February 2009 Vol. 236 No. 2

Features

Process Devised For Monitoring Leak Threats Using GIS

This article shares experiences we have gained working with several customers to geo-code leak and pipe inspection reports and associate them with the pipe segment in the GIS that best matches the address location and other attributes of the report.

It also discusses GIS data-modeling considerations to facilitate data capture and presentation on the map. Finally, we suggest several on-going business process improvements to ensure threats are captured with sufficient quality and timeliness to support the DIMP regulations.

Over the past several years, Pipeline Integrity Management regulations have been implemented for pipelines transporting hazardous liquids (Title 49 CFR Part 192) and natural gas (Title 49 CFR –Subpart O). In June 2004, the DOT Inspector General, in testimony before Congress, recommended that the Office of Public Safety within DOT’s Pipeline and Hazardous Materials Safety Administration (PHMSA/OPS) require operators of natural gas distribution pipelines to implement an enhanced safety program similar to those used on hazardous liquids and natural gas transmission. As a result, a multi-phase action plan was initiated to proactively gather industry and stakeholder input about any potential safety programs.

The first phase of the action plan was completed in 2005 by four multi-stakeholder work/study groups, and resulted in a report on distribution integrity for Gas Distribution Pipelines (PHMSA, December 2005). The report concludes that it would be appropriate to modify existing pipeline safety regulations to convey the concept of a risk-based distribution integrity management process (PHMSA, December 2005). It recommends a “high-level, flexible federal regulation, in conjunction with implementation guidance, a nationwide education program, and continuing research and development.”

The report concludes that significant differences between transmission and distribution pipeline systems, as well as diversity among gas distribution pipeline operators, make it impractical to establish prescriptive requirements. Rather, operators may be required to maintain their own tailored integrity management programs that address seven key elements:

1. Establish a Written Integrity Management Plan.
2. Pipe Infrastructure Knowledge.
3. Threat Identification.
4. Risk Assessment and Prioritization.
5. Risk Mitigation.
6. Performance Measurement and Adjustment.
7. Regulatory Reporting.

The second phase began in January 2006 and involved development of draft regulations by PHMSA in parallel with preparation of guidance materials and standards by industry and government organizations, including the Gas Piping Technology Committee (GPTC).

As part of the Pipeline Inspection, Protection, Enforcement, and Safety (PIPES) Act signed in December 2006, congress mandated a federal rule on distribution integrity management by the end of 2007. In a departure from the usual approach aimed at maintaining closer alignment between various stakeholders and the final regulations, the GPTC guidance materials were drafted prior to publishing the final regulatory language, and were essentially ready for initial release by fall 2007.

After some delays for cost-benefit analysis and other processes between PHMSA, the Office of Management & Budget, and Congress, the NPRM on Distribution Integrity Management was published in the Federal Register on June 25, 2008. Operators will have 18 months to write and implement their DIMP plans after the final rule is enacted (expected sometime in 2009).

Section IV.D (Findings Relative to Leak Management) of the NPRM specifically highlights the importance of an effective leak-management system as part of an overall distribution integrity management program. According to input gathered from stakeholder groups, the essential elements of an effective leak-management program (with a clever embedded acronym) are:

• Locate the leak.
• Evaluate its severity.
• Act appropriately to mitigate the leak.
• Keep records.
• Self-assess to determine appropriate additional actions.

This article focuses especially on the first element – “locating your leaks,” particularly in light of overall DIMP requirements to “understand your infrastructure and monitor threats.”

The proposed DIMP regulations specifically require distribution operators to establish an effective leak-management program (generally as described above). Among other things, the draft GPTC Guidance Materials suggests the following activities relative to leak repair and pipe-inspection reports:

• Information about an existing system should be updated when new or better information becomes available, including during existing operating or maintenance activities. (3.1(e)).
• When an operator inspects the pipe whenever it is exposed, the operator should use the occasion to record and evaluate any distribution system unknowns that are available at that location (3.1(f)).
• Identify threats based on location, such as leak clusters, damage clusters, cathodic protection history, soil resistivity, or localized materials, operating conditions, or environment (4.2(a), 4.3).

Leak repairs represent the most frequent opportunity to examine and collect information about distribution pipes and the surrounding environment. The draft regulations also require operators to identify additional information that will be collected in the future about the condition of pipes and the surrounding environment to support risk assessment. Integrating information collected during leak repair activities with existing infrastructure knowledge can add significant value and fidelity to on-going risk assessments. For example, by collecting information about the environment surrounding a particular leak, as well as the exact geographic location of the leak, it is possible to incorporate that environmental information into risk assessments on other nearby facilities.

