Search eceee proceedings

Energy and data - how data analysis can contribute to reducing consumption at the scale of the SNCF

Panel: 2. Energy management and corporate culture

This is a non-peer reviewed extended abstract. No paper available.

Author:
Jack Suddaby, AREP, France Antoine Hubert, AREP, France

Abstract

Train stations are important public infrastructures that consume large amounts of energy to provide comfort to passengers. In France, train stations, run by SNCF Gares et Connexions vary in consumption from 100 MWh per year, a large city station, to over 10 000 MWh for an international station. Train stations have unique processes, such as lighting in pedestrian circulations and elevator equipment, such as lifts and escalator. These types of energy use are not covered by standard building practice. Up until recently in France these stations were rarely required to limit their energy consumption as they were not covered by the laws typically applied to offices and housing. Recent SNCF internal policy and a new French law (Décret Tertiaire), motivated by both the energy crisis and greenhouse gas emissions, have introduced targets to reduce the consumption of train stations. Therefore, Energy Management of stations has become an important theme to reduce their consumption.

AREP, a subsidiary of SNCF Gares et Connexions, has been working closely with energy managers for the group to implement new types of analysis and procedures. The strategy adopted is divided into three main areas of study:

– Identification – understanding how energy us being used in the station for: lighting, heating, cooling or other, and defining different typologies of station to hierarchise interventions.

– Detect and improve – analyse the current performance of the station and propose improvements that will reduce energy consumption. These improvements are oriented around “quick-wins” that involve minimal interventions at low cost whilst providing a reduction in energy use, such as lighting astronomical timers and out-of-hours extinction.

– Monitoring – after these interventions and improvements are implemented in stations it is important to monitor if and by how much the energy consumption was reduced to give feedback to energy managers.

In this paper we introduce several examples from the different stages of this strategy that have contributed to significant energy savings in France. These analyses have been performed on electricity consumption curves that give the energy consumption in kWh for 10 minutes intervals for every electricity meter of the SNCF:

– To identify the different types of energy use of train stations the hourly energy use of stations was compared to the local weather conditions for that given hour, including air temperature in °C and the incident solar radiation in W/m², as well as the opening hours of the station. Using an algorithm to correlate the different sources of data it was possible to decompose the energy consumption into that linked to heating and cooling, linked to the air temperature, and lighting, linked to the presence of the sun and the incident solar radiation. The opening times also allowed for an understanding of how efficiently different equipment was programmed to run in the station. The energy consumption of over 300 stations was analysed using this method to demonstrate that for an annual energy consumption of 200 GWh, 47% was linked to heating demand, 3% linked to cooling, 1% to lighting and 49% to other.

– An important tool in the understanding of energy consumption is the visualisation of data. In the “Detect and Improve” phase a novel visualisation technique for the annual hourly energy consumption of stations allowed the lighting linked to platforms to be easily identified visually. Furthermore, it was possible to compare different lighting profiles to understand when platform lighting was not reduced or shut down at night. An analysis of over 230 stations allowed several malfunctioning systems to be corrected with a potential annual energy saving of over 2000MWh, representing roughly 130tCO2eq.

– Finally, after the intervention has been performed the energy consumption is reanalysed to verify that the measure was well installed and to quantify the real energy gain of the action.

Downloads

Download this presentation as pdf: 2-021-23_Suddaby_Pres.pdf