Inspection Risk Notifications at Oil and Gas Facilities in Colorado

How to cite

Gazze, Ludovica; Greenstone, Michael; Covert, Thomas; Rostapshova, Olga. 2021. "Inspection Risk Notifications at Oil and Gas Facilities in Colorado." AEA RCT Registry. AEARCTR-0008405. https://doi.org/10.1257/rct.8405-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Ludovica Gazze* , Michael Greenstone , Thomas Covert , Olga Rostapshova

* Primary investigator

Registry ID
AEARCTR-0008405
DOI
https://doi.org/10.1257/rct.8405-1.0
First registered
Last updated
Status
In development
Trial dates
2021-10-01 – 2025-12-31
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/8405

Abstract

In 2014, Colorado introduced regulations requiring oil and gas companies to identify and to fix methane leaks, and to install equipment to limit or prevent emissions. Yet firms and regulators might be unable to comprehensively monitor these emissions, and still rely on infrequent and expensive on-site facility inspections to detect leaks. This study plans to leverage predictive analytics and remote sensing technologies to test the effectiveness of new targeting and inspection strategies on reducing methane emissions.

Experimental design

The aim of this experiment is to determine whether notifications to oil and gas operators regarding individual facilities’ likelihood of displaying unintended methane leaks and related higher inspection risk reduce facility-level emissions levels. Facilities’ likelihood of displaying unintended methane leaks is estimated by a predictive analytics model. Operators will receive a notification from the regulator informing them that certain facilities under their purview have been flagged by the model as being at a higher risk of displaying methane emissions and are therefore subject to a higher inspection likelihood during the upcoming inspection year.

Design details

Randomization method
Randomization will be completed in an office using a computer.
Randomization unit
Randomization will be completed at the facility level.
Clusters
1,105 facilities
Observations
1,105 facilities, but we might have multiple observations per facility depending on remote sensing capabilities
Arms
553 treatment facilities and 552 control facilities
Primary outcomes
The primary outcome of the experiment is a measure of methane emissions that can reasonably be attributed to an oil and gas facility based on meteorological conditions from a remote sensing gas analyzer, as well as binary indicators of emissions found at a facility from remote sensing and regulatory inspections.

Cited by

No citation data available for this registration yet.

Keywords