Most would agree that the traditional centralized electrical distribution model will evolve to a distributed generation (DG) model. When this occurs, and to what degree remains to be seen. Regardless, a smart grid communications infrastructure is essential in the safe, reliable and efficient management of a DG infrastructure.
For the past couple of years, WireIE has worked in collaboration with the University of Ontario Institute of Technology (UOIT) in developing a model for a smart grid distribution system of the future. Faculty in the university’s Electrical Engineering & Applied Science program, along with their students, have modeled a number of distributed generation scenarios from the utility’s perspective. One of the many outcomes of this exercise has been a clearer specification of communication network requirements to support these distributed generation scenarios.
Communication Network Requirements
A smart grid communications network must support a number of applications, some mission critical, while others are comparatively forgiving. As our UOIT colleagues specify, the operation of taking a distributed generation source on or off line demands execution of the transition in no more than 5 – 6 cycles, or 80 – 100 milliseconds. In contrast, other administrative functions such as a dispatch applications may be tolerant of a number of seconds delay.
With UOIT’s DG scenarios in mind, our most critical communications network specification is latency. Latency is defined as the time taken for an element of data to transcend a link, or series of links, in a data communications network. We therefore need to factor in the very stringent latency requirements of DG while also recognizing that our smart grid communications network will be handling significant volumes of less time-sensitive administrative traffic.
Communications Network Architecture
A smart grid communications network must support protection and control functions at DG interconnection points. These sites include facilities on the grid itself, along with businesses and residences where alternative energy may also to be available to the grid. With a clear delineation between mission-critical operations and those more tolerant of latency and throughput variations, a dual or potentially multi-layered, communications network is envisioned.
One can think of the bottom layer of the network being administrative and housekeeping oriented. It is designed for high reliability but it also has comparatively high forgiveness of latency, along with other network performance variations. Geographically, this layer covers a wide area – potentially all of a Local Distribution Company – and is appropriately referred to as a Wide Area Network (WAN). In contrast, the top layer is composed of several Local Area Networks (LANs). All LANs connect to the WAN so that communication can take place between the Operations Centre on the WAN and remote sites on the network.
Mouse Over the Image to Reveal the LAN Layer
The Drawing Assumes an IEC 61850 Interface as a Demarcation Between Electrical Utility and Communication Network Assets
While this basic topology is by no means revolutionary, the mission-criticality of many protection and control functions will require unprecedented robustness and redundancy – particularly on the LAN layer, and often at the network edge. As is the trend with many modern networks, edge oriented data processing and storage yields significant bandwidth efficiencies, along with a commensurate improvement in network performance and service reliability.
The LAN’s primary purpose is to execute time-sensitive, mission-critical protection and control operations such as a DG source switch-over. It should be noted that DG operational decision making is not the same thing as the actual execution of the operational decision. This distinction is important in that business and operational policies and decision-making do not occur on the LAN. Instead, a centralized operations facility, or perhaps a collection of regional operations centres, are located on the WAN. Among other things, these centres are where operational decisions are made and subsequently delivered to the appropriate LAN. Once an instruction is delivered to the appropriate LAN, local sensing and measuring equipment determine whether conditions are conducive to actual execution on the instruction. The outcome of the instruction (executed successfully, failed) is then delivered from the LAN to the operations centre via the WAN.
Why not consolidate the WAN and LAN layers? The main reason relates to the wide range of expectations placed on the smart grid communication network as a whole. As previously mentioned, protection and control functions are comparatively demanding of the network in terms of reliability and low latency, whereas administrative functions are quite forgiving.
As a self-contained network within a larger ‘network of networks’, the local aspect of a LAN has some very important attributes in supporting protection and control. As a topologically simple, self-contained local network, a LAN is very fast – an essential characteristic in executing protection and control operations. Not only are communication link distances short in a LAN, there are fewer hops (a linear collection of communication links) per communication channel. Multiple hops introduce aggregate latency. An additional inherent benefit of the LAN’s simplicity is reduced points of failure within the LAN itself. In fact in most situations, the LAN can operate autonomously should there be either a planned or unforeseen disconnection from the WAN. Predefined operational policies would stipulate the degree to which the LAN can operate autonomously in the event of a disconnection from the WAN.
Communications Network Technology Considerations
Many DG sources are in locations where limited or no communications infrastructure exists. In these cases deployment of digital radio, or a digital radio/fiber optic hybrid is both attractive and pragmatic.
WireIE’s Transparent Ethernet Solutions™ (TES) are built with exceptionally low latency characteristics – all backed up by a Service Level Agreement (SLA). WireIE TES can be deployed in a point-to-point, or point-to-multipoint topology. For access, Long Term Evolution(LTE) promises very attractive latency characteristics, well within the requirements set out by our friends at UOIT. WiMAX(Worldwide Interoperability for Microwave Access) also shows potential as a Smart Grid access technology — particularly WiMAX 802.16m, recently approved by the ITU.
Single hop latency in a WiMAX or LTE link measured from base station to CPE (customer premises equipment), is typically equal to or less than 10 milliseconds. Aggregate latency must therefore be kept safely below 50 milliseconds on all protection and control paths. Again, containing execution of distributed generation activities to a LAN ensures latency thresholds are not exceeded.
WireIE TES, LTE and WiMAX offer a number of sophisticated capabilities over and above impressive latency characteristics. All employ dynamic radio link quality management capabilities. Throughput is traded off for link robustness in the event the quality of a radio path should deteriorate. The reverse is also true as radio path quality improves. The mechanism facilitating throughput verses robustness is known as adaptive modulation.
It is essential that each digital radio link be engineered to exceptionally strict path propagation specifications because of the mission critical nature of smart grid protection and control applications. This entails exhaustive path analysis and a subsequent network design to ensure that every radio path is never at risk of engaging a modulation scheme below a carefully calculated threshold. As a fixed network, radio link reliability can be achieved with a high degree of predictability. That said, best-of-breed engineering is an essential ingredient from a reliability and performance perspective. In addition, network redundancy and/or diversity must be incorporated into the design, thus enhancing overall reliability and equally important, allowing for any and all network failure scenarios. Further protection against communication network failures must also be addressed as the application layer.
A properly engineered LAN using digital radio technologies such as WireIE’s TES, LTE and WiMAX will provide a safe and reliable platform by which to execute critical protection and control operations such as a DG switch-over. The underlying WAN provides the necessary communications foundation to administer such activities. The WAN also supports the broader administrative, ‘house keeping’ activities envisioned for smart grid.