Conceptual framework for blockchain-based metering systems

Abstract

The smart grid environment requires the enhancement of various computational tools, especially for routine tasks of data acquisition and system monitoring. This paper presents the building blocks of a conceptual framework to be used as the basis for the construction of novel distributed remote metering systems with utilization of the cutting edge Blockchain technology. The proposed methodology is suitable for processing a large volume of data aimed at monitoring modern electric power distribution grids. As a proof of concept, a collaborative metering system is conceived based on the Blockchain technology, being primarily capable of: dealing with the entirety of the collected data (conveniently stored and filtered); assuring data integrity by means of cryptography; optimizing implementation/operation costs of the telecommunication services involved. Simulation results concerning the reliability and performance of the designed distributed remote metering system are presented.

Keywords

Blockchain collaborative networks power systems smart grids

1. Introduction

The electric power systems of the twenty-first century are changing gradually, with the incorporation of innovations to build smart grids (SGs) – efficient and reliable power networks that integrate a wide range of cutting-edge digital technologies, capable of automating many functionalities. Modernizing the grid can reduce the frequency/duration of power outages, restoring the system to service faster. In the smart grid arena, problems related to complex automation and monitoring systems, devoted to ensure the quality of service requirements, have been recurrently studied [1, 2]. Usually, utilities track distribution reliability indices to evaluate how the system is performing. This has driven the industry to find technical solutions to remote metering problems (e.g., capillarity, security, robustness). In addition, there is a clear consensus on the necessity to strengthen, transform, and improve the electricity infrastructure to ensure access to reliable, secure, and clean energy sources. To achieve this goal, newly emerging challenges– concerning the introduction of distributed energy resources (DERs), system security, and privacy of consumers’ information – are added to the conventional power quality and economic aspects. To address all the challenges present in the near-term agenda of experts, energy generation and distribution systems have been modernized and, to some extent, switched to smart grids.

There is an ongoing technology revolution in the power industry that can be considered substantially transformative. This is particularly related to the establishment of an integrated and pervasive communication network, covering the entire grid and supporting automated intelligent transactions that will make the new grid smart. So, one of the pillars of smart grids is the advanced metering infrastructure (AMI), which includes smart meters, communication networks, meter data management systems, and means to integrate the collected data into software application platforms/interfaces.

Smart grids use data collection and computer technologies to gather information on consumption by individual consumers, interpret it, and optimize operations to match different patterns of demand. Power companies are modernizing their grids in terms of their infrastructure, absorbing intelligent digital technologies, which includes automatic meter reading. A massive amount of data is collected and analyzed, being stored for reference and reuse. This is not an easy task, requiring high-performance database systems in conjunction with automated metering and other supporting technologies.

Considering that the smart grid effort will continue to evolve, this paper proposes a novel approach to deal with huge datasets collected for monitoring smart grids. New data transportation, management and information security schemes are proposed to improve the reliability of the remote metering process.

The architecture of a collaborative metering system is conceived based on the blockchain technology [3]. To date, no publication reporting blockchain-based metering systems (henceforth, referred to as BBMS) for monitoring smart grids is found in the technical literature. As a proof of concept, a BBMS is designed to be capable of managing all the collected data (conveniently validated, stored and filtered) from the power grid and handling data integrity and security using cryptography.

The remainder of this paper is organized as follows. Section 2 presents a literature review containing publications that represent important information and issues related to the smart grids and blockchain. Section 3 shows the integration of smart grid and blockchain systems. Basic elements of a blockchain metering system are presented in Section 4. Numerical results of simulation studies are included in Section 5. Finally, Section 6 presents the conclusions of the paper.

2. Literature review

This section indicates in a synthesized way the main specialized publications that represent the background and research developments related to smart grids: standards; advanced metering; monitoring; data management; cyber security; communication networks; challenges. In addition, references containing very useful information on blockchain and related issues are provided. The readers interested in a more detailed approach to smart grids can rely on the books of Momoh [4] and Borlase [5].

The AMI technologies, subsystems, elements and challenges are fully described by Mohassel et al. in [6] as well as the meter data management systems (MDMS) that presents different software platforms and interfaces based on the vendor’s market concept.

Hassan and Larik, in [7, 8], respectively, review basic concepts, technologies, benefits, and challenges related to smart grids especially about data security and specialist systems to support grid operation. The existing and recommended standards used for smart grids projects are found in [9, 10]. Vasconcelos emphasizes the potential benefits of smart meters and give a concise review of the legal framework for governing policies and metering activities in Europe [11]. An industry perspective for the smart distribution system is presented in [12] and identified the technologies that can be applied in future researches on the field. Yan et al. in [13] and Chen in [14] present a review on smart grid cyber security issues. Gungor discusses communication networks (architecture and requirements) for power systems in [15] and other researchers review the types of communication suitable for deployment in smart grids in [16]. They also examine aspects of network communication implementation and challenges in the power system settings. Wang et al. [17] shows the future with consumer participation and illustrates the internet as the main network backbone integrating substations, power plants and low voltage metering networks. A comprehensive survey exploring the technologies used in smart grids also can be found in [18].

Considering the blockchain technology, an interesting work presented in [19] provides a review on the architecture, key characteristics, and mechanisms involving its use. Similarly, one can find in [20, 21] overviews on blockchain, emphasizing big data areas and industrial applications. Finally, the systematic reviews presented in [22, 23] merit attention, since they contain a sort of topics that papers deal with when proposing the use of blockchain.

