Software Defined Digital Grid on a P2P Network — On Systems of Autonomous Energy Cells

Prof. Rikiya Abe (CEO&CTO, Digital Grid™ Platform Corporation), 
Dr. Carsten Stöcker (CEO Spherity GmbH)

Abstract, TL;DR

Renewable energy resources such as solar and wind have increased their capacity significantly. To support their penetration to the conventional grid, people try to increase the transmission capacity of the grid, which will strengthen grid synchronization. However, with a highly synchronized grid and variable renewable generation, a small grid failure can easily start cascading outages, resulting in a large-scale blackout. Smart grids and micro grids are part of the synchronous grid and do not solve the grid resiliency problem.

The Digital Grid™concept has been introduced in 2011[1], where large synchronous grids are divided into smaller segmented grids and connected asynchronously, via multi-leg IP addressed AC/DC/AC converters, named Digital Grid Routers (DGRs).

DGR can remodel traditional transmission and distribution grids from a tree-like network to mesh networks, solving the problem of grid resiliency.

DGR has the unique feature that it can record all the power generated, transmitted, stored and consumed along with properties such as location, time, power supply, price (including CO2 value) etc.

The Digital Grid Controller (DGC) communicates with the DGR and the smart meter, predicts the electricity supply and demand, and bids by buying and selling energy through the peer-to-peer (P2P) network using Blockchain.

The DGR receives the bidding result through the DGC and executes transmission and reception of electric power.

DGC reads the smart meter and reports the execution result to the blockchain, so that it receives the virtual currency as a result.

Matching of buying and selling the energy through the market becomes a balancing mechanism of demand and supply of electric power as it is, so that control and finance can be unitarily processed.

Software defined Digital Grid will not only convert the conventional electric power system into a highly reliable mesh network but also change to the P2P autonomous market network. The combination with AI/ML transforms the electric power gird from a tree-like, centrally controlled system into a System of Autonomous Energy Cells that are transacting among each other and self-stabilizing the power system on a local level.

Index Terms — smart grid, Digital Grid, System of Autonomous Energy cells, renewable energy, AC/DC/AC converters, Back to Back (BTB), P2P, Blockchain, Decentral Energy Markets, Autonomous Economic Agents, Digital Twins, Fintech


Renewable energy such as solar and wind power has characteristics quite different from conventional generators.

They do not need fuel. This means that the marginal cost of generating electricity is nearly zero. If you can control the flow of electricity, it will be the winner of the power market.

However, the power generation fluctuates, neither stable nor reliable.

In order to stabilize the grid, it is necessary to control the amount of renewable electricity generation in response to instantaneous fluctuations in demand. Sometimes it is better to divide a large conventional grid into smaller grids (this is called a “cell”). It is beneficial for both the main grid and the cell grid when these cells control the flow of power to the main grid without being affected by fluctuations in renewable energy.

In a conventional electric grid, multiple synchronous generators rotate at the same frequency and their behavior is fairly dynamic. As soon as a small fault occurs, all grids will be affected and cascading blackouts may occur. Therefore, in order for the grid to maintain frequency and grid stability, matching of supply and demand every moment is the most important task. Adjusting the amount of power generation in relation to the frequency change rate keeps the supply and demand consistent.

Our motivation is to make it possible for renewable energy to be fully released from such electrical constraints, and to implement supply-demand correspondence based on future market mechanisms.

All bids must be executed accurately by autonomous electrical equipment (so called autonomous economic agents) or by autonomous energy cells.

Measurements and verifications are also performed accurately and logged as data audit trails on the blockchain. The blockchain will provide authenticity, integrity and time stamping for asset IoT data in form of so called Digital Twins.

A tokenised digital currency is used to perform direct counter party financial transactions among parties that are previously unknown. These mechanisms are autonomous, software driven and linked to IoT data about physical transfers. The hybrid structure of conventional power systems and the Digital Grid cells will enable improved renewable power systems.

These transactions mechanism do not replace the traditional grid architecture, rather it is realized through cooperation and enabled by a much simpler market design.

In this paper, we explain how this new concept of software driven Digital Grid on P2P network works.


A. Digital Grid

The “Digital Grid” concept was introduced by R.Abe in 2011[1].

In this paper, it says that “the “Digital Grid” where a wide-area synchronized power system is subdivided (“digitalized”) into smaller or medium sized power systems. Subdivided grids called “Digital Grid Cells” (simply called “cells,” from now on) are connected asynchronously via “Digital Grid Routers” (DGR). … All electrical devices, including generators, batteries, even smaller home electronics, etc., in a Digital Grid can be IP addressed and controlled by or through Digital Grid Controllers (DGCs).”

