Crash after takeoff due to icing, Continental Airlines, Inc., Flight 1713, McDonnell Douglas DC-9-14, N626TX, Stapleton International Airport, Denver, Colorado, November 15, 1987 Micro-summary: This Douglas DC-9-14 crashed shortly after takeoff, due to ice contamination. Event Date: 1987-11-15 at 1415:43 MST Investigative Body: National Transportation Safety Board (NTSB), USA Investigative Body's Web Site: http://www.ntsb.gov/ Cautions: 1. Accident reports can be and sometimes are revised. Be sure to consult the investigative agency for the latest version before basing anything significant on content (e.g., thesis, research, etc). 2. Readers are advised that each report is a glimpse of events at specific points in time. While broad themes permeate the causal events leading up to crashes, and we can learn from those, the specific regulatory and technological environments can and do change. Your company's flight operations manual is the final authority as to the safe operation of your aircraft! 3. Reports may or may not represent reality. Many many non-scientific factors go into an investigation, including the magnitude of the event, the experience of the investigator, the political climate, relationship with the regulatory authority, technological and recovery capabilities, etc. It is recommended that the reader review all reports analytically. Even a "bad" report can be a very useful launching point for learning. 4. Contact us before reproducing or redistributing a report from this anthology. Individual countries have very differing views on copyright! We can advise you on the steps to follow. Aircraft Accident Reports on DVD, Copyright © 2006 by Flight Simulation Systems, LLC All rights reserved. www.fss.aero TRANSPORTATION SAFETY BOARD WASHINGTON, D.C. 20594 AIRCRAFT ACCIDENT REPORT CONTINENTAL AIRLINES, INC., FLIGHT 1713 McDONNELL DOUGLAS DC-9-14, N626TX STAPLETON INTERNATIONAL AIRPORT DENVER, COLORADO NOVEMBER 15.1987 UNITED STATES GOVERNMENT CONTENTS .......................................... EXECUTIVE SUMMARY v FACTUAL INFORMATION ........................................... History of the Flight 1 ............................................. InjuriestoPersons 4 ........................................... Damage to Airplane 4 ................................................ Other Dama ? e 4 ......................................... Personnelln ormation 4 ................................................... Thecaptain 4 ............................................... The First Officer 5 .......................................... Airplane Information 7 .................................... Meteorological Information 8 ............................................. Aids to Navigation 9 .............................................. Communications 9 ....................................... Aerodrome Information 9 ................................ Continental Deicing Procedures 12 ............... Denver Control Tower Procedures During Deicing 13 .............................................. Flight Recorders 15 ....................................... Cockpit Voice Recorder 15 ........................................... Flight Data Recorder 15 ............................. Wreckage and Impact Information 16 .......................... Medical and Pathological Information 20 Fire ......................................................... 20 .............................................. Survival Aspects 20 ......................... Interior Damage and Occupant Injuries 20 .................................... Crash/Fire/Rescue Activities 22 ............................................ Tests and Research 23 En8 i neTeardowns ............................................ 23 .................................................. Le Engine 23 ................................................. Right Engine 23 ....................... Airplane Systems Teardowns and Testing 23 .............................................. Electrical System 23 ......................................... Flight Control System 24 ....................................... Additional Information 24 ................................. Preflight Activities of the Crew 24 ................................. FAA Flow Control Into Denver 25 ..... Effects of Airframe Contamination on Airplane Performance 25 ........................................... Anti-Iceprotection 27 ............................... Wingtip Vorticesand Flight 1713 27 ANALYSIS ..................................................... General 28 ..................................... Continental DC-9 Training 29 .................. The First Officer's Initial Operating Experience 29 ........................................ Airport Snow Removal 29 ............................................. Wingtipvortices 30 ........................ Wake Vortex Factor 1: Airplane Weight 30 .............. Wake Vortex Factor 2: Vortex Generation Altitude 30 .................. Wake Vortex Factor 3: Atmospheric Instability 31 iii Preceding page blank EXECUTIVE SUMMARY On November 15, 1987, Continental Airlines, Inc., flight 1713, a McDonnell Douglas DC-9-14, N626TX, was operating as a regularly scheduled, passenger-carrying flight between Denver, Colorado, and Boise, Idaho. The airplane was cleared to take off following a delay of approximately 27minutes after deicing. The takeoff roll was uneventful, but following a rapid rotation, the airplane crashed off the right side of runway 35 left. Both pilots, 1 flight attendant, and 25 passengers sustained fatal injuries. Two flight attendants and 52 passengers survived. The National Transportation Safety Board determines that the probable cause of this accident was the captain's failure to have the airplane deiced a second time after a delay before takeoff that led to upper wing surface contamination and a loss of control during rapid takeoff rotation by the first officer. Contributing to the accident were the absence of regulatory or management controls governing operations by newly qualified flightcrew members and the confusion that existed between the flightcrew and air traffic controllers that led to the delay in departure. The safety issues discussed int his report include: 0 pilot training; aircraft deicing procedures; and wingtip vortex generation and lifespan. Recommendations concerning these issues were addressed to the Federal Aviation Administration, the National Fire Protection Association, the American Association of Airport Executives, the Airport Operators Council International, and Continental Airlines, Inc. AIRCRAFTACCIDENT REPORT CONTINENTAL AIRLINES, INC. FLIGHT 1713 McDONNELL DOUGLAS DC-9-14, N626TX STAPLETON INTERNATIONAL AIRPORT DENVER, COLORADO NOVEMBER 15,1987 1. FACTUAL INFORMATION 1.1 History of the Flight On November 15, 1987, Continental Airlines, Inc., (Continental) flight 1713, a McDonnell Douglas DC-9-14 registered in the United States as N626TX, was a regularly scheduled, but delayed, passenger-carrying flight between Denver, Colorado, and Boise, Idaho. The original departure time of 1225 was adjusted due to adverse weather conditions at Denver. The flight was to be the first of a 3day sequence of flights for the captain and first officer. Continental flight 1713 was to be the beginning of the captain's third trip sequence as a DC-9 captain and the first officer's second trip sequence as a line first officer in the DC-9. The captain commuted to Denver from San Diego, California, arriving at 1118, about 12 minutes before his scheduled "show time" of 1130. The first officer commuted to Denver from Houston, Texas, on the previous day. Between 1200 and 1230, the captain signed a dispatch flight release for flight 1713. The captain indicated on the release that he was a "high minimumsm1c aptain. The captain also asked for a weather update from the Continental weather clerk a short while later. While the crew was in the gate area of the terminal awaiting the arrival of the airplane, one of the flight attendants asked the captain who was going to make the landing at Denver on the return leg. The captain replied that he would be making the landing. The flight attendant later stated that she was concerned because she had heard that the weather was supposed to remain poor at Denver and she knew that the first officer was new to the company. At 1303, the first officer contacted clearance delivery and received a routine clearance to Boise Airport; however, the flight did not request taxi clearance from air traffic control even though their path to the deice pad crossed a designated airplane movement area. Denver Tower is not equipped with Airport Surface Detection Equipment (ASDE) and visibility was about 2,000 feet. Consequently, air traffic control was unaware that flight 1713 had taxied from its gate. The flight then proceeded tot he deice pad where it was deiced between snorkels 1 and 2 after the crew shut down the engines. 'Title 14 CFR 121.652 states, in part, that: "If the pilot in command of an airplane has not served 100 hours as pilot in command in operations under this part in the type of airplane he is operating, the MDA [minimum desent altitude] or OH [descent height] and visibility landing minimums in the certificate holder's operations specification.. .are increased by 100 feet and one-half mile (or the RvR [runway visual range] equivalent). The MDA or DH and visibility minimums need not be increased above those applicable to the airport when used as an alternate airport, but in noevent may the landing minimums be less than 300 and 1. " The operator of a deicing truck that assisted in the deicing of flight 1713 stated that the trucks had been ordered to spray the tail surfaces of every airplane going through the deice pad. He characterized some