The smart grid requires the integration of complex electrical, communications and information technology systems. High levels of reliability, cost efficiency and customer satisfaction are required from each layer of the smart grid, yet the companies involved — telecoms and electric utilities — historically have not collaborated effectively. As the word “networking” implies, these two industries will have to cooperate to improve future communications capabilities of the smart grid.
Difference in Approach
Electric utilities and telecom operators approach smart grid networking issues differently. As a former telecom chief strategy officer, I take it for granted all network engineering decisions are made from an integrated perspective, including consideration of any inter-network interface, transition or standards issues. With few exceptions, telecom operators have a common understanding that achieving competitive-scale economics requires sending as many digital bits over shared physical communications networks as possible.
In contrast, it appears that while utility engineers apply similar principles to the electric grid itself, they often append communications network decisions to each specific construction project, which can lead, over time, to a fragmented collection of differing technologies. Even if each decision is optimal for a particular project, the result can be a collection of telecom networks that are suboptimal as a whole. Using the National Institute of Standards and Technology's smart grid framework, the secure communication flows should be as carefully engineered and integrated as the electrical flows.
Applying telecom strategic perspectives does lead to a few — inevitably oversimplified — observations about smart grid network variations.
Public and Private Networks
Smart grid publicity most often focuses on smart meters and the advanced metering infrastructure (AMI) networks enabling communication between the customer and the utility. AMI traffic volumes will be thousands of times larger than they have been in the past, especially as demand management is extended, through home or building area networks, to smart appliances and thermostats, networked industrial engines, plug-in electric vehicles and distributed generation facilities. That volume is still trivial compared to the capacity of public wireline and wireless broadband networks.
This high-volume shared capacity causes public telecom networks to have lower costs than dedicated AMI solutions. It seems inevitable that, as costs are publicly scrutinized, more utilities will adopt public communications network solutions for smart meter communications.
Many utility engineers object to dependence on public networks. Their points are valid when discussing the supervisory control and data acquisition (SCADA) portion of smart grid communications networks. For mission-critical elements of the electric grid linking bulk generation, transmission (substations), distribution (transformers or reclosers), operator and other domains, the accountability for reliability, safety, security and control implies that most utilities will continue to prefer private network solutions. This is especially true for high-bandwidth, low-latency applications that utilities can afford to serve with fiber technologies.
My expectation of increased public network adoption of AMI communications is based on the perception that no single smart meter is material to the management of the electric grid and that safety is ensured by a “dumb grid” fallback crisis mode. If some industrial demand management or distributed generation site becomes critical to grid operations, then the utility should consider extending its private network to this site.
Another less developed meaning of smart grid involves distribution automation, especially the instrumentation of the electric grid through the insertion of many monitoring devices. This communications need involves many new links similar to AMI networks, which tend toward low-cost public solutions, but also may be important to electric grid functionality similar to SCADA networks, which tend toward controlled private solutions. An integrated approach to smart grid network design suggests that evaluation of the optimal solution by each utility should be holistic, considering the incremental cost of adding the communications of new instruments to preexisting public AMI or private SCADA networks.
Bill Blessing (firstname.lastname@example.org) is a former senior vice president of corporate strategy and development for Sprint and Embarq telecommunication service providers. He serves on the board of Clearwire and other organizations, and has consulted on smart grid strategy.
The National Institute of Standards and Technology's smart grid conceptual framework defines seven important domains.
The Issues of Public or Private AMI Communications Networks
Both electrical utilities and telecom operators claim impressive uptimes exceeding 99.9%. Ironically, the biggest cause of extended wireless network outages is loss of electrical power, suggesting the benefits of coordinated engineering.
Utility communications can be segregated onto secure virtual private networks, since billions have been spent by telecom networks to ensure the security of all their Internet traffic. Because use of the public Internet inherently increases risks, especially when customers elect to share data with third-party service providers, many utilities may choose to use it only for non-mission-critical communications.
The Federal Communications Commission (FCC) says public broadband services are available to more than 96% of U.S. households, with more than 90% having multiple network options. When utilities serve areas so rural that competitive telecom operators do not find it economical to serve them, the high cost of providing AMI communications should be considered as part of any smart meter business case.
Comparing public network pricing to internal costs is further complicated because many private utility wireless networks use unlicensed spectrum, while public operators consistently prefer licensed spectrum that adds to their costs. This means that interference and other quality differences need to be considered, and it illustrates why the FCC is averse to dedicating valuable spectrum that could be shared by multiple users to any single-purpose user, even one as valuable as utility networks.
If utility regulations provide for profits based on private network investments, but only allow recovery of public network costs, utilities might logically prefer the former even when less economical. The need to correct such incentives, and to provide support for energy-efficiency initiatives, is why many advocate for decoupling and other regulatory reforms.