6 years after publication of this paper, DGR has developed marks I, II, III of several kW class. 40 kW machines combining commercial inverters are also being tested demonstration projects.

Digital Grid Platform Corporation and University of Tokyo is currently developing a 10 kW class high-speed control model based on a 1 kW prototype model called “PROTON”. In order to hybridize the power system and the cell, we retrofitted the DGR and DGC with Blockchain including Blockchain identities instead of the IP addresses in order to build a decentral power trading market. The use of Blockchain for applications such as power control and financial transactions is being tested.

Fig. 1 Asynchronous connection of AC circuit by DC bus

A typical asynchronous connection of an AC circuit is a back-to-back (BTB) connection, as shown in the left hand side of Figure 1. BTB uses the DC side of two bidirectional AC / DC inverters as a common bus to block frequency and voltage information and transmit only active power in both directions. It is the principle used in systems such as frequency conversion stations and high-voltage, direct current electric power transmission (HVDC).

The Digital Grid Router (DGR) is an invention which added one or more bidirectional AC / DC inverters to the common DC bus of BTB as shown in the figure on the right side of Fig. 1. Power can be exchanged between multiple AC terminals regardless of frequency or voltage, even handles DC from sources such as batteries or photo voltaic power cells (PV).

Fig.2 Three (3) Leg DGR image by Half Bridge Inverter

FIG. 2 shows details of the three leg DGR. Each leg consists of a typical half bridge inverter. This configuration is equivalent to the DGR shown on the right side of FIG. The DGR leg shown in FIG. 2 is controlled by the gate signal of the gate of the switching element in the DGR arm.

In PROTON, hysteresis feedforward control is used. Even if the physical hardware of the half bridge inverter is the same, any power supply such as AC 60 Hz, AC 50 Hz, DC, etc. can be created only by changing the control mode. You can also convert energy smoothly by connecting to any of these power supplies.

In a real application, you can substitute a power converter system (PCS) by connecting one leg to the PV panel. Likewise, one leg can be connected to the battery to substitute the battery charger / discharger. Single phase inverter for grid interconnection, three phase inverter, motor control inverter, DC / DC inverter, uninterruptible power supply (UPS) can also be substituted. By changing the mode of the software you can have different functions using the same half-bridge architecture.

So, DGR is a “software defined” multi-purpose, multi-terminal inverter that is steered by software in a DGR control module (Digital Grid Controller, not depicted in Fig. 2).

Since the DGR control module steers

  • the magnitude of power,
  • time and
  • direction

it is possible to set a power clump of constant magnitude of 15 minutes or 30 minutes as one power packet. If you do so, you will be able to send energy packets from point A to point B and to treat electricity like physical products.

The Digital Grid is able to take combine commercial and physical transactions when sending dozens of power packets from point A to point B, instead of distinguishing physical and commercial power flows in today’s markets.

Power transmission is sent at the speed of electromagnetic waves. Within the same synchronous power grid, DGR can instantly transmit individual power packets anywhere via existing transmission lines.

The conventional electric power system can be regarded as a high-speed transportation road of goods, and the cell can be regarded as a production base and a consumption base.

The “Digital Grid Controller” (DGC) will be the command tower directing DGR to absorb, consume and generate individual power packets related to solar, wind, load, and energy storage.

Fig.3 Power Packets Transaction
Table 1 Image of Power Packet Transaction Record

Table 1 shows the record of such power packet transactions. The power packet P1 is transmitted from Blockchain address (identity of DGR A) “0xbaf03ec …” to “0x6c009c …” (identity of DGR B) from 11/17/2017 11:20:00 to 11:30:00 and its power source is PV and has a carbon value of 0.

The power packet P2 is transmitted from Blockchain address “0x6c009c…” to “0x8af0ce…” from 11/17/2017 11:20:00 to 11:30:00 and its power source is coal with carbon value of 80.

This is just an example, but we understand that the power flow of “0x6c009c …” will be canceled if P1 and P2 are about the same amount.

In other words, it can be said that the power flow does not flow from P1 to the reception of P1 and from P2 to the reception of P2, but rather a power flow flows from P1’s feed to P2’s receipt and from P2’s feed to P1’s receipt, then swapping financial settlements would be the same, if P1 and P2 are about the same amount.

This fact leads to an extremely important idea that the actual flow of electric power is taken as a flow of reality and the transaction as a commercial flow is securitized and separated.

B. From Tree-like Distribution Network to Meshed Networks

Conventional grid architecture is “Tree-like distribution network” as shown in the left drawing (i) of Fig. 4.

In Fig. 4, the main generator (G) drawn on the lower side supplies all the electricity to the local demand (black spot) via a transmission / distribution network similar to a tree trunk or branch. Power flow is unidirectional from generator to demand. The tips of branches and twigs are radial. This architecture is sometimes called “tree-like distribution network”.