The IEEE vision for Smart Grid control can be found in [24]. In this respect, some issues are mandatory to the evolution of the energy grid control namely the increasing of information security, intelligent data management, direct meter access by consumers and demand management information using existing communication infrastructure. The blockchain methodology applied in power systems can help to design a better AMI system capable of addressing all these mandatory issues.

3. Integration of smart grids and blockchain systems

Smart Grids (SGs) are complex systems built for small, medium, large, isolated or interconnected networks. The electric sector is focusing on building smart grids that leverage digitally based advanced technologies capable of automating power system functions. In simple words, the emerging smart grid can be described as the convergence of power system and information technologies.

The main goals of data communication in smart grid projects go through availability, reliability, and security of collected data. The available technologies and some concepts applied in the data management and transportation problem are addressed below.

3.1 Standard advanced metering infrastructure (AMI) solutions

The AMI provides a computer-based sensor system extending from energy consumers to a service provider [25], enabling bi-directional communication. In addition, AMI allows the implementation of many important functions that were not previously available (or were performed manually), such as energy consumption metering (automatically and remotely), service connection/disconnection, detection of tampering in metering devices, identification/isolation of faulted equipment, and voltage monitoring.

To date, AMI communications connecting metering devices [26] to the utility’s control center are numerous: RF mesh [27], point-to-multipoint [28], general packet radio service, etc. The big mass of data to be transmitted impose a great challenge to communication systems. Hence, it is important to seek efficient solutions that adequately address requirements associated with real-time communication, as well as data integrity and security.

The smart grid’s metering systems use electronic devices called smart meters capable of measuring electric quantities (e.g., energy consumption, active/reactive power, etc.) and operating the supply circuit remotely. There are currently many ways of collecting data from smart meters. From the most basic methods with visual assessment, to the most advanced technologies, such as the use of smart meters in AMI networks. Figure 1 shows two typical AMI technologies used to collect data from smart meters, namely the wide spreaded RF Mesh (e.g., ZigBee) and Power Line Carrier (PLC) networks [29]. The PLC uses alternative current (AC) power line as data network carrying all information over the AC wave, working as a carrier wave. This technology still faces important issues about multiple non-interoperable standards in the same power line and mutual interferences between devices [30].

Figure 1.

Standard advanced metering infrastructure solutions.

Usually, smart grids collect measurement data from Smart meters using local collectors located along the cover area and distribute all data to the concentrators. The concentrators are connected to the data service providers using different technologies (e.g., 3G, 4G, WiMAX, GPRS, others) that will deliver the data packages to the specific data server.

Technically, the whole process is critical and there are many points of failure namely the collectors, concentrators, data service and data servers. The collectors and concentrators, usually, has no redundancy and hardware failure are not rare. There is no intelligence whatsoever in these devices that work as a communication gateway, consequently, no qualitative validations are possible.

Choosing an appropriate communication network is a difficult task that involves key aspects: the expressive amount of data transfer, restriction in accessing data, access to information of customer’s consumption, the authenticity of data, confidentiality of sensitive data, facilitation of future expansions, and cost-effectiveness. In addition, it is important to consider that large urban centers may have the so-called high-risk areas (often without adequate infrastructure) that are obstacles for meter reading and maintenance. Existent communication infrastructures, which contemplate internet hotspots and mobile cellular networks, should not be ignored since it provides a way of exchanging data efficiently. Smart meters can use mobile devices, in-vehicle communication networks [31], and wi-fi hotspots as gateways to transmit data from electric power distribution systems to control centers [32].

3.2 The blockchain approach

The blockchain technology entails the management of an intrinsically safe and decentralized database system [33], avoiding human interference during the transaction of receiving and delivering data blocks. Although blockchain has been implemented in very complex applications, such as cryptocurrencies (e.g., Bitcoin and Ethereum), machine-to-machine (M2 M) interactions [34], supply chain controls [35], and energy trades [36], the concepts involved are relatively simple. They are used to establish three main stages: data delivery, mining, and delivery of results. Initially, the source device sends data blocks with a unique identification (ID) and time stamp to the mining stage (carried out by miners), at which all incoming data are encrypted. The idea is to process these data blocks using the so-called Smart Contracts (SCs), in which there are preset rules and metrics to be satisfied. All data transactions performed are stored and their traceability and cryptography of the blockchain process ensure reliable data (free of external manipulation).

Idealized by Szabo [37], the smart contract is a computer program to validate established terms/con- ditions, mitigating the action of intermediaries. The smart contract is a powerful programmable tool used as a deterministic and independent process of the blockchain technology. The versatility of smart contracts is demonstrated for instance in systems concerning renewable energy trades [38] and Internet of Things (IoT) [39]. smart contracts are algorithms implemented into the miners, being a fundamental element of the framework proposed in this paper. They are responsible for verifying/qualifying all blocks sent by agents and smart devices in order to approve and address all data to the right target. At this stage, contractual clauses are checked and consistency analyses involving electrical measurements made. Each transaction performed will be appropriately recorded in accordance with the adopted smart contract.

Figure 2.

Blockchain-based metering systems approach.

Figure 2 shows the same remote metering problem presented in Fig. 1 in a proposed blockchain approach. Starting the diagram analysis in the smart meter side, all measurement data can be collected by the collaborative agents and/or smart meters by itself. Using existing communication interfaces such as wi-fi hotspots and cellular networks, the data flow to the blockchain miners. After mining process (refer to Section 4.3), the electrical measurement data go to the utility data servers. The control center can take all data and address it to the proper final client (e.g., operation, billing, maintenance, etc.).

