AM/FM/GIS Modeling to support Power System Planning
John W. Hardin Intergraph Public Safety Mailstop LR23A1 One Madison Industrial Park Huntsville, Alabama 35894-0001 Karen Bachmeyer Intergraph Public Safety Mailstop LR24B2 One Madison Industrial Park Huntsville, Alabama 35894-0001 Abstract The GIS contains a warehouse of geographic based information that is tracked in numerous systems throughout the utility. With some common sense modeling techniques, a basic appreciation of circuits, and knowledge of the vendor specific analysis package, this data can support electrical analysis. Often, the facility data is simply extracted from the GIS to an analysis package to perform load flow analysis, fault analysis, motor start analysis, capacitor placement optimization, and switching optimization. With current architectures, these capabilities can be built into the GIS providing the distribution engineer a set of powerful analytical tools in a single environment. In addition, switching status from the outage system and metered data from SCADA/MV90 can be leveraged to analyze summer and winter, heavy and light-loading situations. Transformer load statistical analysis output can be used to derive load data for the electrical analysis. Modeling GIS Data to support Analysis The GIS contains a warehouse of geographic based information that is tracked in numerous systems throughout the utility. With some common sense modeling techniques, a basic appreciation of circuits, and knowledge of the vendor specific analysis package, this data can support electrical analysis. Common Sense Modeling Techniques Most GIS facilities represent items of plant, though some are only logical. When it is possible, model electrical devices as they exist in the field. This will simplify the maintenance of the data in the GIS. Also, it increases the likelihood that you will later be able to look-up electrical properties for analysis using standards books. Try to keep things as simple as possible. For example, let us examine the modeling of switchgear. Typically, switchgear is purchased as a single unit. However, the connectivity of switchgear is relatively complex and the switches and/or fuses contained in switchgear may be replaced individually. If you model the switchgear as a composition of buses, fuses, and switches, modeling the connectivity becomes simple. To avoid confusion in the GIS, most companies place a logical facility representing the switchgear cabinet to tie everything that composes the switchgear back together though the switchgear cabinet is not part of the electrical model. A good question to ask yourself as you are designing the GIS facilities for electrical modeling is, “How will this data be maintained?”. Since much of the information you need may be known by field personnel, you may need to examine your work processes to make sure that this information is maintained in the GIS whenever feasible. Part of determining the feasibility of collecting the data is weighing the cost of maintaining the data against any benefit you may achieve. If you determine you cannot maintain the data over the long haul, then you need to re-think what you are doing. A Basic Appreciation of Circuits It sounds obvious, but connect facilities in the GIS as you need them to be connected for the electrical analysis you plan to perform. In a node-edge model, series devices need to be modeled with two nodes (e.g. a conductor, switch, or fuse) and shunt devices (e.g. a shunt capacitor) can be modeled with a single node. Typically, source and spot load objects are generated from other GIS facilities. Source objects are used to model where power is being injected into the network and spot loads indicate where power is being absorbed from the network. For example, a substation transformer may be used to model a source, while a secondary meter or distribution transformer may be used to model a spot load. If a series device needs to break the electrical flow, then it should always be modeled with two nodes. In other situations, the decision may not be quite so evident. Consider a bus for a moment. A bus acts a connection point, so you may consider modeling it as a single node facility, which implies that the electrical properties of the bus are inconsequential. However, if you consider electrical properties of the bus significant, you may deem it appropriate to model the bus with two nodes and treat it similar to a conductor for analysis. By most estimates, buses are relatively short and have minimal impedance, so it is acceptable to model them as single node facilities in the GIS and neglect them in your analysis. Secondly, consider a distribution transformer. You could model the transformer and all of the devices on the secondary side of the distribution transformer right down to the secondary meters (which you could model as load objects), but it greatly increases the size of the network that you are to analyze. Another approach would be to model the transformer and associate a load object to the secondary side of the distribution transformer to account for the load absorbed by all of the devices downstream of the distribution transformer. An even simpler approach would be to simply model a distribution transformer as a load. In either of these last two scenarios, you will need to create a relationship between the