decision Exhibit , CCA argument, paragraph 19. Exhibit , CCA argument, pages 6 and AUC Decision (January 27, 2014)

31 112. However, given the Commission s findings on the potential benefits of adopting a POD-specific allocation based on hourly load data, the Commission directs Fortis to further explore the costs and benefits of this approach at the time of its next Phase II application In Decision , the Commission approved a full dollar-for-dollar flow-through and true-up of AESO transmission costs commencing in The transmission costs filed as part of Fortis PBR compliance filing (Application No , Proceeding ID No. 2130) are the transmission costs that will be allocated to the various rate classes and reflect the latest forecast of AESO volumes and prices The Commission has considered Fortis transmission cost allocation methodology, which has not changed since the last Phase II application. In Decision , the Commission found that that the allocation of transmission costs using Fortis transmission allocation methodology is reflective of cost causation and consistent with previous Commission decisions. The Commission continues to find that the allocation of transmission costs adequately reflects cost causation because the allocators are consistent with the basis upon which the AESO charges Fortis. Therefore, the Commission approves Fortis methodology and allocation of transmission costs, as filed. 7 Load forecast (billing determinants) 115. Among the approvals that Fortis requested were the following: (a) Approve the 2013 billing determinant forecast on existing rate structures, and as filed in the PBR Compliance Filing in Proceeding ID No. 2130, for purposes of establishing the target 2013 PBR revenue. (b) Approve the converted 2013 billing determinant forecast, reflecting the proposed new rate structures, for purposes of establishing the newly structured 2013 rates, recognizing that if Alternative 2 is accepted, then the Farm Rate 21 billing determinants will not require conversion. 51 Commission findings 116. In Decision , the Commission approved the 2013 billing determinant forecast on existing rate structures filed in the PBR Compliance Filing in Proceeding ID No for purposes of establishing the target 2013 PBR revenue. The Commission approves the converted 2013 billing determinant forecast, for purposes of establishing the newly structured 2013 rates, as discussed in Section 15.1 of this decision. 8 Revenue-to-cost ratios 117. The Commission considers that rates should generally be based on a revenue-to-cost ratio of 100 per cent, to the greatest extent possible, because inter-class subsidies are eliminated when customers in a rate class are required to pay all of, and only, those costs incurred to provide their service. Failure to achieve revenue-to-cost ratios at or near 100 per cent may result in some 51 Exhibit , Fortis argument, paragraph 172. AUC Decision (January 27, 2014) 25

32 customer classes bearing a disproportionate and potentially onerous share of costs. However, in some circumstances, the public interest may be served by allowing for some inter-class subsidies between specific rate groups. In addition, when revenue-to-cost ratios are altered significantly to achieve a 100 per cent revenue-to-cost ratio for all rate classes, rate shock may result for some rate classes, or the rates for some rate classes may become relatively unaffordable. Accordingly, the Commission has generally determined that it is desirable to gradually move the rates for all rate classes to within a 95 per cent to 105 per cent revenue-to-cost ratio Fortis proposed to continue to reflect a 100 per cent revenue-to-cost ratio for all rate classes for transmission For distribution, Fortis had initially proposed to move all rate classes to 100 per cent revenue-to-cost ratios in 2014, but also stated that it recognized that some classes, particularly irrigation and farm, would experience significant bill impacts The bill impacts resulting from Fortis initial proposal to implement rates based on a 100 per cent revenue-to-cost ratio for all distribution rate classes are set out in the following table based on Exhibit 15, Schedule 4.1-A3. Table 11. Revenue-to-cost ratio changes Rate Class Current 53 proposed Originally 26 AUC Decision (January 27, 2014) Distribution increase (%) Total transmission and distribution increase Residential (3) (2) FortisAlberta farm REA farm (58) n.a. FortisAlberta irrigation Exterior lighting (5) (5) Small general service Oil and gas General service (21) (8) Large general service (42) (8) Transmission connected (10) n.a In the application, Fortis noted the Commission s previous findings regarding the desirability of moving all rate classes to within a 95 per cent to 105 per cent revenue-to-cost ratio. 54 Fortis also submitted that the Commission had stated that it was not imperative that all rate classes be in that range prior to the commencement of a PBR scheme Fortis stated that, in accordance with cost causation principles, revenue-to-cost ratios should be set at 100 per cent or, if rate stability was a concern, should be transitioned over time to a reasonable range such as 95 to 105 per cent. 56 Fortis further stated that its previous cost allocation study was subject to a 20 per cent cap on increases by rate class and that this had resulted in little movement in revenue-to-cost ratios. Therefore, the 2010 cost allocation and rate Exhibit 15, Schedule 4.1-A3. Existing revenue-to-cost ratio based on 2013 revenue collected on 2013 PBR compliance rates. Exhibit 1, paragraph 39. Exhibit 1, paragraph 38. Exhibit 1, paragraph 45.

33 design had resulted in revenue-to-cost ratios that ranged from 77 per cent to 108 per cent and Fortis estimates that its revenue-to-cost ratios in 2012 will range from 72 per cent to 109 per cent In its application, Fortis indicated that it does recognize that some classes, particularly Irrigation and Farm, would experience bill impacts that are not insignificant in order to achieve 100% R/C ratios, if implemented as a single step increase in As such, if the Commission and/or parties still consider the proposed changes as excessive, FortisAlberta is open to considering alternatives to mitigate the rate impact for such classes in The Potato Growers Association submitted a letter to the Commission stating its concern about the very large bill impacts of the applied-for rates to its customers. These concerns are further discussed in the irrigation rate design section of this decision In response to a Commission information request, Fortis proposed three alternatives to help mitigate the rate impacts for farm and irrigation customer rate classes in In summary these alternatives were: (a) Impose a transmission and distribution (T&D) total rate cap of 15 per cent in each of 2014 and 2015 and retain existing farm rate structure. (b) Impose a T&D total rate cap of 10 per cent in each of 2014 and 2015 and retain existing farm rate structure. (c) Impose a T&D total rate cap of 0 per cent for Farm in 2014 and 10 per cent in 2015 and 10 per cent for both classes in 2015 and move ahead with the proposed farm rate structure changes in Fortis suggested that Alternative 2 (10 per cent T&D rate cap in each of 2014 and 2015) should be considered by the parties and the Commission as a reasonable, and possibly preferred, option for mitigating potential bill impacts to farm and irrigation customers In argument, Fortis noted that IPCAA had recommended: For 2014, under Alternatives 1, 2 and 3 there are two rate classes below 100% R/C ratio and the remaining rate classes are at or above 100%. Another approach under each alternative would be to calculate the R/C ratios for Farm and Irrigation classes with the rate caps described and then spread the cross-subsidy as an equal percentage of revenue across all rate classes i.e. set each rate class at the same R/C ratio In argument, Fortis noted that in rebuttal evidence it had submitted that: While, on the surface, such an approach may have some appeal in terms of simplicity, it does not accord with how such cross-subsidies have been funded in Phase II proceedings over the past decade; an approach that has been consistently approved by the Commission and its predecessor since the DTA. Further, the approach currently applied and approved by the Commission and subsequently proposed in this Exhibit 1, paragraphs 46 and 47. Exhibit 1, paragraph 17. Exhibit 43.01, AUC-FAI-001(a). Exhibit , Fortis argument, paragraph 63. Exhibit , Fortis argument, paragraph 46. AUC Decision (January 27, 2014) 27

34 Application, is the same approach that was suggested by IPCAA in the DTA. It was subsequently agreed to by FortisAlberta s predecessor, and approved by the Commission and applied since that time Fortis submitted secondly that, Table 4 of the IPCAA evidence included only one of the two transition years. By looking at a multi-year period, it becomes apparent that the new suggested approach by IPCAA is problematic in terms of rate variability for individual classes. 63 Fortis submitted that IPCAA s suggested approach would cause instability in rates and revenue to-cost ratios for some rate classes, because they move up and down rather than consistently toward the target 100 per cent ratio Fortis addressed the UCA s submission that farm customers will already see significant increases in order to move to 100 per cent revenue-to-cost ratios. The UCA submitted that this is not the time to also introduce a new rate design. As such, the UCA would support retaining the current rate design for farm customers Fortis recommended that Alternative 2 be approved by the Commission for Phase II adjustments for the years 2014 and Fortis submitted that the advantages of this alternative are: (a) It spreads the bill impact of movement towards 100 per cent revenue-to-cost ratios over a number of years instead of trying to achieve it in one large change while providing some certainty that the goal of minimizing cross-subsidization will be carried out over time. (b) With the approval of the distribution revenue allocators for 2014 and 2015, the mechanism is simple and aligns with the approved PBR process, which includes the annual filing for PBR rates and approval of billing determinants each year. (c) It does not cause additional intra-class bill impacts for Farm Rate 21 due to a changing rate structure while in the midst of overall rising rate levels for the class The UCA indicated that it was not opposed to Alternative The CCA submitted that the subsidy to farm and irrigation customers has been in place for many years and must cease as soon as possible. Of concern is that, under Fortis Alternative 1, the subsidy will continue to until 2015 in the case of farm and beyond that for irrigation. The CCA concurs with the UCA that it is not desirable to impose further increases arising from some structural changes proposed by Fortis to these rate classes. In lieu of such changes which would raise the rates for these two rate classes even further, the CCA suggested instead that, in the interest of getting these customer classes to 100 per cent revenue-to-cost ratios as soon as possible, the constraint on the 15 per cent T&D cap be lifted to allow the revenue-to-cost ratios to reach 100 per cent by no later than the end of Exhibit 90.01, Fortis rebuttal evidence, page 3. Exhibit , Fortis argument, paragraph 50. Exhibit , Fortis argument, paragraph 52. Exhibit , Fortis argument, paragraph 56. Exhibit , Fortis argument, paragraph 65. Exhibit , Fortis argument, paragraph 68. Exhibit , UCA argument, paragraph 18. Exhibit , CCA argument, paragraph AUC Decision (January 27, 2014)

35 134. IPCAA submitted that its recommended approach provides equal and fair allocation of farm and irrigation subsidization among all customers Fortis submitted that this newly suggested IPCAA approach, as further modified in its Argument (Exhibit ), is clearly unworkable for this proceeding, as it leaves open and uncertain how much revenue is allocated to particular rate classes for purposes of designing 2014 and 2015 rates. This would be left undetermined until such time as IPCAA explains what new method it suggests to refine or further adjust for a smooth transition Fortis recommended that IPCAA s proposed change to the method of revenue allocation be rejected by the Commission Fortis noted in reply argument that the evidence of the PGA appears to further support Alternative 2 as the option which would best mitigate potential bill impacts for irrigation customers such as the PGA In reply argument, Fortis continued to support Alternative 2 (T&D rate cap of 10 per cent and 10 per cent) for all classes in each of 2014 and 2015 and retention of the existing farm rate structure Fortis requested approval of its Alternative 2 recommended distribution revenue allocation percentages to be used for purposes of setting distribution rates for 2014 and 2015 in the annual PBR rate filings. In conjunction, the initially proposed changes to the farm rate structure would not be implemented at this time. 75 Commission findings 140. In response to AUC-FAI-050, Fortis explained the difficulty in determining the actual revenue-to-cost ratios for Fortis stated: In summary, if one makes the assumption that, with above-average load growth in the rate class, there should be a corresponding above-average growth in costs, then it is likely that the Rate 61 R/C ratio in 2013 would likely fall between 121 per cent and 126 per cent. However, without a revenue requirement or cost allocation study for the 2013 PBR test year, that assessment cannot be proven Fortis proposed, as an alternative, to cap the total transmission and distribution rate impact that results from moving the revenue-to-cost ratios towards 100 per cent. The Commission is not opposed to a capping mechanism to mitigate rate shock which may result from adjusting the revenue-to-cost ratios towards 100 per cent Given the limitations of the CAM model discussed previously and the impact of changed assumptions including the enhancements to the model and the exclusion of the Y factor portion of load settlement costs, the Commission considers the revenue-to-costs ratios to be indicative of Exhibit , IPCAA reply argument, page 5. Exhibit , Fortis reply argument, paragraph 29. Exhibit , Fortis reply argument, paragraph 32. Exhibit , Fortis reply argument, paragraph 37. Exhibit , Fortis reply argument, paragraph 43. Exhibit , Fortis reply argument, paragraph 44. Exhibit , AUC-Fortis-50(b). AUC Decision (January 27, 2014) 29

36 a directional change, but not sufficiently precise to allow for the proposed adjustments to achieve a 100 per cent revenue-to-cost ratio for all rate classes Accordingly, although the Commission considers that revenue-to-cost ratios should be adjusted toward 100 per cent, the Commission must also consider possible rate shock as a result of adjusting revenue-to-cost ratios. The Commission directs Fortis to adjust the revenue-to-cost ratios for all rate groups as close as possible to the generally accepted 95 per cent to 105 per cent revenue-to-cost ratio range, such that the average total bill impact for each rate class does not exceed a ten per cent increase. 9 Transmission rate design 144. Fortis proposed to continue to set the total revenue-to-cost ratios for the transmission component of each rate class to 100 per cent. Fortis proposed to retain its transmission rate design to recover all AESO transmission tariff demand charges through a single ratcheted demand charge, for those rate classes that are billed based on demand. Fortis indicated that designing rates in such a manner recognizes the intended flow-through nature of transmission costs, as well as the ongoing increases to the AESO transmission tariff and the corresponding proposed increase to the Transmission Component of FortisAlberta s rates in FortisAlberta s PBR Compliance Filing (Proceeding ID. 2130) In response to CWSAA evidence, Fortis proposed an alternative rate design for Rate classes 44, 45, 61 and 63 to recover bulk transmission charges. The Commission discusses this proposal in the following section. 9.1 Recovery of bulk transmission charges 146. CWSAA considered that Fortis bulk transmission cost of service study accurately reflected the current AESO tariff structure. These costs are allocated to rate classes based on each rate class average MW/month of billing capacity. As this allocation is consistent with the AESO tariff, changes are neither necessary nor appropriate. Notwithstanding, Fortis current tariff applies a twelve month ratchet to all demand charges, including bulk transmission charges and as a result, customers with seasonal load variation pay considerably more than their actual bulk transmission cost of service Customers without demand meters pay bulk transmission costs through an energy charge. Using an energy charge to recover bulk transmission costs from all customers would provide consistency across rate classes and would not require a new billing determinant. However, transferring this cost recovery to an energy charge would create significant rate dislocations across a large number of customers. This approach would also not be consistent with the principle of cost recovery based upon cost causation, since energy is not the basis on which these costs are charged to Fortis by the AESO CWSAA pointed out that the highest metered demand in the billing period is a billing determinant which is already available, used in billing processes, and more closely matches the way in which bulk transmission demand charges flow to Fortis. Since this demand based billing determinant is quite consistent with the existing ratcheted billing determinant, its application creates less rate dislocation than an energy based billing determinant. 77 Exhibit 1, application, paragraph 155, PDF page AUC Decision (January 27, 2014)

37 149. A new demand determinant would require changes in Fortis billing system, and attendant changes in tariff bill files and retailer billing processes. Fortis indicated that its costs to implement the change proposed by CWSAA would not be large and are immaterial in comparison to the magnitude of cross subsidization embedded in current rates In principle, customers with interval meters could be billed on demand at the time of the system peak, which would accurately reflect the AESO tariff. However, individual customer demand at the time of system peak is not currently used as a billing determinant for distribution-voltage customers. It would be administratively complex to use this transmission system peak demand determinant in large scale customer billing Recovery of bulk transmission costs through an un-ratcheted demand charge, based on the highest metered demand in the billing period (i.e. kw of billing period capacity) would provide consistent treatment for all distribution level customers and would be administratively simple to implement. CWSAA therefore requested that the Commission direct Fortis to implement this change in its distribution tariff to more fairly and accurately reflect cost causation Noting that the benefits to CWSAA customers would be materially greater under the energy-based design proposed by Fortis, CWSAA requested that the Commission direct Fortis to implement whichever of these two options the Commission considers most appropriate under the circumstances While Fortis proposal to collect bulk transmission charges on an energy basis would remove charges for months when seasonal loads do not operate, it creates significant cost shifts between high and low load factor loads for Rate 61 and 63 customers. IPCAA submits these impacts can be significant and they have not been examined in evidence. Furthermore, load factor (and therefore total energy) has no relationship to AESO bulk transmission charges. Accordingly, IPCAA argued it is not an appropriate basis for billing bulk transmission charges. As a result, IPCAA strongly opposed Fortis proposal to collect bulk transmission charges on an energy basis. IPCAA is supportive of CWSAA s position to collect bulk system charges on a non-ratcheted NCP demand basis Fortis submitted that its offer of a rate design alternative, to collect bulk system CP demand charge costs through an energy charge, is a reasonable way to resolve CWSAA s legitimate concern around having un-ratcheted demand costs collected through a ratcheted demand charge Fortis recommended that the Commission approve and direct Fortis to use the alternative energy-based rate design, proposed in CWSAA-FAI-005, to collect bulk system CP demand charge costs for all rates that have a demand charge in the transmission component of Fortis distribution tariff. This rate design would be implemented for purposes of Fortis next annual PBR rates filing for 2014 (assuming a Phase II decision prior to that filing), and/or any other future compliance filings with respect to this application After evaluating the options with respect to bulk system CP demand charges, CWSAA requested that all of rates 61 and 63 be redesigned to include an additional transmission billing determinant and charge based on un-ratcheted demand to accommodate the particular cost 78 Exhibit 118. AUC Decision (January 27, 2014) 31

