The Board arrived at the Savannah River Site (SRS) on April 4, 1995, and took responsibility for the Security Rappel Tower and the equipment associated with the accident. Wackenhut Services Incorporated–Savannah River Site (WSI–SRS) had posted a 24-hour guard at the entrance to the Advanced Tactical Training Academy-Rappel Tower shortly after the accident. Because rain and wind were predicted, WSI–SRS had also erected a tent and covered the Rappel Tower gates with clear-plastic sheeting to preserve the scene of the accident, as shown in
The gate lock-pin housing that was in contact with the rope at the time of the separation is shown in Figure H-4. Rope fibers can be seen next to the edge where the rope abraded prior to the separation.
Figure H-1.
The Board concurred with the actions taken and agreed that WSI–SRS should continue to maintain security at the Security Rappel Tower until it was returned to them for training operations.
Based on a review of meteorological conditions at the time of the accident, the Board determined that weather was not a factor in the accident. On the day of the accident, the sky was clear, the temperature did not exceed 70 degrees Fahrenheit, and the wind was from the north at 9 miles per hour.
The Security Rappel Tower was examined by the Board, as was all of the accident-related equipment, both on the tower and at the base of the tower.
The Security Rappel Tower was examined by the Board, as was all of the accident-related equipment, both on the tower and at the base of the tower.
When the accident occurred, the elasticity of the rope caused it to recoil from the gate at the top of the Security Rappel Tower. The rope and the area in which it came to rest following the accident are shown in Figure H-2. The gate had been opened after the accident, before this photograph was taken.
A closeup view of the rope separation at the top of the tower is shown in Figure H-3. The gate lock-pin housing that was in contact with the rope at the time of the separation is shown in Figure H-4. Rope fibers can be seen next to the edge where the rope abraded prior to the separation.
The impact area of Rappeller and the Buddy, at the base of the Security Rappel Tower, is shown in Figure H-5, as seen from the top of the tower. Another view of the impact area is shown from ground level in Figure H-6. The rope used for single-person rappels was still in place on the tower at the time this photograph was taken. Another view of the impact area and the rope in use at the time of the accident is shown in Figure H-7. A closeup of the separated rope, as it came to rest after the accident, is shown in Figure H-8.
On April 7 and 10, 1995, respectively, the Assistant Secretaries for Environmental Management (EM) and Defense Programs signed memoranda suspending all rappelling operations conducted for such purposes as training, initial qualification, requalification, certification, and competition at all EM sites. (See Appendix Q in Volume II of this report.) This suspension did not prohibit currently qualified individuals from rappelling during emergency lifesaving operations. The Board concluded its examination of the Security Rappel Tower and returned it to WSI–SRS for normal range operations, with the exception of rappelling, on April 17, 1995.
The Security Rappel Tower is approximately 40 feet high, and three levels are used for training purposes. The tower was constructed in 1983. Originally, the Security Rappel Tower had barrier chains installed at each level as fall protection. However, during a Savannah River Operations Office (SR) Safety Division occupational safety and health (OSH) inspection of the tower, the chains were judged to be inadequate to meet Occupational Safety and Health Administration (OSHA) fall-protection requirements. Correction of this deficiency was a factor in the rappelling accident.
Addressing the OSHA requirement for better fall protection in lieu of barrier chains led to the installation of new safety railings on the Security Rappel Tower.
Addressing the OSHA requirement for better fall protection in lieu of barrier chains led to the installation of new safety railings on the Security Rappel Tower. These railings were designed as gates that could be opened for training exercises. The fatal buddy rappel was conducted over these gates although they were not designed for that purpose.
Most rappel ropes are constructed using twisted, braided, or plaited nylon-yarn fibers. This process is illustrated in Figure H-9. The rope being used for the rappel training exercise at the time of the accident was a 7/16-inch diameter, "Military Green Line," twisted-nylon, mountaineering-operations rope that measured 93 feet, 4 inches in length. The rope was 120 feet long when purchased and had apparently been shortened by 26 feet, 8 inches. The shortening had no bearing on the accident, as the Board found the length of the rope was not a relevant factor.
The rope was purchased from the Defense Industrial Supply Center under Federal Stock Number 4020-00-931-8793. This is consistent with Revision 2 to WSI–SRS Standard Procedure 1-5600, Rappelling, which defines the requirements for rappel rope.
