“Engineering” the Collapse: Making the Structure Safe

“Engineering” the Collapse: Making the Structure Safe

BY DAVID J. HAMMOND

Each FEMA USAR task force has six specially trained civil/structural engineers, two of whom deploy with the team in any six-hour period. These volunteer engineers, called “structural specialists (S. Spec.)” are required to attend a one-week special course developed by the Army Corps of Engineers in addition to normal USAR training. The objective of the Corps course is to familiarize the S. Spec. with the disaster environment. Among the information presented in the course is that USAR most often occurs in multistoried, fully or partially collapsed, very dangerous structures that contain heavy debris and have a high potential for additional collapse. USAR teams with specialists trained in search, rescue, and medical care must work in this dangerous environment when the recovery of live victims is probable. Engineers trained and experienced in evaluating damaged buildings can help reduce (or at least better define) the risk to these teams and victims. To function effectively, these engineers also must be well prepared to make difficult value judgments in an environment that is very different from the orderly design office. In contrast, the team`s search, rescue, and medical members are asked to make rapid, high-pressure decisions as a normal part of their occupations.

As with most engineering problems, the first job is to identify the problem. Some of the basic questions to be addressed are the following:

What types of structures are involved?

What hazards are present–falling objects, collapse, other?

What are the locations and conditions of the remaining voids in the structure?

Where are the previous access openings in the structure?

What tools and shoring/stabilizing methods are available?

What are the needs of the search, rescue, and medical functions?

The S. Spec. is trained to use this information to do the following:

perform structure triage (quick prioritization) if many buildings are involved;

provide detailed hazard assessment, including drawings, of individual structures;

work with search and rescue leadership to provide alternate strategies to reduce risk; and

design and help provide mitigation measures.

It is relatively easy for engineers to recognize and report on the hazards; it is much more difficult to devise creative, immediate ways to mitigate them. The goal is to have each task force staffed with an S. Spec. who will become a trusted voice in determining the best course of action–not engineers who just provide negative assessments. The level of acceptable risk may be high during the initial hours of a disaster, but it will change with time. Most hazard mitigation slows rescue, thereby placing great pressure on the engineer/rescue relationship. Mutual trust and understanding are essential. The engineer`s role is to provide information, not make critical rescue decisions.

EFFECTS OF BLAST ON STRUCTURES

Explosions normally cause a supersonic shock wave and a buildup of pressure due to expanding gas. If the blast occurs within a confining structure, both effects can cause damage. When explosions originate outside of buildings, the gas can expand in the air and most of the damage is caused by the shock wave. According to Dr. Eve Hinman, specialist in blast-resistant design at Failure Analysis Associates in Menlo Park, California, the shock wave resulting from the bombing of the Alfred P. Murrah Federal Building in Oklahoma City radiated from the detonation point with a very large, initial pressure in the range of 5,000 psi. Although this pressure dissipates as a function of the cube root (a location twice the distance from the blast will feel only one-eighth the pressure), structures can be severely shattered or shredded by its effect, especially the upward pressure.

Natural gas or industrial accident explosions occurring within a structure blow the weakest parts of the enclosures away as the pressure escapes. If the wall and roof coverings in frame structures are well attached to the structure, the frame may be blown away as well. In large explosions, precast concrete walls, slabs, and even columns can be blown out, leading to conditions that will produce a progressive collapse. Progressive collapse starts with the initial damage caused by the force of the explosion: The blown-out vertical support makes the structure unstable, and the gravity load then drives the structure to the ground. It takes a very large explosion to damage a reinforced concrete building as long as there are enough lightly enclosed openings to allow the pressure to escape. However, buildings constructed of lighter wood, steel, and precast concrete can be leveled by large explosions as horizontal and vertical enclosing planes are blown out, leading to an overall loss of stability.

When large bombs have been placed immediately adjacent to structures, the effects also have been devastating. The shock waves radiate out in all directions, causing lateral and upward pressures in nearby buildings. All kinds of missiles–such as glass, building contents, and small building parts–are created; they can have devastating effects, especially on humans. Since, however, the pressure dissipates rapidly and has the greatest effect on objects in its direct path, a strong, resistant structure can provide shading from these effects. For example, Michael Rogers, the manager of the Murrah Building, was standing in front of the first-story east bank of elevators at the time of the blast. He was about 100 feet from the detonation point, but his view of the truck containing the explosives was obstructed by the corner of the concrete wall that formed the elevator shaft. He was not injured, but the west bank of elevators immediately behind him and in the shock wave`s direct path were severely damaged. Damages included the blowing in of their metal doors. (These elevators were used during the rescue effort after auxiliary power was provided.)

Most of the damage to structures more than a block from the immediate street front in the Oklahoma City incident was limited to nonstructural items (windows, ceilings, partitions, and the like). When standing in the upper stories of the Murrah Building, it was apparent that most of the damage occurred within a “line of sight” from the bomb origin.