38 causation concerns of unique seasonal customers. Fortis did not support such a change. In response to AUC-FAI-043(f), Fortis stated: FortisAlberta does not disagree that the AESO tariff Bulk System CP charges are incurred using a non-ratcheted billing determinant, whereas these costs are recovered for some rate classes using a ratcheted demand billing determinant. However, the AESO tariff structure is overly complex with many components, billing determinants and blocks, and to attempt to mirror the tariff precisely through the Transmission Component of the Distribution Tariff would be unworkable. For example, residential customers do not have demand meters, and as a result, all AESO Tariff charges are recovered via an energy charge in FortisAlberta s rates CWSAA requested Fortis to discuss the pros and cons of implementing a billing determinant change, and the potential timing of any such implementation, particularly in respect of rate classes other than rates 61 and 63. Fortis responded in CWSAA-FAI-004(f): Introducing a non-ratcheted demand billing determinant for these rate classes for the Transmission Component alone does increase the complexity of the rates and quantity of data required to be provided to retailers in the tariff bill files. One alternative, if the objective is to collect Bulk System CP Demand Charge costs through a non-ratcheted billing determinant is to continue to allocate to rate classes based on each rate class forecast monthly CP demand but rather than classifying the costs as a demand-related cost component, classify the allocated costs as energy related for FortisAlberta DT purposes. In other words, collect the allocated CP Demand Charges through an energy charge, which would not require an additional billing determinant Fortis offered to use the current energy billing determinant for purposes of collecting AESO transmission bulk CP demand charges, and recommended that the Commission not accept CWSAA s recommendation to add another demand billing determinant to the transmission component for all Fortis Rate 61 and 63 customers. Such a recommendation would complicate the distribution rate structures in an attempt to mirror an overly complex AESO tariff structure Fortis considered that its proposed solution would not introduce any more billing determinants into the rate structure and provides a reasonable middle-ground rate solution that addresses CWSAA s concern to reflect cost causation Fortis also proposed that, if accepted by the Commission, the energy recovery approach to collect bulk CP demand charges be consistently applied to rates 41 and 45, in addition to rates 61 and 63. CWSAA s approach would result in different and inconsistent rate recovery treatment of these costs among rate classes Should the Commission not accept Fortis proposed energy charge solution, then Fortis recommended that the status quo approach continue to be applied in rate design as applied for in the application. Commission findings 162. CWSAA argued that there is a mismatch between the AESO tariff and the Fortis distribution tariff. Fortis did not disagree that the AESO tariff bulk system CP demand charges are incurred using a non-ratcheted billing determinant, whereas these costs are recovered for some rate classes using a ratcheted demand billing determinant. 32 AUC Decision (January 27, 2014)

39 163. The Commission has reviewed the 2013 AESO distribution forecast of connected load costs as set out in Schedule 3.1A of the application, net of under-frequency load shedding credit and interchange charges and has extracted the information in the following table: Table 12. Commission summary of 2013 forecast AESO charges to Fortis Charge basis Amount ($000) Percentage (%) Charge type Bulk system charge Coincident metered demand($/mw/month) Coincident demand 94, Metered energy ($/MWh) Energy 20,938 6 Local system charge Billing capacity ($/MW/month) Ratcheted demand 38, Metered energy ($/MWh) Energy 8,769 2 Point of delivery charge Substation fraction ($/month) Fixed charge 14,660 4 MW of billing capacity ($/MW/month) Ratcheted demand 114, Operating reserve charge Pool price ($/MWh) Energy 54, Voltage control charge ($/MWh) Energy 9,127 3 Other system support service charge ($/MW/month) Energy 2,171 1 TOTAL 358, Of the total AESO tariff charges of $358,536,000, bulk system charges based on CP demand were $94,930,000, which accounts for 26 per cent of the total. Further, 43 per cent of the total AESO charges are ratchet- based, with 27 per cent being energy based, and four per cent of the total AESO charges being fixed Given the apparent mismatch between transmission billings from the AESO to Fortis and the collection of transmission costs from Fortis customers, the Commission considers that the issue before it is whether the transmission component included in Fortis distribution rates for Rate 61 and Rate 63 customers mirrors the AESO s transmission charges to the greatest extent possible. The Commission agrees that there is a mismatch requiring further consideration To address the mismatch, three alternatives were proposed to the Commission: (a) CWSAA proposed that a third billing determinant of non-ratcheted demand to the transmission component of Fortis distribution rate structure for Rate 61 General Service and Rate 63 Large General Service customers to collect CP demand charges be implemented. (b) Fortis proposed to collect the allocated CP demand charges through an energy charge for Rate 41 Small General Service, Rate 45 Oil and Gas Service, Rate 61 General Service and Rate 63 Large General Service customers. (c) In the event that the Commission did not accept either of these proposals, Fortis recommended maintaining the status quo In considering these options, one concern is rate dislocation or intra-rate class impacts. CWSAA summarized Fortis evidence that the dislocations arising from an energy determinant would be considerable. While IPCAA submitted that these impacts had not been examined in evidence, the Commission recognizes that in Exhibit 105, Fortis provided the impacts on Rate 61 AUC Decision (January 27, 2014) 33

40 and 63 customers of using a third billing determinant as recommended by CWSAA and Fortis alternate proposal of allocating costs based on energy While CWSAA used a threshold of 10 per cent increase or decrease on the total bill for Rates 61 and 63 to demonstrate the intra-rate class impacts and argue against the use of an energy based billing determinant, the Commission considers that this analysis may not be sufficiently sensitive and has repeated the analysis using a five per cent threshold. The Commission has recast the charts presented by CWSAA showing the impact of energy and demand as billing determinants adding a five per cent threshold. Table 13. Impacts of energy as billing determinant Commission 5% Threshold CWSAA 10% Threshold Rate 61 Rate 63 Rate 61 Rate 63 Sites with an increase greater than 5% or 10% TOTAL sites with an increase % sites with an increase greater than 5% or 73% 78% 49% 51% Commission 5% threshold CWSAA 10% threshold Rate 61 Rate 63 Rate 61 Rate 63 Sites with a decrease greater than 5% or 10% TOTAL sites with a decrease % sites with a decrease greater than 5% or 10% 82% 81% 63% 70% Table 14. Impacts of demand as billing determinant Commission 5% threshold CWSAA 10% threshold Rate 61 Rate 63 Rate 61 Rate 63 Sites with an increase greater than 5% or 10% TOTAL sites with an increase % sites with an increase greater than > 5% or 10% 65% 49% 0% 0% Commission 5% threshold CWSAA 10% threshold Rate 61 Rate 63 Rate 61 Rate 63 Sites with a decrease greater than 5% or 10% TOTAL sites with a decrease % sites with a decrease greater than> 5% or 10% 73% 69% 55% 47% 169. Based on this analysis, it appears that the use of energy as a billing determinant will result in slightly higher intra-rate class impacts. The Commission concludes that customers with identical peak demands can have different load factors and would experience quite different impacts if charges are transferred to an energy determinant The Commission also considered whether the proposed determinants reflect how the AESO levies these charges to Fortis. For bulk system CP demand charges, the Commission recognizes that CWSAA s proposal only partially reflects the basis on which Fortis receives these charges. In contrast, Fortis proposed energy charge would not be consistent with the manner in which the AESO levies these charges, as bulk system CP demand charges are not 34 AUC Decision (January 27, 2014)

41 based on energy. Thus, using an energy determinant would be inconsistent with the goal of having rates based on cost causation Another factor considered was the administrative burden required to implement the proposals. Fortis recommendation to base bulk system CP demand charges on energy would likely be administratively easier to implement, because Fortis would simply classify these costs as energy-related. In contrast, using non-ratcheted, non-coincident, metered demand as a billing determinant would require billing system changes that would take some six weeks to implement at an approximate cost of $10,000 to $15,000. Fortis also indicated that CWSAA s proposal will increase the complexity of the resulting rates and volume of data to be provided to retailers 172. In making its recommendation, Fortis indicated that, for consistency, the use of energy to recover bulk CP charges should also apply to customers in rates 41 and To the extent that AESO charges can be mirrored in rates charged by Fortis to customers, the rate structure would be consistent with cost causation. Fortis submitted that the AESO s tariff structure is overly complex with too many components, billing determinants and blocks, such that an attempt to mirror the tariff precisely through the transmission component of the Fortis distribution tariff would be unworkable. Among other things, it proposed that mirroring would require Fortis to bill customers on their contribution to the coincident transmission peak in Alberta. Customers would not know in advance when the peak would occur as it would depend on the loads of all other customers in the province IPCAA, representing Rate 63 customers, supported the change recommended by CWSAA The Commission concludes that mirroring of the AESO s charges in Fortis rates is consistent with cost causation and directs Fortis, in its refiling, to reflect CWSAA s proposed new billing determinant of un-ratcheted kw of billing period capacity for rates 61 and 63. The Commission recognizes that this direction will require billing system changes with an approximate cost of $10,000 to $15,000. However, the Commission considers the benefits of adopting CWSAA s proposal outweigh the related costs The Commission is not convinced that CWSAA s proposal should apply to rates 41 and 45 at this time as application to those customer classes has not yet been sufficiently explored. The Commission directs Fortis to consider whether CWSAA s proposal should apply to Rate 41 and Rate 45 in its next Phase II application The Commission continues to accept Fortis proposed transmission rate design for the remainder of its transmission rates. The Commission agrees that a 100 per cent revenue-to-cost ratio supports the flow through nature of transmission costs from the AESO to customers and leaves both customers and Fortis revenue neutral with respect to changes in transmission costs. This approach also mirrors the demand ratchet rate structure of the AESO s transmission charges. Accordingly, the rate design for Fortis transmission rates other than Rate 61 and Rate 63 is approved as filed. AUC Decision (January 27, 2014) 35

42 10 Distribution rate design 178. In the following section proposed changes to the distribution charges rate design for each rate class are discussed Rate 11 - residential service 179. Fortis applied to restructure its residential distribution rate to /kilowatt hour (kwh) for the first 30 kwh per day and /kwh for all daily consumption over 30 kwh. The higher charge for usage above 30 kwh per day reflects the additional costs caused by customers with larger loads, such as a larger transformer capacity. 79 These charges would be in addition to the distribution service charge applied-for increase from $64.69/day to $0.789/day. The rate structure would help to minimize intra-class cross-subsidization better supporting the user-pay principle Fortis clarified that in implementing the charge greater than 30kWh/day, all consumption greater than 900 would be billed at the second block rate The CCA submitted that the proposed change added complexity to the residential rate structure and that customers have neither requested these changes, nor are the proposed changes the result of a Commission order or direction In considering the proposed change, the CCA pointed out that Fortis has not completed any studies that demonstrate that the proposed changes will cause customers to be more responsive to prices or determined whether other jurisdictions in North America have similar rate structures. Further, the proposed change appeared contrary to the guiding principles used by Fortis in designing rates for residential customers, which were: little responsiveness to prices generally believe that same price should apply for same end-use prefer simple rates very sensitive to rate impacts more intra-class averaging is acceptable 183. The CCA submitted that it appeared that the only rationale in support of these changes is to increase the relative portion of revenues collected from fixed charges. Further, as a greater portion of revenue would be derived from fixed charges, it appeared that the proposed rate structure would reduce Fortis exposure to weather-related earnings variability by recovering revenues from the fixed charge Fortis responded that, for residential customers, the average kwh/day consumption has a close relationship to their average daily demand and that this validates the use of consumption as a proxy for demand. The following graph for the residential rate class provided in response to AUC-FAI-024 (c) indicates that approximately 80 per cent of residential customers have an average demand less than five kw, and average consumption of 30 kwh/day Exhibit 1, paragraphs 164 and 166. AUC-FAI-23(a). 36 AUC Decision (January 27, 2014)

43 185. Choosing a lower 20 kwh/day breakpoint would cause more than half of these customers to be billed for part of their consumption on the second energy block. On this basis, Fortis considered that a rate block starting at 30 kwh/day would promote monthly rate stability Fortis provided by the number of customers that would be subject to the second block charge if this rate had been in place, for the months from January 2011 to December The number of customers that would have been impacted ranged from 13 per cent to 35 per cent Fortis submitted further that its current minimum transformer size is 15 kva. Customers who use greater than 30 kwh/day have a rapidly increasing average demand which quickly approaches 10 kva. This suggests that, as a customer s average usage increases above 30 kwh/day, there would be an increased likelihood that an upgrade to a larger and more costly transformer size would be required. The distribution facilities required to serve residential customers, such as poles and conductors, do not change with customer demand. However, high usage customers do require larger transformers and their rates should reflect this fact. Therefore, Fortis argued, increasing a customer s cost when the customer consumes more than 30 kwh/day reflects cost causation Fortis argued that the proposed rate structure supports the guiding principles highlighted by the CCA as follows: Exhibit , Fortis reply argument, paragraph 49. UCA-FAI-12(b). Exhibit , Fortis reply argument, paragraph 49. AUC Decision (January 27, 2014) 37

44 (a) Little responsiveness to prices. Most customers have relatively low consumption. 80 per cent of customers use less than 30 kwh/day or 900 kwh/month, meaning there is little opportunity to significantly reduce usage. However, high usage customers (with consumption in excess of 1200kWh per month), which account for 10 per cent of the total number of residential customers, have would be more suited to responding to price signals. The proposed rate structure gives an appropriate signal to these high usage customers to reduce their loads, thereby promoting conservation. (b) Generally, the same price should apply for the same end-use. Eighty per cent of customers would be billed on the first energy block and therefore would see relatively similar prices. Further, high usage customers with consumption over 2,000 kwh/month require transformers much larger than those typically used to serve other residential customers. Customers with very large consumptions may have extra living quarters, heated indoor pools, hockey rinks, carriage houses, etc., which would not be considered typical by most residential standards. Therefore, these high energy usage customers should be charged more to recover the additional costs incurred to serve them. (c) Simple rates are preferable. With regards to residential customers preference for simple rates, most customers are unaware of the rate structure they are billed on. These customers typically only see a Distribution Charge line item on their monthly electric bill from their retailer. If a residential customer requests more detail regarding the customer s bill, the proposed rate is easy to describe and calculate. (d) Customers are very sensitive to rate impacts. Fortis noted that there has been an increased awareness on how mass market customers value predictability in their electricity rates, as highlighted in the Retail Market Review Committee report prepared for the Alberta government in This concern aligns with the fact that distribution costs are also fixed and do not vary by month or by the kwh consumed. The rate structure proposed by Fortis will result in more rate predictability for the majority of residential customers. (e) More intra-class averaging is acceptable. The proposed rate averages costs for the majority of customers. The impact to intra-class averaging due to the rate structure change will result in the relatively few high usage customers bearing less of the cost required to serve them than is borne by lower usage customers While the CCA argued that the proposed rate change would reduce revenue variability, Fortis considered that variability might increase by putting more revenue at risk in the second variable energy block, which has a higher cent/kwh charge. Commission findings 190. The Commission has considered the proposed rates in comparison to the rates that were approved in Decision Exhibit , Fortis reply argument, paragraph 50. Decision : 2012 Performance-Based Regulation Second Compliance Filings, April 1, 2013 Interim Distribution Rates for each of AltaGas Utilities Inc., ATCO Electric Ltd., ATCO Gas and Pipelines Ltd., EPCOR Distribution & Transmission Inc. and FortisAlberta Inc., Application No , Proceeding ID No. 2477, March 22, AUC Decision (January 27, 2014)

