Emergence of the Energy Internet (EI) demands restructuring of traditional electricity grids to integrate heterogeneous energy sources, distribution network management with grid intelligence and big data management. This paradigm shift is considered to be a breakthrough in the energy industry towards facilitating autonomous and decentralized grid operations while maximizing the utilization of Distributed Generation (DG). Blockchain has been identified as a disruptive technology enabler for the realization of EI to facilitate reliable, self-operated energy delivery.
IoT devices, including smart meters and sensors, communicate real-time measurement data of the large-scale participation of the Distributed Generation (DG) . This is envisaged to facilitate autonomous operation of energy grids, benefiting seamless integration of DG without the involvement of a third party compared to the conventional counterpart. Implementation of EI grids is proposed as an overlay of four layers: namely, physical layer, communication and control layer, application layer and data analysis layer . The former two are comprised of IoT devices and beyond 5G communication technologies, enabled through edge computing respectively. The latter two layers of the novel architecture incorporate the applications of the envisaged EI grid and data analysis technologies supported through big data management . Applications of EI span beyond offering dynamic energy prices to the consumers and obtaining their contribution in DR initiatives. They also exhibit prospects in multi-dimensional aspects, including: (1) Peer-to-Peer (P2P) energy trading; (2) plug-and-play interfacing for DERs; (3) microgeneration; (4) Demand Side Integration (DSI); (5) automation and management of distribution networks; and (6) management of energy data. Figure 1 illustrates the interrelationship of these applications.
Figure 1. Overview of the bi-directional energy and information flows in the applications of the envisaged EI grids.
Together with these applications, EI envisages autonomous grid operation where the central authority governing the grid under the current context will be overlooked 
. Further, this would be protruding as consumers begin to gain liberalization in the energy market and actively participate in power production. Human intervention in the decision-making process will be automated through smart contracts, facilitated by Artificial Intelligence (AI) and Machine Learning (ML) algorithms. However, as a consequence of delegation of authority among stakeholders and alleviating the contribution of the intermediary, trust establishment would become a key consideration regarding EI grids.
EI will be enabled through blockchain, which is a Distribute Ledger Technology (DLT) with inherent features including immutability, transparency, distributed verification/storage and decentralized authority over a peer-to-peer network 
. Security features and privacy-preserving techniques incorporated with blockchains offer solutions to mitigate cyber-physical attacks and privacy violations within the operations of the EI grid. Smart contracts enable autonomous operations of the EI grid with the execution of programmed scripts upon the fulfilment of the defined prerequisites 
. Some processes could be automated through the utilization of smart contracts implemented upon blockchain platforms. These include billing for the energy consumption, invocation and revocation of certificates to authorize heterogeneous DER integration, authorizing payments upon energy trading, dynamic price signalling and monitoring IoT devices to identify node tampering 
. Realization of EI is also envisaged to be facilitated through the developments of beyond 5G and 6G communication networks through inherent features. These include DLT/blockchain, ultra-massive machine-type communication, extremely low-power communication, extremely reliable low-latency communication, AI and ML, big data management and distributed processing through edge intelligence 
The significance of inherent features of the blockchain, utilized towards realization of the next generation of smart grids in the identified directions, have been illustrated in Figure 2.
Figure 2. Key directions of blockchain utilization in EI realization.
2. Energy Sustainability through Heterogeneity
2.1. Key Challenges
Seamless grid integration:
How can we facilitate seamless grid integration of DERs in order to achieve a decentralized, customer-centric grid architecture 
How is it possible to achieve a decentralized marketplace with dynamic price signaling and maximized consumer satisfaction 
Secure communication: How can we manage secure communication links and storage for large energy data aggregation arising from the increasing number of grid interconnections in a decentralized and secure platform?
Transient and dynamics:
How can we mitigate the transient over voltages and undesirable dynamics resulting from bi-directional energy routing and uncoordinated grid interconnections of DG, ESS and EV 
Improving interoperability: How can we achieve interoperability of heterogeneous energy sources that adopt different grid interconnection standards?
