The smart grid is opening the floodgates on customer data, just as consumers are getting comfortable with retail self-service and mobile apps. With dynamic rates, distributed generation and...

## New Directions in Distribution Management

Advanced systems turn ‘event-driven’ binary schemes into hybrid hierarchical controls.

and voltage vary in proportion to one another across the network, solutions to accurately compute voltage and current flow at any node are iterative. Some software uses a closed-form voltage drop computation, but those are generally estimated and based on assumptions that can limit accuracy. For DMS to work well in a smart grid context, the loadflow calculations must be accurate.

Power system engineers use the term “state estimation” to mean the ability to monitor certain points in the network for things like voltage and current, and solve those parameters for other, non-telemetered points of interest. This technique is non-trivial, and has been applied to the sparse networks of transmission systems for many years. For distribution networks, the problem becomes much more intensive and state estimation has to work hand-in-glove with the loadflow method chosen for analysis.

For DMS to do its job, it must be able to to accept real-time data from SCADA systems and other sources, and incorporate that information into its network solver. This capability is closely related to the state estimation function described above and adds another level of complexity. That’s because as real-time parameters change, the state estimator must be capable of solving and resolving the network, deciding which, if any, of the monitored parameters justify a re-running of the loadflow calculations.

Among the most important functions for a modern DMS are fault location and service restoration. This capability is an extension of basic short-circuit current and voltage computations. The DMS uses fault calculations to help system operators plan protective schemes, analyze system failures and plan service restoration.

In addition to these core capabilities, there are two more important characteristics for DMS. North American style networks require that network solvers be capable of solving three-phase, unbalanced systems. In European networks, a three-phase balanced solution is adequate, since nearly all distribution systems operate with all three phases in balance. But North American networks often consist of segments of single-phase circuits, so that the simplifying assumptions of the balanced solution do not apply. Most importantly, modern DMSs must have the ability to quickly solve large, complex networks, often with hundreds of thousands of nodes or more. In the complex equation of loadflow + state estimation + real-time integration, solving a large network in near-real-time is a challenge. But high performance is the price of effective control of the distribution grid.

**Capturing Dollar Benefits **

It might come as a surprise to some, but tightening the control of the operation of medium- and low-voltage systems for greater efficiency can lead to significant savings for the utility and its ratepayers. Operating efficiencies can come through a wide array of techniques, including balancing phasing, managing system voltage and power factor, and optimizing network configuration. To take one example, most utilities today operate the distribution grid in a fairly static configuration. At most, they may be able to reconfigure feeders through switching once a year or so to optimize performance as seasonal loads change. As the grid gets smarter, with more points of switching control and more places where operating parameters are measured, more