45 191. The Commission has also considered the structural change proposed for the residential rate, as noted in the table below: Table 15. Fixed charge First 30 kwh/ day (912 kwh /month) (Just one block on existing rates) Additional kwh Illustration of residential rate and rate structure changes 2013 interim Proposed 2014 Increase /day 78.9 /day 22% ($19.68/month) (24.00/month) /kwh /kwh (for all kwh) minus 48% /kwh (for all kwh) /kwh 192. The Commission finds that there is insufficient information on the record of this proceeding to support the proposed change to the rate design for Rate 11. The Commission is not opposed in principle to a multi-block rate design, however the proposed breakpoint and the split between the proposed fixed and variable charge has not been adequately supported The Commission is not convinced that Fortis proposed residential Rate 11 rate structure and the resulting amounts are justified at this time. On this basis, the Commission directs Fortis to maintain the existing residential rate structure In its next Phase II application, should Fortis propose changes to the residential rate structure, it is directed to provide analysis regarding the need for rate structure changes and a sensitivity analysis of the customer impacts of different block thresholds Rate 21 - farm service 195. Large increases were proposed for farm customers due to the results of the cost of service study and particularly, the proposal to move the revenue-to-cost ratio to 100 per cent Fortis also proposed to change its rate design to no longer charge differentially based on farm breaker size. In response to AUC-FAI-021, Fortis indicated that there is minimal cost difference between different breaker sizes. For farms, breaker size has been historically used in rate design as a proxy for demand. With the introduction of AMI it became apparent that breaker size is a poor proxy for demand. Further, there is a very close relationship to farm consumption and average demand. The proposed rate structure will reduce intra-class subsidization such that farms with the same consumption will pay the same distribution charge Fortis proposed to differentiate the rate depending on whether the customer was demand metered or not. 2% 86 AUC-FAI-021(d). AUC Decision (January 27, 2014) 39

46 198. The applied-for structural changes on each element of the farm rate are set out in the tables below: Table 16. Proposed farm rate structure as revised March 4, 2013 Restructured 2013 Rates S ummary (Rounded) ($/Day) Filed January 18, 2013 Blk 1 Blk 2 Blk 3 $/site Blk 1 Rate Blk 2 Rate Blk 3 Rate Size Size Size R11 - Residential 0.79 $/site 7.89 /kwh <=30 kwh >30 kwh R Farm (kwh Metered) 1.32 $/site 7.81 /kwh All kwh R Farm (kva Metered) 1.32 $/site 0.53 $/kw 0.37 $/kw >3-15 kw >15 kw Restructured 2013 Rates S ummary (Rounded) ($/Day) Revised Blk 1 Blk 2 $/site Blk 1 Rate Blk 2 Rate Size Size R11 - Residential 0.79 $/site 0.94 /kwh 1.86 /kwh <=30 kwh >30 kwh R Farm (kwh Metered) 2.79 $/site 3.56 /kwh All kwh R Farm (kva Metered) 2.79 $/site 0.42 $/kw >3 kw 199. Interveners expressed concerns that the change could result in cross-subsidization of the farm rate class by other rate classes. In argument, Fortis recommended that, considering the concerns of interveners, the existing farm rate structure should be retained, subject to a 10 per cent transmission and distribution rate cap for all customers. 87 Commission findings 200. As the proposed rate structure will reduce intra-class subsidization such that farms with the same consumption will pay the same distribution charge, the Commission approves the proposed rate structure change for farm service Rate 26 - irrigation service 201. Large rate increases were proposed for irrigation customers. The increases were due to the results of the cost of service study which, as explained in Section 5 above, resulted in allocating significantly more costs to irrigation customers as compared to the previous cost allocation study. Table 17. Illustration of irrigation rate and rate structure changes 2013 interim Proposed 2014 Increase Demand charge /kw-day 19.4 /kw-day 78% ($3.3093/kW- month) ($5.901/kW- month) Energy charge per kwh No energy charge Minus 100% 202. Fortis indicated that it proposed minimal changes to this rate structure because there was a large increase due to the revenue-to-cost ratio increase The PGA noted that the proposed increase is 79 per cent. The proposed demand charge increases by about 78 per cent but was mitigated by the elimination of the energy charge in the Fortis proposal Exhibit 109, Fortis argument, paragraph 65. Exhibit 1, application, paragraph AUC Decision (January 27, 2014)

47 204. The PGA stated that: The Alberta Potato Industry not only has to compete on a National level, but we are also competing with a huge U. S. market as well. Over 70% of our processing product goes to the Unites States on an annual basis. Currently, we are able to remain somewhat competitive with this market as our Cost of Production remains manageable. We are, unfortunately, at a climate disadvantage compared to the Pacific Northwest where yields are in excess of 30 tons/acre, compared to our 18 tons/acre. Our quality is what has allowed us to remain competitive, along with our management practices The PGA submitted that: Every year our growers are faced with increased input costs and what appears to be an ever-increasing request for an increase in power costs. Electricity costs alone are now between 5-6% of their total operating costs. If this trend continues, Alberta runs the risk of losing an industry that is generating close to one billion dollars annually. If we cannot compete with our American counterparts then our industry will cease to exist The PGA did not comment on the changes to the rate structure by which the charge per kwh is proposed to be eliminated. Commission findings 207. Fortis has applied to remove the energy through-put charge in the distribution charges. The Commission finds that there is insufficient information on the record of this proceeding to support the proposed change to the rate design for Rate 26. Therefore, Fortis proposal to remove the energy through-put charge in the distribution charges is denied Rates 31, 33 and 38 - lighting service 208. Fortis proposed to remove the energy charge portion of distribution charges for lighting services, which was a fixed charge based on fixture size. The distribution charges would then be fixed for lighting. Commission findings 209. The Commission considers the proposed change to the distribution rate structure to be a change in terminology, rather than a change in structure. Because these services are unmetered, it is reasonable for lighting service distribution charge to be fixed based on fixture size. The Commission approves the proposed change for the lighting rate class Rate 41 - small general service 210. Fortis proposed to restructure Rate 41 as illustrated in the table below. The applied-for rate structure increases the service charge and demand charge (over 3kW minimum) and removes the energy charge. The consequence is that customers with low energy consumption in relation to demand will face higher charges Fortis indicated that there have been concerns about the complexity of this rate which retains its unique kwh/kw of capacity/day (load factor) rate structure from the 1990s. The proposed rate is designed to align with other rates that have a service charge based on a rate Exhibit 95.01, Potato Growers of Alberta Evidence. Exhibit 95.01, Potato Growers of Alberta Evidence. AUC Decision (January 27, 2014) 41

48 minimum of 3 kw, to recognize the fixed nature of costs for standard customers. Additional charges for loads greater than this threshold amount are applied consistent with other rates. The current ratchet structure remains unchanged. 91 Table 18. Illustration of small general rate and rate structure changes Service charge (Presented as a demand charge but is effectively a service charge due to the 3 kw minimum) Demand charge (over 3 kw min.) 2013 interim /day ($28.82/month) 2014 proposed $1.1836/day (36.00/month) Increase /kw-day / kw-day 28% (5.846/kW-month) (7.46/kW-month) Energy charge for first kwh per kw per day /kwh No energy charge minus 100% Energy charge for additional kwh No charge No charge Commission findings 212. Fortis has applied to remove the energy charge portion of distribution charges for Rate 41. The Commission finds that there is insufficient information on the record of this proceeding with respect to cost causation to support the proposed change to the rate design for Rate 41. Therefore, Fortis proposal to remove the energy charge in the distribution charges is denied Rate 45 - oil & gas service 213. Fortis proposed to reduce the demand charge block to 3kW from 5kW of capacity to be consistent with other rate classes. The proposed rate structure for Rate 45 is set out in the following table: 25% Table 19. Illustration of oil & gas rate and rate structure changes 2013 interim 2014 proposed Service Charge $1.747/day $2.071/day ($53.13/month) ($63.00 /month) Demand Charge (Next kw over 3 kw /kw-day / kw-day min.) ($17.709/kW- month) ($15.120/kW-month) Applies for 4th and Applies for 4th through 15th 5th kw only kw only Demand Charge (All additional kw) / kw-day /kw-day ($11.909/kW-month) ($14.288/kW-month) Energy charge No charge No charge Increase 19% minus 15% on 4th and 5th kw 27% on 6 th through 15th kw 12% on all kw over 15 Commission findings 214. The Commission supports the alignment of the block size with other rate classes and finds the rate structure changes proposed for the oil and gas rate class to be reasonable. The rate structure for Rate 45 is approved as filed. 91 Exhibit AUC Decision (January 27, 2014)

49 10.7 Rate 61 - general service 215. Fortis proposed to leave the rate structure for Rate 61 unchanged. The Commission agrees that it is reasonable to retain the rate structure for Rate 61. The Commission approves the rate structure for Rate 61, as filed Rate 63 - large general service 216. In its application, Fortis proposed to restructure Rate 63 by adding a fixed customer charge and changing the per kilometre charge to a kilometres times demand charge to recognise that both demand and distance impact the cost allocation In response to AUC-FAI-033, Fortis explained that this change would favor customers with lower demand while customers with higher demand would see a relative increase in the distance component of their bill In its evidence, IPCAA discussed its understanding of how various costs were of a fixed nature rather than variable nature. Factors that contribute to the cost of providing service include demand and distance of the dedicated line, or distance from the transmission system. It submitted that neither the existing rate design nor the proposed rate design is ideal In its rebuttal evidence, Fortis indicated that it did not disagree with IPCAA s characterization. Given IPCAA s concerns about the widely varying impact of the rate structure changes on various customers, Fortis indicated that it was open to retaining the existing rate structure, albeit with the addition of a customer charge In response to AUC-FAI-044, Fortis provided a revised rate schedule for Rate 63 which retained the existing rate structure but incorporated a customer charge at the level suggested by IPCAA. Fortis provided a graph which indicated that the impact of the rate structure change on individual customers ranged from about minus eight per cent to about five per cent IPCAA recommended that: [c]ustomer charges should be reviewed to ensure alignment with the principle of cost causation. Costs that are asset-related but charged on a $/customer basis should be removed and recovered elsewhere in the rate structure. The recommended Level of Customer Service charge allocation to Rate 63 is $842, IPCAA recommended that only transformer costs and series capacitor costs should be collected through the demand charge. The cost of dedicated overhead lines and underground facilities ought to remain in the distance related charge The CCA submitted that, if the Commission were to approve any changes to the structure or design of Rate 63, Fortis should be directed to reflect corresponding changes in the structure and design of the investment in customer facilities for Rate Fortis submitted that IPCAA s recommendation under-allocates information and technology (IT) costs and corporate general and administration (G&A) costs to the customer Exhibit , IPCAA argument, page 4. Initially the proposed customer rate had been $1.705 million including a customer service charge of $842,000, a IT cost of $175,000 and general and administrative expenses of $688,000. Exhibit , IPCAA argument, page 5. Exhibit , CCA argument, paragraph 48. AUC Decision (January 27, 2014) 43

50 charge and that a portion of these costs should be allocated to the distance charge to better reflect cost causation. Fortis submitted that IT and corporate G&A costs support customer service related functions that are appropriately allocated to the customer charge. The method used to allocate corporate G&A and IT costs to the customer charge and distance charge in AUC-FAI-044(d) 95 is reasonable and appropriate. Accordingly, Fortis continues to recommend approval of the alternate rates filed in AUC-FAI-044(d) as stated in its argument Fortis submitted that the difference in IPCAA s recommended allocated costs to the customer charge and the allocated costs determined by Fortis in response to AUC-FAI-044(d) is $475,000 ($1,317,000 - $842,000) which is less than five per cent of the rate class total allocated cost. Fortis considered the creation of a customer charge, even if its allocated costs are understated, as a positive step in having the rate better reflect cost causation, which is a principle IPCAA also recommends. Accordingly, if the Commission is persuaded by IPCAA s argument, Fortis would not be opposed to IPCAA s recommendation to allocate $842,000 to the customer charge, for the purpose of this application. Fortis proposed that this could be easily accomplished by the Commission directing it, as part of a compliance filing, to replace Fortis recommended $1,317,000 customer charge allocation amount in attachment AUC-FAI , Column K line 3 97 with IPCAA s recommended $842,000 customer charge allocation amount in order to determine the revised billing components percentages to be used for rate design Fortis agreed with the CCA that any change from the distance charge to distance multiplied by demand charge should have a corresponding change to investment structure. 99 Table 20. Illustration of large general rate and rate structure changes Service charge (Presented in the current rates as a demand charge but is effectively a service charge due to the 2,000 kw minimum) Demand charge (Effectively applies above 2,000 kw) 2013 interim $39.40 per day for minimum demand ($1198 per month) 1.97 /kw-day (59.92 /kw- month) Proposed 2014 (as modified in Exhibit ) $ per day plus $21.28 per day for minimum demand charge. Total $ per day ($1342 per month) kw-day (32.36 /kw- month) Charge for each contact kilometre $ /km-day $8.6428/km-day ($520.42/km-month) ($262.89/km-month) Energy charge No charge No charge Increase (%) 15 minus 46 Minus 49 Commission findings 227. The initial application for the service charge included IT and G&A components, as well as a customer service component. Based on input by interveners, Fortis revised its position to eliminate the IT and G&A components from the proposed service charge. Fortis expressed the view that the introduction of a service charge (at a reduced level) was preferable to no customer charge. The Commission considers it reasonable to charge customers a fixed fee related to the customer service component. The Commission finds the alternative rate structure proposed by Exhibit Exhibit , Fortis reply argument, paragraph 65. Exhibit Exhibit , Fortis reply argument, paragraph 66. Exhibit , Fortis reply argument, paragraph AUC Decision (January 27, 2014)

51 IPCAA and accepted by Fortis to be reasonable. The Commission approves the introduction of a service charge based on $842,000 of customer service costs Fortis also applied to change its per kilometre charge to a kilometres times demand charge. The Commission considers there is not sufficient information on the record to support the applied-for change. There is no information on the record of the proceeding on the extent to which local property cost differences and distance from the transmission system have been reflected in customer contribution. Likewise, there is no information on the extent to which distance from the transmission system is relevant to cost causation in the case of customers with dedicated feeders and customers that tap into existing distribution lines. Accordingly, the Commission denies the proposed rate structure for Rate 63, and directs that the existing rate structure be retained, with the exception of the proposed change to the service charge approved above. 11 Distribution bill (rate) impacts 229. Fortis provided bill (rate) impacts for each rate class and for different components of a typical customer s bill based on the proposed adjustment to 100 per cent revenue-to-cost ratios by rate class and the effects of the proposed rate designs. The bill impacts discussed in this section are for the distribution component only. Commission findings 230. The Commission directs Fortis in its refiling to submit updated bill impact schedules comparing the proposed 2014 rates, as adjusted to comply with directions in this decision, to the current rates. 12 Distribution adjustment rider 231. In addition to base distribution rates, Fortis advised that a distribution adjustment rider (DAR) will be required in order to recover revenues associated with the rate adjustments related to PBR to account for approved K, and potentially Z factor amounts. Under the PBR framework, Fortis proposed to recover all other PBR related amounts through a single DAR, expressed as a percentage of base distribution rate revenue, by rate class. The directions issued in decisions and considered and approved Fortis approach to recovering K, Y and Z factor amounts in a DAR expressed as a percentage of base distribution revenue The CCA recommended that Fortis should be directed to recover all PBR-related amounts (K,Y and Z factor amounts) by way of PBR rates for the upcoming year, to achieve a consistent approach among Fortis and other electric distribution utilities regulated by the AUC. Commission Findings 233. Given that the Commission approved the estimated amounts to be recovered by way of the 2014 DAR and the proposed 2014 DAR rate as filed in Decision , the Commission considers that no further directions are required with respect to the CCA s recommendation. 100 Decision : 2012 Performance-Based Regulation Second Compliance Filings, AltaGas Utilities Inc., ATCO Electric Ltd., ATCO Gas and Pipelines Ltd., EPCOR Distribution & Transmission Inc. and FortisAlberta Inc., Application No , Proceeding ID No. 2477, July 19, AUC Decision (January 27, 2014) 45

