A 2021/2022 Leak Survey of Philadelphia


In the summer of 2021 and 2022, HEET commissioned Gas Safety Inc. to perform two gas leak surveys in Philadelphia using a high‐precision vehicle‐mounted methane analyzer. 

HEET analyzed the leak survey data and found a total of 383 leaks along 1,533 miles of roads, for an average of 4 leaks per mile. Similar surveys of Boston in 2013 and Washington, DC in 2021 found 4.2 and 4.7 leaks per mile, respectively. Both these cities are of similar gas infrastructure age and density to Philadelphia.

The surveys did not grade the leaks for safety. HEET selected at random 10 leak locations found during surveying and Gas Safety verified that these leaks were associated with gas pipes and that the locations were accurate.

The map below shows the two surveys as snapshots in time as of the day of the survey.  Any one leak could have been repaired or eliminated  and new leaks may have developed since the surveys were performed. Use the map to explore the leak locations, noting the following :

  • Readings obtained in this survey are not intended to serve as indicators of safety or hazardousness. Safety determinations are made by Philadelphia Gas Works using site‐specific criteria for identifying safety risks.
  • Leaks in the map represent a snapshot in time as covered by the duration of the surveying. Since the surveying, leaks may have been repaired, pipe replacement may have occurred, and/or new leaks may have developed. 
  • Leaks are on mains or service lines under streets, not inside buildings.
  • No significant correlation was found between the socioeconomic status of the areas covered and the number of leaks per mile. 


Leaks from natural gas infrastructure release methane, CH4, into the atmosphere. Methane is a powerful climate-damaging gas that causes climate change. The leaks create potential safety risks, degrade air quality and damage trees. Identifying the locations of leaks so that they can be prioritized for repair can help quantify and reduce greenhouse gas emissions.

Most gas leaks in the distribution system are associated with old, leak‐prone pipes. Some pipes over a century old can be found, especially in older cities. In 2021, Pennsylvania had 48,753 miles of gas distribution pipe of which 9,148 miles (or 19%) was leak prone, ranking it 6th worst state in the nation1. Gas in Philadelphia is served exclusively by Philadelphia Gas Works (PGW) which is the largest municipally-owned gas utility in the country. In 2021, PGW reported a total of 3,046 miles of mains, 60% of which were leak prone, and 70% of which were installed over 40 years ago.

In May and July of 2021, Gas Safety Inc worked with Sierra Club, HEET and other groups and individuals to conduct a leak survey of the City of Philadelphia, focusing on areas of higher-density gas distribution infrastructure found typically along larger streets containing larger diameter pipes with more pipe connections. The survey found leaks everywhere across the survey route for a total of 1001 leaks or an average of 6.5 leaks road mile across the approximately 154 miles surveyed.

In June and July of 2022, HEET performed a second mobile leak survey covering a diverse but less infrastructure-dense cross-section of Philadelphia city streets (see Appendix 1 for more details). The survey found leaks to be everywhere across the survey route for a total of 532 leaks with an average of 2.3 leaks per mile across the 229 miles surveyed. No significant correlation was found between average socioeconomic status of the census tracts covered and the number of leaks found per mile.

More information from Philadelphia Gas Works (PGW) would be required to investigate or establish any relationships between leak patterns or prevalence and planned or performed pipe replacement locations.

Note that these surveys did not grade the leaks for safety or hazardousness. Philadelphia Gas Works uses their site‐specific criteria to identify safety risks.

During surveying, each road was driven twice wherever possible to verify stationary repeatable leak locations and to eliminate false positives. Air methane concentrations were collected using a high‐precision vehicle‐mounted methane analyzer (Picarro G2311-f). The resulting raw survey geospatial emissions data was processed into leak locations with a new and improved algorithm developed for the Washington, DC project in 2021. The Philadelphia survey identified 532 leaks attributable to natural gas pipes, or 2.3 leaks per mile. 

1 PHMSA – Gas Distribution, Gas Gathering, Gas Transmission, Hazardous Liquids, Liquefied Natural Gas (LNG), and Underground Natural Gas Storage (UNGS) Annual Report Data

Verification of Leak Sources and Locations

Methane can come from sources other than natural gas pipelines, including broken sewer mains, landfills, and wetlands. During the 2022 survey, HEET selected at random ten leak locations found during surveying and Gas Safety Inc. verified that these leaks were associated with gas pipes and were close to gas mains, valves or service lines. The locations of all of these leaks was also verified as accurate.

2022 Surveying Route Selection Criteria

The 2022 road survey was limited to 500 driven road miles of the city. We developed and used the criteria below in guiding the selection of the survey areas. Within the selected areas, driven surveying routes were randomly chosen. 

  • Ensuring a diversity of socioeconomic status areas to facilitate exploration of whether some communities are relatively more or less exposed to leaks:
    • The NSES index was used to provide census-tract based socio-economic index information which used data from the 2011-2015 American Community Survey (ACS).
    • The index incorporates median household income, educational attainment, percentage of individuals with income below the Federal Poverty Line, unemployment rate and the percentage of households with children under the age of 18 that are “female-headed”.
  • Planned and completed gas pipe replacement locations, which could reveal whether or not the encompassing areas have a relatively higher prevalence of leaks:
    • As part of Pennsylvania’s Long Term Infrastructure Improvement Plan (LTIIP), Philadelphia Gas Works is required to submit annual asset optimization plans  (here) each year for the duration of LTIIP. These plans include locations of pipe replacement projects both completed in the previous year and to be performed in 2022.
    • The plans were checked by HEET who noticed there were some inconsistencies which were then corrected by PGW and resubmitted to the PUC. Using the corrected updates, sufficiently detailed locations were extracted, geocoded and mapped for inclusion in route planning.
    • PGW also provides the public information about pipe replacement locations though a pipeline improvement map.
  • Specific neighborhoods relative leak prevalence:
    • We included surveying some specific neighborhoods based on recommendations from meetings with groups and individuals active in the area such as 350.org and POWER Interfaith. These included neighborhoods in South Philadelphia (an area that includes 31 neighborhoods), Hunting Park and Strawberry Mansion.
  • Building on the 2021 city survey:
    • In 2021 Gas Safety Inc. worked with Sierra Club, HEET and other groups and individuals to perform a strategic leak survey of the city. This survey focused on areas of higher density gas mains infrastructure – typically larger streets along which larger diameter infrastructure with more connections tends to be located.
    • Whilst the 2022 survey revisited some areas from that survey to see how leak prevalence may have changed, the 2022 survey visited new and less infrastructure-dense areas to continue building the picture up of leaks across the city.

Other criteria to be explored in future surveys might include: 

  • Detailed gas system characteristics  such as pipe age, pressure, diameter and material.  These would allow for comparison of  leak rates across these different attributes.  Unfortunately, we were not able to find infrastructure details for the system.
  • Council members’ home addresses: these were used in the 2021 survey but not in the 2021 survey because we didn’t have capacity to coordinate with on the ground organizations to use the leak information tactically.
  • Locations of public housing: transitioning public housing areas away from leaky gas infrastructure could have a high impact, but we did not have time to explore these areas.