Communication Networks and Interoperability

Introduction

A smart grid is not only an electrical infrastructure; it is also a large distributed information system. Sensors, intelligent electronic devices (IEDs), smart meters, protection relays, distributed energy resources, control centres, electric vehicles, and customer devices must exchange information accurately and securely. Communication networks provide the transport mechanism, while interoperability standards define how devices represent, interpret, and act on exchanged information.

Traditional power systems used relatively isolated, proprietary communication systems. Smart grids require more open and integrated communication because electricity generation is increasingly distributed, loads are more controllable, and operational decisions depend on measurements collected across wide geographical areas. Communication failures can therefore affect reliability, market operation, protection, and consumer services.

This chapter compares wired and wireless networks, explains home-area, neighbourhood-area, and wide-area network structures, introduces Internet of Things (IoT) connectivity, and examines latency, reliability, bandwidth, and quality-of-service requirements. It also discusses IEC 61850, DNP3, Modbus, IEEE 2030, and the Common Information Model (CIM).

Learning Objectives

After studying this chapter, the reader should be able to:

1. Distinguish wired and wireless communication technologies used in smart grids.
2. Explain HAN, NAN, and WAN architectures and their typical applications.
3. Relate smart-grid applications to latency, reliability, bandwidth, and quality-of-service requirements.
4. Describe the functions and limitations of IEC 61850, DNP3, Modbus, IEEE 2030, and CIM.
5. Explain why interoperability requires agreement on communication, information, and semantic layers.
6. Evaluate a communication architecture for a practical distribution-grid application.

1. Communication Requirements of the Smart Grid

A smart-grid communication system must support several classes of traffic:

• Protection traffic: trip, interlocking, and fault-clearing messages. These require very low delay and high dependability.
• Supervisory control: measurements, status indications, and control commands exchanged with substations and control centres.
• Metering traffic: periodic energy readings, tariff information, remote connection and disconnection, and power-quality data.
• Asset monitoring: transformer temperature, dissolved-gas analysis, vibration, partial discharge, and equipment alarms.
• Consumer and IoT traffic: smart appliances, electric-vehicle chargers, photovoltaic inverters, and home energy-management systems.

Network design is therefore an engineering trade-off. A system may need high availability but only modest data rates, or high bandwidth but less stringent latency. For example, a relay trip message may contain only a small number of bits but must arrive predictably, whereas a substation video stream requires substantial bandwidth but may tolerate a delay of several hundred milliseconds.

Important performance measures include:

• Latency: elapsed time between transmission and successful reception, normally measured in milliseconds.
• Jitter: variation in packet delay.
• Throughput: useful data delivered per unit time, measured in bit/s.
• Packet-loss ratio: fraction of transmitted packets that are not successfully delivered.
• Availability: proportion of time during which a service performs its intended function.
• Reliability: probability that a communication function succeeds under specified conditions.
• Scalability: ability to expand the number of devices and traffic sources without unacceptable degradation.

For a communication service observed over a period T, an operational availability estimate is:

where A is dimensionless availability, T is the observation period, and T_down is the total time during which the service is unavailable. The formula assumes that downtime can be identified consistently and that the observation period represents normal operating conditions. If availability is expressed as a percentage, A is multiplied by 100.

A useful reliability model for a component with a constant failure rate λ is:

Here, R(t) is the probability of surviving without failure until time t, λ is the constant failure rate in failures per unit time, and t is time. This exponential model is appropriate only when the failure rate is approximately constant; it does not adequately describe aging equipment with increasing failure rates.

For independent components connected in series, where every component must operate, system reliability is:

For two independent redundant components connected in parallel, where either component can provide the service:

These expressions illustrate why redundant communication paths can improve dependability, although real networks may contain common-cause failures that violate the independence assumption.

2. Wired and Wireless Networks

2.1 Wired communication

Wired networks use a physical transmission medium. Common media include copper twisted pair, coaxial cable, and optical fibre.

Twisted-pair Ethernet is economical and widely available in substations, buildings, and control rooms. It can carry industrial Ethernet and Power over Ethernet in suitable installations, but electromagnetic interference, grounding differences, cable length, and environmental conditions must be considered.

Coaxial cable has good shielding and was historically used in industrial and utility networks. It is less common in new general-purpose installations than Ethernet over twisted pair or fibre.

Optical fibre provides high bandwidth, electrical isolation, resistance to electromagnetic interference, and long-distance transmission. It is particularly suitable for substations, transmission corridors, and wide-area utility backbones. Its disadvantages include higher installation cost, the need for optical transceivers, and specialized maintenance.

Wired networks usually provide predictable performance and strong physical security. However, cables can be damaged by excavation, storms, fire, rodents, or construction activity. A wired network is not automatically secure: unauthorized access to an Ethernet port can be as serious as interception of a wireless signal.

2.2 Wireless communication

Wireless networks use radio communication and avoid the need for a continuous physical cable. Technologies include Wi-Fi, cellular networks, IEEE 802.15.4-based low-power networks, mesh radio, microwave links, and utility-oriented radio systems.

Wireless systems are attractive where devices are geographically dispersed, cabling is expensive, or rapid deployment is important. They can support mobile assets, temporary installations, and large numbers of customer devices. Their limitations include radio interference, changing propagation conditions, shared-spectrum contention, obstruction, limited battery life, and greater exposure to eavesdropping and jamming.

A wireless mesh can improve coverage because devices forward traffic for one another. Nevertheless, additional hops may increase delay and create alternative failure modes. Cellular networks offer broad coverage and professionally managed infrastructure, but the utility may have limited control over congestion, routing, or service outages.

2.3 Comparison

The choice should be based on the required service, not on the medium alone. Hybrid architectures are common: fibre may form the utility backbone, cellular networks may connect remote sites, and short-range wireless links may connect sensors inside a facility.

3. HAN, NAN, and WAN Architectures

3.1 Home-area network (HAN)

A HAN connects devices within a customer premises. Typical devices include smart meters, thermostats, controllable appliances, photovoltaic inverters, batteries, electric-vehicle chargers, and a home energy-management gateway.

The gateway may translate between local protocols and the utility or aggregator interface. HAN traffic is generally low power and moderate in volume, but privacy and usability are important. A customer should not be required to understand complex network configuration merely to operate an appliance or charger.

HAN technologies may include Wi-Fi, Ethernet, Bluetooth Low Energy, Zigbee, Thread, and other low-power protocols. The correct selection depends on range, power consumption, interoperability, cybersecurity, and the required control response.

3.2 Neighbourhood-area network (NAN)

A NAN aggregates traffic from multiple premises or field devices. In advanced metering infrastructure, it may connect smart meters to a data concentrator, access point, or utility communication gateway. NANs may use RF mesh, cellular, fibre, power-line communication, or combinations of these technologies.

NAN design must account for pole-top equipment, transformer locations, urban density, terrain, interference, and the number of devices per concentrator. It must also support bidirectional communication because modern meters and distributed-energy resources both send measurements and receive commands.

3.3 Wide-area network (WAN)

A WAN connects substations, control centres, generation sites, regional operations centres, and external market or service platforms. WANs often use utility fibre, leased telecommunications circuits, microwave radio, cellular networks, or multiprotocol label switching.

A utility WAN commonly includes redundant paths, routing controls, firewalls, virtual private networks, precise time distribution, and network monitoring. Operational technology traffic should be segmented from corporate information technology traffic, while carefully controlled interfaces are provided where data must cross between them.

The three levels are logical rather than universal physical boundaries. A device may communicate directly through a cellular WAN without passing through a conventional NAN, and a building-scale microgrid may contain its own local-area network that performs functions similar to a HAN.

4. IoT Connectivity in Smart Grids

The Internet of Things describes networks of physical objects that contain sensors, processing capability, software, and communication functions. In smart grids, IoT devices include line sensors, transformer monitors, weather stations, inverter controllers, water-heater controls, and electric-vehicle charging stations.

An IoT communication chain normally contains:

1. Sensing or actuation: a physical variable is measured or a device is controlled.
2. Edge processing: data may be filtered, compressed, time-stamped, or analysed locally.
3. Connectivity: information is transmitted through a local, neighbourhood, cellular, or utility network.
4. Platform services: data are stored, correlated, visualized, or supplied to applications.
5. Control or decision: an operator, automation system, or algorithm issues an action.

Protocols such as MQTT and CoAP are common in general IoT systems. MQTT uses a publish–subscribe model in which clients exchange messages through a broker. CoAP is designed for constrained devices and resembles a lightweight web transfer protocol. Neither protocol by itself guarantees the deterministic performance required for every protection application. Application selection must therefore follow the operational requirement.

Edge computing is valuable when sending every raw measurement to a cloud platform would create excessive bandwidth use, delay, or privacy exposure. An edge device can detect abnormal vibration locally and transmit an alarm plus selected evidence rather than a continuous raw data stream. AI methods can assist with anomaly detection and predictive maintenance, but their outputs require validation, explainability, monitoring for data drift, and safe fallback behaviour. AI should not silently replace deterministic protection logic.

5. Latency, Reliability, and Quality of Service

Communication requirements should be written as measurable service-level requirements. A useful specification identifies the maximum allowable end-to-end delay, acceptable packet-loss ratio, availability target, jitter limit, security controls, time synchronization accuracy, and recovery time after failure.

The total end-to-end delay can be represented conceptually as:

The terms represent processing, queueing, transmission, propagation, and any recovery or retransmission delay. This is a decomposition rather than a universal fixed formula; some systems define recovery separately, and retransmissions may alter all terms. Transmission delay for a message of L bits on a link with rate C bit/s is:

This expression assumes that the entire message is sent over one link at rate C and excludes protocol overhead, queueing, propagation, and processing. For example, a 2,000-bit message sent at 10 Mbit/s has a transmission time of 0.2 ms, before other delays are included.

Quality of service (QoS) mechanisms include traffic prioritization, admission control, bandwidth reservation, time synchronization, path redundancy, packet marking, and network segmentation. Protection and control traffic may receive priority over routine meter reads. However, priority does not compensate for inadequate capacity, insecure devices, or a single point of failure.

Typical qualitative requirements are:

Exact limits depend on the protection scheme, utility standard, network design, and regulatory context. It is unsafe to assign one latency threshold to all smart-grid services.

6. Interoperability: Layers and Meaning

Interoperability is the ability of systems from different suppliers or organizational domains to exchange information and use it correctly. It has at least three dimensions:

• Technical interoperability: physical links, addressing, transport, and message exchange function correctly.
• Syntactic interoperability: systems agree on message structure, encoding, and data types.
• Semantic interoperability: systems assign the same meaning to data, including units, status, quality, time, and relationships.

A network can be technically connected but semantically incompatible. For example, two systems may exchange a value labelled “voltage” while disagreeing about phase, line-to-line versus line-to-neutral measurement, scaling, or units.

Interoperability also requires conformance testing, version management, device profiles, cybersecurity policies, and clearly defined ownership of data. Open standards reduce vendor dependence, but they do not eliminate engineering configuration.

7. Major Protocols and Standards

7.1 IEC 61850

IEC 61850 is a family of standards for communication networks and systems in substations and related power-utility automation applications. Its central contribution is not merely a packet format; it defines an information model, services, communication mappings, and engineering concepts.

Important concepts include:

• Logical nodes: standardized functional objects representing protection, measurement, control, and other functions.
• Common data classes: structured definitions for attributes such as status, measured values, commands, and quality.
• Abstract Communication Service Interface (ACSI): services presented independently of a particular network protocol.
• Manufacturing Message Specification (MMS): commonly used for client–server supervisory communication.
• Generic Object-Oriented Substation Event (GOOSE): fast event messaging, often used for peer-to-peer status and trip-related functions.
• Sampled Values (SV): time-critical transmission of digitized analogue measurements.
• Substation Configuration Language (SCL): machine-readable descriptions used to support engineering and configuration.

IEC 61850 can improve interoperability and reduce hardwired connections, but successful implementation requires disciplined system engineering. Vendor-specific extensions, inconsistent naming, configuration errors, and cybersecurity weaknesses can still undermine interoperability.

7.2 DNP3

The Distributed Network Protocol version 3 (DNP3) is widely used in supervisory control and data acquisition (SCADA), especially in electric utilities. It supports remote terminal units, intelligent electronic devices, event data, time-stamped information, unsolicited responses, and communication over serial or IP networks.

