Global optimization: Gas mainline replacement planning

Global optimization: Gas mainline replacement planning

Mark Thornton
National Fuel Gas Distribution Corporation
10 Lafayette Square
Buffalo, NY 14203

Ray Boy
National Fuel Distribution Corporation
10 Lafayette Square
Buffalo, NY 14203

Abstract
An on-going challenge faced by National Fuel Gas Distribution Corporation (NF) is to identify gas mainline most in need of replacement, given the corporate goal of maintaining a safe and reliable system within budgetary considerations. The process of assisting Engineering to evaluate many possible mainline replacement project scenarios throughout NF’s service territory was dubbed “Global Optimization.” NF’s challenge was to utilize existing AM/FM/GIS data and develop a low cost analysis tool using off-the-shelf products requiring minimal customization. NF met this challenge by employing an available program that would utilize existing graphic and tabular leak data from different sources. This data was then spatially analyzed to highlight ranked clusters of leak and water interruptions. These leak clusters were then grouped into “super-clusters” to help identify areas for NF’s systematic replacement program. To date, NF has plotted various leak points throughout its New York and Pennsylvania service territories. This tool is fully functional and has been used to facilitate project planning for calendar year 2000. The project has been deemed successful by its user community in that what was envisioned was in fact built, that the total cost of development was minimal, and that the time to bring up the system was reasonable.

National fuel
National Fuel Gas Company (NYSE: NFG), incorporated in 1902, is a diversified energy company with its headquarters in Buffalo, New York. The Company’s assets are distributed among six business segments: Exploration and Production, Pipeline and Storage, Utility, International, Energy Marketing and Timber. National Fuel Gas Distribution Corporation sells or transports natural gas to over 733,000 customers through a local distribution system located in western New York and northwestern Pennsylvania. National Fuel Gas Supply Corporation provides interstate natural gas transmission and storage for affiliated and nonaffiliated companies through an integrated gas pipeline system that extends 3,065 miles from southwestern Pennsylvania to the New York-Canadian border at the Niagara River. It also owns and operates 29 underground natural gas storage areas and is co-owner and operator of four others. NFG Supply Corporation and NFG Distribution Corporation are involved in the AM/FM/GIS project.

AM/FM/GIS project background
In 1992, National Fuel Gas Distribution Corporation (NF) set out to build an AM/FM/GIS system that would automate our key mapping records and provide a base for future enhancements. An important design criterion was and continues to be cost. The unique solution adopted by our team is a hybrid raster/vector mapping system. NF scanned, warped, and edge-matched nearly 5000 maps at a scale of 1”=200’ in New York and 2000 maps in Pennsylvania. A two-year backlog of as-built gas mainline updates was drafted electronically on to the raster base maps. The vector gas mainline symbols carry attribution such as project number, size, test pressure, and other important information. NF completed map conversion of the New York and Pennsylvania distribution maps this past year. Once completed the team began to look for opportunities to enhance the basic system with applications that could provide quick payback. Global Optimization is one such application that is built on the existing AM/FM/GIS system.

The system
NF’s AM/FM/GIS system is built on Intergraph’s FRAMME running on Windows NT 4.0. The system was moderately customized with a graphical front end to our work management and cost tracking system. The FRAMME application was the most costly part of application development since it required custom programming. The resulting customized application handles our mapping and workflow needs quite well. View only users access the system using Intergraph’s Field View program. For map view users, we chose to use the out-of-the-box Field View functionality and keep the cost of development very low. These two applications formed the base of the system that is used for updating, printing and viewing our operating maps.

In 1999 we added Intergraph’s Geomedia program for spatial analysis of various gas and non-gas map features. Our team also intended to deploy this tool with minimal customization and the correspondingly low cost. GeoMedia was chosen because it offered built-in spatial query capability that would have required custom programming to develop in FRAMME.

Other software that NF integrated into the system is from the Microsoft Office suite of applications. The previous FRAMME version ran on UNIX, which had limited third party software support. With the migration to Windows NT, we can use MS Access and MS Excel as part of the analysis process in optimizing and planning gas mainline replacement. These are very low cost tools with powerful analysis capability that can be used as they are delivered. Only some simple macros and Visual Basic programs are needed to perform the Global Optimization process.

