TRANSITION: naTuRal gAs efficieNcy inveStments Impact quanTificatION
DOMX is an innovative startup company established in Thessaloniki in 2019 and develops integrated software and hardware systems for cost-effective and universal energy management. DOMX IoT products integrate seamlessly with legacy building systems, delivering detailed monitoring, optimal management, improved energy efficiency and promoting long term energy behavior change through nudging interventions. Key business stakeholders (energy suppliers, maintenance companies, facility managers) get access to valuable energy data and demand management services offered on top of their customers’ assets, by enabling real-time monitoring and management at scale, for both the electricity and natural gas energy vectors. DomX members bring significant research expertise through their academic background and participation in EU research projects of the energy domain.
DOMX products and services target European Energy suppliers, which can benefit by improving their customer loyalty and satisfaction, while achieving their National Energy and Climate Plan (NECP) targets, thus leading to optimal implementation of energy efficiency investments in terms of costs and results. Through the TRANSITION project, DOMX aims at delivering a new service to energy suppliers for automatically quantifying the impact of energy efficiency investments for the two main energy vectors of electricity and natural gas. The new product will be packaged as a service and dashboard interface able to assist decision makers in monitoring their consumer portfolio performance in terms of various indicators, such as energy (savings, efficiency improvement), financial (cost reduction, ROI), environmental (reduced emissions), social (consumers reached), health (climate comfort) and others. The measured benefits can be automatically aggregated for supporting the EEOS of obligated parties, including governmental bodies and grid operators, beyond energy suppliers.
The existing EU building stock represents the main challenge for efficient energy use, considering that buildings account for 40% of the total energy consumption and almost 75% of them are energy inefficient. The EU policy response included the 2018 amended Energy Efficiency Directive (EED - Article 7) that forces EU countries to achieve new end-use annual energy savings of at least 0,8%, by establishing national energy efficiency obligation schemes (EEOS). The recent global and EU energy crisis have also led to a 3x increase in energy bills, especially for natural gas that accounts for 32.1% or residential EU consumption. The REPowerEU plan has also set out a series of measures to rapidly reduce the dependence on Russian fossil fuels and fast forward the green transition.
In this rapidly deteriorating context, energy suppliers & utilities, governmental bodies and other key energy stakeholders are actively investigating cost-effective measures for improving the energy efficiency and comfort in residential buildings and reducing the costs of end-use. In order to achieve the above targets, it is important for decision makers, e.g. Heads of Energy Management Departments, to be able to automatically quantify the evolution of important KPIs (technical, financial, social, environmental), during the planning, implementation and evaluation phases of energy efficiency investments for buildings.
The target KPIs for portfolio level performance need to include various types of indicators, such as energy (savings, efficiency improvement), financial (cost reduction, ROI), environmental (reduced emissions), social (consumers reached), health (climate comfort) and others. The utilization of big data techniques, powerful deep learning models and automation frameworks has the potential to properly handle the collected data and to perform the required complex simulations in a coordinated and efficient way, for calculating various KPIs under multiple different configurations in parallel.
An excellent contributor towards improving energy efficiency and reducing costs of end-use, is the integration of cost-effective home-IoT equipment with heavy consuming legacy building appliances. DOMX offers a cost-effective and universal retrofit IoT solution that integrates seamlessly with legacy heating systems, providing for optimized space heating and circumventing the need to install dedicated metering equipment for consumption monitoring. The DOMX heating controller builds on widely applicable protocols, to optimally manage the heating system’s operation, by considering the users’ comfort limits as set by the room thermostat, and local climate conditions as captured by device sensors. The DOMX smartphone application enables end consumers to understand how energy is consumed and to achieve long-term consumer behavior change through nudging interventions. The edge devices connect over Wi-Fi to the Internet and the company’s cloud energy management platform, enabling adaptation to building characteristics (envelope, boiler), climate variations (indoor, outdoor) and user schedules-preferences.
The DOMX heating controller is able to deliver up to 30% of savings for space heating when attached with legacy heating systems, without affecting the occupants' thermal comfort. A plurality of energy (system load, efficiency improvement, savings, etc.) and non-energy (temperature variations, climate comfort, etc.) can be collected in real-time. Key energy stakeholders (suppliers, facility managers, governmental bodies, etc.) get access to the data collected in real-time by their assets, belonging to consumers of their portfolios over secure APIs, for enabling real-time monitoring and management at scale.
DOMX has developed an ML-based framework enabled by trained LSTMs that produces the digital twin representation of connected heating systems (as synthetic time-series), emulating their operation under different configurations. The tool is particularly useful for comparing the performance of a considered set of configurations (heating system settings, building type, weather conditions, user comfort limits) that have real data associated with them, versus the digital twin model when employing different configurations. For instance, the tool is able to quantify the amount of energy savings that can be attained over a given period (e.g. 24h), by operating under energy efficient settings (lower outlet temperature) or different outdoor temperature conditions or different user consumption profiles (lower target room temperature), versus its baseline performance, in accordance with the IPMVP protocol. The LSTM models have been trained with data collected from >100 domestic buildings in the span of two previous winters, i.e., seasons 2020-21 and 2021-22.
At DOMX, we aim at exploiting the existing ML-based framework towards introducing a new service for decision makers of EEOS obligated parties (energy suppliers & utilities, governmental bodies). The envisioned solution will be able to quantify the impact of applying our energy efficiency upgrade solution for legacy heating systems to large consumer portfolios, under a wide range of settings (portfolio dimension, consumer profiles, weather conditions, building types, etc.). The overall aim is to package the service in a dashboard interface to assist decision makers in monitoring the evolution of the target KPIs, during the planning, implementation and assessment of energy efficiency investments.