Leveraging GIS Beyond Mapping
John A. Wakefield Section Lead, Information Technology Systems, Field Operation Systems Washington Gas Light Compan Andy Benedict President Vantage Management Solutions, Inc. Introduction Over 25 years ago, Washington Gas had a vision of an integrated enterprise information system that was capable of connecting our physical assets, our customers, and our employees in a new way. At its center, we envisioned a graphical system where users could manage the customer relationship, portray our assets accurately and manage them efficiently, and plan and execute work – truly bringing all aspects of the enterprise together. Any Star Trek fan will recognize that back then, this vision was akin to using worm holes to connect multiple parallel universes. It wasn’t called “GIS” back then, but then, the systems available at that time were capable of doing little more than “Automated Mapping.” Nevertheless, our vision was persistent, and we always managed to evolve with and beyond the existing technologies. This paper describes how Washington Gas has achieved this vision and how we have leveraged GIS well beyond mapping. A brief history As of the mid-1970’s, Washington Gas maintained over 900,000 manual records, coordinate maps, valve sheets, service record cards, regulator station, and construction drawings. The process of updating and handling all of these manual records took a force of more than 50 draftspersons and file clerks. While these visual records were adequate for facility description and location, the process of compiling any type of information for analytical, statistical or regulatory reporting purposes took great effort and cost. Studies conducted at that time identified that significant benefits would be gained by automating the mapping process, and providing better means for extracting critical strategic management information. With the full blessing of upper management, WG undertook the design and construction of an state-of-the-art facilities mapping system, aptly named “FMS”. The task of building a “first generation” GIS is surprisingly similar to what one would need to do today, with the exception that none of it came out of a shrink-wrapped box. These included:
The state of the art at that time consisted of mini-computer graphics systems with dedicated work stations. The primary data itself resided on a main-frame, with a tape-drive transfer mechanism to handle active data updates. We selected Calma Corporation’s graphic hardware system, due to its data model flexibility, and ability to interface with our IBM host computer. The graphics system had a whopping 25 megabytes of storage capability. Incidentally, Calma was best known for its graphics design capabilities in the manufacturing of electronic circuit boards. Data Model Development Our data model design consisted of three primary partitions – our land base, cultural data, and our main and services piping system. The land base included building outlines, streets, sidewalks, driveways and water features. The cultural data augmented the land base with street names, addresses, and landmarks. Our piping system included distribution and transmission mains, foreign piping, valves, regulators, and services. Land Base Development Since a high level of accuracy was desired, we chose high resolution aerial photography, which needed to be conducted during perfect weather with no foliage. Photographing our 1000 square mile service area took more than four years. In those days, digital storage was more expensive than labor, so digitizing the photos (rather than using a raster land-base layer) was the only way to go. The digitizing of the land-base was completed in mid-1982. During that process, we decided to scrap the idea of including sidewalks and driveways. Pipe and Service Asset Development Also by 1982, we began converting our gas mains, valves, and regulators from paper maps and drawings into the main and services portion of our FMS. By 1983, we started converting our service record cards. As many utility companies discovered at this time, the conversion process was tedious, expensive, and for some, overwhelming. But rather than abandon our vision and investment, we upgraded to a newer and much faster generation of graphics hardware workstations, and contracted out the remainder of the pipe and service asset conversion. Over the next five years, we completed the conversions, and made several improvements to the system configuration and hardware, evolving into our “second generation” GIS. Early spatially spatially intelegent solutions Ten years after we started, we were finally publishing and distributing copies of electronic maps to our engineering, operations, and field personnel, realizing the initial part of our vision. But the part about connecting systems and users to better manage the assets and customer relationship still remained. Network Analysis. One of the first “beyond mapping” solutions we developed was a direct interface between our FMS and our Stoner Network Analysis model. By 1988, we were adding over 10,000 customers per year to our system. This interface enabled us to extract portions of our as-built piping network to the model, and better examine our growth alternatives. Graphic Inquiry. In 1988, we expanded FMS access to users via a new graphic interface that no longer required dedicated workstations – they now had direct desk-top PC access. We also provided several graphic inquiry tools that enabled users to access facility information easily. For example, if a customer called reporting a leak, a customer service representative could call-up the location within the FMS, and direct leak response personnel. A Critical Internal Study By 1990, a number of factors that drove our decade long development of FMS had changed. Deregulation, competition, and perhaps the most critical of all, upper management. A new management team had called for a re-examination of all our FMS. This study not only revalidated our vision, but garnered additional support from our upper management team. The study identified several additional “beyond mapping” solutions that needed to be developed at an accelerated pace. On-Line Service Updates. Due to the volume of new services being added, we enabled a “decentralized” approach to performing and validating these updates. Prior to this time, all facilities modifications were performed on the graphics workstations. Now, they could be performed on PC’s. Safety Sector Analysis. Our piping system model provided a number of benefits. By performing traces of our model, we were able to define and locate those