DNP3 was designed for utility environments with relatively constrained and sometimes unreliable links. Its object groups and event mechanisms help a control centre obtain changes rather than repeatedly transferring every value. Secure Authentication and related security practices can strengthen protection, but the base protocol should not be assumed secure merely because it is widely deployed.

7.3 Modbus

Modbus is a simple and widely implemented request–response protocol. Modbus RTU commonly operates over serial links, while Modbus TCP operates over TCP/IP. Data are organized using coils, discrete inputs, input registers, and holding registers.

Its advantages are simplicity, low cost, and broad device support. Its limitations include a relatively weak semantic model, limited native security, and dependence on external documentation to explain what register values mean. Modbus is therefore often suitable for local equipment integration but requires gateways, segmentation, authentication controls, or secure tunnels when used across untrusted networks.

7.4 IEEE 2030

IEEE 2030 is a framework and guide for smart-grid interoperability. It addresses the interaction of power, communications, and information technology systems. It is better understood as an interoperability framework and reference architecture than as a single field communication protocol.

The framework encourages analysis across domains, layers, interfaces, data flows, and actors. It helps engineers identify where standards must cooperate and where responsibilities lie. It does not replace application-specific standards such as IEC 61850 or DNP3.

7.5 Common Information Model (CIM)

The Common Information Model is an object-oriented information model standardized primarily through IEC 61970 and IEC 61968-related work. It represents power-system resources, connectivity, assets, measurements, work activities, and other utility information in a common form.

CIM is especially important for enterprise and control-centre integration, system planning, distribution management, outage management, and exchange between applications. It addresses the meaning and relationships of information rather than specifying every physical communication link. A CIM-based exchange may be serialized using technologies such as XML or RDF, but the model and the encoding are distinct concepts.

7.6 Comparison of standards

8. Security and Governance

Smart-grid communication networks are critical infrastructure. Threats include unauthorized control, false-data injection, denial of service, malware, credential theft, eavesdropping, and supply-chain compromise. Wireless interfaces and remote maintenance channels deserve particular attention, but internal networks must also be treated as potentially hostile.

Good practice includes defence in depth, asset inventories, network segmentation, least-privilege access, strong authentication, cryptographic protection, secure boot where appropriate, patch management, logging, intrusion detection, backup communication paths, and tested incident-response procedures. Safety and availability must be considered alongside confidentiality. A security control that causes an unsafe loss of control during an emergency is poorly designed.

Governance issues include data ownership, customer privacy, retention periods, third-party access, regulatory reporting, and responsibility for automated decisions. Smart meters can reveal occupancy patterns, while distributed-energy data may expose business operations. Data minimization and role-based access should therefore be incorporated into architecture rather than added after deployment.

9. Illustrative Case: Distribution Feeder with Distributed Energy Resources

Background

A utility operates a 13.8-kV distribution feeder with rooftop photovoltaic systems, battery storage, smart meters, voltage regulators, and an automated recloser. Rapid changes in solar output cause voltage excursions, while the control centre has limited visibility of low-voltage conditions.

Problem

The utility must obtain reliable measurements, coordinate voltage control, and avoid using a communication architecture that fails whenever a single access point or cellular connection is unavailable. Customer privacy and cybersecurity must also be protected.

Analysis

The application contains several traffic classes. Smart-meter energy data are delay tolerant, whereas recloser status and control are operationally important. Inverter measurements may require more frequent updates during abnormal conditions. A single protocol is unlikely to be optimal for every interface.

Approach

The utility adopts a hybrid architecture:

• Smart meters communicate through a NAN using a managed RF mesh, with cellular backup for selected critical sites.
• The recloser and voltage regulator connect to a substation gateway through fibre or hardened Ethernet.
• Local control uses an appropriate utility automation protocol, while the gateway maps field information into the control-centre environment.
• IEC 61850 is used within the substation for structured IED data and fast local events.
• DNP3 or another approved SCADA interface connects selected remote devices to the control centre where required by the existing operational environment.
• CIM-based information exchange supports distribution-management and enterprise applications.
• Network segmentation separates protection, operational control, customer metering, and corporate services.
• Edge logic maintains safe local voltage control if the WAN is temporarily unavailable.

