This is the second of five chapters of an investigation supported generously by the Fund for Investigative Journalism and the George Polk grants for investigative reporting administered by Long Island University.

As amazing as it seems now, there was resistance to investigating Sept. 11. Twenty-two House Republicans voted against authorizing a study by the National Institute of Standards and Technology (NIST). President George W. Bush resisted the creation of the National Commission on Terrorist Attacks Upon the United States, a.k.a. the 9/11 commission. Mayor Michael Bloomberg took no questions from the 9/11 commission when it came to New York to investigate the city's response, and for a time he blocked the release of official records to the NIST investigators and the 9/11 commission, because the city was fighting an ultimately successful New York Times lawsuit to make the records public.

Meanwhile, neither the FDNY's Safety Battalion nor the National Institute for Occupational Safety and Health (NIOSH), which routinely probe firefighter deaths, investigated the World Trade Center fatalities. "At the time, I just think because of the magnitude of the event, nothing was done," says Tim Merinar, who leads NIOSH's firefighter fatality investigations. "I don't really think any organization did a true fatality investigation of the incident." (In fairness, the scale of the losses would have made such investigations, which detail the precise circumstances of each death, difficult to complete.)

Despite these obstacles, three reports—the NIST study, the 9/11 commission investigation and an FDNY-authorized report by the consultant McKinsey & Co.—managed to, very delicately, spell out Sept. 11's painful lessons for the city and its fire department.

Death in a disaster usually has multiple causes. You could talk about the Titanic and focus on the iceberg, not the lifeboats. You could discuss Hurricane Katrina and omit any mention of the levees. You could blame the Hindenburg only on static electricity. But none of these would give a full accounting of why people ended up injured or dead.

Sept. 11 would never have happened if terrorists hadn't decided to kill Americans. But it might have taken a lower toll on the fire department if problems with communications and personnel had been averted.

What was the mission?

Fighting fires in high-rise buildings is very different from battling blazes in shorter structures. Firefighters can't leave a tool in their truck and run down to get it or easily step outside for a new tank of air. Most important, they usually have to use a water system built into the high-rise building to put out the blaze. Even when that water is available, some high-rise fires are simply too large to put out. To extinguish even a flaming half-floor of the WTC would have required, by NIST's calculations, 1,250 gallons of water a minute, a deluge that might take 10 engine companies to provide.

Rescuing people in high-rise buildings is complicated too. The time it takes a person wearing at least 50 pounds of gear to climb dozens of flights of stairs could be longer than the time a civilian can survive trapped amid toxic smoke.

All these challenges were exacerbated on Sept. 11. Not only were multiple, huge floor areas in flame, the fires were also whipped by wind pouring in through the buildings' shattered sides. The standpipes in both buildings were believed to have been severed by the aircrafts' impact, meaning no water could reach the upper floors. The elevators were—with a single exception in each tower—rendered immobile, robbing firefighters of a method for reaching the fire faster.

So FDNY commanders decided early on at the World Trade Center that they would mount a rescue operation, not a firefighting one. "At best it would take hours to establish meaningful firefighting operations on the upper floors of the buildings," the NIST report found. "It was likely that many of the occupants trapped at or above the impact zone would die before help could get to them."

Despite this decision, the NIST report and individual firefighters' oral histories reveal that some fire companies were ordered to head to the impact zone and set up a post for what NIST dubbed "rescue and firefighting operations." That report found that "as the senior command level strategies were communicated to the lower levels, the concepts appeared to take hold at a slower pace at the next level down. … Some firefighters at the company level were disturbed by the operations order that signaled a change toward assisting with the evacuation. They wanted to go up and put the fire out." In the FDNY's oral histories, a number of firefighters recalled being told to prepare to extinguish the blaze, or being given vague orders to simply head upstairs.

Doing so was extremely difficult. One unit took an hour to reach the 31st floor. Many fire-fighters went 10 or 12 floors, rested, climbed five or six more, took another "blow," then scaled three or four additional flights. Several firefighters reported chest pains. Large groups of exhausted FDNY men were seen resting in a north-tower elevator lobby just before the building came down.

Despite the confusion and physical strain, firefighters doubtless saved lives by helping people evacuate. But late-arriving fire companies found few civilians left in the building

Keeping track of a tragedy

According to NIST, the fire department's personnel-tracking system "generally worked well for the first 30 minutes" but then became overwhelmed with the large number of units and personnel arriving at the scene. Sixty-one percent of the city's engine companies, 43 percent of ladder units and nearly half of all chiefs were dispatched to the World Trade Center, but all the incident commanders had to track them on was a suitcase-size "command board" that used a whiteboard and magnets to keep sense of who was being sent where.

There's a misconception that dozens of FDNY units deployed to the WTC on their own. In fact, only four did. But according to the 9-11 Commission, more fire units were dispatched to the scene than commanders asked for. Some fire trucks and engines "rode heavy," with fire-fighters who were going off duty leaping aboard to join the shift that had just come in. Several firefighters arrived on the scene as individuals, asking to help. McKinsey found that "as these units approached the area, many failed to report to the staging areas and instead proceeded directly to the tower lobbies or other parts of the incident area. As a result, senior chiefs could not accurately track the whereabouts of all units."