IEEE Power & Energy Magazine - May/June 2020 - 76

In North America, a second load
research alliance, the Western Load
Research Association (WLRA), was
formed in the late 1970s. The WLRA is a
regional organization for the western part
of North America whose most important
function is information exchange. Some
utilities participate in both the WLRA
and the AEIC. Whereas the WLRA is a
regional association of primarily western U.S. utilities with two conferences
per year, the AEIC is a formal member
association that hosts workshops, training sessions, and conferences and maintains the load research manual. In recent
years, the WLRA fall conference and
AEIC annual load research conference
have been jointly held.

Analytical Methods
During the early years, the cost of installing and maintaining metering and
data collection equipment was both

formidable and challenging. The costper-sample point could easily exceed
US$5,000, meaning that sample design
efficiency was a primary consideration.
However, the minimiz a tion of the sa mple
conf l ict e d w it h t h e
reliability of the early
data collection networks,
which were not dependable. Load researchers
would often make the
planning assumption
of a 20% data loss, increasing sample sizes
to account for missing
data. The sample design
strategy was tied directly
to the sample extrapolation approach.
In the beginning, most
companies were using a simple mean
per unit (MPU) expansion strategy,

W. Cochran's
book, Sampling
Techniques,
became the
go-to sampling
and analysis
book of the
load research
community.

IS YOUR REPUTATION FOR
RELIABLE SERVICE AT RISK?

BirdguarD™
Wildlife Power Outage Protection
BirdguarD™ Material Qualities:
UL94 V-0 Flammability Rating
Material Will NOT Support Combustion
IEEE 1656 & ASTM Tested to 72kV
UV Resistant
ISO 9001:2015 Certified
Women Owned Business - WBENC
Public Utility Commission Certified

INNOVATIONS IN PLASTIC

BirdguarD™ Delivers a Triple Win:
1. Consumer Reliability
2. Utility Asset Protection

Customer Support
P (801) 972-5400 | (888) 658-5003

www.KADDAS.com

ieee power & energy magazine

where the average for the sample was
calculated and extrapolated based on
the number of customers in the population. MPU estimation could be used with
a simple random sample
or within a stratification
framework. Stratification
was primarily done using
Dalenius and Hodges'
minimum variance stratification with Neyman
allocation. These were all
well-established statistical
methods used for estimation and sample design
and formed the original
statistical methods deployed in load research.
In the mid-1970s, Consumers Power Company
was presented with a
challenge by the Michigan Public Service Commission. Although the company had a long history of practicing
load research, the commission staff
requested that the samples be significantly increased. As mentioned previously, adding sample points was a
formidable challenge and an expensive
endeavor. The company took the challenge to Roger Wright at the University of
Michigan. Wright examined the challenge and quickly realized that the load
factor seemed to be a very important
element in the analysis schema. He also
noticed that utilities had significant
amounts of useful information about
how their customers use energy, namely, the monthly billed energy for its full
population of customers. Wright speculated that if this information could be
used in the sample design, then the
sample size requirements could be reduced by leveraging the correlation
between the known billing information
and the variable(s) of interest, such as
system peak demand or the 12-month
coincident peak demands. This gave
rise to the use of an approach called
model-based statistical sampling, which
was merely a modified application of
stratified ratio estimation.
W. Cochran's book, Sampling Techniques, became the go-to sampling
and analysis book of the load research

3. Positive Environmental Impact
Kaddas Enterprises, Inc. 2019. Not for Personnel Protection

may/june 2020

http://www.KADDAS.com

IEEE Power & Energy Magazine - May/June 2020

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - May/June 2020