Methane is a potent greenhouse gas that tends to leak from equipment at oil and gas (O&G) sites. Conventional leak detection and repair methods for fugitive methane emissions are labor-intensive and costly because they involve time-consuming close-range, component-level inspections at each site. This has prompted duty holders to examine new methods and strategies that could be more cost-effective. We examined a cooperative model in which multiple duty holders of upstream O&G sites in a region use shared services to inspect on-site equipment using optical gas imaging camera or Method 21. This approach was hypothesized to be more efficient and cost-effective than independent inspection programs by each duty holder in the region. To test this hypothesis, we developed a geospatial simulation model using empirical data from 11 O&G-producing regions in Canada and the United States. We used the model to compare labor cost, transit time, mileage, vehicle emissions, and driving risk between independent and co-op leak inspection programs. The results indicate that co-op leak inspection programs can generate relative savings in labor costs (1.8%–34.2%), transit time (0.6%–38.6%), mileage (0.2%–43.1%), vehicle emissions (0.01–4.0 tCO2), and driving risk (1.9%–31.9%). The largest relative savings and efficiency gains resulting from co-op leak inspection programs were in regions with a high diversity of duty holders, which was confirmed with simulations of fictitious O&G sites and road networks spanning diverse conditions. We also found reducing leak inspection time by 75% with streamlined methods can additionally reduce labor cost by 8.8%–41.1%, transit time by 5.6%–20.2%, and mileage by 2.60%–34.3% in co-op leak inspection programs. Overall, this study demonstrates that co-op leak inspection programs can be more efficient and cost-effective, particularly in regions with a large diversity of O&G duty holders, and that methods to reduce leak inspection time can create additional savings.

Reducing methane (CH4) emissions from the fossil fuel industry is important for meeting climate targets and net-zero commitments (UN Environment Programme, 2021). Fossil fuels account for an estimated 30%–35% of total global anthropogenic CH4 emissions (Saunois et al., 2020). Although agriculture accounts for a large proportion of anthropogenic CH4 emissions, the fossil fuel industry, particularly the oil and gas (O&G) production segment, is better positioned to reduce CH4 emissions in the near future because abatement technologies already exist. According to the International Energy Agency (IEA, 2022), there are 3 approaches to reduce CH4 emissions from O&G operations: (i) replacement of devices or components that emit CH4 (e.g., conversion from gas-driven to air or electric pneumatics), (ii) installation of new devices that reduce or eliminate vented CH4 emissions (e.g., gas capture), and (iii) leak detection and repair (LDAR). Among these approaches, emission reductions from LDAR are the least well understood because CH4 leaks arise stochastically at O&G sites (Brandt et al., 2016), which suggests that frequent inspections should be implemented to reduce emissions (Ravikumar et al., 2020; IEA, 2022). Accordingly, many jurisdictions have implemented LDAR requirements as part of their CH4 emissions regulations for the O&G sector (Environment and Climate Change Canada [ECCC], 2018; Pennsylvania Department of Environmental Protection [PA DEP], 2018; Colorado Department of Public Health & Environment [CDPHE], 2019; European Union [EU], 2020; New Mexico Environment Department [NMED], 2021; U.S. Environmental Protection Agency [U.S. EPA], 2021; Alberta Energy Regulator [AER], 2022).

LDAR uses techniques and instruments to locate and repair leaking components. The extent of fugitive emission reductions achievable by an LDAR program is significantly influenced by its leak detection efficiency. Conventional leak detection work practices are time-consuming because they rely on close-range component-by-component inspections to identify CH4 emissions or other signs of leakage. Common conventional leak detection work practices include the U.S. EPA’s M21 and alternative work practice using an optical gas imaging (OGI) camera. A suite of new technology platforms and methods has emerged in the last decade to supplement M21 and OGI: satellites, aircraft, drones, vehicles, and continuous monitors (Fox et al., 2019; Ravikumar et al., 2019). The potential to reduce leak inspection costs is a key driver motivating the development and implementation of these technologies and methods. Cost savings could incentivize more frequent surveys to increase emissions reductions (Ravikumar et al., 2020) or free up capital for additional abatement technology and mitigation efforts. Recent work has shown the potential of using aircraft and vehicle systems to increase the efficiency of LDAR (e.g., Schwietzke et al., 2019; Barchyn and Hugenholtz, 2022). However, modeling suggests that cost savings from implementing new technologies are not guaranteed (Fox et al., 2021). Even as emerging technologies are being deployed or considered, there will still be a role for on-site inspections, as OGI cameras and M21 are still the most common regulatory LDAR technology across North America and required to isolate the exact component needing repair (ECCC, 2018; PA DEP, 2018; CDPHE, 2019; EU, 2020; NMED, 2021; U.S. EPA, 2021; AER, 2022).

Mechanisms for lowering leak inspection costs need not rely exclusively on the application of new technology and methods. A report commissioned by the AER (2017) suggested that inspection costs with conventional leak detection methods could be reduced by as much as 50% if performed by O&G facility operators instead of third-party service providers. However, some duty holders (i.e., the party(ies) responsible for developing and conducting LDAR and ensuring regulatory compliance) may not want to invest capital up-front for the technologies and expertise required to perform in-house leak inspections. Another approach to reduce cost is through network-based cooperative services, whereby multiple duty holders of O&G sites in a region cooperate to execute one leak inspection program for the region through shared services, rather than independent leak inspection programs by each duty holder. This approach is called a “co-op leak inspection program.” This concept is primarily motivated by regulations, geography, and allied research from the logistics sector (Cruijssen et al., 2007; Fernández et al., 2018). LDAR regulations establish a unifying goal or requirement to encourage cooperation through shared leak inspection services. Furthermore, many O&G-producing regions are characterized by clusters of wells and facilities operated by different companies. Instead of leapfrogging nearby facilities owned by other duty holders during leak inspections, mobilization costs and transit times between facilities could be reduced through shared leak inspection services at a regional level, which could increase efficiency and lower program costs. In the logistics sector, extensive studies have shown that cost savings from cooperative route planning are promising and vary depending on the purpose of collaboration in different regions (Guajardo and Rönnqvist, 2015; Muñoz-Villamizar et al., 2015; Chabot et al., 2017; Gansterer et al., 2017; Mancini et al., 2021). Therefore, we hypothesized that savings in labor, mileage, and transit time could be achieved using this approach in the context of leak inspections with co-benefits in reducing windshield time, associated driving risk, and vehicle CO2 emissions.

