Broken Arrow - Project Crested Ice: the Thule Nuclear Accident (U) (Accident 30, B-52 Thule, Greenland)

Page 1

PROJECT CRESTED ICE: THE THULE NUCLEAR ACCIDENT (u) HISTORIC & RESEARCH DIVISION HEADQUARTERS STRATEGIC AIR COMMAND

Page 2

This study is concerned with the USAF recovery and decon tamination effort precipitated by the crash of a B-52G near Thule, Greenland, on 21 January 1968. It is largely the story of Strategic Air Command's Disaster Control Team during the three-month period of Project Crested Ice. The contributions to this operation of other USAF commands 'and other governmental agencies were substantial, and it is not the intention of the historian to slight them. It must be understood, however, that this history was written from documentation either prepared by SAC or made available to it. Consequently, the history ends with the completion of SAC's decontamination work at Thule early in April. Only mention is made of the Air Force Logistics Command's activities at Thule in the summer of 1968, which included preparing the contaminants for shipment and loading them aboard ships of the Army's Military Sea Transport Service. The final report pre pared by San Antonio Air Materiel Area on its activities is included in the supporting documents. These documents are on file in the SAC History and Research Division and in the Air University Historical Archives. Also appended to this history, in Volume II, is a collec tion of photographs. They clarify dramatically the conditions under which the men of the Disaster Control Team worked.

Page 3

TABLE OF CONTENTS Part I: The Accident......................................... 1 Rescue and First Visitsto the Scene ................ 4 Part II: Recovery Operations................................ 8 The Disaster ControlTeam Staff.......... 9 Phase I: Picking Upthe Pieces...................... 15 Phase II: Removing theContamination................ 20 LIST OF SUPPORTING DOCUMENTS..................................... 51 LIST OF ILLUSTRATIONS Sketch of Crash Scene......................... 5 SAC Disaster Control Team Organizational Chart, 2 February 1968. 11 Photographs................ .................. following page 24 Organizational Chart for Phase II of Crested. Ice.......... .. . 30

Page 4

Strategic Air Command B-52s began flying airborne alert in 1958* In August 1961, the CINCSAC, General Thomas S. Power, gave the operation a new dimension when he ordered that one B-52 orbit continuously near the Thule, Greenland, site of the Ballistic Missile Early Warning System. The aircraft would be in a position to determine quickly the nature of any communications failure between the site and warning centers in the United States. On 1 January 1968, the Strategic Aerospace Wing (SAW) at Plattsburgh AFB, New York started its regular 45-day tour of monitoring the Thule BMEWS site. The Thule monitor mission actually started when the aircraft reached TI^OO'n/ST^O'W. The crew was to fly the aircraft at 35,000 feet in a "bow-tie" pattern, as depicted on the map on the following page. On 21 January 1968 crew E-l4 of the 528th Bomb Squadron, with one qualified substitute navigator and a required .third pilot oh board, flew mission Junky 14. Before takeoff the third pilot, assigned to occupy the instructor navigator’s (IN) position, placed three clothcovered foam rubber cushions under the seat and on top of the hot air spray tube. Shortly after takeoff a fourth cushion was placed under the seat, and a metal box was placed in front of and against the cushions. After about five hours of flight, the aircraft's cabin temperature became uncomfortably cool. To provide more heat, the third pilot (now occupying the co-pilot's seat) increased the cabin heat by switching to the emergency right hand inboard position of 2the cabin's heating and pressurization system. This drew hot air from the engine's manifold into the heating system. The temperature of the air bleeding off the manifold increased rapidly to about 428

