Air Force Bases

Titan II Accident Searcy AR, August 9 1965

On August 9, 1965, the Titan II Launch Complex 373-4, located near the town of Searcy, Arkansas, was undergoing a modification program called Project YARD FENCE. Project YARD FENCE was part of a modification program designed to facilitate maintenance and increase reliability, and improve hardness against a nearby nuclear blast. Per plan, the missile remained in the launch duct fully loaded with propellant, but with the reentry vehicle removed. There were over 50 local workmen on site. Work was in progress simultaneously throughout the silo and on the surface of the site. Hydraulic System 2, the system that operated the launch duct work platforms and silo blast valves was being flushed. A hydraulic reservoir and pump were located on the surface with pressure and return lines fed through a reopened 8-inch-diameter construction access port in the silo closure door apron. The silo closure door was closed. The silo equipment area hardening modifications required oxy-acetylene cutting torches and electric arc welding equipment on Levels 2,3,5,6, and 7 of the silo equipment area. Work within the launch duct included painting the access hatches for the silo closure T-lock wells on Level 1 and installation of steel covers on acoustic modules on and just above Level 7.

At 1309 laborer Gary Lay was standing with 12 coworkers near the emergency escape ladder that connected silo equipment area Level 2 to Level 3. Lay felt a rush of warm air on his back and turned to see flames directly behind him near the water chiller shock isolation platform. The lights went out, the emergency lighting system came on, and the group rushed for the escape ladder. Lay decided to try to go through the fire area and out to the cableway. Hurbert Sanders, a painter working in the launch duct on Level 1, smelled smoke and left the launch duct. Just as the silo lights went out, he found the access ladder to silo equipment area Level 2 and descended. There he met Lay, and they both ran down the the launch control center with smoke billowing behind them.

In the launch control center, the first indication of a problem came with the illumination of the FIRE DIESEL AREA indicator on the Launch Control Complex Facilities Console in the launch control center. As the klaxons sounded throughout the complex, Capt David Yount began the silo fire emergency checklist and ordered the evacuation of the silo area. Personnel on the surface notices smoke coming out of the hose access area. As the surface warning control lights and sirent activated, all power was lost in the complex. At 1311 Lay and Saunders, the only two survivors from within the silo equipment and launch duct areas, entered the launch control center. Lay had numerous burns on his hands and face, while Saunders was suffering smoke inhalation.

Yount called the wing command post requesting a Missile Potential Hazard Team (MPHT) be formed. At 1320 the base hospital was alerted and ambulances were requested to be send the the launch complex. At 1316 MSgt Ronald Wallace BMAT and A1C Donald Hastings, MFT, donned air packs and proceeded to Level 2 of the silo. They reported extreme heat and smoke but did not see flames and returned to the launch control center. At 1407, Col Charles Sullivan, the 308th SMW commander, requested a physical count of personnel on the site and was told that 53 workers were missing.

Firefighters and equipment were dispatched from the main base at 1411 by helicopter followed by medical vehicles. From 1440 to 1630 additional support equipment was dispatched. At 1800 the missile combat crew attempted to open the silo closure door to permit ventilation of the launch duct, but the door failed to open. At 1915 no fire was visible in the silo equipment area on Level 2, but smoke and fumes were still too dense to conduct further rescue operations. One hour later, a rescue team was able to reach Level 5. Finding no survivors and with their air supplies running out, they returned for air replenishment and proceeded to Level 8 at 2030. At 2240 SAC Deisaster Control Center reported 53 fatalities, with two hospitalized survivors. The last casualty was removed from the silo at 0530 on August 10th 1965. The 53 fatalities were distributed throughout the silo. Tweleve were on Level 2, 24 on Level 3, 8 on Level 4, 1 on Level 5, and 4 each on Levels 6 and 7.

