How The Globe analyzed thousands of inactive well sites

There are two main stages in the life of any oil or gas well: the active, production stage and the inactive stage.

“Inactive” actually seems like a misnomer: Once a well becomes inactive, its oil and gas will no longer flow, but the well itself may still move on companies’ balance sheets as it is traded, purchased or sold.

Ideally, such wells are supposed to be cleaned up and plugged. That process makes them environmentally safe and returns the land around the well to its original state.

Faced with the costs of cleaning up these post-production wells, some companies have packaged them together with low-production active wells, shuffling them off their books and onto those of smaller producers, many of which have folded since the economic downturn in the oil and gas industry.

To puzzle out the liability hustle in Canada’s oil patch, The Globe and Mail analyzed records for more than 600,000 wells reported to three provincial energy regulators: the Alberta Energy Regulator (AER), the BC Oil and Gas Commission (BCOGC) and Saskatchewan’s Ministry of Energy and Resources. We created a database of wells that was pulled together from a combination of public provincial sources and freedom of information requests. The database includes information such as well status, fluid type and the facility and operator for each well, all of which formed a basis for our analysis. Most of the information was provided at no cost, though The Globe did pay the Saskatchewan Ministry of Energy and Resources $630 for its data sets.

These data sets are updated at different rates. For example, the raw list of wells is a daily report, while supplementary data sets such as the liability management rating (LMR) report, which is used to measure a company’s asset-to-liability ratio, are reported monthly or quarterly. To address the inconsistencies caused by the different publishing periods, we sought out reports in which the published dates lined up as closely as possible across the different data sets. Finally, the provincial data were “cleaned up” to line up the varying well status classifications and dates the wells became inactive.

To determine the number and status of wells in each province, we first had to define the life cycle of a well. In addition to the two main stages outlined above, there are also the plugged and reclaimed stages, when a well is sealed and then cleaned up. We also developed a separate classification for orphan wells, which lack a fiscally viable owner due to insolvency and are thus considered abandoned.

We excluded all the wells that had been cancelled, were classified as confidential or were still in the pre-drilling stage. Ultimately, our database logged 615,855 wells spanning Alberta, British Columbia and Saskatchewan, of which 122,456 were inactive. The overall wells count includes all orphan assets.

We focused much of our analysis on inactive wells because they can remain idle indefinitely in Western Canada. This creates increased environmental and financial risk, as companies often delay the cleanup of these wells.

While the AER maintains a separate list of inactive wells with a singular inactive date for each well, the other two provincial regulators track multiple inactivity dates. For each inactive well in B.C. and Saskatchewan, we used the earliest available date after the end of the well’s production cycle. As a well’s status can flip between active and inactive multiple times throughout its lifespan, the inactive date captured in our database only represented the last time a well changed to inactive status.

All of the analysis described above concerns the current status of the three provinces' wells.

By using the AER’s facility-operator history data set, we were able to unlock a well’s past. A well facility is all of the equipment that exists between a well and the pipelines or trucks that make up the resources’ transportation system. We used this historical information on well-facility operators to link a well and all the companies that have ever operated it. Though being the operator of a well’s facility is not the same as full ownership, we regarded it as a strong indicator of the connection between a well and a company; in fact, most of the time, an operator was also that well’s owner.

By using the start and end dates of these facility-company linkages, we were able to determine the operator of a well in any year between 1994 and 2018 and, as a result, any changes in operators. We applied a similar method when tracking the operator changes in Saskatchewan and British Columbia, with a well-owner linkage list provided by BCOGC. The three provinces’ combined result allowed us to map out the transactions and the players: the sellers, the buyers and the number of wells that have changed hands.

Our method doesn’t capture all transactions, unfortunately. Some wells can’t be linked to facilities, often because they aren’t in production stage and thus won’t reveal themselves in our analysis. In other words, our analysis of wells that have changed hands since 2015 errs on the conservative side.

The data sets we had did not show cumulative counts of inactive wells for previous years. We requested those numbers directly from the regulators. Saskatchewan was not able to provide the data before our publication deadline.

To measure well liabilities, we used a methodology developed by the C.D. Howe Institute to examine Alberta’s wells and expanded it to measure the combined cleanup costs for all three provinces, breaking out the estimates into low-cost and high-cost scenarios. These estimates are based on the number of wells that could potentially become a liability multiplied by the potential cost to clean up each one. We included the current orphan wells and wells owned by companies with an LMR less than or equal to 2. For the cleanup cost estimate, we used $80,000 and $20,000 respectively for the cost of plugging and reclamation in the low-cost scenario, then doubled those for the high-cost scenario. We subtracted a total of $487-million from the cleanup costs, which, according to the most recent LMR reports, was the overall security deposit collected to offset potential suspension, abandonment, remediation and reclamation costs. In the end, our cleanup estimates range between $13.2-billion and $26.9-billion.

The Globe arrived at its findings by using a programming language and set of statistical tools that allowed for analyses to be tweaked and rerun continuously during the investigation. All of the code developed for the analysis was verified and fact-checked by data journalists not previously involved with the project.