MAISY Profiler Incorporates Load Profiles to Assess Customer and Segment Profitability in Competitive Markets

Service Area Profiler Extensions

The MAISY Database Profiler is a software system designed to assist MAISY State and Utility Service Area Database users with customer and market hourly load data development. Special applications of the Profiler address specific issues such as DG and profitability analysis.

The Service Area Profiler has now been extended to provide profitability results with dynamic load profiles from distribution utilities and with actual customer hourly loads. Actual loads, which are extracted from the MAISY databases, are appropriate for analysis of customers who have hourly meters and to determine which non-interval metered customers can benefit from hourly meters. Actual load data is also important in revenue management analysis of distribution utilities.

Dynamic load profiles are required for profitability analysis of customers who do not have hourly meters including most residential and small commercial customers. Background information on dynamic load profiles in profitability analysis is provided in a separate section below.

The Service Area Profiler applies load profiles from the distribution utility to individual utility customer monthly energy use extracted from the MAISY database to determine load profiles used in profit calculations. Typical weather for the service area is used for baseline analysis; however, monthly heating and cooling degree day parameters can be used to assess profitability under alternative weather conditions. Changes in weather parameters change both customer load profiles and the dynamic load profiles of the distribution utility.

A sample Profiler screen is shown below.

Distinguishing Between Customer Hourly Load Applications And Load Profiling Applications

MAISY Utility Customer Databases provide hourly load data for each of the 156,000 sample records. These actual customer loads are used in a variety of market analyses and marketing applications. The extension to the Service Area Profiler described on this page incorporates another load profile for each customer based on monthly billing energy and load profiles specified by the distribution utility. Terminology for these distribution-utility profiles varies as does their actual development.

Applications Where Actual Customer Loads are Preferred

• Analysis of customers with hourly meters (e.g., medium and large commercial customers)
• Identification of customers who will benefit from installation of hourly meters
• Determination of load contributions of customers and customer segments to distribution utility system loads

Applications Where Distribution Utility Load Profiles are Preferred

• Analysis of monthly metered customers whose settlement is determined by distribution utility load profiles (residential and small commercial)
• Assessment of cream-skimming and gaming threats to distribution utilities

Background on the Role of Dynamic Load Profiles in Profitability Analysis

Dynamic load profiles are used in competitive markets to translate utility customer monthly billing energy use to hourly energy scheduling requirements and monthly settlements with the distribution utility. Monthly energy use is distributed to hours of the month based on load profiles of the appropriate load class and hourly temperatures.

The limited nature of load research samples at most utilities typically requires broad customer categories based on rate classes. Consequently, profit potential exists for energy service providers who are able to differentiate between customers or customer segments that reflect significantly different cost-of-service characteristics within the same rate classes.

While technical details on load profiling differ from utility to utility, the following characterization illustrates profit issues which arise under the dynamic load profiling process. The chart below shows hourly energy use for a January and August weekday for two residential customers and for the entire residential class.

Total energy use for the two days for each customer is the same as shown in the table to the right. Customer A uses much more energy on a summer day while Customer B uses slightly more energy on a winter day.

Customer A Customer B January Weekday kWh 21 43 August Weekday kWh 58 36 Total 79 79

If these customers were on interval (hourly) meters, their energy service provider (ESP) would be responsible for supplying enough energy to meet each customer's needs on an hour by hour basis throughout each day of each month. Substantial variation in the cost of generating power from hour to hour throughout the day and across seasons of the year, lack of differentiation in rate structures with respect to customer cost of service and the daily and seasonal variations in loads which occurs across customers result in variations in customer profitability. ESP's can evaluate the cost of service and revenue for customers when hourly load histories are available. When historical customer data are not available it is possible to analyze customer-detailed profitability issues with analysis and models of customer profitability based on information sources like the MAISY Energy Marketing Databases (see Assessing Utility Market "Headroom": Pitfalls of Traditional Analysis and Pennsylvania Profitability Analysis Identifies Individual High-Value Electric Utility Customers With 83-96% Accuracy ).

However, since most residential and small commercial customers do not have hourly meters, load profiling is used to assign hourly energy use to customers based on their monthly meter readings. Energy suppliers are required to provide power based on these load profiles which are dynamically determined to reflect hourly weather conditions. Thus ,for customers who are not interval metered, actual load profiles are not directly relevant in profitability analysis (actual load profiles are still important in terms of revenue management for the distribution utility and in decisions concerning the use of interval metering for selected customers and customer segments as indicated above).

The chart below illustrates the load characterizations relevant for these two customers under load profiling.

As the chart indicates, load profiling removes individual customer hour-to-hour variations in load profiles used for scheduling and settlement. Differences in monthly energy use are still maintained however; that is, Customer A still uses less energy than the average in January and more in August and Customer B reflects the opposite seasonal pattern. Variations in monthly energy use among customers along with variations in energy supply costs still generate variations in customer profit even with dynamic load profiling.

The potential for profit variation is suggested by further examination of our example customers. Each day's energy use is multiplied by 31 days to provide more meaningful monthly figures in the table below. Average monthly supply costs of $0.02/kWh and $0.045are used in this example. Multiplying the supply costs by monthly kWh yields a monthly electricity supply cost. Adding the January and August monthly supply costs shows that the total supply costs for these two months varied between these two customers by $16 even though each customer used exactly the same total amount of electricity for the two months.

		Customer A	Customer B
January	Supply Cost	$0.020/kWh	$0.020/kWh
	Monthly kWh	656	1,331
	Monthly Supply Cost	$13	$27
August	Supply Cost	$0.045/kWh	$0.045/kWh
	Monthly kWh	1,783	1107
	Monthly Supply Cost	$80	$50
Both Months	Total kWh Supplied	2438	2438
	Supply Cost	$93	$77
	Supply cost/kWh	$0.0381	$0.0316

This example illustrates the potential for substantial profit variations across customers and customer segments, even after inter-hour profit variations are removed with the use of dynamic load profiling by distribution utilities.

A recent analysis of the Texas estimated residential customer commodity profits of $75 per customer/year and small-mid market commercial customer profits of $1,500/year for the most profitable quintile of customers in each class (customer losses are of about equal magnitude for the least profitable customer quintiles in each class). Since profit scoring models can distinguish as much as 70 percent of the variation in profit among customers, prior to acquisition, energy service markets will soon reflect the same kind of frenetic marketing activity that accompanied the early years of telecommunications market deregulation.