52 13 Review of Fortis Rate 44/45 charges 234. On June 4, 2012, the Commission received a complaint (the complaint application) from Harvest Operations Corp. (Harvest) 101 about billing by Fortis. Harvest and Fortis had been unable to resolve issues surrounding the appropriate billing of 166 accounts in the Wainwright area that had been transitioned from Oilfield Rate Schedule 44 (Rate 44) to Oilfield Rate Schedule 45 (Rate 45) in 2011 The Commission established Proceeding ID No to deal with the complaint application. For the purposes of this Phase II application, the entire record of Proceeding ID No was available to parties in the complaint application Background of the transition of customers from Rate 44 to Rate In its Distribution Tariff Phase I application ( application), Fortis applied for approval of a business case for the installation of metering equipment at approximately 10,500 unmetered oilfield sites, enabling the transition of customers served on the unmetered Rate 44 to the metered Rate 45. Fortis stated in its application that the upgrade would provide oilfield customers with consumption data consistent with all other Fortis customers, and would ensure appropriate billing based on actual consumption and increased load settlement accuracy for all customers. 103 Although Decision does not explicitly reference this project, the complainant asserts in the complaint application that, because the business case was filed with the application, Commission approval was implicit Fortis subsequently closed Rate 44 to new customers, so that Rate 44 is now only available to existing oil and natural gas field services that are unmetered or have demand meters only Rate 44 is a capacity (demand) charge only rate. Under Rate 44, charges for unmetered accounts are billed on the basis of a formula that utilizes the horsepower (hp) of connected motors to estimate demand. 105 Rate 44 also applies to customers with demand only meters. For all Rate 44 customers, the Rate 44 terms define billing demand, referred to as kilowatt (kw) of capacity, as the greater of: 1. for unmetered and energy metered services, the sum of all connected motors and equipment (1 horsepower equals kw); 2. for demand metered services, the highest Metered Demand in the 12-month period including and ending with the current billing period; or 3. the Rate Minimum of 3 kw Rate 45 is available to oil and natural gas field services including pumping and related operations such as rectifiers, cathodic protection and radio transmitters and to water pumping services. Rate 45 is available to services with operating demands less than 75 kw that have a Proceeding ID No. 2006, Harvest was represented by GenAlta Power Inc. (GenAlta) who acted as consultant to Harvest only. Exhibit 40.01, AUC correspondence, response to FortisAlberta Inc. letter regarding information request R44 - FAI-001. Proceeding ID No. 2006, Exhibit 4, Attachment 3 to the application, Fortis 2010/2011 Phase I/II distribution tariff application, pages Proceeding ID No. 2006, no specific reference in Decision of approval of the business case, however Fortis did file this business case as part of its Distribution Tariff Phase I application. The business case has been filed on the record of this proceeding as Exhibit 2, Attachment 1 to the application. Proceeding ID No. 2006, Exhibit 5, Attachment 4 to the application, Rate 44 and Rate 45 charges and terms. Proceeding ID No. 2006, Exhibit 5, Attachment 4 to the application, Rate 44 and Rate 45 charges and terms. 46 AUC Decision (January 27, 2014)

53 demand and energy measurement meter. 107 Unlike Rate 44 customers, Rate 45 customers are charged on the basis of both energy (kwh) and demand (kw), with billing demand or kw of capacity set as the greater of: 1. the highest Metered Demand in the billing period; 2. 85% of the highest Metered Demand in the 12 month period including and ending with the current billing period; 3. the Rate Minimum of 3 kw The descriptions of Rate 44 and Rate 45 above are descriptions from Fortis terms and conditions of service and rate schedules that were approved and in effect at the time of the complaint application. The Rate 44 and Rate 45 rate schedules that were approved and effective at the time of the complaint application are attached as Appendix 3 to this decision. Fortis has since applied to the Commission in this proceeding to revise its terms and conditions of service Proceeding ID No and AUC Decision The Commission established Proceeding ID No to deal with the complaint application. In the complaint application, Harvest submitted that Fortis interpretation and application of Rate 45 resulted in service that was unduly discriminatory and contravened established administrative tribunal precedents. Harvest further stated that it did not agree with the methodology that Fortis employed to calculate the demand charges during the transition from Rate 44 to Rate 45 and that it was now clear, with demand and energy meters installed for these accounts, that the unmetered accounts were being under-billed for energy and over-billed for demand In the complaint application, Harvest submitted that Fortis assumed the historic Rate 44 demands, which were based on connected hp, would be equal to the historic Rate 45 measured demands, if they had been measured. 111 Because billing demand based on the connected hp formula of Rate 44 had been deemed by Fortis to be the highest metered demand for the prior 12-month period, the application of the 85 per cent ratchet provision in Rate 45 determined billing demand for the unmetered customers transitioned to Rate 45 who initiated the complaint. The use of deemed demand resulted in higher demand charges under Rate 45 for the first 12 months of the transition from Rate 44 to Rate 45 than would have occurred if demand for Rate 45 billing purposes was based on current metered demand for the customers that filed the complaint In the complaint application, Harvest submitted that it was unfair and inappropriate for Fortis to use historic Rate 44 connected hp capacity values in the determination of the Rate 45 billing demand given that the Rate 44 demands were significantly larger than the actual demand values measured following the transition to Rate Proceeding ID No. 2006, Exhibit 5, Attachment 4 to the application, Rate 44 and Rate 45 charges and terms. Proceeding ID No. 2006, Exhibit 5, Attachment 4 to the application, Rate 44 and Rate 45 charges and terms. Exhibit and Exhibit 49.13, UCA-FAI-16. Proceeding ID No Exhibit 1, application, Harvest s position, page 4. Proceeding ID No. 2006, Exhibit 1, application, background, page 2. Proceeding ID No. 2006, Exhibit 1, application, background, page 2. Proceeding ID No. 2006, Exhibit 1, application, background and Harvest s position, pages 2-5. AUC Decision (January 27, 2014) 47

54 243. In the complaint application, Harvest requested that the Commission direct Fortis to: 1. Retroactively adjust the Rate 45 demand charges to remove the ratchet calculation based on historic Rate 44 kw of capacity values for every Rate 45 account billed to every FortisAlberta customer where the account was moved from Rate 44 to Rate 45 following the installation of a demand and energy meter as part of the meter installation program approved by the Commission under Decision Not to use Rate 44 billing determinants in the determination of Rate 45 charges for any account moved from Rate 44 to Rate 45 in the future following the installation of a demand and energy meter as part of the meter installation program approved by the Commission under Decision In the complaint application, Fortis maintained that its billing approach had been consistent with how billing and ratchets had historically been applied whenever a customer site was transitioned from unmetered to metered service. That is, unmetered demand present in months when the site was billed on the unmetered rate had been carried forward as the demand in that month until the 12-month demand ratchet expired. 115 Fortis argued that while not explicitly documented, this treatment was consistent with the wording of Rate 44 that states [t]he kw of Capacity must bill for a minimum of 12 consecutive months before being reduced In the complaint application, Fortis explained that in both Rates 44 and 45, the billing demand (kw of capacity) was calculated based on 12 months of demand data, thereby creating a 12-month transition period when customers are switched from Rate 44 to Rate 45. When a service is switched from Rate 44 to Rate 45, the demand based on connected hp is used as a proxy for metered demand in months where metered demand is not available in the 12-month period, including and ending with the current billing period; that is, until a full 12 months of actual demand data is recorded. As a result, throughout the 12-month transition period, where the new monthly metered demand is lower than the previous assessed connected hp, the demand charge portion of Rate 45 may be based on 85 per cent of the demand level applied for billing purposes over the prior 12 months In the complaint application, Fortis stated that its revenue forecast implicitly assumed that billing would be applied consistent with historical practice. Fortis submitted that in reality it had actually experienced a revenue under-collection in rates 44 and 45 when compared to forecast. Therefore, Fortis submitted that if Harvest were to be granted the relief it seeks, it would need to reflect this new interpretation and recast its Rate 44 and Rate 45 demand billing determinants. Fortis further submitted that all Rate 44 and Rate 45 customers would be affected as a result of the recalculation of the billing determinants and the corresponding changes to Rate 44 and Rate 45 that would occur after updating the rates to reflect the revised billing determinants Fortis stated that the complaint application could potentially affect the tariffs in 2011, 2012 and future years, given that the oilfield metering project is ongoing. Fortis stated that, should the Commission determine that one or more aspects of Harvest s complaint have merit Proceeding ID No. 2006, Exhibit 7.01, GenAlta Power letter responding to AUC, Section A, page 1. Proceeding ID No. 2006, Exhibit 10.01, Fortis comments to AUC, page 1. Proceeding ID No. 2006, Exhibit 18.02, Fortis argument, paragraph 6. Proceeding ID No. 2006, Exhibit 10.01, Fortis comments to AUC, pages 1 and 2. Proceeding ID No. 2006, Exhibit 10.01, Fortis comments to AUC, pages 2 and AUC Decision (January 27, 2014)

55 and require further review, then the scope and magnitude of the issue, secondary implications and impacts would be greatly expanded and the most efficient process to deal with the issue would be in its upcoming Phase II application In Decision , 120 the decision resulting from Proceeding ID No. 2006, the Commission found the following: 35. For the reasons set out below, upon further examination of the evidence on the record of this proceeding, the Commission considers that its determination on the issues raised in the complaint application could have wider impacts than intended and may directly affect other Fortis customers and Fortis tariff in a manner beyond the scope of a complaint application. The Commission considers that Harvest s complaint application has raised important questions about the proper interpretation and application of Fortis tariff, and that those questions require further review, including the opportunity to determine how these issues may potentially affect other Fortis customers and Fortis tariff, before a proper determination can be made. Therefore, the Commission finds that it is not reasonable to consider the complaint application in isolation Therefore, in Decision , the Commission deferred its consideration of the complaint application to Fortis next Phase II application and ordered the following: (1) The Commission will not make a determination on the relief requested by Harvest Operations Corp. in this decision. (2) FortisAlberta Inc. is to include additional information as directed in this decision in its upcoming Phase II application Fortis response to AUC directions from Decision Decision set out the following three directions for Fortis next Phase II application: 43. The Commission directs Fortis to address the issues raised in the complaint application as part of its upcoming Phase II application, in particular that Fortis: Provide evidence demonstrating all assumptions used when setting its billing determinant forecast for accounts transitioned from unmetered to metered service, including what ratchet would be applied to transitioned unmetered accounts. Provide two rate designs for Rate 45, one based on current approved billing determinants for Rate 45 and the other based on forecast billing determinants for Rate 45 modified to reflect the relief requested by Harvest and extended to include other similarly affected customers. Prior to filing its Phase II application, Fortis should provide notice to all customers that have been or could potentially be impacted by a transition from unmetered to metered service. The notice should inform impacted parties that the issues from this complaint application will be canvassed in the Phase II application Proceeding ID No. 2006, Exhibit 10.01, Fortis comments to AUC, pages 3 and 4. Decision : Harvest Operations Corp., Application to Review FortisAlberta Inc. Rate 44/45 Charges, Application No , Proceeding ID No. 2006, January 3, Decision , paragraph 43. AUC Decision (January 27, 2014) 49

56 251. With respect to directions a) and b), Fortis responded to these directions in Section 4.3 and Appendix 5 of the application, respectively. Fortis stated that it had provided two 2013 billing determinant forecasts based on current rate structures, with or without Harvest s relief being granted, in Appendix 5 of the application. Also, Fortis provided two additional 2013 billing determinant forecasts that reflected the newly proposed rate structures for rates 44/45, with or without Harvest s relief being granted With respect to direction c) in the application, Fortis submitted that it was in the process of assembling a list of all companies which have sites billing on Rate 44/45 with the intent of notifying all Rate 44/45 customers, preferably through Later in IR responses, Fortis confirmed that it had complied with this direction and provided notification to affected parties by mail. 124 Commission findings 253. The Commission has reviewed Fortis evidence and supporting documentation and is satisfied that Fortis has complied with these three directions from Decision Rate 44/ The Rate 44/45 group was comprised of Harvest Operations Corp., Husky Energy, Pengrowth Corp., Talisman Energy, Canadian Natural Resources, and Direct Energy Resources. 125 The Rate 44/45 group retained Desiderata Energy Consulting Inc. (DESI) to represent them in this proceeding. DESI submitted that the Rate 44/45 group members are oil producers with oilfield facilities, primarily oilfield pumping equipment, served on Fortis Rate 44 and Rate In this proceeding, DESI confirmed that the relief the Rate 44/45 group was seeking was the same as the relief asked for by Harvest in Proceeding ID No In this application, DESI submitted that Fortis had applied its tariff inappropriately by applying the Rate 45 demand ratchet to accounts moved from Rate 44 and had therefore collected incremental revenues. DESI stated that the Rate 45 demand ratchet should not apply for those accounts moved from Rate 44 to Rate 45 under the meter installation program. 128 DESI recommended that: 129 a) FortisAlberta be directed to refund to Rate 45 customers tariff charges which were calculated using historic Rate 44 connected horsepower (hp) values, for all accounts moved to Rate 45 as part of FortisAlberta s meter installation program for the period from January 17, 2011 to the implementation date of the tariff approved in this proceeding. b) The refunded amounts not be levied on and collected from Rate 44 or Rate 45 customers via a deferral account or rate adjustment. The refunded amounts should be to the account of FortisAlberta s shareholders. c) The tariff approved in this proceeding should specifically exclude, in the case of unmetered Rate 44 sites which are metered and subsequently billed on Rate 45, the use Exhibit 1, paragraph 227. Exhibit 1, paragraph 236. Exhibit 43.02, AUC-FAI-037. Exhibit 62.01, page 1, Q2. Exhibit 62.01, pages 1-2, Q3. Exhibit 82.01, AUC-DESI-1(c). Exhibit 62.01, pages 9-12, Q16. Exhibit , page AUC Decision (January 27, 2014)

57 of the historic Rate 44 connected hp values in the application of the Rate 45 demand ratchet, and for all future accounts moved from Rate 44 to Rate 45 resulting from FortisAlberta s meter installation program. d) FortisAlberta be directed to cease its practise of billing Rate 44 accounts that have had a constant billing demand determinant for 12 months not be subject to an additional 12 months on the prior billing demand determinant when the connected hp is reduced or when moved to idle service DESI stated that the terms of one rate cannot be deemed to be applicable to another rate, even when an account is moved from one rate to another. The kw of Capacity must bill for a minimum of 12 consecutive months before being reduced clause is not included in Rate 45. The Rate 44/45 group submitted that any argument that this Rate 44 clause justifies applying the Rate 45 ratchet to the Rate 44 connected hp value is inappropriate. 131 DESI submitted that if Fortis wanted to rely on the kw of Capacity must bill for a minimum of 12 consecutive months before being reduced clause to impose the Rate 45 ratchet, then the clause should have been included in Rate DESI stated that the Rate 44/45 group evidence clearly demonstrates that connected hp values are not equivalent to metered demands. DESI submitted that in practice, Rate 44 accounts are unmetered meaning the highest metered demand term is meaningless. Therefore, the kw of capacity or billing demand is the greater of the connected hp converted to kw and three kw. Rate 44 is billed on the electric motor manufacturer s rating of the maximum hp capability of the motor. 133 Connected hp values will always be higher than metered demands as the pump jack motor must be large enough to lift the produced fluids to the surface. A connected hp value is the maximum capability of an electric motor, whereas a metered demand reading is the actual utilization of an electric motor. 134 When converting from Rate 44 to Rate 45, Fortis applied the historic connected hp value (from Rate 44) as being equivalent to the Rate 45 metered demand for the purpose of the ratchet clause. DESI stated that Fortis practice of applying Rate 44 connected hp values under Rate 45, which is based on metered demand, is simply wrong DESI submitted that there is no evidence to suggest that oilfield operators knew that Fortis had the billing system capability to apply Rate 44 connected hp values under the Rate 45 demand ratchet provision between 2007 and 2009, nor that Fortis intended to apply Rate 44 connected hp values after the meter installation program was approved. 136 DESI also submitted that Fortis had not advised customers that it would be using Rate 44 connected hp values in the determination of Rate 45 charges. DESI submitted that on average from 2007 to 2009, fewer than 10 accounts per month were moved from Rate 44 to Rate 45. Fortis provided the following table showing the number of accounts moved from Rate 44 to Rate 45 as part of the meter installation program: Recommendation (d) from DESI was not explicitly stated as requested relief until DESI s argument submission. Exhibit , paragraph 17. Exhibit , paragraph 18. Exhibit 62.01, page 2, Q6. Exhibit , paragraph 24. Exhibit , paragraph 23. Exhibit , paragraph 44. AUC Decision (January 27, 2014) 51

