public document – not public data has been excised · 2017-11-14 · bryce c. haugen rider roll...

Otter Tail Power Company Before the

North Dakota Public Service Commission

Application for Authority to Increase Electric Rates in North Dakota

Case No. PU-17-

November 2, 2017Volume 2C

Direct Testimony and Supporting Schedules

PUBLIC DOCUMENT – NOT PUBLIC DATA HAS BEEN EXCISED

1/3

Volume 2C

Direct Testimony and Supporting Schedule

Otter Tail Power Company

North Dakota General Rate Case Documents

Case No. PU-17-

Volume

1 Notice of Change in Rates – Interim Rate Petition

Index

Filing Letter

Notice of Change in Rates

Alternative Petition for Interim Rates

Interim Supporting Schedules and Workpapers

Summary of Present and Interim Revenue

Interim Tariff Sheets – Legislative

Interim Tariff Sheets – Non-Legislative

2A Direct Testimony and Supporting Schedules

Index

Bruce G. Gerhardson

Policy

Stuart D. Tommerdahl

Allocation Factors

Compliance Items

Bryce C. Haugen

Rider Roll in

Gina S. Ice

Jurisdictional and Class Allocators

Class Cost of Service Study

Class Revenue Responsibilities

Tyler A. Akerman

Rate Base

Operating Statement

Capital Budgeting Process

Christine L. Petersen

Budget Process

Operations and Maintenance Expenses

Pension and Post Employment Expenses

2B Direct Testimony and Supporting Schedules

Index

Kevin G. Moug

Financial Soundness

Capital Structure

Cost of Capital

Robert B. Hevert

Return on Equity

Kirk A. Phinney

Big Stone AQCS and Hoot Lake MATS Capital Projects

David G. Prazak

Rate Design



Case No. PU-17-

Volume

2C Direct Testimony and Supporting Schedules

Index

Brian H. Draxten

Sales and Revenue Forecast

Peter E. Wasberg

Employee Compensation

2D Proposed Legislative and Non-Legislative Tariff Sheets

Index

Proposed Tariff Sheets – Legislative

Proposed Tariff Sheets – Non-Legislative

3 Supporting Information

Index

Supporting Information

A. Jurisdictional Financial Summary Schedules

1. Summary of Revenue Requirements – 2018 Test Year

2. Summary of Revenue Requirements – Jurisdictional

B. Rate Base Schedules

1. Rate Base Summary

2. Rate Base Components – 2018 Test Year

3. Rate Base Components - 2018 Test Year to Most Recent General

Rate Case

4. Cash Working Capital

5. Rate Base Adjustments

6. Summary of Approaches and Assumptions Used

7. Rate Base Jurisdictional Allocation Factors

C. Operating Income Schedules

1. Jurisdictional Statement of Operating Income

2. Statement of Operating Income - Jurisdictional

3. Statement of Operating Income – 2018 Test Year

4. Statement of Operating Income – 2018 Test Year to Most Recent

General Rate Case

5. Computation of Federal and State Income Taxes

6. Computation of Deferred Income Taxes

7. Development of Federal and State Income Tax Rates

8. Operating Income Statement Adjustments Schedule

9. Summary of Approaches and Assumptions Used

10. Operating Income Statement Allocation Factors

D. Rate of Return / Cost of Capital Schedules

1. Summary Schedule

2. Cost of Long-Term Debt

3. Cost of Short-Term Debt

4. Common-Equity



Case No. PU-17-

Volume

3 E. Rate Structure and Design Information

1. Test Year 2018 Operating Revenue Summary Comparison

2. Test Year 2018 Operating Revenue Detailed Comparison

3. Class Cost of Service Study

F. Other Supplemental Information

1. Annual Report

2. Gross Revenue Conversion Factor

4A Work Papers

Index

A. 2018 Test Year Workpapers

1. Jurisdictional Cost of Service Study (JCOSS)

2. Class Cost of Service Study (CCOSS)

3. Functionalization

4. Input Summary

5. 2018 Test Year Adjustments

TY-01 – Normalized Plant in Service

TY-02 – Rate Case Expenses

TY-03 – Normalized Plant Outage

TY-04 – Removal of PTC’s

TY-05 – Economic Development

TY-06 – Prorate ADIT

B. 2018 Base Year Workpapers

1. Jurisdictional Cost of Service Study (JCOSS)

2. Functionalization

3. Input Summary

4. Work Papers A-D, ND

C. Interim Cost of Service Study

D. Hevert Cost of Capital Workpapers

4B Lead Lag Study

Index

Lead Lag Study

5 Budget Documentation

Index

O&M Budget Process

Capital Budget Process

1/5

Volume 2C

Direct Testimony and Supporting Schedules:

Brian H. Draxten

Before the North Dakota Public Service Commission

State of North Dakota

In the Matter of the Application of Otter Tail Power Company

For Authority to Increase Rates for Electric Utility

Service in North Dakota

Case No. PU-17-

Exhibit___

ENERGY SALES FORECAST

Direct Testimony and Schedules of

BRIAN H. DRAXTEN

PUBLIC DOCUMENT – NOT PUBLIC (OR PRIVILEGED)

DATA HAS BEEN EXCISED

November 2, 2017

TABLE OF CONTENTS

INTRODUCTION AND QUALIFICATIONS................................................................... 1

PURPOSE AND OVERVIEW OF DIRECT TESTIMONY ............................................. 1

CUSTOMER SALES AND FORECASTS ........................................................................ 3

Sales Trends ............................................................................................................ 3

Methodology and Inputs ......................................................................................... 7

2018 TEST YEAR CLASS SALES FORECASTS .......................................................... 10

Residential ............................................................................................................. 10

Farms ..................................................................................................................... 12

Small Commercial ................................................................................................. 13

Large Commercial ................................................................................................. 14

Other Public Authority .......................................................................................... 16

Street Lighting ....................................................................................................... 17

Unclassified ........................................................................................................... 18

CONCLUSION ................................................................................................................. 19

ATTACHED SCHEDULES

Schedule 1 – Draxten Resume

Schedule 2 – North Dakota Population Growth by County

Schedule 3 – OTP Sales and Revenue Forecast Methodology

NOTE: The data files used to create the sales forecast and revenue forecast are voluminous and

have been provided on accompanying disks.

1 Case No. PU-17-

Draxten Direct

PUBLIC DOCUMENT – NOT PUBLIC (OR PRIVILEGED) DATA HAS BEEN EXCISED

INTRODUCTION AND QUALIFICATIONS 1

Q. PLEASE STATE YOUR NAME AND OCCUPATION. 2

A. My name is Brian H. Draxten. I am employed by Otter Tail Power Company (OTP) as 3

its Manager, Resource Planning. 4

5

Q. PLEASE SUMMARIZE YOUR QUALIFICATIONS AND EXPERIENCE. 6

A. I have worked for OTP for 36 years in various positions, including as a Rates Analyst, 7

Manager of Market Research, and Manager of Budget. I have served in my current position 8

as Manager of Resource Planning since January 2008. In that position, I am responsible 9

for the Load Forecasting and Resource Planning functions at OTP. I oversee the 10

development of the retail sales forecast, the related demand forecast, and the resource plan. 11

I also manage the related analyses for these processes. 12

I have a Bachelor of Arts degree in accounting with a minor in business finance 13

from Moorhead State University in Moorhead, Minnesota. A copy of my resume is 14

included as Exhibit ___ (BHD-1), Schedule 1. 15

PURPOSE AND OVERVIEW OF DIRECT TESTIMONY 16

Q. WHAT IS THE PURPOSE OF YOUR DIRECT TESTIMONY? 17

A. I will discuss OTP’s energy forecasting process, including the 2018 Test Year forecast 18

customer count and energy sales to be used in setting rates in this proceeding. 19

20

Q. PLEASE PROVIDE A BRIEF OVERVIEW OF YOUR DIRECT TESTIMONY. 21

A. OTP is forecasting 1,810 gigawatt hours (GWh) of North Dakota sales in the 2018 Test 22

Year. The class-by-class 2018 Test Year sales are shown in Table 1 below. 23

2 Case No. PU-17-

Draxten Direct


Table 1 1

2018 Test Year Forecasted Sales 2

3

4

The 2018 Test Year sales are 2.5 percent higher than sales from OTP’s 2008 Rate 5

Case (Case No. PU-08-862), which equates to approximately 0.25 percent annual average 6

growth over the 10-year period.1 OTP witness Mr. Bruce G. Gerhardson discusses the 7

impact of the relatively small change between the 2007 Test Year sales and the 2018 Test 8

Year sales in his Direct Testimony. 9

10

Q. HOW IS YOUR DIRECT TESTIMONY ORGANIZED? 11

A. In Section III, I discuss trends in OTP’s North Dakota sales and our forecasting 12

methodology. In Section IV, I discuss our class-by-class forecasts. Section V contains my 13

conclusion. 14

15

1 OTP’s 2008 Rate Case used a 2008 historical Test Year and sales were based on 2007 actual (not weather

normalized) sales with one adjustment for a large industrial customer coming online in 2008. This resulted in 2007

Test Year sales of 1,766 GWh.

Class kWh Percent of Total Sales

Residential 613,950,110 33.9%

Farms 39,840,654 2.2%

Small Commercial 267,760,078 14.8%

Large Commercial 858,966,988 47.5%

Other Public Authority 16,459,056 0.9%

Streetlighting 12,795,273 0.7%

Total 1,809,772,159

Unclassified 3,874,985

3 Case No. PU-17-

Draxten Direct


CUSTOMER SALES AND FORECASTS 1

Sales Trends 2

Q. PLEASE DESCRIBE THE RECENT TRENDS IN OTP’S NORTH DAKOTA SALES. 3

A. As noted in Figure 1 below, OTP’s North Dakota Sales are basically flat since 2015. The 4

average annual growth rate between 2008 actual weather normalized sales and the 2018 5

Test Year forecast is 1.02 percent. 6

7

Q. IS THE PATTERN OF RELATIVELY FLAT SALES CONSISTENT ACROSS ALL 8

CLASSES? 9

A. Yes. As demonstrated in Figure 1, all classes show minimal growth from 2016 levels, and 10

all classes except Large Commercial2 have seen little to no growth from 2008 levels. 11

12

Figure 1 13

14

15

2 As noted above, the 2007 Test Year included one adjustment for a large industrial customer coming online in 2008.

It took several years for that customer’s sales to materialize, which is the largest driver of the increase in Large

Commercial sales between 2008 and 2010. The customer’s sales have remained relatively consistent from 2010

forward, though at levels that are significantly lower than those assumed in the 2008 Rate Case.

4 Case No. PU-17-

Draxten Direct


Q. WHAT FACTORS ARE CONTRIBUTING TO OTP’S RELATIVELY FLAT SALES? 1

A. One factor is little to no customer growth. OTP’s North Dakota customer count has grown 2

by less than 0.25 percent annually since the 2008 Rate Case. Another factor contributing 3

to the flat sales is the agriculture industry. Recent dry years and low crop prices have 4

decreased sales to farm customers and to commercial and industrial customers in related 5

agricultural industries. 6

7

Q. HAVE THE AREAS SERVED BY OTP GROWN SINCE OTP’S LAST RATE CASE? 8

A. No. North Dakota’s population has grown significantly, with total population growing by 9

12.7 percent between 2010 and 2016.3 That growth, however, has been concentrated in the 10

western portion of the State that is not served by OTP. Many of the counties served by 11

OTP have actually experienced population loss over the last several years. This can be 12

seen in Figure 2 below, which is a reproduction of Exhibit ___ (BHD-1), Schedule 2. 13

3 https://www.census.gov/quickfacts/fact/dashboard/ND/PST120216.

5 Case No. PU-17-

Draxten Direct


Figure 2 1

North Dakota 2010-2016 Population Change by County4 2

3

4 5

Q ARE THESE TRENDS EXPECTED TO CONTINUE INTO THE FUTURE? 6

A. Yes. Population projections developed by the North Dakota Census Office in January 2016 7

forecast that the trend of western population growth outpacing eastern population growth 8

is expected to continue into the future.5 9

10

4 https://www.census.gov/quickfacts/fact/map/ND/PST120216. 5 North Dakota Census Office Population Projections of the State, Regions and Counties 2016 (Jan. 19, 2016),

available at https://www.commerce nd.gov/uploads/8/2016PopulationsProjectionsFinal.docx.

6 Case No. PU-17-

Draxten Direct


Q. ARE 2018 TEST YEAR CUSTOMER COUNTS CONSISTENT WITH THESE 1

POPULATION FIGURES? 2

A. Yes. Customer counts for the 2018 Test Year sales forecast are consistent with the 3

historical trends of little to no population growth in the areas served by OTP. Table 2 4

below shows the class-level growth in customer counts from 2016 actuals. 5

6

Table 2 7

Comparison of 2018 Test Year Customer Count to 2016 Actuals 8

9

10

Q. IS THE RELATIVELY LITTLE GROWTH ASSOCIATED WITH THE 2018 TEST 11

YEAR SALES FORECAST CONSISTENT WITH REGIONAL AND NATIONAL 12

TRENDS? 13

A. Yes. Midcontinent Independent System Operator, Inc. (MISO) is using a 0.5 percent 14

baseline 20-year energy growth rate for the midpoint assumption in the 2018 MISO 15

Transmission Expansion Planning study.6 The U.S. Energy Information Administration 16

similarly projects growth in electricity use will remain relatively low into the future based 17

on a variety of factors: 18

In recent history, the growth in electricity demand has slowed as older 19

equipment was replaced with newer, more efficient stock, as efficiency 20

standards were implemented and technology change occurred, particularly 21

in lighting and other appliances. The demographic and economic factors 22

6 See MISO MTEP18 Futures, Planning Advisory Committee, p. 13 (June 14, 2017), available at

https://www misoenergy.org/Library/Repository/Meeting%20Material/Stakeholder/PAC/2017/20170614/20170614

%20PAC%20Item%2002a%20MTEP18%20Futures.pdf.

Year 2016 2018 % Change

Residential 45,892 45,851 -0.09%

Farm 1,001 1,005 0.40%

Small Commercial 10,518 10,472 -0.44%

Large Commercial 981 973 -0.82%

Other Public Authority 287 290 1.05%

Streetlighting 203 203 0.00%

Unclassified 77 79 2.60%

7 Case No. PU-17-

Draxten Direct


driving this trend included slowing population growth and a shifting 1

economy toward less energy-intensive industries.7 2

Thus, OTP’s experience of almost no sales growth and forecast of relatively little growth 3

in the 2018 Test Year is consistent with regional and national expectations. 4

Methodology and Inputs 5

Q. PLEASE SUMMARIZE OTP’S SALES FORECAST METHODOLOGY. 6

A. OTP forecasts both a class customer count and a class Use per Customer (UPC). The class 7

customer count is multiplied by the class UPC to arrive at the class kilowatt hour (kWh) 8

sales forecast. OTP uses multiple econometric models to create forecasts for each class of 9

customers. Weather data, economic data, customer counts, and historical usage are all 10

inputs to the models. For each class, multiple economic variables are tested and those that 11

have a statistical relevance to the model are used. When all models are complete, they are 12

compared to actual historic data for reasonableness. Additional detail regarding our 13

customer and sales forecast procedure is included in Exhibit ___ (BHD), Schedule 3. 14

15

Q. DOES OTP USE THIS FORECASTING METHODOLOGY FOR ALL CLASSES? 16

A. Generally, yes, though OTP does manually forecast sales to several of our largest 17

customers. The manual forecasts involve discussions with individual customers to obtain 18

their expected future energy usage as well as our own assessment of various economic, 19

regulatory, legislative factors, and specific industry trends. 20

In North Dakota, we manually forecast sales for pipeline pumping customers, 21

ethanol producers, and certain large manufacturers. We also manually forecast sales to oil 22

pipeline pumping customers in Minnesota. South Dakota has no manually forecasted 23

customers. 24

25

7 U.S. Energy Information Administration, Annual Energy Outlook 2017, p. 76, available at

https://www.eia.gov/outlooks/aeo/pdf/0383(2017).pdf.

8 Case No. PU-17-

Draxten Direct


Q. WHAT IS THE BASIS FOR THE 2018 TEST YEAR SALES FORECAST? 1

A. The 2018 Test Year sales forecast was completed in July 2017. The historical data used to 2

develop the forecast runs through May 2017. 3

4

Q. WHAT IS THE SOURCE OF HISTORICAL CUSTOMER COUNTS AND ENERGY 5

USAGE DATA? 6

A. Historical customer counts and usage data for the sales forecast are from our Customer 7

Information System (CIS). This dataset is internally referred to as “CIS/A”. We have had 8

this dataset in place since the early 1990s. This dataset applies billing corrections in the 9

month that the actual bill occurred, as opposed to including the corrections in the month 10

that they are made within CIS. This dataset more accurately reflects actual usage during 11

the applicable timeframe. The historical customer counts are used to forecast customer 12

counts, while historical usage is used in the UPC forecast. 13

14

Q. DOES OTP USE ECONOMIC DATA IN ITS FORECAST? 15

A. Yes. OTP used economic data for North Dakota, Minnesota, and South Dakota purchased 16

from Woods and Poole Economics. This data is at a county level. OTP serves a large 17

service territory and in some instances, OTP serves only a small percentage of the 18

population in any given county. To prevent interference from counties where OTP only 19

serves a small percentage of the total county population, economic data from any county 20

where OTP serves less than 10 percent of the population of the county is excluded. 21

22

Q. HOW IS ECONOMIC DATA USED IN THE SALES FORECAST PROCESS? 23

A. Economic data is generally used to forecast customer counts. It is not used in the UPC 24

forecasts because economic growth does not have a significant impact on usage and what 25

little growth that may occur as a result of economic growth is offset by the improvement 26

in energy efficiencies of equipment and household products. 27

28

9 Case No. PU-17-

Draxten Direct


Q. PLEASE DESCRIBE HOW WEATHER DATA IS USED IN THE SALES FORECAST. 1

A. Weather data is the primary driver used in the UPC forecasts. OTP uses 20 years of 2

historical weather data paired with 20 years of usage data to develop a relationship between 3

weather and UPC. This relationship is used in the UPC model to predict future UPC. 4

OTP also uses “normal” weather to determine what historic sales would have been 5

during a normal weather year. By applying “normal” weather to historical data, OTP is 6

able to remove most of the effects of non-normal weather on historical sales during those 7

years. Similarly, the 2018 Test Year sales forecast is based on a normal weather year. 8

Q. HOW DOES OTP CALCULATE NORMAL WEATHER? 9

A. OTP serves an extremely large geographic foot print, covering over 70,000 square miles. 10

OTP uses data provided by the weather division of Schneider Electric (www.schneider-11

electric.com). They capture weather measurements from fourteen weather stations situated 12

throughout OTP’s service area. Schneider Electric goes through multiple data “cleansing” 13

processes to ensure the data is correct and missing values are filled. 14

OTP has North Dakota customers near eight of these weather stations, and these 15

stations are used in the North Dakota weather normalization process. OTP uses data from 16

these eight weather stations to calculate daily average Heating Degree Days (HDD) and 17

Cooling Degree Days (CDD) for North Dakota. HDD and CDD are measurements 18

designed to reflect the energy needed to heat or cool a building. It is derived from 19

measurements of outside air temperature. The heating and cooling requirements for a given 20

structure at a specific location are considered to be directly proportional to the number of 21

HDD and CDD at that location. The daily average temperature is compared to a base 22

temperature to determine HDD and CDD. The base temperature used in OTP’s weather 23

normalization was 55 degrees Fahrenheit for HDD and 65 degrees Fahrenheit for CDD. 24

These are calculated for each month and are then averaged over a 20-year period to create 25

normalized HDD and CDD. 26

27

10 Case No. PU-17-

Draxten Direct


Q. HOW DOES OTP DETERMINE WHICH DATA INPUTS ARE STATISTICALLY 1

RELEVANT IN ITS FORECAST MODELS? 2

A. Various independent variables were used in the creation of the models. Generally, 3

variables with a t-statistic value of the absolute value of 2 or greater are considered good 4

predictor variables, and OTP attempted to only select variables that met this criterion. 5

Some exceptions do exist – for example, some monthly binary variables do not meet this 6

criterion for some of the months but were still used in the model because there are enough 7

weather differences month-to-month to warrant the inclusion of these variables. 8

In addition to monthly binary variables, other binary variables are used in the 9

models. While OTP uses a large amount of historical data to better predict future sales, 10

recent trends in usage patterns may be much different than usage patterns earlier in the 20-11

year period. For example, there is a clear change in usage patterns of some classes starting 12

around 2014 due to extreme weather and economic issues related to commercial customers. 13

Both the number of customers and usage patterns often change during periods of economic 14

change, and binary variables help reflect that change. 15

OTP also uses other statistical tests (the R-squared, Mean Absolute Percent Error 16

(MAPE), and the Durbin Watson Statistic) to check the reasonableness of the models. Each 17

model has also been reviewed graphically by various areas within OTP, compared to 18

historical data, and found to be reasonable. 19

2018 TEST YEAR CLASS SALES FORECASTS 20

Residential 21

Q. WHAT ARE THE MAIN VARIABLES IN FORECASTING RESIDENTIAL CLASS 22

SALES? 23

A. North Dakota Residential class sales are the product of forecasted customer count and UPC 24

for the Residential class. Forecasted customer count is primarily a function of the number 25

of households and the historical population. Residential class UPC is driven mostly by 26

weather and historical UPC. 27

11 Case No. PU-17-

Draxten Direct


1

Q. PLEASE DISCUSS HOW WEATHER IMPACTS THE FORECAST OF 2

RESIDENTIAL CLASS SALES. 3

A. This class is extremely weather-sensitive, so weather is an important predictor of sales. The 4

objective in forecasting this class is to forecast it as though it is a “normal” weather year – 5

neither extremely warm nor extremely cold. OTP does have a significant number of 6

electric heating customers, so this class has more seasonal deviation than most other 7

classes. 8

9

Q. HOW DO 2018 TEST YEAR RESIDENTIAL CLASS SALES COMPARE WITH 10

RECENT ACTUAL, WEATHER NORMALIZED RESIDENTIAL CLASS SALES? 11

A. The 2018 Test Year Residential sales are 0.16 percent lower than actual 2016 weather 12

normalized Residential sales. This is generally consistent with the recent past. Between 13

the years of 2007 and 2014, this class was growing at a rate of 1.89 percent. Growth ended 14

in 2015, with sales decreasing by 1.51 percent in 2015 and by 1.85 percent in 2016. While 15

customer counts were up very slightly in 2015 and 2016, the growth in customer count was 16

not enough to counter a decrease in UPC, which was likely related to energy efficiency 17

seen in household appliances, electronics, and LED lighting. Figure 3 shows weather 18

normalized historical sales and forecasted sales for the Residential class. 19

20

12 Case No. PU-17-

Draxten Direct


Figure 3 1

2

Farms 3

Q. WHAT ARE THE MAIN VARIABLES IN FORECASTING FARM CLASS SALES? 4

A. North Dakota Farm class sales are the product of forecasted customer count and UPC for 5

the Farm class. OTP has found the number of households to be a significant predictor of 6

Farm customer counts. Farm class UPC is mainly driven by weather. 7

8

Q. PLEASE DISCUSS HOW WEATHER IMPACTS THE FORECAST OF FARM CLASS 9

SALES. 10

A. Much of the Farm class load is grain drying. When and whether grain might require drying 11

is difficult to predict and is largely dependent on precipitation levels. Grain drying 12

sometimes occurs in August, September, October, or November – or not at all. 13

Q. HOW DO 2018 TEST YEAR FARM CLASS SALES COMPARE WITH RECENT 14

ACTUAL, WEATHER NORMALIZED FARM CLASS SALES? 15

A. We forecast 2018 Test Year Farm class sales to be 5.2 percent greater than actual 2016 16

weather normalized Farm class sales, echoing the steady growth that has occurred in this 17

class since 1997. As shown in the figure below, recent sales to the Farm class have been 18

quite variable: for example, in 2014, sales grew by 26.73 percent from the prior year, while 19

13 Case No. PU-17-

Draxten Direct


in 2015, sales fell by 16.06 percent. The 2018 Test Year forecast is consistent with the 1

longer-term (since 1997) upward trend in sales for this class. Farm class sales since 2008 2

are shown in Figure 4 below: 3

4

Figure 4 5

6

7

Small Commercial 8

Q. WHAT ARE THE MAIN VARIABLES IN SMALL COMMERCIAL CLASS SALES? 9

A. North Dakota Small Commercial class sales are the product of forecasted customer count 10

and UPC for the Small Commercial class. Gross Regional Product is the main predictor 11

of customer count in this class. The main driver of UPC is weather. A downturn in the 12

agricultural economy and a spike in propane prices for the winter of 2013-14 also had 13

significant impacts on this customer class. See a more detailed discussion of these factors 14

in the Large Commercial customer class below. 15

14 Case No. PU-17-

Draxten Direct


Q. HOW DO 2018 TEST YEAR SMALL COMMERCIAL CLASS SALES COMPARE 1

WITH RECENT ACTUAL, WEATHER NORMALIZED SMALL COMMERCIAL 2

CLASS SALES? 3

A. OTP forecasts 2018 Test Year sales for this class to be 0.78 percent higher than actual 2016 4

weather normalized Small Commercial class sales. The slight increase in 2018 Test Year 5

sales is due to an increase in UPC. The 2018 Test Year Small Commercial sales are 6

generally consistent with sales to this class since 2010.8 Small Commercial class sales 7

since 2008 are shown in Figure 5 below: 8

9

Figure 5 10

11

12

Large Commercial 13

Q. WHAT ARE THE MAIN VARIABLES IN LARGE COMMERCIAL CLASS SALES? 14

A. Except for a few manually forecasted customers, North Dakota Large Commercial class 15

sales are the product of forecasted customer count and UPC. The Large Commercial Class 16

8 OTP experienced extreme weather in 2013 and 2014. OTP’s internal research indicates that this extreme weather

is the reason 2013 and 2014 Small Commercial sales are above the 2010-2016 trendline.

15 Case No. PU-17-

Draxten Direct


customer count model is based on historical customer counts. The UPC model uses 1

weather and historical sales to forecast the UPC. 2

Q. HOW DO 2018 TEST YEAR LARGE COMMERCIAL CLASS SALES COMPARE 3

WITH RECENT ACTUAL, WEATHER NORMALIZED LARGE COMMERCIAL 4

CLASS SALES? 5

A. The 2018 Test Year Large Commercial class sales are 0.75 percent lower than actual 2016 6

weather normalized Large Commercial class sales. Large Commercial class sales since 7

2008 are shown in Figure 6 below: 8

9

Figure 6 10

11

12

Q. PLEASE EXPLAIN HOW CHANGES IN CUSTOMER COUNT CAN AFFECT 13

SALES TO THIS CLASS. 14

A. Given these are very large customers, slight changes in customer count or adding one or 15

two very large customers can dramatically affect the total sales to this class. For example, 16

in 1999, seven customers were added to this class, one being a very large customer (much 17

larger than the class average). This caused actual 1999 weather normalized Large 18

Commercial sales to grow by approximately 7 percent from actual 1998 weather 19

16 Case No. PU-17-

Draxten Direct


normalized Large Commercial sales. Similarly, a large ethanol operator began taking 1

service from OTP in late 2008 and ramped up production over the next two years. The 2

addition of this operator was the primary driver of OTP’s actual, weather normalized Large 3

Commercial sales increasing by 9 percent in 2009 and 9 percent again in 2010. 4

5

Q. HAS OTP PERFORMED ANY OUTREACH TO BETTER UNDERSTAND ITS 6

SALES TO LARGE COMMERCIAL CUSTOMERS? 7

A. Yes. OTP conducted an informal survey of North Dakota Large Commercial customers 8

last year to try to better understand the decrease in sales between 2014 and 2015. The 9

informal survey indicated that some former customers that left the system went out of 10

business and are not likely to return. We were also told by customers that the agricultural 11

economy is down, and this has affected production in some areas. Additionally, propane 12

prices spiked during the winter of 2013-14. Between 2008 and 2013, propane prices were 13

generally in the $1.50-$2.00 per gallon range. During the 2013-14 winter, prices went as 14

high as $4.50 per gallon. Starting in 2015, prices dropped generally to $1.00-$1.50 per 15

gallon. This significant price increase had the effect of customers switching some of their 16

heating and processing load to electricity in 2014. As propane prices returned to lower 17

rates, the customer load followed. OTP feels that after this one-year downturn in sales in 18

2015, Large Commercial sales will be generally consistent with the long-term trend. 19