This study examines the potential cost savings and efficiencies of regional co-op leak inspection programs compared to programs executed independently by each duty holder. The research approach involved a geospatial simulation model using both empirical and artificial input data. Empirical data were obtained from 11 O&G-producing regions in Canada and the United States, spanning different road network densities, duty holder diversities, and site densities. In each region, the relative difference in labor cost, mileage, and transit time between co-op and independent leak inspection programs was determined. Additional simulations using artificial O&G sites, duty holders, and road networks were performed to determine the most influential geospatial factors. In addition, a sensitivity analysis to examine the effects of variations in key parameters was performed.

Geospatial simulation model

The geospatial simulation model was developed to simulate leak inspections of O&G sites, whereby a survey of equipment is completed with an OGI and tags are placed on leaking components. The repair portion of LDAR was not included because it is difficult to model in the absence of empirical data.

The simulation model was developed based on finding the local optimal solution of the traveling salesperson problem (TSP), which is an optimization method that determines the shortest driving route to visit predefined locations and return to the point of origin (Applegate, 2007). For this study, the predefined locations were O&G sites (representing wells or facilities) in the region requiring leak inspections and the point of origin was the nearest town or service center where the leak inspection crew was based during the survey campaign. The routing problem belongs to the TSP with fixed time windows (Boland et al., 2017), because daily route planning is constrained by the workday duration of the leak inspection crew. Therefore, simulations were structured to minimize the number of days required to complete the leak inspections by minimizing travel time and maximizing the number of O&G site inspections per day. Geospatial inputs to the simulations included the coordinates of O&G sites requiring leak inspections, roads, and the coordinates of the point of origin. Fixed parameters included the workday duration, survey frequency, number of service centers, driving speed, study area, and the number of leak inspection crews. The only parameter that varied was time on site to complete the leak inspection.

An overview of the simulation workflow and pseudocode is shown in Figure 1. The simulation starts by creating a daily pool of sites requiring leak inspection and identifying the point of origin (i.e., the service center), which is set as the first departure point (DP1). The sites included in the pool depend on the type of program being simulated. For a co-op leak inspection program, all sites within the region are included in the pool. For simulations of independent leak inspection programs, only the sites within the region owned by the specified duty holder are included in the pool, and routes are planned accordingly. Once the site pool is created, the next step determines the site closest to DP1 using Dijkstra’s shortest path algorithm (Dijkstra, 1958). This site is the first leak inspection target of the day and is set as the first arrival point (AP1). The arrival time from DP1 to AP1 is estimated based on the speed limits of the intervening road segments. The travel times and associated distance from DP1 to AP1 are recorded. The daily clock is updated to account for the travel time and the duration of the leak inspection at AP1. Once the leak inspection at AP1 is complete, the next-nearest site in the pool (AP2) is determined, and the leak inspection crew advances. The process is repeated until the end of the workday; however, if the estimated travel time and/or leak inspection time for the next site exceeds the remaining duration of the workday, the site is skipped, and the travel time and distance back to DP1 are recorded. Each day, total distance and travel time are recorded, and the sequence is repeated until the site pool is empty.

Figure 1.

Simulation workflow overview.

Figure 1.

Simulation workflow overview.

Close modal

Empirical data and analysis

We selected 11 O&G-producing regions (Table 1) in Canada and the United States to simulate co-op and independent leak inspection programs and compare the mileage, transit time, and labor costs. These regions were selected because they represent different spatial densities of upstream O&G sites, road networks, and duty holder diversities (i.e., number of duty holders in a region). Each study region was defined as a 2,827 km2 circular area (30 km radius) centered on the towns or cities labeled in Figure 2. These towns and cities served as the points of origin and departure points for the simulations. For each of the 11 regions, 20% of the onshore active wells were randomly selected for simulations. This was done for computational efficiency. Well-level data were selected for sampling because facility-level data do not have consistent geographic data for all regions. O&G well data were obtained from the Homeland Infrastructure Foundation-Level Data (2020) Platform for the study regions in the United States, and the Alberta Energy Regulator’s (2021) ST37 dataset for the study regions in Canada. O&G wells in Lloydminster, SK, were obtained from the Saskatchewan GeoAtlas. Road and trail networks were obtained from OpenStreetMap (OSM) for route planning. The Google Distance Matrix API was applied to calculate the travel time and distance for O&G wells that were inaccessible through the OSM road and trail networks.

Table 1.

Characteristics of upstream oil and gas (O&G) sites, duty holders, and roads in the 11 case study regions

Service CentersTotal # of Duty Holders [20%]Total # of O&G Wells [20%]Total Rural Road Length (km)Duty Holder Diversity [20%]Site Density (sites/km2) [20%]Rural Road Density (km/km2)Data Source
Brooks, AB, Canada 39 [18] 9,002 [1,800] 3,449 0.00 [0.01] 3.18 [0.63] 1.22 Alberta Energy Regulator ST37 
Drumheller, AB, Canada 34 [22] 3,288 [658] 4,665 0.01 [0.03] 1.16 [0.23] 1.65 Alberta Energy Regulator ST37 
Grande Prairie, AB, Canada 45 [24] 565 [113] 4,976 0.08 [0.20] 0.2 [0.04] 1.76 Alberta Energy Regulator ST37 
Lloydminster, AB & SK, Canada 41 [27] 1,176 [235] 6,276 0.03 [0.11] 0.42 [0.08] 2.22 Alberta Energy Regulator ST37 and Saskatchewan GeoAtlas 
Rocky Mountain House, AB, Canada 52 [38] 2,001 [400] 4,806 0.03 [0.12] 0.71 [0.14] 1.70 Alberta Energy Regulator ST37 
Bakersfield, CA, USA 98 [75] 18,199 [3,640] 1,4870 0.01 [0.02] 6.44 [1.29] 5.26 Homeland Infrastructure Foundation-Level Data Platform 
Farmington, NM, USA 72 [52] 4,401 [880] 10,545 0.02 [0.06] 1.56 [0.31] 3.73 Homeland Infrastructure Foundation-Level Data Platform 
Greeley, CO, USA 44 [30] 14,397 [2,879] 17,527 0.00 [0.01] 5.09 [1.01] 6.20 Homeland Infrastructure Foundation-Level Data Platform 
Odessa, TX, USA 324 [173] 13,427 [2,685] 11,506 0.02 [0.06] 4.74 [0.95] 4.07 Homeland Infrastructure Foundation-Level Data Platform 
Robstown, TX, USA 394 [154] 3,597 [700] 9,244 0.11 [0.22] 1.27 [0.25] 3.27 Homeland Infrastructure Foundation-Level Data Platform 
Indiana, PA, USA 118 [84] 11,775 [2,355] 13,004 0.01 [0.03] 4.17 [0.83] 4.60 Homeland Infrastructure Foundation-Level Data Platform 
Service CentersTotal # of Duty Holders [20%]Total # of O&G Wells [20%]Total Rural Road Length (km)Duty Holder Diversity [20%]Site Density (sites/km2) [20%]Rural Road Density (km/km2)Data Source
Brooks, AB, Canada 39 [18] 9,002 [1,800] 3,449 0.00 [0.01] 3.18 [0.63] 1.22 Alberta Energy Regulator ST37 
Drumheller, AB, Canada 34 [22] 3,288 [658] 4,665 0.01 [0.03] 1.16 [0.23] 1.65 Alberta Energy Regulator ST37 
Grande Prairie, AB, Canada 45 [24] 565 [113] 4,976 0.08 [0.20] 0.2 [0.04] 1.76 Alberta Energy Regulator ST37 
Lloydminster, AB & SK, Canada 41 [27] 1,176 [235] 6,276 0.03 [0.11] 0.42 [0.08] 2.22 Alberta Energy Regulator ST37 and Saskatchewan GeoAtlas 
Rocky Mountain House, AB, Canada 52 [38] 2,001 [400] 4,806 0.03 [0.12] 0.71 [0.14] 1.70 Alberta Energy Regulator ST37 
Bakersfield, CA, USA 98 [75] 18,199 [3,640] 1,4870 0.01 [0.02] 6.44 [1.29] 5.26 Homeland Infrastructure Foundation-Level Data Platform 
Farmington, NM, USA 72 [52] 4,401 [880] 10,545 0.02 [0.06] 1.56 [0.31] 3.73 Homeland Infrastructure Foundation-Level Data Platform 
Greeley, CO, USA 44 [30] 14,397 [2,879] 17,527 0.00 [0.01] 5.09 [1.01] 6.20 Homeland Infrastructure Foundation-Level Data Platform 
Odessa, TX, USA 324 [173] 13,427 [2,685] 11,506 0.02 [0.06] 4.74 [0.95] 4.07 Homeland Infrastructure Foundation-Level Data Platform 
Robstown, TX, USA 394 [154] 3,597 [700] 9,244 0.11 [0.22] 1.27 [0.25] 3.27 Homeland Infrastructure Foundation-Level Data Platform 
Indiana, PA, USA 118 [84] 11,775 [2,355] 13,004 0.01 [0.03] 4.17 [0.83] 4.60 Homeland Infrastructure Foundation-Level Data Platform 

Square brackets indicate values for 20% of O&G sites randomly sampled from each region for the model simulations.

Figure 2.

Selected oil and gas (O&G)-producing regions for the present study. The color gradation represents the number of active O&G wells within each 30 km × 30 km grid cell. The 11 study regions are highlighted with black circles. The sources of the data used in this map are described in the text.

Figure 2.

Selected oil and gas (O&G)-producing regions for the present study. The color gradation represents the number of active O&G wells within each 30 km × 30 km grid cell. The 11 study regions are highlighted with black circles. The sources of the data used in this map are described in the text.

Close modal

We used the following fixed parameter settings in all simulations for commensurability: 8-h workday, 1 service center (departure point) in each region, 1 leak inspection crew, and 1 leak inspection survey per site. To simulate real world leak inspection times, we sampled the leak inspection time for each site from a probability distribution. This distribution was obtained from an anonymous LDAR service provider (Figure 3).

Figure 3.

Probability distribution of optical gas imaging (OGI) camera leak inspection times. The red curve indicates the probability density function of the leak inspection times probability distribution. Each underlying data point is associated with the OGI inspection time of one site.

Figure 3.

Probability distribution of optical gas imaging (OGI) camera leak inspection times. The red curve indicates the probability density function of the leak inspection times probability distribution. Each underlying data point is associated with the OGI inspection time of one site.

Close modal

Table 1 provides an overview of the characteristics of the 11 study regions based on spatial metrics affecting the mileage and labor costs of leak inspection programs: duty holder diversity (DHD), site density (SD), rural road density (RRD), and site clustering per duty holder (SCDH). SCDH describes the clustering of LDAR sites for each duty holder. Mathematically, it is the average proportion of sites within 5 km of each site that have same LDAR duty holder. The equations for calculating these 4 factors are described as follows:

DHD=Ndh1Ns ,
1
SD=NsA ,
2
RRD=LLu A ,
3
SCDH= jNsNsDH_/Ns_Ns,
4

where Ndh is the total number of duty holders in the region, Ns is the total number of sites, A is the area of the region (2,827 km2), L is the total length (km) of roads and trails in the region, Lu is the length of roads and trails (km) within the census boundary of the service center, Ns_ is the number of sites within 5 km of site j, and NsDH_ is the number of sites in Ns_ that have the same duty holder as j. A 5 km search radius was used for all study regions to ensure equal boundary effect from the spatial clustering analysis. RRD is positively correlated to SD, because each site has a connecting road or trail segment.

Performance metrics

Time- and distance-based metrics were used to evaluate the performance of the independent and co-op leak inspection programs in each study region. These measurements were recorded daily and summed for each region to determine the total time and distance to complete the leak inspections at the end of each simulation. The time required to complete the leak inspection program was used as a proxy for labor costs (Lbc). Travel time and distance were used to estimate the mileage (M) and transit time (Tt) required to complete the LDAR program. The relative savings in labor cost (ΔLbc), mileage (ΔM), and transit time (ΔTt) are given by percentages:

ΔLbc=(iNtdh_itsiNtdh_i)×100,
5
ΔM=(iNMdh_iMsiNMdh_i)×100,
6
ΔTt=(iNTdh_iTsiNTdh_i)×100,
7

where tdh_i, Mdh_i, and Tdh_i are the estimated working hours, mileage, and transit time to complete leak inspection programs for duty holder i; N is the number of duty holders; and ts, Ms, and Ts are the estimated working hours, mileage, and transit time, respectively, required to complete the co-op leak inspection programs.

We also estimated the potential driving risk exposure density (RED) of each leak inspection program in the 11 regions using the metric developed by Rolison and Moutari (2017):

RED=TtM×Df,
8

where Df is driving frequency. Since we assume crews work every day in the model, Df is equivalent to the number of days required to complete leak inspections. The potential risk reduction (RR) is given by:

RR=(iNREDdh_iREDsiNREDdh_i )×100,
9

where REDdh_i is the estimated risk exposure density associated with the completion of leak inspections for duty holder i and REDs is the estimated risk exposure density associated with the completion of leak inspections in the co-op program.

Vehicle CO2 emissions reductions from the co-op leak inspection program were calculated based on an average vehicle emission of 252.5 gCO2 km−1 (U.S. EPA, 2018). Total potential vehicle CO2 emissions reductions were determined by multiplying 252.5 gCO2 km−1 by the difference in mileage between the simulated independent (iNMdh_i) and the co-op leak inspection programs (Ms).

