Self-Healing in the Age of Big Data

Back to Top

Lighted Cityscape

Big data has arrived, and most utilities likely didn’t notice. Depending on their total customers served and the number of substations and grid-connected devices they have, electric utilities manage from tens of thousands to millions of DNP3 and/or IEC 61850 data points in their grid operating systems. Managers want to have SCADA visibility to most or all grid devices. But add in smart-meter, customer-service, and internal utility-support systems for day-to-day operations, and we’re talking terabits of data.

The challenge is how to sift through the information, identify the important (actionable) data points, and take action. Here are three options for managing data for utility grid operations:

Conduct a manual analysis of the data and the resulting actions to be taken. For example, in a fault situation where there’s SCADA control center visibility and manual control, the console operator must interact with the crew reports to identify the fault location, isolate the fault, manually calculate loads to be transferred to other sources, determine whether an alternate source has enough capacity, and issue switching commands to field crews. This system offers little benefit because it is very time-consuming and difficult to keep up with. The only solution is to have a control center with a significant number of operators, increasing the operating costs of the utility.

Use a system that evaluates the data and has an output of “suggested actions” that shine a spotlight on items of significant importance. For example, a self-healing system will analyze the outage situation and suggest the next action for the operator. The operator performs his own analysis and confirms or rejects the suggested action. Users of this system typically cast doubt on the suggested action, reject full autonomy, and desire manual confirmation. The benefit of this system over a fully manual system is it filters out non-actionable data and allows the operator to focus on actionable information. The system’s drawback is that, after an issue is brought to the operator’s attention, it becomes a manual system and carries with it the same negative aspects of manual operations.

Fully automate the system to evaluate the data and to act on items requiring attention. For example, a fully automated self-healing system detects the fault location, determines the transferable loads, and issues switching commands, all while the SCADA center sees the operations performed. The system must be mature and well-tested to eliminate any error potential. As such, it must be pressure-tested on various fault scenarios through factory acceptance testing prior to implementation. The testing should involve actual controls and use system logic with programmed circuit arrangements so the self-healing system is placed under the conditions in which it will operate. The benefit of this system is speed of action, which results in improved reliability indexes, less impact on customers, and lower operational costs for the utility.

Because data availability will continue to grow and customers will have increasingly high expectations for electric service continuity, have you given thought on moving your system from manual operation to a fully automated self-healing system?

I’d be interested in learning your thoughts on this issue in the Comments below.

专家

Soren Varela

出版日期

十月 21, 2019