Communications System Design

Electric power delivery is among the most communications-centric industries in existence. But even for the largest electric utilities, the telecom planning and operations effort is intense, complex and multifaceted. The Utilities Telecom Council pointed out the need for utility planners to first develop utility-wide strategic communications architectures (UTC Journal, 2010 Special Edition), and this makes perfect sense.

Today, much of the time utilities spend on communications operations is focused on the coordination and maintenance of multiple applications-specific subnetworks. More emphasis needs to be placed on developing a communications strategy. For most large utilities in North America, the initial development of smart grid plans must be based on a forward-looking multipart, multitiered communications strategy focused primarily on private utility ownership of telecommunications assets, which will ensure a highly reliable, secure and robust communications system design. This design will be the foundation on which smart grid applications can be built.

A strategic overhaul of the telecommunications infrastructure for utilities often involves updating an existing multitiered approach to such a “comms” — communications or telecom — design, at least for the largest electric utilities. For midsize and smaller utilities, the architecture design is likely to be less complex, less sophisticated and less costly with fewer tiers. There is an urgent need for a comprehensive look into existing communications capabilities at major and midsize utilities. Such an evaluation will likely yield opportunities for gains in efficiencies, opportunities for consolidation and a chance for a systematic overhaul and redesign.

Further complicating the communications strategy considerations are the emerging requirements among transmission utilities, regional transmission organizations and transmission system operators evaluating national, regional or multi-utility approaches to coordinating phasor measurement data acquisition and the developing role of teleprotection for ultra-high-speed relay coordination.

Role of Private Telecom Networks

The debate on private versus public ownership of utility telecommunications networks is not going to end any time soon. However, it is reasonable for a large proportion of utility networks to remain essentially private, supplemented by the judicious application of commercial telecommunications services for less critical activities. A spate of comments has been made by several telecommunications industry executives and senior officials from the metering industry regarding inefficiencies within utility-operated private networks. While there are likely to be some cost advantages to using commercial services, there is not the same level of availability, reliability and security inherent in commercial-grade public networks as there is in today's private utility telecommunications networks.

A Brilliant Grid

U.S. Army Lt. Gen. Russel L. Honoré (Ret.) recently questioned why the electric power industry would only attempt to construct a smart grid, when what is really needed is a brilliant grid to be prepared for any emergency. He further suggested that cell towers and certain other public telecommunications infrastructure components were among the first to fail during Hurricane Katrina. Whether for a brilliant or smart grid, it is important to consider the key aspects of communications system design as a starting point for a realistic approach to grid development.

Wireline or Wireless?

Clearly, utilities must choose the best of both wireline and wireless technologies. A combination of these approaches will provide what is needed for a multitiered communications platform that meets requirements for operational data communications. Fiber wireline applications are currently centered on synchronous optical networking (SONET) in North America and synchronous digital hierarchy (SDH) internationally. Leased digital service lines of T1 or higher bandwidth are also popular.

Other wireline approaches in use throughout North America include more digital subscriber line (DSL) options, broadband over power line (BPL), power line carrier/distribution line carrier (PLC/DLC) and dial-up telephone services.

Some wireless approaches to utility communications have been around for 60 years or more, beginning with radio telemetry applications in the 1940s. Today's wireless options include microwave, licensed radio, satellite and, more recently, cellular and mesh networks based on unlicensed spectrum.

A brief overview of the North American communications segmentation for utility planners might include:

• Tier 1: Backbone network infrastructure
• Tier 2: Backhaul networks
• Tier 3: Field area networks
• Tier 4: Neighborhood area networks/home area networks.

If premises-based communications systems are to be owned by the utility (which is unlikely in North America), they could be viewed as a fifth tier.

Operational smart grid applications are primarily wireline based for a large control center to substation, protection systems and substation automation systems. These are core system-protection applications, and almost all large urban utilities rely on fiber and leased lines for the backbone design and backhaul transmission of key operational data.

For distribution automation, voltage-stability applications centered on the monitoring and control of devices on poles and lines represents a major shift to wireless communications. Field workforce communications is also wireless and is likely to remain as licensed radio because of the need for reliable and secure communications under the most adverse conditions.

Consumer-side smart grid applications are primarily wireless for metering data acquisition, with some international use of wireline methods employing PLC and BPL technologies. Wireless is also used for in-premises communications for a host of consumer applications, including energy management and demand response.

Following is a summary of what has been observed in studies on substation communications in the North American marketplace. This list is just a sampling of the firms active in specific segments:

• PLC /DLC

  Who's Who in Communications

  This is a tried-and-true low baud rate (narrow-bandwidth) method for communicating to and from remote devices, primarily meters, and is sometimes used in load control applications. These firms are active in this segment: Amperion, Current Technologies, Grid Stream, Echelon, Landis & Gyr and Aclara.

• Cellular

  Cellular-based technology — whether mesh-based or point-to-multipoint — is popular with several advanced metering infrastructure (AMI) firms and their communications partners. Sensus/Telemetric and SmartSynch are key players here.

• Satellite

  Convergence of Communications

  Satellite data acquisition plays a huge role in the oil and gas industry, maritime ship tracking, GPS and other segments. Globally, utilities spend more than US$175 million for remote site data acquisition. Leaders in satellite services for electric utilities include Hughes, Spacenet, LBiSat and iDirect.

• Worldwide Interoperability for Microwave Access (WiMAX)

  This newer communications protocol provides fixed and mobile Internet access. The current WiMAX working group developments have enabled speeds approaching 40 MB/sec. The new IEEE 802.16 (version m) release provides 1 Gb/sec rated throughput. Trilliant, Clearwire and Grid Net are three key participants offering WiMAX services to electric utilities.

• 900-MHz Multiple Address System (MAS), licensed radio

  Silver Spring Networks leads the way for applying 900-MHz MAS to the AMI marketplace. However, most supervisory control and data acquisition (SCADA) firms manufacturing remote terminal units and other intelligent electronic devices for utilities provide 900-MHz MAS compatibility.

• 900-MHz MAS , unlicensed spectrum

  This is less costly than licensed radio, but the big problem is noise. Co-leaders in this utility communications market segment include GE MDS, FreeWave and Motorola.

• Public network operators and their partners

  Communications Protocols

  These technology giants — AT&T, Verizon, T-Mobile and Sprint — are the commercial service providers for most landline-based and cellular telephones in use today. These and other firms partner with utilities to provide coverage in portions of service areas not covered by private utility networks. More importantly, these companies are fighting for a larger role in the new era of advanced distribution automation (ADA) and automatic metering applications. These firms are working to build out fourth-generation (4G) capabilities to their networks. The firms are now deciding whether to go ahead with either long-term evolution (LTE) or WiMAX as their foundation for 4G offerings. While North America's larger utilities will likely make selective and limited use of these new commercial service offerings, they will become quite popular in the developing nations of the world, where utility capital is not available to build out and maintain private utility telecom networks.

As the AMI market develops and matures, the related communications network infrastructure is a lot like “dark fiber” in that there is — or can be — significant unused excess capacity that can be put to work for some, or all, of an ADA network. Of course, the choice will be determined by the perceived adequacy of security, reliability, resiliency and the appropriate level of bandwidth and latency as required for various distribution automation (DA) data operations.

In some cases, smart meter manufacturers have partnered with one or more of the aforementioned communications companies, the reason being that this is the part of the utility service area where private network ownership does not exist. Total solutions providers gain a marketing advantage by offering a solution that includes metering hardware and software as well as communications services.

Newton-Evans Research expects that each of the four key segments of the DA market — broadband communications, smart field devices and equipment, device controllers and applications software — will grow over the 2010-2018 period. Dedicated DA-centric telecommunications spending will likely increase, but — because of the shared nature of utility infrastructure communications (Table 1) — only dedicated expenditures, amounting to about $65 million in 2010, will be separated out. In fact, additional tens of millions of dollars will be allocated for the development of a 21st century distribution network-wide communications infrastructure to serve DA, metering, related customer premises data acquisition and reporting, and equipment/device diagnostic information retrieval.

Communications for ADA

Newton-Evans also studied substation automation, providing information relative to the mix of communications technologies used to link substations to control centers (Fig. 1).

Protocol usage within North American electric power substations remains predominantly DNP3, with Modbus next in importance, according to a study by Newton-Evans Research. There is little likelihood of significant change in protocol use over the next two to three years, based on responses received from very key utilities of each type and in each size range (Fig. 2). More than one-half of the Top 30 North American utilities (by numbers of customers, by line miles, by number of transmission substations and by number of distribution substations) are represented in this study.

The highest percentage of subgroups indicating current use of DNP 3 (serial and local area network) was investor-owned utilities followed by cooperatives. Modbus serial was especially popular among Canadian respondents. If anything, the utilities using legacy or other protocols may make some moves to DNP 3 over the next three years, but there is very little evidence at this time of any groundswell of support for migrating to IEC 61850 among this representative group of North American utilities through 2012.

Slightly more than one-half of the international respondents reported some use of IEC 61850 for intra-substation communications. More than one-third was using either/both DNP3 and/or Modbus. One-third reported ongoing use of legacy protocols.

In the United States, it is likely that DNP3 will continue to be widely deployed for utility-installed sensors, switches, pole-top devices and some types of line status devices. However, within customer premises, it is likely that smaller, lower-cost, limited-capability/functionality devices will communicate via ZigBee protocols in so-called home area networks and up to the next level of neighborhood area networks.

The DA communications market segment will move from a serial to broadband approach during the 2010-2015 era, and the new communications infrastructure will have strong cyber security defenses as mandated by various federal entities.

The market for broadband communications in a distribution automation environment will be mostly wireless, with a need for approaches that may involve multiple communications technologies and methodologies. The approaches to be taken will consist of utility-owned and -operated wireless and wireline infrastructure, likely to be supplemented with commercial carrier services.

The Big Picture

Device density, topography, spectrum licensing issues, security and communications technologies becoming available will have an impact on the telecommunications decisions for particular DA applications. Routing of DA information itself may become a basis for partial communications network redesign. Alternatives include three key options in DA design: control center based, substation based and field based.

In a January 2011 client study of more than 30 of the nation's major utilities serving close to 20% of all customers, the majority of respondents (87%) reported then-current use of DNP3 serial protocol for DA and smart grid applications. Fifty-eight percent had already migrated to a local area network version of DNP3 for these applications.

Regarding communications networks for feeder automation, two approaches garnered more responses than the other options. These were licensed private radio (34%) and fiber (31%). Twenty-two percent reported a preference for unlicensed private radio, and 19% cited public cellular. Nine percent indicated mesh technology with user datagram protocol/Internet protocol (UDP/IP) as a preference. Importantly, 25% of the respondents wrote in “other” preferences.

Transmission control protocol/Internet protocol was the prevalent method being used at the time of the survey to transport DA protocols/data streams, reported by nearly two-thirds of the respondents. Thirteen percent cited some use of UDP/IP. One-third of the survey base reported no use of IP.

Respondents indicated that some installed reclosers (84% citing at least some) and sectionalizing switches (68%) had been configured with communications capabilities for DA applications. More than one-half of the respondents also cited capacitor bank controllers, distribution switchgear, load tap changers and voltage regulators as also having communications capabilities for DA applications.

Table 2 provides a view of communications being used or planned for the most important T&D operational smart grid applications and consumer-level smart grid applications. Fig. 3 shows the large role of utility telecommunications and the expenditures involved with combined voice and data communications, accounting for more than $13 billion worldwide, with North America representing about 26% of the total.

With such large investments being made in the smart grid, it is crucial that utilities take the next right step and develop a forward-looking multipart, multitiered communications strategy that will ensure a highly reliable, secure and robust communications system design, one on which smart grid applications can be built.

Chuck Newton, president of Newton-Evans Research Co., has been a 35-year career-long researcher of information technology products, markets and trends.

Table 1. The Shared Nature of Distribution Automation Communications Networks

		Distribution automation
	Substation to control center		Substation to endpoint device		Substation to substation		Remote engineer maintenance access		Remote vendor access
	4Q 2008 (%)	YE 2010 (%)	4Q 2008	YE 2010	4Q 2008	YE 2010	4Q 2008	YE 2010	4Q 2008	YE 2010
Leased lines	41	1	9	2	20	2	12	0	7	20
Dial-up telephony	9	3	4	5	2	5	40	2	40	13
Frame relay	22	1	5	9	2	7	10	5	7	20
PLC/BPL	3	2	7	4	28	2	2	5	0	33
Fiber SONET	47	5	25	2	68	3	36	3	7	20
T-1 or other multiplexer	24	1	7	2	17	0	17	3	0	33
Internet (TCP/IP)	11	8	20	5	7	5	19	12	27	13
Microwave	32	3	11	0	25	0	10	7	0	33
Spread-spectrum multiple address radio	25	1	27	2	17	2	9	3	0	33
Licensed-spectrum radio	39	2	18	2	15	3	9	9	0	40
Satellite	8	0	2	5	2	5	2	5	0	33
Celllular (CDMA)	4	2	4	9	2	7	5	9	13	20
Cellular (GSM)	2	2	0	9	0	8	0	9	7	20
Cellular (UMTS)	0	6	0	9	0	8	0	9	0	33
802.11n WLAN	0	6	2	9	0	8	2	7	0	33
Total responses	93		56		60		58		15
Source: “Distribution Automation: Trends, Developments and Retrospectives: 2007-2018,” Newton-Evans Research Co.

Table 2. Application of Communications Methods for Key Smart Grid Applications

Smart grid application	Wireline fiber/DSL/dial-up	Wireless-1 satellite/microwave licensed-spectrum radio	Wireless-2 cellular/mesh unlicensed-spectrum radio	Communications tier
T&D operations
System protection	Primary	Secondary	No	Core backbone-1
EMS/SCADA/DMS	Primary (urban)	Secondary (primary rural)	Minor use	Core backbone-1 Backhaul-2
Substation automation	Primary (urban)	Secondary (primary rural)	No	Core backbone-1 Backhaul-2
Advanced distribution automation	Some (urban)	Primary	Some	Backhaul-2 Field area net-3
Consumer applications
Advanced metering infrastructure	No	Some	Primary	Neighborhood area network/home area network (NAN/HAN) or personal area network (PAN) Tier-4 Field area net-3
Premises energy management services	Low-voltage wiring (some)	No	Primary	NAN/HAN or PAN Tier-4
Demand response	No	Some	Primary	NAN/HAN or PAN Tier-4 Field area net-3
Active load control	No	Primary	Some	Tier-4 inward
Source: Newton-Evans Research Co.