SSP provides utility companies with multiple data-centric offerings such as work order posting, QA/QC, source record validation, and more. Lately, SSP has seen an increase in requests from companies for source record validation services; some are large in both size and scope while others are narrowly focused, such as the scenario discussed below. The catalyst for some of these requests has been the new Gas Mega Rule, but source record validation is important in maintaining the integrity of data within a GIS.

What is the Source Record Validation? Source record validation can be defined as the process of verifying the accuracy and completeness of facility/feature attributes through researching and analyzing original installation documents and updating data based on the results. In some workflows, the source record is attached or linked to the attributes it verifies.

This post will give a brief overview of a source record validation workflow SSP used to verify, and update as needed, the attributes of a gas service record. While a variety of attributes can be queried within a service record to flag conflicting data, the scenario below involves a segment of service pipe flagged due to conflicting material type and install date.

Scenario

Service pipe segment 2 was flagged due to a discrepancy between the Polyethylene (PE) material type and the install date 2/15/1952 because the PE service pipe was not installed as early as 1952. The service record originally listed 3 segments of service pipe: 20’ of 3/4” Steel (ST) from main to stop box, 12’ of 5/8” PE from stop box to lot line, and 30’ of 5/8” PE from the lot line to building wall.

Main to lot line

Lot line to building

SSP’s task was to verify the installation date and material type for the 12’ segment of 5/8” PE. The validation process began with searching for relevant service installation documents to determine which field (or fields) were in error; in this case, Install Date or Material Type.

Searching for Documentation

The source documents had varying levels of information that were captured and used to create smart tags that were searchable in a web-based archive database, e.g., Year, Address Number, Address Street Name, Order Number, etc. Consequently, users had to cast a wide net. Using the database search tools, multiple criteria could be set to find related documents. From the scenario above, here are some of the ways we can search for documentation pertaining to the 12’ segment of 5/8” PE.

• Order # of the service pipe in question or the order # of an existing segment.
• Address Number and street name.
• Main Order #. Reviewing the main installation documents can provide service replacement/reconnect project orders.
• Opening a neighboring service record and searching the installation orders can lead to finding large document files (sometimes 1000+ pages) of all the services replaced/reconnected along that street during the given project.
• Reviewing all field documents associated with the address for notes of related orders.

Scanning Document/Verifying

Upon finding pertinent source records, the address and service information were verified as correct to confirm the installation document was associated to the service record in question. Once confirmed, the entire document was then reviewed. Although the service pipe segment section of the document may be the most relevant for our research in this scenario, there may be field notes elsewhere on the document providing crucial information as well. In some cases, service pipe segments that were installed more recently contained order numbers with relevant information for the previously installed pipe. Searching those newer order numbers can provide answers as the data may have been corrected on the physical document during the most recent fieldwork, but the changes were not applied to the service record resulting in it being flagged for review.

If no relevant documents were found, or if documents were found but had inconclusive information, recording detailed notes of what was found can be helpful for future research efforts. A placeholder number may be left in one of the segment fields so the service record will be easily flagged in future queries, e.g., setting install date to 1/1/1900.

In this example, per field comments written on installation order 450352, the service was replaced with 42’ of 5/8” PE from stop box to building wall. When the data was originally entered into the service record, the 12’ of 5/8” PE in the main to lot line section was incorrectly recorded as being installed on 2/15/1952, when in fact the pipe was part of the 1989 install. This discrepancy was likely due to a misinterpretation of the information written on the field document.

Resolution

With the necessary documentation now located and correct install date verified, segment 2 was combined with segment 3 resulting in two segments of pipe total. Segment 1 being the 20’ of steel pipe from main to stop box, and segment two being the 42’ of 5/8” PE from stop box to building wall.

Main to building

While this may seem like a minor fix, the data contained in these records is used to generate reports which are used by business leaders to make important decisions. Having accurate data at the source leads to accurate decisions being made in the future.

Scrubbing Existing Record

With the service segment in question corrected, the service record is then “scrubbed” to update the existing data to the current standards. Along with updating the service record, the quality of existing data mapped in the GIS is improved as well. When needed, service lateral lengths and paths were corrected along with building footprints and address points.

Conclusion

This has been a look into just one of many source record validation tasks SSP has been asked to perform. SSP has an exceptionally talented data team with extensive experience in researching, analyzing, and validating attributes for gas facilities ranging from service lines to high-pressure transmission mains. Now is a good time to utilize source record validation to strengthen the integrity of your data.