At 1010 on August 10th, members of the Air Force Aerospace Safety Missile Accident Investigation Team arrived at the complex. Although the investigation team did not take over control of the site until 2000, team members participated in the penetration of the silo during a preliminary investigation. This fist team noted a slight soot residue distributed partially down the cableway from the launch control center to Level 2 of the silo. The silo equipment floor was covered with a film of what was later determined to be hydraulic fluid. The floor plan of the silo launch duct area is divided in to 4 quadrants, quadrant 1 and 2 meet at the cableway entrance to the silo equipment area. Proceeding past the quadrant 2 portion of the exhaust duct and into quadrant 3, the investigators found the first evidence to substantial fire damage. Burned electrical cables and fire debris was noted on the floor. There was only mild soot deposits in quadrant 4.

Down one level to Level 3, light soot was noted in quadrant 1 and 4, past the diesel generator. The floor was more heavily covered with hydraulic fluid, about 1/4-inch thick. At the far end of the generator, the aluminum partitions between Motor Control Center 1 and the generator were burned out and melted in several places. Past these partitions, in quadrant 3, massive fire damage to the Motor Control Center 1 was apparent with meter dial faces broken or melted, a wood work table heavily charred, and two areas of light concrete spalling. In the area of the pipe race, evidence of extreme heat was apparent. Hydraulic fluid was present on nearly all surfaces. Burned electrical cable insulation covered the floor. Past the Motor Control Center in quadrant 2, soot on all surfaces was the only damage noted.

The results of the preliminary investigation showed that the missile was undamaged and the Missile Engineer Technician Team, working with the Missile and Launch Complex Group Team investigators, prepared for propellant offload and removal of the missile from the site while not compromising the investigation. At 0825 a new missile combat crew reentered the launch control center. By 1258 the silo was reported clear of carbon monoxide except for a reading of 25 parts per million in the collimator room which housed the Azimuth Alignment Set used to align the guidance system and which was opposite the launch duct from the site of the fire. At 1700 missile tank venting began and was terminated at 1853. The Missile Potential Hazard condition was terminated at 2000 on August 10th.

The results form the first inspection and statements from the survivors, rescue personnel, and medical personnel responsible for locating that removing the casualties strongly pointed to an intense fire of short duration on Levels 2 and 3 in the silo equipment area. The accident investigation team split into groups to evaluate the possible major contributory subsystems of the complex and to be sure the missile, still loaded with propellant, was in a safe condition. Since complex power was still down, the launch duct air conditioning system was inoperative. The launch duct air temperature was above 70 degrees fahrenheit, the boiling point of the oxidizer onboard the missile. Thermal mass was sufficient to prevent tank rupture due to expansion from the increased temperatures, but not for very long. Oxidizer tabk pressures increased from about 10 psi to 65 psi before venting took place. A second inspection was made by the Explosive Material and Fire Pattern Group as well as by the Facility Pneudraulic Team (compressed air and hydraulic systems). The focal point of the flame pattern was found to be located at quadrant 3 on Levels 2 and 3. They discovered a fresh, unfinished weld near a ruptured hydraulic line in quadrant 3, Level 2. Inspection of the ruptered hydraulic hose revealed that the rupture had been caused by a weakening of the stainless steel braiding covering the hose due to possible contact with the electrode of an arc welding tool. The Life Sciences Group began an intensive survey of the casualties, emphasizing location and analysis of cause of death. Conclusive evidence was found by this group that clearly placed a welder in the area of suspected flame origin.

Project YARD FENCE modifications included the flushing of Hydraulic System 2, located on Level 6 of the sile. The flushing system had been operating at 500 pounds per square inch pressure with a flow of 110 gallons per minute through a pari of hoses leading from the surface hydraulic reservoir and pubp. At the time of the accident, these hoses were arrached to the Hydraulic System 2 panel on Level 2 quadrant 4. The hoses ran within 14 inches of a welding operator who was attaching a triangular stiffener plate to the existing web stiffener on a support for the Motor Control Center 1 platform. The contractor personnel locator board showed the welder to be on Level 3. The location of the weld was in an extremely awkward position that was only accessible working from Level 2, kneeling on the floor, leaning through the guardrails, and reaching around the hydraulic lines to the stiffener plate. A hardhat located on Level 2 at the welding operation site confirmed that the welder had been on Level 2.