2.2. Role of Blockchain
Smart contracts can be utilized to ensure seamless integration of DERs through invocation of certificates and revocation of them upon request. Blockchain establishes a trusted environment in a decentralized marketplace, which provides confidence for the prosumers to engage in P2P energy trading while avoiding the risk of non-repudiation and double spending 
. Microgrid implementations such as Brooklyn microgrid 
and Power Ledger 
in Australia could be identified as promising, decentralized blockchain-based solutions facilitating P2P trading. Furthermore, Share & Charge in Germany and Juice Net of North America are milestone projects in dencentralized energy markets for P2P EV charging 
. Inherent cryptographic encryption of blockchain would enable user authentication without third-party intervention. Blockchain could further function as a secure and reliable data communication link and storage platform for efficient management of large data sets associated with these diverse EI applications. Energy sustainability is assured with transparent exchange of heterogeneous forms of energy 
However, mitigation of transient overvoltages and issues related to interoperability, which arise while the implementation of heterogeneous EI grids cannot be facilitated through a blockchain platform itself. Adequate standards need to be implemented with the intention of sustaining desirable grid operations.
2.3. Future Directions
The incentive-based benefits obtained through real-time dynamic electricity pricing are receiving attention in the grid integration of DERs beyond small scale capacities at different voltage levels of the grid 
. This will be facilitated through 5G and beyond 5G technologies, offering ultra-reliable, low-latency communication and edge intelligence supporting remote communication for intermittent connectivity respectively. Blockchain would provide an overlay with distributed storage facility. Precision decision-making, incorporating AI along with reliable communication and data processing through edge devices, as envisaged with 6G, would enable the realization of next-generation electricity networks 
Decentralized energy trading is a well-established research area with several approaches proposed and real-world scenarios being implemented 
. Examples include P2P trading of renewable energy 
. Communication links are expected to be made secure through proposed blockchain integrated architecture, which, however, has room for improvement with novel Smart Grid 2.0–specific security threats to be addressed 
. Meanwhile, facilitating seamless grid integration of DGs 
and their interoperability 
are the challenges with future research prospects considering the existing work, which would require collaborative technological approaches facilitated through blockchain.
3. Improved Trust, Security and Privacy
3.1. Key Challenges
How can we prevent tampering and unauthorized accessing of smart meters and smart sensors to ensure integrity of the obtained energy measurements 
Man-in-the-Middle attacks in EI grids: How can we establish a secure communication link between the prosumer and the consumer during energy trading and prevent Man-in-the-Middle attacks causing data manipulation?
DDos attacks in EI grids:
How will it be possible to detect Distributed Denial of Service (DDoS) attacks causing deliberate traffic of energy requests and depriving the legitimate users from consuming energy 
Privacy issues: Can a consumer participate in DSI initiatives while preserving the privacy of energy consumption data which can trace back to the behavioral patterns of the user?
Authentication: How can the identity of a node in the energy grid be verified in a decentralized architecture without revealing the connection between the energy signature and the owner’s name and location?
AI and ML related-attacks:
How can we mitigate data poisoning attacks related to integration of AI and ML techniques in predictive data analysis 
3.2. Role of Blockchain
Blockchain platform inherently establishes trust with minimal external interventions while offering a secure and transparent mechanism to create a reliable link between the nodes participating in energy trading. Transactions are recorded in an immutable and transparent format while each node holds a copy of the current ledger 
, preventing data modification and false data injection. Smart contracts could automate processes such as billing and finance settlement without the requirement of a trusted third-party intervention at a cost, while blockchains with the inherent use of Public Key Infrastructure (PKI) would enable identity authentication with pseudo-anonymity, privacy preservation of the participating nodes and protection of data integrity 
. The Lightning Network and Smart Contract (LNSC) model proposed in 
offers a security model comprising of registration, scheduling, authentication and charging phases. This integrates security options for user authentication, facilitating secure mechanisms for charging and discharging EVs. Guardtime, a US-funded project, has utilized a keyless authentication scheme for scalable EI grids with hash-function cryptography and digital signature authentication 
Cryptanalysis, in which breach of encryption algorithm is observed, can be addressed through the digital signatures incorporating private-public key pair, which is unique to each stakeholder. AI and ML models introduce a new set of adversaries, including data poisoning attacks, model evasion, extraction and inversion–related ML techniques utilized in EI grid realization 
. Data poisoning can be mitigated through the incorporation of blockchain distributed data storage, while alternatives such as adversarial machine learning, moving target defence and defensive distillation would provide resilience against adversaries identified in ML models 
Even though tampering of IoT devices and undesirable data traffic in communication channels cannot be fully addressed through blockchain initiatives, such platforms can be utilized to monitor the scenario and execute corrective measures to minimize the damage. Further, the existing security and privacy-preserving mechanisms incorporating cryptography and 51% attacks on the blockchain-based applications are vulnerable to advancements in quantum computing, thus demanding for quantum-resilient security alternatives 
3.3. Future Directions
Under the current blockchain context, immutability of the distributed ledger has been exploited to establish the trust factor and verify information security. However, the developments emerging with 6G technology facilitate distributed computing utilizing edge devices, which could further enable consolidation of resources to achieve computational efficiency 
. Such collaborations would pose a risk of accumulating 51% authority over the peer nodes, thereby gaining capabilities for modification of the past records. Such adversaries should be addressed in the envisaged grid operation.
