Institutional Repository
Technical University of Crete
Technical University of Crete
EN
|
EL
| URI: | http://purl.tuc.gr/dl/dias/E21479BC-1189-4BC8-993B-D526C1E071C4 | ||
| Year | 2021 | ||
| Type of Item | Diploma Work | ||
| License |
|
||
| Bibliographic Citation | Konstantinos Kouletakis, "Energy profile forecast of the Souda port for the year 2030", Diploma Work, School of Chemical and Environmental Engineering, Technical University of Crete, Chania, Greece, 2021 https://doi.org/10.26233/heallink.tuc.89674 | ||
| Appears in Collections |
Ports are characterized by the geographical concentration of activities of high energy demand due to their proximity to metropolitan areas, as well as their function as central hubs for the transportation of raw materials. In recent decades, the need for a better understanding and monitoring of energy-related activities, which takes place near or in port, has become more apparent as a result of the growing relevance of energy transactions, public awareness of the environment and focus on industrial energy efficiency. Aiming at a long-term development strategy, estimating future energy demand is essential which is implemented through forecasting models.In the present dissertation, the project of creating a state-of-the-art model for forecasting the electricity consumption profile that the port of Souda is expected to have for the year 2030, was studied. To fulfill this purpose, 11 forecast models were created, using real port energy consumption data from the five-year period 2015-2019 as well as meteorological data of the region, divided into three main categories - hourly, daily and monthly. Simple Linear Regression, Decomposition Method, Box-Jenkins ARIMA method, as well as Machine Learning algorithms and Artificial Neural Networks, were used as methods.From the conduct of the study and the analysis of the results, it was initially revealed that the main factor of electrical demand is the lighting equipment of the port area, with higher consumption during the night and the winter period. Then, the best model for predicting energy consumption, according to the mean square error MSE (4.86 kWh2), was that of the Exponential GPR machine learning algorithm, based on hourly data, which estimates annual energy consumption at 1,027,649.94 kWh, an increase of about 54% compared to 2019. At the same time, all forecast models estimate electricity consumption of over 1,000,000 kWh, which requires immediate action now to address this demand and highlights the need for further research in the field of creating reliable forecasting models.
Copyright © DIAS 2013
