LEVEL 1 ACCIDENT INVESTIGATION REPORT OF AUGUST 17, 2004 FATAL AIRCRAFT ACCIDENT ON THE GRAND COULEE-BELL NO. 6 500 kV LINE FINAL REPORT OCTOBER 1, 2004 BPA Level 1 Aircraft Accident, August 17, 2004 in Mead, WA FINAL REPORT TABLE OF CONTENTS ____________________________________________________________________________________ 1. Scope of Investigation 2. Summary 3. Facts 4. Probable Cause and Contributing Factors 5. Findings and Recommendations 6. Board Authority 7. Signatures 8. Members 9. Appendices 1) Timeline 2) Operator and Pilot History 3) Accident Photos 4) Wreckage Map 5) Cargo Hook and Suspension System 6) Crash Sequence 7) Site Map 8) BPA Rotorcraft-Load Combination Flight Manual (Part 133) 9) Definitions Page 1 of 91 BPA Level 1 Aircraft Accident, August 17, 2004 in Mead, WA FINAL REPORT SUMMARY INTRODUCTION On August 17, 2004, at approximately 0940, a Bonneville Power Administration (BPA) pilot was killed in the crash of a Bell 206BIII helicopter while stringing "sock line" to enable the subsequent stringing of new conductors and static wire on the Grand Coulee-Bell #6 500-kV line between tower 84/2 and BPA’s Bell Substation in Mead, Washington. (See Appendix 7, Site Map.) On August 17, 2004, Terry Esvelt, BPA Senior Vice President, Employee & Business Resources, appointed a Level I Accident Investigation Board (AIB) to investigate the helicopter accident in accordance with Delegation Order No. 0204-161, signed by the Secretary of Energy on October 7, 1996, and with BPA Manual Chapter 181, "Accident Investigation and Reporting". ACCIDENT DESCRIPTION On Tuesday, August 17, with the Spokane TLM Foreman III acting as overall job supervisor, a crew was present to string “sock line” onto the final two spans of the Grand Coulee #6 500-kV line into Bell Substation. Sock lines would later be used by the line crew to pull in the heavier phase conductors and overhead ground wire. A helicopter was requested so that the sock lines could be passed over nine normally energized lines that crossed under the new Grand Coulee-Bell #6 line. The helicopter would fly the sock lines to the substation dead-end bay where linemen in the structure would secure them. At 0700, the aircraft landed near tower 84/2 located west of Bell Substation. The pilot shut down the aircraft and attached a 25-foot long line with a 31-lb. ballast weight and a remote hook to the aircraft’s belly hook. The remote hook would then be connected to the sock line. After a crew tailboard meeting and helicopter equipment examination, the helicopter operation began at approximately 0850 with three conductor sock-line pulls from a triple-drum puller/tensioner in the power- payout mode, in which the line is reeled out at a speed set by the operator’s hydraulic controls. This mode was chosen to increase control and avoid making contact with normally energized lines beneath the pull. The three pulls went smoothly except for a momentary tug on the first sock line pulled that resolved itself. At approximately 0935, the pilot maneuvered the aircraft to a position above the substation dead-end bay structure to initiate the remaining static-wire sock-line pulls. The operation would now involve a different puller/tensioner, a single-drum machine with different operating characteristics. The crew had expected the pilot to land before the static-line pulls and was not prepared. A crewmember in the tower scrambled up to the goat peak to connect the sock line to the helicopter long line. Then the aircraft began to move backward in a westerly direction when the lineman in the tower noticed that the equipment operator was readjusting his radio speaker and was not ready. The lineman radioed the pilot to hold the pull. The pull continued when communication was re-established. The aircraft had traveled approximately 700 feet from the puller/tensioner when a snag occurred on the reel. The snag caused the sock line to double back onto the reel for 36 feet. This line reversal, along with the opposing motion of the helicopter, removed all slack from the sock line and the aircraft’s 25-foot long line and rocked the aircraft back on its tail. The resulting jolt may have unplugged the power source to the remote-hook release, preventing the pilot from jettisoning the sock line remotely (the pilot’s primary method of releasing the load). With the sock line taut and the aircraft now heading westward, physical evidence indicates that the aircraft’s belly hook was pulled aft and the long line electrical/solenoid release Page 2 of 91 BPA Level 1 Aircraft Accident, August 17, 2004 in Mead, WA FINAL REPORT mechanism jammed against the belly-hook attachment frame and became inoperable (the second option for releasing the load). The aircraft quickly lost altitude and collided with the de-energized Kaiser 13.8- kV power line below before crashing to the ground. Crewmembers immediately called 911, rushed to the accident scene and provided excellent emergency response. Using their training, experience and knowledge of the hazards, they responded quickly to shut off the fuel switch, extinguish a small fire, winch the helicopter upright to extricate the pilot and start two- person CPR. Unfortunately, the pilot had sustained fatal injuries. FACTS The job involved TLM crews stringing new conductor on the final two spans of the newly constructed Grand Coulee-Bell #6 500 kV line going into the Bell Substation. The rest of the 84-mile line was being constructed by a contractor as part of BPA’s infrastructure reinforcement project. The job plan for these two spans began several months before the accident. Since the job’s manpower needs exceeded the capacity of the Spokane-based Bell TLM line crew, the Foreman III supplemented the crew with employees from BPA line crews headquartered in Kalispell, Montana and Ellensburg, Washington. On May 17, 2004 the Spokane TLM Foreman III requested a helicopter from BPA Aircraft Services for the job. Sometime in mid-July the BPA Chief Helicopter Pilot decided to use the Spokane-based aircraft and pilot for this operation. August 12, 2004 Witnesses stated that the Spokane helicopter pilot met with the Bell line crew to discuss the sock-line pull. At the meeting the pilot and ground crew looked at the triple-drum puller/tensioner that would be used for the sock-line pull. The pilot discussed with the crew the direction and sequence in which he would pull the sock lines from the triple-drum puller/tensioner. During the briefing, the pilot indicated to the crew that he would be using a 50-foot long line (with a 31-lb. weighted ball, and remote cargo hook), which would be attached to the aircraft’s cargo hook. The pilot expressed concern to the crew and later to the Chief Helicopter Pilot about setting the sock line into the traveler without fly arms installed. The fly arms could be used only on the two outside conductors of tower 84/3 because the other three lines had to be passed through the structure. The crew informed the pilot that they would be in the tower anyway and could assist in placing the sock lines in the travelers. The triple-drum operator and the pilot discussed the operation of the puller/tensioners and whether to free-wheel or power pay out the sock line. The pilot had been told by the BPA Chief Helicopter Pilot that as long as the line would come off the reel at a fast walking pace, it would be adequate for the operation. The Chief Helicopter Pilot later testified that he intended the reel to be freewheeling and assumed the pilot understood. The equipment operator paid out the sock line under power from the triple-drum puller/tensioner as the pilot walked with it at a fast pace. As a result of this test, the pilot was satisfied with the speed of power payout. However, this test was not performed on the TSE puller/tensioner used for the static sock-line pull. The initial decision to power payout rather than free-wheel the puller/tensioner was based on the fact that the sock line would be passing over a number of energized power lines. It was believed that the power-payout method allowed the puller/tensioner operator to maintain better control of the sock line with the equipment available for this job. This would prevent the Page 3 of 91 BPA Level 1 Aircraft Accident, August 17, 2004 in Mead, WA FINAL REPORT sock line from drooping onto the energized lines. Based on discussions with experienced linemen and outside helicopter operators, the practice of paying out under power while pulling sock lines over energized lines with a helicopter is not uncommon in the industry. However, some operators, utilities and jurisdictions exclusively require freewheeling rather than power payout when using a helicopter. August 16, 2004 (Day Before the Accident) The pilot and TLM crewmembers attended an in-depth briefing of the work at a “tailboard” meeting. This included status of normally energized lines and equipment, Clearances and Hold Orders, electrical safety, communication channels and procedures, hardware, and technical specifications. Job assignments, work procedures and equipment locations were also discussed. Most of this information was contained in packets that were handed out. In addition, the pilot and crewmembers physically walked the job site. The BPA Foreman III in charge of the job gave the helicopter safety briefing that mainly covered ground crew safety. This included approaching the helicopter while it is on the ground, loose equipment precautions, and cautions on static discharge from the helicopter’s remote cargo hook. The pilot was present and agreed with the content of the briefing. There was no testimony to indicate that topics such as how to respond to emergency situations like sock-line hang-ups, other hazards specific to Class C loads1, or other detailed helicopter operational procedures were discussed by the pilot at the general meeting. However, individual conversations between the pilot and the equipment operators and the pilot and the linemen designated as signalmen did take place, at which time some of these issues were discussed. Testimony from many of the witnesses who had attended the August 16th tailboard meeting indicates the pilot had concerns about the job and was nervous about the sock-line pull. However, statements from other witnesses indicate that the pilot was not concerned about the work and that he expressed no nervousness about the operation. These include testimony from a contractor helicopter pilot who had been pulling sock line on the contractor-built portion of the same line, another BPA helicopter pilot with whom the pilot talked on the day before the job, and the mechanic who worked with the pilot setting up the long line and testing the emergency release on the morning of the accident. These conflicting statements and a lack of any testimony that the pilot had expressed specific concerns to any witness made it impossible for the Board to lend credence to suggestions relating to the pilot’s demeanor and attitude regarding the job. In addition, the Chief Helicopter Pilot, Aviation Safety Manager and Aircraft Services Manager testified that the pilot had not expressed any concerns to them regarding the job. 1 Class C rotorcraft-load combination means one in which the external load is jettisonable and remains in contact with land or water during the rotorcraft operation Page 4 of 91 BPA Level 1 Aircraft Accident, August 17, 2004 in Mead, WA FINAL REPORT August 17, 2004 That Tuesday, with the Spokane Transmission Line Maintenance (TLM) Foreman III acting as overall job supervisor, a combined Spokane and Ellensburg, Washington, and Kalispell, Montana TLM line crew was present to string "sock line" on the final two spans of the Grand Coulee-Bell #6 500-kV line into Bell Substation. The sock line is a small, light rope unwound from a ground-based rope puller/tensioner machine that was intended to be strung through travelers mounted on towers 84/2 and 84/3 and the Bell Substation dead-end bay structure. Sock lines would later be used by the line crew to pull in the heavier phase conductors and overhead static lines. A helicopter was requested so that the sock lines could be passed safely over nine normally energized lines that crossed under the new Grand Coulee-Bell #6 line. The first part of the work plan for the three conductor positions involved pulling sock line from the triple-drum puller situated west of tower 84/2 through 84/2 and 84/3 and terminating at the substation dead-end bay structure. Linemen would be positioned in each tower to thread the sock lines through the travelers on each tower. The helicopter would then fly the sock lines to the substation dead-end bay where linemen in the structure would secure them. The second part of the job entailed stringing sock lines for pulling the two overhead static lines from the substation bay west to tower 84/2. Each of these sock lines would be pulled from a separate single-drum puller/tensioner. One lineman in each structure would be equipped with a radio and designated as signalman for communicating with the pilot and the equipment operators. On August 17, 2004, at approximately 0655, aircraft N34698, a BPA Bell 206B III helicopter, departed Spokane International Airport (KGEG) enroute to Mead, Washington to pull the sock-line. This was a Class C rotorcraft external-load operation. The sole occupant on board the aircraft was the pilot-in- command (PIC). At 0700, the aircraft landed near tower 84/2 located west of Bell Substation. The pilot shut down the aircraft and attached a 25’ longline (a ½” steel cable, 21000 lbs. tensile strength) with a 31-lb. ballast and remote hook, to the aircraft’s belly hook. The photo at right depicts BPA’s helicopter rigging (the helicopter is not pulling sock line in depiction). The remote hook would then be connected to the sock line lead rope (a 3/4” synthetic fiber, 9450 lbs. tensile strength). The lead rope is attached to a swivel connected to the sock line (Example, TSE sock line 7/16” synthetic fiber, 16500 lbs. tensile strength). See photo at left. The pilot remained at the landing site while the line crewmembers attended a brief tailboard meeting in the Bell District “Bull Room”. The Foreman III reiterated the high points of the previous day's detailed tailboard meeting and included the direction and sequence of the sock-line pulls. After the tailboard meeting, the pilot and the Spokane heavy mobile equipment mechanic (HMEM) physically Page 5 of 91 BPA Level 1 Aircraft Accident, August 17, 2004 in Mead, WA FINAL REPORT checked the range of motion in all directions on the helicopter’s cargo hook (belly hook), with the longline attached, to ensure that it would not interfere with the electrical connection to the remote cargo hook. The pilot and a lineman then tested the remote hook to ensure that it operated electrically and mechanically, but the pilot did not wrap tape around the remote cargo hook’s electrical connection at the aircraft. The pilot and an equipment operator discussed the operation of using the triple-drum puller/tensioner with the helicopter. The triple-drum puller/tensioner was located on the west end of the job and would be used for pulling sock line into the conductor location. The BPA Aircraft Service’s technique for pulling sock line with a Bell 206B, with the pilot flying from the right seat, is to position the aircraft with the nose 15 degrees to the left of the centerline of the pull and then fly rearwards. This puts the aircraft attitude slightly nose up. A single-drum TSE puller/tensioner (see below) was positioned 258 feet east of the substation dead-end bay. In preparation for the static-line pull, a lineman and the equipment operator were preparing to remove a pin from the puller/tensioner that would allow the machine to operate in the free-wheel mode. In the free-wheel mode, the reel is disengaged from the hydraulic drive system of the puller/tensioner, allowing the sock line to be manually pulled off the reel. At this time, the designated signalman at this end of the job (east end) informed crewmembers that the pilot wanted the puller/tensioner in the power- payout mode where, the reel is connected to the hydraulic drive and the sock line is paid out by the machine’s motored drive. In this mode the speed of payout is dependent on the speed set by the operator’s hydraulic controls. The ground crew then walked the sock line off as the operator power paid it out to establish the adequacy of the payout speed. The ground crew pulled the sock line out three (3) times the distance to the substation dead-end bay, approximately 700 feet. While walking out the sock line, a ground crew member experienced a tug on the line that spun him around and pulled him back towards the TSE puller. The equipment operator told the ground crewman he had a “snag” in the line, causing it to double back and pull in onto the reel. Such a snag is not uncommon in the use of puller/tensioner equipment and when not using a helicopter is not considered a safety hazard. The operator stopped the pull and reversed the reel to the position of the snag to jerk it free. The crew then fed the sock line through the traveler in the "goat peak," or top peak, on the dead-end bay structure. At this point, they realized they had pulled out too much slack sock line and had to reel in between 100 to 150 feet. Page 6 of 91 BPA Level 1 Aircraft Accident, August 17, 2004 in Mead, WA FINAL REPORT The helicopter operation began at approximately 0850 with the three conductor sock-line pulls from the triple-drum puller in the power-payout mode. Sock lines were fed from the reels directly to the traveler in the tower without passing through the level wind, a guide sitting on the front of the tensioner, manually controlled by the operator, that allows the line to be reeled in evenly onto the reel. At the beginning of the first pull there was a momentary tug on the helicopter, prompting the pilot to ask, “What was that?” The equipment operator and the linemen in the tower determined that all was clear, so the pull continued. The rest of this pull and the next two were uneventful and went “smoothly.” Then the plan was to start with the south static position at the substation dead-end bay. The sock line would be pulled toward the West. At approximately 0935, the pilot maneuvered the aircraft to a position above the substation dead- end bay structure to initiate the static-sock-line pulls. Several crewmembers, including the Foreman III, were surprised, since they had expected the pilot to land after the first three pulls to discuss and coordinate the next two pulls, although this was not discussed at any tailboard meeting. The crew at the substation dead-end bay was still completing its work from the first three pulls and was not quite ready for the static pulls. As the helicopter hovered above the substation dead-end bay, an apprentice lineman in that structure scrambled to the goat peak. He positioned himself to connect the sock line to the helicopter’s longline. After hooking the sock line to the longline remote hook, the apprentice climbed down from the goat peak to a safe position while the helicopter hovered above the tower. At this point, the helicopter was approximately 140 feet above ground level (AGL) and 25 feet above the substation dead-end bay. The aircraft then began to move backward in a westerly direction while the lineman on the ground fed the line through his hands, holding tension until the slack on the ground, approximately 100 feet, was removed. In the meantime, the lineman in the tower noticed that the equipment operator was not ready to go because he was readjusting his radio speaker. The lineman radioed to the pilot to hold the pull. After communication was re-established between the equipment operator, the lineman in the tower and the helicopter pilot, the pull was continued. The aircraft had traveled approximately 700 feet from the puller/tensioner when a snag on the reel occurred. The snag caused the sock line to double back onto the reel. Instead of paying out toward the helicopter, even though the reel rotation was in the proper direction, the line was being pulled in. The reel operator sensed the machine lugging down and then noted the snag. He immediately attempted to shift the power lever to neutral, but he overshot the neutral detent (lever stop position) and went into reverse. At the same time he radioed, “Hold the pull.” Evidence indicates that for approximately six seconds, the reel machine had pulled about 36 feet of sock line back onto the reel. Then, when the TSE reel operator shifted the lever through neutral to reverse, the rope once again snagged and doubled back on itself, pulling in an additional 9 ft. 7 in. of sock line, before the TSE was shut down. (The Accident Investigation Board (AIB) believes the last 9 ft. 7 in. of rope was reeled in after the helicopter’s longline was cut by the Kaiser 13.8-kV overhead static line.) This was the condition in which the AIB found the equipment, see photo at left. Page 7 of 91 BPA Level 1 Aircraft Accident, August 17, 2004 in Mead, WA FINAL REPORT The aircraft was moving at the same speed that the line was paying out, (approximately 4.0 miles per hour), in rearward flight (westerly direction), when the first snag occurred and the puller/tensioner began pulling the sock line back in. The puller/tensioner was then pulling line back onto the reel at approximately 4.0 mph. Subsequent analysis indicates that it likely took almost two reverse wraps to develop enough friction to hold the line fast on the reel. Then two additional wraps were reeled in the same direction. These reverse wraps, along with the opposite motion of the helicopter, removed all the slack from the sock line and the aircraft’s 25-foot longline, causing a sudden jolt that rocked the aircraft back on its tail (see drawing to right). The longline acts as a lever along the helicopter’s vertical axis therefore the nose-up motion was due to the helicopter’s aerodynamics and not pilot input. This jolt more than likely unplugged the power source to the remote- hook release, preventing the pilot from jettisoning the sock line remotely (the pilot’s primary method of releasing the load). Witnesses reported that the aircraft then nosed down. The AIB determined through examination of the wreckage that the aircraft then turned about the mast centerline to the left. The nose of the aircraft was now heading westerly. At this point, the longline was taut against the belly of the aircraft, trailing directly aft, because the sock line was still attached to the longline as well as to the reel on the puller/tensioner. Physical evidence indicates that when the aircraft’s belly hook is pulled full aft, the manual release knob for the longline will jam against the belly-hook attachment frame rending the pilot’s electrical release inoperable (the second option for releasing the load). The aircraft was losing altitude when the left skid near the aft cross tube struck the upper wooden cross arm of the de- energized Kaiser 13.8-kV power line, see drawing next page. Physical evidence indicates that the longline was stretched across the west static wire of the Kaiser line. Page 8 of 91