Failure rate estimation for transformer population based on health index through Markov model approach

This work examines the failure rate of the transformer population through the application of the Markov Model (MM) and Health Index (HI). Overall, the condition parameters data extracted from 3,192 oil samples were analysed in this study. The samples were from 370 transformers with the age range bet...

Full description

Bibliographic Details
Main Authors: Shariffuddin, Nor Shafiqin, Azis, Norhafiz, Mohd Selva, Amran, Yahaya, Muhammad Sharil, Jasni, Jasronita, Ab Kadir, Mohd Zainal Abidin, Talib, Mohd Aizam
Format: Article
Language:English
Published: Universiti Putra Malaysia Press 2021
Online Access:http://psasir.upm.edu.my/id/eprint/98151/
http://psasir.upm.edu.my/id/eprint/98151/1/42%20JST-2691-2021.pdf
_version_ 1848862791872544768
author Shariffuddin, Nor Shafiqin
Azis, Norhafiz
Mohd Selva, Amran
Yahaya, Muhammad Sharil
Jasni, Jasronita
Ab Kadir, Mohd Zainal Abidin
Talib, Mohd Aizam
author_facet Shariffuddin, Nor Shafiqin
Azis, Norhafiz
Mohd Selva, Amran
Yahaya, Muhammad Sharil
Jasni, Jasronita
Ab Kadir, Mohd Zainal Abidin
Talib, Mohd Aizam
author_sort Shariffuddin, Nor Shafiqin
building UPM Institutional Repository
collection Online Access
description This work examines the failure rate of the transformer population through the application of the Markov Model (MM) and Health Index (HI). Overall, the condition parameters data extracted from 3,192 oil samples were analysed in this study. The samples were from 370 transformers with the age range between 1 and 25 years. First, both HIs and failure rates of transformers were determined based on the condition parameters data of the oil samples known as Oil Quality Analysis (OQA), Dissolved Gas Analysis (DGA), Furanic Compounds Analysis (FCA) and age. A two-parameter exponential function model was applied to represent the relationship between the HI and failure rate. Once the failure rate state was obtained, the non-linear optimisation was used to determine the transition probability for each age band. Next, the future failure rate of the transformer population was computed through the MM prediction model. The goodness-of-fit test and Mean Absolute Percentage Error (MAPE) were utilised to determine the performance of the predicted failure rate. The current study reveals that the future state of the transformer population and failure rate could be predicted through MM based on updated transition probabilities. It is observed that the MAPE between predicted and computed failure rates is 7.3%.
first_indexed 2025-11-15T13:22:39Z
format Article
id upm-98151
institution Universiti Putra Malaysia
institution_category Local University
language English
last_indexed 2025-11-15T13:22:39Z
publishDate 2021
publisher Universiti Putra Malaysia Press
recordtype eprints
repository_type Digital Repository
spelling upm-981512022-08-13T00:33:10Z http://psasir.upm.edu.my/id/eprint/98151/ Failure rate estimation for transformer population based on health index through Markov model approach Shariffuddin, Nor Shafiqin Azis, Norhafiz Mohd Selva, Amran Yahaya, Muhammad Sharil Jasni, Jasronita Ab Kadir, Mohd Zainal Abidin Talib, Mohd Aizam This work examines the failure rate of the transformer population through the application of the Markov Model (MM) and Health Index (HI). Overall, the condition parameters data extracted from 3,192 oil samples were analysed in this study. The samples were from 370 transformers with the age range between 1 and 25 years. First, both HIs and failure rates of transformers were determined based on the condition parameters data of the oil samples known as Oil Quality Analysis (OQA), Dissolved Gas Analysis (DGA), Furanic Compounds Analysis (FCA) and age. A two-parameter exponential function model was applied to represent the relationship between the HI and failure rate. Once the failure rate state was obtained, the non-linear optimisation was used to determine the transition probability for each age band. Next, the future failure rate of the transformer population was computed through the MM prediction model. The goodness-of-fit test and Mean Absolute Percentage Error (MAPE) were utilised to determine the performance of the predicted failure rate. The current study reveals that the future state of the transformer population and failure rate could be predicted through MM based on updated transition probabilities. It is observed that the MAPE between predicted and computed failure rates is 7.3%. Universiti Putra Malaysia Press 2021-10-28 Article PeerReviewed text en http://psasir.upm.edu.my/id/eprint/98151/1/42%20JST-2691-2021.pdf Shariffuddin, Nor Shafiqin and Azis, Norhafiz and Mohd Selva, Amran and Yahaya, Muhammad Sharil and Jasni, Jasronita and Ab Kadir, Mohd Zainal Abidin and Talib, Mohd Aizam (2021) Failure rate estimation for transformer population based on health index through Markov model approach. Pertanika Journal of Science & Technology, 29 (4). pp. 3029-3042. ISSN 0128-7680; EISSN: 2231-8526 http://www.pertanika.upm.edu.my/pjst/browse/regular-issue?article=JST-2691-2021 10.47836/pjst.29.4.42
spellingShingle Shariffuddin, Nor Shafiqin
Azis, Norhafiz
Mohd Selva, Amran
Yahaya, Muhammad Sharil
Jasni, Jasronita
Ab Kadir, Mohd Zainal Abidin
Talib, Mohd Aizam
Failure rate estimation for transformer population based on health index through Markov model approach
title Failure rate estimation for transformer population based on health index through Markov model approach
title_full Failure rate estimation for transformer population based on health index through Markov model approach
title_fullStr Failure rate estimation for transformer population based on health index through Markov model approach
title_full_unstemmed Failure rate estimation for transformer population based on health index through Markov model approach
title_short Failure rate estimation for transformer population based on health index through Markov model approach
title_sort failure rate estimation for transformer population based on health index through markov model approach
url http://psasir.upm.edu.my/id/eprint/98151/
http://psasir.upm.edu.my/id/eprint/98151/
http://psasir.upm.edu.my/id/eprint/98151/
http://psasir.upm.edu.my/id/eprint/98151/1/42%20JST-2691-2021.pdf