Page 75 - ECLECTIC MARCH-2022 INSIDE PAGES_p2.indd
P. 75
ORITIZATION OF LOW VOLTAGE CASES
USING AI / ML TECHNIQUES FOR CAPEX
OPTIMIZATION
ESHITA ROY, SAYAN KOLEY
PLANNING
Abstract
A stable and reliable electric power supply system is an inevitable prerequisite for the technological and
economic growth of the country. One of the main challenges for electricity distribution companies in the MV/
LV public distribution network is Low Voltage. Indeed, being one of the most important parts of the distribution
network, MV/LV public distribution network needs critical attention to solve this issue. Moreover, the reliability of
the supply has become a subject of discussion at di erent levels of society. Every year thousands of low voltage
complaints have been docketed in call-center database. Some cases are resolved by local rectifications like
replacement of cut-out/fuse, change of isolation etc. with little or less involvement. However, multiple calls from
an area warrant major investments like Overhead conductor replacement, AB cabling for parallel network or
DTR commissioning. Presently, the allocated budget to resolve Low voltage issues is disbursed among several
cases based on the previous history, gut feeling, availability of space, public grievance, and political influence
without any prioritization in terms of financial benefits. A data-centric approach to optimize the investment is
the need of the hour technological and economic sustainability. This paper delineates a novel approach to cook
thousands of low voltage calls and to generate a priority list based on critical predefined weighted parameters
& AI/ML techniques for CAPEX optimization. It concludes by providing information on the action plan and future
prospects of this method in the CESC’s distribution system.
Keywords: CESC (Erstwhile Calcutta Electric Supply Corporation), DTR (Distribution Transformer), LV (Low
Voltage), Capex Optimization, Prioritization, O/H (Overhead), AB Cable (Aerial Bunched Cable), ROI (Return on
Investment), AI (Artificial Intelligence), ML (Machine Learning).
INTRODUCTION economic growth of the country. One of the main
challenges for electricity distribution companies in the
Electric utilities are confronted with a myriad of challenges MV/LV public distribution network is Low Voltage.
that include aging infrastructure, enhanced expectation
of reliability, and coping effectively with uncertainties Low voltage is a basic concept that refers to lower than
and changing regulation requirements. normal levels of electric power entering buildings. The
effects of this include poor appliance performance,
A stable and reliable electric power supply system is dim lighting and intermittent outages. As per supply
an inevitable prerequisite for the technological and code of WBERC 2013, the variation of voltage should
75
USING AI / ML TECHNIQUES FOR CAPEX
OPTIMIZATION
ESHITA ROY, SAYAN KOLEY
PLANNING
Abstract
A stable and reliable electric power supply system is an inevitable prerequisite for the technological and
economic growth of the country. One of the main challenges for electricity distribution companies in the MV/
LV public distribution network is Low Voltage. Indeed, being one of the most important parts of the distribution
network, MV/LV public distribution network needs critical attention to solve this issue. Moreover, the reliability of
the supply has become a subject of discussion at di erent levels of society. Every year thousands of low voltage
complaints have been docketed in call-center database. Some cases are resolved by local rectifications like
replacement of cut-out/fuse, change of isolation etc. with little or less involvement. However, multiple calls from
an area warrant major investments like Overhead conductor replacement, AB cabling for parallel network or
DTR commissioning. Presently, the allocated budget to resolve Low voltage issues is disbursed among several
cases based on the previous history, gut feeling, availability of space, public grievance, and political influence
without any prioritization in terms of financial benefits. A data-centric approach to optimize the investment is
the need of the hour technological and economic sustainability. This paper delineates a novel approach to cook
thousands of low voltage calls and to generate a priority list based on critical predefined weighted parameters
& AI/ML techniques for CAPEX optimization. It concludes by providing information on the action plan and future
prospects of this method in the CESC’s distribution system.
Keywords: CESC (Erstwhile Calcutta Electric Supply Corporation), DTR (Distribution Transformer), LV (Low
Voltage), Capex Optimization, Prioritization, O/H (Overhead), AB Cable (Aerial Bunched Cable), ROI (Return on
Investment), AI (Artificial Intelligence), ML (Machine Learning).
INTRODUCTION economic growth of the country. One of the main
challenges for electricity distribution companies in the
Electric utilities are confronted with a myriad of challenges MV/LV public distribution network is Low Voltage.
that include aging infrastructure, enhanced expectation
of reliability, and coping effectively with uncertainties Low voltage is a basic concept that refers to lower than
and changing regulation requirements. normal levels of electric power entering buildings. The
effects of this include poor appliance performance,
A stable and reliable electric power supply system is dim lighting and intermittent outages. As per supply
an inevitable prerequisite for the technological and code of WBERC 2013, the variation of voltage should
75
