Digitalisation in electric distribution systems is perhaps the most significant trend in the evolution of smart grids.
Distribution systems have been undergoing the transition to the automation control paradigm since the beginning of the digital era in the late sixties. Nevertheless, the scope of work has been mostly limited to substations and mainlines because of the focus on engineering economics and reliability considerations.
In more recent years, the pace and scope of the digitalisation of the power system have grown, due to the confluence of favourable market developments and technology advancements. On one hand, demand and supply characteristics are changing, becoming more dynamic and uncertain, which in turn necessitates faster and better situational awareness, analytics, and controls.
On the other hand, technology advancements on both sides of the electric meter are making it possible to add new digital technologies and integrate them into the ever-evolving utility automation systems.
The electric meter typically defines the ownership boundary between a utility and the customer (or prosumer) where many technical and financial ‘handshakes’ occur to keep the overall system functioning.
Often dubbed the grid edge, this is where digitalisation is fuelling many of the Internet of Things (IoT) applications for the utility sector.
While IoT in general encompasses a myriad of hardware and software technologies, edge computing and analytics in particular are driving the state of digitalisation in distribution systems.
Edge computing and analytics
Why edge and why now? There are a number of technical and business reasons for the synergistic application of edge analytics with cloud technologies, such as:
1. Edge computing and storage are becoming increasingly cost-effective and powerful, giving rise to the emergence of intelligent devices that are cyber secure by design and remotely accessible and manageable. These field devices provide new capabilities to process streaming data at the grid edge and relay the resulting information to the next appropriate automation and control layer.
2. Edge computing provides gateway communications closer to field sensors and controllers (the end-points), enabling both wired and wireless communications links to data sources that were not possible in the past. Development and practice of industry standards for field interoperability are key to expand these capabilities and support future automation needs.
3. Sending all field data to the cloud and having it analysed there can translate into steep storage and compute costs. Performing more computational steps at the grid edge is an effective way to reduce cloud-computing costs. There is also a time value associated with edge data that can be lost when analytics is limited to data-at-rest.
4. Edge computing reduces analytics latency. In some cases, such as closing a local control loop on a detected anomaly, round-trip latencies to the cloud would not be acceptable.
5. Localising data cleansing, preparation, filtering, aggregation, desensitising, and analytics at the edge reduces bandwidth requirements for data transport to the cloud.
6. Aggregating, encrypting, and/or anonymising sensitive data at the grid edge can be an effective strategy to mitigate data privacy and security concerns.
7. Edge computing provides a new degree of resiliency when cloud connectivity is disrupted – either intentionally (e.g., cyberattack) or unintentionally (e.g., act of nature).
8. It provides new opportunities to optimise grid analytics where machine learning model development occurs in the cloud and model execution, scoring and updates take place at the edge.
Future applications and deployment
As the distribution systems evolve towards an active network of intelligent loads and generation sources, it will require more intelligent hardware and software for situational awareness, flexibility, and controllability. A new breed of smart devices is needed that feature flexible edge computing, field interoperability and communications as intrinsic attributes of their design.
The state of digitalisation in distribution systems will be advanced by the cost-effective combination of intelligent hardware and connectivity solutions that form the backbone of IoT solutions. Smart sensor technology, for example, is one of the necessary building blocks that provides the ability to know the system state in both steadystate and transient conditions.
The requirements for grid flexibility and controllability may also be addressed by intelligent switches, transformers and other digitalised grid apparatus. Furthermore, this distributed hardware will also support advanced analytics that employ machine learning, sensor fusion, and other techniques for proactive network management (e.g. by anticipating and addressing operational considerations before they become an issue).
Benefits and impacts
As computing hardware, communication, and sensor technologies continue to advance at reducing costs, the proliferation of intelligent devices and applications for edge computing will continue to grow.
The optimum placement of these devices in the field will also yield capital cost savings, as sensors and monitoring devices will not be required at all edge endpoints.
Many use cases for these types of devices are in progress that generally support the evolution of smart grids, smart cities, building automation and transactive energy. Continued digitalisation will result in more operating efficiencies, less downtime for customers and/or equipment, and faster response to issues that occur close to the grid edge.
Running analytics software engines on top of this hardware will make the devices and applications even more effective in providing benefits to prosumers and service providers.
The path to deployment of edge computing and analytics involves a number of technical and non-technical challenges that requires the attention of smart grid advocates.
Chief among these are standards and technologies for field interoperability, flexibility in deployment and management of edge applications, and availability of a thriving ecosystem of educators (for talent), early-adopters (for technology de-risking) and visionaries (for economical de-risking). With these pieces in place, smart grids will be positioned to realise the full potential of digitalisation at the grid edge. SEI
This article was originally published in the IEEE Smart Grid newsletter and is republished with kind permission.
About the authors
Mirrasoul (“Mir”) Mousavi. Dr Mousavi spearheads strategic initiatives and technology development projects. His current professional interests are energy systems automation, DER integration, and machine learning/grid analytics applications. He holds a PhD degree in electrical engineering from Texas A&M University. He was formerly with ABB and is now with Sentient Energy.
James Stoupis is manager of the Grid Digitalisation and Intelligent Systems group in ABB Corporate Research in Raleigh, NC. He oversees technology development focusing on mostly grid-related applications. He holds a BS degree in electrical engineering from the University of Illinois at Urbana-Champaign, and a MS degree from Virginia Tech.