Simulations with fictitious O&G sites, roads, and duty holders

The goal of these simulations was to verify and expand on results from the empirical case studies by isolating the effects of key geographic factors on program efficiencies and cost-savings. The procedures were identical to the workflow in Figure 1; however, instead of using real O&G sites and roads to simulate leak inspection programs, we used fictitious sites and roads as inputs. These data were designed to examine different levels of site and road densities, and DHD. We also examined the effects of road network structure, noting that irregular road networks connecting O&G sites in forested settings or complex terrain differ substantially from the regular, grid-based networks common in many agricultural and non-forested settings in North America.

The simulation area was a square, 30 km × 30 km (900 km2) flat region with a local coordinate system and one service center located at the center (Figure 4). For the gridded road network (Figure 4a), we added 14 evenly spaced highways (7 N–S and 7 E–W) to create 36 grid squares. Backroads were added to connect every O&G site to the nearest highway. The speed limits for the highways and backroads were 80 km/h and 30 km/h, respectively. For the irregular road network (Figure 4b), we added 4 highways from the service center to the edges of the study region in each cardinal direction to create 4 subregions. Within each subregion, one secondary road was added to connect 2 highways and multiple backroads were added to connect the backroads with each site. The maximum speed limits for highways, secondary roads, and backroads were 80 km/h, 50 km/h, and 30 km/h, respectively. Note that an elevation model was not included in the simulations; road network geometry is the only difference between non-forested and forested settings.

Figure 4.

Design of fictitious oil and gas (O&G) sites and roads for simulations: (a) gridded road network, common in agricultural and non-forested settings; and (b) irregular road network, common in forested settings and complex terrain. An elevation model was not used in the creation of road networks.

Figure 4.

Design of fictitious oil and gas (O&G) sites and roads for simulations: (a) gridded road network, common in agricultural and non-forested settings; and (b) irregular road network, common in forested settings and complex terrain. An elevation model was not used in the creation of road networks.

Close modal

Two groups of experiments were designed to simulate the independent and co-op leak inspection programs using artificial O&G sites and roads. The first group examined the effect of DHD using 500 randomly distributed O&G sites with connecting backroads in regular and irregular road networks. The inspection time of O&G sites was sampled from the OGI inspection time distribution in Figure 3. Simulations were run using 1–100 duty holders randomly assigned to each site, producing duty holder diversities from 0.0 to 0.2. For each simulation, the time (i.e., the number of days), mileage (km), and transit time (hours) required to complete the leak inspection programs were recorded and used to calculate the metrics in Equations 5, 6, and 7. In this group of experiments, site location, DHD, and the inspection time of each site were random variables. The simulations for every scenario were repeated 100 times to minimize the impact of the random sampling of location, duty holder, and inspection time for O&G sites.

The second group of simulations was used to isolate the effect of SD. DHD was held constant. We did not design simulations specifically to determine the effect of road density because it interacts with SD; road density increases as SD increases in the model. We started with a random sample of 100 sites, randomly assigned to 5 duty holders, which resulted in a DHD of 0.05. In this configuration, we simulated 1 co-op leak inspection program and 5 independent leak inspection programs. Savings in labor cost, mileage, and transit time were calculated for the co-op leak inspection program and summed for the 5 independent leak inspection programs. Subsequently, the number of candidate leak inspection sites was increased to 200, producing a SD of 0.22. The number of duty holders was increased to 10, which maintained the DHD at 0.05. For the new sites, roads were created to ensure site accessibility. Savings in labor cost, mileage, and transit time were calculated for the shared services program and summed for the 20 independent leak inspection programs. This process was repeated until the number of sites and duty holders were incremented to 1,000 and 100, respectively. This established the upper bound of SD at 1.11 sites per km2. Similarly, we repeated the simulations 100 times to minimize the impact of the random sampling of location, duty holder, and inspection time for O&G sites.

Simulations with empirical O&G sites, road networks, and duty holders

Simulations of the 11 O&G-producing regions indicate that co-op leak inspection programs reduced mileage by 0.2%–43.1%, labor cost by 2.5%–34.2%, and transit time by 0.6%–38.6%, compared to the sum totals of independent leak inspection programs (Figure 5a). These results are in the same range of reductions (4%–46%) reported from shared services analyses (i.e., co-ops) in the logistics sector (Guajardo and Rönnqvist, 2015). A wide variation was observed between regions, with relative savings of 3 metrics from each region ranging from approximately 1% in Bakersfield, CA, to approximately 37% in Robstown, TX.

Figure 5.

Results from empirical simulations: (a) relative savings in mileage, labor, and transit time for co-op leak inspection programs in 11 study regions; and (b) potential reductions in risk-exposure density (brown) and vehicle CO2 emissions reduction (blue) for co-op leak inspection programs in 11 study regions.

Figure 5.

Results from empirical simulations: (a) relative savings in mileage, labor, and transit time for co-op leak inspection programs in 11 study regions; and (b) potential reductions in risk-exposure density (brown) and vehicle CO2 emissions reduction (blue) for co-op leak inspection programs in 11 study regions.

Close modal

Simulated reductions of vehicle CO2 emissions and risk for co-op leak inspection programs in the 11 O&G-producing regions are shown in Figure 5b. These metrics are increasingly important for duty holders and service providers when assessing their risk portfolio and scope 1 emissions. The pattern is similar to Figure 5a. Co-op leak inspection programs in these regions reduced potential driving risk by 1.9%–31.9% and vehicle emissions by 0.01–4.0 tCO2 compared to the sum totals of the independent leak inspection programs. The maximum potential reduction (4.0 tCO2) is slightly lower than the annual emissions of 1 passenger vehicle (4.6 tCO2) based on combustion of 2,000 liters of gasoline. Grande Prairie, AB, and Robstown, TX, stand out with the largest reductions in driving risk in co-op leak inspection programs. These 2 regions have the highest duty holder diversities in the simulations (Table 1). Odessa, TX, and Indiana, PA, are 2 regions where co-op leak inspection programs resulted in the largest reductions of vehicle CO2 emissions. These 2 regions exhibited higher rural road and site densities compared to most study regions (Table 1).

Figure 6 suggests that the variations in labor cost, mileage, and transit time savings between regions may be related to geographic factors. Two geographic factors (independent variables) that stand out based on R2 values are DHD and SCDH. The correlations are positive between all 3 performance metrics and DHD (Figure 6a), with R2 values ranging from 0.32 (P < 0.1) for transit time to 0.91 (P < 0.01) for labor cost. The latter is notable and reflects the increase in the number of daily site inspections completed in co-op leak inspection programs as the number of duty holders in a region increases. In this case, leapfrogging sites in independent leak inspection programs increases the transit time and the total time it takes to complete leak inspections at all sites in the region. The result is a higher total number of work hours and labor cost compared to co-op leak inspection programs.

