COSA Results for 2020

June 17, 2019

Based on your financial forecast for 2020, and all of the other data provided, we have completed our initial COSA and a summary of the results are included in this memo. This reflects several changes resulting from our discussion with the Risk Management Committee. The results are a starting point for rate design, which is not addressed in this memo. 

Revenue Requirements

 The revenue requirements show the 2020 forecast results and will be the test year for the COSA.

The following are the revenue requirements assumptions: 

• Power supply costs are based on the 2019 budget amounts for power purchases and transmission costs as a starting point.  The resulting cost per kWh for power supply, and $-kW-month cost for transmission, were then applied to the 2020 forecast loads.  The resulting 2020 forecast is higher than what is in your 2020 financial forecast because it is calculated out based on the loads each month rather than based on 2019 numbers plus a % increase.  This is important because we want the power costs and revenues to both be based on the detailed load amounts.
• The transmission costs were increased by the 53% proposed by Northwestern.  The resulting annual cost for 2020 is projected at $728,241.
• Other costs were based on the 2020 financial forecast, split into the various detailed FERC accounts based on the 2018 actual costs by detailed FERC account.
• The revenue requirement includes depreciation, interest and operating margins as provided in the financial forecast.  The margin was adjusted to equal the exact amount collected after the revenues and costs included in the COSA.  It is slightly different than what was in the financial forecast.
• The revenue requirement deducts revenue for other sources and therefore only reflects the revenues that need to be collected in retail rates.

The following table provides a summary of the projected revenue requirements.  There is no need for a rate increase for 2020.  The amounts in the 2020 financial forecast are also shown for comparison purposes.  The difference from the financial forecast is that power supply costs are higher by $240,446, partly due to higher transmission charges and partly due to the way the cost was calculated based on loads.  The revenues are higher by $300,332 based on the detailed revenue calculations.  This results in a margin that is higher by $59,887.

Table 1

2020 Revenue Requirement

Cost of Service Study

A cost of service analysis (COSA) is concerned with the equitable allocation of the revenue requirement to the various customer classes of service.  The following are the assumptions used in the COSA:

• Power supply purchases are allocated on the basis of energy. Wholesale transmission costs are allocated on the sum of the 12 monthly coincident peaks.
• BEC’s internal transmission costs are allocated on the basis of the annual system coincident peak.
• Both the 100% demand and minimum system methods are used for allocating distribution costs, and a non-coincident peak is used for allocation.
• Customer service and accounts costs are allocated on the basis of the number of customers.
• The Southern Exit amortization listed as a separate item under A&G and is classified as customer-related and allocated on the number of customers weighted for the average kWh per customer.
• Most other A&G costs are allocated on the basis of assignments developed by BEC.  Based on the total of the various assignments made by BEC, the result is 5% distribution and 95% customer-related.  The distribution portion was allocated on the basis of the NCP while the customer portion was allocated on the basis of customers weighted for the average kWh per customer.
• Depreciation and interest payments are assigned on the basis of plant.
• Margin is allocated on the basis of all revenue.
• AMI data for some residential customers during the winter months was used to develop load factors and coincidence factors for the residential class.  Hourly data for the Belfry substation was used to help develop load factors and coincidence factors for the irrigation class.
• Revenues were based on specific rate components times the load forecast of customers, demand and energy amounts.  It differs from the financial forecast because of the more detailed approach.

 2020 COSA results are summarized for the minimum system approach in Table 2 and for the 100 percent demand approach in Table 3. 

When looking at the revenue to cost ratios it is important to reflect the uncertainty inherent in any COSA due to methodology approaches and uncertainty in the load data used to allocate costs.  For that reason, a range of 90% to 110% is typically used to measure whether customer classes are paying an appropriate share of costs. 

Based on the results under the minimum system methodology, all classes are in the 90-110% range and do not need an increase or decrease.  In the 100% demand case, all classes except Irrigation are in the 90-110% range. 

For the Irrigation class, the COSA shows that they are still underpaying their cost of service in the 100% Demand case but are above 90% in the Minimum System case.  This is not an unexpected result due to the nature of the service.  On the power supply side, the costs of purchased power do not differ based on season and so there is no benefit to Irrigators related to their summer usage.  Power costs may have been more seasonal in the past, but BEC has changed its power supply source and rates are flat across all hours.  Because they do not peak at the time of the annual peak for BEC they are not allocated any of the transmission-related costs, however, these costs are small for BEC. 

The issue is with the distribution costs.  Distribution facilities are designed based on the maximum peak load that the customer might have throughout the year.  For this reason, the distribution costs are allocated on the basis of the non-coincident peak.  The Irrigators receive a full allocation based on their annual peak loads that occur in the summer months.  Unlike other classes, these distribution costs must be spread out over a 4-5 month period rather than the full 12-month period used by other types of customers. 

If the Board determines that is appropriate to increase Irrigation rates, it may be difficult to determine who would get a corresponding rate decrease.  Under the 100% demand approach, it would go to the residential and small commercial class.  Under the minimum system study, it would go to the large commercial and industrial classes.  However, given the small amount of revenues associated with the Irrigation class, any offset to the other rates would be small.

Unit Costs

Based on the COSA, Tables 4 and 5 (At the end of the memo) show the unit costs for each customer class.  This will be used to assist in developing rate options.

Next Steps

Once we make any adjustments in the revenue requirements and COSA, we can finalize the results and use the per unit costs as an input in the rate design process.  Based on the revenue to cost ratios resulting from the COSA, we will need policy direction from the Board as to whether there needs to be an overall rate increase or decrease to any of the classes.  Ideally this decision would be made prior to rate design so we know what our revenue targets should be.  It might be possible to do rate design first with revenue set to equal current revenues by class, and then have any overall increases or decreases applied after that is done.

Given the discussion among Board members and staff, the rate design step will be much more interactive and take more time than the COSA itself.  In that step, we will identify several options for each class and compare them in terms of design, revenue impacts and bill impacts for customers.  The per unit costs from the COSA, BEC analysis of costs by category, and rate designs for other utilities will all play a role in rate design.  It is likely that there will be several iterations of this process before we reach a final rate design proposal.