Outcome

The hybrid design avoids forcing a low-bandwidth meter network to carry protection traffic. Local automation reduces dependence on WAN latency, while redundant paths improve service continuity. The principal remaining risks are configuration errors, inconsistent device models, wireless interference, and insecure legacy equipment.

Learning Points

1. Communication requirements must be derived from the application.
2. Interoperability involves models and meanings, not only physical connectivity.
3. Local fallback control is important for resilience.
4. Gateways are useful but can become bottlenecks or single points of failure.
5. Security, privacy, and maintainability should be designed from the beginning.

10. Best-Practice Design Method

A practical design process is:

1. Identify devices, users, control functions, and data flows.
2. Classify each flow by latency, reliability, bandwidth, timing, and security requirements.
3. Select physical media and network topologies, including redundancy where justified.
4. Select protocols and information models appropriate to each interface.
5. Define addressing, naming, time synchronization, QoS, and gateway behaviour.
6. Separate safety-critical and business traffic through segmentation and access control.
7. Test interoperability using realistic devices, failure conditions, and configuration versions.
8. Monitor performance and security after commissioning.
9. Plan lifecycle maintenance, certificate renewal, patching, spare equipment, and technology migration.

Challenges and Contemporary Developments

Legacy equipment remains in service for decades, making gradual migration more realistic than complete replacement. Protocol gateways can connect old and new systems, but every translation may lose semantics, timing information, quality flags, or security context.

Private 5G networks, time-sensitive networking, software-defined networking, and satellite connectivity are developing options for selected utility applications. Their suitability depends on coverage, deterministic performance, lifecycle support, and regulatory conditions. Digital twins and AI-assisted operations can improve planning and maintenance, but they increase dependence on accurate models and trustworthy data.

Future systems will increasingly coordinate millions of small devices. Interoperability testing, secure device identity, automated certificate management, and machine-readable information models will become more important. Engineers must evaluate not only nominal performance but also behaviour during congestion, cyberattack, equipment failure, and loss of external connectivity.

Chapter Summary

Smart-grid communication networks connect devices ranging from household appliances to transmission control centres. Wired networks generally provide predictable performance and strong immunity when fibre is used, while wireless networks offer flexibility, mobility, and economical connectivity over dispersed areas. HAN, NAN, and WAN structures organize communication at customer, neighbourhood, and utility-wide scales, although practical systems may combine or bypass these levels.

Communication design must match application requirements for latency, jitter, throughput, packet loss, reliability, availability, security, and recovery. IoT technologies and edge computing extend observability and control, but they do not automatically satisfy critical-grid requirements. IEC 61850 provides rich models and services for substation automation; DNP3 supports utility SCADA and event communication; Modbus offers simple device integration; IEEE 2030 provides an interoperability framework; and CIM supplies common semantic models for utility information exchange.

Interoperability is technical, syntactic, and semantic. Gateways, common models, conformance testing, security controls, redundancy, and lifecycle governance are all necessary. A resilient smart grid uses hybrid communication architectures, local fallback functions, carefully controlled interfaces, and continuous monitoring.

Key Terms

Availability: Proportion of time that a service performs its intended function.

CIM: Common Information Model used to represent utility information and relationships.

DNP3: Utility-oriented protocol used widely for SCADA and remote-site communication.

HAN: Home-area network connecting devices within a customer premises.

IEC 61850: Standard family for power-utility automation communication and information models.

Interoperability: Ability of different systems to exchange and correctly use information.

IoT: Network of physical objects containing sensing, processing, and communication functions.

Jitter: Variation in packet or message delay.

Latency: Time taken for information to travel from source to destination.

Modbus: Simple register-based industrial communication protocol.

NAN: Neighbourhood-area network aggregating traffic from multiple premises or field devices.

Packet-loss ratio: Fraction of transmitted packets that fail to arrive successfully.

QoS: Quality-of-service mechanisms and requirements governing communication performance.

Reliability: Probability that a specified communication function succeeds under stated conditions.

SCADA: Supervisory control and data acquisition system.

Semantic interoperability: Agreement on the meaning, units, relationships, and context of exchanged data.

WAN: Wide-area network connecting geographically distributed utility sites.