accumulations of snow on airplanes as 1 inch, but he did not specifically remember the upper surface accumulation on flight 1713. He recalled an icelslush buildup on the nose gear of the airplane, which he removed. The cockpit voice recorder (CVR) indicated that the last sound of deicing spray hit the airplane at 1346:22. At 1347:30, one of the flightcrew members stated "Blast off," and at 1349:01, a flightcrew member called "Commence start." Meanwhile, at 1346:58 Continental flight 594 called clearance delivery with a request to go to the deice pad. The clearance delivery controller told the flight to monitor ground control. At 1347:33, the ground controller told the flight to "taxi to the pad." Continental flight 594 stated that they were prevented from taxiing because they were blocked by another Continental airplane. The ground controller then told flight 594 to let him know when they could taxi, and they replied "Wilco." Several minutes later, the flightcrew stated that they were ready to taxi to the deice pad, and the controller replied "Continental 594, watch for two companies inbound to there, taxi to the north side of the runup 35 left." The flightcrew responded with "594." The crew, did not question the controller's instructions, and contrary to those instructions, taxied to the deice pad to be deiced. At 1351:12 the crew of flight 1713 contacted clearance delivery for the second time with the radio call ". . .taxi from the ice pad." The clearance delivery controller acknowledged with the radio call "Continental 1713 monitor ground twenty-one nine." The clearance delivery controller later stated when he received this transmission he thought flight 1713 was still at its gate and was asking for clearance to the deice pad. He did not note that the captain used the word "from" in his radio transmission. Several seconds later, the ground controller contacted flight 1713 with the radio call "Continental 1713 left side taxi to the pad give way to two companies on the south side of Delta goin' into there it's an Airbusand a ah MD-80." The crew initially responded with "Roger," but after some cockpit discussion about the intent of the instructions, requested clarification. The ground controller responded, "Yeah behind the Airbus. I think ah he's just got out of the alleyway now. They're goin' northbound." At 1358:51, the captain called for the taxi checklist which was accomplished shortly afterward. The flightcrew taxied the airplane from the deice pad to the ramp area near the end of runway 35L where they awaited takeoff clearance. At that juncture, the order of flights on the north side of the runup area for runway 35L was as follows: Continental flight 1617, Continental flight 65, and Continental flight 1713. A short while later, Continental flight 875 taxied in behind flight 1713. The order of aircraft on the south side of the runup area at that time was as follows: United flight 227 and TWA flight 124. At 1400:56, the local controller cleared Continental flight 1617 into position to hold. The CVR revealed that a crewmember of flight 1713 then said "We're next." This was not a radio transmission. Shortly afterward, the local controller attempted to contact Continental flight 594 to clear it onto the runway but received no response. At that time, flight 594 was still in the deice pad with engines and radios off. At 1405:14, Continental Flight 65 took off. Shortly afterward Continental 875 acknowledged a radio check from the tower controller. At 1405:53, the captain of flight 1713 then prompted the first officer to advise the tower controller that they were in the number one position on the north side. Between 1402:46 and 1404:59, the flightcrew of flight 1713 mentioned a runway visual range call of 2,200 feet that they overheard on tower frequency and briefly discussed the captain's status as a "high minimums captain." They also stated that the flaps should be set to their final setting because there was not "much slop between here and the end [of the runway]," and they mentioned that the adverse weather at Denver may remain for some time. At 1411:08 they talked about running the engines up to a high power setting every 10 minutes. No mention was made of airframe surface contamination after the completion of deicing on the CVR tape. Between 1408:23 and 141 1:08, the captain and first officer engaged in nonpertinent social conversation. At 1405:26, an arriving general aviation airplane, N706PC, reported that he was on the ground. The next arriving airplane in the landing stream was about 6 miles behind N706PC. At this point, flight 1713 was physically in the number one position and the crew was ready to take off. At 1406, the first officer of flight 1713 called the tower controller but received no response. The tower controller then inquired if Continental 875 could get around a company MD-80, referring to what he