N61MDEurocopter AS 365 N2 DAUPHIN · 1993

On February 11, 2010, about 2236 Eastern Standard Time, a Eurocopter AS 365 N2, N61MD, operated by the Maryland State Police Aviation Command (MSPAC), was substantially damaged when it contacted a snow bank during operations on the ground-level helipad at Prince George's General Hospital (1MD4), Cheverly, Maryland. The commercial pilot and the pilot-rated flight paramedic were not injured. The public medical evacuation flight was being operated in night visual meteorological conditions, and no flight plan was filed.

About 2112, the helicopter departed Andrews Air Force Base (ADW), Camp Springs, Maryland, under visual flight rules (VFR), in order to retrieve and transport two patients from an automobile accident site to 1MD4. The helicopter landed at the automobile accident site, the patients were loaded, and about 2200, the helicopter departed for 1MD4. While en route to 1MD4, the pilot was advised by hospital personnel to land on the ground-level helipad, due to the presence of ice on the hospital's elevated helipad.

According to the pilot, due to noise abatement procedures and the reported winds at ADW, he approached 1MD4 from the south-southeast. He stated that when he was on final approach to the pad, he observed that it was clear of snow, but that it had snow banks around the perimeter. He also noted that a paved path was cleared of snow to allow the hospital receiving team to access the helipad from the hospital building. The pilot stated that during the final approach, he was concerned about creating "white out" conditions (blowing snow which would restrict visibility) as the helicopter neared the ground. In addition, he wanted to position the helicopter so that the clearance between the snow banks and the front of the helicopter would permit the hospital receiving team to maneuver around the helicopter.

The pilot stated that "the approach was made to a high hover with straight let down," and that it "was completed to the center of the ground pad to ensure safe clearance for the hospital personnel to safely maneuver to unload the patients." The helicopter landed about 2212, facing slightly south of west. The pilot stated that "the landing...was accomplished with no incidents detected." After landing, the flight paramedic and the receiving team offloaded and transported the patients to the hospital emergency department, while the pilot shut down the helicopter. The pilot reported that the shutdown was normal, and that after he secured the helicopter, he joined the flight paramedic in the hospital.

About 2229, when the pilot and flight paramedic returned to the helicopter, the pilot noticed that the aircraft's fenestron was resting on top of a snow bank that was approximately 3 feet high. He inspected the fenestron and did not detect any damage. According to the flight paramedic, he suggested to the pilot that it might be appropriate to remove some of the snow beneath the fenestron, in order to provide more clearance for the takeoff.

The pilot determined that "since there was no damage" and that "the tail rotor itself was clear of the snow, a straight up take off with no yaw movement would be made." The engine start was completed, with no anomalies noted. About 2236, the pilot initiated "a slow deliberate takeoff," and when the helicopter "became light on the wheels, but prior to takeoff, a vibration was detected, emanating from the rear of the aircraft." In response to the vibration, the pilot lowered the collective, shut down both engines, and applied the rotor brake.

The pilot stated that after he shut down and secured the helicopter, he and the flight paramedic visually inspected the helicopter. The pilot stated that the inspection revealed that "the fenestron was found to be severely damaged by the tail rotor blades, the right tail rotor gear box cap appeared to have been ingested by the tail rotor, and numerous tail rotor blades were damaged."

Post accident examination by MSPAC and Federal Aviation Administration (FAA) personnel revealed substantial damage to the tail rotor blades, the fenestron, the tail rotor gear box, and its mounting structure and drive coupling. Some portions of the fenestron and other components in that area were of composite construction, of light-colored material.

FAA records indicated that the pilot held a commercial pilot certificate with rotorcraft helicopter and instrument helicopter ratings. His most recent FAA second-class medical certificate was issued in June 2009. The MSPAC reported that the pilot had 2,452 total hours of flight experience, including 360 hours in the accident helicopter make and model. He had 25 hours of flight experience in the accident helicopter make and model in the 90 days prior to the accident, and 6 hours in the 30 days prior to the accident.

FAA records indicated that the flight paramedic held a private pilot certificate with multiple ratings, including rotorcraft helicopter.