Based on a reading of the draft DIMP rule and associated guidance materials, and experience in the industry, the authors believe that capturing the geographic location of leak repairs and pipe-inspection reports, and associating those reports with the specific pipe segment on which the leak or inspection occurred, can significantly improve the overall effectiveness of a distribution integrity management program. While not specifically mandated by the proposed rules, the authors believe that investing in a process and systems to accomplish this objective will help operators illustrate to regulatory authorities that they have implemented an effective leak-management system (as part of their overall DIMP plan).

Thus, when reference is made in this article to “Locating Your Leaks,” the authors mean knowing both the geographic location of the leak (relative to other facilities in the system) as well as the specific facility record on which the leak occurred.

UGI Infrastructure

UGI Utilities, Inc. distributes natural gas to customers in eastern and northeastern Pennsylvania. Until recently, we have operated a little more than 5,000 miles of main and served over 300,000 customers. Having just purchased PG Energy (now called Penn Natural Gas), we have increased the size of our business by about 50% so that we are now serving almost 500,000 customers and operate almost 8,000 miles of main. In the combined company, 76% of the services are plastic and 13% of them are unprotected metallic services. The remaining metallic services (11%) are cathodically protected in some way.

In March 2008 UGI signed an agreement to purchase PPL Gas Utilities Corp. Pending PUC approval, the new company will be called UGI Central Penn Gas. This acquisition will increase UGI customer count by 16% to almost 550,000 and increase its miles of main by another 49% to over 11,500 miles. The distribution of service material type in the combined company will be about the same, percentagewise.

DIMP Plans

UGI has completed extensive re-engineering of the processes and tools used to collect and maintain data about leaks and repairs on its distribution system. Its previous process was:

• Leaks were found or re-inspected by roving leak surveyors or those called out on odor calls. Details were captured on a paper leak form.
• Repairs were documented on separate paper form.
• Non-leak excavations did not fit this process well.

For each repaired leak, the following basic information was collected:

• Field assigned leak identification.
• Main identification determined by office staff.
• Political subdivision.
• Street address.
• Survey method.
• Leak “location” (in street or behind curb).
• Classification (leak grade).
• Sketch with readings.
• Repair action.

The paper forms were keyed into a mainframe system. The original paper was filed by map grid in each leak surveyor’s office. There are multiple leak surveyor offices throughout the operation area. Some of the shortcomings of this process were that leak locations were not apparent to field personnel; surveyors had to manage paper docs; and old documents were archived in a “vault,” that is, they were not readily accessible.

New Process

UGI’s new process, started in 2006, was one portion of a multi-year asset management project named “FLAME” (Field Level Asset Management Environment). In the new process, leak, leak repair, and pipe inspection data is captured electronically, in the field, on new mobile GIS platforms. Initial leak entry is still completed by roving leak surveyors or those called out on odor calls. Repair crews then enter repair details and pipe inspection information on an electronic repair form, accessed from a button on the Leak form. The system also supports non-leak inspections and exposed pipe reports.

In addition to the data that was being collected, crews now capture the following additional information:
Political subdivision (one click based on spatial location).
Leaking facility, including valves or services (not just mains). The suspected facility is selected by field user for open leaks and the actual facility is confirmed by the repair crew.
Geometry created to display leak/repair/inspection to all GIS users.
Leak-reading values.
Sketch captured as mini CAD drawing.

The new system stores all leak and repair data in our enterprise RDBMS (Oracle). Records are linked to the geographic location in our enterprise GIS (Smallworld), and sent out to all mobile users the next day. Field sketches are converted to PDF and stored in the Document Management System (Documentum). The sketches are available to all users whether using the mobile system or in the office.

During the upgrade, legacy paper documents were scanned and associated with the applicable leak or repair location. Some of the important challenges we had to overcome are listed in Table 1.

[inline:Fig 1 first graph Risk_Score_Distribution.jpg=Risk Score Distribution]

Challenges UGI Had To Overcome In Re-Engineering The Leak Data Capture Process

A process had to be developed to scan the legacy paper documents and properly associate them with the new digital records.

• We were lucky that the old process used people reviewing the leak and repair reports to determine a Main ID. This was keyed in and available digitally for conversion.
• We elected to manually review “active” leaks and have a GIS technician place geometry based on the paper document.
• Repair geometries were automatically placed at the midpoint of the indicated main id. Engineers contemplating repair/replace can review the documents associated with the repairs and correct the positions if needed.