58 Table 21. Number of accounts moved from Rate 44 to Rate 45 Year Site IDs moved from flat to metered , , DESI expressed the view that prior to the implementation of the meter installation program in 2011, very few or no oilfield operators were aware that Fortis would be applying the Rate 45 demand ratchet provision to accounts moved from Rate 44 to Rate DESI submitted that the 1996 TransAlta proceeding 138 approving Fortis current practise of using Rate 44 billing determinants under Rate 45 does not provide a precedent justifying Fortis use of connected hp values under Rate 45 as argued by Fortis. Nowhere in the 1996 documents does it suggest that Rate 4400 Oil and Gas capacity service connected hp values should be or could be applied under the newly established Rate 4500, Oil and Gas Energy Service. 139 Further, the 1996 documents suggest that in the few instances where a Rate 4400 metered account could be moved to Rate 4500, the historical metered demand billing determinants could be applied under Rate DESI contended that deeming connected hp values under Rate 44 to be demand values under Rate 45 had resulted in higher billing determinants and in Fortis collecting approximately $0.5 million more in revenue. 141 DESI submitted that the 85 per cent ratchet clause results in a higher billing demand when the Rate 45 metered amount is lower than 85 per cent of the Rate 44 connected hp value in any billing period. For example, if after a meter is installed and the metered peak demand in the billing period is seven kw, and the former Rate 44 connected hp value was 10 kw, the 85 per cent ratchet clause could result in a higher billing demand of 8.5 kw. 142 DESI asserted that Fortis s claim that no incremental revenues were collected from Rate 45 customers is based on the measure of actual rather than forecast revenues. DESI submitted that this is not the appropriate test to use. DESI stated that applying the Rate 45 demand ratchet using Rate 44 connected hp values had resulted in increased Rate 45 revenues, over and above what would have been collected had the Rate 45 ratchet not been imposed Fortis evidence in its application was that the conversion from Rate 44 to Rate 45 would be revenue neutral: Given that rate structures for the distribution component for Rates 44 and 45 are identical (i.e. two block kw of Capacity charges), the proposal is designed to be revenue neutral regardless of any customer movement between rate classes. FortisAlberta submits that this proposal is an appropriate measure to ensure that basic consumption and load data is Exhibit , paragraph 34 and Exhibit 62.01, page 7. Exhibit , TransAlta Utilities, 1996 Phase II Refiling Pursuant to Board Decision U99035, October 4, 1999, page 9. Exhibit , paragraphs Exhibit , paragraph 7. Exhibit , paragraph 26. Exhibit 82.01, AUC-DESI-4 (a). Exhibit , paragraphs 15 and AUC Decision (January 27, 2014)

59 captured on all new services to accurately reflect a site s load level for purposes of distribution tariff billing and load settlement DESI argued that Fortis did not keep its commitment that the meter conversion program would be revenue neutral for accounts moved from Rate 44 to Rate 45 and that on average, Rate 44 customers are worse off after demand meters are installed and accounts are moved to Rate 45. The meter conversion could have been considered to be revenue neutral if Fortis had not used the Rate 44 connected hp values in the calculation of Rate 45 tariff charges If the Commission were to grant the relief requested, DESI submitted that the refund should be to the account of Fortis. DESI stated that Fortis had applied its tariff inappropriately and that the revenues collected by applying connected hp values under the Rate 45 demand ratchet should be fully refunded In response to the billing complaint and the relief requested by the Rate 44/45 group, Fortis responded that: 147 The Commission should deal with this matter expeditiously as a customer complaint and deny the requested relief of resetting the demand ratchet upon transition from Rate 44 to Rate 45, by confirming that Fortis tariff is being billed consistently and in a non-discriminatory manner as approved. Fortis also noted that on a go-forward basis, proposed changes to the Rate 45 rate sheet have been proposed in the application to remove any doubt with how customers moving from Rate 44 to 45 are billed in terms of application of the lower 85 per cent ratchet; and while Rates 44 and 45 have separate rate sheets, they are essentially the same customer rate group, with the only difference being the application of metering. In the alternative event of any of the relief being granted, the Commission should direct Fortis to use the Rate 44/ billing determinant forecasts and associated rate designs for relief being granted, for purposes of establishing final 2013 rates and newly structured 2014 rates in the application, and/or any other forthcoming application for purposes of establishing final 2013 and 2014 rates; and approve a onetime Rate 44/45 charges deferral account, which will capture any amounts, once calculated by a reconfigured billing system and refunded to the customers seeking relief; to be recovered from all Rate 44/45 customers in a future distribution adjustment rider In this application, Fortis stated that it had maintained that its billing approach had been consistent with how billing and ratchets have historically been applied whenever a customer site was transitioned from unmetered to metered service. That is, unmetered demand that was present in months when the site was billed on the unmetered rate has been carried forward as the demand in that month until the 12-month demand ratchet expired. Also, while not explicitly documented, this treatment was consistent with the wording from Rate 44 that states the kw of Capacity must bill for a minimum of 12 consecutive months before being reduced FortisAlberta 2010/2011 Phase I/II Distribution Tariff Application, AUC Application No , Proceeding ID No. 212, pages 8-43, emphasis added, (see Proceeding ID No. 2006, Exhibit 4, pages 8-43). Exhibit , paragraphs Exhibit , paragraph 86. Exhibit , paragraphs Exhibit 1, paragraph 216. AUC Decision (January 27, 2014) 53

60 267. Fortis submitted that given it had consistently billed demand in the same manner for many years, it did not provide any specific notification to parties who were impacted by the transition from unmetered accounts to metered accounts. 149 Further, Fortis stated that, in Proceeding 2006, it had submitted: FortisAlberta is not aware of any of any references in its approved tariff that specifically address the treatment of the demand ratchet when an account is moved from Rate 44 to 45. However, FortisAlberta s consistent historical practice has been that whenever a customer site is transitioned from unmetered to metered, the unmetered demand that was present in the months when the site was billing on the unmetered rate is carried over as the demand in that month until the 12 month demand ratchet window has expired Fortis stated that while investigating this issue, it did find a reference to the issue of the demand ratchet being applied to unmetered and metered services under Rate 4400 in the 1996 TransAlta Phase II refiling: Pursuant to Directive 32, TransAlta has removed the designation of the existing Oilfield Pumping/Water Pumping Rate 73X (now Rate 4400) as a closed rate. The billing capacity has been set equal to the rated capacity or the highest metered demand in the current or previous 11 months. The Board indicated an 85% demand ratchet is appropriate for loads supplied through the 25 kv distribution system, in Directive 20. Since Rate 4400 is predominantly an unmetered rate, the use of rated capacity provides an implicit 100% demand ratchet. To provide parity with the unmetered services on this rate, a 100% demand ratchet is appropriate for the few metered services on this rate. Pursuant to Directives 30 and 32, TransAlta has introduced the new Oil and Gas Energy Rate 4500 (non-time-of-use) available to all new oil and Gas services. Pursuant to Directive 20, the billing capacity includes a demand ratchet of 85% for [to] this rate Fortis submitted, from its reading of the above, it appeared that at the time there were still a few metered services on the unmetered Rate 4400 and that a 100 per cent ratchet was being applied to these services Fortis submitted that the metering of oilfield sites program was proposed and approved in its DTA and that the upgrade to metered services provided oilfield customers with consumption data consistent with all other Fortis customers, would ensure appropriate billing based on actual consumption and increased load settlement accuracy for all customers. Fortis submitted that any assertion that it had forced this change would be misleading, and that there was never any obligation to ensure that the metered customers were not worse off. Metering of these customers and the transition to Rate 45 ensured appropriate billing based on actual consumption and increased load settlement accuracy for all customers Fortis submitted that it did not over-collect revenue from Rate 44/45 customers. Instead, Fortis stated that it had experienced a shortfall in revenue of $2.3 million in 2011 and $0.9 million in Therefore, Fortis stated that no incremental or additional revenue was Exhibit 43.02, AUC-FAI-035 (b). Proceeding 2006, Exhibit 17.02, HARVEST-FAI-001 (e). Exhibit 107, PDF page 34. Exhibit , paragraph 107. Exhibit , paragraph AUC Decision (January 27, 2014)

61 collected from Rate 44/45 customers in 2011 and If ratchet relief is granted, the estimated refund amount would be approximately $500,00 for sites converted from 2011 to 2012 on approved interim rates but, as the conversion of metering of unmetered sites is ongoing and final rates for 2012 and 2013 have yet to be approved, the total amount cannot be determined. Fortis submitted that it would consider this amount to be material, especially considering that any relief granted would alter Fortis billing practice of its approved tariff for such cases and would alter the basis upon which billing determinants were forecast and rates designed In response to DESI s recommendation that Fortis should not be allowed to collect any refunded amounts from its customers by way of a deferral account or any other tariff mechanism, Fortis submitted that it would be inappropriate to deviate from consistent billing practices relating to a transition from an unmetered to metered service and reduce the rate class revenues even further, without subsequent recovery of those amounts from the class that incurred the costs allocated to them The CCA had concerns with DESI s requested relief, specifically with DESI s recommendation (c). The CCA submitted that this request could potentially result in cross subsidies not only within rate classes 44/45 but also in relation to other rate classes because the CAM method uses demand parameters to allocate costs among customers at shared feeders. The CCA noted that in Proceeding ID No. 2006, Fortis had stated: Fortis noted that it has not assessed a buy-down or Payment in Lieu of Notice (PILON) to sites impacted by the meter conversion project. Because the metering project and resulting transition to Rate 45 was caused by Fortis installation of meters at affected sites, Fortis has not subjected customers impacted by the meter installation to an investment buy-down charge that would otherwise apply when metered demand falls below their previous assessed demand. Customers whose actual metered demands are lower than the previously assessed demand will see a 15 per cent reduction in demand charges for the 12 month transition period after the rate transition has occurred, and even lower demand charges thereafter, once the 12 month window has expired. These reduced demand charges benefit the customer regardless of the fact that the size of their service capacity and the level of investment, which were based on the customer s initial assessed demand upon servicing, are not similarly reduced The CCA recommended that all customers moving from Rate 44 to Rate 45 be subject to PILON by virtue of the fact their billing demand is effectively being reduced. In other words, oilfield customers for whom investments in customer facilities were made by the customer on the basis of a 100 per cent demand ratchet under Rate 44 are now, under Rate 45, being subjected to an 85 per cent demand ratchet. These Rate 44 customers would have received a lower level of customer investment had the 85 per cent ratchet been in place from inception when they contracted under Rate 44. The imposition of the PILON is intended to transition Rate 44 customers to Rate 45 with a reduced demand ratchet The CCA submitted that any PILON revenue should be credited to the relevant asset accounts in Fortis books and AESO demand charges (which may include ratchets) so that the allocations of assets and AESO charges would not be unduly distorted by the transition of Exhibit , paragraphs Exhibit 43.02, AUC-FAI-036 (a) and (b). Exhibit , paragraph 116. Proceeding ID No. 2006, Exhibit 18.02, paragraph 10 and Exhibit 10.01, page 2. AUC Decision (January 27, 2014) 55

62 Rate 44 customers to Rate 45. Further, the CCA submitted that if PILON is implemented as per the CCA s recommendation, then the CCA would have no issue with the AUC approving the Rate 44/45 customers' request that once a Rate 44 customer transitions to Rate 45, it be subject to the 85 per cent demand ratchet applicable to that rate class Fortis did not support the CCA s PILON recommendations, but did note that billing the twelve month 85 per cent ratchet on the previous unmetered demand, as the basis for investment, could be considered conceptually similar in effect. 158 DESI also did not support the CCA s recommendation and stated that in its interpretation, Fortis tariff, specifically Section 7.3.2, does not provide for the application of a PILON for accounts moved from Rate 44 to Rate 45 because the connected or operating loads did not change and because under the meter conversion program customers did not provide notice of a load reduction as there was no load reduction and the meter installation was imposed by Fortis. 159 Commission findings 277. The Commission finds that the Rate 44/45 group s complaint has merit and that relief is warranted. The Commission will discuss the specific relief approved and its reasons for approving relief in the sections below as follows: interpretation of rate schedules 1996 TransAlta proceeding billing determinants revenue impacts relief approved payment in-lieu of notice (PILON) billing adjustment Interpretation of rate schedules 278. Fortis submitted that its billing approach has been consistent with how billing and ratchets have historically been applied whenever a customer site was transitioned from unmetered to metered service and stated that while not explicitly documented, this treatment is consistent with the wording of Rate 44 that states [t]he kw of Capacity must bill for a minimum of 12 consecutive months before being reduced Having reviewed the rate schedules in effect at the time of the complaint, the Commission finds that Fortis tariff does not provide sufficiently clear direction as to how unmetered customer accounts may be billed when transitioned to metered service i.e., there is no explicit transition provision in Fortis tariff that clearly explains to customers how unmetered accounts will be transitioned for billing purposes to accounts that utilize the meters necessary for the application of Rate 45 (see Appendix 3 to this decision for Oilfield Rate Schedules 44 and 45 in effect at the time of the complaint application). Although, Fortis Rate 44 rate sheet states [t]his rate is no longer available for new installations and existing services are being transitioned to Rate 45 as metering is installed, the Commission finds that the billing transition provisions provided in the tariff are inadequate to impose liability on customers for charges being billed by Fortis during this transition. Fortis acknowledged that the consistently applied Exhibit , paragraph 110. Exhibit , paragraphs AUC Decision (January 27, 2014)

63 billing practice upon metering of a service has never been explicitly inserted into the Rate 44/45 rate sheets Fortis submitted that while rates 44 and 45 have separate rate sheets, they are essentially the same customer rate group with the only difference being the application of metering. 161 The Commission considers that a customer account can only be on one rate schedule at a time and that a rate schedule should stand alone for billing purposes. Although metered demand is a capitalized term both in Rate 44 and in Rate 45, the Commission finds that metered demand does not apply to unmetered accounts under Rate 44.Therefore, the Commission finds that customers who were unmetered customers under Rate 44 have no metered demand for the purposes of Rate 45. For example, for a new customer who signed up for rate 45, such that there was no history of demand, demand history would build gradually one month at a time The Commission finds that the language used in Fortis tariff does not support its historical billing practice with respect to Rate 44 and Rate 45 accounts and that Fortis cannot rely on either a historical practice or an implied, but not expressed, interpretation of its tariff TransAlta proceeding 282. The Rate 44/45 group complained regarding Fortis treatment of the transition of unmetered accounts to metered services. The Commission does not consider that the 1996 TransAlta proceeding is a precedent for how unmetered accounts should be transitioned to metered accounts. In that proceeding, TransAlta stated: Pursuant to Directive 32, TransAlta has removed the designation of the existing Oilfield Pumping/Water Pumping Rate 73X (now Rate 4400) as a closed rate. The billing capacity has been set equal to the rated capacity or the highest metered demand in the current or previous 11 months. The Board indicated an 85% demand ratchet is appropriate for loads supplied through the 25 kv distribution system, in Directive 20. Since Rate 4400 is predominantly an unmetered rate, the use of rated capacity provides an implicit 100% demand ratchet. To provide parity with the unmetered services on this rate, a 100% demand ratchet is appropriate for the few metered services on this rate. 162 [emphasis added] 283. The Commission finds the TransAlta case is distinguishable from this complaint because the Board was referring in the TransAlta case to transition of the few metered services on Rate 4400 and its direction applied specifically to those accounts that were metered on Rate Billing determinants 284. The Rate 44/45 group had submitted that connected hp values are not equivalent to metered demands and that Fortis had inappropriately taken advantage of the meter installation program to increase revenues. In reply, Fortis stated that while it accepted as a fact that connected hp values will always be higher than metered demand, the Rate 44/45 group s assertions that it had collected additional revenue were in error Exhibit , paragraph 97. Exhibit , paragraph 98. Exhibit 107, PDF page 34. Exhibit , paragraph 99. AUC Decision (January 27, 2014) 57

64 285. The Commission understands that due to the nature of a pump jack motor, connected hp values can be higher but not lower than metered demands The Commission also understands that one of the expressed purposes of the meter installation program was to provide Rate 44/45 customers with actual consumption information. Fortis stated that: 164 The upgrade to metered services provides oilfield customers with consumption data consistent with all other FortisAlberta customers, and would ensure appropriate billing based on actual consumption and increased load settlement accuracy for all customers In the specific case of the Rate 44/45 group complaint and all similarly affected customers, the Commission finds the use of historic Rate 44 connected hp values in the application of the Rate 45 demand ratchet to be unsupported. Fortis itself stated: 165 Since there are many variables that can cause a customer s bill to change when transitioning from unmetered to metered, such as changes in weather, rate changes, changes to capacity factor, and operational patterns, it is difficult for FortisAlberta to make an overall general conclusion as to the changes in consumption. However, the billing data does support that after the sites were metered, the average energy (kwh) increased, while average demand (kw) decreased. [emphasis added] 288. The Commission considers that using historic Rate 44 connected hp values in the application of the Rate 45 demand ratchet does not align with Fortis expressed purposes of the meter installation program Revenue impacts 289. Fortis had submitted that in designing the transition of unmetered accounts to metered accounts there was never any obligation to ensure that the metered customers were not worse off. However Fortis stated that from the utility s viewpoint, it had designed the transition from unmetered accounts to metered accounts to be revenue neutral regardless of any customer movement between rate classes. Therefore, the Commission does not agree that Fortis intentionally collected additional revenues by applying its historical billing practice when transitioning unmetered accounts to metered accounts Notwithstanding, the Commission finds that applying the Rate 45 demand ratchet using Rate 44 connected hp values has resulted in increased Rate 45 revenues, over and above what would have been collected had the Rate 45 ratchet not been imposed by Fortis. Further, the Commission agrees that Fortis will continue to receive a net increase in revenues (over and above what would be collected if the Rate 45 ratchet were not imposed) if the Rate 45 ratchet continues to be applied for 12 months to customers whose actual measured demand has been shown to have decreased The Commission does not accept Fortis submission that should any relief be granted the Commission ought to allow Fortis to recalculate its Rate 44 and Rate 45 parameters using revised billing determinants and that the refunded amounts should be recovered from all Rate 44 and Rate 45 customers. The Commission finds that Fortis should bear the inherent risk that billing determinant forecasts, as approved in the general tariff application proceeding on which Exhibit , paragraph 89. Exhibit 43.02, AUC-FAI-039(c). 58 AUC Decision (January 27, 2014)

65 its tariff is based, may not materialize in practice. The Commission considers that a revenue reduction caused by a billing adjustment arising from a successful complaint to the Commission falls within this inherent forecast risk Relief approved 292. The Rate 44/45 group requested the following relief: a) FortisAlberta be directed to refund to Rate 45 customers tariff charges which were calculated using historic Rate 44 connected horsepower (hp) values, for all accounts moved to Rate 45 as part of FortisAlberta s meter installation program for the period from January 17, 2011 to the implementation date of the tariff approved in this proceeding. b) The refunded amounts not be levied on and collected from Rate 44 or Rate 45 customers via a deferral account or rate adjustment. The refunded amounts should be to the account of FortisAlberta s shareholders. c) The tariff approved in this proceeding should specifically exclude, in the case of unmetered Rate 44 sites which are metered and subsequently billed on Rate 45, the use of the historic Rate 44 connected hp values in the application of the Rate 45 demand ratchet, and for all future accounts moved from Rate 44 to Rate 45 resulting from FortisAlberta s meter installation program. d) FortisAlberta be directed to cease its practise of billing Rate 44 accounts that have had a constant billing demand determinant for 12 months not be subject to an additional 12 months on the prior billing demand determinant when the connected hp is reduced or when moved to idle service For all the reasons discussed above the Commission grants relief to the Rate 44/45 group as requested above in a), b) and c). The Commission notes that the relief requested in d) above was not explicitly requested in this proceeding until argument and therefore the Commission considers parties did not have sufficient notice of and opportunity to make representations on this request. The Commission considers that the request in part d) above naturally follows from the relief granted in parts a) and c) above The Commission directs Fortis to reflect the Commission s findings with respect to the complaint by the Rate 44/45 group and the relief approved in this decision in its compliance filing. The Commission directs Fortis to provide revised rate design, bill impacts and rate schedules as applicable. The Commission further directs Fortis to provide in its compliance filing any additional schedules showing how the refund to affected Rate 44/45 customers has been determined Payment in lieu of notice 295. The CCA was concerned that if the Commission granted the relief recommended by the Rate 44/45 group referred to in c) above in paragraph 292 of this decision, this relief could potentially result in cross subsidies not only within rates 44 and 45 but also in relation to other rate classes because the CAM method uses demand parameters to allocate costs among customers at shared feeders The Commission notes that in the application, Fortis confirmed that if the requested relief was granted, then all rate impacts would be contained within the Rate 44 and Rate 45 rate 166 Recommendation (d) from DESI was not explicitly stated as requested relief until DESI s argument submission. AUC Decision (January 27, 2014) 59

66 class. 167 The CCA s concern and payment in lieu of notice (PILON) 168 recommendation were not raised in this proceeding until the argument stage and therefore parties did not have sufficient opportunity to make submissions on this issue. Further, the Commission finds that neither the Rate 44/45 group or Fortis supported the CCA s request that all customers moving from Rate 44 to Rate 45 be subject to PILON by virtue of the fact that their billing demand is effectively being reduced For these reasons the Commission does not approve the CCA s recommendation Billing adjustment 298. In Decision , the Commission stated: 44. [T]he Commission finds that Section 11.8 of Fortis Customer Terms and Conditions of Electrical Distribution Service applies to the complaint application, and to any relief that may be granted as a result of any determinations made in Fortis upcoming Phase II application. Section 11.8 specifically addresses circumstances where Fortis overcharges or undercharges on a bill as a result of a billing error including, but not limited to, incorrect meter reads or any calculation, rate application or clerical errors. [emphasis added] 45. The time limitation (and therefore, the limit on the amount) for billing adjustments is set out in Section 11.8, which indicates that billing errors resulting in overcharges or undercharges will be calculated and charged or refunded up to a maximum of 12 months immediately preceding the month in which the billing error was discovered. The Commission agrees with Fortis that the date at which the asserted misapplication of Rate 45 was brought to Fortis attention was January 17, Consequently, because Section 11.8 applies, Harvest s requested relief will not otherwise be prejudiced by the Commission s consideration of these issues in Fortis Phase II application. Any billing adjustment, if and when approved, should be applied to Harvest based on the notification date of January 17, The Commission reconfirms that the relief granted is subject to Section 11.8 of Fortis Customer Terms and Conditions of Electrical Distribution Service and that the relief should be determined effective as of the initial notification date of the complaint being January 17, EQUS/NPP jurisdictional issue 300. In argument EQUS and NPP indicated their interest in this proceeding as follows: The allocation of Fortis distribution costs to all REAs within the Fortis service territory, including EQUS and NPP, for the purposes of the Fortis distribution tariff for the period from 2012 through 2014.; The development and use of rates and tariffs to be applied to all REAs within the Fortis service territory for recovery of distribution costs and usage charges. The use of CAM model for its cost allocation study, including the allocation of distribution costs to all REAs within the Fortis service territory Exhibit 43.02, AUC-FAI-036 (d). Payment in lieu of notice (PILON) is an investment buy-down charge that is applied when metered demand falls below previously assessed demand. 60 AUC Decision (January 27, 2014)

67 The method of allocation of Y, K and Z factors for the period of Fortis approved PBR plan Based on these interests, EQUS and NPP sought the following rulings from the Commission, as set out in their response to AUC-EQUS and NPP-1(a): It will not direct or request Fortis to implement new distribution charges upon REAs or its members. In the alternative, that it has no jurisdiction to approve a distribution tariff of Fortis that imposes a charge to REAs for the use by REAs of the Fortis distribution system, or that otherwise imposes any charge to REAs for electric distribution service, including EQUS/NPP, or to any member of EQUS/NPP EQUS and NPP also wanted the Commission to confirm the following: That the matter of allocation of costs between REAs and Fortis for utilizing each other s distribution systems is a private, commercial and negotiable arrangement between Fortis and the REAs. That EQUS and NPP and Fortis have private, commercial agreements in place that address the allocation of costs when utilizing each other s distribution systems In support of their position, EQUS and NPP reiterated the position of the Central Alberta REA in Proceeding ID No. 362 that the Commission does not have the jurisdiction to approve a distribution tariff of Fortis which imposes distribution/usage charges upon REAs and/or their members. In asserting this position, EQUS and NPP adopted the Central Alberta REA motion filed in Proceeding ID No EQUS and NPP stated that load settlement and flow-through AESO related charges are the only proper components of a Fortis distribution tariff that would apply to REAs. Further, there has been no change in the applicable legislation upon which EQUS and NPP rely in making their submissions asserting lack of Commission jurisdiction to impose distribution/usage charges upon REAs Fundamental to this question of the Commission s jurisdiction to impose distribution/usage charges upon REAs, the Commission confirmed in Decision that persons can provide distribution services to themselves, and not from Fortis, by becoming members of an REA as noted in the following excerpts: Allowing some customers in specific circumstances to supply distribution services to themselves is also part of the legislative scheme. [paragraph 65] for qualified persons to choose to become a member of an existing REA is a form of customer self-supply and is therefore consistent with the legislative scheme. [paragraph 77] An REA is not simply an owner of an electric distribution system, its customers are the owners of the electric distribution system and as such, this membership requirement puts 169 Decision : Central Alberta Rural Electrification Association Limited, Application for a Declaration under the Hydro and Electric Energy Act, Application No , Proceeding ID No. 886, July 4, AUC Decision (January 27, 2014) 61

68 these customers in a similar position of self-supply as that of other customers who are choosing to serve themselves under other statutory provisions. [paragraph 78] The legislative scheme also provides for operating agreements to address the potential for inefficiency in the provision of distribution services and also provides customers with the ability to choose to serve and supply themselves by becoming members of an REA. 78 [paragraph 104] 78 Memorial Gardens Association (Canada) Ltd. v. Colwood Cemetery Co. [1958] S.C.R. 353 at page 4. The meaning in a given case must be ascertained by reference to the context and to the objects and purposes of the statute in which it is found EQUS and NPP submitted that the imposition of distribution/usage charges by Fortis upon REAs or their members vitiates and is contrary to the intent of a self-supply framework confirmed by the Commission in Decision Making REAs or their members subject to the tariffs of another owner of an electric distribution system was argued to also be contrary to the public interest aspects of the REA regime noted by the Commission in Decision and inconsistent with the legislative scheme confirmed by the Commission in that decision In the case of EQUS, it said that the imposition of Fortis tariffs for distribution/usage charges is clearly duplicative of the negotiated charges already in place in the 1997 agreement regarding connection to and use of the Fortis system. Those charges were freely negotiated by the parties to that agreement, and for the Commission to approve duplicative charges would not only be unfair to EQUS but would also inhibit or interfere with EQUS freedom to contract In the case of all REAs, it said that charges for usage of the system of another electric distribution system owner are matters of integrated operations agreements that are provided for under the Roles, Relationships and Responsibilities Regulation, AR 169/ The issue for consideration in this Proceeding is considered to be whether a proper tariff includes distribution/usage charges upon REAs and what is the proper jurisdiction of the Commission to approve such tariffs. EQUS and NPP argued that the Commission only has the jurisdiction to approve load settlement and flow-through AESO related charges. Further, the only costs that Fortis can properly allocate to them and that the Commission can approve in a Fortis distribution tariff are those contained within these two categories of charges Fortis argued that the jurisdictional fulminations by EQUS and NPP are academic in this proceeding and that the essence of the EQUS and NPP requests is that the Commission does not, in this proceeding, make determinations that would fetter any consideration in a later proceeding, when and if the particular facts of that proceeding required such a determination by the Commission. While EQUS and NPP considered that Fortis has ducked the issue of jurisdiction, Fortis considered there was no issue Fortis pointed out that EQUS and NPP acknowledged that cost allocation may be a component of a distribution tariff and the only allocated costs included in the tariff for charging to REAs are costs that they do not dispute Fortis said it continues to allocate distribution costs to REAs using the same cost allocation methods that were used and approved in the last cost allocation study and, as in the 62 AUC Decision (January 27, 2014)

69 last proceeding, Fortis has continued to propose a distribution charge to REAs to recover only costs related to transmission and load settlement Throughout this proceeding, Fortis emphasized that it has clearly stated that neither Fortis nor any other party has proposed to charge REAs for net distribution costs. The filed REA rate has been designed only to recover load settlement and transmission related costs Fortis responded extensively with regard to the Commission s jurisdiction in response to REAs-FAI-023 and in paragraphs 155 to 157 of its argument in chief. Commission findings 316. It appears to the Commission that there is no dispute that the only costs proposed to be allocated to REAs in this proceeding are load settlement costs and transmission related costs The Commission finds the jurisdictional issue put forward by EQUS and NPP is not relevant to the Commission s determinations in this proceeding. 15 Incorporating the 2012 Phase II results in 2012, 2013 and 2014 rates 15.1 Finalizing 2012 and 2013 distribution rates 318. Fortis noted in the application that it did not have final rates approved for The 2012 going-in rates used to establish its proposed 2013 PBR rates in the PBR compliance application had been approved on an interim basis only On October 28, 2011, in Proceeding ID No. 1534, Fortis submitted its 2012 distribution tariff rates filing for the approval of 2012 rates, riders and terms and conditions of electric distribution service to be effective January 1, On December 21, 2011, the Commission issued Decision on Fortis 2012 distribution tariff, where the Commission stated: 36. Accordingly, the Commission considers that the determination of final rates and cost allocation methodologies arising from the 2012 NSA should be addressed in the proceeding to deal with Fortis s Phase II tariff application in This will provide the Commission and interested parties an opportunity to review the cost of service study, proposed cost allocation methodologies and rate design to ensure that they are fair and reasonable Subsequently, the Commission issued Decision on Fortis 2012 Phase I distribution tariff application, approving the 2012 distribution revenue requirement of $398.2 million. The Commission approved Fortis negotiated settlement agreement (NSA) and further reaffirmed its position in Decision : (2) FortisAlberta Inc. s request for final rates for 2012 to be determined in this proceeding is denied. The Alberta Utilities Commission directs FortisAlberta Inc Exhibit 1, paragraph 18. Decision : FortisAlberta Inc., 2012 Distribution Tariff Rate Filing, Application No , Proceeding ID No. 1534, December 21, Decision : FortisAlberta Inc., Application for Approval of a Negotiated Settlement Agreement in respect of 2012 Phase I Distribution Tariff Application, Application No , Proceeding ID No. 1147, April 18, AUC Decision (January 27, 2014) 63

70 to provide, within 30 days of the issue date of this decision, an anticipated filing date for its Phase II application In Decision , PBR Compliance Filings and in Decision , 174 PBR Second Compliance Filings, the Commission approved Fortis proposed 2013 PBR rates on an interim basis, until all remaining 2013 placeholders (including any potential adjustments to 2013 rates resulting from Fortis Phase II application in Proceeding ID No. 2363) have been determined. When these placeholders are resolved, the 2013 rates will be finalized and any required true-up adjustments will be directed by the Commission. 175 The approved 2013 PBR interim rates were calculated using the 2012 base rates or going-in rates, which, in turn, were based on the 2012 interim distribution rates approved in Decision Fortis stated that while this application, specifically the cost allocation study, was filed in respect of the year 2012, it did not propose to alter its 2012 rates by rate class from those that were in effect on an interim basis in Accordingly, Fortis requested that the Commission approve on a final basis those 2012 distribution tariff rates that were in effect on an interim basis from January 1, 2012 to December 31, In response to AUC-FAI-4, Fortis further proposed that declaring the 2012 rates as final would have no further implications, since the 2012 interim rates were designed on a forecast basis to recover the distribution revenue requirement of $398.2 million as approved in Decision Fortis explained that if the 2012 rates were adjusted to reflect the Phase II results, Fortis would be required to reallocate the approved distribution revenue requirement of $398.2 million and design a set of rates for each rate class based on the Phase II structures proposed in this application. The adjustments by rate class would then need to be collected through a prospective, zero sum, DAR to be applied in 2014 in addition to the proposed Phase II adjustments for The impact on rate classes of such reallocations is presented in the table below Decision , paragraph 133 (2). Decision : 2012 Performance-Based Regulation Second Compliance Filings, AltaGas Utilities Inc., ATCO Electric Ltd., ATCO Gas and Pipelines Ltd., EPCOR Distribution & Transmission Inc. and FortisAlberta Inc., Application No , Proceeding ID No. 2477, July 19, Decision , paragraph 93. Exhibit , Fortis argument, paragraph 33. Exhibit 43.02, AUC-FAI AUC Decision (January 27, 2014)