In the (i) Tree-like distribution networks, each twig and branch of the tree have orderly voltage level from the high end at the generator to the low end of the demand.

Fig. 4 Distribution Network Evolution

When the number of distributed generators (branch (G)s) such as solar cells, wind power generators, fuel cells, batteries, gas engines increases, they become as shown in the middle view (ii) in Fig 4.

As shown in the view (ii), when the distributed power supply is installed at the end of the power distribution network, the voltage fluctuation is likely to occur at the local power distribution end. Voltage level has to be maintained within a certain range for customers’ electric appliances.

The difficulty of the voltage level control will become a significant issue. Power companies, however, did neither attempt to measure nor control the voltage level of each customer.

Meshed networks — a System of Energy Cells — in Fig.4 (iii) will become much attractive and easier to maintain voltage level, if we can control power flow through these multiple paths in the meshed network.

In the actual distribution network of the city, redundant distribution networks are installed and it is a mesh like network. However, it is not allowed to flow the power flow in the form of a mesh. Since it is only to route either route by cutting the switch in between, it is effectively a radial network

It is because there is an electrical problem that electric power cannot form in a meshed network. Meshed connections in electrical grids are prohibited.

When the connected line makes a loop, a slight voltage difference makes a loop current flow constantly, resulting in power loss and possibly heat, resulting in damage to the electric wire. Once a small fault occurs, a large fault current will flow into the fault point, resulting in a big failure and sometimes cause a fire.

It will be a good solution to have a back to back (BTB) inverter connection to be inserted among meshed networks (or the leaves of a tree in the tree-like network). A loop current or a fault current will be controlled at BTB.

A multi-terminal BTB, that is, the DGR would be an ideal solution for the meshed networks.

The DGR uses multiple AC/DC/AC connection, each pair of which act as back to back (BTB). The fault current is reduced by reactor coil and cut in 1,000 to 10,000 times faster than ordinary circuit breakers. The so-called loop current is also automatically suppressed by DGR. This means that it is relieved from constraints of the power system.

Instead, deliberate power flow based on the market results is created by DGR sending out or pulling in to the network. Mesh networks are particularly suitable for distributed generation for renewable power generation where demand is dispersed.

Meshed networks controlled by DGR have flexibility in the direction of power flow, have a redundant backup mechanism, and become a power system with high robustness even when renewable power generation is abundant.

How to manage such a mesh network is a big challenge.

It is not realistic to intensively analyze the enormous number of power flows obtained from DGC and to control DGR, and security problems may also become serious.

C. Decentral Technology, Blockchain and Modern Cryptography

We propose to use an emerging new field of information technology called decentral information technology for the Digital Grid Transaction System.

Decentral information technology uses peer-to-peer networks such as consensus algorithms, blockchain, DAG tangle, P2P communication protocols, P2P file and data storage, trusted computation as well as modern cryptogryphy to

  • secure IoT data,
  • establish data audit trails,
  • provide time stamping,
  • and provide (trusted) computing resources and
  • enable direct counter party transactions among previously unknown parties without the need to involve intermediaries.

These direct counter party transactions facilitate physical and financial transaction at the same time, with (almost) zero transaction fees, credit risks and double spending risk, with no intermediary involved. Direct counter party machine-to-machine transactions among DGR, DGC and Digital Grid Assets can be fully automated while being executed based on data audit trails, trust and reputation vectors about historic and actual physical transactions of a machine.

In this paper we will not elaborate the variety of existing Decentral Information Technology developments. For simplicity and focus we concentrate primarily on the use of blockchain.

Definition: Blockchain Peer-to-Peer (P2P) module — A Blockchain Peer-to-Peer (P2P) module connects a (Digital Grid) device to a the P2P network in order to provide the device with an identity and to enable it to transact via the P2P network. A Blockchain P2P Module can be either:

  1. A μcontroller with a crypto chip (or trusted enclave) storing a unique public private key pair and enabling signing, hashing and / or encryption. The public key is registered in the blockchain P2P network representing the unique identity of the device and providing a blockchain wallet. The μcontroller is connected to a P2P node operated somewhere in the internet for connecting the device to the blockchain transaction layer.
  2. As an alternative physical unclonable functions (PUF) or SW solution such as whitebox encryption or sMPC can be used to connect the μcontroller with a P2P network and provide identity, wallet, key management and a secure access to a decentral transaction system.
  3. A light Blockchain client that is optimized to run on IoT devices.
  4. A full Blockchain client.