The smart grids deal with a huge amount of data, sent remotely by smart meters, and requires protections against cyber threats. The data security in the blockchain methodology uses cryptography as the main tool, which can guarantee the data source, integrity and assertiveness required in smart grids.

Sending all data remotely measured to the clouded blockchain can reduce drastically the costs in the communication infrastructure, considering existent hotspots and cellular networks. This well-established network can help smart grids to achieve the desired robustness using multiple communication gateways (e.g., public/private hotspots and cellular antennas) and significant data redundancy, as will be explained in next sections.

One of the main differences between the BBMS and the standard AMI solutions presented is on the data processing and filtering. While there is no intelligence associated with the standard solution, the BBMS introduces a data validation process concept, discarding or approving the data block. This validation avoids duplicated data and has an effectible fraud detection for corrupted measurements blocks. The overall result is a metering system that delivery only useful data and report any undesired deviation to the utility. Consequently, the distribution management system (DMS), advanced distribution management system (ADMS) and meter data management systems (MDMS) can benefit from increased performance and quality in data processing.

4. Blockchain-based metering system (BBMS)

The smart grid success depends on the robustness and reliability of data communication related to energy consumption. Actually, not only energy measurements are useful. All electrical quantities available are significant for power system operation, especially when computational intelligence algorithms are implemented (e.g., alarm management, state estimation, etc.) [40, 41]. The BBMS is designed to support all kind of electrical quantities that integrate one single blockchain data block. All data blocks travel through the proposed system trying to reach to the final target (energy utility).

Figure 3 illustrates the complete cycle of a data block throughout a BBMS. Part 1 is related to the smart metering devices (data source) and any system agent – e.g., a person or vehicle into the communication range using a mobile device with a data collector application. Smart meter devices can send spontaneously their own data to the system, such as the moderns AMIs, or through these local agents. These devices and agents should have the capability to pack and send through the Interne all data blocks collected, working as gateways, addressing the data packages to all miners, as depicted in part 2. They use an application responsible for collecting, encrypting the data and sending packages upstream. Only miners can decrypt a data block and validate its contents.

Owing to the recent increase in the development of collaborative networks, the number of agents tends to grow and so the data reading redundancy (number of duplicated measurements). Such duplicated data should be detected, filtered, and discarded by the miners, avoiding over processing. Each miner processes the collected data using a given validation algorithm (refer to Section 4.3.1). As illustrated in Fig. 3 (parts 3 and 4), when validated by a mining process and its smart contracts, the accepted block is added into the chain of blocks. The block is delivered already encrypted (part 5), ensuring the information privacy [42], accordingly to the National Regulatory Authorities (NRAs), traceable to a particular destination (e.g., energy utility).

The BBMS created in the paper shares the miner structure among companies that cover a given area. The miners’ intelligence applied in the proposed solution is able to process data blocks that belong to any utility and send them to the aimed destination. A miner performs the validation tasks in the utilities using different smart contracts, with specific terms and conditions. Therefore, a collaborative system can be proposed enabling a cost sharing-based metering network to the open market of distribution systems.

Figure 3.

Blockchain steps of the Blockchain-based metering systems.

Figure 4.

Data block fields.

4.1 Data block fields

A blockchain system basically starts with the creation of data blocks. In the case of BBMS, each data block contains the set of fields relevant to the electrical measurement values information from the smart meter and provides the information needed for the blockchain mining process. All data blocks have exactly the same data structure that is sufficient to record the information coming from the measuring device and measuring agent.

The blocks mainly carry information necessary to be used by the energy utility specially in the billing area and for the operation and maintenance of the network. The Fig. 4 illustrates a block with all of its respective fields discussed in detail as to their importance and use throughout the process.

4.1.1 Security key

This field is provided by the source device, e.g., smart meter, RTU, etc. This work used the well-known 256 bits secure hash algorithm (Sha-256) [43], created by the national security agency of the United States (NSA), is usually adopted in key encryption. This algorithm creates a unique value for the Security Key field, which is used in the internal consistency security analysis and the chain of blocks cryptography, as will be seen in Section 3B. The cryptography is used considering the composition of the following fields in a single string:

$\displaystyle\textit{SecurityKey}=sha256(\textit{GIS},\textit{TimeStamp, SGEID% , DataCat, Data})$ (1)

where GIS is the smart grid element’s geographic location; Time Stamp refers to the source device; SGEID is a unique identifier for the smart grid element; DataCat is the data category (type), e.g., measurement or alarm; and Data contains all electrical measurement values and alarm information. Figure 5 illustrates the elements of the Security Key field.

4.1.2 Alarms and measurements

Measurement data are those obtained by a typical smart meter, such as voltage (V), current (I), active power (P), energy (E), and power factor (PF). The refresh rate of each measurement value depends on the metering device technology employed, varying between 1 s and 15 min (typical energy integration time). The alarms are composed by type (e.g., internal faults) and numerical values that represent the alarm description from the source or smart contract process.

Figure 5.

Security key field.

Figure 6.

Simplified data block workflow.

4.2 Data workflow

Data received by the miners have one or more targets, depending on the data category. If the block is a simple alarm, the target will be the control center and maintenance area. If the block is related to the measurements, it will be delivered to the control center and to the billing area. The data routing strategy is a utility duty, although the system will distribute each information in its own blockchain database as shown in Fig. 6.

