Black Night at Black Rock

© tom Moloney

An unlikely confluence of circumstances causes a well-trained crew to perish during a rescue operation.

Operators who conduct high-consequence flights under difficult conditions have learned to construct multitiered systems of risk management: redundant equipment, sophisticated automation, multiple sources of flight-critical information, rigorous adherence to standardized procedures, and a constant cycle of review and improvement.

The various tiers vastly reduce the chance that any single error, omission, or failure will put aircraft and crew in jeopardy. Yet, it seems impossible to eliminate risk altogether. Exactly the wrong confluence of overlooked information, mistaken assumptions, and an inhospitable environment can still yield tragic consequences.

The Mission

At 9:39 pm on Mar. 13, 2017, the captain of a fishing vessel about 140 nautical miles (nm) west of the Irish coast contacted the Marine Rescue Sub Centre (MRSC) in Malin Head, Ireland, to report that a crew member had suffered the traumatic amputation of most of one thumb. After consultation with medical staff, the MRSC decided to request helicopter evacuation of the injured sailor.

Under contract with the Irish Coast Guard, Sikorsky S-92A helicopters are operated from four search-and-rescue (SAR) bases around the country: Dublin on the east coast under the designation “Rescue 116” or R116; Shannon and Sligo on the west (R115 and R118, respectively); and Waterford on the southern coast (R117). All are qualified for helicopter emergency medical services as well as SAR missions, including offshore rescues in degraded visual environments (DVE). Sligo was the base nearest to the vessel’s location, and the R118 SAR duty pilot accepted the mission.

The commander of R116 had gone home for the evening when the Dublin Marine Rescue Coordination Centre contacted her to request that her crew fly “top cover” on the rescue flight, following R118 to the scene after a delay to provide assistance in the event of difficulty with the extraction or an in-flight emergency.

While driving back to the airport, the commander called the Sligo base to coordinate plans, speaking to R118’s winch operator. He suggested they could refuel at the helipad adjacent to Blacksod Lighthouse, the westernmost fuel stop available along their route. After reaching the base, however, she commented that she expected the weather to be too low to get into Blacksod and planned to refuel at Sligo instead.

The four-person crew boarded, and 900 L (238 gal.) of fuel were added to bring the total load to 5,000 lb. R116’s multi­­­­purpose flight recorder (MPFR) registered the engine start at 10:55 pm.

The Crew

Both pilots held European Union airline transport licenses issued by the Irish Aviation Authority with S-92A type ratings. Both had qualified to fly as captains.

On this flight, the 45-year-old mission commander was the pilot flying. Since beginning her training in 1990, she’d logged 5,292 hours of flight time. She earned her instrument and multi-engine ratings in 1993 and all-weather SAR qualification the following year.

She flew the Sikorsky S-61N for 19 years, upgrading to captain in 2000. In 2013, she transferred from Waterford to Dublin and completed transition training to the S-92A. Her 825 hours in that model included 725 as pilot-in-command (PIC).

The 51-year-old copilot began learning to fly helicopters in 1996 and got his first professional job in 1999, flying the Bell 206 in western Ireland. In 2001, he earned his multi-engine and instrument ratings, S-61N type rating, and all-weather SAR qualification. He was upgraded to captain in December 2007 after assignment to the Dublin base and obtained his S-92A type rating in 2013. Of his 3,435 hours of flight experience, 795 had been flown in the S-92A, 695 of them as PIC.

Both rear crew members held dual ratings as winch operators and winchmen, having gained their initial SAR experience in military service. The 53-year-old winch operator had begun his civilian career as a senior SAR crew member in 1998, while the 38-year-old winchman entered civilian service in 2004. Like the pilots, both completed their S-92A transition training in 2013 during the operator’s fleet upgrade from the S-61N.

In addition to their scheduled flight-
competency checks, all four crew members were up-to-date on six recurrent training modules, including emergency and safety equipment, helicopter underwater egress, crew resource management, and fatigue risk management. The pilots had also completed recurrent training in controlled flight into terrain less than six months earlier.

The Aircraft

The helicopter operated as Rescue 116 was manufactured in 2007 and registered in Ireland in 2013 as EI-ICR. It was equipped with a Rockwell Collins avionics suite that included five multifunction displays (MFDs); a Universal Navigation UNS-1 flight management system (FMS) with GPS; a Honeywell Enhanced Ground Proximity Warning System (EGPWS) modified for SAR flight profiles; a weather radar, also from Honeywell; a forward-looking infrared (FLIR) camera integrated with the optical camera and searchlight systems; two Honeywell AA-300 radar altimeters; and a Euroavionics EuroNav 5 moving map display. A Zone 4 50-hour inspection had been carried out four days and three flight hours before the aircraft’s assignment to the top-cover flight.

The FMS included a multimission management system programmed with six search patterns and eight additional flight profiles (four descent/approach, three hover, and one departure). All could be modified by the pilots as the situation required.

The operator’s standard procedure was to fly the helicopter coupled, with all mode selections, changes, or adjustments called out by one pilot and confirmed by the other.