Remote Nodes: A simple technical set-up of (1.) or (2.) would be sufficient to assign a DGR or a DGC an identity and to register the device as a so called “smart asset” on the Blockchain. The DGR / DGC is connected to either a gateway running a blockchain client or another device running the blockchain client to read the Blockchain and / or send transactions to the blockchain.

On Device Blockchain Clients: For more advanced implementation the device will run its own blockchain client including a wallet described in (3.) or (4.). This enables for instance transaction analytics, service discovery, context learning, autonomous economic agents or direct transaction validation.

As soon as two devices are registered on the Blockchain they have a machine identity (public key). These two devices (Allice and Bob) can check the public key on the Blockchain and start to directly exchange secure information or messages among each other.

As long as the cryptographic primitives are not broken (e.g. by Quantum Computing) blockchain serves as a Next Generation Public Key Infrastructure.

Data Security and P2P Communication: This blockchain infrastructure provides a variety of data security features without the need of a central party for orchestration. Bob receives a messages from Allice that is signed and validated to prove authenticity and hashed to prove data integrity. In addition messages can be encrypted. This first feature serves as a protection against attack vectors such as man in the middle or message injection in the Digital Grid.

With authenticity and integrity of data a full digital representation can be established about grid assets including DGRs. All meta data and sensor data can be stored in a so called digital twin of the asset including asset reputation, a data audit trail to the creation of sensor data and time stamps (Fig. 5).

Physical transactions are fully digitized. A financial transaction can be done simultaneously among these two assets by either using smart contract escrow payment mechanisms or P2P token streaming.

Fig.5 Energy Exchange and Financial Transaction in the Physical and Digital World

A second new feature is data and energy provenance. Devices connected to a blockchain are registered with attributes on the blockchain (e.g. a feed-in device is registered at a DGR as a PV cell including its idnetity and provenance as provenance data are already recorded by the manufacturer on the blockchain as well).

With these attributes in place energy gets a provenance. Renewable Energy Certificates can be created and tracked. An energy package in the Digital Grid can be earmarked with its specific provenance. The recipient of the energy package, Bob, can be sure that he receives the type of energy directly form Allice that he bought. Further attributes such as voltage levels, frequency or response behavior to grid control signals can be stored as well in order configure a DGR router properly or to pre-qualify a flexibility by measuring responses to control data sets. Reputation and trust vectors about pre-qualified flexibility assets are written to the blockchain to enable trusted flexibility transactions. Reputation and trust vectors can be created with machine (context) learning algorithms that are integrated with the blockchain.

A third new feature is that data (or their hash values) can be recorded in an immutable Blockchain ledger and time stamped.

With a fourth new feature — using smart contracts for access management — secure access to assets can be granted and orchestrated. Access to assets can be either access to a data stream (sensors) or access to physical behavior of assets (actors). In addition ownership of assets can be securely transferred among parties .

A fifth new feature is the the creation of decentral (blockchain) markets for establishing local market prices, P2P transactions, P2P trading, audit-proof transaction histories and P2P payment settlement.

Fig.6 Economically Independent Machines

The combination of an IoT devices (such as a DGR or PV cell) with a blockchain P2P module and analytics results in the concept of an economically independent machine that makes economic decisions while enabled to do physical and financial transactions. Each economically independent machine has its own machine P/L statement (Fig. 6).


A hierarchy of aggregation layers exists in the Digital Grid. Each aggregation layer represents either an individual physical or virtual market for participants conducting direct counter party transactions. These virtual markets can also be understood as energy cooperatives as the participants equally share assets and aggregation benefits among each other eliminating the need of a third party such a utility.

Fig. 7 depicts the following aggregation layers with the respective markets:

  • Digital Grid
  • Digital Sub-Grid
  • Individual Cells (Micro Grids)
  • Devices

Each aggregation layer can be a either a physical system in which DGR and DGC transact or a sub-set of DGR, DGCs or Devices that transact across physical systems in a joint virtual aggregation layer or market.

Components of the Digital Grid can then dynamically decide in which market they participate even is this outside (local) physical boundary. These markets can be dynamically defined, on-demand markets.

Fig. 7 Aggregation Hierarchy and Market Layers


A. Micro Grid

Fig.8 Micro Grid, Digital Grid, Semi Digital Grid

The micro grid, as shown in the left figure in Fig.8, is a group of local generators and loads, connected to the main grid and receiving electric power supply, at the same time, it releases the connection point and power is supplied to its own load only by its own generator It is a small-scale electric power system that has a self-sustaining operation ability to stably supply.

When the micro grid is decoupled from the main grid, the micro grid must operate in island mode with generator and load. This is a big challenge for micro grid.

When the generator was linked to the grid, it was unnecessary to control the voltage and frequency, but when it is separated, it must be self-sustaining and maintain frequency and voltage. This is a much more difficult control.

