Report reveals fires and explosions underreported by Colorado operators

April 2014 — A storage tank exploded approximately 1,800 feet away from Legacy Elementary School in Frederick, CO. Students and teachers sheltered in place. Luckily, no one was injured.

A new report from researchers at the Colorado School of Public Health shows that statistics on oil and gas fires and explosions are underreported because of the state’s lenient system of allowing operators to self-report incidents.

Is reporting “significant damage”’ transparent? Assessing fire and explosion risk at oil and gas operations in the United States is published in the July edition of Energy Research & Social Sciences. Dr. Lisa McKenzie and Dr. John Adgate co-authored the report along with Benjamin Blair and William Allshouse.

The report comes in the aftermath of recent incidents in Colorado’s oil & gas fields that have raised alarm over the proximity of oil and gas operations to homes.

An April 17th home explosion in Firestone that killed homeowner Mark Martinez and his brother-in-law Joey Irwin, and seriously injured his wife Erin, was caused by a methane leak from an uncapped pipeline at a nearby well owned by Anadarko.

On May 25, an explosion at a tank battery (also owned by Anadarko) near Mead, killed one worker and injured 3 others. That same day about 2 hours later a blowout occurred at a well near Peetz (owned by East Cheyenne Gas Storage, a subsidiary of Midstream Energy Holdings of Houston). No injuries were reported although 25 people were evacuated.

In response to the Firestone home explosion, Governor Hickenlooper ordered a review of all state oil and gas wells. The first phase of inspections has been completed and the second phase is expected to be completed by June 30.

While the tragedy in Firestone and the other two explosions in May were not part of the report’s analysis, researchers found that between 2006 and 2015, the COGCC database did report a total of 18 incidents that had 10 or more residences within one mile of a fire or explosion.

Colorado has a growing urban/suburban population that is increasingly residing near oil and gas development. As the report notes, “this highlights the close proximity of these incidents to homes,” and indicates the potential for serious incidents to occur in the future. Large multi-well pads are increasingly common in populated areas, which compounds the potential risks.

Colorado is the nation’s seventh-highest crude oil producer. For this new report, researchers looked at the Denver Julesburg basin along the Front Range and the Piceance basin on the West Slope, the state’s most active basins.

Researchers compared the rates of fires and explosions at oil and gas sites in Colorado and Utah, which is the eleventh-highest producer, and then cross-checked the data with the property’s proximity to occupied homes.

They found that from 2006 to 2015, there were 116 fires and explosions reported in Colorado, and 67 in Utah. While the number of reported explosions is higher in Colorado, the rate of incidents per number of active wells in Colorado (0.03 percent of active wells), is significantly lower than in Utah (0.07 percent of active wells).

Colorado’s rate of incidents per active well represents an average of one fire or explosion for every 3,690 active wells. The main causes of the recorded incidents in the COGCC database were equipment failure (20%), lightning strikes (14%), and operator error (9%). For nearly half the incidents — forty-two percent — the causes were listed as unclear, unspecified, or under investigation.

Utah requires mandatory reporting of all incidents involving fires and explosions, but Colorado requires self-reporting only of fires or explosions that have caused harm “to a member of the general public which requires medical treatment” or “significant damage to equipment or well site” according to COGCC rules.

“While the rate of fires and/or explosions were higher in Utah compared to Colorado, this is likely due to the mandatory reporting of all incidents in Utah,” said John Adgate, PhD, senior author and chair of the Department of Environmental and Occupational Health at Colorado SPH. “In Colorado, the judgment on whether significant damage or injuries to the public have occurred is left to the operator’s discretion.”

Researchers found that 40 percent of fires or explosions in Colorado required response from local fire departments, and close to two-thirds of the fires and explosions occurred in the Denver Julesburg and Piceance basins.

In 2013, Colorado changed the setback requirement for new well sites near occupied residences to 500 feet. However that means thousands of existing active wells are closer than that, as the research and recent incidents have shown.

The research shows that approximately 19 percent of people living within the Denver Julesburg basin live within one mile of an oil and gas well, including several pads that contain multiple wells.

The report also looked at the number of homes within one mile of the explosion sites, and found that many of the sites had homes within a one-mile radius. In the Denver Julesburg basin, on average there were 31 homes located within one mile of each explosion or fire site. One of the sites had 819 residences within that one-mile radius. Eighteen of the fires or explosions had more than 10 homes within the one-mile radius.

