Cooperative Wireless Networking for Smart Grid
Written by Hao Liang, Weihua Zhuang and Xuemin (Sherman) Shen
With cooperative wireless networking—a model allowing for coordinated operation among two or more wireless networks outfitted with dual or multi-mode communication devices—utilities are able to improve power system operation efficiency and reliability by acquiring more accurate and timely information. At the same time, customers can benefit from reduced energy bills and higher power quality.
A smart grid’s communication network can be conceptualized as generally having three levels, each with distinct technologies. The base level is a home area network (HAN), interconnecting electric devices such as household appliances within the customer’s premises. HAN communication requirements are relatively low-cost and short-range, and so technologies like ZigBee, WiFi and powerline communications are suitable.
Information acquired by the HAN is aggregated at the smart meter for the second-level communications via a neighborhood area network (NAN). The intermediate-level neighborhood network is also used to interconnect the field components of the power distribution system such as the power quality monitoring devices and the control devices for distributed generation units. Low-cost communication technologies like WiFi and ZigBee tend to dominate here as well, as power distribution systems are cost-sensitive by nature. Powerline communications is a relatively more attractive candidate only if electromagnetic interference from coexisting networks can be well addressed at the neighbourhood level. Cellular technologies like GPRS and 3G, offering ubiquitous coverage and low latency, are alluring candidates but are substantially more costly, mainly because they rely on licensed frequency bands.
Information acquired in electrical distribution systems via NANs is aggregated at the distribution substation, a typical utility asset. Communications with other utility assets such as power plants, transmission lines and distribution substations are handled in a wide area network (WAN). Wireline communication technologies such as fiber optics are widely used in WANs to ensure high efficiency and reliability for utility-level monitoring and control.
Given the diversity of communication technologies within the smart grid, there is no certainty when a communication standard will be developed. The smart grid of the future is expected to be a heterogeneous network environment in which a rich variety of communication technologies coexist. In comparison with the designs of the home area network and the wide area network, which are cost- and performance-driven, respectively, the neighborhood area network’s design is challenging because both cost and performance issues must be addressed. In the future smart grid, the NAN will be deployed in a power distribution system characterized by a large-scale integration of distributed generation (DG) units and electric vehicles (EVs) with highly random output and mobility, respectively. Given all the design challenges, relying on a single network infrastructure is neither efficient nor economic.
An emerging technology known in the communication research community as “cooperative wireless networking” offers an unprecedented opportunity to address these challenges. Cooperative wireless networking enables coordinated operation among two or more wireless networks based on the use of dual-mode or multi-mode communication devices so the advantages of each network can be fully exploited to maximize the network utilization. However, when diverse communications technologies are harnessed in the context of the smart gird, information received about data delivery from existing networks may not be valid if the characteristics of power system applications were not taken into account (see IEEE Journal on Selected Areas in Communications special issue on Cooperative Networking - Challenges and Applications Part I and Part II). Specifically, when and how to cooperate are still open questions that need to be answered.
Cooperation among wireless networks is called for when quality-of-service requirements of smart grid applications cannot be satisfied by a single network. The most important requirements concern network coverage, capacity and delay.
The coverage of a smart grid communication network defines a set of communication network nodes that can be reached by certain monitors or controllers. The coverage provided by certain short-range communication devices (such as ZigBee and WiFi devises) may not be ubiquitous. Although deploying more relay nodes can improve network coverage, the coverage improvement cannot be ensured if some of the critical relay nodes on the multi-hop link fail. Based on cooperative wireless networking, the cellular network can be used as an alternative to bypass the outage link. Activating cellular communication devices would need to be mandatory for link reestablishment. Otherwise, full monitoring and control of all nodes in the smart grid cannot be guaranteed.
Given a specified channel bandwidth, a smart grid communication network’s capacity is defined as the maximum transmission rate that can be achieved over the channel. Since decentralized medium access control is typically used by ZigBee and WiFi networks, the NAN’s capacity may be limited due to high channel contention, especially when the NAN coexists with public or private wireless local area networks that are operating on the same frequency channel. If the smart grid controller collects insufficient information because of the capacity limitation, its control actions may not be accurate. In this case, some NAN nodes can activate the cellular communication devices so the controller can collect more information. But the activation is not mandatory if the smart grid can still maintain stable operation with the information acquired via ZigBee and WiFi networks, given the high cost of using a cellular network.
