Unintended biases of an electricity demand forecast based on a double-log regression

Chi Keung Woo; J. Zarnikau; Kang Hua Cao

doi:10.1016/j.tej.2020.106866

Unintended biases of an electricity demand forecast based on a double-log regression

Chi Keung Woo, J. Zarnikau^*, Kang Hua Cao

^*Corresponding author for this work

Research output: Contribution to journal › Journal article › peer-review

3 Citations (Scopus)

34 Downloads (Pure)

Abstract

This paper identifies the unintended biases occasionally not recognized when using a double-log regression to make an electricity demand forecast. It shows that ignoring the stochastic nature of forecasts of income, price and weather can vastly overstate the forecast's precision, potentially causing inadequate resource procurement for reliable service at least cost. Fortunately, the overstated precision is readily avoidable because its correction uses information available when making the forecast.

Original language	English
Article number	106866
Journal	Electricity Journal
Volume	33
Issue number	10
DOIs	https://doi.org/10.1016/j.tej.2020.106866
Publication status	Published - Dec 2020

Scopus Subject Areas

Business and International Management
Energy (miscellaneous)
Management of Technology and Innovation

User-Defined Keywords

Double-log regression
Electricity demand forecast
Forecast bias
Forecast precision

Access to Document

10.1016/j.tej.2020.106866

Full TextAccepted author manuscript, 233 KB

Cite this

@article{7622156b954a4ed8b4af0408b7d9c364,

title = "Unintended biases of an electricity demand forecast based on a double-log regression",

abstract = "This paper identifies the unintended biases occasionally not recognized when using a double-log regression to make an electricity demand forecast. It shows that ignoring the stochastic nature of forecasts of income, price and weather can vastly overstate the forecast's precision, potentially causing inadequate resource procurement for reliable service at least cost. Fortunately, the overstated precision is readily avoidable because its correction uses information available when making the forecast.",

keywords = "Double-log regression, Electricity demand forecast, Forecast bias, Forecast precision",

author = "Woo, {Chi Keung} and J. Zarnikau and Cao, {Kang Hua}",

note = "Funding Information: C.K. Woo{\textquoteright}s research is funded by research grants ( #4388 and #4400 ) from the Education University of Hong Kong . J. Zarnikau{\textquoteright}s contribution to this paper is part his ongoing research on electricity economics at UT Austin. Funding Information: C.K. Woo's research is funded by research grants (#4388 and #4400) from the Education University of Hong Kong. J. Zarnikau's contribution to this paper is part his ongoing research on electricity economics at UT Austin.",

year = "2020",

month = dec,

doi = "10.1016/j.tej.2020.106866",

language = "English",

volume = "33",

journal = "Electricity Journal",

issn = "1040-6190",

publisher = "Elsevier Inc.",

number = "10",

}

TY - JOUR

T1 - Unintended biases of an electricity demand forecast based on a double-log regression

AU - Woo, Chi Keung

AU - Zarnikau, J.

AU - Cao, Kang Hua

N1 - Funding Information: C.K. Woo’s research is funded by research grants ( #4388 and #4400 ) from the Education University of Hong Kong . J. Zarnikau’s contribution to this paper is part his ongoing research on electricity economics at UT Austin. Funding Information: C.K. Woo's research is funded by research grants (#4388 and #4400) from the Education University of Hong Kong. J. Zarnikau's contribution to this paper is part his ongoing research on electricity economics at UT Austin.

PY - 2020/12

Y1 - 2020/12

N2 - This paper identifies the unintended biases occasionally not recognized when using a double-log regression to make an electricity demand forecast. It shows that ignoring the stochastic nature of forecasts of income, price and weather can vastly overstate the forecast's precision, potentially causing inadequate resource procurement for reliable service at least cost. Fortunately, the overstated precision is readily avoidable because its correction uses information available when making the forecast.

AB - This paper identifies the unintended biases occasionally not recognized when using a double-log regression to make an electricity demand forecast. It shows that ignoring the stochastic nature of forecasts of income, price and weather can vastly overstate the forecast's precision, potentially causing inadequate resource procurement for reliable service at least cost. Fortunately, the overstated precision is readily avoidable because its correction uses information available when making the forecast.

KW - Double-log regression

KW - Electricity demand forecast

KW - Forecast bias

KW - Forecast precision

UR - http://www.scopus.com/inward/record.url?scp=85092409233&partnerID=8YFLogxK

U2 - 10.1016/j.tej.2020.106866

DO - 10.1016/j.tej.2020.106866

M3 - Journal article

AN - SCOPUS:85092409233

SN - 1040-6190

VL - 33

JO - Electricity Journal

JF - Electricity Journal

IS - 10

M1 - 106866

ER -

Unintended biases of an electricity demand forecast based on a double-log regression

Abstract

Scopus Subject Areas

User-Defined Keywords

Access to Document

Other files and links

Fingerprint

Cite this