While researchers found that operators’ self-reported accident analysis was “valuable,” the report reveals:

However, the self-reporting by operators in both states regarding these incidents is also a notable limitation. For example, non- or underreporting of incidents at these sites is possible, even in the case of an emergency response. Similarly, incidents at inactive sites, such as abandoned or plugged wells, may not be reported. Previous research has also found that dishonesty and underreporting is prevalent with self-reporting of industrial accidents. In addition, the accident reports require a short written description of the incident and did not require any specifics such as cost of property damage, severity of incident, occurrence of explosions or extreme hazards, or damage to surrounding property. As a result of limited instructions for the operators to complete the accident reports, the information describing the fire or explosion were generally brief. To illustrate the under-reporting of incidents in Colorado, a lightning strike led to a tank battery fire with multiple explosions propelling the tanks into the air, caused an estimated $1.5 million dollars in damage, evacuation of homes in the area, and took many hours to control by the fire department, as reported by the news media. In contrast, the entirety of the accident description on the accident report stated “Due to a Lightening [sic] strike on Friday 4/17/15 there was a loss of the Tank Battery. An Investigation of the extent of loss is in progress. The
Facility is shut in and there were no injuries.” This highlights how the incidents are required to be reported initially, but follow up reports, which may also include more information on the cause, may have been submitted but not through a form that met our search criteria.

The report also exposes how difficult it is to obtain accident data in a timely manner from the COGCC website due to a process that requires significant computing and analysis to access the information. The authors said that gathering the data for their report was a “time consuming and costly endeavor.” They noted that Utah’s data were easily accessible.

According to the report’s authors the state’s reporting system could be improved because they do not require operators to include much detail in their reports, nor are operators required to document the actual wellhead site associated with each incident, even though some incidents occur a distance from the actual wellhead.

“Based on the incomplete information in many reports, we recommend that the minimal data required on a fire report should document the fuel source and estimated volume, cause, damage to other property, and injuries to workers and citizens,” the authors wrote.

Other information that could improve reporting includes the duration, intensity, exact location of the incident, proximity to residences or other buildings, inventory of combustible or explosive chemicals and supplies on site at the time of the incident, an estimate of economic costs, and the type and extent of emergency response.

Furthermore, the authors recommend increasing transparency of reporting and communication of these incidents. Compiling these incidents in a form that is accessible and understandable to the general public and policymakers will help with future prevention efforts.

“Based on these findings, it is our judgment that many incidents are likely not being reported to the COGCC in Colorado,” the report states. “Utah reported an incident rate that is on average 2.49 times greater than Colorado for these incidents per well and this potentially represents the magnitude of the unreported events in Colorado.”

The report concludes:

To our knowledge, this is the first state-level analysis of fires and explosions at O & G operation sites and the first comparing the occurrence of these incidents in states with different reporting requirements. Two state level datasets were compiled to compare the rate of fires and explosions at oil and gas development sites in Colorado and Utah over a 10 year period. Our judgement is that the observed differences are likely explained by the varying reporting requirements for these incidents. In Colorado, we also found areas with significant numbers of residences within 1609 m of a reported fire or explosion, which shows the potential for evacuation or direct risk from these incidents for populations living in close proximity to O & G operations. Based on these findings, we present policy recommendations, including the need for standardized and transparent reporting requirements. Given the frequent occurrence of fires and explosions at O & G operation sites, these incidents are a concern for populations living and working in close proximity to these operation sites. This study provides a baseline for understanding the reporting of these events and provides insight into the fire and explosion risks at both urban and rural O & G operation sites.

This research was completed by the Colorado School of Public Health at the University of Colorado Anschutz Medical Campus, with support from the AirWaterGas Sustainable Research Network funded by the National Science Foundation.

Colorado Public Radio interviewed the study’s author, John L. Adgate:  New Study Looks At Frequency Of Oil And Gas Explosions In Colorado

Read the report: Is reporting “significant damage” transparent? Assessing fire and explosion risk at oil and gas operations in the United States

, , , , , , , , , , , , , , , , , , , , , , ,

Get From the Styx delivered

Subscribe to our RSS feed and social profiles to receive updates.

No comments yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: