Building data scientist to help us dive deep into the very large amount structured time series data pertaining to building energy use

Información del proyecto

EnergySequence

Identificador del acuerdo de subvención: 739834

DOI

10.3030/739834

Proyecto cerrado

Fecha de la firma de la CE 16 Junio 2017

Fecha de inicio 1 Septiembre 2017

Fecha de finalización 31 Agosto 2018

Financiado con arreglo a

• INDUSTRIAL LEADERSHIP - Innovation In SMEs

Coste total

€ 68 000,00

Aportación de la UE

€ 68 000,00

Coordinado por

BETTERGY SL
Spain

Periodic Reporting for period 1 - EnergySequence (Building data scientist to help us dive deep into the very large amount structured time series data pertaining to building energy use)

Período documentado: 2017-09-01 hasta 2018-08-31

Resumen del contexto y de los objetivos generales del proyecto

Nearly 40% of final energy consumption and 36% of greenhouse gas emissions are produced by buildings. The actual process of improving building energy efficiency is time-consuming and cost intensive. Energy experts spend a lot of time walking around buildings, collecting information, and using outdated tools to identify and evaluate energy savings. The solution to improve the process of improving energy efficiency can be Virtual Energy Assessment and automated evaluation of building saving potential. Within the EnergySequence project, it was aimed to develop a web intelligence platform that creates a building energy model for each building to identify energy saving opportunities and be able to effectively evaluate opportunities in energy efficiency management.
The overall objectives can be described as follows:
• Perform holistic building energy audits remotely and identify the building saving potential based on building energy modelling and analysis
• Analyse the state-of-the-art in the context of energy disaggregation and building energy modelling based on existing building information
• Develop different building energy models for energy efficiency analysis using inference models, physical modelling, reverse engineering, predictive modelling, and statistical analysis for the assessment of building energy saving potential
• Develop an energy efficiency score (EScore) applicable to every building expressing the annual energy saving potential of commercial and non-commercial buildings.
• Validate the automated remote energy audit results versus standard detailed onsite energy audit

Specific objectives:
• Developing EnergySequence a web-based intelligence platform that will drive deeper building energy efficiency savings at 5-10x lower cost and faster speed from traditional approaches
• Being able to unlock energy savings at scale. The key factor to unlock energy savings at scale is being able to preliminary assess energy efficiency opportunities at low cost, that means remotely without going on site, targeting the buildings with the most energy saving potential, therefore maximizing energy savings and at the same time minimizing upfront investments.
• Open up the market for energy efficiency assessments at large scale helping large corporations, municipalities, retail chains, taking energy efficiency to their portfolios in a cost-effective way.
• Help utilities with customer segmentation accelerating engagement, and program participation leveraging the vast amount of energy consumption data they gather through smart meters for billing their customers
• Increase the technology readiness level (TRL) of EnergySequence. By means of the developments within the project, we aim to reach a TRL 8.

Trabajo realizado desde el comienzo del proyecto hasta el final del período abarcado por el informe y los principales resultados hasta la fecha

The WP1 covered the recruitment process and the initial training of the innovation associate. The outcomes of the first work package were the signed work contract of the innovation associate and the elaboration of the detailed training plan.

Within the WP2 the focus was on the analyse of the state-of-the-art in the context of energy disaggregation and building energy modelling based on existing building information such as energy bills, smart metering data, location and weather data. The main outcomes were the two state-of-the art surveys in the mentioned context a selection of potential approaches for specific use cases in the context of virtual building audits.

The WP3 covered the design and development of different building energy models for energy efficiency analysis and virtual building audits using predictive modelling, statistical analysis and physical modeling. The aim of the building energy model was the assessment of energy saving operational measures, like building energy demand, load base reduction, consumption reduction by energy end-use, peak shaving, load shifting, optimization of building start-up and shutdown, ghost loads, abnormal consumptions, optimization of the power hired, etc. The main outcome of WP3 is the building energy model, which allows us to provide and sell new intelligent services.

The WP4 focused on the design and development of an EScore applicable to every building expressing the annual energy saving potential buildings. The outcomes are a method to calculate the EScore for a certain building that supports the client to take informed decisions about the retrofitting and refurbishment actions.

The objective of WP5 was the design of advanced analytics that can be utilized for the virtual assessment of buildings. The outcomes of the WP5 are basic data analysis methods (e.g. data pre-processing) and more use case focused analytics such as load profiling and prediction, energy-end-use disaggregation and energy and cost saving analysis.

The general objective in WP6 was to validate the automated virtual energy audit (VEA) results versus an onsite energy audit. The basis for validation is a number of real onsite energy audits that have been carried out according to the European Energy Directive 27/2012 Standard by our energy experts. We have generated VEA for 700 installations and have compared them with a number of standardized onsite energy audits for validation.

Avances que van más allá del estado de la técnica e impacto potencial esperado (incluida la repercusión socioeconómica y las implicaciones sociales más amplias del proyecto hasta la fecha)

On the market there exist a number of web-based platforms for monitoring the energy consumption of the building. Through the outcomes of the project the EnergySequence platform (ES) is beyond the state of the art by adding intelligent service, which are based on data analysis and artificial intelligence. ES platform with the new developments has the following socio-economic impact:
• The scale up of energy efficiency by evaluating opportunities with simplified building energy models to achieve more savings in less time and at lower cost.
• Improve customer satisfaction, also increasing conversion rates and new service sales revenues.
• Targeting more facilities and conversion of more projects with remote advice, virtual energy analysis and continuous monitoring.
• Complete analysis of the energy consumption to improve the satisfaction of the final client, we reduce its cost and time up to 80%, improving the depth of the analysis x2
• Energy efficiency measures and renewable generation alternatives to improve customer satisfaction, generate new sources of revenue and reduce the cost of service with the customer
ES allows end customers to know energy saving measures and achieve greater energy efficiency with a cost and time that is 80% lower than manual methods and other monitoring platforms. It offers all that information and relevant values in the analysis of energy consumption of the company. The results are based on expert knowledge and artificial intelligence, which allow e.g. reduction of energy demand, optimization of energy contracts and adjustments of consumption habits and activity schedules.
ES offers also solutions to the utility of the future, it helps to achieve a change in the current model, modernizing utilities. The results are based on the knowledge of expert energy analysts, in addition to artificial intelligence, which can identify and implement recommendations to boost energy efficiency, minimize operating costs, and increase revenues. By means of ES the utilities can identify the target customers and increase the participation of customers with individualized messages that improve your satisfaction, offer energy efficiency recommendations that increase conversion and revenue.

EnergySequence - Energy Data Analytics

EnergySequence - Recommendation System

EnergySequence - Building Energy Model

EnergySequence - Overall concept