In the case of Digital Grid, as shown in the middle figure in Fig.8, even if it is connected to the grid or separated, the generator control will not change and only the current will be sent in to DGR.

This control is autonomous.

Semi-Digital Grid, as shown in the right figure in Fig.8, will be applied in small cells, such as houses and shops to reduce the number of legs of DGR and to suppress the conversion losses.

B. Digitized Public Grid

Fig.9 Cell and Service Provider interacting on Blockchain Markets

The connection between the cell and the power system is asynchronized by DGR, and electrical restriction is reduced.

The lack of electrical constraints means that where you are on the electrical system will have little meaning.

As long as there is no electrical constraint, the DGC in the cell will be free to buy and sell electricity.

Cell sizes vary from large to small. The service provider shown in Figure 9 gathers the large and small cells and connects them to the market. The service provider is like a securities company in stock trading or an autonomous agent. DGC can also predict its own power generation and consumption and issue orders. It is also possible for service providers to substitute orders under certain rules, like investment trusts.

Multiple service providers can build a mechanism to bid on the blockchain market by collecting the order of each customer.

The blockchain market in electricity is like a trading market like the New York Stock Exchange or NASDAQ. When “sell order” matches “buy order” at a certain price, in the meantime, electricity supply — demand matching is also satisfied. The price is determined economically according to supply-demand balance.

Matching such market-driven supply and demand has its own unique characteristics. Solar power can not be output unless someone purchases solar power, even if the sun shines. Unless selling is realized, DGR has only to reduce the output power of the selling leg. At that time, the PV leg also limits the output or is stored in the battery whether it meets the demand.

Of course, in addition to such a market settlement system, we can think of proprietary power matching method using blockchains.

For example, it is conceivable to match in order from the closest node in the propagation process of a peer-to-peer network.

When registering a DGC or DGR in a node on a blockchain, those that are physically close will be registered in the same node. In that case, the matching of supply and demand is first determined preferentially within the same node and propagated to nearby nodes on a peer-to-peer basis. The rest is settled in this way. In this way, it is possible to create a mechanism in which demand close to physically and supply are preferentially matched.

However, this method is not a very good way for financial settlement. Because the price of buying and selling is decided by prioritizing the closeness of distance, it is not economically rational. Furthermore, proximity of physical distances is not necessarily meaningful from the viewpoint of electrical engineering.

Paradoxically speaking, even if matching between selling and buying is done between DGCs at long distances, regardless of that, the selling electricity output from one DGR is absorbed by the closest buying DGR. This is because electric power flows into the place where the impedance of the electric power system is the smallest.

Since the flow of physical electricity is automatically determined in this way, the financial settlement can be separated from the physical flow and swapped.

C. Use Case: Urawa Misono Project

Digital Grid Platform Corporation and University of Tokyo have just started “Urawa Misono Project” with a three year’s budget about 6.5 million in terms of US dollars, funded by the Ministry of Environment in Japan.

This project is the joint work with Tateyama Kagaku Group Inc. and the University of Tokyo, and Kansai Electric Power Co. Ltd, Tokyo Electric Power Company Holdings Inc., Hitachi EI system, Tessera Technology Inc. USD, NTT Data Inc.

Fig.9–2 Urawa Misono Project(2017–2019)

In this project, we have five household prosumers, 10 household consumers and PV producer in a large shopping center AEON (Fig. 9–2).

Each prosumer has a DGR of 5kW PV, 12 kWh battery, 10kW sub-grid, 6kW selling and 6kW buying legs, and a DGC that communicates with buying and selling smart meters and DG net (Blockchain market).

Each prosumer has only DGC that communicates with buying smart meter and DG net (Blockchain market).

PV producer will install 60kW PV, which power conversion equipment (PCS) is replaced with the same DGR as in the prosumers’, only with its connection and control mode are slightly changed. The producer has a DGC that communicates with selling smart meter and DG net (Blockchain market).

Fig.10 Sell-Buy Matching on Blockchain Market

Each DGC can predict demand and power generation and bid for selling and buying to different markets made on the Blockchain in units of 30 minutes to 24 hours and even 1 week and 1 month.

Many markets made in the 30-minute frame, such as 30 minutes after 1 hour and one and a half hours later, simultaneously contract and sell and buy in “Zaraba” method, which is the first come first served basis.

The reservation result is recorded in the Blockchain and sent back to each DGC.

DGC tells it to DGR. When the reserved time comes, DGR will output or consume (or stored) power based on the promised result.

DGC further reads the value of the smart meter and verifies whether the correctly promised result has been executed and reports the result to the Blockchain.