The FMS was also programmed with the operator’s proprietary low-altitude arrival routes to off-airport refueling sites and landing zones, many based on user-
defined waypoints. These weren’t charted instrument approach procedures, but lateral course guidance for use in visual meteorological conditions.

Hard-copy documentation was provided in a series of three-ring binders in which overlays of the navigation waypoints on scans of standard aeronautic charts were matched with text descriptions of the headings and distances plus comments, if any. No vertical profiles were specified, though some comments did suggest minimum altitudes for individual segments.

The aircraft’s progress along the route could be displayed on one of the MFDs against either the EuroNav moving map or the EGPWS database, and could also be superimposed on any number of marine, topographic, or aeronautic charts or even on road maps displayed on a laptop at the SAR operator’s console.

The Flight

R116 took off from Runway 16 of Dublin Airport (EIDW) at 11:02 pm, turning to a heading of 300 degrees. After handoff, Dublin Departure assigned a heading of 270. The helicopter climbed to 3,000 ft. and was handed off to the lower-north sector of the Dublin Area Control Centre (ACC). At 11:11, a member of the rear crew made radio contact with R118, about to land at Blacksod. They reported conditions at the pad as “fine … kind of some low cloud approximately 500 ft. up to the north while we were inbound through Broadhaven Bay.”

Two minutes later, R116 turned direct to Sligo; the crew then began calculating whether stopping at Blacksod instead would offer a time or fuel advantage. After double-checking computations showing that doing so would save 30 minutes and 700 lb. of fuel, they advised the Malin MRSC and Dublin ACC at 11:20 of a change in destination. Dublin handed them off to the north sector of the Shannon ACC, which confirmed the helicopter was operating under IFR, would shortly climb to 4,000 ft., and was going to Blacksod rather than Sligo.

After R116 leveled at 4,000 ft., the pilot told the crew she was entering the APBSS (approach to Blacksod from the south) routing into the FMS. Initial radio contact with the helipad elicited a report of west-southwest winds of 25 to 33 kt. and 2 nm visibility under ceilings of 300 to 500 ft.

The pilots initiated the DVE approach checklist as the helicopter crossed the Mayo coast westbound at 4,000 ft. Having confirmed that the ship was over open water and the DVE checklist was complete, the pilot switched the FMS to altitude preselect to descend to 2,400 ft. The copilot advised Shannon of their descent and was instructed to report again when airborne.

On reaching 2,400 ft., the pilot requested the APP1 approach profile, which commanded a 500-ft.-per-minute (fpm) descent to 200 ft. as measured by the radar altimeter while reducing airspeed to 90 kt. Descending through 2,000 ft., R116 crossed BLKMO, the initial waypoint on the APBSS arrival route. As Ireland’s Air Accident Investigation Unit (AAIU) noted in its report, “BLKMO was almost coincident with Black Rock.”

The pilot turned 10 degrees right to facilitate turning back onto the approach route. Descending through 700 ft., she asked the copilot to “confirm that we’re clear on radar and EGPWS.” The copilot responded, “You are clear ahead on … 10-mile range.”

After leveling off at 200 ft., the copilot turned the heading bug from 291 to 137 degrees over the course of 14 seconds to direct the aircraft back toward BLKMO. Strong southwesterly winds tightened the turn, and the pilots slowed to 75 kt. airspeed in anticipation of the tailwind. They switched the FMS back to “NAV … or Search” mode and completed the landing checklist, the pilot interrupting the final step to report visual contact with the ocean.

At 12:45:37, the copilot called out “small targets at six miles, 11 o’clock … large out there to the right.” Three seconds later, an altitude alert caused the FMS to first climb and then descend at 125 fpm as they crossed a pair of rocks that reduced radar altitude to 171 ft. The captain said, “There’s just a small little island that’s BLKMO itself.”

At 12:45:56, the winchman said he was “looking at an island directly ahead of us now, you guys; you wanna come right.” The helicopter was closing on BLKMO at a groundspeed of 90 kt. The captain asked for confirmation: “OK, come right; confirm?” and the winchman responded, “20 degrees right, yeah.”

The captain instructed the copilot to “come right, select heading … select heading.”

At 12:46:04, the copilot replied, “Roger … heading selected,” but less than one second later, the winchman urged, “Come right now … come right … COME RIGHT!”

The helicopter pitched up and rolled right, hit the western end of Black Rock (also known as Blackrock Island), and crashed into the sea.

At 1:08 am, Blacksod Lighthouse staff asked the Malin MRSC if they could determine the whereabouts of R116. At 1:13, Malin relayed a MAYDAY call to all stations. After hoisting aboard the injured sailor from the fishing vessel, R118 flew to Black Rock to initiate a search. They arrived at 2:10 and almost immediately saw strobes in the water, a life raft, and an apparent casualty the winchman was unable to recover in rough seas.

The Achill Island lifeboat reached the scene and pulled the captain from the water at 2:37. She was unresponsive and couldn’t be revived. The body of the co­pilot, still strapped into his seat, was subsequently recovered from the wreckage by naval divers. Despite extensive searches both underwater and along the shore, the remains of the two rear crewmen have not been found.

The Investigation

Several pieces of debris, including fragments of the intermediate gearbox fairing and casing, most of the horizontal stabilizer, fragments of a wheel rim, and several tail-rotor blade tips were found in the vicinity of Black Rock Lighthouse and the adjacent helipad. Floating wreckage included the center portion of the right sponson containing the right fuel tank, fragments of the shattered left sponson, the cargo hatch, main ramp, and forward-sliding cabin door.

The approximate location of the MPFR’s underwater locator beacon was determined the day after the accident. One week later, on Mar. 22, a submersible remotely operated vehicle (ROV) found the main wreckage on the ocean floor at a depth of 40 m (131 ft.) and compiled a detailed photographic and videographic survey. Irish Naval Service divers recovered the MPFR and the memory card from the aircraft’s health and usage monitoring system (HUMS). Most of the wreckage was brought to the surface in April in a series of operations by heavy-lifting vessels, during which the cockpit separated from the major portion of the fuselage. Additional ROV dives on Jul. 22 retrieved the FLIR system’s control panel, including the digital video recorder (DVR).

Despite the units’ extended immersion in salt water, data from the MPFR, HUMS, and DVR were eventually downloaded with the assistance of laboratories in the United Kingdom and the United States. The DVR proved to have been set to “Play” rather than “Record” and captured no images during the accident flight. The MPFR, however, recorded the communications and flight data that investigators used to reconstruct the accident narrative.

Exhaustive examination of the wreckage disclosed no evidence of failure of any powerplant, rotor, flight control, or avionics component before impact. Instead, the AAIU’s admirably detailed report identified a series of oversights, both organizational and individual, that allowed an expert and experienced crew to remain unaware of the hazard that lay ahead.

The operator’s arrival routes were “base-centric,” not only in the sense that crews were most familiar with the arrival routes in their usual operational areas, but also in that each base was considered responsible for defining and revising the routes it “owned.” Yet, no formal process had been established for validating or approving these routes or conducting test flights under different weather conditions using different combinations of flyover and flyby waypoints.

Neither R116 pilot was familiar with the APBSS route. While still in cruise flight at 4,000 ft., the captain remarked, “God, I’d say I haven’t been in Blacksod in about 15 years,” and the copilot agreed: “No, not recently; been awhile.”

More than half an hour later, as they briefed the winchman (responsible for monitoring the FLIR) on the approach course, the captain acknowledged that “it’s been donkey’s years since I’ve been in here.”

The operator provided the AAIU with a copy of the approach route. Black Rock wasn’t shown on the visual depiction, apparently obscured by the symbol for the BLKMO waypoint, though the height of the obstruction was noted in the comments. No minimum crossing altitude was suggested. Each pilot briefed the approach route separately, but the MPFR didn’t record either pilot mentioning the 282-ft. pinnacle at the initial fix.

The voice circuits of the MPFR did record the captain complaining that the “bloody lights in this thing drive me mad,” and the copilot agreeing, “Yeah, eh, they’re atrocious,” which the AAIU suggested hindered their reading of the comments on the approach. They selected a later waypoint for flyby or “smart turn” anticipation but kept BKMO as an overfly waypoint.

The EGPWS database didn’t include Black Rock, and the EuroNav 5 moving map ended just to the east, showing BLKMO against a blank blue background. The radar, set to the high-­sensitivity terrain-mapping mode, showed both the terrain and the aircraft’s routing in magenta, obscuring the obstruction beneath the waypoint symbol.

There is no evidence that any member of the crew either saw the beacon of the Black Rock light or expected to. The FLIR relies on temperature differences relative to the background that had diminished many hours after sunset. By the time the winchman identified Black Rock in the FLIR, the helicopter was a scant 0.3 nm away and closing at 90 kt., leaving just 12 seconds in which to avoid the collision.

The crew followed their employer’s standard procedures, which called for heading or altitude changes below 500 ft. to be requested by the pilot flying and performed by the pilot monitoring while continuing to fly coupled. The final pitch and roll inputs suggest the pilot might have responded to the winchman’s urgent warnings just a moment too late.

The Takeaway

Prudent operators protect their crews with multiple layers of equipment and procedures, minimizing the risk of single-point failures, lapses in memory, or errors in judgment. The record shows that the crew of Rescue 116 was disciplined, professional, and thorough, flying the mission by the book.

They completed the correct checklists at the appropriate times, confirmed their position and scanned for obstacles before descending, conducted a detailed landing-site briefing to determine the airspeed and altitude necessary for a successful go-around if one engine failed, and repeatedly cross-checked the fuel on board against that required to divert if they couldn’t get into Blacksod.

But as the investigation concluded, they probably believed that the design of the APBSS arrival route “would provide adequate terrain separation if the FMS was used to follow the route, and that obstacles only need be considered if going off the route.”

The AAIU’s final report includes extensive comments on organizational flaws at multiple levels, from the operator’s informal procedure for establishing and proving routes to the Irish Coast Guard’s lack of a safety management system. But most immediately, it was the highly unlikely convergence of missing data, incomplete charts, indistinguishable colors, and an obstruction noted only in comments difficult to read in the “atrocious” cockpit lighting that sent R116 flying straight and level toward a tall black rock in a pitch-black night.