Integration of asset management and GIS

Integration of asset management and GIS

Gene Kindrachuk
Convergent Group
6399 South Fiddler's Green Circle, Suite 600
Greenwood Village, CO 80111
(303) 741-8400

What is asset management?
It is important to define and understand what asset management is in the context of the modern utility. A utility must have a total inventory of its assets both from a financial and an operational perspective. Tracking assets from a financial perspective is necessary to know what the original and depreciated values of assets are. For current assets, this is required for capital budgeting purposes. For existing assets, it is necessary for depreciation and retirement purposes. There are often additional assets that are tracked for maintenance and reliability purposes that are not individually tracked for financial purposes, generally because costs are low and quantities are large (i.e., switches and fuses). Their costs are loaded in as overhead on other assets that are tracked. From a utility distribution (and transmission) operational perspective, assets are not only those that represent a particular financial interest, but also those against which any type of work may be performed, or those for which historical operation, inspection, and maintenance information needs to be maintained.

Asset management is, therefore, the maintenance of detailed information about various types of equipment that support distribution operations, and the processes associated with analyzing this information to plan work to be performed on these assets to maintain them at an operational level in a cost-effective fashion.

Asset Management Systems are generally implemented as an integral part of a work management system in the operational business unit of a utility. Many of the commercially available work management systems provide capabilities to record inspection and operational histories, and to establish specific criteria, which when met or exceeded can automatically trigger the creation of work orders against that equipment. These criteria are typically:

based - fixed period of time since last work performed
based - quantified levels of equipment condition-recorded from physical inspection or remote monitoring
Operations based - a preset number of operations of a particular type, again determined from physical inspection or remote monitoring

Depending on the type of information available from the inspection or from the last work performed, the system may be able to identify the type of work to be performed. Once identified, the work can then scheduled and managed by the work management system until it is complete, and the results of the work are recorded as historical information against the asset, which in turn may trigger additional work in the future.

What is GIS?
Modern Geographic Information Systems used in utilities today are much more advanced than the computerized mapping systems of the past. They maintain a realistic model of a utility's distribution network on a geographic base. There are a few key concepts here that are critical to understanding how GIS and asset management can work together.

Today's GISs provide the capability to build connected models of a utility's distribution facilities. A facilities model is a computerized representation of the functional organization of the network. It describes individual facilities in terms of the function they perform, their engineering characteristics, and their relationship to other facilities. This facilities model is typically the nucleus of a large and diverse set of applications the utility can use to support its operations.

The most obvious of these applications is the retrieval and presentation of graphic representations of the model, either as graphic displays or as physical hardcopy plots. As the electronic dissemination of graphic data becomes more prevalent, the need to hardcopy plots is diminishing. Two key applications supported by a network facilities model are network analysis (load flow modeling) and outage management systems. These both rely on a connected network model that contains sufficient engineering attribution on the facilities to define the characteristics of the facility objects. In most cases, these applications are themselves separate systems or products that operate within their own environments and rely on the GIS to provide a "snapshot" of the electric or gas (or water) network at one point in time. Other applications can be developed directly on the GIS application to support network as well as spatial analyses. These include areas such as leak survey management, tree trimming, transformer load management, route calculation, and optimization. The common data requirements that the majority of these applications share is knowing what specific function is being performed at a point in the network and what the generic characteristics of this function point are, or what the engineering characteristics of the conductor are that connect to functional points on the network. What is generally not required is specific knowledge of the physical device that is performing that function such as manufacturer, age, serial number, warranty information, and so forth.

The GIS also has the inherent capability to place facility objects in a geographic space and identify, either precisely or aesthetically, their location. This implies that the GIS can then ascertain various spatial relationships between facility objects, as well as between facilities and other spatial constructs such as lines (roads), and polygons (administrative areas).

GIS systems that are used today to address utility operations have a sophisticated capability known as "long transaction management." While different vendors address this in different ways, the capability essentially allows a complex change (as a collection of a large number of individual changes) to the database to be defined over a long period of time, and then applied as a single transaction. Again, there are implementation differences among GIS vendors, but some version management capability related to the long transaction mechanism is provided. These capabilities provide critical support for the engineering design process, for as-built posting, and for engineering "what if" types of analyses.

It is important to note that traditional databases, relational, object-oriented, and others are short transaction databases. GIS network models reside in a long transaction, version-managed database. There is a challenge in integrating information managed within these two disparate database technologies.