FAA records indicated that the helicopter was manufactured in 1993, and registered to the MSPAC in 1995. It was equipped with two Turbomeca Ariel turboshaft engines, a single main rotor system, and a shrouded tail rotor referred to as a "fenestron." The landing gear configuration was tricycle-style, with retractable wheel-and-tire gear. The overall color of the helicopter was dark brown or black. Manufacturer's published technical data indicated that the overall length of the helicopter, from the most-forward extent of the main rotor blade to the aft end of the fuselage, was 45.1 feet. The main rotor diameter was 39.2 feet, the fuselage length was 38.2 feet, the fenestron ground clearance was 2.2 feet, and the pilot's station was situated about 32 feet forward of the fenestron. Documentation provided by the MSPAC indicated that the helicopter was within applicable weight and balance limits for both its arrival at, and attempted departure from, the hospital helipad.

The 2255 automated weather observation at ADW, located 7 miles south of 1MD4, recorded winds from 310 degrees at 9 knots, 10 miles visibility, clear skies, temperature -2 degrees C, dew point -7 degrees C, and an altimeter setting of 29.88 inches of mercury.

FAA database information for 1MD4, which consisted of facility data provided by the hospital, indicated that the hospital was equipped with two separate helipads; one at ground level, designated "H1," and one on the hospital rooftop, designated "H2." According to the FAA and operator's databases, the ground-level pad was equipped with perimeter lights and a lighted wind indicator. The databases indicated that the pad was a square which measured 67 feet on each side. The databases indicated that the rooftop pad was a square that measured 40 feet per side. Helipad elevation was listed as 297 feet above mean sea level.

The accident occurred on the ground-level pad. A north-south brick wall of the hospital structure was located about 20 feet west of the west edge of the paved pad surface. The paved surface was black in color, and was devoid of any markings except a perimeter stripe. A white stripe, approximately 3 feet wide, was painted around the entire perimeter of the pad; its outer edge was coincident with the outer edge of the pad pavement.

Notices to Airmen (NOTAMs) obtained by the pilot prior to the flight included citations of snow banks at several local airports, but did not include warnings about snow banks at 1MD4.

Hospital Surveillance Imagery

A hospital surveillance camera that viewed the lower helipad recorded the arrival and attempted departure of the helicopter. The camera was situated west of the pad, and viewed the pad looking down and to the east. The imagery was reviewed to determine the sequence and timing of events, as well as possible details regarding the nature of the accident. The local time (in hours, minutes, seconds, and decimal seconds) was displayed in the upper left of the frame. In general, the imagery corroborated the pilot's recount of events.

The pad area appeared to be well illuminated, but no in-pavement or other lighting was visible in the image frame. The approach of the helicopter was first detected at 2211:21, as indicated by some blowing snow. The blowing snow remained near ground level; the opacity and amount was limited, and did not generate a "whiteout" condition. The helicopter approached from the southwest, and rotated to a nose-west alignment (facing the camera), while it was about 10 to 15 feet above the pad. It then descended to the pad as it translated forward several feet. At 2212:02.5, just as the helicopter touched down, a white cloud of unknown origin and composition (snow and/or tail rotor/fenestron debris) appeared at the fenestron and moved quickly off to the north (helicopter right/starboard). Since the helicopter faced the camera, the exact proximal relationship of the fenestron and the snow bank could not be determined.

About 30 seconds after touchdown the unloading of the patients began, and the main rotor ceased its rotation at 2217:05. The flight paramedic was not seen again (he had apparently entered the building), and at 2217:37, the pilot exited the helicopter to enter the building.

At 2229:45, the two crew members reappeared in the image frame, and they walked directly back to the left (port) side of the fenestron. The pilot climbed partially up the snow bank; both crewmembers remained about 5 feet from the fenestron, and spent about 10 seconds looking at it. No flashlights or other auxiliary lighting was observed to be used by the crew when they were near the fenestron. At 2230:25 the pilot opened the right cockpit door to enter the helicopter. At 2232:19, the main rotor began to rotate, and at 2234:00, the flight paramedic entered the helicopter.