The ‘tool box’
What NF was looking to do was to combine a number of existing and new methods of pipeline project analysis and come up with a set of tools from these to optimize mainline replacement planning. These tools were developed separately in time and then combined to improve the evaluation process. These are PREP, Systematic Replacement, Leaks, and Global Optimization, from the mainframe and AM/FM/GIS systems. We wanted to use existing data and programs, albeit from different platforms, to efficiently evaluate opportunities and focus capital replacement spending. At the same time we did not want to engage in writing any expensive custom programs and interface sets. Global Optimization would combine these existing programs and data into a ‘tool box’ for a very reasonable cost.

Pipeline replacement evaluation program
A challenge faced by NF is to scope and prioritize capital expenditure allocations to the gas mainline most in need of repair while continuing to maintain a safe and reliable system. The Pipeline Replacement Evaluation Program (PREP) was developed in 1994 to assist in evaluating which potential projects should get first priority. The PREP system was implemented on the IBM mainframe.

PREP input screens collect leak information, replacement cost, and O&M cost for a given section of gas mainline over 250 feet in length. Existing leaks are input and the system calculates projected new leaks over a five-year period. The net present O&M cost for patrolling and repairing leaks is calculated. (NFGDC, 1995)

Next the net present value for capital replacement for five years is calculated. The system then calculates a ratio of O&M expense and capital requirements (Figure 1). Ideally, the ratio exceeds one, meaning the cost to replace is less than the cost of maintaining the main over a five-year period. The ratio may be less than one, but the PREP score provides a means to rank all proposed replacement work against the others when other factors apply.

Figure 1
PREP by itself is a useful tool to focus capital spending on projects with the highest value. But, PREP is useful only after an engineer has identified all possible project areas. This previously was a time consuming process that involved manual data collection and subjective evaluation. However, PREP serves as a tool in the ‘tool box’ of Global Optimization. PREP does not provide any spatial capability, but it does a good job in economic analysis.

Systematic replacement
Systematic replacement was conceived from the recognition that it costs less per foot to replace larger scale footages of gas main than it does to replace a series of shorter sections over time. Economy of scale applies since the fixed costs such as moving equipment, crews, and material to a job site are about the same for both large and small jobs. Systematic Replacement is not a computer application; rather it is a design philosophy.

Using this basic concept, the object is to chain together a number of shorter projects to build larger scale jobs with the resulting economies of scale. Once an area is determined to have sufficient sections with a sufficient PREP score, the object is to incorporate these in an overall large replacement project. Also, opportunities exist to convert low pressure areas to medium pressure resulting in projects with less total mainline of smaller diameter. The philosophy of Systematic Replacement is used in conjunction with PREP, leak analysis, and Global Optimization.

Leaks
The next step in the progression of better analysis capability was leak plotting. The Leak Reporting System (LRS) is an existing comprehensive text based system that tracks the life cycle of leaks. This mainframe based application and data was only available to the Engineering staff in report format. However, everything that we needed for spatial analysis except coordinates was in this database. The leak enhancement phase of Global Optimization required building an interface to the mainframe data and adding a leak symbol and attribution to our AM/FM/GIS.

The interface was customized in FRAMME by Intergraph. Using a mainframe application to dump out LRS transactions and an automated file transfer utility, leak data is loaded into our Oracle leak table.

The second part required was adding new mapping functionality to place leak symbols and update the attribution. This functionality automated placing the leak symbols since the source data does not have mapping coordinates, but does have address and measurement information. A labor-intensive three-month effort was required to place known leaks in New York and Pennsylvania. A batch function updated the attributes (Figure 2) that are interfaced to the LRS system giving the leak symbol the intelligence needed for spatial analysis.

Once the data was local to FRAMME and the leak points were plotted, a secondary application updates the leak symbol with the attributes loaded from the mainframe. This changes the symbology for each leak type and the status.

