by Kevin Cornish, Enspiria Solutions Inc.
Supporting smart grid solutions deployed or envisioned by North American utilities requires a new approach to utility communication network design.
Utilities have been among the largest creators of private communication networks supporting varied corporate and operational information technology requirements.
The result has been vast networks of copper pairs, leased analog and frame relay circuits, fiber, microwave and radio networks.
Many of these solutions are not well-equipped to support the requirements of the evolving smart grid landscape, and those that do rarely provide complete coverage across a utility’s service territory.
New approaches are required for the communication networks that will support smart grid field devices. No single network can meet all the smart grid requirements.
Instead, a network of networks can be combined to deliver necessary bandwidth and latency requirements for the wide variation of applications envisioned.
Remote system monitoring, new distribution automation (DA) devices, expanded substation automation coverage, local distribution system grid control and smart meter solutions all have different quality of service (QoS) and performance needs.
The Right Networks
It is possible to use fewer networks to meet applications’ different service requirements, but doing so will entail either the use of networks with higher capability than required–resulting in unnecessary costs–or networks that are ill-suited or provide insufficient capabilities that impact performance and business value.
In both cases, a utility will not have the optimum match of application to network and will not receive the maximum benefits of both.
For example, there has been considerable discussion of using smart meter networks to support additional applications such as distribution system monitoring, automation and local generation control.
Smart meter networks have different bandwidth, deterministic latency and network management capabilities. It makes any generalized assessment difficult.
The Tier 3 smart meter networks might not be ideal candidates for a broad array of automation solutions. In some cases a smart meter network might suffice, but it might not be the ideal approach.
Smart grid applications have different QoS requirements. Figure 1 illustrates the varied bandwidth and latency requirements of different applications.
While not meant to be exact, the graph illustrates that advanced metering infrastructure (AMI), smart metering and demand response (DR) requirements are generally the lowest in bandwidth and latency.
Highest is substation video security and transmission supervisory control and data acquisition (SCADA). In between are the distributed generation (DG), distribution automation (DA), real-time switching and volt/VAR optimization (VVO) solutions.
Using a solution designed to meet the QoS requirements of AMI and DR to fulfill the communications needs of higher requirement applications might not be a good fit. It would be logical to take different communications networking approaches for this disparate solution requirement set.
The Important Criteria
The goal of an intelligent grid is one that monitors, senses and takes action without human direction or intervention.
By design it must collect, manage, analyze and store vast quantities of data faster and more reliably than previous solutions. A view of the primary performance measures for smart grid communications solutions includes the following five categories: bandwidth, latency, reliability, security and standards.
For each category, the smart grid applications have different requirements. Table 1 examines these categories and suggests performance ranges for each criterion.
In meeting the diverse needs of the five performance criteria, multiple and varied communications solutions are required. The network of network approach can be viewed as a tiered network design.
This is simplistic and, in actual practice, there are hybrid network designs and multiuse networks that blur the distinctions between the layers.
This approach allows a utility to identify available network options and capabilities, classify application requirements and pair the smart grid applications with the most appropriate supporting communications infrastructure.
In a four-tier network design, focusing on the first two criteria, the tiers are defined as follows:
- Tier 1: The high-bandwidth, low-latency backbone of the utility communications infrastructure; the information superhighway for the corporate smart grid initiatives.
- Tier 2: Still providing significant bandwidth and low latency but more focused in delivering backhaul solutions to connect specific applications to the Tier 1 network. These network-level solutions also provide for direct connection for high-value devices or those with more stringent requirements than can be served by Tier 3 and 4 solutions.
- Tier 3: Include local area networks (LANs) or neighborhood area networks (NANs) that support applications such as smart metering networks, localized distribution automation or internal substation communications networks.
- Tier 4: The least well-defined level of the network but envisioned to support home area network (HAN) solutions, providing connectivity between a utility and end-use customers.
Although not exhaustive, Table 2 provides examples of the physical approaches and applications for the suggested four network levels.
The Optimum Approach
Several factors influence a utility’s approach to developing its telecommunications network and its plan to support smart grid solutions.
- Unique Starting Point. Every utility has a different starting point. As with some developing countries that have bypassed extensive copper phone networks to go directly to cellular, utilities without extensive existing systems might be able to jump more easily to the desired end-state solutions.
- Evolving Requirements. How well-defined are the utility’s requirements? Because many utilities are still developing their smart grid strategies and business initiatives, they do not know the extent of their communications requirements. These utilities might need to develop interim steps rather than to move directly to their optimum end-state solutions. These strategies ideally would allow for focused and incremental system upgrades without significant stranded-asset concerns.
- Appropriate Solutions. Is it necessary to have a ubiquitous solution that provides the same level of service and complete system coverage? Attaining this is often an expensive proposition. Geography, availability of public networks where a utility’s private network is not cost-effective and making do with lower-level but economically acceptable solutions instead of ideal or spot solutions might be appropriate.
While much of the smart grid focus has been on the applications and business value derived, it is critical to understand the communications networks issues and options, as well.
A mismatch of application requirements and network capabilities can render an otherwise robust solution design unworkable.
Kevin Cornish is an executive consultant and the smart grid consulting practice lead at Enspiria Solutions, a Black & Veatch company. He has an MBA in marketing and telecommunications management, a master’s degree in electrical engineering and power systems and a bachelor’s degree in electrical engineering and computer science. Reach him at firstname.lastname@example.org.
Past EL&P Issues