Further, integrity and confidentiality of the information, along with the identity of the user, could be compromised through the revealing of the public and private keys used in PKI. This could be mainly due to the prolonged usage of the keys and as a result of the malicious attempts to reveal these cryptographic text patterns utilizing quantum computing, which is the most recent development enabling extensive computational capabilities 
. Ensuring security and privacy with technological progressions would be challenging in future grid implementations 
Blockchain integration with Smart Grid 2.0 has facilitated in mitigating software and network related attacks, including Man-in-the-Middle and DDoS adversaries 
. Further, the existing work has proposed different user authentication and privacy-preserving approaches, which have been implemented through cryptographic techniques used in blockchain 
. The most widely adapted approach could be identified as cryptographic encryption–based digital signatures for user verification 
. However, modern smart grids, which are to incorporate predictive data analytic tools for intelligent decision, are vulnerable to AI and ML–related attacks 
. These adversaries would overlay the security threats governing Smart Grid 2.0, which would require accelerating the existing research initiatives.
4. Ultimate Reliability and Stability
4.1. Key Challenges
How can we facilitate seamless integration of DG to achieve dynamic response in power output to rectify supply-demand mismatch 
Intermittent generation: What will the possibility be of securing stable and reliable grid operation in a decentralized architecture with no third-party involvement and heterogeneous grid interconnections?
Secure communication: How can we facilitate secure communication links to improve the exchange of energy data and control signals between peers to improve stability and reliability of a distributed grid?
Intelligent decision-making: How do we arrive at intelligent decisions for optimal generation allocation to improve grid stability management?
Energy theft: How can we prevent energy theft, ensuring the consumer a reliable energy supply?
Power quality management: How can we mitigate the issues related to non-compliance of power quality standards by the prosumer, DG owner and consumer?
4.2. Role of Blockchain
Blockchain and smart contracts can be utilized to invoke certification for seamless grid integration of renewable energy generation upon fulfillment of the prerequisites. This would decrease the latency and improve grid stability 
. Lightweight blockchain platforms would further provide means of increasing the transaction throughput of the energy grid. Efficient correction of supply-demand mismatch through precision decisions arrived upon predictive analysis would facilitate reliable future grids. Blockchains would be the means for secure storage and broadcasting of energy bids and demand requests, enabling the predictive analysis on aggregated EI data. The integrity of data sets used for the application of AI and ML technologies could be ensured through the cryptographic encryption methods incorporated with blockchain 
. The Spanish renewable initiative Iberdrola is utilizing blockchain for tracking of wind power generation and is expected to contribute in seamless grid integration of DER by issuing origin certification 
. The decentralized solution eliminates the need for a third party as a middle man.
Blockchain, however, cannot be identified as the ultimate solution for ensuring stability and reliability of futuristic decentralized grid architecture with no central authority. Advanced control strategies need to be proposed in this aspect, with blockchain facilitating them from the rear end by providing secure data transmission and storage. Deployment of smart contracts would enable autonomous execution of the control strategies for securing grid stability.
4.3. Future Directions
Grid stability and reliability are expected to progressively advance through AI and ML algorithms, enhancing the capabilities of the future grids beyond expectations. Deployment of Virtual Power Plants (VPP) and Autonomous Vehicles (AV) contributing in Vehicle-to-Grid (V2G) and Vehicle-to-Vehicle (V2V) interactions match the supply with the demand. Meanwhile, smart cities that integrate smart buildings with intelligent appliances would reshape the future grid architecture, with autonomous stability and reliability management approaches being a requisite 
Elimination of energy theft through double auction and other mechanisms has been presented in the existing literature 
. However, securing communication channels to offer a reliable electricity supply while addressing the challenges related to intermittent generation of DGs have research prospects for maximization of renewable energy utilization in Smart Grid 2.0 
. This would enable matching the supply with the demand; however, maintaining the power quality within the allowable limits 
would be another challenge to overcome in the envisaged grid architecture, with little research being carried out related to this aspect.