GOOSE: IEC 61850 event messaging service commonly used for fast peer-to-peer communication.

Review Questions

Short Answer Questions

1. What are the principal differences between wired and wireless smart-grid networks?
2. Define HAN, NAN, and WAN and give one application of each.
3. Why is semantic interoperability important in a smart grid?
4. What is the main purpose of CIM?
5. Why should protection traffic not automatically share the same QoS treatment as routine meter reading?

Long Answer Questions

1. Compare fibre, copper Ethernet, cellular, and wireless mesh communication for distribution-grid applications.
2. Explain the IEC 61850 communication model, including logical nodes, GOOSE, Sampled Values, and SCL.
3. Compare IEC 61850, DNP3, Modbus, IEEE 2030, and CIM in terms of purpose and application.
4. Discuss cybersecurity, privacy, and governance requirements for smart-grid communication networks.
5. Design a conceptual HAN–NAN–WAN architecture for a community containing photovoltaic systems and electric vehicles.

Analytical / Application Questions

1. A 2,000-bit message is sent over a 10 Mbit/s link. Calculate its ideal transmission time, excluding propagation, queueing, processing, and protocol overhead. Explain why this is not the same as end-to-end latency.
2. Two independent communication paths each have reliability 0.98 for a specified operating interval. Calculate the reliability of the parallel arrangement and state the independence assumption.
3. A utility wants to use Modbus devices across a public cellular network. Evaluate the technical and cybersecurity risks and recommend an appropriate gateway and segmentation strategy.

Multiple Choice Questions with Answers

Question 1

Which characteristic most strongly distinguishes a NAN from a HAN?

A. A NAN connects devices within one appliance

B. A NAN aggregates communication from multiple premises or field devices

C. A NAN is used only for protection relays

D. A NAN cannot use wireless communication

Correct Answer: B. A NAN aggregates communication from multiple premises or field devices

Question 2

Which medium provides electrical isolation and strong immunity to electromagnetic interference?

A. Unshielded twisted pair

B. Coaxial cable

C. Optical fibre

D. Single-band Wi-Fi

Correct Answer: C. Optical fibre

Question 3

What is the principal purpose of CIM?

A. To define radio modulation

B. To provide a common semantic information model for utility systems

C. To replace all protection relays

D. To provide household wireless encryption only

Correct Answer: B. To provide a common semantic information model for utility systems

Question 4

Which statement about IEEE 2030 is correct?

A. It is solely a serial register protocol

B. It is a framework addressing smart-grid interoperability across power, communications, and IT

C. It is a fibre-optic connector standard

D. It is a replacement for all versions of Modbus

Correct Answer: B. It is a framework addressing smart-grid interoperability across power, communications, and IT

Question 5

A 2,000-bit message is transmitted at 10 Mbit/s. Ignoring overhead and other delays, what is the transmission time?

A. 0.02 ms

B. 0.2 ms

C. 2 ms

D. 20 ms

Correct Answer: B. 0.2 ms

Question 6

Which IEC 61850 service is associated with fast event-oriented peer-to-peer messaging?

A. GOOSE

B. CIM

C. Modbus RTU

D. MQTT broker discovery

Correct Answer: A. GOOSE

Question 7

Which protocol is especially known for simple coils and register-based data exchange?

A. Modbus

B. CIM

C. IEC 61850 SCL

D. IEEE 2030

Correct Answer: A. Modbus

Question 8

If a communication service is unavailable for 2 hours during a 1,000-hour observation period, what is its availability?

A. 98.0%

B. 99.0%

C. 99.8%

D. 100.2%

Correct Answer: C. 99.8%

Question 9

Why can edge computing be useful in smart-grid IoT systems?

A. It eliminates the need for sensors

B. It can reduce latency and bandwidth by processing data near the source

C. It guarantees that wireless links never fail

D. It removes the need for cybersecurity controls

Correct Answer: B. It can reduce latency and bandwidth by processing data near the source

Question 10

Which requirement is most associated with protection and interlocking traffic?

A. Very high delay and low dependability

B. Very low data volume but highly dependable and time-sensitive delivery

C. High entertainment-video bandwidth only

D. No need for time synchronization

Correct Answer: B. Very low data volume but highly dependable and time-sensitive delivery