The workflow illustrated in Fig. 6 can be represented as a pseudocode (Algorithm 1); all the way through the block processing, from the reading to the delivery (target final client) stage. The detailed algorithm of each mining stage will be presented in Section 4.3.

Algorithm 1: The BBMS workflow in pseudocode
1	Start
2	Reading: Metering available in smart meter
3	If smart meter has communication to blockchain
4	Then pack & send measurement/alarm data to the miners
5	If there is an agent close to the smart meter
6	Then collects, pack & send measurement/alarm data to the miners
7	Mining: Miner received new block (Mining Process)
8	Basic security validations (SGID, SecurityKey, etc.)
9	If block is valid
10	Smart contract verification (contractual and semantic)
11	Else
12	Reprove the block and save it to the processed block database
13	If all miners approved the block
14	If all blockchain data base are equal
15	add it to the respective blockchain data base(Measurement/Alarm)
16	Else
17	Reprove the block and save it to the processed block database
18	Else
19	Reprove the block and save it to the processed block database
20	End
21	Delivery: Utility identify new block(s)
22	Collects new block(s) from respective blockchain data base
23	Send block to the target area (billing, control center, maintenance, etc.)
24	End

The control center can use the collected data to feed the operational SCADAs (ADMS/DMS), MDMS or specialist algorithms, such as alarm analysis, state estimators [44], or outage simulators.

4.3 Mining process

The miners usually demand large computational resources frequently exploring parallelism, due to the high-performance computing used in cryptocurrencies and other applications. In the proposed BBMS, the algorithms follow some rules with low computational complexity aimed to achieve the best performance as possible and deliver data in a shorter time to the energy utilities. BBMS is intended to be safe and resilient, avoiding frauds, communication problems, and cyber invasions. To achieve these goals, a cross-check mechanism verifies the security issues, as well as the contents of each block received. This process works with encrypted data and has two parts: (i) each miner validates the incoming block; (ii) later, if the block was approved, it is sent to the other miners with an approval flag. All miners approve the block if its encrypted codes (hashes) match perfectly. The regular path of any block that enters the blockchain system is the cross verification of miners followed by the data delivery to the targeted utility.

4.3.1 Mining algorithm

Three main modules integrate the mining algorithm: mining process, smart contract analysis, and blockchain creation/update. As shown in Figs 5 and 6, the mining logic is divided into two parts (Process 1), at which the following steps are performed:

Process 2. Mining algorithm

Part 1 (Fig. 7): Each miner validates the incoming block

a)
The new block condition (pending or rejected) is checked to avoid reprocessing;
b)
The block is processed, only if it came from the original source (i.e., a smart meter); otherwise, it will be rejected avoiding fraud or error propagation through the mining process;
c)
The validity of some fields is checked. The Security Key will be compared with an internal hash consistency calculation and it has to match perfectly;
d)
According to the target utility the block is addressed;
e)
The block is checked to see if it obeys the corresponding rules of the smart contract;
f)
The accepted block is added to a pending list;
g)
Proceed to Part 2.

Part 2 (Fig. 8): If the block was approved, it is sent to the other miners with an approval flag

a)
Verification of the approval flags and block hashes;
b)
The calculated hashes (even those from other miners) are compared to each other to verify if there are differences that represent some fraud or error in the mining process;
c)
The block is sent to the other miners until all of them approve it. This cross-check procedure guarantees that all miners will receive the block directly from the source device or through a miner that has already analyzed it;
d)
The accepted block will be sent to the blockchain of the target utility.

Figure 7.
Mining algorithm – Step 1.

Figure 8.
Mining algorithm – Step 2.

4.3.2 Smart contract algorithm

The BBMS is not a generic blockchain system – its mining process is totally applied to the power system area. The smart contracts offer this significant difference using the contractual and semantic verification that fit to energy market constraints.

Smart contracts algorithms are implemented in the miners and they are a fundamental part of the metering system. They are responsible for validating all terms, conditions, and semantic constraints on the specific application, in order to approve blocks sent by agents and/or smart devices. The approved blocks are addressed and sent to the target utility. The transactions are recorded in the system in accordance with the rules of the smart contract.

Figure 9.

Smart contract algorithm.

The smart contract algorithm is described by Process 2 and illustrated in Fig. 9.

Process 2. The smart contract algorithm

The category of the block is checked (measurement/alarm);

The block passes through a semantic verification, in order to check if the collected values are coherent (from the electric and application viewpoints). If the block does not satisfy some of the predefined constraints, it is not rejected, returning to its destination tagged as no conform;

The block passes through contractual terms and conditions evaluation, such as those related to contracted power, interruption of power supply, etc.;

Evaluation of the data distribution permissions for the billing, operation, and maintenance areas. Definition of which block goes to which area;

If the block is not of a valid type (alarm/measurement), it is rejected, and flagged with the corresponding time stamp.

4.3.2.1 Semantic verification

The semantic analysis consists in the evaluation of all electrical constraints and rules of the specific application. If some constraints are not satisfied at all or are only partially satisfied, flags will be issued. When this happens, the flagged situations are registered into “Data” field (alarm value) and the block is sent to its destination with such information. It is possible to find one or more of the following results:

•

Approved with no restrictions;

•

Discrepancy between phase $a$ and $b$ amplitudes, based on Maximum Allowable Inter-Phase Discrepancy (MAIPdis), is evaluated by:

$\displaystyle\textit{MAIPdis}<abs(V_{a}-V_{b})$ (2)

•

Detection of the zero voltage value (ZVV), ${ZZV\geqslant V}_{x}$ ;

•

Detection of the zero current value (ZCV), ${ZCV\geqslant I}_{x}$ ;

•

Power discrepancy based on maximum allowable power measurement deviation (MAPMdis):

$\displaystyle\textit{MAPMdis}<\textit{abs}(P-(V*I*PF))$ (3)

•

Energy above seasonal profile based on maximum allowable energy measurement discrepancy (MAEMdis):

$\displaystyle\textit{MAEMdis}<abs(\textit{kWh}-\textit{EnergyProfile})$ (4)

where $V$ , $I$ , $P$ , PF, and kWh are the voltage, current, active power, power factor, and measured energy consumption, respectively.