Figure 6.

Scatter plots comparing relative savings with 4 geographic factors in the 11 study regions: (a) duty holder diversity, (b) site density, (c) rural road density, and (d) site clustering per duty holder. Blue, green, and orange regression lines correspond to labor cost, mileage, and transit time savings, respectively. The R2 and P values between each program metric and geographic factor are listed in corresponding colors.

Figure 6.

Scatter plots comparing relative savings with 4 geographic factors in the 11 study regions: (a) duty holder diversity, (b) site density, (c) rural road density, and (d) site clustering per duty holder. Blue, green, and orange regression lines correspond to labor cost, mileage, and transit time savings, respectively. The R2 and P values between each program metric and geographic factor are listed in corresponding colors.

Close modal

The second geographic factor that stands out in Figure 6 is SCDH (Figure 6d). This refers to the clustering of sites for each duty holder in a region. The significant and negative regressions indicate that regions with highly clustered sites per duty holder have less potential for savings in co-op leak inspection programs, particularly for mileage and transit time (R2 = 0.86, P < 0.001). When sites are more clustered, there is less driving distance and transit time between sites in independent leak inspection programs, so the benefits of co-op leak inspection programs decrease. However, in regions where the sites for each duty holder are more dispersed, the benefits of co-op leak inspection programs become more pronounced. The regression models suggest a reduction of roughly 7%–8% in mileage and transit time for every 20% reduction in site clustering.

We performed multivariate regression to examine how the combination of independent variables influence cost reduction for each dependent variable in the co-op leak inspection programs. RRD was excluded due to multicollinearity with SD and the limited response from the dependent variables. The initial regression models were developed based on the R2 values in Figure 6. Subsequently, we applied the backward stepwise selection and Akaike information criterion (AIC) to eliminate insignificant factors and determine the final regression model. The resultant predictive equations and associated R2 values are:

ΔLbc=157×DHD+0.68, R2=0.96,
10
ΔM=51.8×DHD+2.2×SD33.8×SCDH+28.3,R2=0.87,
11
ΔTt=37.8×DHD+2.3×SD31.6×SCDH+26.7,R2=0.86.
12

These models indicate that DHD is an important predictor of reductions for all 3 dependent variables in co-op leak inspection programs but is the exclusive predictor of labor cost savings. Mileage and transit time savings in co-op leak inspection programs are predicted by combinations of DHD, SD, and site clustering by duty holder. Although the models were determined with the minimum AIC, the coefficients may change if different regions are simulated. Since the models are based on a small sample size, their primary value is in highlighting individual and combinations of geographic factors that appear to play a role in determining the magnitude of savings and reductions in co-op leak inspection programs.

Simulations with artificial O&G sites, road networks, and duty holders

Simulations with artificial sites and road networks were used to examine the effect of SD, DHD, and RRD on program-level savings in labor cost, mileage, and transit time using a co-op approach. Random sampling precluded an evaluation of the effects of site clustering by duty holder. A secondary goal was to examine whether savings in co-op leak inspection programs were affected by road network structure by comparing simulations of 2 distinct road configurations. The latter is apparent from the plots in Figure 7, which consistently show larger savings in simulations with irregular road networks. This is likely related to fewer access options for each site in irregular networks, which has the effect of slightly increasing the time penalty from leapfrogging sites in independent leak inspection programs. As a result, there appears to be a slight advantage for co-op leak inspection programs in irregular networks.

Figure 7.

Simulated relative savings from co-op leak inspection programs based on fictitious sites and road networks. Lines in panels (a–c) and (d–f) indicate the relative savings of co-op leak inspection programs across relative savings of 100 simulations (shaded area) with increases in duty holder diversity and site density, respectively. Since no simulations were performed specifically for the effect of road density, scatters in panels (g–i) represent the relative savings with increasing rural road density derived from simulations of site density (d–f). The horizontal error bars indicate the variations of road densities for different site densities in the model. The vertical error bars correspond to the range of relative savings across 100 simulations.

Figure 7.

Simulated relative savings from co-op leak inspection programs based on fictitious sites and road networks. Lines in panels (a–c) and (d–f) indicate the relative savings of co-op leak inspection programs across relative savings of 100 simulations (shaded area) with increases in duty holder diversity and site density, respectively. Since no simulations were performed specifically for the effect of road density, scatters in panels (g–i) represent the relative savings with increasing rural road density derived from simulations of site density (d–f). The horizontal error bars indicate the variations of road densities for different site densities in the model. The vertical error bars correspond to the range of relative savings across 100 simulations.

Close modal

The largest savings in co-op leak inspection programs were based on simulations that isolated the effect of DHD (Figure 7a–c). The red squares in Figure 7a–c indicate the reference case, where there is only one duty holder in a region and DHD is zero. As the DHD increased in the simulations, savings in labor cost, mileage, and transit time also increased in co-op leak inspection program compared to dependent leak inspection programs. Mileage and transit time reductions follow logarithmic patterns up to the maximum tested DHD (0.2). The near linear increase in labor cost savings for duty holder diversities between 0.05 and 0.2 confirms the strong positive correlations in the empirical analyses. Savings associated with increasing site and rural road densities (Figure 7d–i) are similar and appear to approximate logarithmic trends for the ranges examined. It is notable that the only bivariate relation observed in the simulations consistent with the empirical analyses is between DHD and labor cost reductions. This suggests DHD is a reliable geographic proxy for labor cost savings in co-op leak inspection programs.

Sensitivity analysis

We performed a sensitivity analysis to examine how different parameterizations could impact results. We designed 6 sets of simulations using the One-Factor-At-a-time (OAT) method (Razavi and Gupta, 2015) to examine variations of 6 parameters: area of the study region (400–6,400 km2), number of service centers (1–7), work hours per day (6–12 h), number of inspection crews (1–5), driving speed by road type (highways: 50–80 km/h and backroads: 10–50 km/h for gridded road network; highways: 60–100 km/h, secondary roads: 10–50 km/h for irregular road network), and inspection time on site (10%–90% of sampled inspection time). Each set of simulations was performed with both gridded and irregular road networks. Table S1 of the Supplementary Information lists the detailed parameter combinations used for the 6 simulation sets. We determined the lower and upper boundaries of each parameter based on possible real-world scenarios. For simulating the effect of decreasing inspection time on site while maintaining the same proportion of long and short inspection sites, we recorded sampled inspection times of all leak inspection sites at the start of the simulation. We then decreased inspection times on site by 10% for each simulation. We also calculated the relative percentage of change for all parameters by averaging the simulation results of gridded and irregular road networks.