Only having the leaks plotted proved to be a significant benefit for replacement planning. Staff engineers were able to scan plots of their service territories and quickly identify potential projects. In previous years a laborious process of hand plotting was required. The maps produced were only good for one season and were quickly outdated.

Figure 2
Remembering that most of our facility data is non-intelligent raster, the leaks needed to stand-alone as far as analysis is concerned. There are no links to the vector main, so text-based queries of leaks (correlated to mains) are not possible. So, our next challenge was to develop the final ‘tool’ needed to pull together all of the other ‘tools’ to enhance selection of potential replacement project areas.

Global optimization
Global Optimization is the term coined by the AM/FM/GIS team to represent the combination of new and existing tools used to optimize replacement planning over our entire New York and Pennsylvania distribution service area. Global Optimization is the combination of PREP, leaks, and systematic replacement.

Global Optimization is a combination of software components, namely Intergraph GeoMedia and Microsoft Access. A template is setup in GeoMedia to extract leak data and coordinates. (Azzam, 1999) GeoMedia spatial queries cluster together leaks (Figure 3). We do not import all the raster and vector graphics into GeoMedia because of performance issues. The leak clusters are not overlaid on facility maps. Figure 3 illustrates this graphically. The leak and cluster database information is exported into Access for further analysis. Once in Access, the leak clusters are ranked based on leak types and quantities. A sorted list is then produced identifying clusters with a high probability of becoming a replacement project of priority.

Staff engineers then review each of the clusters, apply PREP and develop projects. The engineer uses Field View to locate and view the resulting clusters.

Figure 3
The benefits of this automated analysis are substantial. Rather than spending large amounts of time gathering data from printed reports and attempting to manually isolate project scenarios, the engineers can focus on evaluating areas already ranked by the system.

The automated process reduces the subjective evaluations and focuses on areas that should be given attention. Global Optimization was used to evaluate projects for 2000. This process will be improved and extended for analyzing projects in 2001.

Active corrosion and beyond
Once we had experience with the tools used for Global Optimization, many other possibilities have become apparent. Using the same tools, we are in the process of evaluating our active corrosion program.

Active corrosion is a mandated program that we evaluate every year. Any gas mainline with five leaks within 500 feet during the past three years requires special maintenance consideration. This is a time consuming review process. Previously, this was also a manual process using leak reports.

We are using the same leak data and spatial query capability as used in Global Optimization to further evaluate leaks. The queries and macros are customized to search for leaks in the density and proximity as defined by active corrosion. These were developed in-house with relatively simple Visual Basic code in Microsoft Excel.

As we continue to move forward, we will look to link other databases to our AM/FM/GIS system for analysis. These include valves regulators, drips, corrosion features, and others. Each of these will hopefully provide new opportunities to replace slower manual analysis or to dramatically reduce the labor to do things that were not practical before.

Conclusion
Many other opportunities exist to enhance the existing system. These can all be accomplished for minimal cost if we continue to make use of existing data and low-cost, off-the- shelf, software tools.

Our new methods of optimizing capital gas mainline replacement projects have been successful. By using our existing AM/FM/GIS and adding low cost tools, we are able to minimize the manual effort that is expended each construction planning season. The most powerful tool we have used to date is GeoMedia’s spatial query functions.

When we first looked at enhancement opportunities to our base AM/FM/GIS system, most of these opportunities were very expensive to develop with custom programming. But, with help from our support partners and some creative team thinking, we were able to implement powerful analysis capability with off-the-shelf software.

Finally, since the beginning, our AM/FM/GIS project the team was focused on converting maps and updating them once completed. Now that most of this work is done, we are able to look at our system in ways we never imagined ten or so years ago by using spatial tools. Linking text based mainframe data to our system may provide many more opportunities.

We have been able to achieve our objectives within our budget by looking beyond our initial project and building upon this with existing tools. Many exciting opportunities are yet undiscovered.

References
Azzam, R., 1999, Global Optimization Project Instructions Document for GeoMedia Use

NFGDC, 1995, Pipeline Replacement Evaluation Program Training Manual