Based on the received report, the Blockchain checks whether there is any difference with the contract result and pays our private virtual currency “Dther” based on the value of the smart meters. (As an alternative asset-backed fiat crypto currencies can be processes as well).

Fig.11 Energy Balance Record for certain period of time

Because the DGR leg is conducting accurate current control, the value obtained by integrating the product of the current and the voltage becomes a considerably accurate electric energy accumulation.

By measuring them on a leg-by-leg basis and aggregating them at regular time intervals, a record of energy balance as shown in Fig. 11 is obtained.

The right column “Input” shows generation of electric power and the left column “Output” shows electric power consumption, including sales to the main grid or sub grid.

All the value should be positive, however, only the battery in the right column can take a negative value when charging, and its value is stored in the battery in the “Stock” in the lower right column.

Non-CO2 values are also recorded on the left side of the bottom row of the same table.

Fig. 11 may show a record of energy balance over a certain period of time, for example, 30 minutes duration, 1day, 1 month, or 1 year.


A. Economic Benefits

Power Balancing and Financial settlement

Financial transactions using Blockchain are generally decentralized and can be constructed cheaply, and it is said that it will be a highly secure mechanism for a relatively small maintenance cost.

If you make a power trading market using a Blockchain, since both selling and buying are established at the same time, the same amount of power generation and consumption as in the electric power system is automatically achieved.

Furthermore, if the Blockchain is used as a mechanism of the financial settlement, the authenticity of settlement will also be secured.

That is, a market settlement type power trading system using a Blockchain will provide an electric power balancing system and a financial settlement system at the same time as an inexpensive system with increased security.

This brings extremely high benefits to all the stakeholders.

Electricity Market vs. Self-consumption

As renewable energy has zero marginal cost, it will be put into the market as a more powerful power supply after depreciation ends. As a result, electricity charges continue to decline, the field of electricity use industry spreads, electric power trading. The market is expected to further expand.

On the other hand, it is generally said that it is desirable to generate electricity in the vicinity of the demand for the distributed generation, and to reduce the usage of transmission and distribution lines as much as possible.

However, considering market transactions using Blockchain, trading the electricity in a wide area will lead to market activation.

Fig. 12 Gross Bidding and Net Bidding

Fig. 12 shows the image of the A) gross wiring for bidding, and B) net wiring for bidding. Gross wiring requires selling leg and buying leg. Each leg controls the sales electricity and buying electricity respectively. This means the market size double or more compared to net wiring.

If we apply net wiring for bidding, one DGR leg is used. In this case, either selling or buying is executed and both are not executed at the same time. In this case, most of the electricity is consumed within the vicinity and never goes to the market.

Therefore, gross bidding should be selected according to the market-oriented basis, however, in general, there is something that has strong persistence of faith in local production and consumption.

This problem can be solved by separating the actual electricity flow and contract based tagging. Contract based tagging is done by swapping of the contract.

One of the important benefits of gross wiring is to solve the problem of grid charge reduction of transmission and distribution networks.

With the net wiring, there was a serious problem that the purchase amount of electricity has a tendency to reduce due to demand side generation, and it becomes difficult to charge the grid charge enough.

But with gross wiring, electricity is always purchased separately with selling electricity, so the grid charge will be guaranteed.

This is beneficial not only for Grid Company but also for market participants.

Sub-Grid Market

The sub grid is not a necessary electric grid immediately.

However, where common underground groove such as shopping malls and industrial estate school residential streets are located, it is a distribution network that can be installed cheaply and maintainability is very high.

The Sub Grid will be used to lend and borrow storage batteries from each other, but if you install a private power generation facility in the future, the market mechanism similar to what we have said so far will work.

Competition is unlikely to occur and it tends to be an exclusive market if the location where the customer receives the service is limited to one receiving point.

Electricity is supplied from either side like the main grid and the sub grid, and inevitably active competition will arise if it gets supplied with power, heat and other services.

This is a very desirable form for customers. DGR will provide asynchronous connection that enables to use both power supplies at the same time.

B. Environmental Benefits

As shown in Fig. 11, DGR can distinguish power generation and record them and also can distinguish and record non-CO2 volumes

If a market is created to sell this by classifying it from electricity, and if the government obliges companies to purchase, the promotion of renewable energy will be accelerated and environmental benefits will be born.

C. Social Benefits — The New Energy Cooperatives

We advocate for open blockchain Digital Grid systems to which new (machine) entities can be easily connected with access to the variety of decentral markets. This will lead to low adoption barriers and equal participation for any new entity in the market dynamics.

The entities in a market, physical or virtual aggregation layer, are registered and their assets can be automatically validated and qualified. As the blockchain SW layer is built upon the principle of distributed ownership the benefits such as asset sharing or aggregation benefits will be equally distributed among the participants.