At 2236:26 the first motion of the helicopter associated with liftoff occurred. At 2236:40 a large whitish cloud of unknown origin and composition (snow and/or tail rotor/fenestron debris) appeared at the fenestron, and the helicopter then simultaneously rolled to the right and yawed nose left. Both angular displacements were on the order of 10 degrees. No pitch changes were apparent during the entire sequence, but the lighting and camera angle were not well-suited to enable detection of small pitch changes. Two seconds later, the helicopter was stabilized on the pad, and had ceased its motion. The pilot then shut the helicopter down, as evidenced by slowing of the main rotor, and deactivation of the helicopter lights.

MSPAC Facility Information and Flight Procedures

Well before the accident, MSPAC had developed and maintained a "Landing Zone Directory," which contained pertinent facility information intended for MSPAC flight crew personnel. The data source for the MSPAC Landing Zone Directory facility data was the FAA database. At the time of the accident, the Landing Zone Directory was produced in hardcopy. Subsequent to the accident, MSPAC began to convert the document to electronic format, in order to make it accessible via the internet.

According to a MSPAC representative, "MSPAC verifies helipad information based on pilot input. The hospitals within our system also contact SYSCOM (our dispatch center) when construction is to take place, repairs are being made, fire suppression systems are to be tested, cranes will be in the area, etc." When new information is obtained, or differences from the published data are detected, MSPAC "immediately" issues a Flight Safety Bulletin (FSB) for distribution to crews. When warranted (such as would be applicable for permanent changes), the Landing Zone Directory is updated to include the revised data.

Five days after the accident, MSPAC issued a Flight Safety Bulletin which stated "Flight Crews are to use extreme caution when operating and landing on Helipads or Landing Zones after heavy snow events, and that they "shall be prepared to avoid...High Snow Banks or Drifts."

After a fatal September 2008 accident involving one of its helicopters, MSPAC designed a mission-specific flight risk assessment tool that pilots were required to use prior to each flight. The tool was designed to classify the risk level as green (low), yellow (medium), or high (red), and calculate a percentage associated with the operational risk. High-risk flights required approval from the director of flight operations (or a designee) before a flight could be accepted. Medium-risk flights that fell near the high end of the yellow range required crews to coordinate with MSPAC SYSCOM, and were subject to monitoring and re-evaluation of conditions and risk levels. FAA Order 8900 (Change 22) "VOLUME 4 AIRCRAFT EQUIPMENT AND OPERATIONAL AUTHORIZATIONS, CHAPTER 5 AIR AMBULANCE OPERATIONS; Section 5 Operational Risk Assessment Programs for Helicopter Emergency Medical Services" was the primary resource for the MSPAC risk assessment tool. While the tool did differentiate between prepared and unprepared landing zones, it did not account for the specific conditions (such as snow banks) at prepared zones. However, existing FAA guidance (see next section) contained extensive dimensional, hazard reduction, and snow-clearing information for prepared landing zones.

Published FAA Heliport Guidance

The stated purpose of FAA Advisory Circular (AC) 150/5390-2B (issued September 2004) was to provide "recommendations for heliport design." According to the AC, adherence to the AC and its guidance was "not mandatory and does not constitute a regulation except when Federal funds are specifically dedicated for heliport construction." No records that indicated that the hospital was required to comply with the AC were discovered.

The AC defined three primary heliport areas (FATO, TLOF and Safety Area) as follows:

• Final Approach and Takeoff Area (FATO): A defined area over which the final phase of the approach to a hover, or a landing is completed and from which the takeoff is initiated.

• Touchdown and Lift-off Area (TLOF): A load bearing, generally paved area, normally centered in the FATO, on which the helicopter lands or takes off. The AC recommended that the TLOF should be a minimum of 1.5 times the overall length of the helicopter; for the accident helicopter, the minimum TLOF dimension was calculated to be 67.5 feet.

• Safety Area: A defined area on a heliport surrounding the FATO intended to reduce the risk of damage to helicopters accidentally diverging from the FATO. This area should be free of objects, other than those frangible mounted objects required for air navigation purposes

Paragraph 212 "Safety Considerations," Section (f) "Winter Operations" specified that "At least the TLOF, the FATO, and as much of the Safety Area as practical should be kept free of snow," and that the "heliport design should allow the snow to be removed sufficiently so the snow will not present an obstruction hazard to either the tail rotor or the main rotor.