Facility locations from legacy base maps are not aligned with true GPS locations.

• We don’t try to capture leak locations using GPS for this reason.
• The mobile units have GPS merely to get the person “close” to the right place.

Some pipes have been retired and are no longer present in the current database.

• We were lucky that records were retained for retired pipe.
• This was an issue for us because retired pipe was given a different Main ID as it was retired, but the new pipe got the main ID of the pipe it replaced. The new system accommodates this by having a unique identifier assigned by the system for each pipe segment. The main ID fields can be anything and not break the database integrity.

Leveraging Data And Software For DIMP

We now have an integrated system for tracking the discovery, re-inspection and repair of leaks as well as tracking exposed pipe inspections. While all this re-engineering was going on, we also updated our main replacement prioritization software (Optimain® DS).

This software was originally used by UGI for estimating the economic viability of a main replacement project. It was using data extracted from our mainframe systems for tracking leaks and mains. As described earlier, the link between pipes and leaks was tenuous at best. The other piece of the puzzle was service data. These records were difficult to link to main segments as it was a field in a totally different mainframe system. All these associations had to be maintained manually each time a main segment was replaced, split and/or renumbered. This happens frequently due to our main numbering scheme for hydraulic network modeling.

As we went through the steps to upgrade to the next version of Optimain, it became apparent that we could do more with the risk score that had been computed all along. The new FLAME data model provided the important data relationships to make the risk model more effective and coherent. We now have a system that estimates the economic viability of a main replacement job as well as computes a relative risk score based on information already in our database. The links between facilities and reports no longer get broken as they are based on clear and definitive unique identifiers.

We can also update most of the important cost and risk values used in calculations to tweak the model as we see fit without needing help from the vendor. The other benefit is that we have data from multiple tables accessible in one place, boiled down to the essentials and related in a way that makes sense. Optimain allows us to visualize the risk on our system with charts likes those shown in Figure 1.
[inline:Fig 1 second graph Risk_Score_Comparison.jpg=Risk Score Comparison]

We intend to expand the Optimain risk model going forward as a major part of our DIMP solution. So far we have only scratched the surface of modeling risks associated with services; we intend to expand in this area by taking advantage of our segment by segment service data records that include size, material, coating and vintage. We are also interested in accounting for the influence of geospatial features to risk scores. Adding buildings and other spatial layers like soils data may be valuable. We are also fortunate that our one-call system provides spatial coordinates of tickets. Frequency of locates in a particular area may indicate a need for increased patrolling of different areas at different times.

Having or developing an integrated, robust system for storing and organizing data will go a long way toward complying with the new DIMP regulations.

Associating Leaks With The Right Pipe

In order to assess current risk associated with various threats to the integrity of a distribution network, it is very important to understand the history of leaks on the system. For many common threat categories, including steel corrosion, cast iron breaks, and joint or coupling failures, the number of prior leaks on nearby pipes or joints is statistically correlated with the likelihood that the same type of failure will occur again (in the same general area). When associating leaks with pipes, it is important to consider capturing the location of prior repaired leaks as well as a process and tools for on-going data entry and quality assurance. Several approaches are available to determine how many and which repaired leaks to process:

• Filter repairs by cause and material to focus on specific threat categories.
• Cover at least one normal leak survey cycle (two is better) across the entire system.

By at least capturing the leaks discovered during the most recent leak survey on every active pipe in the network, a risk-assessment system will initially focus attention on areas with recent corrosion, joint leaks, or other time-correlated threats.

When attempting to associate a leak repair with the correct facility in a geographic information system (GIS), there is a series of challenges that must be addressed. These include poor address quality, no zip code or city, multiple candidate address locations, facilities drawn relative to inaccurate (legacy) land-base maps, and retired facilities that are no longer available in the geographic information system (GIS).

Geocoding Process

To address all of these challenges, it is best to implement a multi-stage approach to leak geocoding and pipe association:

1. Construct Candidate Addresses. For each record, construct a series of possible addresses using each of the possible street aliases and each overlapping zip code.
2. Geocode. Obtain all candidate locations for each candidate address.
3. Find Candidate Facilities. For each candidate location, search (in the GIS) for candidate facilities based on attributes of the repair record (main pipe, service pipe, valve, regulator and material type.). Use a maximum search radius that is based on the relative accuracy of the land-base. Widening the search radius will lead to more false positive matches and will also increase the overall processing time.
4. Compute Confidence Score. For each candidate facility, use a combination of available attributes in the repair record and the facility record to compute a score indicating the confidence that it is the correct facility. Use the actual proximity as part of the confidence score (preference for facilities that are closer to the address location when all other attributes match).
5. Select Best Location. Select the candidate site and nearby facility with the highest confidence score. Place the leak on the facility, at the point closest to the best address location.

