Date
2021-03-19xmlui.dri2xhtml.METS-1.0.item-supplement
doi:10.17632/nwxv38dxsr.1Metadata
Show full item record
Preprint
Annual industrial and commercial heat load profiles: modeling based on k-Means clustering and regression analysis
Abstract
An accurate method to predict annual heat load profiles is fundamental to many studies, e.g., preliminary design or potential studies on renewable heating systems. This study presents a method to predict annual heat load profiles with a daily resolution for industry and commerce, based on an analysis of 800 natural gas load profiles (≥ 1.5 GWh/a). To derive heat load profiles, these natural gas load profiles are normalized and those with a potentially non-linear relationship between heat demand and natural gas consumption are excluded. The heat load profiles are clustered using the k-means algorithm according to their respective dependency on mean daily ambient temperature. The results reveal that the heat demand of most consumers is characterized by a clear dependency on mean daily ambient temperature, even in industry. The assignment of the load profiles to the clusters can be explained by the respective composition of each consumers’ heat sinks. In a regression analysis, individual regressions for each load profile are only slightly more accurate than the regressions for all load profiles assigned to one of the respective clusters. In terms of accuracy and user-friendliness, the developed cluster regression-based correlations for load profile prediction offer a significant improvement on previous methods.
Additional Information
This is the preprint version of the article "Annual industrial and commercial heat load profiles: modeling based on k-Means clustering and regression analysis". Date: 19/03/2021.Sponsorship
This work was supported by the German Federal Ministry for Economic Affairs and Energy within the framework of the 7th Energy Research Program [project: “AnanaS”, grant number 03ETW014A]. The authors would like to express their sincere thanks to the ministry for the financial support and to the involved utilities for providing natural gas load profiles.Citation
@article{doi:10.17170/kobra-202103233576,
author={Jesper, Mateo and Pag, Felix and Vajen, Klaus and Jordan, Ulrike},
title={Annual industrial and commercial heat load profiles: modeling based on k-Means clustering and regression analysis},
year={2021}
}
0500 Oax 0501 Text $btxt$2rdacontent 0502 Computermedien $bc$2rdacarrier 1100 2021$n2021 1500 1/eng 2050 ##0##http://hdl.handle.net/123456789/12700 3000 Jesper, Mateo 3010 Pag, Felix 3010 Vajen, Klaus 3010 Jordan, Ulrike 4000 Annual industrial and commercial heat load profiles: modeling based on k-Means clustering and regression analysis / Jesper, Mateo 4030 4060 Online-Ressource 4085 ##0##=u http://nbn-resolving.de/http://hdl.handle.net/123456789/12700=x R 4204 \$dPreprint 4170 5550 {{Thermische Belastung}} 5550 {{Lastprofil}} 5550 {{Korrelation}} 5550 {{Industrie}} 5550 {{Handel}} 5550 {{k-Means-Algorithmus}} 7136 ##0##http://hdl.handle.net/123456789/12700
2021-04-13T13:06:43Z 2021-04-13T13:06:43Z 2021-03-19 doi:10.17170/kobra-202103233576 http://hdl.handle.net/123456789/12700 This is the preprint version of the article "Annual industrial and commercial heat load profiles: modeling based on k-Means clustering and regression analysis". Date: 19/03/2021. This work was supported by the German Federal Ministry for Economic Affairs and Energy within the framework of the 7th Energy Research Program [project: “AnanaS”, grant number 03ETW014A]. The authors would like to express their sincere thanks to the ministry for the financial support and to the involved utilities for providing natural gas load profiles. eng Namensnennung - Weitergabe unter gleichen Bedingungen 4.0 International http://creativecommons.org/licenses/by-sa/4.0/ annual heat load profiles correlations industry commerce k-means clustering standard load profiles 620 Annual industrial and commercial heat load profiles: modeling based on k-Means clustering and regression analysis Preprint An accurate method to predict annual heat load profiles is fundamental to many studies, e.g., preliminary design or potential studies on renewable heating systems. This study presents a method to predict annual heat load profiles with a daily resolution for industry and commerce, based on an analysis of 800 natural gas load profiles (≥ 1.5 GWh/a). To derive heat load profiles, these natural gas load profiles are normalized and those with a potentially non-linear relationship between heat demand and natural gas consumption are excluded. The heat load profiles are clustered using the k-means algorithm according to their respective dependency on mean daily ambient temperature. The results reveal that the heat demand of most consumers is characterized by a clear dependency on mean daily ambient temperature, even in industry. The assignment of the load profiles to the clusters can be explained by the respective composition of each consumers’ heat sinks. In a regression analysis, individual regressions for each load profile are only slightly more accurate than the regressions for all load profiles assigned to one of the respective clusters. In terms of accuracy and user-friendliness, the developed cluster regression-based correlations for load profile prediction offer a significant improvement on previous methods. open access Jesper, Mateo Pag, Felix Vajen, Klaus Jordan, Ulrike 42 Seiten Kassel, Universität Kassel, Fachbereich Maschinenbau doi:10.17632/nwxv38dxsr.1 03ETW014A Thermische Belastung Lastprofil Korrelation Industrie Handel k-Means-Algorithmus submittedVersion false
The following license files are associated with this item: