BOULDER COUNTY, Colo. — A report released Thursday gives an in-depth look into the chaotic first hours of the Marshall Fire.
The report, known as a facilitated learning analysis, was requested by the Mountain View Fire Department, Louisville Fire Department and Boulder County. It was led by the Colorado Division of Fire Prevention and Control.
The analysis is intended to be an educational and training tool for first responders, helping those who were directly involved in the response understand what happened.
The report does not include the cause and origin of the fire, which is still under investigation by the Boulder County Sheriff's Office.
Two people were killed in the Dec. 30 fire, which was the most destructive in Colorado history. The fire burned more than 6,000 acres in Boulder County, destroying 1,091 structures and damaging 179. It caused more than $2 billion in losses, making it the costliest fire by far in Colorado history.
The report details heroism by first responders, some of whom went door-to-door in the fire's path to save residents.
"Thousands of emergency personnel and equipment poured in from neighboring counties, and a local Type 3 team tried to establish control during the most catastrophic fire in Colorado’s history," the report says. "Water failed in some areas, and the utility crews went to heroic lengths to keep the pressure up for firefighters. Propane tanks detonated over the shriek of the hurricane-force gusts, making the area sound like a war zone. The efforts of these responders was nothing short of extraordinary, with many working through the night and into the next day. Many local responders had homes within the fire perimeter."
The report found there were not enough radio frequencies for all of the resources deployed to fight the fire.
The planned communications setup was already being used for the so-called Middle Fork Fire, which was first reported around 10:30 a.m. that day in the foothills north of Boulder. That fire briefly threatened homes but was contained without causing any damage.
That earlier fire response meant that Marshall Fire responders had to figure out a different setup on the fly.
Crews that came in from other places had their own radio systems, and they didn't always match up.
"Rural departments came in with VHF radios. City departments often had only 800MHz radios. Law enforcement was working directly with fire resources, but many were relying on their home dispatch centers for tracking and information,” the report says.
The report found Avista Adventist Hospital was evacuated faster because an EMS chief used his home dispatch center, in Broomfield County, to request resources, rather than going through mutual aid.
The report says extreme drought was a driving factor in the fire's initial spread, noting the winds that day weren't unprecedented.
"Without the impacts of drought and multiple other environmental factors that set the stage for the fire to ignite and spread, the wind event would have been a non-factor," the report says.
However, the report found, the winds were bad enough that if the fire had started earlier or the winds had lasted longer, there would have been even more destruction.
"Fire spread was halted in urban areas by an abrupt wind shift recorded by line personnel at 0015 December 31, 2021, and by nearby weather stations as having occurred sometime between 2300 and 0100," the report says. "This wind shift brought a cooler and wetter airmass as the rotor broke down and upslope flow enveloped the fire area. Even before this wind shift, responders noted that the intensity of winds lessened, allowing them to begin more effectively suppressing the fire. If the Marshall fire had started earlier in the day, or the downslope winds had maintained through the night, the extent of damage would have been greater, as distances between structures do not appreciably increase to the north and east of the final fire perimeter."
The wind also made it hard for crews to find the fire at first, since there was no clear plume of smoke.
"[A firefighter] went to secure his helmet with the chin strap, but in the time it took to shift one hand from the top of the dome to the side, the helmet was stripped off his head and sent bouncing across the grass and into the hazy smoke," the report says. "He never found it again. Flames were moving horizontally at a rapid rate of spread and the water was ineffective. The flame front had moved past them into a garden. [The firefighter] was grateful that they had used the hose reel so that they could retrieve the hose fully charged with water rather than having to abandon it as they repositioned to chase the fire front."
The report found homes were more likely to be destroyed if they caught on fire because of indirect exposure than direct exposure. "Indirect exposure" means embers or flames from other burning buildings, while "direct exposure" is embers or flames from wildland fuels.
“Of the exposed structures, the most common outcome was no damage in directly exposed structures, occurring 46.93%, but destruction was the most prevalent outcome in indirectly exposed structures (46%)," the report found. "Indirectly exposed structures were more likely to be destroyed than directly exposed ones (46.17% indirect exposure to destruction rate vs. 41.27% direct exposure to destruction rate).”
RELATED: Marshall Fire losses top $2 billion
The report also found wooden fences played a role in the fire's spread.
“Both the Sagamore Subdivision and Subdivisions along McCaslin Boulevard were directly exposed to some intensity of wildfire, spatially oriented the same direction, and had similar spacing between structures. The Sagamore subdivision was completely destroyed, while subdivisions along McCaslin Boulevard were largely spared. Some of the difference appears to be in the distances between wildland fuels and wooden fences next to homes where they existed," the report says. "Both neighborhoods had wooden fences around numerous homes on their periphery, but homes just east of McCaslin Boulevard had a distinct advantage, as a small area (generally 30-50') of maintained lawn was next to these fences. The small embers associated with wildland grasses spotting burn out very quickly, with those landing often able to only ignite small 1-hour fuels such as cured grasses. Because of this, when combined with the effect that McCaslin Boulevard had itself on slowing, but not stopping, overall fire spread, the lower intensity and reduced spotting produced by a small area of wildland fuels next to the maintained lawns was not enough to catch wooden fences on fire, and homes in this area are largely undamaged."
SUGGESTED VIDEOS: Marshall Fire Coverage