Marija Ilic 2010-2011 Seminar Series
December 1, 2010
In this talk we present a vision of Dynamic Monitoring and Decision Systems (DYMONDS) as a possible Information Communications Technology (ICT) framework in support of sustainable energy systems. The concept rests on the idea that much could be gained by relaxing spatial and temporal decomposition assumptions underlying today's hierarchical electric power systems. It is suggested that, as the number of actors and their diversity increases, it becomes necessary to capture and manage both inter-temporal and inter-spatial correlations by: (1) embedding prediction tools into distributed resources, both production and consumption; (2) embedding dynamic look ahead decision-making tools by the resources themselves needed to internalize their highly unique temporal characteristics; (3) minimally coordinating carefully designed multi-layered multi-directional information exchange between different groups and the higher-level aggregators and system operators to manage inter-spatial correlations; and, (4) aggregating the results of minimally coordinated scheduling of resources at multiple layers throughout a complex network system to ensure the system-wide objectives are met while choice is made by the groups of system users. Such an approach leads to qualitatively new Just-in-Time (JIT), Just-in-Place (JIP) and Just-in-Context (JIC) functionalities enabling orderly performance in which tradeoffs between different sustainability attributes are internalized and complexity of decision making is distributed across several industry layers. This approach is used to illustrate proof-of-concept to managing large wind power capacity without requiring major new capital investments. Instead, demand responds to signals sent by the aggregators and system operators to respond JIT and JIP and balance the system. We claim that without correlating temporally and spatially actions by intermittent resources and highly flexible responsive demand, Plug-in-Electric Vehicles (PHEVs) and future storage technologies it is not going to be possible to utilize energy in sustainable ways.
In the second part of this talk, formal relations between the objectives of future energy systems and the role of man-made physical and ICT delivery network systems are attempted by building on the recent multi-disciplinary Socio-Ecological Systems (SES) framework proposed by the recent Nobel Prize winner Elinor Ostrom. In particular the role of interaction variables used in engineering modeling of electric power grids and the second-- and deeper-level variables used in the SES framework are related. Monitoring and managing interaction variables, otherwise shaped by the internal dynamics and decision making, becomes key rationale for defining information exchange patterns within a multi-layered complex energy system. It is claimed that ensuring as sustainable service by aligning temporally and spatially groups of energy resources and users becomes key objective of ICT design. Last, but not least, the sensitivity of system-wide performance on how well is this alignment done depends on the type of starting socio-ecological energy system (SEES). An SEES with many small distributed resources will require perhaps much peer-to-peer collaborative information exchange and very little top-down coordination. An energy SES built around economies of scale will require much more hierarchical coordination.