5. Decentralized Scalability
5.2. Role of Blockchain
Inherent features of blockchain enable the delegation of processing to edge nodes and cloud storage platforms through a trustless trust establishment. Low latency and high throughput can be achieved through distributed processing which further provides solutions to intermittent connectivity in remote areas in a store and forward manner. PETCON, a secure P2P trading mechanism for plug-in hybrid EVs 
, and Guardtime, a permissioned blockchain-based approach, are considered to be scalable solutions for complex data exchange in EI grids.
Further, with the incorporation of off-chains, side-chains, sharding and edge computing to offload the computational burden, better results have been obtained to cater for the increase of nodes connected with minimum trust issues. This would be further empowered with the speculated advancements in communication infrastructure by deployment of 6G technologies, enabling edge intelligence 
. DL and big data analysis applied on the data aggregated in cloud storage would be controlled and managed through a secure mechanism utilizing blockchain platforms. Leakage of information would be shielded and better control over the data can be achieved through such approaches 
Scalability issues related to EI grid implementations, however, have not been fully overcome through the blockchain platforms. These can be identified as future research prospects for the improvement of the applicability of blockchain for energy grid management.
5.3. Future Directions
Achieving distributed scalability at the cost of compromising information security and privacy of the user respectively defies the expectations of the future grids. The blockchain trilemma would receive attention in the pathway towards successful implementation of the futuristic grid infrastructure 
. The vulnerabilities in grid security would expand beyond the current extent with the scalable, decentralized operations. Thus, proper measures would be a necessity to instate the trust in the EI grids.
Further, connectivity has a major impact in reaching the scalability goals, which can be addressed through advancements such as 5G and 6G technologies. The former would enable low-latency communication channels while the latter would address the intermittent connectivity in rural geographical locations by utilizing edge devices 
Scalability of EI grids are partly addressed through off-chain and side-chain implementations 
as well as suitable selection of the blockchain platform 
. This includes scalable initiatives such as NRG Xchange 
and analytical selection of the blockchain platform such as HyperLedger over Ethereum 
. Energy data management with the increasing number of consumer connections would be a challenge to be addressed, while facilitating low-latency grid synchronization of DERs has not been discussed in the existing literature 
6. Advanced Big Data Management
6.1. Key Challenges
Data silos: How can we overcome data silos and establish trust between prosumers, microgrids and large power plants for better coordination?
How can we achieve secure communication channels between smart meters/smart sensor nodes and the Energy Management System (EMS) 
Secure data storage: How can we provide secure, privacy-preserving and scalable storage for the aggregated large data sets containing generation and consumption patterns of consumers and prosumers respectively?
Data integrity protection:
How can we ensure the integrity of the stored energy data utilized for AI model development, training, validation through ML techniques, testing and deployment 
Data ownership: How can we ensure ownership of the aggregated energy consumption/production pattern data to prevent privacy-violations arising from unauthorized trading of these sensitive data to a third party?
Scalable grids: How can we facilitate the management of large data volumes while offering scalabilty for grid expansion with numerous grid integration of prosumers, microgrids, EVs and collaborative consumers participating in DSIs?
6.2. Role of Blockchain
Communication technologies are progressing towards delivering an efficient, ultra-reliable, low-latency service to the consumer through 5G, thereby facilitating the data aggregation process. Further, 6G network operation enables delegation of the computational capabilities for the processing of information to multiple edge devices 
. Blockchain will be an integral part of both of these scenarios and its integration would enable secure data, control signal transmission and trust establishment for distributed processing using edge computing respectively 
Data aggregation further involves secure data storage on cloud-based platforms, in which blockchain would ensure cyber-security, information security and network security, preventing malicious attacks that could modify data. The hash function incorporated in blockchain ensures the integrity of the stored data. Moreover, privacy could be enacted, gaining control over the information through the deployment of smart contracts for data sharing while maintaining anonymity 
Predictive analysis upon the aggregated data will be facilitated through AI and ML, in which blockchains could contribute as a trusted mediator. This would ensure the integrity of the data incorporated and the algorithms compiled 
while deploying autonomous operations through smart contracts.
Blockchain alone, however, cannot facilitate scalable platforms for big data management. This demands alternative distributed storage platforms supported by the blockchain from the back end, which include utilization of off-chain storage and Inter-Planetary File System (IPFS).
