Innovative Energy Programs’ (EPs) modeling and KPIs

To design a new EP, a dynamic pricing model and algorithm are required. Let us consider a system, which consists of a utility and its N clients/energy consumers. Without harm of generality, in the retail electricity market, the utility provides electricity to its clients in order to cover their demand. Thus, utility participates in wholesale electricity markets and purchases the required amount of energy at a certain cost, which is time-variant and also a non-linear function of the aggregated consumption of all N end users (i.e., each incremental energy unit purchased costs more). Generally, the utility can minimize the cost of the energy that it purchases in the wholesale electricity market (i.e., the system cost) by giving incentives to its end users to “harmonize” the aggregated ECC (i.e., the demand curve of its entire customer portfolio) with the wholesale market prices. Utilities and end users (consumers) can mutually benefit from this system’s cost reduction and the stability improvement that behavioral changes in the energy consumption can bring. Modern pricing schemes (or else EPs) should be able to trigger these behavioral changes (e.g., by motivating users to consume less during peak hours and more during non-peak hours). For example, in Real Time Pricing (RTP), prices are analogous to the dynamic ratio between the total energy production cost (i.e. supply) and the total amount of consumption (i.e. demand). A pricing scheme has to achieve an attractive trade-off among the following requirements (KPIs): i) the end user’s satisfaction, ii) the stability of the energy production/ transmission/consumption system, and iii) the utility’s financial profitability. The first requirement is also referred to as ‘user’s welfare’ and is formulated as the difference between a utility function that expresses how much an end user values a specific consumption pattern and the cost of energy that s/he consumes. In the context of comparing different pricing schemes, the user’s welfare expresses which pricing scheme leads to more competitive services in the open market. The second requirement is also denoted as ‘behavioral efficiency’ and expresses the capability of a pricing scheme to achieve the objectives that motivated it in the first place (e.g. load curtailments and shifts). Intuitively, behavioral efficiency of a pricing scheme expresses how friendly it is to a TSO/DSO (addressing issues related to energy network stability, efficiency and costs) and implicitly affects several financial metrics (e.g. investments in RES, energy storage and network upgrades). Usually, it is linked with minimizing the system’s energy cost, The third requirement is also referred to as ‘profit dynamics’ and represents the profit percentage per energy unit and the total revenues of the utility company. In other words, it expresses the financial growth potential of the company that exploits a specific pricing scheme (or else EP).

A wide range of innovative EPs are integrated in SOCIALENERGY platform. In particular, SOCIALENERGY conducts research on the improvement of the behavioral efficiency of the EPs without sacrificing the rest of the aforementioned KPIs. For example, as shown in the figure above, a behavioral change in the aggregated ECC can provide reduced energy cost for the system without sacrificing users’ welfare due to the fact that some of them are flexible enough to undertake the changes in their individual ECCs and in return get reimbursed by the utility. Through SOCIALENERGY platform, the administrative user can perform exhaustive system-level simulations before deciding to release a new EP in the retail market. Similarly, an end user can also exploit SOCIALENERGY platform to dynamically invest (if it is beneficial for her/him) on a new EP that fits his/her updated needs. Finally, an end user can also play the SOCIALENERGY GAME in order to comprehend the optimal behavior that one should have towards harvesting the maximum benefits from a certain EP.

Management of multi-parametric virtual energy communities (VECs)

In SOCIALENERGY system, VECs can be created in a bottom-up (and thus manual) way from the users themselves just like in traditional social network platforms. A VEC leader may also be the one that initiates and coordinates the process just like in web forums and other web 2.0 tools. However, VECs can also be created and dynamically adapted in an automated way via the use of clustering algorithms in order for both users and the utility to optimally exploit the benefits of VEC concept. In particular, a utility’s portfolio can be categorized in several VECs based on qualitative characteristics such as demographics, geographical, socio-economic and other social norms-based metrics Given an already existing social graph, the goal of a clustering algorithm may also be to find such VECs that the total power consumption in each group of users achieves minimum variance VECs can also be created in a way that users’ satisfaction, social network dynamics and the peer pressure that VEC members induce to each other are taken into consideration Other algorithms take into account quantitative metrics for VEC creation problem. For example, the dominant VEC creation criterion can be the similarity factor of Energy Consumption Curves (ECCs) and/or the Flexibility Curves (FCs) of the users. In other words, users with similar ECCs and FCs increase the probability of performing better in a community-based EP. Another criterion would be to put together users that have the minimum deviation between their forecast and real consumption in order to minimize the imbalance penalties of a utility’s portfolio. Finally, for billing purposes, there are also intra-clustering algorithms, which can allocate the costs among the members of a certain VEC by applying various policies according to each end user’s performance and behavioral change.

Multi-parametric VEC creation and dynamic adaptation

All the above-mentioned multi-parametric approaches for VECs’ creation can be easily customized and integrated in SOCIALENERGY platform. A few clustering examples that we currently use in the SOCIALENERGY platform are illustrated in the figure above. What’s more interesting is that the administrative user can set specific thresholds based on which an end consumer can be recommended to switch to a different VEC that better fits his/her updated interests and needs. Users can also play the multi-player mode of the GAME in order to be seamlessly educated about the potential benefits and operation of community-based EPs.

Context-aware data analytics services (reporting/recommendation)

 The goal of SOCIALENERGY platform is to use the recent innovative concepts on e-commerce in order to trigger and possibly facilitate e-governance and consequently social innovation related activities in the future. There are five promising innovation fields towards this goal. The first is the exploitation of information beyond the e-commerce retailer’s site towards personalized and accurate recommendations for products and services. Secondly, SOCIALENERGY envisages cross-domain e-commerce hyper personalized services that will offer a great opportunity to retailers to dispose their products/services beyond their company’s site in a targeted/efficient and non-intrusive manner. Thirdly, SOCIALENERGY targets the increase of e-commerce transactions through automatic and intelligent product/service assortment recommendation services for portfolio extensions that will be highly beneficial for e-commerce retailers. Towards this goal, there will be used information from: i) social network relationships’ and activities’ analysis, ii) communities in GSRN, and iii) LCMS. Fourthly, an innovative service that SOCIALENERGY envisages to offer through GSRN is to advance existing e-commerce paradigm through collective e-commerce services and bottom-up collaborative crowd-funding services.