Page 6

degrees, and there was very little drop in temperature between the manifold and the heating system outlets. The heat coming from the duct under the IN's seat was hot enough to ignite the flammable cushions stuffed around the duct. At 2022Z the EWO reported that he smelled burning rubber. After a search, the smoldering cushions were discovered and fire extinguishers were brought into action. These fire fighting actions were unsuccessful. No attempt was made to pull the cushions away from the hot spray bar. Flames soon blossomed from beneath the seat. The pilot reported the fire to the ground station and requested clearance to descend immediately. During the fighting of the fire, the pilot started the descent. He raised the airbrakes and increased the speed to 370 knots indicated airspeed (KIAS). The EWO opened the sextant port to exhaust the smoke from the cabin. The navigator attempted unsuccessfully to smother the first with an A-3 bag. He reported the fire as uncontrollable. Shortly after 2030Z the bomber's electrical power failed. When the aircraft com mander estimated he was close to Thule AB, he ordered the crew to bail out. The time was 2037Z. Six aircrew members ejected safely, at about 14,000 feet, although all received injuries. The seventh, the co-pilot (it will be remembered that the third pilot occupied his seat at this time), sustained fatal injuries leaving the air craft through a lower hatch. After the crew ejected, the aircraft continued on over the base, made a 180-degree turn, and at 2O39Z crashed on the ice of Wolstenholme Fjord (the point of impact is indicated on the following sketch) .* While making its last plunge the aircraft started to come apart in the air. Later search parties The location of the bomber crash has been described variously in documents as being in Wolstenholme Fjord and in North Star Bay. Since a bay is defined as "a recess or inlet in the shore of a sea or lake between two capes or headlands," and since the crash site was about seven and one-half miles out and not between the capes or headlands of North Star Bay, the crash is most accurately described as being on the ice of Wolstenholme Fjord. A fjord is described as "a long relatively narrow arm of the sea, bordered by steep cliffs."

Page 7

found bomb bay doors, wing panels, and flaps five miles north of the impact point. When it crashed the B-52 still had aboard four ' . _'J. . thermonuclear bombs with a total yield of megatons. Rescue and First Visits to the Scene The USAF organization at Thule AB was the 4683d Air Base Group, part of the Air Defense Command. Its reaction to this emer gency was prompt and effective. The first priority was of course to rescue the crew; speed was essential since the ground temperature was -24 degrees. Helicopters of Detachment l8, 54th Air Rescue and Recovery Squadron, flew nine sorties in the next 19 hours. Eight other sorties were flown by other base aircraft. Base vehicles, in cluding the taxi fleet, assisted in search operations. Fortunately, the aircraft commander had given the order to bail out as the air craft approached the base, and the crew had reacted promptly. Two of them landed among the buildings of the base. All but one came down within two and one-half miles of the base. With one exception, all survivors were rescued within two hours. The navigator, first to leave the aircraft, landed six miles south of the base. He wrapped himself in a parachute and spent an uncomfortable night on an ice flow. He was found after 21 hours by searchers from the base. The body of the copilot was found two and one-quarter statute miles from the base within eight hours.^ The Air Base Group also quickly dispatched a helicopter to the crash scene. Its vice commander. Colonel Paul D. Gopher, reported back that the bomber had crashed on the ice approximately seven and one-half miles from the Thule runway. It was still burning. By 0108z the Thule base operations officer had also gone to the scene in a helicopter. He reported that the aircraft had gone through the ice in water 150 to 600 feet deep.* He estimated the ice to be about * The main reasons for this initial and erroneous conclusion were that neither the aircraft's tail section nor any other major structural pieces were evident. Normally, the tail section survives an accident as an entity> usually whole sections of the wing can be found; and large panel sections are frequently in evidence. But in the case of Junky 14, those first at the scene reported only small bits and pieces of aircraft visible.

Page 9

six feet thick and noted that the impact and burn area had alreadyfrozen over. He had not landed because the rotor downwash of the helicopter caused the snow to blow and limited visibility. Also, the condition of the ice was not known.. During the first four hours after the crash, base personnel had found all but one of the crew survivors, visited the crash site by helicopter, and checked the base and shoreline for radiation (with negative results). .Also, a Danish official of the Greenland Trade Department, Mr. Jens Zinglersen,' set out to warn the Eskimos not to approach the crash site. Thule Air Base immediately reported the B-52 crash to the SAC Command Post. The senior controller then alerted the head quarters Disaster Control Team. It assembled and under the command of Major General R. 0. Hunziker (SAC DCS/Materiel) departed Offutt AFB in a KC-135 at 0125Z, 22 January, arriving at Thule about five hours later. After being apprised of the situation and actions taken prior to its arrival, the DCT established a command post. The Air Base Group provided office space in the Service Club. Office machines and other staff equipment came from base stocks. Telephones were installed. A separate room served as a debriefing office. An adjoining room was made available first to the Accident Investigation Board and later to the Danish scientists who arrived. While the command post was being set up, the DCT commander flew over the accident site in a helicopter. Because of darkness, no photos could be taken, but from what he saw General Hunziker con cluded that most of the bomber had gone through the ice. He saw a tire without a wheel and six engines, but there was no evidence of the empennage or a major piece of the fuselage. The scene also gave evidence-that the aircraft had exploded violently upon impact, prob ably because its JP-k fuel ignited. The only recourse was to send a reconnaissance survey team to the site by dog sled.'*''*'