valves that could safely isolate sections of our system in the event of an emergency. The Safety Sector analysis became part of our overall emergency response plan. Critical Valve Inspection. The isolation valves described above are in effect, “critical valves.” We needed a way to ensure their operability through periodic inspection and maintenance. We developed an application that scheduled inspections, and tracked inspection data as well as the status of any remedial work. Operations Reporting. We created several specialized reports that enabled our Operations personnel to query our FMS in different ways. For example, service replacement budgets could be developed based on the number of services of a particular vintage within a particular area; main pipe surveys could be conducted based on pipe footage; and DOT reports could be developed; among others. Market Enhancement. During the mid-1970’s, a gas moratorium created a situation where thousands of homes and businesses were built in our service area using heat-pumps. During the 1980’s we built our gas mains beyond these areas to serve continuing growth. But by the 1990’s the heat-pumps began wearing out. We were able to use FMS to identify potential customers along our existing gas mains, and develop an active marketing enhancement program. The Next Generation of GIS By 1996, we had a substantial user base and support was becoming a major task. We began taking a good hard look at what we could continue to do with the existing FMS, versus the third generation GIS’ coming out on the market at that time. We also had a weary eye toward Y2K. Our studies show that we would not be able to leverage some of the new technologies becoming available at that time, and that we would be severely impacted by millennium date changes. It simply would not be prudent to continue investment in what was a home-grown, outdated technology. We embarked on the search for a new GIS with the following criteria in mind:
We also made a number of enhancements to the applications described above, which were not possible in our second generation GIS. For example, in support of our Safety Sector application we were able to trace-out the piping network downstream of the regulator and display the maximum operating pressures (MOP’s) within a Safety Sector. Through a link to our CIS Safety Sector analysis is now able to identify specific customers who would be affected by valve closures, enabling safer and expedient outage management. Our Market Enhancement application is now able to identify the true proximity to main pipes, resulting in a more accurate solicitation of potential customers. Recent Asset Management Focus Washington Gas has more than 3 million feet of Cast Iron, 1.3 million feet of Bare Steel, 25 million feet of cathodically protected steel, and 24 million of plastic pipe within its distribution system. We have traditionally spent $45-50 million per year in O&M and Capital programs to maintain or replace this pipe. In the late 1990’s, organizational changes, retirements of key personnel responsible for defining those programs, and Y2K issues caused our System Replacement group to re-evaluate its asset management approach and to investigate externally available pipe asset management systems that could support its new approach. Our asset management business case identified significant savings associated with program budget optimization, as well as reductions in the engineering effort needed to analyze our more than 10,000 miles of distribution pipe. Here we need to digress a moment to answer the question, “Just what is asset management?” We found that to different types of solution providers, it means different things. To a financial systems provider, it’s what keeps track of the pipe in the ground, and provides asset valuation and depreciation information for regulatory and financial reporting. To a work management solution provider, asset management is the ability to manage individual maintenance programs and replacement projects and ensure you are executing them properly. But none of these systems address a more strategic view of what are the most cost effective programs, what is the optimal mix of capital and O&M dollars, and how do we better manage the inherent risk of transporting natural gas to our customers. Vantage Management Solutions, Inc. a consulting and product solutions firm out of Newtown, PA had an approach that matched our philosophy as well as a pipe asset management solution called Optimain. However, this product had never been integrated within the GIS environment. Washington Gas embarked on a joint project with IT, System Replacement, and Vantage to develop our Strategic Asset Management approach, and to assist Vantage in integrating Optimain with our GIS. The project consisted of a number of business process activities and the following GIS related elements:
A key element of pipe asset management is to fully determine the relative risk, current condition, and stability of our existing pipe, and to be able to predict its potential for deterioration over time. While our GIS provided an excellent means for determining such things as the population density, building usage and even average service lengths around specific pipe segments, it had no means to evaluate existing pipe condition. We determined that it was necessary to associate our maintenance history, including leak and inspection records, with the pipe. This decision was supported by the fact that DOT 192 requires gas utilities to maintain pipe leak and inspection records for as long as we maintain the pipe – In essence, similar to a valve or regulator, a leak clamp is a permanent facility in our system. Since our maintenance history is maintained on a mainframe system, we needed to find a way to import leak and maintenance records into our GIS, and geocode these records to the appropriate pipe segments. The project team initially downloaded and analyzed over 200,000 records dating back to 1987. The team filtered and reduced this list to just over 70,000 leak and inspection records that were pertinent to determining pipe condition. The team then needed to build objects within our GIS that were capable of receiving the records and associate them to our pipe facilities. The team created the Pipe Maintenance Tracking System (PMTS), which has the flexibility of tracking more than just leaks. It can link to virtually any external database and associate records to any object within our GIS. The task of geocoding these records was complicated by the fact that the address and street intersection information did not necessarily match