The active, registered participants of a market are representing the members of a (dynamic) ‘cooperative’. This new concept will drive social innovation in how people will organize their participation and contribution to cooperatives.

Members of a cooperative can participate in investements in the local infrastructure such as installation of renewable energy sources, batteries / flexibilities or elimination of grid congestions. The blockchain offers innovative crowd funding and asset sharing methods.

VI. Implementation Outlook

A. Data and Algorithms

Using the blockchain so called data oracles can be established. These data oracles connect the outside world with the Digital Grid on the Blockchain. For instance a weather data or prediction service or a single device can be connected to the Blockchain to provide authenticated data for optimization purposes.

Sensor data can be recorded in a decentral Digital Grid data log. There are technologies such as Blockchainified Databases (e.g. BigchainDB) that allow to write a huge number of Tx per second of (authenticated) sensor data into an immutable database.

Algorithms can trust these data and be either applied to optimize an entire system based on the data stored or the algorithms are deployed at the edges (i.e. DGR, DGC) to conduct local optimization. In the second case algorithms are brought to the data.

Algorithms can run off-chain and connected to on chain transactions.

Concepts of decentral / open AI can be applied to make the system more robust or to optimize the system form a commercial perspective.

Energy routing data can be recorded as well for routing optimization purposes.

B. Optimization and Inverse Game Theory

Optimization can be performed on all aggregation levels of the Digital Grid.

There are two options for optimizing transactions within an aggregation layer:

  1. Markets: Entities within an aggregation layer are engaging via markets on transactions
  2. Algorithms: An algorithms optimizes the resource usage on behalf of the market participants and on a pricing mechanism

The combination of both will be most effective:

  • markets defining economic incentives that drive optimization of the Digital Grid, power generation, consumption as well as deployment and use of flexibilities and
  • algorithms that does optimization on behalf of the individual entities in the sense of a ‘cooperative’. Entities delegate optimization tasks to the cooperative algorithm. Benefits are shared.

A market or benefit allocation design for either defining rules or connecting one entity with other entities in the aggregation scheme is not yet been designed. We propose to apply ‘inverse game theory’ (or mechanism design) to identify mechanisms for the markets and optimization algorithms including incentive schemes for multi-agent systems. Optimization takes into account different aspects, economic, social and technical parameters. Shared social values such as equal participation and benefit distribution can be introduced as constraints for a machine learning mechanism design.

C. Autonomous Agents and “System of Autonomous Energy Cells”

We introduce the concept of decentralized autonomous entities to the Digital Grid: Individual devices, cells or routers can act as autonomous agents. They are registered on the Blockchain and have a wallet. They can physically, financially and regulatory engage with other participants.

Autonomous agents running their own algorithms might decide to share data and / or algorithmic code and / or learnings with other agents to achieve aggregation benefits. This can be understood as agents (dynamically) organizing themselves in a machine learning and data cooperative.

In combination with a cell-type grid structure the autonomous agents managing individual energy cells transform traditional a tree-like, centrally controlled energy system into a System of Autonomous Energy Cells.

These Autonomous Energy Cells are transacting among each other energy and access to flexibilities while the transaction is recorded and settled on blockchain markets. Self-stabilizing is enabled and settled on a cell level providing an effective stabilization means for the entire the power system on a local level.

D. A new protocol: Integration of Consensus and Crypto Economic Approaches among Multi-Agent and Blockchain protocols.

Integration of multi-agent system and blockchain protocols will be a future research and development topic.

This will lead to new protocols as a multi-agent system with a decentral, edge intelligence and machine learning capability will embed consensus algorithms and crypto-economic incentivization schemes to operate effectively and to avoid overshoot reactions in edge cases of a system that might bring down the entire system.

Recent blockchain and AI projects such as are already developing a combination of machine identities, trust vectors, protocols, machine learning, decentral markets and autonomous economic agents to establish a Service Discovery Capability and Machine Language Communication Skills.

These types of protocols enable machines, autonomous agents, to define a need or an offer, to identify counter parties in the real (physical) transaction context, establish a communication and dynamically organize on-demand markets even for entirely new needs while facilitating the machine transactions.

E. Smart Contracts and Transactional Privacy

Algorithms can then be used by either participants or aggregation layers to optimize technical and economic parameters of the energy system. Market mechanisms will be included into or defined by smart contracts to allow the agents to perform economic transactions on the Blockchain, e.g. buying, selling, storing, routing or trading of energy. All transactions can be recorded in the Blockchain as well.

A smart contract market will consists of on chain code (such as ID and access), off-chain code, decentral computation markets, trusted computing hardware solutions or hybrid solutions.