Page 10

Four SAC personnel conducted the first ground recon naissance of the crash scene. They were Major D. R. Gowing and CMS and TSgt J. L. Calhoun, explosive ordnance disposal experts. Gowing and Galhoun constituted one team and Gooderham and Friend the other. The two teams were to survey the general situation and determine the extent of the damage, the amount of area affected, the existence of contamination; attempt to find the missing crew member; and to look for any classified materials. The two parties left the base together at approximately I33OZ on 22 January. They traveled by dog sled driven by Greenlanders and were guided by Mr. Zinglersen. It took about half an hour to reach the crash scene. The two parties then split up and entered the burn area. One party went north and the other south. They traversed the area burned by the crash seeking its boundaries. They met on the opposite side and crossed the burn area moving west to east. The four men spent about two hours at the crash site. Upon their return, the teams reported their findings to General JunziKer and the Disaster Control Team. It will be re called that at first it was feared that the bomber had gone through tne ice. The six engines and thousands of pieces of debris found by the recon teams, and the identification.of three and possibly four weapons chutes, gave the first evidence that in fact it had not.* * Boeing Company experts visited the crash scene on 27 January. From evidence of the wreckage and the patterns made by the debris, they concluded that the bomber was structurally intact until the impact with the ground. They believed it impacted in a left wing low, nose down attitude of approximately k-5 to 55 degrees. They concluded that Junky l4 disintegrated due to impact, explosion of the weapons and fire, and that no part of it had penetrated the ice. (Observation Report, Boeing Company, to Maj Gen Hunziker, n.d., circa 1 Feb 68).

Page 11

Radiation monitoring equipment carried by the teams gave positive readings of alpha and gamma radiation of a type consistent with plutonium contamination, but most significantly, no evidence of nuclear contamination was found.* The teams and their Greenlander sled drivers registered only light contamination when they returned. The teams determined that the ice of the fjord was strong enough to permit construction of a camp adjacent to the crash site. They did not find the missing crew member nor any classified documents. The day following the first ground reconnaissance of the crash site, two RR-^Cs flew the first photo recon sortie over the scene. SAC's U-2s did not have the ability to take usable photos at night, so Major General A. C. Gillem II, DCS/O arranged for the Tactical Air Command to send three RF-J+Cs to Thule from their station at Mountain Home AFB, Idaho. Between 23 and 26 January these air craft flew nine sorties Out of Thule. Each aircraft carried ES-72 cameras. These were activated by photo flash cartridges when the cartridges illuminated to their peak intensity (about 562 billion candlepower). Each aircraft took about 120 pictures a sortie.'*'^ • Armed with the knowledge provided by the recon teams, the SAC Disaster Control Team began planning the work ahead. It based its recovery plans on several assumptions. It was probable that the high explosives in all four nuclear weapons and the fuel on board the aircraft had detonated. The explosion had spread significant amounts of contaminated weapons and aircraft * Interrogation of the B-52 flight crew by the Eighth Air Force Aircraft Accident Investigation Board had definitely verified that the aircrew had taken no action to change any of the prescribed positions of the weapon control switches. All controls had remained in a safe configuration in accord ance with technical data. (Memo, Brig Gen A. J. Bowley,

Page 12

debris over an area about one and one-half by three miles. Recovery of the debris and decontamination of the area would require a maximum of 90 days to complete; exact levels of decontamination would of course have to await the outcome of government-to-government negotiations. Climatological conditions would probably slow down operations and place the recovery crews in some danger. Storms could not only delay the work, but also spread debris and contamination. The policies of disaster control, already set down in reg ulations, determined in large measure the priorities given to the work ahead. Major General Hunziker explained them to his staff on 2k January. First priority called for the recovery of classified documents carried aboard the bomber, specifically materials contained in the combat mis sion folder. Second priority was to find and recover the nuclear weapons or rather the pieces of them. Third was …