thought was Continental flight 594; flight 875 responded "Affirmative." At 1408:07, the flightcrew of flight 594 contacted ground control for clearance to taxi from the deice pad to the end of the runway. Clearance was granted, and the flight taxied into the takeoff lineup shortly thereafter. As flight 875 taxied around flight 1713 and onto the runway, the first officer on flight 1713 again called the tower and stated that they were number one for takeoff. At 1407, the controller inquired if Continental flight 594 was listening and if flight 1713 was an MD-80. Continental flight 594 again did not respond because it was not monitoring tower frequency. The crew of Continental flight 1713 then replied that they were a DC-9. The captain of flight 875 stated that about this time he observed the right side of flight 1713. He later stated that he could discern no visible contaminants on the airplane other than a 4- by 4-foot, square-shaped patch of snow or frost on the fuselage and that he based his takeoff decision on that observation. Flight 875 took off at 1412. Shortly afterward. Continental flight 1713, now correctly identified by the tower, was cleared onto the runway. At 1413, Continental flight 87.5 called the local controller and reported that "there was a little clutter on the runway." At 1414:31, flight 1713 was cleared for takeoff. The winds were reported to be from 360' at 14 knots with a runway visual range (RVR) of 2,000 feet. The captain was making the cockpit callouts and was conducting the nonflying pilot duties. At 1414:51, increasing engine sounds were recorded on the CVR. At 1415:06.7, the captain reported that the power was set at 95 and 93 [N2 engine compressor revolutions per minute in percent]. At 1415.17.1, he announced 100 knots. He called 'Vl" at 1415:28.5, "rotate" at 1415:30.9, and "positive rate [of climb]" at 1415:36.5. Less than a second later, the sounds of nosewheel rotation stopped. At 1415:39.5, the sound of a compressor surge was heard, followed by an exclamation by a crewmember and three more engine compressor surges. The sound of initial impact with the ground was recorded at 1415:43.8. The flight data recorder (FDR) recorded a maximum airspeed of about 165 knots and a maximum G load of + 1.4 during the flight. This information was recorded at 1415:39.5. The accident occurred during daylight hours at 39'46'28" North, 104'53'4!"> West. A fuel-fed flash fire ignited somewhere in the left wing area shortly after the wing began to contact the ground during the impact sequence. A "fireball" associated with the flash fire was momentarily noted inside the cabin by several passengers. After the wreckage came to rest, several small residual fires that caused minor damage to airframe components were quickly extinguished by the first fire department units to arrive on scene. The captain, the first officer, 1 of 3 flight attendants, and 25 passengers died during the accident. Two flight attendants and 52 passengers survived. 1.2 Injuries to Persons - - lniuries Crew Passenaeq Total Fatal 3 Serious 1 - MinorINone 1 Totals 5 1.3 Damage to Airplane The airplane was destroyed by impact forces and small fires following impact. The estimated value of the airplane was $4.5 million. 1.4 Other Damage None. 1.5 Personnel Information 1.5.1 The Captain The captain was hired by Continental on January 6, 1969. He held airline transport pilot certificate No. 1898373, with type ratings for the CE-500 and DC-9, along with airplane multiengine land and commercial privileges for airplane single-engine land, issued April 3, 1987. He received his initial type rating in the DC-9 on April 3, 1987, and his last proficiency check in the DC-9 simulator on October 30, 1987. On his type rating check ride of April 3, the requirement to demonstrate proficiency in recovering from approach to stalls in the takeoff, clean, and landing configurations was waived by the lead Federal Aviation Administration (FAA) flight examiner on board at the time. The approach to stall maneuvers were waived because alloted time in the simulator expired before they could be accomplished. Two extra instrument approaches were flown by the captain because one of the FAA flight examiners mismanipulated the simulator visibility control during the ride. According to Federal Aviation Regulation (FAR) 14 CFR Part 121, Appendix F, "Proficiency Check Requirements," only two out of these three approach to stall configuration maneuvers are waivable. On his proficiency check in the simulator on October 30, the captain demonstrated average performance in recovering from an approach to stall in the takeoff configuration. On this check ride, approach to stalls in the landing and clean configurations were waived by the Continental check airman. His last recurrent training was completed on October 16, 1987. His most recent FAA first class medical certificate was issued on October 8, 1987, with no limitations. He also held flight engineer certificate No 1912062 with ratings for turbojet powered airplane, issued March 3, 1969. He had accumulated approximately 12,125 total flying hours, of which 6,069 hours were pilot time including 3,111 hours as first officer in the B-727 and 133 hours as first officer in the DC-9. All of the captain's DC-9 first officer time was flown after March 13, 1987. He had a total of 33 hours as a DC-9 captain. The captain's duties with Continental began as a second officer in the 8-727 and continued from his date of hire until June 1977 when he became a first officer in that airplane. His service in the B-727 was interrupted during May through August 1973, when he served as second officer on McDonnell Douglas DC-10s. Following this interruption, he again served as first officer in the B-727 until December 1982, when he reverted to second officer status due to company furloughs. He took part in a labor strike against Continental between October 1, 1983, and July 31, 1986, when he returned to duty as a Continental second officer in the 0-727. He completed DC-9 ground school training on March 13, 1987. The captain completed the entire required Continental DC-9 training program with no notable problems. During his initial operating experience (IOE), a period of initial line flying with a Continental instructor pilot which culminated in an FAA observed check ride, he encountered one area of difficulty. The captain did not allow enough spacing between the preceding airplane on a final approach for landing and was forced to execute a missed approach. The instructor then decided to extend his IOE flying time by one trip sequence. However, the instructor did not consider this an "unsatisfactory" performance but thought that the captain would feel more comfortable after several more flight hours with an instructor. The FAA observer on board at the time stated that he considered the entire flight acceptable and would have approved the captain for line flying at that time. One more trip sequence of IOE was then performed by the captain, ending in a successful FAA observed check ride. According to his instructor, on simulator periods 3, 4, 5, and 6, the captain received specific instruction on aircraft icing protection systems and aircraft deicing during those periods. Unique deicing procedures at Stapleton also were taught during these simulator sessions. According to his seventh simulator period instructor, the captain gave a "thorough and professional' briefing on airframe and engine icing and the effects of icing on takeoffs. His instructor during IOE also recalled discussing deicing with the captain as required by Continental company policy. The Continental operations manual states "A repeat visual inspection of aircraft surfaces is required if snow or freezing precipitation is present and 20 minutes have passed since the last inspection or de-icing." The captain was scheduled to attend the company cockpit resource management program but had not done so by the time of the accident. 1.5.2 The First Officer The first officer was hired by Continental on July 20, 1987. He held airline transport pilot certificate No. 463331081, with ratings for BE-300, BE-1900, airplane multiengine land, and commercial privileges for airplane single-engine land, issued November 4, 1986. He completed his initial DC-9 training with a proficiency check on September 14, 1987. His most recent FAA first-class medical certificate was issued on June 11, 1987, with no limitations. He also held flight instructor certificate No. 463331081CF1, issued on September 8, 1986. He had accumulated approximately 3,186 total flying hours, of which 36 hours, the extent of his turbojet experience, were in the DC-9. The Safety Board's investigation of the training and performance history of the first officer examined the 5-year period before his employment by Continental. The first officer received a total of 4.8 hours of initial multiengine flight instruction before a check ride administered by an FAA-designated flight examiner on December 7, 1983. The check ride lasted 0.5 hour with no instrument time logged and one landing accomplished. On February 17, 1984, the FAA revoked the flight examiner's examination authority because the examiner had issued a flying certificate to another pilot without conducting the required flight test items. FAA records revealed that the examiner had been under investigation for "short checking" since May 1983. In March 1985, the first officer's employment as a pilot with an on-demand Part 135 commercial operator was terminated because he "failed [a] Part 135 (PIC IFR Multi) check after 30 hours of training," according to the former president of the company. The logbook of the FAA examiner who administered the checkride indicated that the first officer "failed to properly intercept 7.0 DME arc at GLS [Galveston, Texas]; went below minimums on approach] at stepdown; failed to feather SE [single-engine] (unsat)." The company's chief flight instructor stated that the first officer experienced habitual difficulties in single-engine procedures and directional control and that he made little progress in training because he repeated the same mistakes. He also stated that the first officer had a chronic problem of stepping on the wrong rudder and becoming disoriented, and he described the first officer as tense and unable to cope with deviations from the routine. He recalled that the first officer had failed the checkride on three occasions before his employment was terminated. On March 25, 1985, the first officer was hired by a Part 135 regional airline as a first officer in Beech BE-1900 scheduled commuter operations. During upgrade training in May 1986, instructor comments in company training records documented that the first officer was "weak on memory items on V1 cut, had to miss the first two ILS's" and "approaches are going to need a lot of work, became disoriented in holding, concentrate exclusively on inst. approaches and procedures, continue training." On May 31, 1986, he failed to successfully complete an Airline Transport PiloVBE-1900 type rating flight examination administered by an FAA examiner. On that flight, he passed a designated holding fix at cruise speed before realizing his mistake and also did not perform an ILS approach to specified tolerances, according to the examiner. The first officer underwent additional training and on June 6,1986, he successfully completed that check ride and was upgraded to captain. The FAA requires air carriers to conduct security checks of pilot applicants before employment because they have unescorted access to airport security areas. These checks must include, at a minimum, reference and prior employment histories for verification of employment during the preceding 5 years. There is no requirement to verify previous flight experience or to determine an applicant's FAA accidenvincident history or enforcement history, previous employer's pilot training and performance records, and criminal and driver histories. Although employment verification for the preceding 5-year period is mandated, commercial operators are not required to maintain pilot records for that length of time. The corporate security section at Continental commissioned a background check on the first officer through a private company. According to this background check, which was dated July 28, 1987, the first officer was employed by an on-demand Part 135 commercial operator between February 1984 and March 1985. This operator was the only previous employer mentioned in the report. In answer to the question in the report "Did the subject leave on his or her own accord?" the answer was "Yes." In answer to the question "Would the subject be eligible for rehire?' the answer was "Yes." Lastly, according to the background check, the quality of the first officer's work was described as "Very good." The first officer completed Continental's DC-9 ground school on August 11, 1987, and two cockpit procedures trainer (CPT) periods on August 26, 1987. As part of his ground school training, he attended the company's 2-day cockpit resource management program. He then entered into a series of instructed visual simulator periods. Following the second period, the instructor's written comments were: "SCAN! Need to review (procedures) and profiles." On the third period he was described as ". . . Better, but scan still needs work, and a little jerky on flight controls." After the fourth period, the instructor commented, "Scan still needs work. Pitch control jerky, altitude control when pressure is on is somewhat sloppy. Knowledge of (maneuvers) is good." On the fifth period, administered on September 2, the comments indicated general improvement; "Scan is better. Still a bit jerky on pitch! [the first officer] seems to have caught up with airplane today." The comments from his sixth period on September 8, with a different instructor indicated "Scan is a real problem, completely lost control of airplane with engine out and at 2,000! Went into 45'-60' angle of bank, lost 1,500! Had to be arrested by (instructor). Altitude and airspeed control generally way out of limits.. Some basic procedures still require review." The first officer's unsatisfactory progress in the sixth period necessitated a repeat of the simulator session which was accomplished with the same instructor and only one student. The first officer received 3 hours of training in the available 4-hour block. The instructor did not grade the flight "normal progress," but made the following comments: 1. Needs to review limitations and profiles. 2. Falls behind in planning, also not sure of what to do next--may lack experience.