The Public Power Corporation (PPC) is the biggest energy producer, provider and operator in Greece and one of the largest in Europe, with activities in the generation and distribution of electricity to consumers and it administers numerous energy-related critical infrastructures. Within the core policies of the Corporation lie the reduction of carbon footprint (by decommissioning of lignite mines and power plants and by increasing the renewable energy production), the promotion of green policies and the raise in energy efficiency. Towards the latter, PPC has been engaged in studying consumers’ behavior in energy consumption, exploiting a large database of about 4,000,000 customers. With a dynamic entry into new business areas, the need for promoting digital transformation and new investments strengthening the green portfolio, the consumer is at the heart of the Corporation’s activities.
Horizon 2020 EVIDENT project
PPC participates in the Horizon 2020 project EVIDENT. The European funded project entitled as “bEhaVioral Insights anD Effective eNergy policy acTions”, shortly EVIDENT, aims at providing new insights in the energy efficiency change policy innervations. Through field trials, surveys and games, the project focuses on behavioural biases and heuristics that affect decision making in energy consumption. The results of the analysis will be used to evaluate and propose energy efficiency policy measures that will reduce energy consumption and boost energy efficient technology diffusion. Understanding how individuals make decisions is important for researchers and policy makers concerned with the impact of human behaviour on energy use and the environment. The main challenge is to translate behavioural science insights into scaled interventions for having large environmental and economic returns as it is schematically presented in Figure 1. That is, to answer the question how behavioural interventions can become a more integral component of climate change policy. For accomplishing that, EVIDENT proposes a three-fold strategy: very large participation of end users in the designed experiments, games and surveys, novel analytical tools to accommodate well defined shortcomings in existing methodologies and an integrated platform including methods, datasets, tools and reports for promoting energy efficiency.
The participation of PPC in a such an ambitious project includes several activities: At first, PPC provides consumption data to be analyzed with advanced computational methodologies and machine learning algorithms developed within the project. The target of 100,000 participants set by project’s surveys are only a small fraction of the total database size. It is important to mention that PPC provides data considering and respecting all the GDPR terms for customer personal data protection. Secondly, PPC will adopt the data analysis conducted within the project towards increasing energy efficiency. Finally, PPC contributes to the exploitation and dissemination plan of the project output through relevant information activities and the energy market analysis.
Both the real-time and the past energy consumption data provide the opportunity for a thorough analysis of the initial financial literacy aspects in conjunction with energy consumption behavior. Analysis of these insights is necessary to craft policies that effectively and efficiently achieve the targeted goals.
The main goal is to provide the corresponding data structures to support the multi-domain exchange of information and at the same time, deliver the necessary data services for the analytical scope of the project’s tasks. This includes flexible architecture for scale-in/scale-out infrastructure, suitable for large-scale data collection, storage and processing. Data from prior relevant surveys will also be collected for enhancing the empirical result of the project. The data gathered and provided by PPC to the project undergo a careful and considerate data analysis by their grouping into appropriate classifications such as residential, commercial, and industrial/ manufacturing categories.
An independent PPC consumer-oriented tool for data gathering towards the EVIDENT goals is the “myEnergyCoach” platform which has been developed and incorporated on the electronic customer service of the Company.
The “myEnergyCoach” platform has been designed to gather information regarding the appliances of a household and their use by the customers of PPC. It is a digital and user-friendly platform which consults customers about their consumption within their household and informs them on the energy distribution between the appliances. The feedback of the platform includes personalized tips and suggestions to save energy, as well as advice on potential economic benefits by replacing old appliances with modern, energy-efficient ones. The independent development of the platform is an important tool for the EVIDENT project goals towards consumption monitoring and evaluation, real customer incentives and energy information metrics, since it can collect and provide customers data in a more efficient way.
For the EVIDENT project, the understanding of energy consumption patterns in the residential sector is of paramount importance for the design of new energy management strategies and policies which are based on novel data analyses and communication technologies. In this perspective, smart energy platforms provide considerable opportunities and allows for the assessment of household characteristics, behaviors and routines that drive household electricity loads. However, the handling of large data requires advanced data analytics methodologies and pattern detection algorithms.
EU policies for energy savings and eco-design
→ Energy savings
Eco-design sets minimum standards through the EU to eliminate the least performing products from the market. The energy labels provide a clear and simple indication of the energy efficiency and other key features of products at the point of purchase. This makes it easier for consumers to reduce their household energy consumption with direct economic benefits for them and thus, contributing to the reduction of greenhouse gas emissions across EU.
The EU legislation for energy labels and eco-design has been estimated to bring energy savings of approximately 230 million tons of oil equivalent (Mtoe) by 2030 . For consumers, this means significant savings per year on their household energy bills. Moreover, energy efficiency measures will create extra revenue for European companies.
→ Energy labels
The EU energy label functions as a key to encourage consumers to choose products which are more energy efficient. Also, it motivates manufacturers to drive innovation by using more energy efficient technologies. Manufacturers are keen to see their energy-labelled products in the highest available category when compared to competitors. Therefore, it is likely that manufacturers who sell appliances in the less efficient classes will aim to improve their rating, in order to position their products within the highest category. As a representative example, roughly two-thirds of refrigerators and washing machines sold in 2006 were labeled as class A, whereas over 90% of those sold in 2017 were labeled A+, A++ or A+++ .
As a result of the development of more and more energy efficient products and because the difference between A++ and A+++ is less obvious to the consumer, the EU energy label categories will be gradually adjusted to reintroduce the simpler A to G scale. For example, a product showing an A+++ energy efficiency class could become a class B or lower after rescaling, without any change in its energy consumption. Class A will initially be empty to leave room for more energy efficient products to be developed. This will enable consumers to distinguish more clearly between the most energy efficient products. At the same time, it is meant to encourage manufacturers to continue research and innovation into more energy efficient technologies.
There is a worldwide demand for more efficient products to reduce energy consumption and other natural resources in line with improving overall sustainability. The EU legislation on eco-design is an effective tool for improving the environmental performance of products by setting mandatory minimum standards for their energy efficiency. As a consequence, the least performing products are eliminated from the market and this step significantly contributes to the EU’s energy and climate targets. Eco-design also supports industrial competitiveness and innovation by promoting better environmental performance of products throughout the internal market.
An important addition to the overall eco-design rules, is the inclusion of elements to further improve the reparability and recyclability of household appliances. Several of the new measures include requirements, such as making spare parts more easily replaceable, and ensuring that key parts and repair and maintenance information are available for end-users and professional repairers as appropriate, for a minimum duration of 7-10 years depending on the product.
PPC implements EU legislation and policies
PPC, as the biggest energy producer and provider in Greece, has its own strategies for adopting and respecting the abovementioned EU legislation and policies, such as the developed platform “myEnergyCoach” which is aligned with EU policies. The general direction is to promote energy efficient strategies through consumption monitoring and evaluation. It compares household appliances and distinguishes between the eco and non-eco-friendly ones. Additionally, it enables consumers to set an energy-saving goal and provides personalized tips towards this goal. This is an incentive for any household to be on track with EU legislation.
However, through the participation in the EVIDENT project, PPC benefits a series of advantages in this direction. In EVIDENT, new policy strategies are designed by experienced researchers which will provide a competitive advantage to Policy Makers by offering them an easy way to design and evaluate new policies. Advanced features developed within EVIDENT such as big data analytics and visualization can help policy makers to design more creative and better targeted interventions. Under this framework, trial-and-error setting of policy interventions can be avoided and thus, reduce its cost while increasing its effectiveness. Moreover, previous results, data as well as policy measures stored in EVIDENT platform can be used by policy makers as a reference/best practices in order to design new and more effective policies.
Energy consumption data analysis
→ Data and methods
Energy analytics is the process of gathering data, using software, with the purpose of analysis, supervision, and optimization of energy related KPIs (eg. production costs, consumption or production distribution). An efficient energy analytics software harnesses the nuances of the data produced within the various energy sectors and employs it more beneficially. PPC provides an energy dashboard, which can improve profit margins as well as manipulate and understand large-scale trends in the industry. Applying data analytics to the energy sector provides a deeper insight across all the dimensions and in particular, in energy savings. To get specifics regarding the energy usage, it is potentially required only one tool: an integrated platform with electricity usage or energy consumption monitor that provides precise information about the amount of kWh of a device or home appliance. This integrated tool is the ultimate goal of EVIDENT since it is vital in simulating and visualizing consumer behavior.
Since residential electricity consumption has strong temporal variation, which is not captured with low resolution consumption data such as monthly bills, long-term consumption data are required. By making use of this data, the surveys are linked to the smart meters databases through encoded household meter numbers, for confidentiality reasons and for respecting the GDPR. Combining these two sets of information allows an extensive and coherent big data analysis. By evaluating the survey results for the dwellings in each cluster, it is possible to identify important similarities and differences regarding socio economic determinants, dwellings characteristics and appliances use and ownership, that could explain the different clusters’ aggregation and consumption profiles.
→ Classified energy consumption data analysis
The very common classifications include residential, commercial, and industrial/manufacturing consumption data analysis.
Residential data analysis: involves household characteristics, home energy use and costs, and detailed household data processing. Housing characteristics data collected should include fuels used and end-uses in structural and geographic characteristics, such as electronics, lighting, heating and cooling, household unit, demographics, energy insecurity, as it shown in Figure 2. For primary housing units, energy consumption and expenditures can be calculated with end-uses and fuel consumption. In residential energy consumption investigation, energy-related random data are collected.
Commercial data analysis: involves building characteristics, commercial energy use and costs, as well as detailed buildings’ data processing. Detailed tabulated data may be applied and may contain characteristics of commercial buildings, such as buildings size, age, and their usage, energy sources used in the buildings, even geographic location, and other energy-consuming devices and units, as it is shown in Figure 2.
Manufacturing data analysis: involves the characteristics of consumption of energy for all purposes by an industry consisting of systems, buildings, devices, and other energy consumers.
There are many relevant studies by exploring the literature online:
- Suggested methods for analyzing various building energy use cases by user groups . Recent market, technology and policy drivers have resulted in widespread data collection by stakeholders across the buildings industry. Consolidation of independently collected and maintained datasets presents a cost-effective opportunity to build a database of unprecedented size. Applications of the data include peer group analysis to evaluate building performance, and data-driven algorithms that use empirical data to estimate energy savings associated with building retrofits.
- Research suggesting how to predict building energy requirements and energy supply costs by using sensitivity analysis . Detailed consumption data can be used to offer advanced services and provide new business opportunities to all participants in the energy supply chain. The survey proposes an innovative tool that combines the estimation of the building energy consumption and supply costs with a sensitivity analysis.
- Guidelines on measuring energy saving needs by building energy management projects . They include minimum compliance requirements to insure a fair level of confidence in the savings determination. These requirements are set forth in three specific approaches and include compliance paths for each approach. The approaches include: 1) Whole-building metering, 2) Retrofit isolation metering, and 3) Whole-building calibrated simulation.
- Research on building energy data analysis has been conducted on prediction methods with accuracy guaranteed by applying a model that automatically analyses and predicts schedule data, which applies input data when predicting energy consumption .
The majority of the surveys apply data analytics and clustering analysis to energy consumption data towards energy efficiency, which correlates the nature of the surveys with the EVIDENT’s long-term goals.
 Available at: https://www.dei.gr/en/home/myenergy/myenergycoach/
 Available at: https://www.egalite.org/ecodesign-ed-etichettatura-energetica-leuropa-e-gia-in-ritardo/
 Available at: https://www.eumonitor.eu/9353000/1/j9vvik7m1c3gyxp/vlgqhguavfv8?ctx=vg9pi5ooqcz3&tab=1
 Mathew P., Dunn L., Sohn M., Mercado A., Custudio C., Walter T. (2014). Big-data for building energy performance: Lessons from assembling a very large national database of building energy use, 140,85-93.
 Gruber J., Prodanovic M., Alonso R. (2015). Estimation and analysis of building energy demand and supply costs. Energy Procedia, 83, 216–225.
 ASHRAE. (2002). ASHRAE guideline 14-2002 (1st ed.). ASHRAE.
 Ahn Y., Lee Y., Oh E., Kim B. (2020). Building energy demand prediction model using Fourier transform based schedule analysis algorithm. Korean Journal of Air-Conditioning and Refrigeration Engineering, 32, 386–397.