The maximum recommended load for the rope that was involved in the accident is 315 pounds (0.07 x 4,500). The actual weight of the Rappeller and the Buddy plus their gear was estimated to be 484 pounds.
When a rope is used for rappelling, a working load of 7 percent or less (a safety factor of 15:1) is recommended. The working load is defined as the percent of the tensile strength of the rope that should be used for rappelling. The maximum recommended load for the rope that was involved in the accident was 315 pounds (0.07 x 4,500). The actual weight of the Rappeller and the Buddy, plus their gear, was estimated to be 484 pounds.
Figure H-2. Top of Security Rappel Tower
Figure H-3. Closeup of Rappel Rope
Figure H-5. Impact Area as Seen from Top of Rappel Tower
Figure H-6. Impact Area as Seen from Ground Level
The acceptable minimum tensile strength of the rope, as stated in the WSI–SRS rappel procedure, was 3,840 pounds. This was determined to be inconsistent with the tensile strength of 4,500 pounds specified for the rope purchased under the Federal Stock Number referenced above. The origin of the 3,840-pound tensile strength was found to be taken from a Department of the Army Training Circular, C1, TC 90-6-1, dated 30 September 1976, entitled, Military Mountaineering. This reference states: "Nylon rope is most commonly used in climbing. The rope is 1.1cm (11mm) in diameter and is issued in 36-1/2 meter lengths. The actual separating strength when dry averages 3,840 pounds (+5 percent). The separating strength is reduced by 18 percent when the rope is wet."
Sections of the rope involved in the accident and a new "reference" rope, which was taken from inventory, were subjected to tensile testing by Westinghouse Savannah River Company to determine if the rope met the original specification tensile requirement of 4,500 pounds. Tests were also conducted to simulate the small-radius edge of the safety gate lock-pin housing that caused the rope to separate. In addition, static and dynamic analyses were conducted to determine the actual loads on the rappel rope at various positions of the Rappeller during the descent. Forces in the Rappeller's legs that were required to maintain the combined center of mass at various distances from the Security Rappel Tower wall were also calculated.
The average separating strength of the reference rope taken from inventory was 5,600 pounds. The average separating strength of the rope involved in the accident was 5,370 pounds. Both ropes exceeded the tensile strength requirement of 4,500 pounds found in the rope specification.
The average separating strength of the rope involved in the accident, when subjected to the same conditions, was 783 pounds.
The average separating strength of the reference rope over a small- radius edge simulating the safety lock-pin housing was 767 pounds. The average separating strength of the rope involved in the accident, when subjected to the same conditions, was 783 pounds.
The static rope tension force present during the rappel was the combined weight of the Rappeller and the Buddy (i.e., 484 pounds). The dynamic shock load on the accident rope, due to its sliding from the top rail of the gate to the pin housing, which is a 7.25-inch vertical drop, was calculated to be between 780 pounds and 1,150 pounds, depending on the stiffness of the rope and the position of the Rappeller relative to the tower wall. Assuming an average value of 965 pounds, the dynamic load on the rope exceeded the load capacity of the rope when in contact with the small radius of the lock-pin housing, resulting in separation of the rope.
The Board used several analytical techniques to examine the events that contributed to the fatal accident.
The Board used several analytical techniques to examine the events that contributed to the fatal accident, including change analysis, barrier analysis, fault tree analysis, and events and causal factor analysis.
The static rope tension force present during the rappel was the combined weight of the Rappeller and the Buddy (i.e., 484 pounds). The dynamic shock load on the accident rope, due to its sliding from the top rail of the gate to the pin housing, which is a 7.25-inch vertical drop, was calculated to be between 780 pounds and 1,150 pounds, depending on the stiffness of the rope and the position of the Rappeller relative to the tower wall. Assuming an average value of 965 pounds, the dynamic load on the rope exceeded the load capacity of the rope when in contact with the small radius of the lock-pin housing, resulting in separation of the rope.
The significant departures from normal operating conditions that preceded the accident are discussed in Volume 2 of this report under the section entitled Change Analysis. The performance of barriers and controls that could have prevented the accident are also discussed in Volume 2 of this report under the section entitled Barrier Analysis.
The fault tree analysis was developed using a MORT logic diagram. The MORT logic diagram is a model of the generic events, basic events, and conditions that represent failures in the management control system. See Section 3 of Volume 2 for five fault tree analyses of this accident.