Chapter 4 "Hospital Heliports," Paragraph 409 "Heliport Markers and Markings" specified the recommended surface markings. The AC stated that "Markers and/or surface markings should identify the facility as a heliport. Surface markings may be paint, reflective paint, reflective markers, or preformed material." The recommended marking was a "red H in a white cross, with a white border if required" to denote the TLOF. The H should be oriented on the axis of the preferred approach/ departure path.

Hospital Snow Removal Procedures

According to the Vice President of Support Services for the hospital, the snow removal practices remained unchanged for several years. The individual in overall charge of the snow removal activities had 42 years of experience with the hospital, and the individual who was the primary snow removal equipment operator had 24 years experience with the hospital. The principal snow removal equipment was a Kubota tractor, which was supplemented by snow blowers and hand shovels.

The hospital utilized its security personnel to monitor the snow accumulation on the helipads. Whenever security personnel determined that a pad needed to be cleared of snow, they contacted the hospital dispatcher, who then contacted the hospital facilities department, which was the department physically responsible for the clearing of the snow from the pad(s). The hospital facilities staff received regular briefings prior to the winter weather season, and snow removal for the helipads was a "high priority." The hospital did not produce or utilize any formal guidance regarding the standards for snow removal, including readily-available FAA-produced guidance.

Snow banks that resulted from the clearing of the pad, and that ranged from 2 to 5 feet high, bounded all four sides of the pad. Photographic evidence showed that the snow banks on the west, north and south sides (on both sides and forward of the helicopter, respectively) obscured about half the width of the perimeter stripes. The east perimeter stripe, which was behind the helicopter, was completely obscured by the snow bank. The snow-cleared path that connected the pad and the hospital joined the pad near the south end of the west edge.

Although compliance was not mandatory, the snow removal was not accomplished in accordance with existing FAA guidance. The TLOF was not completely cleared of snow, and the snow that was removed from the TLOF was deposited onto the FATO; the resulting snow banks constituted an obstruction hazard to helicopter operations.

The accident helicopter was the first to use the ground-level pad since it was cleared of snow the previous day. Subsequent to the accident, the hospital facilities management "instructed Staff to push the accumulated snow back toward the street" so that the snow was "no higher than 24 inches around the perimeter" of the pad. Also subsequent the accident, the Safety Management Section of the MSPAC suggested to the hospital's Security Director that the hospital "consider developing a snow and ice control plan and establish snow removal and control procedures" for each helipad, consistent with applicable FAA Advisory Circulars. The MSPAC representative also forwarded electronic versions of the FAA ACs to the hospital Security Director.

Helipad Dimensional Discrepancies

NTSB examination of post-accident photographs revealed that the fenestron was situated above a snow bank that resulted from the pad-clearing activity; the fenestron extended about 5 feet beyond the east boundary of the cleared pad and the snow. There was about 15 feet of pad forward of the nose of the helicopter, and therefore about 8 feet between the forward extent of the main rotor and the west edge of the pad, and about 28 feet between the main rotor and the hospital wall.

Based on the published helipad dimensions, the dimension of the cleared snow parallel to the helicopter's longitudinal axis was about 59 feet. Comparison of the observed clearances between the helicopter and the pad edges with the calculated clearances based using the published pad dimensions revealed an apparent dimensional discrepancy; the actual pad length appeared to be approximately 11 feet shorter than the published dimension. Examination of commercially-available aerial imagery of the helipads corroborated the apparent discrepancy. The NTSB communicated the specifics of the apparent discrepancy to MSPAC and hospital personnel. Subsequent physical measurements by hospital personnel confirmed that the published pad dimensions were in error. The FAA and MSPAC databases were subsequently modified to incorporate the corrected dimensional values.

Based on the published helipad dimensions, the dimension of the cleared snow parallel to the helicopter's longitudinal axis was about 8 feet less than the minimum TLOF dimension specified by the AC. The actual dimension of the cleared snow was nearly 20 feet less than the minimum TLOF dimension specified by the AC. If the actual pad dimension had been the same as the published value, and the helicopter had been landed the same distance from the west edge of the pad and hospital wall as previously, the fenestron would have cleared the aft snow bank by about 6 feet, instead of protruding several feet over the snow bank.