Delay is defined as the time between when a message is generated by an information source and when the message is received by the intended destination in smart grid. The delay consists of several components: channel access time, data transmission time, data buffering time, data processing time and so on. For a multi-hop network, the end-to-end delay is a summation of the delays over all hops. Many protection functionalities in the smart grid have stringent delay requirements to support the operation of circuit breakers and/or protective relays. Although the delay is somehow related to the channel capacity that determines data transmission time, the channel access time can be the dominating factor for safety-oriented services with small size data packets. If the data packet is not delivered in time, the protection functionality cannot be performed, resulting in malfunction or direct damage to electric devices. In this case, a communication link via the cellular network can be activated to create a “shortcut” between the information source and destination. As with the capacity requirement discussed above, whether or not to activate the cellular communication devices depends on the severity of the malfunction/damage and the cost of using a cellular network.
How to cooperate is both system and application dependent. The two cases presented below illustrate the situation with respect to the capacity and delay requirements, respectively. The coverage requirement is not discussed in-depth, since alternative network activation is mandatory and all nodes in the smart grid can be monitored and controlled.
Case 1: Microgrid Operation and Control. Decentralized mechanisms are promising solutions for microgrid operation and control to avoid a single point of failure and fit the plug-and-play nature of distributed generation units and loads in microgrids. In our previous research, we developed a decentralized economic dispatch approach such that each DG unit makes the optimal decision on power generation locally via low-cost short-range communication devices. To avoid a slow convergence speed caused by the limited network capacity, a cooperative networking architecture is used: Some DG units or loads are equipped with dual-mode devices with cellular communication capabilities. The cellular devices can be periodically activated to improve the convergence speed of decentralized economic dispatch.
Case 2: Protection Coordination in Smart Distribution Systems. Take overvoltage protection as an example of one of the most challenging issues arising with randomized power injection by DG units. According to the recommendation of IEEE 929 standard, a solar unit’s operation voltage can be anywhere from 88 percent to 110 percent of the nominal voltage. While an overvoltage in the range of 110 -137 percent results in the trip of a DG unit within 2 seconds, the DG unit should be disconnected immediately (within 0.033 second) if the voltage rises beyond 137 percent. But the overvoltage does not necessarily cause overvoltage at load buses because of the voltage drop along the feeder. Thus, to avoid nuisance trips, the load buses’ voltages should be reported within a specified time limit. WiFi and ZigBee devices can be used for a multi-hop voltage report under normal operation and slight overvoltage conditions (that is, 110-137 percent overvoltage), while the cellular network can be used for an emergency report for a significant overvoltage of 137 percent or more.
In addition to the two cases discussed above, cooperative wireless networking can also facilitate other smart grid applications whenever there is a mismatch between quality of service requirements and quality-of-service provisioning in a single network. For a utility company with multiple options on wireless network infrastructures, fully utilizing each network’s capability via cooperative wireless networking not only improves power system efficiency and reliability but also reduces power system operation cost. The reduced cost leads to lower energy bills for customers, which is another advantage beyond high-quality power delivery.
It is worth mentioning that despite all the advantages, the cost of deploying dual-mode/multi-mode communication devices is not small. Moreover, as cooperative wireless networking technology is in its infancy, it will take time to realize and implement the benefits of cooperative wireless networking in the future smart grid, as demonstrated by academic theoretical studies and real world industry experiments.
Hao Liang, a member of IEEE, is a postdoctoral research fellow in the Department of Electrical and Computer Engineering at the University of Waterloo, Canada. His current research interests include microgrid operation, plug-in electric vehicle energy management and energy-efficient wireless networking. He was the recipient of the Best Student Paper Award at IEEE’s 72nd Vehicular Technology Conference, held in fall 2010 in Ottawa.
Weihua Zhuang, an IEEE fellow, has been with the Department of Electrical and Computer Engineering, University of Waterloo, Canada, since 1993, where she is a full professor and Tier I Canada Research Chair in wireless communication networks. Her current research focuses on distributed network control and service provisioning in wireless communications and on smart grid. She is a fellow of the Canadian Academy of Engineering and the Engineering Institute of Canada, and a member of the Board of Governors of the IEEE Vehicular Technology Society.
Xuemin (Sherman) Shen, an IEEE fellow, is a professor and University Research Chair in the department of electrical and computer engineering, University of Waterloo, Canada. His research focuses on resource management in interconnected wireless/wired networks, wireless network security, wireless body area networks, vehicular ad hoc and sensor networks, and the smart grid. He has taken responsibility for technical program planning for several major IEEE conferences, including Globecom'07, VTC’10. He is an Engineering Institute of Canada Fellow and a Canadian Academy of Engineering Fellow.