Data partitioning
The major challenges in integrating an asset registry with a GIS-based network model are:

The database design to define the partitioning of data between long and short transaction data types
The analysis of data to determine the business process ownership to ensure that no conflicts exist between updates to the long and short transaction databases

To accomplish this, the concept of a facility will be separated from that of an asset. A facility is something owned and maintained in the spatial, long transaction database, and is part of an engineering model of the distribution (and/or transmission) assets of a utility. An asset is a record of a physical piece of equipment, whether or not it is performing a function at a specific location or at a specific time. This is illustrated by the following example of an electric distribution transformer as shown in Figure 1 "Bank/Device Model." These are typically three-phase devices that are installed at some point on the primary electric network to step down voltage to secondary level to provide service directly to individual customers. Based on the primary voltage, secondary voltage, and desired load capacity (as a function of the number and type of services), an engineer defines the engineering characteristics common to the installation as a whole and to the individual phases where these values may differ. The engineer has defined a bank facility and up to three device facilities. At the time of a design, the designer generally has no specific requirement as to manufacturer, nor has any idea as to asset-specific information such as serial number, date of purchase, or cost of purchase. The specific asset to be used to actually perform the function of the device slot is established at a later time. When construction and installation work is actually completed, as-built field reporting will provide that information. At this time, a relationship is constructed in the database between the physical item of equipment (the asset) and where it is performing its function (the phase of a particular bank).

Figure 1 - Bank/Device Model

The particular transformers actually used to perform the function must meet certain engineering characteristics for that class of transformers. When a design is developed, a transformer of a particular class is selected. A physical transformer to actually perform the function would have a matching set of characteristics. These are organized into specification tables that provide a look-up table of legal combinations of engineering characteristics of physical equipment items. When specific items of equipment (assets) are actually installed in the field, the physical asset's engineering characteristics should match those of the specific item (or change specific to specifications) defined for the particular facility.

It is similar for conductors, where within an electric segment is a facility with a particular function, with individual conductors for each phase, that reference particular standard engineering specifications. The Figure 1 example presents a situation where individual conductors are not treated as assets, and as such do not have records within an asset database.

The GIS provides capabilities to manage the connectivity of linear and point facilities through a configurable set of rules that defines how facilities connect to each other. Connectivity rules provide definition in terms of what facilities can or cannot connect to others, where they can connect, and what actions are to be performed when they connect. For example, a transformer connects to a segment at any point along its length without breaking the segment. However, a switch may only connect to the end of a segment, or it must break a segment if it is connected at any point other than its end point. Similarly, additional rules may be applied, such as conductors at the same voltage level are allowed to connect, but are not allowed to connect if they are at different voltage levels.

Figure 1 describes the relationship of an asset to a facility in terms of the asset being installed at a facility to perform a function. However, in reality, the relationship is more complex than that. Assets are installed and removed, and often installed at another facility location at some other time. Asset management applications typically require knowledge of the entire history of an asset, not just where the asset is currently installed.

Figure 2 - Asset Event History

A more realistic representation of this relationship is shown in Figure 2. This example uses a recloser - a somewhat more interesting device than a transformer. There is an event history record between the facility and the asset. This record documents all significant events that occur in the relationship between the two entities. Significant events include installation, removal, repair, inspection, and maintenance activities. Therefore, as a piece of equipment is installed in a location, removed, then re-installed at another location, the event history provides a record of all activities associated with an asset, as well as providing a record of all activities occurring at a facility location. An event-type specific record is maintained to provide details about particular activities. For example, if the event is an inspection record, the values of readings obtained, or a quantified assessment of equipment condition, are recorded. If it is a maintenance record, then details of the work performed would be recorded. By associating these historical records with both the facility and the asset, very robust analyses may be performed on this data. All maintenance work performed on a specific asset can be retrieved, regardless of where it may have been installed. Similarly, all of the maintenance work performed at a facility location can be retrieved, regardless of how many individual items of equipment have been installed at the location. Figure 3 provides some details as to typical attribute information stored in the database at each of the levels discussed above.

Figure 3 - Event History - Typical Attributes

This data organization provides a blueprint for partitioning data between a GIS and the asset database within a work management system, and identifies the types of data that are best suited for management within each environment. The asset database, residing in a traditional relational database management system, contains physical equipment-specific data, a history of discrete events that have occurred related to the assets, and information well-suited to maintenance within a short transaction paradigm. On the other hand, the GIS database manages all data related to the spatial characteristics of data including network topology. The GIS also manages data that needs to be managed using long transaction controls, including version management. This allows the GIS to model facilities that are not yet assets (i.e., designs), as well as facilities that may never be assets (i.e., planning studies and "what if" scenarios).