The accidental contact of the welding rod to the hose caused the failure of the exterior metal braiding. Thus weakened, the braiding no longer prevented the interior teflon hose from rupturing, spraying, and atomizing the fluid into a mist that permeated Levels 2 and 3. The heat for the just welded fixture or the heat from the electrode touching the metal braiding was significantly higher than the 200 degree fahrenheit flash point of the fluid and served as the ignition point.

A set of oxyacetylene cutting torch tanks was located on Level 2 in quadrant 3 as well. While the valves on the tourch were closed, the tank valves were open. Both tank lines were burned through by the initial flames, contributing to the intensity of the fire. The resulting flames instantaneously consumed the oxygen on Levels 2 and 3. With the lack of oxygen, extreme levels of carbon monoxide, and toxic fumes, most of the casualties were due to either asphyxiation or the inhalation of toxins, with death occurring within two to five minutes. Workers on Level 7, 75 feet below the fire, were likely overcome by toxic combustion fumes.

The final report attributed the primary cause of the accident to be "that a welding (contract employed civilian) caused a flexible high pressure line containing flammable hydraulic fluid to rupture by accidentally striking it with a welding rod". Combined with a lighting system that was not strong enough to penetrate the think smoke from the fire and an escape system that was never meant to accommodate 53 people in an emergency. Each of the workers had been issued a face mask for use if a rocket propellant spill occurred, but these masks offered no protection from the fumes of a fire.

A number of additional causes and contributing factors were sited. Inadequate ventilation was the result of partial work completed on the silo blast valves that impeded airflow. The silo elevator did not have an independent power source. The Level 2 collimator room partition blocked access to the cableway on Level 2, forcing exit in only one direction, back through the heat of the oxyacetylene fueled fine. Safety procedures for each of the tasks were well identified, but the combination of several tasks on one level at the same time made for incompatible work safety conditions.

The report ended with 26 individual findings and changes to the launch complex facilties and Project YARD FENCE protocols.

A lack of safety disipline by the contractor contributed to the accident. The hydraulic flushing lines were haphazardly draped, welding blanks and standby fire extinguishers were lacking, cigarettes and lighters and other prohibited articles were found on the workers or within the work areas. Hand tools in the launch duct were not tied to the personnel or support beams to prevent them from falling and striking the missile.

Base firefighting personnel were not familiar with the layout of the launch complex. Critical medical information was lost when no record was made of the injuries of the first 11 casualties removed from the accident area. No record was made of the body locations within the silo by name.

On August 13th all but minor above-ground work with Project YARD FENCE at the 308th SMW was halted at the five launch complexes that had work in progress. The sites were brought back to operational status. Work did not resume until December 1st 1965. A similar work delay occurred at both the 381st and 390th SMW.

In January 1966 SSgt Robert Cunningham, A1C Joseph Rollings, A1C William Hand, and A2C Donald Trojanovich were awarded the Airmans Medal for the heroism "involving voluntary risk of life at Titan II Missile Complex 373-4." Staff Sergeant Cunningham had aided in the recovery of the victims; Airman 1st Class Rolling and Airman 2nd Class Trojanovich entered the launch duct to perform a propellant tank decompression; and Airman 1st Class Hand had entered the complex in an attempt to rescue any survivors.

On September 8th 1966, 13 months after the fire, operational control of the launch complex was returned to the 308th SMW, and on September 29th it was returned to alert status.

On August 9th, 1986, more than 400 people attended a ceremony to dedicate a memorial to the 53 workers killed 21 years earlier. The 7 foot tall granite monument, engraved with the names of the 53 victims, was paid for by private funds raised by Mrs. J Turley, whose father, Willis Bailey, was one of those killed in the fire. The monument was placed near the entrance to Little Rock Air Force Base.

Here is a newspaper interview of Gary Lay from 2000.