The causal factor analysis presented in Table H-1 uses techniques from MORT-based root cause analysis and the Institute of Nuclear Power Operation's Good Practice OE-907, Root Cause Analysis. Causal factors are classified as either probable or root causes. This classification is used to differentiate (1) causes that, if corrected, would not by themselves prevent the accident but are important enough to be recognized as needing corrective action to improve the quality of the process; and (2) fundamental causes and associated corrective actions that, if corrected, will prevent recurrence of an event or adverse action.
The Board performed an analysis of oversight issues to determine those barriers whose failure could have been prevented by the exercise of DOE's oversight responsibility. Table H-2 presents the results of this work. Table H-3 summarizes the change analysis.
Root Causes |
Discussion |
|
Management |
Management did not specifically assign responsibility for hazard evaluation and identification to SRT supervision. As a result, management was not in control of the SWAT competition training process. Management did not provide for organizational communication between WSRC and WSI-SRS on the need for a safety review and risk analysis on the safety railings. The management control system was less than adequate. |
|
Training |
The SWAT competition training program was informal, was not based on approved lesson plans, and improperly exposed personnel to hazardous rappelling activities. The training did not use a graded approach with progression to the hazardous rappelling activities. There was no provision for risk-free failures in the training activities. |
|
Policy Implementation |
Policy and mission requirements conflict with policy implementation in the SR/WSI-SRS contract. This resulted in the WSI-SRS interpretation of rappelling as a mission requirement from the contract. Safety policy implementation was insufficient in that supervision was not specifically responsible for evaluation and identification of hazards and risks encountered in the performance of SRT training duties. Safety responsibilities are not identified in position descriptions. Safety policies did not control the training process in the SWAT competition training. |
|
Risk Assessment and Hazard Analysis |
A risk analysis was not performed during design on the new safety railings on the Rappel Tower. The hazards to rappelling activities were not identified and evaluated, and barriers and controls were not in place to protect team members. |
|
Conduct of Operations |
Conduct of operations as implemented by WSI-SRS was deficient because:
|
|
DOE Oversight |
Oversight did not reveal deficient conditions in WSI-SRS conduct of operations training and safety in rappelling operations. |
|
Probable Causes |
Discussion |
|
Procedures |
Procedures in effect at the time of the rappelling accident did not provide progression of safety responsibility, rope awareness and improperly granted an exception to rappelling on two ropes without an annual safety review. |
|
Safety |
The WSI-SRS ES&H organization did not conduct a safety inspection following the installation of the new safety railings on the Rappel Tower and did not evaluate the risks involved in the rappelling activities. WSRC did not provide a safety review of the safety rail design and did not perform a safety analysis of the interface between the safety rails and rappelling operations. |
|
Supervision |
Although supervisors are assigned responsibility for safety in procedures, the flowdown of this assigned responsibility is not contained in position descriptions of supervisors in the SRT ranks. This caused the safety railings modification not to be inspected by supervisors present at the accident scene. The rappelling procedure does not assign supervision responsibility for inspection of Rappel Tower modifications. |
|
Design |
Failure to conduct a safety review of the safety railings during design resulted in a lack of identification and evaluation of risks to tower fall protection during rappel activities. Barriers were not designed and installed on the safety railings prior to the start of the SWAT competition training. This was caused by a lack of a risk analysis and identification of the safety precedence sequence for the barriers. |
|
Communications |
The intent to train on the hazardous Buddy Rappel was not properly communicated to higher supervision nor was it properly discussed by supervision prior to the training. Information pertaining to the safety railings on the Rappel Tower was not adequately communicated to SRT personnel. |
|
Equipment |
New safety railings were installed on the Rappel Tower without an adequate safety review to identify the interface between rappelling operations, safety, and procedural requirements for inspections. Safety railings were used for rappelling, an activity for which they were not intended. The new safety railings were not inspected by WSI-SRS ES&H Division and SRT supervision. |
|
Operational Readiness |
The safety railings were not ready to be placed into service, and there was not a requirement for an operational readiness review. An operational readiness review with the user, designer, and Safety Division representative could have revealed the interface between rappelling operations and the hazardous safety railings. |
Last Modified: Friday, 28-Feb-97 10:09:00