It is also worthwhile to retain the following information as part of the process:

• Coordinates of the best address location.
• Confidence score.
• Direct relationship of the leak to the best matched facility.

Finally, note that the same basic approach can be applied to match pipe inspection reports or service card locations to the best location in the GIS.

What About The Rest?

Once all leaks have been processed and matched to the best candidate location and facility, there will be a (hopefully small) set of leaks with either no candidate location, or with confidence score below some established threshold. It may be worth manual intervention on some or all of these remaining leaks to attempt to determine where they belong. When considering this remaining manual effort, it is also feasible to filter the list to include only recent leaks or only leaks of certain types. This determination might be based on an overall threat analysis (e.g., manually review all remaining cast iron breaks, but only look at corrosion leaks from the last five years).

Reusable Tools

Many of the tools and algorithms constructed for placing the legacy leaks can be re-used to provide tools to facilitate on-going placement and quality assurance. When mapping technicians are placing new leaks in the GIS, a tool could automatically present candidate locations based on the address and other data in the leak record. The candidate locations can be rank-ordered using the same facility confidence score used above. The user can then snap the leak (or other record) onto the correct facility with just one or two mouse clicks. In fact, it is even possible to completely automate the placement process, and only trigger manual review or intervention when the confidence score is below a certain threshold.

Mobile data capture tools may lead to alternate approaches. Platform restrictions may make it unfeasible to construct the complete facility confidence score on the mobile device. In that case, the score can be computed in the back office as part of a quality control process. Records with low confidence can then be flagged for manual review.

It is also important to consider how open leaks will be handled. Open-leak records will normally not include information about the leaking facility (since the leak has not yet been located or repaired – it is simply an indication of gas odorant in the air). Options include assuming default values or posting only the repaired leaks.

Assume Default Values. It may be best to assume a conservative set of default facility attributes for an open leak. For example, assume an open leak is on the main pipe (not a service). Then match the closest main pipe (regardless of material). If the closest main is cast iron or plastic, assume a joint or coupling leak (since CI breaks and excavation damage normally lead to immediately repaired leaks). If the closest main is steel, assume a corrosion leak. If the steel is allegedly protected, this is also an opportune time to trigger a cathodic protection test.

Only Post Repaired Leaks. It may be reasonable to only post-repaired leaks to the GIS, avoiding this problem altogether. However, this is less desirable if the utility carries a large number of open (non-emergency) leaks. Open leaks are normally more recent and thus have a higher impact on current risk assessments. In addition, a complete economic assessment model needs to consider the on-going O&M costs associated with open leaks (more frequent surveys and eventual cost to repair).

Authors

James Stout is supervisor or pipeline integrity systems integration for UGI Utilities, Inc., Reading, PA. His primary responsibilities are managing leak surveys, administering Optimain from Opvantek and Pipeline Compliance System from American Innovations, contributing to the development of UGI’s Mobile GIS application suite, organizing pipeline integrity data and developing various reporting and data translation processes.

Tony Sileo is product manager for Opvantek, Inc., Newton, PA. He is responsible for establishing product roadmaps based on customer input and market research, directing product development and support activities, managing customer delivery projects, and participating in consultative sales efforts. Recently, he has been working with Opvantek’s LDC customers to expand the capability and coverage of the company’s Optimain DS product to meet all aspects of pending distribution integrity regulations.

References

PHMSA, December 2005, “Integrity Management for Gas Distribution Pipelines, Report of Phase 1 Investigations,”
http://www.cycla.com/opsiswc/docs/S8/P0068/DIMP_Phase1Report_Final.pdf.

BSR GPTC Z380.1-2009 TR05-01-200x, GPTC Guide Appendix G-192-8, Distribution Integrity Management Program (DIMP). at http://www.aga.org/Committees/gotocommitteepages/gaspiping/

Federal Register, Vol. 73, No. 123, June 25, 2008, Proposed Rules, pp. 36015-36034 (DIMP Notice of Public Rulemaking)

Pipeline Inspection, Protection, Enforcement, and Safety (PIPES) Act of 2006, Pub. L 109-468, Dec 29, 2006.