When there is significant penetration of Smart Grid techno­logies – Advanced Metering Infrastructure (AMI), Demand Response (DR), Distributed Energy Resources (DER), including Electric Storages and Plug-in-Electric Vehicles (PEVs), Power Electronics (PE), and advanced communications, the existing ADA applications will need significant upgrades, and
some new applications will need to be developed. With the diversity of the new technologies and their utilization in a
Smart Grid, the modeling of the components of distribution grid operations and their impacts on other power system
domains must be reviewed.

The following ten ADA applications are described in the IntelliGrid project:

1.    Real-time Distribution Operation Model and Analysis (DOMA)
2.    Fault Location, Isolation and Service Restoration (FLIR)
3.    Voltage/var Control (VVC)
4.    Distribution Contingency Analysis (DCA)
5.    Multi-level Feeder Reconfiguration (MFR)
6.    Relay Protection Re-coordination (RPRC)
7.    Pre-arming of Remedial Action Schemes (PRAS)
8.    Coordination of Emergency Actions (CEmA)
9.    Coordination of Restorative Actions (CRA)
10.    Intelligent Alarm Processing (IAP)

In this article we will briefly review the existing design of three major applications (the first three in the above list), and discuss the expected upgrades of these applications needed to meet the Smart Grid requirements. These applications, in the current state, had been implemented in a number of utility pilot projects and in ongoing operations (e.g., BC Hydro, FPL, JEA, Progress Energy Florida, OPPD).

Near-real-time Distribution Operation Model and Analysis (DOMA)
This application provides the situational awareness of distribution system operations. Currently, it is based on input data collected from various corporate databases, SCADA, and operator’s entries. In the Smart Grid environment, the multifunctional AMI system, customer EMS, market and weather IT systems will become significant sources of information support for the ADA applications. DOMA will process these various input data into a near-real-time and short-term look-ahead comprehensive models of distribution operations, to be used as a base for other ADA applications and to provide the operators with analyses of the behavior of the distribution system. A high level illustration of the information flow for DOMA application is presented in Figure 1.

The DOMA functionality is based on the following component models:

•    Model of transmission/sub-transmission system. This model needs to account for the impact of the distri­bution operations on transmission operations.
•    Model of distribution circuit connectivity. This model is supported by the GIS database for nominal connectivity and by SCADA and operator’s input for real-time updates. Improvement in the timely updates and comprehensiveness of GIS databases is needed.
•    Models of distribution nodal loads. At the present time, the nodal load modeling in distribution is based on ‘typical’ real load shapes and expert estimates of the power factors for a number of load categories and on the monthly billing data. In the Smart Grid environment, the concept of ‘typical’ load shape is not applicable due to the diversity of possible behavior of the many small, distributed generators, electric storage devices, plug-in electric vehicles, and demand response means scattered among many customers. The real and reactive load models – individual or aggregated – shall reflect the behavior of these composite loads depending on the known weather, prices, voltage, time of day, and other factors. The AMI data, if properly processed, will provide a much better distribution nodal load models.
•    Models of Distributed Energy Resources and Micro-grids. As the minimum, the DER models should be sufficient to estimate the generated kW and kvars at any given time, the financial attributes, and the capability curves. These models can be supported by SCADA, Customer Information Systems, DER and AMI data management systems, by aggregators, and by weather forecast systems. The models of the Micro-grids can be viewed by the utility as aggregated object at the point of common coupling, or they can consist of the individual element models within the micro-grid. The utility should know the aggregated impact of micro-grid separation on both the utility and on the micro-grid.

•    Models of distribution circuit facilities. These models, in addition to the conventional facility models, include the models of the secondary circuit equivalents. Presently, these models are developed based on expert estimates and may significantly differ from the real objects, resulting in large errors of voltage modeling and, consequently, in reducing the operational tolerances. Based on the voltages and powers measured by Smart Meters, adequate secondary equivalents can be derived.
•    Model of distribution power flow/state estimation. Under conditions of the Smart Grid, the power flow/state estimation will need to additionally model the price-dependent events, and solve radial and meshed networks with multiple generation busses in different modes of operation.

The modeling discussed above will increase the accuracy of the power flow model, thus supporting a better utilization of the distribution systems.

The analysis part of the DOMA application includes the following analyses:

•    Analysis of adequacy of distribution system operations. The adequacy of the operations is defined by the loading of the distribution elements, by the transfer capacity of normally open ties, and by the consistency of the fault currents with the capabilities of distribution facilities and protection settings. Under the Smart Grid conditions, the transfer capacity analysis shall take into account the availability, impacts, and cost of involvement of DER, Micro-grids, DR, PEV, ES, and Feeder Reconfiguration and Volt/var/Watt control applications. The fault analysis will also estimate the impact of the fault on the status and operations of the DER.
•    Power quality analysis. Presently, the power quality analysis of the DOMA application analyzes the voltage deviations and voltage imbalance calculated by the power flow model. In the Smart Grid environment, this sub-function will analyze the voltage deviations, sags and swells measured and collected by the AMI system, will analyze the correlations between higher harmonic levels and operations of shunt devices and power electronics, including converter-based DER devices.
•    Analysis of the economic efficiency. The economic efficiency can be determined in different ways depending on the utility business environment and objectives. The following components of the economic efficiency of distribution operations can be suggested: 1) Evaluation of the incremental cost of delivered energy by components, one of which is the cost of energy losses, and 2) Evaluation of the incremental benefits due to a particular change in distribution operations implemented in the utility. The incremental cost may include the cost of supply from both bulk energy sources and distributed energy sources, the incremental cost of demand response incentives, the cost of losses, the penalties for limit violations, etc. The evaluation of the incremental benefits of “what-if” operations can be done by DOMA in the near-real time mode with pre-defined changes calculating the difference between the actual operations and the “what-if” operations.
•    Determining the dynamic T&D bus voltage limits. Presently, in many cases, the T/D bus voltage limits are constant for an extended time interval. The dynamic optimization of the distribution system operations results in different optimum voltages at the distribution side of the T&D substation. These voltages can be supported within a certain range of the transmission-side voltages. This range defines the transmission-side voltage limits at the time of optimization. There may be another set of dynamic voltage limits: the power quality limits, when the voltage at the buses shall satisfy the standard voltage tolerances at the customer terminals. The dynamic voltage limits defined by DOMA should be submitted to the transmission domain for use in the Wide Area Situational Awareness applications.
•    Determining the available dispatchable real and reactive load at the T&D buses. The significant penetration of DER, Demand Response, and PEVs in combination with Volt/var/Watt control and Feeder Reconfiguration applications will provide wide ranges of dispatchable loads at the T&D buses. These loads will be dependent on a number of conditions, such as real-time energy prices, reliability signals (can be price also), ancillary service conditions, temporary voltage limit for peak load reduction, weather, etc. Hence, the dispatchable loads at the distribution side shall be also based on behavioral models.

•    Determining the aggregated at the T&D buses parameters of remedial action schemes. In many cases the
actuators for load-shedding Remedial Action Schemes (RAS) are located in the distribution system on per feeder basis. In the future, the load shedding could be done in a more refined manner moving it closer to the end users, e.g., using micro-grids, operating in absorbing mode. The Wide Area Measurement and Control System (WAMCS) should define for each moment the amount of load to be armed at different RAS to satisfy the power security requirements. The ADA application should support the model of available loads under different RAS, their interrelationships, and their behavior under different circumstances.

Fault Location, Isolation and Service Restoration (FLIR)
In the current design of the ADA application, the fault location is based on SCADA-supported fault indications, trouble-call systems, and, sometimes, on fault-locating devices. In the Smart Grid environment, the Smart Meters, customer EMS, and fault predictors will become significant sources of information for fault location. The processing of these, sometimes, voluminous data will need to be accomplished in a very short time interval.

The switching orders generated by the application for fault isolation and service restoration, will include, in addition to switching devices and feeder paralleling, separations of micro-grids, synchronization of disconnected DER, and enabling of DR. The solutions should be dynamically optimized based on the expected operating conditions during the time of repair. The FLIR application should be coordinated with other ADA applications, such as MFR, VVWO, RPRC, and CRA.

Voltage, Var, and Watt Optimization (VVWO)
This is a major multi-objective ADA application performing dynamic optimization of the distribution operations taking into account all significant impacts of the application on the operations in different domains (Figure 2).

In the Smart Grid environment, in addition to the current control of voltage controller settings and feeder capacitor statuses, the application should be able to control the reactive power of DER and other dynamic sources of reactive power. Under some objectives, the application should be able to control the Demand response means and the real power of DER. Therefore, the Volt/var optimization becomes a Volt/var/Watt optimization.

As follows from the above discussion, the ADA applications should actively exchange information with applications and IT systems in other power system domains. A high-level information exchange diagram between the Distribution Grid Management and other systems is presented in the EPRI Report to NIST on the Roadmap for Smart Grid Interoperability Standards.

A more detailed illustration of the information exchange just between the DMS and EMS is presented in Figure 3.

Conclusions

•    The existing designs of ADA applications can be used as foundation of the ADA applications in the Smart Grid environment.
•    Comprehensive near-real-time and short-term look-ahead models of the behavior of the Smart Grid components are the basis of ADA applications in the Smart Grid environment.
•    The upgrade requirements for such models may define the specification and prioritization of AMI, DER and DR.
•    Integration of Smart Grid technologies into ADA applications also implies the ability to optimally control some of the new objects.
•    Active exchange of information between the Distribution Operation domain and other power system operation domains will be needed for comprehensive dynamic optimization of power system operations.
 

About the Author

Dr. Nokhum Markushevich has more than 45 years of experience in power system operations, research, and education. He worked for the Latvian Power Company for 29 years and has been with Utility Consulting Inter­national (UCI) since 1992, where he consults in the areas of EMS and DMS. He was the Distribution Operations Domain Leader in the EPRI IntelliGrid project and had actively participated in the development of the EPRI Report to NIST on the Roadmap for Smart Grid Interoperability Standards. Dr. Markushevich has authored several books and more than 100 papers on power system operations, EMS and DA.