Smart contract transactions can be recorded on the Blockchain. With further applications of modern cryptography encrypted data are used.

‘Selective privacy’ can be established with modern cryptography as well. For instance a key can be provided to a 3rd party (such as an authority) to provide access to recorded transactions.

Academic research and practical product development are currently ongoing to use cryptographic methods such as confidential transactions, stealth addresses, one time transaction identities, zero knowledge proof systems or secure multi-party computation to increase the level of privacy.

F. Pricing Mechanism and Priority

Smart contracts allow to orchestrate the planning energy package transfer, the execution of energy package transfers as well as billing and settlement of both, energy supply and grid service. Billing and settlement could be done in real time (even by streaming crypto tokens in sync to the delivered energy).

Different network routes can be priced differently depending on availability or technical parameters. Bottlenecks in an energy system will be assigned a commercial value which defines value of fixing it in a local market or via virtual entities with complementary capabilities.

Priority rules can be defined in the market design as well and agreed upon participants in a smart contract. E.g. in case of emergency some entities have a higher priority, pay a higher priority fee and get energy first while others have to wait in case of a bottleneck or an emergency. This concept is similar to ‘Quality of Service’ in data networks.

G. Decentralised Grid Control Systems

Combination of autonomous agents and algorithms will lead to decentralized gird control systems. These decentralized gird control systems will support today’s central grid control systems that are typically operated by TSOs.

H. Elimination of Balance Groups

With the combination of Digital Grid, Blockchain and decentral markets the concept of “balance group accounting” — which is the foundation of today’s energy market models — can be replaced.

In balance groups the commercial flow of energy is disconnected form the physical flow of energy. This disconnection leads to inefficiencies, complexities, grey areas in regulatory frameworks and unbalanced incentive systems.

The digital grid and Blockchain are connecting commercial and physical energy flows. Many complex and time consuming processes in billing and settlement of grid and power services will be replaced by much more efficient processes doing billing and settlement in (almost) real time.

I. DGR Prototypes

A first 10kW Proton DGR Prototypes (Fig. 13) are currently being built and tested by Hitachi IE (Hardware) and Digital Grid™ Platform Corporation (Firmware). The Proton DGR is planned to be commercially available in 2019.

Fig. 13 10kW Proton DGR. Source: Digital Grid™ Platform Corporation

VII. Conclusion

The fusion of Digital Grid and blockchain technology will enable entirely new, much more efficient energy systems and markets.

In a Digital Grid energy packages can be sent from A to B in a similar way as we send data packages from one device to another device in the internet. The Digital Grid has the potential to transform traditional transmission and distribution grids from a tree-like network to mesh networks, solving the problem of grid resiliency and linking physical energy flows to commercial transactions without the need of balance groups.

Matching of buying and selling the energy through the decentral blockchain markets becomes a balancing mechanism of demand and supply of electric power as it is, so that control and financial transactions can be instantaneously performed.

Software defined Digital Grid will not only convert the conventional electric power system into a highly reliable mesh network but also change to the P2P autonomous market network.

There are several limitations in the blockcain technology such as privacy, performance or scalability. These limitations are areas of ongoing research and development. A variety of potential solutions is proposed by decentral technology research teams.

Therefore we believe that technology advancements will it make soon entirely possible to develop a decentral P2P platform as a transaction system for a transactive digital gird.

The combination of technologies will soon create Software Defined Digital Grid on Blockchain/P2P Network.

About the Authors

Dr. Carsten Stöcker is founder of Spherity GmbH. Spherity is a scalable decentral platform for the fourth industrial revolution providing secure identities and digital twins bridging the physical, biological and digital spheres. He is a physicist by training with a Ph.D. from the University of Aachen. He also serves as a Council Member of Global Future Network for the World Economic Forum. Prior to founding Spherity GmbH, Dr. Stöcker worked for innogy SE, German Aerospace Center (Deutsches Zentrum für Luft- und Raumfahrt, DLR) and Accenture GmbH. | Twitter: @CarstenStoecker

Professor Rikiya Abe is CEO&CTO of Digital Grid™Platform Corporation in Japan. Prior to founding Digital Grid Platform Corporation he worked as Project Professor of Internet of Energy Lab, Graduate Course of Technology Management for Innovation, School of Engineering, the University of Tokyo, 7–3–1, Hongo, Bunkyo-ku, Tokyo, Japan, 113–8656


[1] IEEE Transactions on Smart Grid (Volume: 2, Issue: 2), R.Abe,, April 29, 2011 “Digital Grid: Communicative Electrical Grids of the Future

Like what you read? Give Carsten Stöcker a round of applause.

From a quick cheer to a standing ovation, clap to show how much you enjoyed this story.