71 Table 22. Rate Class Change in 2012 revenue if the 2012 Phase II results had been reflected in the 2012 rates 178 Rate Code 2012 NSA revenue per interim rates $ NSA revenue per restructured rates $ 000 Increase (Decrease) $ 000 Increase (Decrease) % Residential , ,179 (2,319) -2% FortisAlberta Farm ,510 60,903 13,393 28% REA Farm (180) -33% FortisAlberta Irrigation 26 8,246 13,439 5,193 63% Exterior Lighting ,562 19,630 (932) -5% Small General Service 41 49,682 51,426 1,744 4% Oil and Gas ,291 43,042 2,751 7% General Service 61 76,210 62,628 (13,582) -18% Large General Service 63 15,507 9,578 (5,929) -38% Transmission Connected 65 1,163 1,023 (140) -12% Total FortisAlberta 398, , % 324. Similarly, Fortis did not propose to incorporate any of the Phase II adjustments discussed in this proceeding in its 2013 PBR rates. To design its 2013 final PBR rates, Fortis proposed to follow the approach set out in its PBR compliance filings and approved in Decision and Decision As such, Fortis noted that finalization of 2013 rates is expected to occur in due course in other PBR-related proceedings and decisions dealing with the outstanding matters related to the 2013 amounts. 179 As discussed in this decision, Fortis proposed that the cost allocation and new rate design developed in this Phase II application be incorporated in the updated 2014 PBR rates No party objected to Fortis proposed approach to finalizing its 2012 and 2013 rates. Commission findings 326. As the Commission explained in Decision , in a Phase II rate application, individual customer rates are developed by determining how much of the total revenue requirement, approved in the Phase I proceeding, should be recovered from each customer class (residential, commercial, etc.) and on what billing unit basis (monthly charge, per kilowatt hour, etc.). Rates are established by dividing the revenue requirement for each customer class by the billing units In Fortis case, the last year under traditional cost of service regulation was 2012, whereby a Phase I distribution revenue requirement totaling $398.2 million was approved in Decision Fortis 2012 interim rates, approved in Decision , were designed on a forecast basis to recover this approved revenue requirement. Accordingly, incorporating the cost allocation and rate design methodologies proposed in this Phase II application in the Exhibit 43.08, attachment, AUC-FAI Exhibit , Fortis argument, paragraph 34. Decision , paragraph 8. AUC Decision (January 27, 2014) 65

72 rates would result in the re-allocation of revenue requirement among different classes with a zero net impact on the total revenue requirement for the company The same reasoning generally applies in the case of the 2013 PBR rates as well. As discussed in Section 4.1, although incorporating the 2012 Phase II cost allocation and rate design methodologies in the 2013 rates can have an impact on the company s total PBR revenue, Fortis estimated this impact to be minimal. Accordingly, incorporating the cost allocation and rate design proposed in this Phase II application in the 2013 PBR rates would result in the re-allocation of revenue among different rate classes with a minimal net impact on the company s total revenue in Rather than reflecting the new cost allocation and rate design set out in this Phase II application in developing its 2012 and 2013 final rates, Fortis proposed that the Phase II methodologies be incorporated in the updated 2014 PBR rates. For the purposes of this decision, the Commission finds that Fortis proposal will result in regulatory efficiencies, as there will be no need to implement a DAR to true up differences by rate class, without any impact on Fortis total revenue Furthermore, the Commission considers that the DAR adjustments per rate class, when applied in 2014 in addition to the proposed PBR rate adjustments for 2014, could result in rate shock for some rate classes For these reasons, the Commission finds Fortis proposal regarding the finalisation of its 2012 and 2013 rates to be reasonable. The Commission approves Fortis interim rates in effect from January 1, 2012 to December 31, 2012, approved in Decision , as the final rates for The Commission accepts Fortis proposal that its 2013 PBR rates will not be reflective of the 2012 Phase II methodologies developed in this proceeding and will be based on the cost allocation and rate design used in Fortis PBR compliance filings and approved in Decision and Decision Fortis 2013 PBR rates will be finalized in due course when the outstanding matters related to the 2013 amounts, such as the K factor placeholder and the generic cost of capital placeholder, are resolved Establishing distribution rates for 2014 and beyond 333. Fortis proposed that the new cost allocation and rate design proposed in this Phase II application (in particular, the transition to 100 per cent revenue-to-cost ratios by rate class), be incorporated in the updated 2014 PBR rates. However, in order to move to 100 per cent revenue-to-cost ratios by rate class and implement new rate structures in 2014, Fortis also needed to design the notional 2013 distribution rates (which Fortis referred to as newly structured 2013 rates ) which will not go into effect in 2013 but rather, will serve as an intermediate step in calculating the 2014 updated PBR rates. The newly structured 2013 rates are determined by applying the cost allocation methods and allocators from the 2012 cost allocation study to the target 2013 PBR revenue Fortis explained that the target 2013 PBR revenue is a PBR proxy for a revenue requirement under cost of service framework, used for purposes of Phase II cost allocation and 181 Exhibit 1, application, paragraph AUC Decision (January 27, 2014)

73 rate design. 182 Fortis calculated the target 2013 PBR revenue by applying the 2013 PBR rates (reflective of the going-in rate adjustments and the 2013 I-X index) against the approved 2013 billing determinants forecast. Fortis calculated the target 2013 PBR revenue to be $411.7 million as shown in Column D of Schedule Fortis further observed that since not all of the components of the 2013 PBR rates have been finalized, the target 2013 PBR revenue is a placeholder for purposes of this application and may change depending on the outcome of the related PBR proceedings The newly structured 2013 rates were designed to achieve 100 per cent revenue-to-cost ratios by rate class based on the target 2013 PBR revenue of $411.7 million and incorporated the new rate structures as per this Phase II application. Fortis noted that the newly structured 2013 rates will be used as the basis for setting PBR rates for 2014 and going forward. Fortis submitted that this proposal is consistent with the Commission s findings in paragraph 996 of Decision , which reads: 996. The Commission considers that PBR is unrelated to the requirement to periodically update rates through a Phase II process. However, during the PBR term the companies may file applications for Phase II adjustments to their rate design and cost allocation methodologies and the Commission will make a determination at that time as to whether the adjustments are warranted. For purposes of a cost of service study, the companies shall use the revenue requirement resulting from going-in rates adjusted by the PBR formula (including the I-X mechanism, K factors, Y factors and Z factors) and the latest updated billing determinants On October 15, 2013, the Commission issued a supplemental IR to Fortis asking for additional information regarding its approach to developing the updated 2014 PBR rates. Based on the information provided in the application and the IR response, the approach proposed by Fortis to calculate the updated 2014 PBR rates can be described as follows: 186 (a) First, the target 2013 PBR revenue is calculated by applying the 2013 PBR rates to the 2013 forecast billing determinants. (b) Second, allocators from the 2012 Phase II cost causation study are applied to the target 2013 PBR revenue to allocate this revenue to rate classes. (c) Third, the newly structured 2013 rates per rate class are calculated by dividing the revenue allocated to each customer class by the relevant billing units in accordance with the proposed 2012 Phase II rate design. (d) Updated 2014 PBR rates are calculated by escalating the newly structured 2013 rates by the 2014 I-X index and including any approved 2014 K, Y, and Z factor rate adjustments In its supplemental IR to Fortis in part AUC-FAI-49(e), the Commission inquired whether, instead of applying 2012 Phase II allocators to the target 2013 PBR revenue, the Exhibit , AUC-FAI-49(a). Exhibit 9, Schedule 1. Exhibit 1, application, paragraph 49. Decision , paragraph 996. Exhibit , AUC-FAI-049(a) and (c). AUC Decision (January 27, 2014) 67

74 Phase II cost allocation and rate design could be applied to the approved 2012 revenue requirement for purposes of calculating the 2014 updated PBR rates, specifically as follows: (a) First, apply the 2012 Phase II cost causation study and rate design to the approved 2012 revenue requirement and 2012 billing determinants to obtain the revised 2012 rates per rate class. (b) Second, apply the required going-in rate adjustments to the revised 2012 rates per rate class to calculate the revised going-in rates. (c) Third, escalate the revised going-in rates by the 2013 I-X index and include any approved 2013 K, Y, and Z factor rate adjustments to calculate the newly structured 2013 PBR rates. (d) Calculate the updated 2014 PBR rates by escalating the newly structured 2013 rates by the 2014 I-X index and including any approved 2014 K, Y, and Z factor rate adjustments Fortis submitted that the allocated revenue and the resulting PBR rates calculated under the method set out in AUC-FAI-49(e) are different from its proposal in the application. Fortis indicated that the revenues allocated to the farm, small general service and oil and gas rate classes had decreased, whereas the residential, irrigation, lighting and general service allocated revenues had increased. Therefore, the methodology used in AUC-FAI-49(e), resulted in a target 2013 PBR revenue of $410.3 million, 187 compared to the target 2013 PBR revenue of $411.7 million under Fortis method, a decrease of approximately $1.4 million or 0.3 per cent. Fortis explained that the reason for this decrease is that, under the method set out in AUC-FAI-49(e), on average, more 2012 revenues are being allocated to the slower growing rate classes In its IR response, Fortis provided the following arguments to support its applied-for methodology to develop the updated 2014 PBR rates over the method set out in AUC-FAI-49(e): ) Does not Materially Alter PBR Incentives and Risks: The proposed approach allocates the total revenues that would have been received had the prices for all rate classes been uniformly indexed by the PBR Price Cap formula. That is, the proposed Phase II approach and adjustments does not materially alter the PBR incentives and risks to FortisAlberta and its customers. 2) Consistent with paragraph 996 of Decision , the proposed approach uses the revenue requirement resulting from going-in rates adjusted by the PBR formula (including the I-X mechanism, K factors, Y factors and Z factors) and the latest updated billing determinants. 3) Simple and can be applied in any PBR year: The proposed method is simple and can be used and applied in any year of the PBR term. For example, in its Argument and Reply, FortisAlberta has recommended Scenario 2, which makes Phase II adjustments and implements a transitional plan towards rates that achieve 100% R/C ratios by rate class over a two year period 2014 and The proposed method accommodates this, while the method inquired about in part (e) above would likely require a recurring Exhibit , attachment, AUC-FAI Exhibit , AUC-FAI-049(f). Exhibit , AUC-FAI-049(f). 68 AUC Decision (January 27, 2014)

75 backward examination of 2012, establishing growth assumptions by rate class from year to year, and an incremental stepping calculation from year-to-year to the year in question, whether that be 2014, 2015, 2016, 2017 or beyond. 4) Regulatory Streamlining Benefits: Any consideration of the part (e) method, or any other potential method for determining and allocating revenue in a PBR test year, will create an incentive for bias among parties, whether that is the individual customer groups, rate classes, or the Company. That is, by considering different alternatives for dealing with Phase II under PBR in each and every year, rate classes will have the natural incentive to selectively pick the method that is financially beneficial to their rate class in the given year. Further, it may also raise questions around the companies Phase II proposals if the Company has the ability to affect the Target PBR Revenue for a given year by the method it employs to determine that amount. Such potential for bias from Phase II did not exist in Phase II applications under Cost of Service, as the Phase II Revenue Requirement was established independent of the Phase II methods and processes. FortisAlberta expects that such an outcome would work counter to any measure of regulatory streamlining that the AUC had desired under PBR, as it would have the effect of linking and creating a relation between Phase II and PBR. Conversely, if the Commission were to approve FortisAlberta s proposed method for the PBR term, the result is essentially a mechanical exercise. The rates are increased by I-X each year, the rate class billing determinants grow in accord with the Commission approved forecasting method, a Target PBR Revenue results from the multiplication of the two rates and billing determinants, and is then allocated based on the results (allocators) from the most recently approved Cost Causation Study. This would keep the PBR and Phase II processes unrelated. 5) One Year Transition: Given that 2012 is the last year of the Cost of Service before the five year PBR term, the part (e) method is, pragmatically and realistically, only available to be used for 2013 and In any event, the part (e) method does not generate the revenues that would have been received had the 2012 Going-in prices for all rate classes been uniformly indexed by the PBR Price Cap formula and as such is inappropriate from the Company s perspective. That is, the part (e) approach would be counter to the Commission s statement that The Commission considers that PBR is unrelated to the requirement to periodically update rates through a Phase II process In its correspondence of October 29, 2013, the CCA indicated that, for the reasons outlined by Fortis in response to AUI-FAI-49(f), it did not object to Fortis proposed method to apply the allocators from the 2012 cost of service study to the target 2013 PBR revenue using the 2013 billing determinants, as set out in this Phase II application. 190 Commission findings 341. As set out in Section 15.1, for purposes of regulatory efficiency and to mitigate the potential impact of a DAR on customer bills, the Commission accepted Fortis proposal that its final 2012 and 2013 rates will not be reflective of the 2012 cost allocation and the proposed Phase II rate design methodologies developed in the current Phase II proceeding. The Commission found Fortis proposal that the new cost allocation and rate design proposed in this Phase II application be incorporated in the updated 2014 PBR rates and PBR rates going forward, to be reasonable. 190 Exhibit , CCA correspondence dated October 29, AUC Decision (January 27, 2014) 69

76 342. Two methods of calculating the updated 2014 PBR rates were discussed in this proceeding. Under the first method, proposed by Fortis in the application, the 2012 Phase II cost allocation and rate design are applied to the target 2013 PBR revenue and approved 2013 forecast billing determinants to develop the notional, newly structured 2013 rates. These rates will not go into effect in 2013 and will only serve as an intermediate step in calculating the 2014 updated PBR rates. The updated 2014 PBR rates are then calculated by escalating the newly structured 2013 rates by the 2014 I-X index and including any approved 2014 K, Y, and Z factor rate adjustments. 191 This has the effect of re-establishing the PBR rates in 2014 to give effect to the approved Phase II methodologies Under the second method, set out in AUI-FAI-49(e), the 2012 Phase II cost allocation and rate design are applied to the approved 2012 revenue requirement and 2012 billing determinants to develop the notional revised 2012 rates. Then, the newly structured 2013 rates are calculated by applying the going-in rate adjustments and the 2013 I-X index to the revised 2012 rates, and including any approved 2013 K, Y, and Z factor rate adjustments. Finally, the updated 2014 PBR rates are calculated by escalating the newly structured 2013 rates by the 2014 I-X index and including any approved 2014 K, Y, and Z factor rate adjustments The primary difference between the two methods is that the approach set out in AUI-FAI-49(e) would apply the I-X index to the notional revised 2012 rates. This does not take into account any changes in billing determinants. 192 In contrast, Fortis proposed method incorporates changes in billing determinants, effectively re-setting the 2014 PBR rates to incorporate the effects of changes in billing determinants Given that this Phase II application was originally intended to establish the 2012 cost of service rates, the Commission expected Fortis to use the method set out in AUI-FAI-49(e), i.e., to rely on the 2012 cost allocation study and the 2012 billing determinants to set the 2012 rates. The Commission is cognizant that Fortis could not complete the Phase II application in Additionally, the Commission acknowledges that due to the passage of time, this application is unique from past Phase II applications in that, rather than just setting rates by rate class for a single cost of service test year, such as 2012, FortisAlberta must also make proposals for setting rates for 2013 and 2014, the initial years of PBR. 194 Therefore, the Commission accepts Fortis method to apply the 2012 Phase II cost allocation and rate design to the target 2013 PBR revenue and approved 2013 forecast billing determinants Furthermore, Fortis approach re-sets the PBR rates effective January 1, 2014, and accordingly, the approach used by Fortis is consistent with the Commission s determinations at paragraph 996 of Decision on how the Phase II applications will be implemented under PBR. In particular, consistent with the findings in that paragraph, Fortis method, uses the revenue requirement resulting from going-in rates adjusted by the PBR formula (including the I- X mechanism, K factors, Y factors and Z factors) and the latest updated billing determinants. Fortis also pointed out that this method can be used and applied in any year of the PBR term. 195 The CCA submitted it did not object to Fortis proposed method Exhibit 1, application, paragraph 151 and Exhibit , AUC-FAI-049(a) and (c). Exhibit , AUC-FAI-049(f). Exhibit , AUC-FAI-049(f). Exhibit 1, application, paragraph 3. Exhibit , AUC-FAI-049(f), point (3). Exhibit , CCA correspondence dated October 29, AUC Decision (January 27, 2014)