Other Public Authority 20

Q. WHAT ARE THE MAIN VARIABLES IN OTHER PUBLIC AUTHORITY (OPA) 21

CLASS SALES? 22

A. North Dakota OPA class sales are the product of forecasted customer count and UPC for 23

the OPA class. Weather is the only significant forecasting variable used in the UPC model 24

for this class. Economic factors seem to have little or no impact on this class of customers, 25

so none are included in the customer count model as they are in most other models. 26

27

17 Case No. PU-17-

Draxten Direct


Q. HOW DO 2018 TEST YEAR OPA CLASS SALES COMPARE WITH RECENT 1

ACTUAL, WEATHER NORMALIZED OPA CLASS SALES? 2

A. OTP expects no sales growth in this class in 2018, mostly due to slightly declining UPC. 3

Figure 7 shows historical weather normalized sales and the 2018 Test Year forecasted sales 4

for the OPA Class. 5

6

Figure 7 7

8

9

Street Lighting 10

Q. WHAT ARE THE MAIN VARIABLES IN STREET LIGHTING CLASS SALES? 11

A. North Dakota Street Lighting class sales are the product of forecasted customer count and 12

UPC for the Street Lighting class. The primary economic variable used to forecast the 13

customer count for this class is total population. Unlike UPC models in other classes, 14

weather is not a factor used to predict street lighting sales; lighting needs do not change 15

due to hot or cold weather. 16

17

18 Case No. PU-17-

Draxten Direct


Q. HOW DO 2018 TEST YEAR STREET LIGHTING CLASS SALES COMPARE WITH 1

RECENT ACTUAL, WEATHER NORMALIZED STREET LIGHTING CLASS 2

SALES? 3

A. We forecast 2018 Test Year Street Lighting class sales to be 0.63 percent higher than actual 4

2016 weather normalized Street Lighting class sales. Again, this small rate of growth is 5

typical to what we have seen over the past 20 years, and an improvement over recent (since 6

2008) past. Figure 8 details the historical weather normalized sales and 2018 sales forecast 7

for this class. 8

9

Figure 8 10

11

12

Q. DO YOU ANTICIPATE STRUCTURAL CHANGES TO THIS CLASS IN THE 13

FUTURE? 14

A. Yes. We expect to see this load decrease in the future as LED lighting becomes more 15

common. OTP witness Mr. David G. Prazak discusses LED lighting in his Direct 16

Testimony.Unclassified 17

19 Case No. PU-17-

Draxten Direct


1

Q. WHAT ARE UNCLASSIFIED SALES? 2

A. This class is made up of company use accounts such as office and crew buildings, fleet 3

service buildings, and our General Office. It is mainly OTP’s own use of electricity. It 4

makes up less than 0.21 percent of our total kWh sales in North Dakota. Figure 9 shows 5

the historical weather normalized sales and 2018 Test Year sales forecast for this class. 6

7

Figure 9 8

9

10

CONCLUSION 11

Q. PLEASE SUMMARIZE YOUR DIRECT TESTIMONY. 12

A. In my Direct Testimony, I have discussed the forecast methodology used to develop the 13

2018 Test Year sales forecast. I have discussed the development of both the customer count 14

forecast and sales forecast and have supplied the data used in those calculations. 15

The forecast provided in this case is a reasonable estimate of the expected customer 16

counts and kWh sales for the 2018 Test Year, and OTP requests that they be adopted for 17

the purpose of determining the revenue requirements and final rates in this proceeding. 18

19

20 Case No. PU-17-

Draxten Direct


Q. DOES THIS CONCLUDE YOUR DIRECT TESTIMONY? 1

A. Yes, it does. 2

Case No. PU-17- Exhibit___(BHD-1), Schedule 1

Page 1 of 1

BRIAN H. DRAXTEN

EMPLOYMENT

2008 – PRESENT Otter Tail Power Company Fergus Falls, MN

Manager, Resource Planning

2002 – 2008 Otter Tail Power Company Fergus Falls, MN

Manager, Budget / Forecast


Manager, Market and Load Research


Administrator, Market Research


Rate Analyst, Rate Department

EDUCATION

• Bachelor of Arts Moorhead State University

Accounting

Business Finance

Schedule 3

Brian H. Draxten

PUBLIC DOCUMENT – NOT FOR PUBLIC (OR PRIVILEGED) DATA HAS BEEN EXCISED

Case No. PU-17-

Exhibit __(BHD-1), Schedule 3

INTRODUCTION .......................................................................................................................... 3

A. SALES FORECAST ................................................................................................................ 3

1. OVERVIEW ................................................................................................................. 3

a) Customer Model ................................................................................................... 5

b) Use-Per-Customer (UPC) Model ......................................................................... 5 2. MODEL INPUTS ......................................................................................................... 6

a) Sales and Customer Count Historical Data .......................................................... 6

b) Otter Tail Power Company’s Weather Data ......................................................... 8 c) Woods & Poole Economics, Inc. ........................................................................ 11

3. CALENDAR MONTH CALCULATION ................................................................. 11

4. BINARY VARIABLES ............................................................................................. 12

5. USE OF SALES FORECAST IN REVENUE FORECAST...................................... 12

B. REVENUE FORECAST ....................................................................................................... 14

1. OVERVIEW ............................................................................................................... 14

2. REVENUE FORECAST DESCRIPTION ................................................................. 15

3. INPUTS ...................................................................................................................... 19

a) Forecasted Sales ................................................................................................. 19

b) Composite Pricing .............................................................................................. 19 c) Demand Ratios ................................................................................................... 24

d) Ratcheted Demand Ratios .................................................................................. 25

e) Meter Count Forecast ......................................................................................... 26

f) Western Area Power Administration (WAPA) Credit ....................................... 27 g) Manually Forecasted Customer Inputs ............................................................... 28

4. REVENUE MODEL .................................................................................................. 28

a) Manually Forecasted Customers ........................................................................ 28 b) Main Calculation (non-manually forecasted) ..................................................... 29

5. EXCEL REVENUE CALCULATION MODEL ....................................................... 30

a) Model Overview ................................................................................................. 30 b) Input Section ....................................................................................................... 33 c) Calculations Section ........................................................................................... 34 d) Reports Section ................................................................................................... 34

CONCLUSION ............................................................................................................................. 35


Case No. PU-17-


Page 3 of 35

INTRODUCTION

Section A of this filing provides an overview of the process Otter Tail Power Company

(OTP) uses to develop its Sales Forecast. This overview includes the methodologies employed to

develop the forecasts for various classes of customers. Section B provides an overview of the

processes OTP uses to develop various pricing and billing determinants to develop its Revenue

Forecast. This section also provides an overview of a workbook model that combines the Sales

Forecast (section A) and the pricing and billing determinant information (section B). The

workbook generates the 2018 forecast test year Revenue Forecast.

A. SALES FORECAST

1. OVERVIEW

Section A is a description of the process used to develop OTP’s 2018 Test Year forecast for

energy sales. The forecast uses actual sales history through May 2017. The model used by OTP to

produce the forecast is MetrixND® (MetrixND). The files produced in MetrixND are easily

exportable to Microsoft Excel.

Along with this document are numerous workbooks that provide all regression models,

results, and data used to create those test year forecasts.

SALES MODEL DESCRIPTION

The following flowchart is the process OTP follows to create its Sales Forecast.


Case No. PU-17-


Page 4 of 35

Sales Forecast Flow Chart


Case No. PU-17-


Page 5 of 35

a) Customer Model

The Customer Models, designed in MextrixND, forecast monthly customer counts, by state

and by class, based on historical customer counts, economic indicators, and various binary

variables. The economic data is 2017 vintage and from Woods and Poole. The variables most

often used are Number of Households and Gross Regional Product.

The table in Attachment A1 Independent Variables.xlsx shows the variables that are

included in each customer model.

The following attachments contain all Customer Model and statistical information.

- Attachment A2 ND ResCust.xlsx

- Attachment A3 ND FarmCust.xlsx

- Attachment A4 ND SComCust.xlsx

- Attachment A5 ND LComCust.xlsx

- Attachment A6 ND OPACust.xlsx

- Attachment A7 ND SltCust.xlsx

- Attachment A8 ND UnclCust.xlsx

- Attachment A9 MN ResCust.xlsx

- Attachment A10 MN FarmCust.xlsx

- Attachment A11 MN SComCust.xlsx

- Attachment A12 MN LComCust.xlsx

- Attachment A13 MN OPACust.xlsx

- Attachment A14 MN SltCust.xlsx

- Attachment A15 MN UnclCust.xlsx

- Attachment A16 SD ResCust.xlsx

- Attachment A17 SD FarmCust.xlsx

- Attachment A18 SD SComCust.xlsx

- Attachment A19 SD LComCust.xlsx

- Attachment A20 SD OPACust.xlsx

- Attachment A21 SD SltCust.xlsx

- Attachment A22 SD UnclCust.xlsx

b) Use-Per-Customer (UPC) Model

The Use-Per-Customer (UPC) Models, also designed in MetrixND, forecast estimated

monthly UPC as a function of historical usage, weather conditions, and binary variables. Weather

conditions are represented using monthly Heating Degree Days (HDD) and Cooling Degree Days

(CDD) (definitions to follow), with a base of 65 degrees for cooling and 55 degrees for heating. In

some cases, binary variables are included in the equation to account for events in the historical

period.


Case No. PU-17-


Page 6 of 35

Generally, OTP does not use economic indicators in the UPC Models because economic

growth does not have a significant impact on usage. What little growth that may occur is offset by

the improvement in energy efficiencies of household products.

The table in Attachment A1 Independent Variables.xlsx shows the variables that are

included in each UPC model.

The following attachments contain all UPC Model and statistical information.

- Attachment A23 ND ResUPC.xlsx

- Attachment A24 ND FarmUPC.xlsx

- Attachment A25 ND SComUPC.xlsx

- Attachment A26 ND LComUPC.xlsx

- Attachment A27 ND OPAUPC.xlsx

- Attachment A28 ND SltUPC.xlsx

- Attachment A29 ND UnclUPC.xlsx

- Attachment A30 MN ResUPC.xlsx

- Attachment A31 MN FarmUPC.xlsx

- Attachment A32 MN SComUPC.xlsx

- Attachment A33 MN LComUPC.xlsx

- Attachment A34 MN OPAUPC.xlsx

- Attachment A35 MN SltUPC.xlsx

- Attachment A36 MN UnclUPC.xlsx

- Attachment A37 SD ResUPC.xlsx

- Attachment A38 SD FarmUPC.xlsx

- Attachment A39 SD SComUPC.xlsx

- Attachment A40 SD LComUPC.xlsx

- Attachment A41 SD OPAUPC.xlsx

- Attachment A42 SD SltUPC.xlsx

- Attachment A43 SD UnclUPC.xlsx

2. MODEL INPUTS

a) Sales and Customer Count Historical Data

Data: Attachment A44 Sales and Cust Counts.xlsx

Adjustments Made:

Monthly kWh data was graphed and values were checked for errors due to meters not being

billed, being billed twice in one month, etc. As described in detail below, any bill adjustments are

applied to the month in which the error occurred. In most cases the corrections are found and

downloaded during the next monthly update.


Case No. PU-17-


Page 7 of 35

Detailed Information:

Historical kWh data and the number of customers are read from SAS CIS/A data sets. The

SAS data sets are created from extracts of OTP’s CIS, which are downloaded the first day of each

month for the prior month. These datasets are also updated monthly for billing adjustments to

appropriately reflect actual usage and billing impacts in the month of the original bill. Any

changes made in OTP’s CIS are included in the CIS/A download, and the adjustments are made to

the month the error occurred (as opposed to the month the adjustment was made). For example, if

a customer has a bill adjustment made to their July bill, but the need for the adjustment was not

determined or made in the CIS until December, the adjustment in the CIS/A data set would adjust

the July bill, not the December bill.

From the CIS/A data set, the data is written into a totalized SAS dataset called

cisa_s_allyrs. This dataset is input to both the Sales Forecast and the Revenue Forecast. In the

Sales Forecast, the data is downloaded to the workbook referenced in this section (Attachment A44

Sales and Cust Counts.xlsx ). Each record in the dataset is assigned to one of seven classes used in

the forecast, by state (tab kWh). All customer counts and kWhs are divided into one of the

following worksheets (tabs), one for each state and class:

• MNRes

• NDRes

• SDRes

• MNFarm

• NDFarm

• SDFarm

• MNScom

• NDScom

• SDScom

• MNLcom

• NDLcom

• SDLcom

• MNOPA

• NDOPA

• SDOPA

• MNSlt

• NDSlt

• SDSlt

• MNUncl

• NDUncl

• SDUncl


Case No. PU-17-


Page 8 of 35

The variable UPC is created by dividing the monthly kWh by the monthly number of

customers.

b) Otter Tail Power Company’s Weather Data

Data: Attachment A45 HDD CDD.xlsx

Adjustments Made:

OTP graphs hourly monitoring station temperatures each month after downloading the

data. Any missing or obviously bad temperatures are corrected based on temperatures from other

nearby monitoring points or by judgment when necessary.


OTP uses 20 years of historical weather in its 2018 Sales Forecast. This weather was

collected from 1997-2017, from 14 monitoring stations throughout Minnesota, North Dakota and

South Dakota. OTP’s service territory is broken up into 14 geographic divisions. There is one

weather station in each of OTP’s 14 divisions, so that the weather across OTP’s entire service

territory is well represented.

The UPC forecast uses HDD and CDD as inputs – values calculated from dry bulb

temperatures in the weather data referenced above. The following is a definition of Heating and

Cooling Degree Days from The National Oceanic and Atmospheric Administration (NOAA)

(www.noaa.gov):

Degree days are the difference between the daily temperature mean and

55°F (heating) 65°F (cooling). If the temperature mean is above 65°F, we

subtract 65 from the mean and the result is Cooling Degree Days. If the

temperature mean is below 55°F, we subtract the mean from 55 and the

result is Heating Degree Days.

For each weather station, an average dry bulb temperature is calculated for each day. The

HDD are then calculated by subtracting the average daily temperature from 55 degrees (the base).

For example, if the average temperature for the day is 30 degrees, the HDD for that day is 25 (55-

30). CDD are calculated by subtracting 65 (the base) from the average daily temperature. For

example, if the average daily temperature is 70 degrees, the CDD for that day is 5 (70-65).

Table 1 lists each worksheet in Attachment A45 HDD CDD.xlsx, and its

description/purpose. A brief overview of the HDD and CDD calculation follows.