With this semantic verification, measurement deviations from a normal operation state can be identified and informed automatically to the utility, without any human intervention. Other analysis using data availability can be done in smart contracts with attention to the miner’s hardware performance and the impact on the time to deliver the block [45].

4.3.2.2 Contractual verification

The contractual terms and conditions can contain as many clauses as necessary to validate a block. Once the metering device is known (SGEID), even a particular analysis for each kind of client (e.g., residential, commercial, etc.) can be implemented. This verification uses two basic constraints:

•

Power values exceed the contracted power, $P_{meas}>P_{contract}$ ;

•

Interruption of the power supply, $P<$ min-value.

4.4 Construction of the block’s chain

The chain of blocks is created by adding into the chain all unanimously approved blocks if the hash consistency has already been verified. The chain of blocks consists of a limited threaded list that stores all elements until a pre-defined maximum number of elements (defined by the user) has been reached. The steps of the blockchain algorithm are described by Process 3 and illustrated in Fig. 10.

Figure 10.

Block chaining creation/update algorithm.

Figure 11.

Building modules of a Blockchain-based metering systems simulator.

Process 3. Blockchain algorithm

Only new blocks approved by all miners are processed;

The blockchain hash is calculated based on the composition of the last blockchain hash available and the Security Key field of the new block in the chain. The last block contains the last blockchain hash value that is a sha256 function:

$\displaystyle\text{BChash $=$ sha256}\left(\text{Security Key, BChash}\right)$ (5)

A new crosscheck is performed with the blockchain hash value of all miners before creating/updating its own blockchain. If the hashed values match, the accepted block is added in the respective blockchain and it is no longer a pending block;

All databases follow the First in First out (FIFO) method to maintain the storage size limit requirements;

The oldest block is eliminated when a new block is added.

5. Results

This section describes the features of a simulator developed to illustrate the use of the proposed framework for BBMS, implemented using Matlab™ on a PC, Core™ i7-2600, 3.40 GHz, 32Gb RAM. The three main modules of the simulator are shown in Fig. 11: the measurement generator is the first one, followed by the mining module, and finally, the report module that simulate the utility. With this simulator it is possible to assess the performance of the proposed methodology (as a proof-of-concept) regarding block processing times and potential fraudulent data/attackers. It provides the full range of functionalities needed to evaluate the algorithm performance and the security of measurement data within an AMI configuration.

Figure 12.

Block path: Send/receive delay.

Figure 13.

Block path: Fraud in miner B.

5.1 Illustrative examples

A BBMS was designed to be safe (avoiding the propagation of frauds) and resilient to communication problems or punctual cyber invasions. Three scenarios simulating the occurrence of frauds and systemic problems were created.

In a normal workflow, only the block sent by the source (metering device) become a pending block in each miner where the Security Key is verified. The first simulated scenario (refer to Fig. 12) is related to the occurrence of a communication delay between the source and miner C. The block is delivered to miner C after being processed by miners A and B.

The block is sent to miner C by miners A and B. In this case, C ignores the block because it is not present in its pending list (unknown Security Key). When the block is received from the source, miner C approves it and send it to miners A and B, where it has been already approved. The process concludes successfully with a delay. The three blockchains were updated with the same BChash and are ready for the utility.

The second scenario, illustrated in Fig. 13 represents a cyber-invasion into miner B by introducing manipulated values in a given measurement (Data field). In the frauded miner B, the SecurityKey is recalculated with the corrupted values.

Figure 14.

Block path: Fraud in miner B and block data.

The miner tries to disseminate the frauded block among the other miners. Miners A and C will ignore the frauded block because it is not registered in their pending lists (unknown Security Key). The block from the meter (source) is then received by these miners that approve it and send it to miner B.

Due to the differences in Security Keys readings by all miners, the result of this miner corruption is a corrupted block approved unanimously only by miner B. With that, the fraud is not propagated. Miner B blockchain will have a suspicious BChash generated that not match to the other miners, turning it into an alarm situation and causing the elimination of this miner from the mining process.

In the third scenario, illustrated in Fig. 14, a fraud is simulated in miner B and into the block Data field. This fraud changes the measurement values from the data package without any changes in the Security Key generated by the smart device. As a result, only the corrupted miner B will approve the block, because the other miners immediately rejected it after verifying the security consistency. Then, the frauded block is discarded and not propagated to the utility.

5.2 Numerical results

In medium or big cities, the typical low voltage distribution secondary substations can serve hundreds of residential and commercial clients. In a real-life scenario, data readings can be sent directly by the smart metering devices and/or collected through mobile devices used by any kind of agent. The tests imposed a similar number of metering data to the simulated avalanche scenarios, face to the real-world possibilities. Each block has its own unique Security Key and consequently, each block is processed as a unique information.