The results of the sensitivity analysis (Figure 8) indicate that estimates of mileage, transit time, and labor cost were most sensitive to inspection time on site. As inspection time decreases, the number of site inspections completed each day increases in the simulations, on average, which reduces the total number of workdays required to complete the program. As a result, the travel time and mileage also decrease because there are fewer mobilizations between the point of origin (i.e., service center) and candidate O&G sites. By extension, there are additional benefits for reducing driving risk and vehicle emissions. Results were also sensitive to the number of working hours per day. As the working hours increase, the total number of days to complete the leak inspection program decreases, since more sites can be completed each day. The only other notable sensitivity was driving speed, which reduces transit time as speed increases.

Figure 8.

Sensitivity analysis results of 6 parameters used in the spatial model. The colors indicate the relative percentage changes.

Figure 8.

Sensitivity analysis results of 6 parameters used in the spatial model. The colors indicate the relative percentage changes.

Close modal

Based on the importance of inspection time on site in the sensitivity analysis, we investigated the effect of shorter inspection times (25%, 50%, and 75%) for independent and co-op leak inspection programs in the 11 study regions. The calculated relative savings of shorter inspection time were compared to relative savings from the simulations using initial inspection time. Results indicate that additional savings in labor cost, mileage, and transit time can be achieved in co-op leak inspection programs (Figure 9). Reducing inspection time by 75% resulted in an additional savings of 8.8%–41.1% for labor cost, 2.6%–34.3% for mileage, and 5.6%–20.2% for transit time. Correspondingly, additional reductions in driving risk and vehicle CO2 emissions could be achieved as well. In most study regions (i.e., Grand Prairie, Lloydminster, and Robstown), labor cost savings in shared services programs more than doubled when inspection time decreased by 75%. The most notable improvements occurred in regions with the smallest labor cost savings before simulated inspection times were shortened (e.g., Drumheller, AB).

Figure 9.

Simulated effect of shorter leak inspection times on the relative savings for the 11 study regions. The savings from the initial simulations are equivalent to those shown in Figure 5a. Additional savings from reducing leak inspection times by 25%, 50%, and 75% are shown. Blue, yellow, and green color palettes represent potential labor cost, transit time, and mileage savings, respectively.

Figure 9.

Simulated effect of shorter leak inspection times on the relative savings for the 11 study regions. The savings from the initial simulations are equivalent to those shown in Figure 5a. Additional savings from reducing leak inspection times by 25%, 50%, and 75% are shown. Blue, yellow, and green color palettes represent potential labor cost, transit time, and mileage savings, respectively.

Close modal

One strategy that can reduce inspection time involves the use of a vehicle system to screen each site and direct where the OGI camera is used to find and tag emitting components. This work practice is referred to as OGI triage (Barchyn and Hugenholtz, 2022). Vehicle systems equipped with sensors that measure CH4, location, and wind conditions are well established for research and industrial applications in the upstream O&G sector (Fox et al., 2019; Gao et al., 2022). For leak inspections, these systems can be used to accelerate OGI inspections by identifying emitting and nonemitting equipment on site (Barchyn and Hugenholtz, 2020). OGI follow-up inspections can then target only the emitting equipment to identify and tag the source component. The time saving occurs when there is nonemitting equipment on site, which can be skipped as it does not require time-consuming, component-by-component OGI inspection. The maximum time saving occurs at sites where none of the equipment is emitting as there is no requirement to follow-up with OGI. The main limitation of a vehicle platform is that some systems may miss small leaks because they have higher minimum detection limits, although some also have detection limits comparable to OGI (Barchyn and Hugenholtz, 2020). Driving accessibility on site also may impact the effectiveness of the vehicle system. Of course, there are many factors other than sensor platform that affect LDAR performance. For example, survey crew experience or more efficient close-range technology would also be expected to reduce the leak inspection time on site.

We developed a geospatial simulation model with real and artificial data to examine whether a cooperative approach to leak inspection programs at upstream O&G sites could be more efficient and cost-effective than a series of independent leak inspection programs carried out by individual duty holders. The simulations indicate that co-op leak inspection programs can generate savings in labor, mileage, and transit time, with corollary reductions in driving risk and vehicle CO2 emissions. The largest labor cost savings occur in regions with the greatest diversity of site duty holders. In these regions, co-op leak inspection programs are more cost-effective because they reduce the transit time between O&G sites. Instead of leapfrogging sites in independent leak inspection programs because they belong to other duty holders, leak inspection crews in a co-op program can optimize their driving and complete more leak inspections each day if they always target the nearest site. The result is fewer total working days in co-op programs to complete leak inspections at all candidate sites in a region. A sensitivity analysis revealed that the most important parameter in the geospatial simulation model was time on site. Thus, methods to reduce leak inspection times, such as OGI triage using vehicle systems, can produce labor cost savings up to 72.5%, and lower mileage and transit times by up to 65.6% and 58.8%, respectively, in co-op leak inspection programs compared to independent leak inspection programs.

These results suggest that efficiencies are available to leak inspections if a co-op approach is taken. The maximum benefits of shared operations exist in geographic situations with many interspersed duty holders. The results and simulations developed in this study demonstrate that material efficiency gains can be found through operations optimization. Furthermore, this study highlight how new research can contribute to the broader goals of cost-effective emissions reductions for the O&G sector, even within the widely implemented LDAR modality of onsite leak inspection.

Practical implementation of co-op leak inspection program requires some shared management of the duty holder requirements to inspect sites and design surveys to minimize travel time. This type of cooperation is certainly possible and not often limited by regulations. We suggest carefully controlled field studies in regions where coalitions already exist, such as the Sundre Petroleum Operators Group (SPOG) in Alberta, the Marcellus shale collation, and the Permian strategic partnership in Texas. The testing also provides an opportunity to review and improve current regulatory frameworks. Related, it is important to note that we only considered travel time when determining the visiting order of sites. Hierarchical clustering can be added to the Dijkstra algorithm to change the rule of site visiting. For example, a rule can be implemented that prioritizes visits to sites with more super-emitters found in the past or high-emitting sites flagged by vehicle or aircraft systems. With appropriate data, the geospatial simulation tool developed in this study can be extended from ground-based leak inspection to aerially guided leak inspection. Furthermore, this study only confirmed the cost savings of co-op leak inspection programs; more granular economic modeling and cost allocation methods are needed to minimize the costs of all participants in the co-op leak inspection program.