Case No. PU-17-


Page 9 of 35

Table 1

HDD CDD By Division

Worksheets Description

FergusFallsHourlyDB Hourly Dry Bulb/HDD/CDD for Fergus Falls Division

DevilsLakeHourlyDB Hourly Dry Bulb/HDD/CDD for Devils Lake Division

JamestownHourlyDB Hourly Dry Bulb/HDD/CDD for Jamestown Division

MorrisHourlyDB Hourly Dry Bulb/HDD/CDD for Morris Division

OakesHourlyDB Hourly Dry Bulb/HDD/CDD for Oakes Division

WahpetonHourlyDB Hourly Dry Bulb/HDD/CDD for Wahpeton Division

LangdonHourlyDB Hourly Dry Bulb/HDD/CDD for Langdon Division

RugbyHourlyDB Hourly Dry Bulb/HDD/CDD for Rugby Division

CanbyHourlyDB Hourly Dry Bulb/HDD/CDD for Canby Division

BemidjiHourlyDB Hourly Dry Bulb/HDD/CDD for Bemidji Division

CrookstonHourlyDB Hourly Dry Bulb/HDD/CDD for Crookston Division

HallockHourlyDB Hourly Dry Bulb/HDD/CDD for Hallock Division

GarrisonHourlyDB Hourly Dry Bulb/HDD/CDD for Garrison Division

MilbankHourlyDB Hourly Dry Bulb/HDD/CDD for Milbank Division

MNDailyAvgDB MN Daily Average HDD and CDD, weighted by station

NDDailyAvgDB ND Daily Average HDD and CDD, weighted by station

SDDailyAvgDB SD Daily Average HDD and CDD

MNMeterSchedule MN - Calculates Average HDD and CDD by individual billing cycle

NDMeterSchedule ND - Calculates Average HDD and CDD by individual billing cycle

SDMeterSchedule SD - Calculates Average HDD and CDD by individual billing cycle

MNMonthlyBilling MN - Combines individual cycles into billing month HDD and CDD

NDMonthlyBilling ND - Combines individual cycles into billing month HDD and CDD

SDMonthlyBilling SD - Combines individual cycles into billing month HDD and CDD

MNBillingNormal

MN - Combines 20 years of billing month HDD & CDD to create Normal

HDD & CDD

NDBillingNormal

ND - Combines 20 years of billing month HDD & CDD to create Normal

HDD & CDD

SDBillingNormal

SD - Combines 20 years of billing month HDD & CDD to create Normal

HDD & CDD

MNMonthlyCalendar MN - Combines calendar month HDD and CDD

NDMonthlyCalendar ND - Combines calendar month HDD and CDD

SDMonthlyCalendar SD - Combines calendar month HDD and CDD

MNCalendarNormal

MN - Combines 20 years of calendar month HDD & CDD to create normal

HDD & CDD

NDCalendarNormal

ND - Combines 20 years of calendar month HDD & CDD to create normal

HDD & CDD

SDCalendarNormal

SD - Combines 20 years of calendar month HDD & CDD to create normal

HDD & CDD


Case No. PU-17-


Page 10 of 35

To determine the HDD and CDD for North Dakota, the weather stations in or near North

Dakota are weighted by weather-sensitive sales and summed.

ND Daily Heating Degree Days=

[(Station 1 Sales/ND Weather-Sensitive Sales) * Station 1 HDD]+

[(Station 2 Sales/ND Weather-Sensitive Sales) * Station 2 HDD]+

…

[(Station 8 Sales/ND Weather-Sensitive Sales) * Station 8 HDD]

ND Daily Cooling Degree Days=

[(Station 1 Sales/ND Weather-Sensitive Sales) * Station 1 CDD]+

[(Station 2 Sales/ND Weather-Sensitive Sales) * Station 2 CDD]+

…

[(Station 8 Sales/ND Weather-Sensitive Sales) * Station 8 CDD]

This process is repeated for Minnesota and South Dakota.

OTP creates HDD and CDD based on billing month weather and calendar month weather.

The process is as follows:

1. Billing Month HDD and CDD:

Daily HDD and CDD are next added by billing cycle to determine the HDD and

CDD for each cycle and month. Once we have a HDD and CDD value for each

cycle and month, all the cycles are combined into one billing month, averaging

the cycle HDD and the cycle CDD. A HDD value and a CDD value for each

billing month have now been created.

Next, we calculate Normal Billing HDD and CDD. They are calculated by

averaging 20 years of monthly billing HDD and CDD.

2. Calendar Month HDD and CDD:

Daily HDD and CDD are added by calendar month to calculate the HDD and

CDD for each calendar month.

Normal Calendar HDD and CDD are next calculated. These values are used in

the Sales Forecast. They are calculated by averaging 20 years of monthly

Calendar HDD and CDD.

OTP’s Sales Forecast uses weather normalization principally to compare the Sales Forecast

to weather normalized historical data. HDD and CDD are used in all models with the exception of

Streetlighting. All of OTP’s other customer classes have some level of weather sensitivity.


Case No. PU-17-


Page 11 of 35

c) Woods & Poole Economics, Inc.

Data: Attachment A46 Woods and Poole Data.xlsx

Adjustments Made: None


In its 2018 Sales Forecast, OTP uses economic data from Woods & Poole Economics, Inc.

Woods & Poole’s database contains economic and demographic data through the year 2050. OTP

downloads this information by county to use in its Customer Model.

The Sales Forecast uses the following variables from Woods & Poole:

• Number of Households • Farm Earnings

• Gross Regional Product • Farm Employment

• Total Employment • Persons Per Household

• Total Population

OTP does not serve the entire load in the counties within its service territory. This is

especially problematic when OTP does not serve a large city that has a significant impact on the

economy of the county. Some examples are Fargo, Moorhead, Grand Forks and Minot. OTP does

not serve these larger cities, but it does serve small communities surrounding these larger ones. To

reflect this, OTP used econometric data only from counties where OTP serves at least 10 percent

of the population of the county. County population data is downloaded from www.census.gov.

The percentage of the population served by OTP in each county was determined by dividing the

sum of populations of towns served by OTP in each county by the population of the county.

Counties with a percentage of less than 10 percent were not included. Town populations were

obtained from an internal database of towns served. The data is then summed to the state level and

graphed as a reasonability check. Annual Woods & Poole data is converted from annual data to

monthly data by interpolating between annual values with a flat line.

As OTP serves three states with economic differences, using econometric models makes it

possible to utilize the different economic data for each state and determine whether particular

variables are drivers for each state.

3. CALENDAR MONTH CALCULATION

Because historical usage data is, in its purest form, in billing month format, OTP creates all

models using billing month data. After creating billing month sales models, these models are

adapted to calendar month. As weather generally only affects UPC, not the number of customers,


Case No. PU-17-


Page 12 of 35

the calendar month conversion is only applied to the UPC Model. To create the calendar month

UPC forecast, the calendar month HDD and CDD (from Attachment A45 HDD CDD By

Division.xlsx) are substituted for the billing month HDD and CDD resulting in a calendar month

UPC forecast.

4. BINARY VARIABLES

All models that make up the Sales Forecast utilize binary variables. Monthly binary

variables that account for seasonal differences are the most commonly used variables. Annual

binary variables are used to account for the deviations in growth or consumption that are not

expected in the test year. For example, the Residential Customer Model uses annual binary

variables to account for the gradual change in growth rates that occurred before and after the 2008

recession. Other binary variables are utilized as necessary to improve the fit of the model and

statistical significance of the economic and weather variables.

5. USE OF SALES FORECAST IN REVENUE FORECAST

As noted earlier, OTP develops Sales Forecasts for each class within each jurisdiction.

However, to develop an accurate Revenue Forecast, the Sales Forecast within each class needs to

be allocated to a more detailed Rate Group level1. In this manner, OTP can apply appropriate

billing determinants to compute the forecasted revenues.

To allocate the Sales Forecast to each Rate Group, each meter from the CIS/A data set is

first placed into one of 41 Rate Groups. These Rate Groups are created using a Revenue

Class/Rate Code combination. For example, there is a Rate Group for Residential Water Heating,

another for Small Commercial Water Heating, etc. (see full description of these Rate Groups in

Revenue Forecast Model description – Attachment B1 Rate Group Mapping.xlsx).

In general, CIS/A data for each state/year/month/revenue class/rate code is combined into

Rate Group definitions, by state/year/month. The percentage of each Rate Group in the class is

calculated for each state/year/month, and a two-year average of these percentages is computed.

See Attachment A47 Res Sales to Rate Group.xlsx as an example of this process. The

worksheets contained in this spreadsheet are found in Table 2:

1 OTP has 41 Rate Groups that are created using a Revenue Class/Rate Code (level pricing is applied at)

combination. See Attachment B1 Rate Group Mapping.xlsx for a complete listing of the Rate Groups and how the

Classes, Rate Groups, Revenue Class and Rate Codes all fit together.


Case No. PU-17-


Page 13 of 35

Table 2

Residential Sales to Rate Group

Worksheet Description

MN_Res

ND_Res

SD_Res

Contains Sales by Rate Group,

by State,

Year and Month

MN_Res_Ratio_2yr_Avg

ND_Res_Ratio_2yr_Avg

SD_Res_Ratio_2yr_Avg

Calculates Percentage of Sales

by Rate Group,

by State, Year and Month

This same process is followed for each of the following classes: Residential, Farm, Small

Commercial, and Large Commercial. This process is not necessary for the Pipelines, Malting,

OPA, Streetlighting, and Unclassified Classes as the class and the Rate Group are the same.

These monthly percentages are then applied to the Sales Forecast to allocate sales to the

Rate Group level. The Sales Forecast at Rate Group level is a key input into the Revenue Forecast

Model for pricing.

The following attachments create the Sales Forecast by Rate Group:

- Attachment A47 Res Sales to Rate Group.xlsx

- Attachment A48 Farm Sales to Rate Group.xlsx

- Attachment A49 SCom Sales to Rate Group.xlsx

- Attachment A50 LCom Sales to Rate Group.xlsx

The output of the Revenue Forecast is found in Attachment A51 Sales Forecast to Revenue

Forecast.xlsx. This workbook contains five worksheets. The first, titled Input to Revenue

Forecast, contains the majority of the data that goes into the Revenue Model. The next two

worksheets, Manual Pipe & Malt and Manual – Other contain the manually forecasted customers.

The data within these worksheets goes into the Revenue Model as well, to be priced separately

(Attachment B11 Revenue Model.xlsx). The fourth worksheet, Unclassified, contains data that is

not normally priced, as it includes mostly company use sales. Finally, the last worksheet, Total

contains all the previous worksheets’ sales totaled, to present the entire Sales Forecast as a whole.


Case No. PU-17-


Page 14 of 35

B. REVENUE FORECAST

1. OVERVIEW

OTP produces its Revenue Model in the financial modeling software program, Utilities

International, which easily exports to Microsoft Excel. OTP has chosen to provide this data in

Excel format allowing transparency to its model.

Within this section, section B, is a description of the process used to develop OTP’s

Revenue Forecast. The forecast uses up to 5 years of historical customer data from CIS/A and a

workbook containing OTP’s current rate code prices. There are also four SAS programs and

numerous Excel workbooks attached that provide the inputs and a transparent view of our Revenue

Model.

OTP developed its Revenue Forecast2 by applying various composite pricing to the

applicable billing determinants3 derived from the Sales Forecast. OTP uses actual historical billing

determinants to develop composite pricing, demand ratios, ratcheted demand4 ratios, and

forecasted meter counts. The demand and ratcheted demand ratios are multiplied by the Sales

Forecast to acquire the demand and ratcheted demand for each Rate Group5. Once all the billing

determinants at the Rate Group level are computed, they are multiplied by the corresponding

composite pricing to compute revenues for each Rate Group. Subsequently, OTP rolls up the Rate

2 OTP forecasts retail revenue excluding small power producers. 3 Billing Determinants are units needed for billing, OTP’s billing determinants used in the Revenue Forecast include

sales (kWh), demand (kW), ratcheted demand (ratcheted kW), and meter count. 4 Ratcheted demand is the maximum demand over the last 12 months, primarily used for calculating the facilities

demand. 5 OTP has 41 Rate Groups that are created using a Revenue Class/Rate Code (level pricing is applied at)

combination. See Attachment B1 Rate Group Mapping.xlsx for a complete listing of the Rate Groups and how the

Classes, Rate Groups, Revenue Class and Rate Codes all fit together.


Case No. PU-17-


Page 15 of 35

Group level revenues to their respective Cost of Service Class level revenues by state. The Cost of

Service Classes are as follows.

• Residential

• Farms

• General Service

• Large General Service

• Irrigation

• Outdoor Lighting

• Other Public Authority (OPA)

• Controlled Service Water Heating

• Controlled Service Interruptible

• Controlled Services Deferred

Section B of this document will cover the information needed to develop the North Dakota

Revenue Forecast.

2. REVENUE FORECAST DESCRIPTION

The following flowcharts describe the process OTP follows to create its Revenue Forecast.

The remainder of section B of this document is laid out in three main sections; Inputs, Revenue

Model, and Excel Revenue Calculation Model.

Inputs: This section explains how each input to the Revenue Model is calculated. The

initial sections of each input can be read alone or you can use the SAS sections as a

companion to the SAS program, which are attached in pdf format.

Revenue Model: This section is a description of OTP’s Revenue Model and is the link

between the inputs and the final Revenue Forecast results.

Excel Revenue Calculation Model: This section is a detailed description of the Excel

document that calculates the Revenue Forecast. It details the items in each worksheet and

how the worksheets are linked.


Case No. PU-17-


Page 16 of 35

[PROTECTED DATA BEGINS…

…PROTECTED DATA ENDS]


Case No. PU-17-


Page 19 of 35




Case No. PU-17-


Page 20 of 35




Case No. PU-17-


Page 21 of 35




Case No. PU-17-


Page 22 of 35




Case No. PU-17-


Page 23 of 35


…PROTECTED DATA BEGINS]


Case No. PU-17-


Page 24 of 35




Case No. PU-17-


Page 25 of 35




Case No. PU-17-


Page 26 of 35




Case No. PU-17-


Page 27 of 35




Case No. PU-17-


Page 28 of 35




Case No. PU-17-


Page 29 of 35




Case No. PU-17-


Page 30 of 35




Case No. PU-17-


Page 31 of 35




Case No. PU-17-


Page 32 of 35




Case No. PU-17-


Page 33 of 35


…PROTECTED DATA BEGINS]


Case No. PU-17-


Page 34 of 35

34


...PROTECTED DATA ENDS]

NOT PUBLIC DOCUMENT – NOT FOR PUBLIC DISCLOSURE

Case No. PU-17-


Page 35 of 35

35

CONCLUSION

The forecast provided in this filing is a reasonable estimate of the expected sales and

projected revenues under existing rates for the 2018 Test Year. OTP recommends that they be

adopted for the purpose of determining the revenue requirements and final rates in this proceeding.