77 347. For the reasons above, the Commission approves Fortis methodology to calculate its updated 2014 rates as set out in the application and demonstrated in Schedule 1, 197 subject to complying with the Commission s directions elsewhere in this decision Additionally, Fortis noted that the target 2013 PBR revenue used in this application is a placeholder and may change depending on the outcome of the related PBR proceedings. 198 The Commission directs Fortis, in calculating the target 2013 PBR revenue as part of the compliance filing to this decision, to use the 2013 PBR Second Compliance Filings rates and rate riders, as approved in Decision In Decision , the Commission approved, on an interim basis, the 2014 PBR rates for Fortis, which are currently in effect. These interim rates are not reflective of the 2012 Phase II cost allocation and rate design methodologies developed in this proceeding because the current proceeding was still ongoing as of January 1, Following the finalization of the 2012 Phase II methodologies in the compliance filing to this decision, the Commission will direct the implementation of the updated 2014 PBR rates. 16 Terms and conditions of service 350. Fortis did not submit terms and conditions of service for approval with the application and confirmed in its response to AUC-FAI-043 that it had not proposed any changes to its terms and conditions of service in this application. 199 Fortis terms and conditions of service were last approved by the Commission on a final basis in Decision where the Commission stated the following: 30. The Commission is not prepared to approve changes to the T&Cs on an interim and refundable basis. The Commission considers that the T&Cs are approved on a final basis until such time as a change to the T&C s is subsequently approved on a final basis. T&Cs include requirements for dealing with numerous customer specific matters, including applications for service, contract levels, disconnections, and fees for various services such as reconnections, off-cycle meter readings and dishonored payments that are applied on an individual customer basis. If the Commission were to approve the T&Cs on an interim basis, including the 2012 MILs agreed to in the NSA, and then subsequently the Commission found that, as part of its review of the 2012 NSA, that subsequent changes were required to the T&Cs, this would require Fortis to go back and true-up any differences between these changes and the requirements in the T&Cs that were approved on an interim basis. The Commission considers that it may be difficult and potentially costly for Fortis to track this information and that it may be a significant inconvenience to customers. For these reasons, the Commission does not find that it is in the public interest to approve T&Cs on an interim refundable final basis. Accordingly, Fortis s request for interim approval of its amended T&Cs is denied Subsequently in Decision , where the Commission approved Fortis NSA, the Commission stated that it had verified that the only changes Fortis had made to its approved terms and conditions of service were revisions to the maximum investment level tables in Appendix B to reflect the 2012 maximum investment levels agreed to in the NSA. The Exhibit 9, Schedule 1. Exhibit 1, application, paragraph 49. Exhibit 43.02, AUC-FAI-043. AUC Decision (January 27, 2014) 71

78 Commission further stated that the maximum investment levels agreed to in the NSA were effective only on a going forward basis Subsequently in Decision dated December 23, 2014, the Commission approved the customer and retailer terms and conditions of distribution service set out in Appendix 5 and Appendix 6 of that decision, effective January 1, Aside from adjusting the maximum investment levels and Appendix A fees, to reflect I-X increases, Fortis customer and retailer terms and conditions were not modified for The Commission finds that Fortis current term and conditions of service remain in effect AMI meters 354. The Commission issued information requests to Fortis to understand its policy and practices as they relate to AMI meters and the removal of AMI meters for health reasons Fortis indicated that it responds to customer requests to remove AMI meters or converters by providing information about meter reading systems. In most cases, once customers understand that the meter at their location sends the meter read information over power lines and cannot identify or control appliances, customers usually do not continue to pursue the removal of an AMI meter Fortis submitted that its policy currently does not permit the removal of an AMI meter but suggested that the relocation of an AMI meter, if feasible, may be a strategy discussed with customers who want their meter removed. If a customer requests relocation of an AMI meter, the customer is responsible for associated costs under Section cost of optional facilities or Section 5.3 relocation of facilities in Fortis terms and conditions of service Fortis stated that it is not aware of any verifiable evidence that AMI meters cause or exacerbate health conditions and has no evidence that the removal of a meter will improve symptoms of those individuals who assert sensitivity to the low levels of frequency emitted by these meters Fortis indicated that it would require AUC approval of a program to allow customers to opt-out of automated meter reads and the approval of a fee based on the incremental costs associated with an opt-out. 203 Commission findings 359. The Commission is not persuaded that there are likely to be health impacts related to AMI meters. However, it recognizes that some parties may object to AMI meters for health or other reasons. The Commission considers it prudent to provide an alternative for those customers who object to AMI meters and allow them to choose an alternative metering method for Fortis to obtain billing information Decision , Section 4.3.7, Terms and conditions of service and MILs. Exhibit 43.02, AUC-FAI-042. Exhibit 43.02, AUC-FAI-042 (c). Exhibit , AUC-FAI-047 (b). 72 AUC Decision (January 27, 2014)

79 360. Given the lack of evidence regarding health risks caused by AMI meters, the Commission concludes that it would not be in the public interest to treat costs associated with providing such alternatives as system costs. Therefore, those who choose an alternative to AMI meters must bear the cost of any such alternate arrangement made with Fortis. The Commission directs Fortis in the refiling to propose alternative arrangements for metering, including recovery of the associated costs for customers who wish to decline the use of an AMI meter. 17 Order 361. It is hereby ordered that: (1) FortisAlberta Inc. is directed to comply with all the directions in this decision and refile its application by March 3, Dated on January 27, The Alberta Utilities Commission (original signed by) Mark Kolesar Vice-Chair (original signed by) Bill Lyttle Commission Member (original signed by) Kay Holgate Commission Member AUC Decision (January 27, 2014) 73

80

81 Appendix 1 Proceeding participants Name of organization (abbreviation) counsel or representative FortisAlberta Inc. (Fortis) J. Walsh ATCO Electric Ltd. (ATCO Electric) L. Keough K. Worton S. Parhar T. Martino J. Janow B. Yee T. Small L. Kerckhof Canada West Ski Areas Association (CWSAA) L. Manning R. Cowburn D. Lynn Consumers Coalition of Alberta (CCA) J. A. Wachowich A. P. Merani R. Retnanandan Desiderata Energy Consulting Inc.(DESI) W. D. Hildebrand G. Smillie EPCOR Distribution & Transmission Inc. (EDTI) D. Gerke N. Lamers I. Abbasi EQUS REA Ltd. (EQUS) D. Evanchuk P. Bourne Industrial Power Consumers Association of Alberta (IPCAA) M. Forster R. Mikkelsen V. Bellissimo J. Cheng North Parkland Power REA Ltd. (NPPREA) D. Evanchuk G. Nicol Potato Growers of Alberta (PGA) T. Hochstein AUC Decision (January 27, 2014) 75

82 Name of organization (abbreviation) counsel or representative Office of the Utilities Consumer Advocate (UCA) T. Marriott R. Daw B. Shymanski R. Bell The Alberta Utilities Commission Commission Panel M. Kolesar, Vice-Chair B. Lyttle, Commission Member K. Holgate, Commission Member Commission Staff J. Petch (Commission counsel) S. Karim C. Burt S. Allen 76 AUC Decision (January 27, 2014)

83 Appendix 2 Summary of Commission directions This section is provided for the convenience of readers. In the event of any difference between the directions in this section and those in the main body of the decision, the wording in the main body of the decision shall prevail. 1. The Commission has considered the proposed allocators for each Y factor and K factor, as shown in the table above. The Commission is satisfied that the approach adopted by Fortis to allocate the approved K and Y factors is consistent with the Commission s directions in decisions and Accordingly, the Commission approves the allocation of K, and Y factors as proposed by Fortis. Fortis is directed to use these allocators throughout the PBR term for any approved K and Y factors..... Paragraph Given that the feeder representation in the CAM model has not been updated since 2005, despite the growth in the system, and given that Fortis indicated the model is not representative of the system for some rate classes, the Commission directs the following. Fortis shall add a further 20 feeders (or more if required to make the sample representative) to the CAM model prior to the next cost of service study and undertake a statistical analysis to ensure that the updated CAM model is representative.. Paragraph The Commission has reviewed the record with respect to assumptions. For example, Fortis has assumed that all residential customers have the same on-peak utilization factor regardless of whether they have a dedicated or shared transformer. However, for practical reasons of data availability and cost benefit trade-offs, despite the short-comings of this assumption, it may be the best option. As the CAM model has been in use for many years, without a detailed review, the Commission directs that in its next Phase II application, Fortis shall provide a list of all assumptions and explain the rationale for each assumption that impacts the allocation of costs to one or more rate classes by a material amount.... Paragraph The Commission directs Fortis to update its Rate 63 multiplier study for the purposes of its next Phase II application. Specifically, the Commission directs Fortis to undertake a statistical analysis to assess whether the sample of 25 customers is representative of the total rate class. If not, Fortis is directed to increase the sample size to make it representative of the total rate class prior to its next cost of service study.... Paragraph However, given the Commission s findings on the potential benefits of adopting a POD-specific allocation based on hourly load data, the Commission directs Fortis to further explore the costs and benefits of this approach at the time of its next Phase II application.... Paragraph Accordingly, although the Commission considers that revenue-to-cost ratios should be adjusted toward 100 per cent, the Commission must also consider possible rate shock as a result of adjusting revenue-to-cost ratios. The Commission directs Fortis to adjust the revenue-to-cost ratios for all rate groups as close as possible to the generally accepted 95 per cent to 105 per cent revenue-to-cost ratio range, such that the average total bill impact for each rate class does not exceed a ten per cent increase.... Paragraph The Commission concludes that mirroring of the AESO s charges in Fortis rates is consistent with cost causation and directs Fortis, in its refiling, to reflect CWSAA s proposed new billing determinant of un-ratcheted kw of billing period capacity for rates 61 and 63. The Commission recognizes that this direction will require billing system AUC Decision (January 27, 2014) 77

84 changes with an approximate cost of $10,000 to $15,000. However, the Commission considers the benefits of adopting CWSAA s proposal outweigh the related costs.... Paragraph The Commission is not convinced that CWSAA s proposal should apply to rates 41 and 45 at this time as application to those customer classes has not yet been sufficiently explored. The Commission directs Fortis to consider whether CWSAA s proposal should apply to Rate 41 and Rate 45 in its next Phase II application.... Paragraph The Commission is not convinced that Fortis proposed residential Rate 11 rate structure and the resulting amounts are justified at this time. On this basis, the Commission directs Fortis to maintain the existing residential rate structure.... Paragraph In its next Phase II application, should Fortis propose changes to the residential rate structure, it is directed to provide analysis regarding the need for rate structure changes and a sensitivity analysis of the customer impacts of different block thresholds.... Paragraph Fortis also applied to change its per kilometre charge to a kilometres times demand charge. The Commission considers there is not sufficient information on the record to support the applied-for change. There is no information on the record of the proceeding on the extent to which local property cost differences and distance from the transmission system have been reflected in customer contribution. Likewise, there is no information on the extent to which distance from the transmission system is relevant to cost causation in the case of customers with dedicated feeders and customers that tap into existing distribution lines. Accordingly, the Commission denies the proposed rate structure for Rate 63, and directs that the existing rate structure be retained, with the exception of the proposed change to the service charge approved above.... Paragraph The Commission directs Fortis in its refiling to submit updated bill impact schedules comparing the proposed 2014 rates, as adjusted to comply with directions in this decision, to the current rates.... Paragraph The Commission directs Fortis to reflect the Commission s findings with respect to the complaint by the Rate 44/45 group and the relief approved in this decision in its compliance filing. The Commission directs Fortis to provide revised rate design, bill impacts and rate schedules as applicable. The Commission further directs Fortis to provide in its compliance filing any additional schedules showing how the refund to affected Rate 44/45 customers has been determined.... Paragraph Additionally, Fortis noted that the target 2013 PBR revenue used in this application is a placeholder and may change depending on the outcome of the related PBR proceedings. 204 The Commission directs Fortis, in calculating the target 2013 PBR revenue as part of the compliance filing to this decision, to use the 2013 PBR Second Compliance Filings rates and rate riders, as approved in Decision Paragraph Given the lack of evidence regarding health risks caused by AMI meters, the Commission concludes that it would not be in the public interest to treat costs associated with providing such alternatives as system costs. Therefore, those who choose an alternative to AMI meters must bear the cost of any such alternate arrangement made with Fortis. The Commission directs Fortis in the refiling to propose alternative arrangements for 204 Exhibit 1, application, paragraph AUC Decision (January 27, 2014)

85 metering, including recovery of the associated costs for customers who wish to decline the use of an AMI meter.... Paragraph 360 AUC Decision (January 27, 2014) 79

86 Appendix 3 Rate 44 and Rate 45 rate schedules (return to text) Appendix 3 - Rate 44 and Rate 45 rate sche (consists of 3 pages) AUC Decision (January 27, 2014) 80

87 Phase II Distribution Tariff FortisAlberta Inc. Appendix 3 - Rate 44 and Rate 45 rate schedules Page 1 of 3 FORTISALBERTA INC DT RATES FILING RATES, OPTIONS, AND RIDERS SCHEDULES April 1, 2012 Rates Residential Rate 11 Residential Service... 2 Farm Rate 21 FortisAlberta Farm Service... 3 Rate 23 FortisAlberta Grain Drying Service... 4 Rate 24 REA Farm Service... 5 Rate 26 FortisAlberta Irrigation Service... 6 Rate 29 REA Irrigation Service... 7 Lighting Rate 31 Street Lighting Service (Investment Option)... 8 Rate 33 Street Lighting Service (No Investment Option)... 9 Rate 38 Yard Lighting Service Commercial / O&G Rate 41 Small General Service Rate 44 Oil & Gas (Capacity) Service (Closed) Rate 45 Oil & Gas (Energy) Service General Service Rate 61 General Service Rate 63 Large General Service Rate 65 Transmission Connected Service Rate 66 Opportunity Transmission Special Facilities Charge Options Option A Primary Service Option Option C Idle Service Option Option D Flat Rate Option Option I Interval Metering Option Option M Distribution Generation Credit/Charge Riders Rider A-1 Municipal Assessment Rider Municipal Franchise Fee Ride Municipal Franchise Fee Riders Balancing Pool Allocation Rider Quarterly Transmission Adjustment Rider FortisAlberta s Customer and Retailer Terms and Conditions of Electric Distribution Service provide for other charges, including an arrears charge of 1.5% per month. AUC Decision (January 27, 2014)

88 Phase II Distribution Tariff FortisAlberta Inc. Appendix 3 - Rate 44 and Rate 45 rate schedules Page 2 of 3 FortisAlberta Inc DT Rates Filing Rate Schedules RATE 44 OIL & GAS (CAPACITY) SERVICE (CLOSED) Page 12 Effective: January 1, 2012 Availability Rate 44 is available to existing oil and natural gas field services that are unmetered or have demand meters only. These services include pumping and related operations such as; rectifiers, cathodic protection and radio transmitters, and water pumping services. Rate 44 is available to existing services with Operating Demands less than 75 kilowatts. Flat Rate Option D applies to unmetered services. This rate is no longer available for new installations and existing services are being transitioned to Rate 45 as metering is installed. Rate 44 For the first 5 kw of Capacity All additional kw of Capacity Transmission Component Distribution Component Total Distribution Tariff $ /kw-day $ /kw-day $ /kW-day $ /kw-day $ /kw-day $ /kW-day The kw of Capacity is the greatest of: 1. for unmetered and energy metered services, the sum of all connected motors and equipment (1 horsepower equals kw); 2. for demand metered services, the highest Metered Demand in the 12-month period including and ending with the current billing period; or 3. the Rate Minimum of 3 kw. The Metered Demand is the greater of the registered demand in kilowatts or 90% of the registered demand in kilovolt-amperes. The kw of Capacity must bill for a minimum of 12 consecutive months before being reduced. FortisAlberta s Customer and Retailer Terms and Conditions of Electric Distribution Service provide for other charges, including an arrears charge of 1.5% per month. AUC Decision (January 27, 2014)

89 Phase II Distribution Tariff FortisAlberta Inc. Appendix 3 - Rate 44 and Rate 45 rate schedules Page 3 of 3 FortisAlberta Inc DT Rates Filing Rate Schedules Page 13 RATE 45 OIL & GAS (ENERGY) SERVICE Effective: January 1, 2012 Availability Rate 45 Rate 45 is available to oil and natural gas field services including pumping and related operations such as rectifiers, cathodic protection and radio transmitters and to water pumping services. Rate 45 is available to services with Operating Demands less than 75 kilowatts and have a demand and energy measurement meter. Transmission Component Distribution Component Total Distribution Tariff For all kwh delivered /kwh /kwh For the first 5 kw of Capacity $ /kw-day $ /kw-day $ /kW-day All additional kw of Capacity $ /kw-day $ /kw-day $ /kW-day The kw of Capacity is the greatest of: 1. the highest Metered Demand in the billing period; 2. 85% of the highest Metered Demand in the 12 month period including and ending with the current billing period; 3. the Rate Minimum of 3 kw. The Metered Demand is the greater of the registered demand in kilowatts or 90% of the registered demand in kilovolt-amperes. FortisAlberta s Customer and Retailer Terms and Conditions of Electric Distribution Service provide for other charges, including an arrears charge of 1.5% per month. AUC Decision (January 27, 2014)