committee compiles a written transcript of the CVR to be used during the investigation, and examples of such transcripts, edited by the NTSB investigators, are what you are reading, for the most part, in this book. FAA’s air traffic control tapes, with their associated time codes, are used to help determine the local standard time of one or more events during the accident sequence. The transcripts, containing all pertinent parts of the recording, are edited. Anything other than factual information is removed in the knowledge that the transcripts very often contain highly personal and sensitive verbal communications inside the cockpit.
Cockpit voice recorder
| Time recorded | 30 minutes continuous, 2 hours for solid-state digital units |
| Number of channels | 4 |
| Impact tolerance | 3400Gs/6.5 ms |
| Fire resistance | 1100 deg C/30 min |
| Water-pressure resistance | submerged 20,000 feet |
| Underwater locator beacon (ULB) | 37.5 KHz; battery has shelf life of six years or more, with thirty-day operation capability upon activation |
I have chosen the following twenty-one transcripts on the basis of their variety and drama. I make no apologies for what to some might seem ghoulish. In editing these transcripts for publication I have not tried to ‘characterize’ the crew members whose voices are taken directly off the CVR tapes. I do not want these transcripts to read like an airport novel. Whether the captain of the downed aircraft was kind to animals, was married with children, etc—none of this seems to me to be relevant in an accident; the same goes for the passengers whose lives are equally unknown to me. I have tried to give readers a context—of weather, time, numbers of passengers, sights and sound. I have edited some of the crews’ dialogue for clarity and I have qualified some of the pilots’ jargon with bracketed definitions that laymen better understand. I want readers to know that I am not a pilot. I have never been a pilot. I have not edited this book for pilots or for other aviation experts who will almost certainly be better served reading the original versions of these transcripts.
Finally, readers might be advised to imagine themselves, rather than sitting aboard the aeroplanes mentioned in the following pages, tuned to a radio and overhearing the sounds as they happened, and events as they unfolded. Even if you are not able to visualize everything, I know you will agree with me that these transcripts are as dramatic reading as you are likely to find, because they are minute-by-minute, unvarnished accounts of what actually occurred.
Malcolm MacPherson
COLORADO SPRINGS, Colorado, USA 3 March 1991
On 3 March 1991 a United Airlines (UAL) Boeing 737, registration number N999UA, operating as Flight 585, was on a scheduled passenger flight from Denver to Colorado Springs, Colorado. The weather in Colorado Springs was clear, with visibility 100 miles, temperature 49 degrees Fahrenheit, dew point 9 degrees Fahrenheit, winds 330 degrees at 23 knots, gusts to 33 knots, and with cumulus clouds over the mountains to the north-west. The captain was flying the aeroplane and the first officer was working the radio transmissions. The plane was scheduled to arrive in Colorado Springs at 9.46 a.m.
09:41:20 CAPTAIN: Twenty-five flaps.
09:41:23 TOWER: United 585 after landing hold short of runway 30 for departing traffic on runway…30.
09:41:25 [Sound similar to that of an engine power increase]
09:41:30 CAPTAIN: Starting on down.
09:41:31 FIRST OFFICER: We’ll hold short of [runway 30], United 585. That’s all the way to the end of our runway not…doesn’t mean a thing.
09:41:39 CAPTAIN: No problem.
09:41:51 [Sound similar to that of stabilizer trim actuation]
09:42:08 FIRST OFFICER: The marker’s identified. Now it’s really weak.
09:42:11 CAPTAIN: No problem.
09:42:29 FIRST OFFICER: [We had a] ten knot change here.
09:42:31 CAPTAIN: Yeah, I know…awful lot of power to hold that…airspeed.
09:43:01 FIRST OFFICER: Another ten knot gain.
09:43:03 CAPTAIN: Thirty flaps.
09:43:08 FIRST OFFICER: Wow.
09:43:09 [Sound similar to that of an engine power reduction]
09:43:28.2 FIRST OFFICER: We’re at a thousand feet. Oh, God [the aircraft flips over]—
09:43:33.5 CAPTAIN: Fifteen flaps.
09:43:34 FIRST OFFICER: Fifteen. Oh.
09:43:34.7 CAPTAIN: Oh! [Loud exclamation]
09:43:35.5 [Click sound similar to that of a flap lever actuation]
09:43:35.7 CAPTAIN: Fuck.
09:43:36.1 [Click sound similar to that of a flap lever actuation]
09:43:36.5 CAPTAIN: No! [Very loud]
09:43:37:4 [Click sound similar to that of a flap lever actuation]
09:43:38.4 FIRST OFFICER: Oh, my God…Oh, my God! [A scream]
09:43:40.5 CAPTAIN: Oh, no. [Exclaimed loudly]
09:43:41.5 [Sound of impact]
Numerous witnesses reported that shortly after completing its turn onto the final approach to runway 30 at Colorado Springs Airport at about 9.44 a.m., the aeroplane rolled steadily to the right and pitched nose-down until it reached a nearly vertical attitude before hitting the ground in an area known as Widefield Park. The aeroplane impacted relatively flat terrain 3.47 nautical miles south of the south end of runway 30 and .17 nautical miles to the east of the extended centreline of runway 30 at the airport. Everyone on board the flight (the two flight crew members, three flight attendants and twenty passengers) received fatal injuries. The plane was destroyed by impact forces and post-crash fire.
More than sixty witnesses were interviewed during the initial field phase of the NTSB’s investigation and more than a hundred other witnesses came forward during a follow-up visit to the accident site about a year later. The majority of the witnesses indicated that, although the aeroplane was flying at an altitude that was lower than they were accustomed to seeing, it appeared to be operating normally until it suddenly rolled to the right and descended into the ground. Many witnesses reported that the aeroplane rolled wings level momentarily (as it lined up with the runway) and that it rolled to the right until it was inverted with the nose nearly straight down.
Some of the witnesses saw the nose rise during the initiation of the right roll. One elderly couple, reportedly walking through Widefield Park at the time of the accident, stated to another witness that a liquid substance from the aeroplane fell onto their clothing which ‘smelled very bad’. Repeated efforts to find and interview this couple have been unsuccessful. One witness, who was about six miles west of the accident site, reported seeing several rotor clouds (the rotor cloud is a form of lee eddy, often associated with extreme turbulence) in the area of the accident, ten to fifteen minutes before the crash. That witness said that the rotor clouds were accompanied by thin, wispy condensation. Another person, who passed west of the accident site between 8.30 and 9.00, reported seeing ‘torn wispy clouds’ in the area of the accident.
On 8 December 1992 the NTSB issued a final report on the accident. The Safety Board concluded that it ‘could not identify conclusive evidence to explain the loss of United Airlines flight 585’. In its statement as to the probable cause of the accident, the Board indicated