1/5

Volume 2C

Direct Testimony and Supporting Schedules:

Peter E. Wasberg

Before the North Dakota Public Service Commission

State of North Dakota


For Authority to Increase Rates for Electric Utility

Service in North Dakota

PU-17-

Exhibit___

EMPLOYEE COMPENSATION

Direct Testimony and Schedules of

PETER E. WASBERG

PUBLIC DOCUMENT – NOT PUBLIC (OR PRIVILEGED)

DATA HAS BEEN EXCISED

November 2, 2017

TABLE OF CONTENTS

I. INTRODUCTION AND QUALIFICATIONS................................................................... 1

II. PURPOSE AND OVERVIEW OF DIRECT TESTIMONY ............................................. 1

III. OVERVIEW OF OTP’S WORKFORCE AND COMPENSATION PLAN ...................... 2

IV. OTP COMPENSATION LEVELS AND THE COMPETITIVE MARKET ..................... 7

V. EMPLOYEE BENEFITS .................................................................................................. 10

A. Medical and Dental Plans...................................................................................... 10

B. Retirement and Pension Benefits .......................................................................... 12

VI. 2018 EMPLOYEE COMPENSATION COSTS ............................................................... 17

A. Wages and Salaries ............................................................................................... 17

B. Incentive Compensation ........................................................................................ 18

C. Medical and Dental ............................................................................................... 19

D. Pension and Benefits ............................................................................................. 20

VII. CONCLUSION ................................................................................................................. 23

ATTACHED SCHEDULES

Schedule 1 – Qualifications, Duties and Responsibilities of Peter E. Wasberg

Schedule 2 – 2015 Mercer Compensation Benchmark Study NOT PUBLIC

Schedule 3 – 2015 Mercer Executive Compensation Review NOT PUBLIC

Schedule 4 - Estimates of HDHP Savings 2012-2016

Schedule 5a – Summary Description of OTP 401k Plan

Schedule 5b – Summary Description of OTP Pension Plans

Schedule 5c – Summary Description of Retirement Benefits Eligibility

Schedule 6 – 2017 Willis Towers Watson Energy Industry BenVal Study NOT PUBLIC

Schedule 7a – 2014-2016 3-year payouts on the OTP KPA Plan

Schedule 7b – 5-year payouts of the OTP Management Incentive Plan NOT PUBLIC

Schedule 8 – Mercer Estimate for 2018

Schedule 9 – Mercer 2018 Medical Renewal Results

Schedule 10a – 2016 Pension Accounting Expense Report – Mercer NOT PUBLIC

Schedule 10b – 2017 Pension DAMP – Mercer NOT PUBLIC

Schedule 10c – 2017 Post-retirement Medical and LTD Medical DAMP – Mercer NOT PUBLIC

1 PU-17-

Wasberg Direct


INTRODUCTION AND QUALIFICATIONS 1

Q. PLEASE STATE YOUR NAME AND OCCUPATION. 2

A. My name is Peter E. Wasberg. I am employed by Otter Tail Power Company (OTP) as 3

Director, Human Resources and Safety. 4

5

Q. PLEASE SUMMARIZE YOUR QUALIFICATIONS AND EXPERIENCE. 6

A. I graduated from Concordia College, Moorhead, MN with degrees in business 7

administration and psychology and have worked at OTP since 1991. Prior to moving into 8

the Human Resources Department in 2004, I primarily worked within the customer 9

service/operations area of OTP. I have been in my current role since January 2008. A copy 10

of my resume is included as Exhibit___(PEW-1), Schedule 1. 11

PURPOSE AND OVERVIEW OF DIRECT TESTIMONY 12

Q. WHAT IS THE PURPOSE OF YOUR DIRECT TESTIMONY? 13

A. My Direct Testimony describes matters related to employee compensation and benefits for 14

OTP. 15

16

Q. PLEASE PROVIDE A BRIEF OVERVIEW OF YOUR DIRECT TESTIMONY. 17

A. First, I will briefly describe OTP’s current compensation plan, including its two annual 18

incentive plans (collectively the “OTP Annual Incentive Plan”). Second, I will discuss 19

how the compensation levels provided by OTP compare to the general market. Third, I 20

will generally describe the benefits provided by OTP. Fourth, I will summarize OTP’s 21

2018 compensation and benefits costs. 22

23

Q. HOW IS YOUR DIRECT TESTIMONY ORGANIZED? 24

A. In Section III, I describe OTP’s workforce and compensation plan. In Section IV, I discuss 25

OTP compensation levels and the competitive market. In Section V, I provide a description 26

of OTP’s employee benefits. In Section VI, I describe OTP’s 2018 employee 27

compensation costs, and in Section VII, I provide my conclusion. 28

2 PU-17-

Wasberg Direct


OVERVIEW OF OTP’S WORKFORCE AND COMPENSATION 1

PLAN 2

Q. PLEASE BRIEFLY DESCRIBE OTP’S WORK FORCE. 3

A. In 2018, OTP expects to have an average of 777 full-time equivalent employees, including 4

approximately 392 union employees and 385 non-union employees (these numbers are not 5

adjusted for employees of jointly owned plants). The non-union full-time equivalent 6

employees include: (1) the OTP President and 10 other executives; (2) 250 exempt 7

employees; and (3) 124 non-exempt employees. OTP employees provide a wide range of 8

services required to provide electric utility services to our approximately 131,200 9

customers located in North Dakota, Minnesota, and South Dakota. OTP’s workforce 10

supports a substantially larger net plant in service than it did in 2007. OTP witness Mr. 11

Bruce G. Gerhardson discusses the magnitude of infrastructure additions. 12

13

Q. WHAT ARE OTP’S COMPENSATION GOALS FOR ITS WORKFORCE? 14

A. OTP’s compensation goals are to attract, retain and engage employees. OTP’s employees 15

are essential to achieve OTP’s mission, which is to produce and deliver electricity as 16

reliably, economically and environmentally responsibly as possible to the balanced benefit 17

of customers, shareholders, and employees, and to improve the quality of life in the areas 18

in which we do business. OTP focuses on maintaining a compensation program that 19

provides a competitive, performance-based pay system that helps us attract and retain a 20

quality workforce that provides our customers with safe, reliable, and economical service. 21

22

Q. IS OTP FACING CHANGES THAT INCREASE THE IMPORTANCE OF THESE 23

GOALS? 24

A. Yes. We expect to have over 35 percent (over 275) of our employees retire during the next 25

ten years. We also have the potential for another 150 - 200 employees to leave employment 26

during this same ten-year time period based on historic non-retirement attrition. Based on 27

exit interviews conducted with approximately 50 percent of the employees who have left 28

OTP for reasons other than retirement since 2011, we find that some of these non-29

retirement attritions go to other utilities. The reasons cited vary and include increased 30

3 PU-17-

Wasberg Direct


compensation and/or benefits, a desire for a different geographic setting, and in some cases 1

career advancement. As a relatively small utility serving the rural areas of North Dakota, 2

Minnesota and South Dakota, OTP faces challenges in attracting and retaining skilled 3

workers, especially mobile workers who are in the earlier stages of their careers. 4

5

Q. WHAT ARE THE MAIN COMPONENTS OF OTP’S COMPENSATION PLAN? 6

A. The OTP compensation plan consists of: (1) a combination of base salaries or base wages, 7

plus annual incentive compensation for non-union employees; (2) standard employee 8

benefit plans, including health and dental plans, a 401(k) retirement savings plan with an 9

employer match, and an employee stock ownership plan; (3) defined benefit pensions and 10

post-retirement health benefits for employees whose employment began before OTP closed 11

participation in these plans in 2006 (as I describe later in my Direct Testimony); and (4) 12

defined contributions to a 401(k) plan for other employees. Some key OTP management 13

employees are also eligible for long-term incentives. I discuss the cash compensation for 14

OTP employees in this Section III of my Direct Testimony and I will discuss benefits for 15

OTP employees in Section V of my Direct Testimony. 16

17

Q. PLEASE BRIEFLY DESCRIBE THE BASE SALARY AND WAGE COMPONENT OF 18

OTP’S COMPENSATION PLAN. 19

A. OTP has a combination of hourly and salaried employees. Approximately 50 percent of 20

OTP’s employees are represented by unions and 16 percent are non-union employees who 21

receive hourly compensation. OTP has approximately 32 percent salaried (or exempt) 22

employees who are not part of management. All employees receive either a base wage or 23

base salary as part of the employee’s cash compensation. 24

25

Q. PLEASE BRIEFLY DESCRIBE THE OTP ANNUAL INCENTIVE PLAN. 26

A. The OTP Annual Incentive Plan is provided to all regular employees who are not 27

represented by a union and who work at the rate of at least 1,000 hours per year. The OTP 28

Annual Incentive Plan consists of two separate plans for: (1) non-union employees (the 29

4 PU-17-

Wasberg Direct


“OTP Key Performance Award Plan” or “OTP KPA Plan”); and (2) key management 1

employees (the “OTP Management Plan”). 2

3

Q. PLEASE BRIEFLY DESCRIBE THE OTP KPA PLAN. 4

A. The OTP KPA Plan covers approximately 380 OTP non-union employees. The maximum 5

payout level under the OTP KPA Plan is six percent an individual’s employee’s base salary. 6

The OTP KPA Plan is based on: (1) three operating criteria (safety, customer satisfaction, 7

and equivalent plant availability); and (2) one financial criterion relating to the control of 8

operation and maintenance (O&M) costs. Each of these four criteria has a weighting that 9

together comprise the six percent maximum payout. 10

11

Q. PLEASE EXPLAIN THE WEIGHTING OF THE FOUR INDIVIDUAL CRITERIA. 12

A. Of the six percent maximum payout: (1), customer satisfaction has a weighting of 2 13

percent; (2) equivalent plant availability has a weighting of 1 percent; (3) safety has two 14

targets, each worth 0.5 percent; and (4) the O&M cost control criteria has a weighting of 2 15

percent. Payouts under the three operating criteria are not financially tied to the O&M cost 16

control criterion. 17

18

Q. IS OTP PROPOSING A LIMIT ON THE LEVEL OF COST RECOVERY FOR THE KPA 19

PLAN? 20

A. Yes. OTP is proposing to limit the recovery of incentive compensation under the KPA 21

Plan to the average payout level over the past three years, which is approximately 3.5 22

percent, as I will explain in more detail later in my Direct Testimony. 23

24

Q. PLEASE BRIEFLY DESCRIBE THE OTP MANAGEMENT PLAN. 25

A. The OTP Management Plan covers 19 OTP management employees (not including the 26

OTP President who has a separate plan). The OTP Management Plan includes: (1) safety; 27

(2) a number of individual criteria that vary by the employee’s job and responsibilities; and 28

(3) two overall financial criteria relating to OTP. The safety and individual criteria have a 29

cumulative weighting of 50 percent. The financial criteria have a weighting of 50 percent. 30

5 PU-17-

Wasberg Direct


Q. IS OTP PROPOSING A LIMIT ON THE LEVEL OF COST RECOVERY FOR THE OTP 1

MANAGEMENT PLAN? 2

A. Yes. The settlement in OTP’s last general rate case (PU-08-862) included a limitation on 3

recovery of costs of the OTP Management Plan based on 25 percent of base salary. 4

Consistent with that approach, OTP is proposing to limit the recovery of costs of the 5

Management Plan in this case capped at 25 percent of an individual’s base salary. 6

7

Q. PLEASE DESCRIBE THE INCENTIVE PLAN FOR THE OTP PRESIDENT. 8

A. The OTP President is under the Otter Tail Corporation Executive Annual Incentive Plan. 9

Under that plan, the criteria and target incentives are determined by the Compensation 10

Committee of the Otter Tail Corporation Board of Directors. Criteria under the current 11

plan include: (1) OTP net income; (2) OTP return on equity; (3) Otter Tail Corporation 12

earnings per share; (4) individual performance; and (5) OTP OSHA safety case rate. 13

14

Q. IS OTP PROPOSING A LIMIT ON THE LEVEL OF COST RECOVERY FOR THE OTP 15

PRESIDENT’S INCENTIVE PLAN? 16

A. Yes. The OTP President’s target incentive is 50 percent of base pay, but OTP is also 17

proposing to limit the level of incentive compensation recovered in rates to 25 percent of 18

the OTP President’s individual base salary. Again, this is consistent with the treatment in 19

OTP’s last North Dakota general rate case. 20

21

Q. DO THE FINANCIAL CRITERIA OF OTP’S ANNUAL INCENTIVE PLANS 22

PROVIDE BENEFITS TO CUSTOMERS? 23

A. Yes. The financial performance components of the OTP Annual Incentive Plan benefit 24

customers because financial performance depends on the prudent management of costs, 25

which allows electric utility service to be provided at reasonable prices. Combining 26

financial criteria, along with performance measures, ensures that we maintain the balance 27

between reliable service and reasonable prices. We believe in this balance and that the 28

financial criteria provide a benefit to our customers. 29

30

6 PU-17-

Wasberg Direct


Q. HOW DOES THE OTP ANNUAL INCENTIVE PLAN FIT INTO OTP’S TOTAL 1

COMPENSATION PLAN AND COMPENSATION GOALS? 2

A. The OTP Annual Incentive Plan is an important part of our total compensation plan. 3

Without our annual incentive plan, OTP’s total cash compensation would be significantly 4

below the market median of total cash compensation, and OTP would be impeded in its 5

ability to attract and retain essential employees. The annual cash incentive compensation 6

that is part of the OTP Annual Incentive Plan also encourages increased productivity, and 7

enables OTP to reward employees for providing quality service to our customers. Both 8

customers and employees benefit from setting incentives that tie directly to our customers’ 9

needs. 10

The OTP Annual Incentive Plan also helps us to maintain workforce continuity. 11

Reliability of our electrical plants and electrical systems is enhanced when we have 12

continuity in our workforce. The technical knowledge needed, and the years that it actually 13

takes to acquire the specialized skills for our system, are paramount to our ability to reliably 14

and efficiently provide energy to our customers. Our customers also see the advantages of 15

reduced costs associated with the lower recruitment and training requirements associated 16

with a stable workforce. With an expected retirement rate of approximately 35 percent in 17

the next ten years, we see maintaining a competitive compensation package as a critical 18

requirement for our customers and company. 19

20

Q. WHAT IS THE ALTERNATIVE TO THE USE OF ANNUAL INCENTIVE 21

COMPENSATION IN OTP’S COMPENSATION PACKAGE? 22

A. Without annual incentive compensation, the only way to maintain a competitive cash 23

compensation package would be to increase base salaries, which would increase other costs 24

and substantially reduce both flexibility and incentives for performance. 25

26

Q. DOES OTP PROVIDE LONG TERM INCENTIVES. 27

A. Yes. Qualifying management employees may receive long-term incentives in the form of 28

grants of restricted stock units. In 2017, 20 OTP employees (including the OTP President) 29

received grants of restricted stock. Qualifying employees are awarded restricted stock units 30

7 PU-17-

Wasberg Direct


based on salary, job level, and the price of the stock at the date of grant. These long-term 1

incentives are considered along with other components of compensation when we review 2

compensation levels. The purposes of OTP’s long-term incentives are to: (1) link the long-3

term success of OTP to qualifying employee compensation; (2) encourage the retention of 4

management over the long-term; and (3) provide the opportunity to earn competitive total 5

compensation. OTP has included the costs of long-term incentives for OTP employees in 6

the 2018 Test Year, also subject to a 25 percent cap as described above for the Management 7

Incentive Plan. 8

OTP COMPENSATION LEVELS AND THE COMPETITIVE 9

MARKET 10

Q. DOES OTP USE INFORMATION REGARDING COMPENSATION AT OTHER 11

UTILITIES AND BUSINESSES IN CONNECTION WITH ITS COMPENSATION 12

DECISIONS? 13

A. Yes. OTP routinely compares its compensation levels to those of other utilities and some 14

non-utilities, using a number of surveys and information sources, including Towers Watson 15

(Towers), Mercer, and Aon Hewitt (Hewitt). Every three to four years, OTP participates 16

in a benchmark study for non-executive employees, the most recent of which was 17

conducted by Mercer and completed in 2015. We periodically participate in studies of 18

executive compensation, which I discuss in Section VI of my Direct Testimony. We also 19

routinely review collective bargaining agreements from other utilities throughout the 20

region. 21

22

Q. PLEASE EXPLAIN HOW OTP USES BENCHMARK STUDIES. 23

A. OTP uses benchmark studies as the framework for formulating its compensation programs. 24

Salary surveys are reviewed and analyzed to find positions that correspond with the 25

essential job duties, skills, and functions of OTP’s positions. OTP strives to set 26

compensation at or near the median of the survey data. While the market-based 27

compensation for a position is based on the median, it is not limited to the single data point 28

of the median. Rather, the relevant market for a position includes a range above and below 29

8 PU-17-

Wasberg Direct


the median. The compensation for OTP non-union employees is ultimately determined by 1

a combination of market data and the employee’s responsibilities, performance, and 2

experience. 3

4

Q. PLEASE FURTHER EXPLAIN THE 2015 MERCER BENCHMARK STUDY. 5

A. The 2015 Mercer Benchmark Study covered non-executive employees and included 6

compensation information for a broad sample of positions. The 2015 Mercer Benchmark 7

Study examined base salaries and annual incentives. OTP compensation levels were 8

compared using a combination of general industry, energy/utility industry, and North 9

Central regional data to reflect the labor markets in which OTP competes for employees. 10

A copy of a portion of the 2015 Mercer Compensation Benchmark Study is provided as 11

NOT PUBLIC Exhibit___(PEW-1), Schedule 2. 12

13

Q. WHAT DID THE 2015 MERCER BENCHMARK STUDY INDICATE? 14

A. The 2015 Mercer Benchmark Study showed that: 15

1. OTP base salaries overall were six percent above the base salary medians; 16

2. OTP incentive compensation structure was seven percent below the market median; 17

and 18

3. OTP total cash compensation levels (base salary plus annual incentive) overall were 19

one percent above the market medians. 20

Of the 100 non-union positions that were reviewed, some fell below the market median 21

and some were above the market median. 22

23

Q. WHAT DID YOU CONCLUDE FROM THESE RESULTS? 24

A. Taken together, the 2015 Mercer Benchmark Study confirms that OTP is neither a market 25

leader nor significantly below market in its non-executive compensation levels. The 2015 26

Mercer Benchmark Study showed that, overall, the OTP compensation plan is competitive 27

with the market and reasonable, especially considering the average tenure of our employee 28

group. It also indicated that our compensation structure included in our rate request for 29

non-executive employees is fair and reasonable. 30

9 PU-17-

Wasberg Direct


Q. HAS A RECENT STUDY OF EXECUTIVE COMPENSATION BEEN CONDUCTED? 1

A. Yes. In 2015 Mercer conducted an Executive Compensation Review for OTP which 2

evaluated a number of compensation components, including: (1) Base Salary; (2) Total 3

Cash Compensation; and (3) Total Direct Compensation (including Base Salary, Annual 4

Incentive Compensation, and Long-Term Incentive). The 2015 Mercer Executive 5

Compensation Review utilized multiple data points including: (1) Mercer 2015 US 6

Executive Remuneration Suite; (2) the Towers Watson 2015 CDB Energy Services 7

Executive Compensation Survey; (3) the Mercer 2015 Total Compensation for the Energy 8

Sector Survey; (4) the Mercer 2015 US Short Term Incentive Survey; and (5) the Mercer 9

2015 US Long Term Incentive Survey. Mercer utilized survey data from the Energy and 10

Utility industry sectors where possible and applied discounts and premiums when the 11

survey data reflected companies that were significantly larger or smaller than OTP. A copy 12

of the 2015 Mercer Executive Compensation Review is provided as NOT PUBLIC 13

Exhibit___(PEW-1), Schedule 3. 14

15

Q. WHAT DID THE 2015 MERCER EXECUTIVE COMPENSATION REVIEW SHOW? 16

A. The Study showed the following: 17

1. OTP executive Base Salaries were four percent below the market median. 18

2. OTP’s actual executive Total Cash Compensation was five percent below the 19

market median. 20

3. OTP executive Total Direct Compensation was 12 percent below the market 21

median. 22

23

Q. WHAT DID YOU CONCLUDE FROM THESE RESULTS? 24

A. I concluded that OTP’s executive Total Cash Compensation and executive Total Direct 25

Compensation are generally below the market median and that all the components of OTP’s 26

executive Total Direct Compensation are reasonable and necessary. 27

10 PU-17-

Wasberg Direct


EMPLOYEE BENEFITS 1

Q. PLEASE SUMMARIZE THE BENEFITS PROVIDED BY OTP. 2

A. As I will explain in more detail, OTP provides: (1) employee medical/dental benefits; (2) 3

retirement benefits, including a defined benefit pension plan and a defined contribution 4

401(k) plan; and (3) other post retirement employee benefits. 5

6

Q. ARE THE BENEFITS PROVIDED BY OTP A REASONABLE PART OF EMPLOYEE 7

COMPENSATION? 8

A. Yes. OTP’s benefits contribute to attracting and retaining its skilled workforce. As 9

discussed above, OTP and its customers are directly benefited by having a stable, long-10

term workforce. In an industry where multiple years of training can be required for 11

employees to work independently, it is more fiscally prudent to have a workforce with a 12

low turnover rate. As we retire over 35 percent of our workforce within the next ten years, 13

it will be increasingly important for OTP to attract and retain a workforce that will continue 14

to provide electricity to our customers in a safe, reliable, and efficient manner. 15

A. Medical and Dental Plans 16

Q. PLEASE DESCRIBE OTP’S MEDICAL AND DENTAL BENEFITS. 17

A. OTP provides employees with the option to elect group medical and dental insurance 18

benefits. At present, we have three high deductible health plans (HDHP) for our non-union 19

employees and a separate single HDHP option for our union employees. A comparison of 20

the health care plans is provided in Table 1 below. Premiums for the medical and dental 21

plans vary based on the plan(s) chosen, whether dependents are covered, and differences 22

between current non-union and union plans that will be further described. We are currently 23

in negotiations with our union employees and will continue to work with them to help 24

maintain reasonable costs for OTP while providing quality health care plan options for our 25

employees. 26

27

11 PU-17-

Wasberg Direct


Table 1 1

2017 Medical Plans

Nonunion Employees Union Employees

Premier 80 Advantage 70 Basic 100 Blue Saver 100

Annual

Deductible

$1,500 – Employee

$3,000 – Other Tiers

$3,000 – Employee


$6,550 – Employee


$2,600 – Employee


Co-Insurance 80% after deductible 70% after deductible 100% after deductible NA

Annual Out of

Pocket Max

$3,000 – Employee


$6,000 – Employee

$9,000 – Other Tiers*

$6,550 – Employee


$2,600 – Employee


* While the OOPM is not embedded, ACA rules limit the individual annual liability to $7,150. This amount is subject to change annually.

2

Q. HAS OTP TAKEN STEPS TO CONTROL THE COST OF MEDICAL AND DENTAL 3

BENEFITS? 4

A. Yes. In 2012, OTP moved to a HDHP for all employees. The move to the HDHP, 5

sometimes called a consumer-driven health plan, has enabled our employees to be better 6

health care consumers. Exhibit___(PEW-1), Schedule 4, details multiple scenarios, 7

comparing our previous health care plan to the HDHP. Moving to the new HDHP resulted 8

in estimated savings of $4 million to over $12 million over the 2012-2016 period. In 2017, 9

we took a further step with our non-union employees and replaced the existing HDHP with 10

three new HDHP options that included different deductibles, premiums, and these plan 11

options also introduced coinsurance. Coinsurance keeps employees in the decision-making 12

process longer and that generally provides for increased consumerism. Although health 13

care costs continue to be a challenge for companies, we believe our actions to this point 14

have enabled us to be better consumers, and we expect that our emphasis on preventive 15

health care and wellness will provide long-term benefits for our employees and for 16

Company health care costs. 17

18

Q. HAS OTP TAKEN OTHER STEPS TO CONTROL THE COST OF MEDICAL AND 19

DENTAL BENEFITS? 20

A. Yes. In 2017, OTP also increased the cost-sharing ratio of employee to employer share of 21

the health care premium. This ratio is now at 15 percent employee/85 percent employer of 22

12 PU-17-

Wasberg Direct


the total premium cost in the Advantage 80 Plan. Prior to this, the ratio was 4 percent 1

employee/96 percent employer of the total premium cost. The employer share/employee 2

share for gross health costs, which includes total spend for both OTP and employees, 3

moved from approximately 80 percent employer/20 percent employee to 70 percent 4

employer/30 percent employee on all three of the new HDHPs. Our union employees 5

remain at a 4.45 percent employee/95.55 percent employer of the total premium cost with 6

a gross health cost ratio of approximately 80 percent employer/20 percent employee. With 7

our changes for non-union employees in 2017, the Company also added a monthly tobacco 8

surcharge of $40 and we eliminated health care eligibility for spouses who were able to 9

obtain health care insurance from their own employer. As previously noted, we are 10

currently in negotiations with our unions and will continue to look for options that will 11

help to keep our health care spend on a reasonable trajectory. Some of these changes are 12

difficult on our employees and we will continue working on providing quality health care 13

options for our employees while being responsible with the overall costs associated with 14

the benefit. 15

B. Retirement and Pension Benefits 16

Q. PLEASE SUMMARIZE OTP’S RETIREMENT SAVINGS AND PENSION BENEFITS 17

A. OTP provides 401(k) defined contribution retirement plans (collectively the “401(k) Plan”) 18

for all employees and defined benefit pension plans (collectively the “Pension Plan”) for 19

certain employees depending on the date they were hired. OTP also provides a 401(k)-20

matching plan. 21

22

Q. PLEASE DESCRIBE OTP’S 401K PLAN. 23

A. For employees who are not eligible to participate in OTP’s Pension Plan, OTP has 401(k) 24

Plans that include three different defined contribution amounts made by OTP: (1) a 4 25

percent contribution for non-union employees; (2) a 7 percent contribution for our Coyote 26

Station union employees; and (3) a 5 percent contribution for our other union employees. 27

A summary description of OTP’s 401(k) Plan is included in in Exhibit___(PEW-1), 28

Schedule 5a. 29

30

13 PU-17-

Wasberg Direct


Q. DOES OTP PROVIDE MATCHING FOR EMPLOYEE CONTRIBUTIONS TO THE 1

401(k) PLAN? 2

A. Yes. For the majority of OTP employees, OTP will match 50 percent of the first five 3

percent contributed by an employee (2.5 percent maximum). For our Coyote Station union 4

employees, and for our other union employees not eligible to participate in the Pension 5

Plan, OTP will match 50 percent of the first six percent contributed by an employee (three 6

percent maximum). 7

8

Q. ARE EMPLOYEES ELIGIBLE TO PARTICIPATE IN OTP’S PENSION PLAN ALSO 9

ELIGIBLE TO PARTICIPATE IN THE MATCHING PORTION OF OTP’S 401(K) 10

PLAN? 11

A. Yes. Pension-eligible employees may make contributions to OTP’s 401(k) Plan, and we 12

strongly encourage them to do so, but they do not receive OTP defined contributions. 13

Rather, they are eligible to receive the OTP employer match, subject to the maximum 14

matching provisions I described above and as noted in Exhibit___(PEW-1), Schedule 5a. 15

16

Q. PLEASE DESCRIBE OTP’S PENSION PLAN. 17

A. OTP has a defined benefit pension plan that has two benefit formulas: (1) a benefit formula 18

for non-union and union employees, other than Coyote Station union employees; and (2) a 19

separate benefit formula for union employees at Coyote Station (collectively the “Pension 20

Plan”). A summary description of OTP’s Pension Plan is provided in Exhibit___(PEW-21

1), Schedule 5b. An overall retirement benefits eligibility summary is provided in 22

Exhibit___(PEW-1), Schedule 5c. 23

24

Q. DOES OTP PROVIDE OTHER POST-RETIREMENT EMPLOYEE BENEFITS 25

(OPEBS)? 26

A. Yes. OTP also provides a medical benefits program and life insurance for a limited 27

number of eligible retirees and a long-term disability medical plan for a limited number of 28

disabled employees. The life insurance benefit is no longer available to new employees as 29

a company-paid expense. Eligibility is restricted to those employees who had 25 years of 30

14 PU-17-

Wasberg Direct


service with OTP as of January 1, 2003, and to Coyote Union employees who were 1

employees at the time OTP took over the operations of Coyote Station in 1998. 2

3

Q. HAS OTP TAKEN STEPS TO MANAGE THE COST OF ITS PENSIONS AND OPEBS? 4

A. Yes. Since 2006, OTP has made significant changes to retirement benefits to control costs. 5

The first significant change eliminated eligibility to participate in the OTP Pension Plan 6

for employees hired after certain dates. This change is sometimes referred to as a “soft 7

freeze.” The effective dates of the soft freeze differ depending on bargaining-unit status. 8

For non-union employees, the soft freeze eliminated participation in the OTP Pension Plan 9

for employees hired after August 31, 2006. For the Coyote Plant bargaining-unit 10

employees, the soft freeze eliminated participation in the OTP Pension Plan for employees 11

hired after December 31, 2008. For our other bargaining-units, the soft freeze eliminated 12

participation in the OTP Pension plan for employees hired after October 31, 2010. 13

Employees who are not eligible for the OTP Pension Plan are eligible to receive the OTP 14

contributions to their 401k plan I described earlier in my Direct Testimony. 15

The second significant change was the elimination post-retirement medical benefits 16

for new employees. Like the soft freeze on the OTP Pension Plan, this change means that 17

employees hired after certain dates are not eligible for post-retirement medical benefits. 18

Also like the soft freeze on the OTP Pension Plan, the effective dates differed based on 19

bargaining unit status. For our non-union employees, this soft freeze was effective for 20

employees hired after August 31, 2006. For the Coyote Plant bargaining unit, this soft 21

freeze was effective for employees hired after December 31, 2008. For our other 22

bargaining-units, this soft freeze was effective for employees hired after October 31, 2013. 23

For all of these groups, participation is also limited to persons who are age 55 or older at 24

retirement with 10 or more years of service and eligible for or enrolled in the OTP medical 25

program as of retirement. 26

27

15 PU-17-

Wasberg Direct


Q. WHAT HAS BEEN THE EFFECT OF THE SOFT FREEZE ON THE OTP PENSION 1

PLAN? 2

A. The soft freeze, in combination with retirements, is significantly reducing the number of 3

OTP employees participating in the OTP Pension Plan. As a result of the soft freeze on 4

the OTP Pension Plan, we now have 30 percent of our active employees who are not 5

participating in the OTP Pension Plan. We also have another 30 percent of our employees 6

who are age 55 or older. Our average retirement age is slightly below 62 years of age, so 7

we would expect to have another 30 percent of our employee population retire within the 8

next seven years. As a result, we expect to move from 30 percent of OTP employees 9

outside the OTP Pension Plan to over 60 percent of our employees not participating in the 10

OTP Pension Plan within the next seven years. 11

12

Q. HAVE COSTS CONTINUED TO INCREASE FOR THE OTP PENSION PLAN AND 13

OPEB DESPITE THESE EFFORTS? 14

A. Yes. Cost increases related to the Pension Plan (FAS 87), Retiree Medical (FAS 106), and 15

Long-Term Disability Medical (FAS 112) have continued to increase and reflect the effects 16

of: (1) our employee population, with over 45 percent of our employees at 50 years of age 17

and older; (2) lower-than-average discount rates; (3) changing mortality tables; and (4) cost 18

increases within the medical industry. The OTP Pension Plan is designed to provide more 19

significant benefits to longer-term employees, and the post-retiree medical benefits do not 20

start accruing until an employee reaches age 45. As a result, we will not see short-term 21

cost savings as a result of the changes to the Pension Plan and OPEBs, but we will see 22

mitigated cost increases for our customers in the long-term along with reduced volatility. 23

We have also seen, in prior years, increases in Long Term Disability Medical due to our 24

aging workforce. 25

26

Q. HAS OTP BEEN PRUDENT IN ESTABLISHING AND THEREAFTER 27

CONTROLLING THE COSTS OF ITS DEFINED BENEFIT PENSION PLAN? 28

A. Yes. Defined benefit pension plans were very common when OTP established its defined 29

benefit pension plans in 1975, and OTP’s decision to establish such a plan was prudent and 30

16 PU-17-

Wasberg Direct


consistent with industry compensation and benefit practices. Since then, OTP has 1

continued to be prudent in managing its pension plans and other retirement benefits, as I 2

explained earlier. Most importantly, it has frozen participation in the Pension Plan. 3

4

Q. ARE OTP’S BENEFIT PLANS REASONABLE COMPARED TO THE MARKET? 5

A. Yes. Although there are no overall studies that compare total cash compensation plus 6

benefits between companies, the compensation and benefit studies that are available 7

demonstrate that the OTP compensation and benefit programs are reasonable. In addition 8

to the studies of cash compensation that I discussed earlier in my Direct Testimony, OTP 9

also routinely participates in the Willis Towers Watson Energy Industry BenVal Study. A 10

copy of the 2017 Willis Towers Watson Energy Industry BenVal Study (2017 BenVal 11

Study) is provided as NOT PUBLIC Exhibit___(PEW-1), Schedule 6. 12

13

Q. PLEASE DESCRIBE THE 2017 BENVAL STUDY. 14

A. The 2017 BenVal Study focused on new hires and included 41 participating energy services 15

companies. The companies were further defined by revenue size. The 2017 BenVal Study 16

showed that OTP’s relative benefit value for non-bargaining unit employees was 89 17

percent, or 11 percent below the average of the participating companies. For bargaining 18

unit employees, it was 99 percent, or one percent below the average of the participating 19

companies. 20

21

Q. WHAT ARE THE IMPLICATIONS OF HAVING A BENEFIT PACKAGE WITH A 22

LOWER BENEFIT VALUE? 23

A. It is possible that having a lower benefit value could inhibit OTP’s ability to attract, retain 24

and engage the talented workforce needed to deliver safe and reliable electric service to 25

our customers. It is imperative that we continue to offer a compensation and benefit 26

package that is competitive for our employees and a good value to our customers. We will 27

continue to monitor our benefit package going forward to ensure it doesn’t become a 28

detriment to our ability to attract, retain and engage quality employees. 29

17 PU-17-

Wasberg Direct


2018 EMPLOYEE COMPENSATION COSTS 1

Q. PLEASE IDENTIFY THE COSTS OF OTP’S COMPENSATION PLAN THAT ARE 2

INCLUDED IN THE 2018 TEST YEAR. 3

A. The 2018 Test Year costs include: (1) OTP wages and salaries; (2) the OTP Annual 4

Incentive Plan, with the KPA Plan limited by average payout levels from 2014-2016, and 5

the Management Plan subject to a cap based on 25 percent of employees’ salaries; (3) OTP 6

employee health and dental benefit plans; and (4) OTP costs relating to the Pension Plan, 7

the 401(k) Plan and OPEBs. 8

A. Wages and Salaries 9

Q. HOW WERE 2018 WAGE AND SALARY LEVELS DETERMINED? 10

A. The 2018 base wages and salaries are based on a three percent increase over the base wage 11

and salary level for 2017 for both union and non-union employees. The wage and the 12

salary components for 2018 have also been adjusted for the projected employee census. 13

The three percent increase reflects what OTP believes is necessary to remain competitive 14

in the labor market in which it competes. Accordingly, its primary purpose is to maintain 15

competitive compensation levels. It also addresses changes in the cost of living. 16

17

Q. PLEASE FURTHER EXPLAIN THE 2018 INCREASE FOR NON-UNION 18

EMPLOYEES OTHER THAN MANAGEMENT. 19

A. OTP budgeted for a three percent increase for non-union employees effective 20

April 1, 2018. The three percent increase is cumulative and individual employee increases 21

will vary depending on performance, market data, and where the employee’s wage falls 22

within the respective wage range. 23

24

Q. PLEASE FURTHER EXPLAIN THE 2018 INCREASE FOR UNION EMPLOYEES. 25

A. OTP has two collective bargaining agreements (CBAs) representing approximately 392 26

union employees: (1) a CBA with four International Brotherhood of Electrical Workers 27

(IBEW) Local Unions representing approximately 331 employees; and (2) a CBA with one 28

IBEW Local Union representing approximately 61 Coyote Station employees. The Coyote 29

18 PU-17-

Wasberg Direct


Station CBA, effective September 1 for each year, is currently undergoing contract 1

negotiations. The CBA for the other four IBEW Local Unions, effective November 1 for 2

each year is currently undergoing contract negotiations. 3

4

Q. PLEASE FURTHER EXPLAIN THE 2018 INCREASE FOR MANAGEMENT AND 5

EXECUTIVE EMPLOYEES. 6

A. The three percent non-union wage increases effective April 1, 2018 include management 7

and executive employees. The three percent increase is cumulative and individual 8

employee increases will vary depending on performance, market data, and where the 9

employee’s salary falls within the respective wage range. 10

B. Incentive Compensation 11

Q. HOW WAS THE OTP ANNUAL INCENTIVE PLAN COMPONENT OF THE 2018 12

TEST YEAR REVENUE REQUIREMENT DETERMINED? 13

A. The OTP 2018 Test Year revenue requirement includes amounts for the OTP KPA Plan 14

that was based on a 3.5 percent payout level, which is the three-year average payout level 15

for the OTP KPA Plan for 2014 through 2016, as shown on attached as Exhibit___ (PEW-16

1), Schedule 7a. The OTP 2018 Test Year revenue requirement also includes an amount 17

for the OTP Management Plan, based on our 2018 budget forecast, less any amounts in 18

excess of 25 percent of wages for each individual within that plan. NOT PUBLIC 19

Exhibit___(PEW-1), Schedule 7b, illustrates application of the 25 percent cap for the years 20

2012-2016. The 25 percent cap reduced the 2018 Test Year expense by $143,261 (OTP 21

Total) $58,034 (OTP ND) from our 2018 budget for the OTP Management Plan and from 22

the plan for the OTP President. 23

24

Q. WHY IS IT APPROPRIATE TO INCLUDE THE COSTS OF THE OTP ANNUAL 25

INCENTIVE PLAN IN THE 2018 TEST YEAR REVENUE REQUIREMENT? 26

A. The costs of the OTP Annual Incentive Plan should be included in the 2018 Test Year 27

revenue requirement for several reasons. First, OTP’s Annual Incentive Plan is an 28

important part of OTP’s total compensation plan; without it, OTP’s total cash 29

19 PU-17-

Wasberg Direct


compensation would be significantly below market, making it harder for OTP to attract 1

and retain essential employees. Second, as I have explained earlier in my Direct 2

Testimony, the OTP Annual Incentive Plan includes an appropriate range and balance of 3

factors that provide benefits to customers. Third, including annual incentive plans in total 4

compensation packages is an established utility market practice. Fourth, OTP has applied 5

appropriate controls to the costs to be recovered in the revenue requirement. 6

When the 25 percent cap is applied, OTP’s annual Total Cash Compensation and 7

Total Direct Compensation levels for executives are even further below competitive market 8

levels. The 2018 forecasted level for the Management Plan, adjusted to remove any 9

amounts over a 25 percent cap on individual employee incentives, is appropriate for 10

determining OTP’s 2018 Test Year revenue requirement. Using a three-year average 11

payout level for the OTP KPA Plan fairly represents OTP’s typical payout levels over time. 12

C. Medical and Dental 13

Q. WHAT IS THE BASIS FOR OTP’S 2018 EMPLOYEE MEDICAL/DENTAL, PENSION, 14

DEFINED CONTRIBUTION AND OPEB COSTS? 15

A. OTP’s employee medical/dental, pension, 401(k), and OPEB costs (sometimes collectively 16

“Benefit Costs”) are based on projected 2018 cost levels. OTP’s 2018 pension (ASC 715) 17

costs and OPEB (ASC 715 and ASC 712) costs are based on estimates by Mercer. OTP’s 18

pension costs are based on our demographics and standard actuarial assumptions, using 19

and applying ASC 715 and ASC 712 accounting standards. Actuarial projections are 20

provided formally each January and we will make that information available on request. 21

Copies of 2018 Mercer estimates are attached as Exhibit___(PEW-1), Schedule 8. 22

23

Q. HOW DO OTP’S ACTIVE MEDICAL COSTS COMPARE BETWEEN 2016, 24

PROJECTED 2017, AND 2018? 25

A. Active medical costs decreased between 2016 and projected 2017 by ($37,367) (OTP 26

Total) / ($15,137) (OTP ND), but overall Group Insurance increased by $419,999 (OTP 27

Total) / $170,139 (OTP ND). Active medical costs increased between projected 2017 and 28

2018 by $548,200 (OTP Total) / $222,073 (OTP ND), and they account for the majority of 29

20 PU-17-

Wasberg Direct


increases within overall Group Insurance $587,169 (OTP Total) / $237,859 (OTP ND). 1

Medical costs continue to rise and remain a common challenge across the United States. 2

Q. DOES OTP USE PROJECTIONS TO DETERMINE ITS ACTIVE MEDICAL COSTS? 3

A. Yes. OTP has utilized third parties to determine active medical costs for many years and 4

has determined it to be an effective manner in which to project future active employee 5

medical costs. The 2018 Renewal Results, prepared by Mercer, are provided as 6

Exhibit___(PEW-1), Schedule 9. 7

8

Q. HOW DOES THE PROJECTED INCREASE FOR ACTIVE MEDICAL COSTS FOR 9

2018 COMPARE TO RECENT EXPERIENCE? 10

A. The move to high deductible health plans along with an increase to the employees’ cost 11

share of the premiums has helped to keep the increase in overall medical costs below 12

current medical cost trends. The projected active medical costs increase between 2016 and 13

2018 is 5.6 percent (OTP Total), with costs actually falling by 0.4 percent between 2016 14

and 2017. The size of our health care plans does create some volatility, year over year, but 15

we know our move to the HDHP in 2012 means we have spent less than we would have 16

had we stayed on our previous health care plan. Nationally, health care costs continue to 17

trend upward and OTP will continue to look at plan design changes, premium structures 18

and other alternatives to mitigate future health care spending. Other alternatives include 19

the continued promotion of preventive health care and wellness which help to reduce health 20

care costs over the long-term. OTP does, however, intend to provide a quality health care 21

benefit at a reasonable price for our employees, in our effort to attract, engage and retain 22

quality employees in the coming years. 23

D. Pension and Benefits 24

Q. HAVE PENSION COSTS INCREASED BETWEEN 2016 AND 2018? 25

A. Yes. The projected pension costs have increased by approximately $3.2 million (OTP 26

Total) / $1.3 million (OTP ND) between 2016 and 2018. 27

28

21 PU-17-

Wasberg Direct


Q. DOES MERCER REGULARLY PERFORM ACTUARIAL ANALYSES FOR OTP? 1

A. Yes. Mercer performs annual analyses of OTP’s pension and OPEB costs. Annual 2

actuarial analyses of pension and OPEB obligations are performed to satisfy several legal 3

requirements including: (1) ERISA, which requires annual actuarial reports regarding 4

pensions be filed with the Department of Labor; (2) the pension Benefits Guarantee 5

