Blockchain-based orchestration of distributed assets in electrical power systems

Our society depends on a reliable power supply. Therefore, the electrical power system (EPS) is traditionally planned and operated hierarchically and centrally. However, this does not reflect the increasingly decentralized and volatile energy landscape. The rising number of highly heterogeneous actors in the EPS results in an increased total complexity of the system. Many of the newly installed distributed assets generating renewable electricity are owned by citizens or small companies and are thus not controlled centrally.

A bottom-up approach like the cellular concept (Benz et al., 2015) is able to face the increasing complexity in the system. According to the cellular concept, locally limited areas consisting of electrical producers, consumers and storages form an energy cell. Many of these cells form a larger cell, in which the individual cells organize themselves and balancing is done at the lowest possible level. In this context many recent studies and publications attribute a high potential to peer-to-peer (P2P) approaches based on distributed ledger technologies (DLTs) (BDEW Bundesverband der Energie und Wasserwirtschaft e.V., 2017; Dütsch & Steinecke, 2017; Merz, 2016). Therefore, we investigate if and how an orchestration of assets in the EPS is possible according to the cellular concept and based on DLTs, such as blockchains (Nakamoto, 2008). This extended abstract presents the current stage of my according PhD project including the definition of the research question, related work, the proposed methodology, first results and a conclusion.

The earliest and probably best-known blockchain-based local energy market (LEM) project is The Brooklyn Microgrid (Mengelkamp et al., 2018a). This pilot was followed by research projects like the LAMP project (Mengelkamp et al., 2018b). The new project Pebbles aims for an integration of grid friendliness in LEMs (Allgäuer Überlandwerk GmbH, 2018). The Enerchain project tackles large-scale electricity trading in Europe (Merz, 2016). The Energy Web Foundation aims at accelerating blockchain technology across the energy sector. Therefore, they build upon Ethereum and develop a downward compatible blockchain suitable for energy related applications (Energy Web Foundation, 2018). The Token NRGcoin (Mihaylov et al., 2014) enables decentralized electricity trading as well but in addition allows for demand response. Equally, Power Ledger (Power Ledger, 2018) provides a decentralized platform for P2P electricity trading.

In terms of evaluation metrics considering blockchain technology a good starting point is offered in (Wüst & Gervais, 2017). Furthermore, a performance analysis for private blockchains is presented in (Pongnumkul et al., 2017) and first benchmarks considering private blockchains are described in (Dinh et al., 2018). They conclude that there is still a big performance gap between blockchains and current databases.

This PhD-project bases upon preliminary research concerning simulation of and algorithms for an efficient and decentralized provision of ancillary services by conventional virtual power plants (VPPs) with a central information technology (IT) infrastructure (Schlund & German, 2017; Schlund et al., 2017; Schlund et al., 2018a). The project intends to analyze existing and think of new use cases, to develop an unbiased evaluation framework for such use cases, to identify the most suitable blockchain designs and to quantify the extra effort of a blockchain-based approach in comparison to centralized approaches for the most promising use cases. Therefore, it complements and fits in well in the overall picture.

The general idea of the PhD-project is to simulate the use cases of interest with different blockchain networks in order to evaluate them according to a defined evaluation framework and to use prototypical implementations for the validation of the simulation models. Therefore, the proposed methodology consists of a theoretical and an empirical part. The theoretical part includes literature reviews, an analysis of current research developments and analytical work. In the quantitative part, prototypical implementations and experimental test runs are used for data collection and extensive simulation and modeling is used for studying the use cases. In order to mind both IT and EPS layer probably hybrid simulation including discrete and continuous simulation will be used. An overview of the proposed methods for answering the RQs is visualized in Fig. 1.

In this section RQ1 is partially answered and first ideas considering RQ2 are summarized.

Considering RQ1, use cases of interest have been identified as visualized in Fig. 2. The use cases are clustered into three subgroups presented successively in the following.

The first subgroup is P2P electricity trading. A central trusted third party (TTP) contradicts the idea of P2P trading and obviously storing state is necessary for being able to trade. In addition, a trading partner might be unknown and therefore possibly untrusted. This is the area in which most research and development is done at the moment and in which recent studies see a high potential, mainly because local trade is currently not possible due to the complexity in current electricity trading processes. Specifying, the use cases of interest are electricity trading on different levels, each with different requirements. The most investigated use case at the moment is LEMs (or communities), in which P2P trading (or sharing) is facilitated between end consumers or producers in a locally limited area. The same concept is also applicable on other levels, from device level up to distribution system operator (DSO) or transmission system operator (TSO) level. Besides trading the electricity, trading of certificates of origin or CO₂ emissions is also possible. This is mainly motivated by the properties of blockchains considering transparency and immutability.

The second subgroup is the provision of flexibility as an ancillary service for the EPS. In comparison to the market-based approaches summarized under P2P electricity trading, this approach faces the need for flexibility in the EPS in a more straightforward way. So far, such flexibility is either provided by central power plants, which will become less in numbers in future, or by centrally controlled VPPs. A new blockchain-based approach for such services is a decentralized virtual power plant (DVPP), which is a self-organized VPP and thus without the need for a central operator. This way there are no costs for an always-online TTP and the participants can profit from the full economic benefit. The DVPP has thus the potential of organizing and mobilizing small-scale distributed flexibility sources, which are not part of a VPP and do not have access to a flexibility market so far. In analogy to a VPP, possible applications are frequency or voltage stability or congestion management. Such a DVPP is currently implemented in our lab and a pre-stage of is published in (Schlund et al., 2018b).

The last identified subgroup is battery electric vehicle (BEV) charging. This is chosen as it is expected that the share of BEVs will increase drastically in the next years and because uncoordinated charging might cause problems in distribution grids. A TTP is undesirable as there are many different charging station operators and energy providers involved. As of today, problems already exists considering the interoperability between different providers.

In total, the cost of disintermediation is valuated by restraints in performance, additional holistic costs and limitations in scalability. In addition to the quantitative evaluation, the subordinate qualitative metrics are the security of the EPS, data integrity, transparency and privacy. These are the first ideas considering the evaluation framework. However, the metrics need to be formally defined to be able to quantify and evaluate them at a later stage of the PhD-project. This formal definition is the next step in the PhD-project.

This extended abstract describes the current stage of my PhD-project. The aim is to provide a foundation for deciding how to orchestrate distributed assets in a future electrical power system empowered by renewable energies and storages under the hypothesis that a central control is not desirable. Specifically, the main research question is to investigate how large the necessary extra effort of distributed approaches based on blockchain technology is compared to a centralized approach from a holistic point of view. An overview of the related work is given and this project is placed in the research landscape. Planned and partly already applied methodologies are literature reviews, lab setups and hybrid simulation.

The use cases of interest are identified as different P2P electricity markets, distributed flexibility provision, BEV charging and its coordination. With a DVPP, a concept for self-organized distributed flexibility provision is proposed. However, the DVPP implementation still needs to be formally described, refined and analyzed more in detail. Besides that, a second lab setup of a peer-to-peer electricity market with a focus on grid-friendly operation is currently in work. First ideas for an evaluation framework are provided here, its formal definition is still pending. Considering blockchain design, permissioned blockchains or hierarchical hybrid public and consortial proof-of-authority multi-chains seem promising at the current stage of the project. Next steps are investigating analytical relationships between blockchain designs and the quantitative evaluation metrics and building simulation models, which will be used in the last step to extensively answer the central research question.