the GIS. After several iterative passes of cleansing address information, comparing attributes between maintenance and pipe records, and transforming locations, the team was able to successfully geocode better than 95 percent of the records. The team built an application that automatically updates the GIS with completed maintenance work, and renders the work locations within the GIS. There are currently over 100,000 PMTS work location records within the GIS. Users can select a particular PMTS, a pipe or draw a polygon to see the entire maintenance history of any facility within our system, all within a single PMTS “History Browser” form. Pending Work Order Tracking In early 1999, we identified that our mainframe system used to track pending distribution work (e.g. planned leak repairs and re-surveys) was impacted by Y2K issues. The Pipe Asset Management team was assigned the task of developing a new GIS-based Pending Work Order (PWO) management system. The team built a system that is capable of user entry of pending work data, as well as automated creation via referrals from completed work records. Using the PMTS, all pending work locations are rendered within our GIS. The PWO system is capable of both time and geographical work assignment and management. For example, our distribution personnel are required to repair or re-survey leaks based on regulatory commitments. The PWO system generates reports identifying work to be completed prior to the commitment date. Supervisors are also able to perform geographic queries that enable them to identify other pending work in the same area, and can select and assign those jobs as well, reducing “windshield” time. Strategic Pipe Asset and Risk Management (Optimain) As stated previously, one of the keys to pipe asset management is to evaluate the current condition and stability of our pipe assets, and to be able to predict the potential for deterioration over time. In early 2000, Vantage successfully integrated Optimain to our GIS, acquiring all pertinent PMTS, main and service pipe, street, building, and location specific information. Optimain uses a number of statistical, financial and risk evaluation algorithms, which were created with the input of our engineering, field, finance and regulatory personnel. These algorithms enable a detailed evaluation of a significant amount of data to be performed within seconds. Our System Replacement personnel use Optimain to evaluate our pipes’ current conditions and risk, and to identify pipes with the potential for deterioration. Optimain expresses its results in terms of the economics of maintaining versus replacing the pipe, and the relative risk of one pipe segment over another. Within the GIS, users can build a “candidate project” evaluation by simply clicking on a pipe, or drawing a polygon, then activating the Optimain application. Users can also create pipe segment projects automatically. Users can view all projects individually, or as budgeted program groups. Optimain supports our strategic pipe asset and risk management program by identifying improved programs for maintenance and replacement, and defining optimized O&M and Capital budgets: When transporting natural gas, one can never reduce the risk to zero, but we can help ensure that we achieve the greatest reduction in risk and future O&M dollars for each replacement dollar spent. Jurisdictional Paving Management (JPM) In early 2001, our Nation’s capitol made headlines for an unusual reason: its streets had fallen into an abysmal state of repair, and the DC government needed to take drastic action. One of the steps was to impose strict restrictions regarding the planning and timing of street work as well as requirements for street restoration when work was completed. Now, a simple leak repair requires us to perform a minimum one-lane-wide by 20-foot long concrete sub-surface with asphalt overlay. If we break into a street within five years of its resurfacing, we must repave the entire street. While we have always made a special effort to coordinate our street work with our jurisdictions, we decided to take a proactive approach. By monitoring the various jurisdictions’ street improvement project schedules within our GIS, we could not only better coordinate work, but would save a significant amount of dollars in street resurfacing costs. A number of jurisdictions provide us their paving schedules in a spreadsheet. JPM was designed to accept this as well as user input, and automatically update street attributes within the GIS – a paving project has a distinctive visual style. We also integrated the JPM with PMTS, PWO and Optimain. Maintenance personnel are notified to accelerate leak or cathodic protection work if necessary, and System Replacement personnel can immediately analyze the economic benefits of replacement ahead of paving, resulting in a 20 to 35 percent project savings. Marketing also uses the paving schedules to focus its efforts on potential non-gas customers along existing mains. Ongoing/Future Efforts We have a number of ongoing projects that continue to evolve our vision and take us well “beyond mapping.” These include additional spatial solutions, interfaces to other data sources and applications, and ways to extend the information to more and more users. For example: Cathodic Protection Area (CPA) Enhancement. We are currently updating our manual CPA process, and will begin storing CPA’s in our GIS. Our entire solution will enable better management of CPA surveys, identification of areas that are out of protection, and simplify compliance reporting. Enhanced User Access. We are continuing to roll-out lap-top based GIS information to all service and distribution personnel, as well as providing internet access to the GIS. Facility Survey Unification. We now perform a number of different facility and premise surveys. This will put them under a single umbrella. Our GIS will be the primary user interface, with links to our Customer Information System (CIS), and compliance status databases. Enhanced CIS Links. By tying our GIS to CIS, we will enable a direct link to our large load, interruptible, and transmission customers. Our GIS users will have limited access to CIS customer records and be able to contact customers who may be impacted by operations or replacement projects. Conclusion We have truly “leveraged our GIS beyond mapping,” we recognize that our effort is a constant evolution. We are still trying to “bring all aspects of the enterprise together.” But our persistent vision of connecting users, assets and customers is closer to full achievement than ever. | ||
|
|