Corporation, which also requires annual actuarial reports regarding Pension funding status; 6

(3) the Internal Revenue Service, which requires filing of an annual actuarial report as part 7

of Form 5500; and (4) the Securities Exchange Commission, which requires disclosure of 8

actuarial analyses of pensions and OPEBs in annual Form 10-K filings. These legal 9

requirements underscore the fact that the Mercer actuarial analyses are reliable and 10

independent assessments of anticipated pension and OPEB expenses. The data relied on 11

in this proceeding to estimate 2018 pension and OPEB expenses conforms to these 12

standards. 13

14

Q. WHAT ARE THE PRIMARY SOURCES OF THE INCREASE IN ASC 715 PENSION 15

COSTS (FORMERLY KNOWN AS FAS 87) BETWEEN 2016 AND 2018? 16

A. The most significant factor is the decrease in the discount rate, or the interest rate, used to 17

determine present value of future benefit obligations. The lower discount rate means a 18

lower interest rate and results in both an increased service cost and projected benefit 19

obligation. The lower discount rate requires more cash today to meet future benefit 20

obligations within the pension plan. 21

Another factor contributing to the increase is a decreased mortality rate, reflecting 22

increased life expectancies. A decreased mortality rate results in an expectation that 23

pension benefit payments will occur over a longer period, increasing expense. 24

The decrease in the Estimated Return on Assets (EROA) decreased the offset to 25

pension expenses while the increase in the pension asset (resulting from both growth and 26

the increase in the prepayments by OTP) limited the increase in the expense. 27

28

22 PU-17-

Wasberg Direct


Q. HOW DO OTP’S COSTS FOR ASC 715 POST-RETIREMENT MEDICAL/LIFE 1

(FORMALLY KNOWN AS FAS 106 RETIREE MEDICAL) AND ASC 712 POST-2

EMPLOYMENT (LTD) MEDICAL (FORMALLY KNOWN AS FAS 112 LONG TERM 3

DISABILITY) COMPARE BETWEEN 2016, 2017, AND 2018? 4

A. Post-Retirement Medical/Life costs increased between 2016 and projected 2018 by 5

approximately $1.0 million (OTP Total) / ($421,177 (OTP ND), with much of the increase 6

occurring between 2016 and 2017. Long-Term Disability costs decreased by 7

approximately $242,447 (OTP Total) / ($98,214 (OTP ND), with much of the decrease 8

occurring between 2016 and 2017 9

10

Q. WHAT WAS THE BASIS OF OTP’S PROJECTED ASC 715 POST-RETIREMENT 11

MEDICAL INCREASES FROM 2016 TO 2018? 12

A. OTP experienced the majority of the increase in Post-Retirement Medical/Life expense in 13

2017, primarily related to aging demographics within our employee group, along with 14

medical cost expectations. The 2018 budget for Post-Retirement Medical/Life is slightly 15

higher than 2017, but is still an increase over 2016 actuals. 16

17

Q. ARE ASC 712 POST-EMPLOYMENT MEDICAL (LTD)EXPENSES SENSITIVE TO 18

PARTICIPANT COUNTS? 19

A. Yes. Long Term Disability medical liability is very sensitive to participant counts. Just 20

one or two participants can have a significant impact on plan liability. Moreover, OTP is 21

required to recognize actuarial gains and losses immediately; changes in liability based on 22

participant count changes and claims experience are fully recognized in the year they occur. 23

These factors are considered as part of Mercer’s actuarial projections for ASC 712. 24

25

Q. HAVE YOU INCLUDED THE MERCER REPORTS ON WHICH YOU HAVE BASED 26

2018 COSTS FOR PENSION, POST RETIREMENT MEDICAL, AND LONG-TERM 27

DISABILITY BENEFITS? 28

A. Yes. I have attached Exhibit___(PEW-1), Schedule 10a – 2017 Pension Accounting 29

Expense Report – Mercer; Exhibit___(PEW-1), Schedule 10b – 2017 Pension DAMP – 30

23 PU-17-

Wasberg Direct


Mercer; and Exhibit___(PEW-1), Schedule 10c – 2017 Post-retirement Medical and LTD 1

Medical DAMP – Mercer which provide the basis for those 2018 costs. 2

CONCLUSION 3

Q. PLEASE SUMMARIZE YOUR DIRECT TESTIMONY. 4

A. OTP has structured a reasonable total cash compensation and benefits package that is 5

needed in order to attract and thereafter retain and engage talented employees necessary to 6

provide high quality electric service to our customers. This includes base salaries, active 7

and post-retirement benefits, and incentive compensation. OTP’s proposed compensation 8

and benefit costs are fair and reasonable, and should be included in rates. 9

10

Q. DOES THIS CONCLUDE YOUR DIRECT TESTIMONY? 11

A. Yes. 12

Case No. PU-17- Exhibit___(PEW-1), Schedule 1

Page 1 of 1

PETER E. WASBERG

EMPLOYMENT

2008 – PRESENT Otter Tail Power Company Fergus Falls, MN

Director, Human Resources & Safety


Manager, Human Resources

2002 – 2004 Otter Tail Power Company Crookston, MN

Area Manager, Crookston & Bemidji

1997 – 2002 Otter Tail Power Company Hallock, MN

Division Manager, Hallock & Crookston Divisions Crookston, MN

1995 – 1997 Otter Tail Power Company Bemidji, MN

Office Manager, Bemidji Division

1991 – 1995 Otter Tail Power Company Milbank, SD

Division Accountant, Milbank Division

EDUCATION

• Bachelor of Arts Concordia College, Moorhead

Business Administration

Psychology

• Management Institute University of Wisconsin – Madison

Coaching & Counseling

Full-Range Leadership

• Executive and Organizational Development The Levinson Institute

PROFESSIONAL AFFILIATIONS

• Crookston National Bank Director, since 2001

• University of Minnesota, Crookston All-College Advisory & Advancement

Board, since 2002

• Society of Human Resources Management (SHRM) Member

© MERCER 2015 1

T A B L E O F C O N T E N T S

• Introduction 2

• Benchmarking Methodology 4

• Market Analysis by Grade 6

• Market Analysis by Family 19

• Market Incentive Analysis 30

Page


PUBLIC DOCUMENT - NOT PUBLIC (OR PRIVILEGED) DATA HAS BEEN EXCISED Page 2 of 43


PUBLIC DOCUMENT – NOT PUBLIC (OR PRIVILEGED) DATA HAS BEEN EXCISED Page 3 - 43



Schedule 2 – 2015 Mercer Compensation Benchmark Study

to

Direct Testimony of Peter E. Wasberg


For Authority to Increase Rates for Electric Utility Service in North Dakota

THE REMAINDER OF THIS DOCUMENT IS NOT PUBLIC IN ITS ENTIRETY

© MERCER 2015 1

T A B L E O F C O N T E N T S

• Introduction 2

• Executive Summary 4

• Benchmarking Methodology 6

• Compensation Data 11

• Appendices

– Survey Job Match Descriptions 18

– Incumbent Exhibits 29

Page







Schedule 3 – 2015 Mercer Executive Compensation Review

to




THE REMAINDER OF THIS DOCUMENT IS NOT PUBLIC IN ITS ENTIRETY

Case No. PU-17-

Exhibit___(PEW-1), Schedule 4

Page 1of 2

Actuals

2010 2011 2012 2013 2014 2015 2016 2012 - 2016

Actual

compared to

scenario

Total cost $9,222,610 $10,868,667 $7,274,435 $8,541,747 $9,404,875 $9,458,226 $9,335,644

Less ee premiums* $1,383,392 $1,595,434 $293,283 $389,819

Company cost $7,839,218 $9,273,233 $9,164,943 $8,945,825

Plus H.S.A. contributions* $943,450 119,850 $151,100 $70,000 $79,360

Actual $7,839,218 $9,273,233 $8,217,885 $8,661,597 $9,555,975 $9,234,943 $9,025,185 $44,695,585 $0

Scenarios had we stayed on the plan in use prior to 2012

6% increase in medical trend and no changes to the employer/employee premium cost share 2012-2016

Total Cost 6% $10,868,667 $11,520,787 $12,212,034 $12,944,756 $13,721,442 $14,544,728

Employee Premium 15% $1,595,434 $1,728,118 $1,831,805 $1,941,713 $2,058,216 $2,181,709

Employer Premium $9,273,233 $9,792,669 $10,380,229 $11,003,043 $11,663,225 $12,363,019 $55,202,185 -$10,506,600

6% increase in medical trend and moving the premium cost share from 85/15 to 75/25, starting in 2012 2012-2016

Total Cost 6% $10,868,667 $11,520,787 $12,212,034 $12,944,756 $13,721,442 $14,544,728



7% increase in medical trend and no changes to the employer/employee premium cost share 2012-2016

Total Cost 7% $10,868,667 $11,629,474 $12,443,537 $13,314,584 $14,246,605 $15,243,868



7% increase in medical trend and moving the premium cost share from 85/15 to 75/25, starting in 2012 2012-2016

Total Cost 7% $10,868,667 $11,629,474 $12,443,537 $13,314,584 $14,246,605 $15,243,868



Mercer e-mail (11/17/2015)

Average Annual Trends Medical Rx

2016 Combined

2015 6.00% 10.50% As for future trends, we (Mercer) are using 7% overall trend (combined Medical and Rx) for

2014 6.75% 8.08% the next two to three years. We (Mercer) also do expect that Rx trend will continue to

2013 7.00% 6.75% increase due to the significant uptick in the number and cost of specialty medications.

2012 7.25% 6.75%

These are midpoints from the Mercer national

6.80%

Case No. PU-17-

Exhibit___(PEW-1), Schedule 4

Page 2of 2

annual trend guidelines table

Had we made no changes to our health care plan that existed prior to 2012, and assumed a conservative medical trend of 6%, we will save our

customers $10,506,600, in the years 2012 - 2016 with our move to the HDHP in 2012.

Even with a very conservative estimate, assuming a 6% increase in medical trend over the period 2012-2016, and getting our Unions to agree to a 75/25

company/employee split, we will save our customers $4,012,226, in the years 2012 - 2016 with our move to the HDHP in 2012.

Assuming a more realistic estimate, with a 7% increase in medical trend over the period 2012-2016, and no changes to the our health care plan that

existed prior to 2012, we will save our customers $12,150,773, in the years 2012 - 2016 with our move to the HDHP in 2012.

Assuming a more realistic estimate, with a 7% increase in medical trend over the period 2012-2016, and an assumption that we could have negotiated

a 75/25 company/employee premium split, we will save our customers $5,462,966, in the years 2012 - 2016 with our move to the HDHP in 2012.

Case No. PU-17- Exhibit___(PEW-1), Schedule 5c

Page 1 of 1

Otter Tail Power Company – Retirement Benefits Eligibility

Employee Group Eligible for Pension

Eligible for 401(k) Contribution – 4%



Eligible for 401(k) Match; 50% of the First 5% (max – 2.5%)

Eligible for 401(k) Match; 50% of the First 6% (max – 3.0%)

Eligible for Retiree Medical

Non-union hired prior to 9/1/2006

X X X

Non-union hired 9/1/2006 or after

X X

Coyote Union hired prior to 1/1/2009

X X X

Coyote Union hired 1/1/2009 or after

X X

Other Union hired prior to 11/1/2013*

X X X

Other Union hired 11/1/2013 or after

X X

*Employees in this group, hired after 10/31/2010, are not eligible for Retiree Medical


PUBLIC DOCUMENT - NOT PUBLIC DATA HAS BEEN EXCISED Page 1 of 63





Schedule 6 – Report of the 2017 Willis Towers Watson Energy Industry Benefits (BenVal)

Study

to




THE REMAINDER OF THIS DOCUMENT IS NOT PUBLIC

IN ITS ENTIRETY

Case No. PU-17-

Exhibit__(PEW-1), Schedule 07a

Page 1 of 1

Year Eligible Earnings # of Participants (1) Target Target % Maximum Max % Budget Actual Actual %

2014 25,222,574.51$ 374 -- -- $1,513,354.47 6% $0 756,677.19$ 3.00%

2015 25,134,936.00$ 367 -- -- $1,508,096.16 6% $0 754,048.00$ 3.00%

2016 26,423,107.00$ 386 -- -- $1,585,386.42 6% $0 1,189,040.00$ 4.50%

3YR Average (2) 25,593,539.17$ 376 $1,535,612.35 6% $0 899,921.73$ 3.50%

Note: (1)

Eligible participants include nonunion group only - union are excluded from plan.(2)

Actual does not inculde FICA. Ms. Petersen's testimony includes FICA and is reporting $927,000.

Key Performance Plan - History

Case No. PU-17-

Exhibit___(PEW-1), Schedule 7b

PUBLIC DOCUMENT - NOT PUBLIC (OR PRIVILEGED) DATA HAS BEEN EXCISED

Page 1 of 1

Year

# participants

(including

President) Budget/Target Actual (1) % of Target

Incentive >

25% of base

pay

20122013201420152016

5-Year Average5-Year Average reduced by 25% cap

(1) Actuals do not include FICA. Ms. Petersen's testimony includes FICA and is reporting $706,575

2018

2018 COSS

Adjustment Bonus

cap at 25% Total

Management Incentive Plan - Payout History (Including President)

OTP Management Incentive



PU-17- Exhibit___(PEW-1), Schedule 8

Page 1 of 5


Page 2 of 5


Page 3 of 5


Page 4 of 5


Page 5 of 5


Page 1 of 2


Page 2 of 2

Case No. PU-17- Exhibit___(PEW-1), Schedule 10a



Case No. PU-17- Exhibit (PEW-1), Schedule 10a


Schedule 10a – 2016 Pension Accounting Expense Report – Mercer

to


In the Matter of the Application of Otter Tail Power Company For Authority to Increase Rates for Electric Utility Service in Minnesota


IN ITS ENTIRETY


Case No. PU-17- Exhibit___(PEW-1), Schedule 10b



Case No. PU-17- Exhibit (PEW-1), Schedule 10b


Schedule 10b – 2016 Pension DAMP – Mercer

to




IN ITS ENTIRETY



Case No. PU-17- Exhibit (PEW-1), Schedule 10c


Schedule 10c – 2016 Post-retirement Medical and LTD Medical DAMP – Mercer

to




IN ITS ENTIRETY