Tests were performed considering four avalanche scenarios and five sets of miners, each of them with a different number of miners available (3, 5, 7, 9, and 11). The avalanche scenarios considered 100, 1,000, 5,000 and 10,000 blocks for each one of the five sets of miners. The blocks were sent at the same time by the simulated agent, which represents smart metering devices and/or any other agent collecting data with a mobile device. This volume of blocks reflects a real-life scenario in which multiple readings are taken from meters that belongs to a single secondary substation area.

The processing time for each block was depicted in Fig. 15. The broadcast time between the source and the miners was not considered because the performance analysis was focused on the mining process.

The total block processing time (Tt) is composed by the mining crosschecking time Tc, the smart contract time Tsc, and the blockchain update time Tbc:

$Tt=Tc+Tsc+Tbc$ (6)

Table 1

Miners x block processing time (s)

Miners	Min (100 pts)	Max (10 kpts)	Average
3	0.0143	0.0960	0.0271
5	0.0249	0.1326	0.0480
7	0.0378	0.1969	0.0732
9	0.0546	0.2776	0.1103
11	0.0606	0.3386	0.1246

Figure 15.

Miners x block processing time(s).

The time to process each block observed in the simulations when considering each set of miners is shown in Table 1. It is a patent fact that the larger the database the longer the time to process a block. This is because the query time to find a block into the table of stored blocks has an impact on Tc and Tbc; fortunately, unimportant in professional database implementations – meaning that the average block processing times (Table 1) in the 100 and 10 k points of avalanche scenarios are close.

Due to the crosscheck algorithm, the processing time is directly affected by the increase in the number of miners due to the $O(n^{2})$ algorithm order. Intuitively it would be reasonable to think that an increase in the number of miners would increase the safety and resilience with respect to the process of corrupted data. However, the proposed BBMS was developed in such a way that it does not depend on the number of miners. The BBMS functionality can be guaranteed even if only 3 miners are available. Although a cyber-invasion is easier over 3 miners than 11 miners, the cost-benefit must be analyzed to take a decision on the number of miners to employ.

As shown in Table 2 and Fig. 16 the average time to process one block is 0.2052 s when an avalanche of 10,000 blocks is considered and 11 miners are available. In such case, the time expected to deliver (Td) the last block into the blockchain would be:

$\displaystyle Td=0.2052*10,000=2,052∼{}{\rm s}\sim 34∼{}{\rm min}$ (7)

On the other hand, when considering the same scenario with 3 miners, the time expected to deliver the last block into the blockchain would be:

$\displaystyle Td=0.0440*10,000=440∼{}{\rm s}\to 7.3∼{}{\rm min}$ (8)

Figures 15 and 16 show the number of blocks that can be processed by the BBMS simulator for the scenarios in which avalanches of 100 blocks and 10,000 blocks were considered. The integration time of energy consumption is typically in periods of 15 minutes. In the best scenario (Fig. 17), considering a 15 minutes time window, the simulated BBMS processed almost 13,000 blocks with 11 miners and more than 55,000 blocks with 3 miners.

Table 2

Avalanche scenarios x average block processing time (s)

Numb points	3	5	7	9	11
100	0.0162	0.0276	0.0403	0.0617	0.0695
1,000	0.0186	0.0333	0.0514	0.0907	0.0881
5,000	0.0296	0.0519	0.0797	0.1216	0.1354
10,000	0.0440	0.0790	0.1214	0.1671	0.2052

Figure 16.

Avalanche scenarios x average block processing time (s).

On the other hand, in the worst scenario (Fig. 18), when considering a 15 minutes time window, the simulated BBMS processed more than 4,300 blocks with 11 miners and more than 20,000 blocks with 3 miners.

Figure 17.

Scenarios (best case) x average block time (s).

Figure 18.

Scenarios (worst case) x average block time (s).

5.3 General comments

The results obtained for the four avalanche scenarios were obtained using a simple platform developed in Matlab. The BBMS performance in such conditions cannot be considered good enough to be used in real time operation. However, it can be useful in the ADMS/DMS to feed specialist programs (i.e., state estimator, outage analysis, etc.). Additionally, the billing area can receive all measures to process automatically all bills without any human interference. It is also possible to benefit analysis aimed to solve common problems such as interruption of power supply and theft of energy [46], which are relevant for the maintenance point of view.

It is possible to envisage the application of the proposed BBMS algorithm to real-time monitoring if it is processed using powerful computational resources. In such a situation, even the operation of the low voltage system could be benefited.

System robustness can be improved with more agents collaborating, reducing the risk of having lacks of readings. Reward strategies can be created to motivate agents to participate in the collaborative network [47], by becoming active reading agents with their mobile devices. This action can increase the number of measurements collected and the quality of the system observability.

6. Conclusion

The use of blockchain in smart grids is a greenfield that demands deep multidisciplinary studies. This paper proposed an innovative approach, based on reliable blockchain-based metering system applied to smart grids. It allows that all resources between the utilities that serve the same area can be shared, scaled, and clustered, according to the electrical network topology. Basic concepts were used to demonstrate the applicability of blockchain methodology, as a solution to be explored in the remote metering of distribution systems. The main characteristics of the proposed framework are:

•

Metering coverage increased in areas of poor communication infrastructure;

•

Sharing of common resources between utilities;

•

The possibility of basic or advanced pre-analysis using special algorithms in smart contracts;

•

Data security and integrity from the source device to the utility.

