Skip to main content

Beyond SCADA for Electric Power Infrastructures

Researcher: Marija Ilic


Beyond SCADA for Electric Power Infrastructures

This proposal is motivated by the need to assess and design information technology (IT) framework for provable performance of complex physical network infrastructures.  At present, the IT is pursued somewhat disjointedly from the objectives of the system performance.  In particular, we suggest that a model-based approach to defining information (i) type, (ii) location, and (iii) rate within a complex network infrastructure is needed in order to enhance network performance during extreme events.  While many of the network infrastructures are studied by the non-overlapping communities, they share striking similarities with respect to their structure and objectives.  Examples of such infrastructures are gas-, electric power-, water-, and several other network systems.  We have recently formulated a rigorous R&D agenda in a recently submitted ITR proposal to the National Science Foundation [1].  However, we would like to start this work sooner, and this is why internal seed money is being sought as well.

The seed grant will be used to critically assess information currently used for real-time monitoring and decision making in electric power networks, in particular.  This system known as SCADA has inherent characteristics that do not enable the electric power grid to respond reliably under unusually stressful conditions [2] – [4].  In particular, current SCADA relies on excessive data redundancy (all line flows and inputs are measured):  because of this excess data and its inherently static nature (the methods used do not exploit trends in system changes), the SCADA system (i) is non-robust with respect to topological changes and large variations in operating conditions, and (ii) offers results that are hard to use because of the excessive output data which is not post-processed into transparent information needed for control actions.

Despite these features, SCADA has worked well during normal operating conditions.  The objective of our proposed work is to generalize operating paradigms currently used in electric power grids in order to ensure control logic during extreme events.  The overall approach is the one of multi-layered model aggregation for establishing the most critical measurement locations and for designing flexible control.