Why Outage Management Must Change


Smart grid advancements call for a new approach to restoration.

Fortnightly Magazine - April 2014

In the not-too-distant past, utility distribution networks were managed by two non-competing technologies: outage management systems (OMS) and distribution SCADA (supervisory control and data acquisition). Today, this model is changing rapidly. Smart grid advancements have brought new combinations of advanced technologies, often working together in a real-time operation or a real-time network. In the smart grid, the technologies that are applied must work together dynamically; by their very nature, they must adapt to changes in the grid in a coordinated fashion. Changes to the network brought about as a result of one technology or system usually will have an effect on another system, but there remains only one real-time network state at any given time.

For example, two common smart grid automation technologies, feeder self-healing and power optimization, must be able to interact effectively. Many smart grid solutions with power quality improvement as their business objective are applying technologies such as loss minimization (LM), which performs capacitor switching to reduce losses. In other cases, conservation voltage reduction (CVR) is applied by reducing voltage through coordinated switching of regulators and load-tap-changing transformers. The coordinated combination of each for integrated volt-VAR control (IVVC) provides additional benefit.

These applications can't assume to know the real-time network connectivity status of the devices they're controlling. Operation of self-healing systems can result in the switching of regulators and capacitors from one feeder to another, so related smart grid systems must be able to adapt to these network changes. The problem becomes more pronounced as distributed generation or distributed energy resources are applied to the feeder operation, since this will affect the voltage profile for the downstream feeder. Likewise, these smart grid applications all must adapt their operations in a coordinated fashion. For optimum efficiency and safety, they must operate from a common topology model where all applications maintain awareness of the real-time network state, and they in turn must use the real-time state in their respective solutions.

The traditional OMS now must operate in this new dynamic distribution network environment. In a way, the OMS competes with other operational systems for information, overlapping functionality, visualization, and maintaining the network topology.

Integrated Real-Time System

Many utilities build a smart grid infrastructure without considering the cost to maintain it. The largest single cost, accumulated daily over the life of the system, can be data maintenance. This maintenance includes the database itself, the detailed graphical map display and the network model.

The utility executive must ensure that the smart grid applications use a unified network model unencumbered by the traditional segmentation of data into operational silos. For example, existing network connectivity models previously were maintained only in the GIS (geospatial information system). As a result, many OMS vendors and utilities built a system tightly coupled with the GIS model. In this case, customer operational data, such as trouble calls, is sent to the OMS. Data from the utility's AMI system also is sent to the OMS. The integration of SCADA systems into this equation typically wasn't a concern, since the network model wasn't needed for distribution operations prior to the advent of the smart grid.

With the introduction of advanced distribution management systems (DMS), the question becomes, "Where should the DMS functionality reside? In SCADA, where the real-time field telemetry and control exists? Or in the OMS, where the topology model and trouble calls are managed?" The DMS needs data from both. Driven by the importance of real-time telemetry and control requirements, the DMS - and a more involved network model with enhanced load flow analysis capability - increasingly are being interfaced with distribution SCADA.

Now what about the OMS? A utility desiring a true smart grid operational system can't afford to maintain different subsets of the network model in both the OMS and DMS. Nor can it afford to segregate operational data between two distinct systems, where any one system doesn't possess all of the critical network information. Utility operations personnel simply can't afford to deal with siloed data in three different systems, or hope to synchronize them effectively (or cost-effectively) through complex and customized data links.

Nevertheless, many utilities still send SCADA events to the OMS system via an external data link, only to discover that the OMS is incapable of handling a real-time flood of events in a storm scenario. When the system is needed the most, the overburdened (non-real-time) OMS will simply crash. This is why some utilities "pull the plug," or break the link, between the systems to prevent this occurrence when faced with a crisis situation such as an impending Hurricane, but this is no way to operate a smart grid control center; it's certainly not a strategy for success.