Asset Analysis
One of the key aspects of using a database of asset information is for the actual management of those assets, i.e., performing analysis on the information in the database to help make decisions about what work needs to be done on what assets. Traditionally, this analysis will take the form of queries returning assets of particular types that have attributes that fall within or outside of specific criteria. These criteria may be global in nature (i.e., installation date), asset-type specific (manufacturer), or asset-item specific (number of operations that triggers maintenance work). These types of queries can locate and group collections of assets for which some decision must be made or some activity needs to be identified and scheduled. The end result of these types of analysis is to create groups of assets against which a set of activities can be defined that will then be applied to each member of the group through one or more work orders.

With a relationship between a GIS and the asset database, the asset analysis processes can be significantly enhanced. Current GIS systems have standard capabilities to interface to external RDBMSs. Through this mechanism, data contained in an asset database can be made available to a GIS in one of at least two ways. First, distributed database technology along with gateways and views can provide direct access to the operational asset database. Alternatively, a snapshot of the asset database into a GIS-accessible data warehouse can be made. All of the asset and asset event history data can be made available to the GIS. In its simplest form, the interface can provide a GIS user the asset attributes for a selected facility be retrieved and displayed. More robust capabilities include the ability to construct queries that combine the spatial query capabilities of the GIS with traditional attribute criteria. A user would be able to highlight on a display or plot the facility locations of assets that meet specific criteria. This allows the user attempting to create groups for work purposes to spatially analyze the groups for spatial proximity, locate outliers, and identify additional equipment that is spatially within a group that can be added for efficiency of executing the work.

Having the spatial component of facilities readily available also enhances the data maintenance process of the asset database. Typically, asset records maintain some amount of attribute information that is spatial in nature. Examples of this may be a municipality code, an administrative area, or an environmental zone. As these boundaries are modified or completely reorganized, updating the asset records with the new spatial attribute values can be an extremely difficult and lengthy process, given the limitations of a tabular forms-oriented data entry system. With the GIS, the equipment lying within the new polygon boundaries can be quickly identified, and simple SQL scripts can be developed to bulk update this attribute information.

Another spatial application that can be applied to the management of assets is the identification of the spatial location of assets requiring work for the purpose of developing optimized scheduling and routing of the work to be performed.

Similar to the spatial component enhancing the asset database, potentially even greater benefits can be derived from the application of network connectivity information onto the assets. The performance of equipment on the network is often affected by its electrical environment including what devices are upstream or downstream, as well as what the loading characteristics of the network are at the point of attachment of the asset. Information about the operating environment of the asset provides additional information regarding factors that affect its performance. Enhanced analysis leads to applications such as protective device coordination and transformer load management. This type of analysis is generally performed by external applications that perform load flow analyses based on a connected network model. Based on the raw data and network connectivity model, these analyses can predict conditions under which devices on the network operate. This information can then be related to the physical assets, and provide additional condition factors that can then be used to adjust the maintenance and inspection schedules for the physical assets.

A logical extension of the above discussion is the area of reliability centered maintenance (RCM). Time-based maintenance assumes that all failures within the network have essentially similar consequences. It is known that a significant number of failures are essentially random and unpredictable. As such, maintenance activities need to concentrate on points within the network where they would have the greatest impact. RCM is based on an analysis of the impact of failures on the overall network in terms of customers affected, lost revenue, criticality of customers affected, and similar criteria. By having a connected network model, with associated customer identification superimposed on this network, RCM coordinators can now assess the criticality of assets and adjust maintenance and inspection schedules accordingly. The knowledge of where on the network assets are located, along with an assessment of their criticality, is essential for an effective RCM utility strategy.

Conclusions
In summary, a carefully designed and portioned asset database coupled with the effective use of a GIS can significantly enhance the effectiveness of asset management systems in many areas including:

Maintaining spatial attributes and groupings
Analyzing routing for inspection and maintenance work
Applying network environmental factors to assets
Assessing the impact of equipment failures on distribution networks
Forming a basis for a more effective reliability centered maintenance program
Tracking full asset lifecycle performance and history

Traditionally, asset information has been localized within single monolithic databases. The introduction of GIS technology to support asset management can significantly extend the reach of maintenance history information to apply more effective analysis tools to better support the many aspects of asset management.