Footnotes

Authors’ Bios

Eric Zanghi received the B.Sc. degree in electrical engineering from the CEFET-RJ, M.Sc. degree in computer science, from the Fluminense Federal University, Niterói, Brazil and doing his PhD in the same university. He was entrepreneur and worked in Automation market for more than 22 years. Worked at GE Digital Energy (2012–2016) as Commercial Manager and since 2016, he has been working at INESC TEC (Portugal) as Researcher in Power Systems (eric.zanghi@inesctec.pt). His research interests include Smart Grids, dynamic simulation and DMS software.

Milton Brown Do Coutto Filho received the D.Sc. degree in electrical engineering from the Federal University of Rio de Janeiro, Brazil. His employment experience includes the position of an Associate Professor at PUC/Rio (EE Dept., 1977–1994), and a Visiting Scholar at the Northeastern University (Electrical and Computer Engineering Department, 1992–1993), Boston, MA, USA. Since 1994, he has been at the Fluminense Federal University, Niterói, Brazil, where he is currently a Professor (mbrown@ic.uff.br). His research interests are computer applications in power system analysis.

Julio Cesar Stacchini de Souza received the D.Sc. degree in electrical engineering from the Catholic University of Rio de Janeiro, Rio de Janeiro, Brazil, in 1996. From 1988 to 1989, he was with General Electric Company, Fairfield, CT, USA. Since 1992, he has been with Fluminense Federal University, Niterói, Brazil, where he is currently an Associate Professor (julio@ic.uff.br). In 2014–2015, he was an Academic Visitor with Imperial College London, London, U.K. His current research interests include power system and state estimation, computer methods applied to power systems and intelligent system applications to power systems.

References

Clarizia

Gallo

Landi

Luiso

and Rinaldi

, Smart meter systems for smart grid management, (2016), IEEE Int Instrumentation and Measurement Technology Conf Proc Taipei, Taiwan, (May 2016), 1–6.

Fan

et al., Smart grid communications: Overview of research challenges, solutions, and standardization activities, in: IEEE Communications Surveys & Tutorials 15(1) (2013), 21–38.

The Economist, Blockchain: The next big thing, Special Report (May 2015), Available at: https://www.economist. com/special-report/(2015)/05/09/the-next-big-thing.

Momoh

, Smart grid: Fundamentals of design and analysis,IEEE Press Hoboken, NJ: Wiley, 2012.

Borlase (editor)

, Smart grids: Infrastructure, technology, and solutions, Boca Raton, USA: CRC Press 2013.

Mohassel

R.R.

Fung

A.S.

Mohammadi

and Raahemifar

, A survey on advanced metering infrastructure, International Journal of Electrical Power & Energy Systems Volume 63 (2014), 473–484.

Hassan

and Radman

, Survey on smart grid, Proc of the IEEE SoutheastCon 2010 Concord, USA (2010), 210–213.

Larik

R.M.

Mustafa

M.W.

and Qazi

S.H.

, Smart grid technologies in power systems: An overview, Research Journal of Applied Sciences – Engineering and Technology 11(6) (2015), 633–638.

Samuelsson

Hemmingsson

Nielsen

A.H.

Pedersen

K.O.H.

and Rasmussen

, Monitoring of power system events at transmission and distribution level, IEEE Trans Power Syst 21(2) (2006), 1007–1008.

10.

Rohjansand

Uslar

Bleiker

González

Specht

Suding

and Weidelt

, Survey of smart grid standardization studies and recommendations, IEEE SmartGridComm, Gaithersburg, (Oct 2010), 583–587.

11.

Vasconcelos

, Survey of regulatory and technological developments concerning smart metering in the European Union electricity market, EUI RSCAS PP (2008), Available at: http://cadmus.eui.eu/handle/(1814)/9267.

12.

Brown

H.E.

and Suryanarayanan

, A survey seeking a definition of a smart distribution system, 41st North American Power Symposium, Starkville, USA, (Oct 2009), 1–7.

13.

Yan

Qian

Sharif

and Tipper

, A survey on cyber security for smart grid communications, IEEE Communications Surveys & Tutorials 14(4) (2012), 998–1010.

14.

Chen

T.M.

, Survey of cyber security issues in smart grids, Cyber Security Situation Management and Impact Assessment II; and Visual Analytics for Homeland Defense and Security II (part of SPIE DSS 2010), Orlando, USA, (April 2010), 77090D-1–77090D-11.

15.

Gungor

V.C.

and Lambert

F.C.

, A survey on communication networks for electric system automation, Computer Networks 50(7) (2006), 877–897.

16.

Akyol

B.A.

Kirkham

Clements

S.L.

and Hadley

M.D.

, A survey of wireless communications for the electric power system, Report of Pacific Northwest National Laboratory (2010), 1–73.

17.

Wang

and Khanna

, A survey on the communication architectures in smart grid, Computer Networks 55(15) (2011), 3604–3629.

18.

Fang

Misra

Xue

and Yang

, Smard grid – a new and improved power grid: A survey, IEEE Communications Surveys & Tutorials 14(4) (2012), 944–980.

19.

Zheng

Xie

Dai

H.-N.

Chen

and Wang

, Blockchain challenges and opportunities: A survey, Int J Web and Grid Services 14(4) (2018), 352–375.

20.

Karafiloski

and Mishev

, Blockchain solutions for big data challenges: A literature review, IEEE EUROCON (2017) – 17th International Conference on Smart Technologies, Ohrid, Macedonia, (July 2017), 763–768.

21.

Ahram

Sargolzaei

Daniels

and Amaba

, Blockchain technology innovations, 2017, IEEE Technology & Engineering Management Conference (TEMSCON), Santa Clara, USA, (June 2017), 137–141.

22.

Conoscenti

Vetrò

and De Martin

J.C.

, Blockchain for the internet of things: A systematic literature review, 2016, IEEE/ACS 13th Int Conf of Comp Systems and Applications (AICCSA), Agadir, Morocco (2016), 1–6.

23.

Yli-Huumo

Choi

Park

and Smolander

, Where is current research on blockchain technology? – A Systematic review, PLoS ONE 11(10) (2016), 1–27.

24.

Annaswamy

, IEEE Vision for Smart Grid Control: (2030), and Beyond Roadmap, in: IEEE Vision for Smart Grid Control: (2030), and Beyond Roadmap (Oct 2013), 1–12.

25.

Yan

Qian

Sharif

and Tipper

, A survey on smart grid communication infrastructures: motivations, requirements, and challenges, IEEE Communications Surveys & Tutorials 15(1) (2013), 5–19.

26.

Zheng

Gao

D.W.

and Lin

, Smart meters in smart grid: an Overview, (2013), IEEE Green Technologies Conference (GreenTech), Denver, USA, (April 2013), 57–64.

27.

White

, Mesh Networks, Motorola Inc. (white paper) (2006), Available at: http://www.motorola.com /innovators/ pdfs/mesh-ntwks-wp-7.24.06.pdf.

28.

Kim

Fonseca

Dutta

Tavakoli

Culler

Levis

Shenker

and Stoica

, Flush: A reliable bulk transport protocol for multihop wireless networks, Proc SenSys ’07, Sydney, Australia (2007), 351–365.

29.

Berger

L.T.

Schwager

and Garzás

J.J.E.

, Power Line Communications for Smart Grid Applications, Journal of Electrical and Computer Engineering 2013 Article (3) (2013), 1–16.

30.

Galli

Scaglione

and Wang

, Power line communications and the smart grid, IEEE SmartGridComm, Gaithersburg, (Oct 2010), 303–308.

31.

Bronzi

Frank

Castignani

and Engel

, Bluetooth low energy for inter-vehicular communications, (2014), IEEE Vehicular Networking Conference (VNC), Paderborn, Germany, (Dec 2014), 215–221.

32.

Intanagonwiwat

Govindan

and Estrin

, Direct diffusion: A scalable and robust communication paradigm for sensor networks, Proc IEEE/ACM MobiCom, Boston, USA, (Aug 2000), 56–67.

33.

Peck

M.E.

, Blockchains: How they work and why they’ll change the world, IEEE Spectrum 54(10), 26–35.

34.

Sikorski

J.J.

Haughton

and Kraft

, Blockchain technology in the chemical industry: machine-to-machine electricity market, Applied Energy Journal 195 (2017), 234–246.

35.

Tian

, An agri-food supply chain traceability system for China based on RFID & Blockchain technology, (2016), 13th International Conference on Service Systems and Service Management (ICSSSM), Kunming, China, (June 2016), 1–6.

36.

Mengelkamp

Notheisen

Beer

Dauer

and Weinhardt

, A Blockchain-based smart grid: towards sustainable local energy markets, Computer Science – Res and Devel 33(1–2) (2018), 207–214.

37.

Szabo

, Formalizing and securing relationships on public networks, First Monday 2(9) (1997), Available at: http:// firstmonday.org/ojs/index.php/fm/article/view/548/469.

38.

Rutkin

, Blockchain-based microgrid gives power to consumers in New York, Daily News, (March 2016), [Online], Available at: https://www.newscientist.com/ article/2079334-blockchain-based-microgrid-gives-power-to-consumers-in-new-york/.

39.

Christidis

and Devetsikiotis

, Blockchains and smart contracts for the internet of things, IEEE Access 4 (2016), 2292–2303.

40.

Fallah

S.N.

Deo

R.C.

Shojafar

Conti

and Shamshirband

, Computational intelligence approaches for energy load forecasting in smart energy management grids: State of the art, future challenges and research directions, Energies 2018 11(3) 596.

41.

Santofimia-Romero

M.J.

del Toro-García

and López-López

J.C.

, Artificial intelligence techniques for smart grid applications, CEPIS Upgrade, Green ICT: Trends and Challenges 12(4) (2011), 41–44.

42.

Dimitriou

and Karame

, Privacy-friendly tasking and trading of energy in smart grids, SAC’13: ACM Proc of the 28th Annual ACM Symp Applied Computing, Coimbra, Portugal, (March 2013), 652–659.

43.

Courtois

N.T.

Grajek

and Naik

, Optimizing SHA256 in bitcoin mining, Proc Int Conf Cryptography and Security Systems (CCSS 14), Lublin, Poland, (Sept 2014), 131–144.

44.

Gomez-Exposito

Abur

de la Villa Jaen

and Gomez-Quiles

, A multilevel state estimation paradigm for smart grids, Proceedings of the IEEE 99(6) (2011), 952–976.

45.

Taylor

M.B.

, The evolution of bitcoin hardware, IEEE Computer 50(9) (2017), 58–66.

46.

Coma-Puig

Carmona

Gavaldà

Alcoverro

and Martin

, Fraud detection in energy consumption: A supervised approach, (2016), IEEE Int Conf Data Science and Advanced Analytics (DSAA), Montreal, Canada, (Oct 2016), 120–129.

47.

Camarinha-Matos

L.M.

, Collaborative smart grids – a survey on trends, Renewable and Sustainable Energy Reviews 65 Elsevier (2016), 283–294.