Load forecasting analysis for estimating transformer capacity of Karangkates Substations using Holt-Winters method in Python

• Published in: International Journal of Power Electronics and Drive Systems
• Main Author: Rahmawati Y.; Kaki G.P.M.L.; Aripriharta; Sujito; Afandi A.N.; Wibawa A.P.; Purwatiningsih A.; Bagaskoro M.C.; Omar S.; Rosmin N.
• Format: Article
• Language: English
• Published: Institute of Advanced Engineering and Science 2024
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85208429209&doi=10.11591%2fijpeds.v15.i4.pp2222-2233&partnerID=40&md5=04c0506af9d5c6bb462de47269b0fd92

In the six years from 2010 to 2015, the peak load in the East Java region increased by an average of 284 MW per year. Karangkates Substation is part of an interconnected electrical system that supplies Java Island. To ensure a high level of reliability in its service, it is necessary to prepare for load growth to make sure that it does not exceed its ideal conditions, therefore special analysis of transformer capacity is needed. Using the Holt-Winters (HW) method as a reference for processing the data can be used as a reference in planning and anticipating the growing electricity demand. The results of this study are with the accuracy of the HW method with mean absolute percentage error (MAPE) = 2.645%, while the accuracy of the fuzzy time series (FTS) method = 6.399%. A forecast result done with HW methods shows the transformer at the substation Karangkates reached its normal working capacity in March 2018 at 99.583% of installed capacity and exceeded the maximum capacity in April 2018 at 101.493% of installed capacity. © 2024, Institute of Advanced Engineering and Science. All rights reserved.