Data used for simulations and raw results can be found in https://doi.org/10.7910/DVN/CMCM8P. The geospatial simulation model and analysis were programmed in Python with standard packages. The results can be reproduced by employing the equations, explanation, and parameters provided in the main text. Additional code and data are available upon request.

The supplemental files for this article can be found as follows:

Table S1. Parameters combination used for six groups of controlled simulations.

The authors acknowledge that the concept for this study was based on early discussions with Wayne Hillier.

The authors thank the University of Calgary’s Global Research Initiative for funding support to MG.

The authors have declared that no competing interests exist.

Contributed to conception and design: MG, CHH.

Contributed to acquisition of data: MG.

Contributed to analysis and interpretation of data: MG, CHH.

Drafted and/or revised the article: MG, CHH, MS, TEB, TRG, CV, ZX.

Approved the submitted version for publication: MG, CHH, MS, TEB, TRG, CV, ZX.

Alberta Energy Regulator
.
2017
.
Methane abatement costs: Alberta
[online].
Alberta Energy Regulator
.
Available at
https://static.aer.ca/prd/documents/DelphiAlbertaMethaneAbatementCostStudy.pdf.
Accessed February 20, 2023
.
Alberta Energy Regulator
.
2021
.
ST37: List of wells in Alberta
[online].
Alberta Energy Regulator
.
Available at
https://www.aer.ca/providinginformation/data-and-reports/statistical-reports/st3.
Accessed February 20, 2023
.
Alberta Energy Regulator
.
2022
.
Directive 060
[online].
Alberta Energy Regulator
.
Available at
https://www.aer.ca/regulating-development/rules-and-directives/directives/directive-060.html.
Accessed February 20, 2023
.
Applegate
,
DL.
2007
.
The traveling salesman problem: A computational story
.
Princeton, NJ
:
Princeton University Press
.
Barchyn
,
T
,
Hugenholtz
,
C.
2020
.
University of Calgary rapid vehicle-based methane emissions mapping system (PoMELO) single-blind testing results from the Methane Emissions Technology Evaluation Center (METEC)
[online]. Available at
https://pomelo-methane.ca/UCalgary_PoMELO_METEC_Detection_2020.pdf.
Accessed February 19, 2023
.
Barchyn
,
T
,
Hugenholtz
,
C.
2022
.
Complex multi-source emissions quantification results for the PoMELO vehicle measurement system, test results from the CSU METEC facility
. DOI: https://dx.doi.org/10.31223/x5xp7b.
Boland
,
N
,
Hewitt
,
M
,
Vu
,
DM
,
Savelsbergh
,
M.
2017
.
Solving the traveling salesman problem with time windows through dynamically generated time-expanded networks
, in
Salvagnin
,
D
,
Lombardi
,
M
eds.,
Integration of AI and OR techniques in constraint programming
.
Cham, Switzerland
:
Springer
:
254
262
. DOI: https://doi.org/10.1007/978-3-319-59776-8_21.
Brandt
,
AR
,
Heath
,
GA
,
Cooley
,
D.
2016
.
Methane leaks from natural gas systems follow extreme distributions
.
Environmental Science & Technology
50
(
22
):
12512
12520
. DOI: https://dx.doi.org/10.1021/acs.est.6b04303.
Chabot
,
T
,
Bouchard
,
F
,
Legault-Michaud
,
A
,
Renaud
,
J
,
Coelho
,
LC.
2018
.
Service level, cost and environmental optimization of collaborative transportation
.
Transportation Research Part E: Logistics and Transportation Review
110
:
1
14
. DOI: https://dx.doi.org/10.1016/j.tre.2017.11.008.
Colorado Department of Public Health & Environment
.
2019
.
(Colorado) Regulation No.7: Control of ozone precursors and control of hydrocarbons via oil and gas emissions
[online]. Available at
https://cdphe.colorado.gov/aqcc-regulations.
Accessed February 20, 2023
.
Cruijssen
,
F
,
Cools
,
M
,
Dullaert
,
W.
2007
.
Horizontal cooperation in logistics: Opportunities and impediments
.
Transportation Research Part E: Logistics and Transportation Review
43
(
2
):
129
142
. DOI: https://dx.doi.org/10.1016/j.tre.2005.09.007.
Dijkstra
,
EW.
1959
.
A note on two problems in connexion with graphs
.
Numerische Mathematik
1
(
1
):
269
271
. DOI: https://dx.doi.org/10.1007/bf01386390.
Environment and Climate Change Canada
.
2018
.
Regulations respecting reduction in the release of methane and certain volatile organic compounds (Upstream Oil and Gas Sector) (SOR/2018-66)
[online]. Available at
https://laws-lois.justice.gc.ca/eng/regulations/SOR-2018-66/.
Accessed February 20, 2023
.
European Union
.
2020
.
EU strategy to reduce methane emissions
[online]. Available at
https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A52020DC0663.
Accessed February 19, 2023
.
Fernández
,
E
,
Roca-Riu
,
M
,
Speranza
,
MG
.
2018
.
The shared customer collaboration vehicle routing problem
.
European Journal of Operational Research
265
(
3
):
1078
1093
. DOI: https://dx.doi.org/10.1016/j.ejor.2017.08.051.
Fox
,
TA
,
Barchyn
,
TE
,
Risk
,
D
,
Ravikumar
,
AP
,
Hugenholtz
,
CH.
2019
.
A review of close-range and screening technologies for mitigating fugitive methane emissions in upstream oil and gas
.
Environmental Research Letters
14
(
5
):
053002
. DOI: https://dx.doi.org/10.1088/1748-9326/ab0cc3.
Fox
,
TA
,
Hugenholtz
,
CH
,
Barchyn
,
TE
,
Gough
,
TR
,
Gao
,
M
,
Staples
,
M.
2021
.
Can new mobile technologies enable fugitive methane reductions from the oil and gas industry?
Environmental Research Letters
16
(
6
):
064077
. DOI: https://dx.doi.org/10.1088/1748-9326/ac0565.
Gansterer
,
M
,
Hartl
,
RF.
2018
.
Collaborative vehicle routing: A survey
.
European Journal of Operational Research
268
(
1
):
1
12
. DOI: https://dx.doi.org/10.1016/j.ejor.2017.10.023.
Gao
,
M
,
Hugenholtz
,
CH
,
Barchyn
,
T.
2022
.
Development and validation of a route planning methodology for vehicle-based remote measurements of methane and other emissions from oil and gas wells and facilities