Out of necessity, the next generation smart grid system must become an integrated real-time system performing SCADA, DMS, and OMS functions using a single, common representation of the distribution network. This framework is critical for accuracy of the data maintenance, as well as cost efficiency and operational effectiveness. The new system must be able to process enormous amounts of data from an increasing number of grid sensors and intelligent electronic devices (IED), as well as from AMI systems and customer call centers. It also must accurately represent the network state in real time, providing operations personnel with enhanced situational awareness, whether they're in a control center or in a bucket truck.

An advanced DMS is operated using a real-time map - often a GIS-based map - of the distribution network. Many operators currently use a map in the OMS. However, the OMS typically doesn't support many of the necessary control center operational functions. System operations, such as tagging, present important safety concerns since they include hold and lockout tags on devices in order to protect crews. Safety tags can't be placed on the system simply with the hope that valid data transfer will occur between systems (e.g., an OMS with a data link to the SCADA system). Such a scenario isn't only dangerous; it's expensive. The operational system of record - the SCADA system or DMS - must be the one used to create and enforce this safety tag.

Likewise, the telemetered analog values also can be critical for safety when used to indicate, for example, that a circuit is de-energized. OMS doesn't support the ability to present real-time telemetry on a network display that's needed on the feeder or at the substation.

Another safety concern is presented by the operation of the system using two distinctly different user interfaces (i.e., DMS and OMS), with neither showing the operator a truly complete picture of the data. This generally wasn't an issue for the classical OMS approach, when telemetry and remote control use was confined primarily within the substation and didn't extend to the distribution feeder. The evolution of the smart grid has changed that.

So how can the network be managed effectively during a crisis when widespread damage has occurred? Under this scenario, the normal rules of engagement no longer apply. The operational practices and rules must transcend the situation in order to manage the system during this time. The smart grid OMS must be a real-time system and must be capable of managing the restoration of segmented damaged areas, from areas where only typical disruption occurred. In case of major damage from a storm event, outage calls from customers have little meaning because the prediction engine will indicate that the substation breaker is open, since all customers on the affected feeder (or feeders) are out of service. In this case the mode of operation and reporting must change to match the crisis. The combination of SCADA, DMS, and OMS must be able to handle the extra load, using non-traditional processes in order to help utility personnel effectively complete reconstruction and restoration.

A modern, real-time DMS and OMS must be designed to handle the massive number of real-time events in a crisis and must be able change the way it manages and presents system data - not only to the operators and field staff, but also to customers. In these situations, the OMS is no longer the lone player in the restoration process; they now require real-time performance capability. What's evolving from this reality is a public management system, providing for customer safety, security, and service.

Toward Live Testing

While every industry depends on simulation studies, it's rare to find a distribution utility that owns a transient network simulator or other training tool that can reproduce actual operating situations accurately in a real-time setting. Future critical infrastructure protection standards likely will necessitate a testing platform for all new smart grid deployments on the distribution network.

Such a distribution system simulator is more than a training system. It's a tool for the simulation, analysis, and verification of the performance of smart grid technologies of all types. An effective simulation is performed with the same feeder network model as the utility's production DMS, using both historical and real-time system data, as well as current geospatial data from the GIS. This information is integrated in a relational database that uses an accurate connectivity model to enable simulation with a variety of applications, such as load estimation and volt-VAR control.

Users of the new generation of smart grid OMS aren't restricted to the control center. Mobile users such as field crews require not only outage information, but work assignments, geographic network topology, and real-time data from the network. The information required by these users goes well beyond what's available from the typical OMS. It's a blend of data from many sources including AMI, real-time telemetry, calculations, and customer calls. A unified tool for switching orders integrating SCADA, DMS, and OMS is evolving that will benefit not only traditional users like field crews, but also critically important and demanding new users - the utility's customers. A truly integrated operational system is able to serve not only the utility but also the public at large with a blend of useful and informative data from all available sources.

The smart grid is an integrated set of applications and technologies. If those various systems don't coordinate, they won't function together as a team. The hybrid approach to OMS is the next stage of our industry's evolution. It changes the way we perform under normal, abnormal, and crisis scenarios, using new tools and a greater degree of information in an intuitive way.