distribution transformers and the secondary meters they feed if you plan to derive load information from these secondary meters. Your short-term needs should be your highest priority, but you need to consider what analysis you may want to pursue in the future. There may be some simple things that you can do in your GIS model now that will avoid a lot of re-work and data collection later. For example, though it may not be necessary for analysis of your primary network, you may want to go ahead and maintain connectivity between your secondary devices if you plan to analyze your secondary network in the future. All things considered, try to keep your model as simple as possible. Knowledge of the Vendor Specific Analysis Package Fundamentally, we are trying to model a physical situation using mathematical techniques. It is not always practical to collect/maintain the data necessary to support the more complex/sophisticated algorithms, so we have to strike a delicate balance between the complexity and accuracy of the analysis. For example, let’s examine the phase rotation of conductors. The impedance of the circuit can be more accurately determined knowing the phase rotation of the conductors; however, most utilities find it impractical to gather and maintain this information and therefore choose to ignore this affect. All electrical analysis packages need the same basic information, and generally large portions of the electrical properties of the connected network are determined using catalog tables. To support analysis using your GIS data, you model the conductors and devices with some basic attribution, but you retrieve many of the electrical properties of the items from catalog tables. This avoids complexity and redundancy in your GIS data storage. Leveraging Enterprise Data Further leveraging the wealth of information in the enterprise, switching status from the outage system and metered data from SCADA/MV90 can be used to analyze summer and winter heavy and light-loading situations. Transformer load statistical analysis derived from customer billing information can be used to derive load data for the electrical analysis. Typically, only the design status or planned configuration of switching devices is represented in the GIS. This is fine for some studies, but it may not be OK for others. If you have an outage management system (OMS), you can get the operations configuration of switching devices at defined points in time. However, be aware this may not represent the “real” network configuration if cuts, jumpers, or other devices that change the connectivity are inserted in the OMS that are not represented in the GIS. Also, consider that the GIS may have updates that affect the network configuration that may not yet be reflected in OMS. In the end, you will have to weigh the inaccuracies that may be introduced by this approach against the possible benefits. Another enterprise data source that may be leveraged is the voltages and currents being measured by SCADA/MV90. You may have to do some work to tie the SCADA/MV90 data to facilities in the GIS, and you need to consider what the actual network configuration was when the readings were taken. You may be able to derive satisfactory load data from your customer billing information. The monthly reading can be used as a weighting factor. Another weighting factor can be generated based on the customer category and the time of day. Lastly, the data can be scaled using SCADA/MV90 voltage/current values where you have meters. This will not give you a precise reading at a point in time for a device, but it may be satisfactory. Complimentary Application Architectures Often, the facility data is simply extracted from the GIS to an analysis package to perform load flow analysis, fault analysis, motor start analysis, capacitor placement optimization, and switching optimization. Through complementary application architectures, these capabilities can be built into the GIS to provide the distribution engineer a set of powerful analytical tools in a single environment. The integrated environment is beneficial in many ways. The user has access to all GIS functions without the mental and physical context switch between applications. He has the ability to easily include proposed changes – near term adjustments as well as long-range planning information. The user can immediately adjust his designs based on feedback from the integrated analysis. In separate applications, the user would have to make notes of the changes to make and then find that data within the GIS. In the integrated workflow, the user is in the right context immediately. Additionally, the integrated environment provides a better geographic presentation of the data including backdrop data (landbase and street context). Finally, enterprise data is readily available without the prerequisite extract and import into the analysis tool. Summary You can build a partnership between your GIS application and your analysis program. By considering aspects of both applications as you configure your data model, you can overcome the daunting level of details. You must strive for simplicity of implementation so that you build a maintainable integration. Enterprise data can augment the intelligence of the analysis. This partnership will reap long-term benefits for your organization. | ||
|
|