.
Journal of the Air & Waste Management Association
72
(
2
):
1279
1289
. DOI: https://dx.doi.org/10.1080/10962247.2022.2113182.
Guajardo
,
M
,
Rönnqvist
,
M.
2015
.
A review on cost allocation methods in collaborative transportation
.
International Transactions in Operational Research
23
(
3
):
371
392
. DOI: https://dx.doi.org/10.1111/itor.12205.
Homeland Infrastructure Foundation-Level Data
.
2020
.
Oil and natural gas wells
[online].
Homeland Infrastructure Foundation-Level Data
.
Available at
https://hifld-geoplatform.opendata.arcgis.com/datasets/geoplatform::oil-and-natural-gas-wells.
Accessed February 20, 2023
.
International Energy Agency
.
2022
.
Methane emissions from oil and gas operations—Analysis
[online]. Available at
https://www.iea.org/reports/methane-emissions-from-oil-and-gas-operations.
Accessed February 20, 2023
.
Mancini
,
S
,
Gansterer
,
M
,
Hartl
,
RF.
2021
.
The collaborative consistent vehicle routing problem with workload balance
.
European Journal of Operational Research
293
(
3
):
955
965
. DOI: https://dx.doi.org/10.1016/j.ejor.2020.12.064.
Muñoz-Villamizar
,
A
,
Montoya-Torres
,
JR
,
Vega-Mejía
,
C.
2015
.
Non-collaborative versus collaborative last-mile delivery in urban systems with stochastic demands
.
Procedia CIRP
30
:
263
268
. DOI: https://dx.doi.org/10.1016/j.procir.2015.02.147.
New Mexico Environment Department
.
2021
.
New Mexico methane strategy
[online]. Available at
https://www.env.nm.gov/new-mexico-methane-strategy/.
Accessed February 20, 2023
.
Pennsylvania Department of Environmental Protection
.
2018
.
Reducing volatile organic compounds and methane in Pennsylvania
[online]. Available at
https://www.dep.pa.gov/Business/Air/pages/methane-reduction-strategy.aspx.
Accessed February 20, 2023
.
Ravikumar
,
AP
,
Roda-Stuart
,
D
,
Liu
,
R
,
Bradley
,
A
,
Bergerson
,
J
,
Nie
,
Y
,
Zhang
,
S
,
Bi
,
X
,
Brandt
,
AR.
2020
.
Repeated leak detection and repair surveys reduce methane emissions over scale of years
.
Environmental Research Letters
15
(
3
):
034029
. DOI: https://dx.doi.org/10.1088/1748-9326/ab6ae1.
Ravikumar
,
AP
,
Sreedhara
,
S
,
Wang
,
J
,
Englander
,
J
,
Roda-Stuart
,
D
,
Bell
,
C
,
Zimmerle
,
D
,
Lyon
,
D
,
Mogstad
,
I
,
Ratner
,
B
,
Brandt
,
AR.
2019
.
Single-blind inter-comparison of methane detection technologies—Results from the Stanford/EDF Mobile Monitoring Challenge
.
Elementa: Science of the Anthropocene
7
:
37
. DOI: https://dx.doi.org/10.1525/elementa.373.
Razavi
,
S
,
Gupta
,
HV.
2015
.
What do we mean by sensitivity analysis? The need for comprehensive characterization of “global” sensitivity in Earth and Environmental Systems models
.
Water Resources Research
51
(
5
):
3070
3092
. DOI: https://dx.doi.org/10.1002/2014wr016527.
Rolison
,
JJ
,
Moutari
,
S.
2017
.
Risk-exposure density and mileage bias in crash risk for older drivers
.
American Journal of Epidemiology
187
(
1
):
53
59
. DOI: https://dx.doi.org/10.1093/aje/kwx220.
Saunois
,
M
,
Stavert
,
AR
,
Poulter
,
B
,
Bousquet
,
P
,
Canadell
,
JG
,
Jackson
,
RB
,
Raymond
,
PA
,
Dlugokencky
,
EJ
,
Houweling
,
S
,
Patra
,
PK
,
Ciais
,
P
,
Arora
,
VK
,
Bastviken
,
D
,
Bergamaschi
,
P
,
Blake
,
DR
,
Brailsford
,
G
,
Bruhwiler
,
L
,
Carlson
,
KM
,
Carrol
,
M
,
Castaldi
,
S.
2020
.
The global methane budget 2000–2017
.
Earth System Science Data
12
(
3
):
1561
1623
. DOI: https://dx.doi.org/10.5194/essd-12-1561-2020.
Schwietzke
,
S
,
Harrison
,
M
,
Lauderdale
,
T
,
Branson
,
K
,
Conley
,
S
,
George
,
FC
,
Jordan
,
D
,
Jersey
,
GR
,
Zhang
,
C
,
Mairs
,
HL
,
Pétron
,
G
,
Schnell
,
RC.
2019
.
Aerially guided leak detection and repair: A pilot field study for evaluating the potential of methane emission detection and cost-effectiveness
.
Journal of the Air & Waste Management Association
69
(
1
):
71
88
. DOI: https://dx.doi.org/10.1080/10962247.2018.1515123.
UN Environment Programme
.
2021
.
Emissions Gap Report 2021
[online].
UN Environment Programme
.
Available at
https://www.unep.org/resources/emissions-gap-report-2021.
Accessed February 20, 2023
.
U.S. Environmental Protection Agency
.
2018
.
Greenhouse gas emissions from a typical passenger vehicle
[online].
United States Environmental Protection Agency
.
Available at
https://www.epa.gov/greenvehicles/greenhouse-gas-emissions-typical-passenger-vehicle.
Accessed February 19, 2023
.
U.S. Environmental Protection Agency
.
2021
.
EPA proposes new source performance standards updates, emissions guidelines to reduce methane and other harmful pollution from the oil and natural gas industry
[online].
United States Environmental Protection Agency
.
Available at
https://www.epa.gov/controlling-air-pollution-oil-and-natural-gas-industry/epa-proposes-new-source-performance.
Accessed February 19, 2023
.

How to cite this article: Gao, M, Hugenholtz, CH, Staples, M, Barchyn, TE, Gough, TR, Vollrath, C, Xing, Z. 2023. A cooperative model to lower cost and increase the efficiency of methane leak inspections at oil and gas sites. Elementa: Science of the Anthropocene 11(1). DOI: https://doi.org/10.1525/elementa.2023.00030

Domain Editor-in-Chief: Detlev Helmig, Boulder AIR LLC, Boulder, CO, USA

Guest Editor: Arvind P. Ravikumar, Petroleum and Geosystems Engineering, The University of Texas at Austin, Austin, TX, USA

Knowledge Domain: Atmospheric Science

Part of an Elementa Special Forum: Oil and Natural Gas Development: Air Quality, Climate Science, and Policy

This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC-BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. See http://creativecommons.org/licenses/by/4.0/.

Supplementary data