6.3. Future Directions
Blockchains in collaboration with smart contracts could extend the utilization of the aggregated data set where individual stakeholders (consumer, prosumer and individual power producers) would trade the information to potential investors, asset managers such as financial institutions, utilities and policy makers to obtain financial benefits. The secure and transparent link will be established through 6G architecture and the blockchain platform 
Future EI grids will be integrated with AI and ML for predictive data analysis, giving rise to a new set of challenges which were not encountered in previous generations of smart grids 
. Blockchain integration with EI grids have facilitated data integrity protection through cryptographic hashing 
and is well addressed in the existing literature. However, addressing the challenges, including data silos 
, facilitating secure, scalable data communication and storage with privacy-preserving data ownership 
, would require further research attention.
7. Grid Intelligence
7.1. Key Challenges
Data manipulation: How can we mitigate manipulation of energy input data (electricity consumption and production data obtained through smart meters) and validate the authenticity of the information?
How can we prevent the model inversion, poisoning pertaining to training and deployment of ML models, used for adaptive decision-making processes in automated generation allocation of EI grids 
Ethical data aggregation: How can we ensure ethical use of aggregated energy production/consumption data for AI model training and prevent unauthorized data sharing with compliance to privacy preservation?
Transparency: How can we improve transparency in model development, training, testing and deployment, resulting in algorithms that are reliable for diverse applications with grid integration of heterogeneous energy sources?
How can we assure security in AI-based automation of network control and orchestration with it 
How can we establish trust among stakeholders participating in energy trading in EI grids and improve transparency in process automation through the deployment of AI models 
Accountability: How do we ensure the accountability of the AI algorithms for automated decision-making processes responsible for generation coordination, distribution network management and fault recovery?
7.2. Role of Blockchain
Intelligent grids arrive at decisions based on the aggregated data set obtained through IoT devices and incorporation of AI and ML techniques 
. The integrity of the data used in ML models and preventing data poisoning would be the key consideration, which directly impacts the decision-making process. With blockchain being an immutable DLT, this would ensure data integrity by preventing data manipulation, injection and corruption 
. Privacy preserving of the large data sets can be achieved through distributed edge computing, facilitated by blockchain platforms where raw data will remain closer to its origin, assuring confidentiality 
Nevertheless, the authenticity of the algorithms generated through the incorporation of AI and ML cannot be verified through the blockchain platform. Security and privacy preservation in AI based systems, however, cannot be fully facilitated by blockchain platforms and, thus, will require AI-resilient measures. At the same time, this would affect the control decisions obtained for the stable and reliable operations of the decentralized architecture. Standardization and regulatory enforcement are required to overcome such issues arising in the envisaged electricity grids.
7.3. Future Directions
The advancements in big data and computing technologies have facilitated the emergence of DL techniques for pattern recognition from the aggregated information 
. This would, thereby, enable accurate demand forecasting for optimal load scheduling and load balancing to reduce peak electricity demand, facilitate energy trading and sharing, state estimation of the power grid and perform grid diagnostics for the detection of energy theft. The higher precision achieved in the energy demand speculation would benefit in maximizing the utilization of DER to cater for the demand requirements while maintaining grid stability 
. Further, fast blockchain-based data-feeding models need to be introduced to facilitate the development of efficient and accurate ML models.
Initiatives towards collaborative model development approaches using AI and ML techniques have given rise to Federated Learning (FL), in which individual data storage on a decentralized network is encouraged 
. This facilitates predictive data analysis through training of a shared model using different data sets, offering capabilities of generalized model development with the benefit of privacy preservation of the user data. Blockchain facilitates trust establishment and prevents data poisoning, improving transparency in training, validation, testing, deployment and storage of training data sets for such shared models. Smart contracts enable the automation of iterative processes, improving the flexibility of model development and data analysis. However, challenges raised through the blockchain architecture relating scalability and smart contract security, need to be addressed further to expand the capabilities of FL techniques 
xAI (explainable AI) is gaining attention in the current context of AI integration with IoT for predictive data analysis. This could improve the transparency of the AI models associated with demand forecasting, evaluation of energy usage patterns, renewable energy modeling and automated grid operation with optimum generation allocation and load scheduling. xAI unravels the black box AI models, thereby improving transparency in predictive data analysis and establishing trust in the decisions arrived through such approaches 
Grid intelligence would dominate the future autonomous and the challenges arising from AI-integrated smart grids are seldom addressed through the existing literature 
. Improving transparency to ensure accountability of ML models 
and trust management in the decisions arrived through the models 
would be EI grid–specific challenges to be addressed in future research.