THE MURRAH BUILDING

The Murrah Building, which housed federal offices, was a nine-story reinforced-concrete structure with no basement. Since the adjacent street slopes, the first-floor slab on ground is one-half story higher than the sidewalk at its west end and one-half story below street level at its east end. (This led to some confusion regarding nomenclature of first floor vs. basement.) The typical floor and roof consisted of a six-inch-thick slab spanning 20 feet in the east-west direction set between wide beams spanning 35 feet in the north-south direction. The three main rows of concrete columns, therefore, were 35 feet apart in the north-south direction. Each row contained 11 columns spaced at 20 feet over center. Cylindrical eight-inch-thick duct shafts were exposed on all four corners of the building. The main structure had a 220-foot 3 75-foot plan dimension. On the south side of the building, the stair and elevator towers projected from the main structure; their walls provided the principal lateral bracing. Precast concrete panels provided architectural window treatment at the south wall, and granite veneer panels were used to infill the east and west walls.

The north wall had a special architectural feature that proved to be particularly vulnerable to the effects of the explosion: To provide street access to the first floor, the second floor was held back 10 feet from the north building face and was supported by one-story wall sections spaced 40 feet over center. At the third floor, a large transfer beam, supported by two-story-high columns spaced 40 feet over center, carried the load of the columns above it, spaced at the normal 20 feet over center. From the third floor to the roof, the north wall was glazed with a full-height window/wall system. The first- and second-story windows were set back from the street.

Another construction feature that made the Murrah Building vulnerable to the blast`s uplift pressure was the use of discontinuous top reinforcing at both the slabs and beams. The beams were 48 inches wide by 20 inches deep and had only a few vertical stirrups spaced at nine inches over center at what proved to be a very important joint at the interior column support.

One-story structures were to the east and west of the main building. Only inches separated the structures from the main building. A four-level subterranean parking structure filled the remainder of the block to the south. The roof of the garage provided the main entry plaza at the second floor.

THE EFFECTS OF THE BOMB BLAST

The truck carrying the explosives was parked at the curb within 10 feet of the north wall line, between Column Lines 20 and 22. This was the only side of the main building where parking was possible; unfortunately, this was the most vulnerable side. The blast probably destroyed at least one of the four concrete, two-story columns at the north wall and produced enough force to shatter at least the adjacent second and third floors, including the transfer beam at the third floor. Lift also could have caused many other beams supporting the second and third floors to experience enough distress to collapse, but the pressures were probably not high enough when they reached the center columns (Line F) to shatter them. This upward pressure of the blast caused reversal of the gravity forces, leading to severe cracking at the tops of beams near midspan and cracking at beam bottoms near the face of the interior column along with the pull-out of the poorly embedded bottom-reinforcing steel.

Once the shock wave passed, gravity caused the rapid collapse of most of the front half of the structure due to lack of vertical support at the north face and inadequate resistance at the interior column. In addition, the blast damaged and caused the partial collapse of the second- and third-floor beams and slabs between Lines 16 and 22, south of the interior column line but just north of the stairs and elevators. Columns F22 and F20 were left standing without lateral support between the first floor and the bottom of the fourth floor and had cracked joints where the second and third floors had been ripped away. This area bounded by Columns F18, F22, E18, and E22 became known as “the Pit.” The remaining second- and third-floor beams and slabs, south of Columns F14, F16, and F18, were badly cracked and poorly attached to these columns.

In addition, Column F24 had been compromised, leading to the collapse of all floors in the area between Lines 22 and 26, all the way to the south wall of the building (Line E). This column initially may have been left standing after the collapse of its adjacent second- and third-floor beams (similar to F20 and F22) but then buckled under the load of the upper floors. The overall effect was to leave the building with most of its front half, plus an additional bay, missing; nine layers of badly shattered, compact, reinforced concrete piled in front; and many large pieces of beams and slabs hanging from the remaining floors at the face of the surviving columns.

This collapse is quite different from what most often occurs in earthquakes, where the seismic shaking will cause columns to crack and lose support at their joints with floors, leading to the pancaking of relatively intact slabs. These collapsed slabs can often bridge over projecting beams, columns, and other material to form survivable voids, but these voids may be more difficult for rescuers to penetrate.

HAZARD-MITIGATION ISSUES

As in the case of most building-collapse disasters, the initial victim rescue efforts proceeded without too much regard for stopping to assess and mitigate risk. A secondary bomb scare temporarily halted the rescue efforts, but significant hazard mitigation did not start until the second day (April 20). Where bracing and shoring could be installed without slowing the rescue efforts, it was easily justified and proceeded without much discussion. Some projects, such as removing hanging slabs, would cause curtailment of adjacent search and rescue and, therefore, became a source of future conflict between engineers and rescue.

Early on Day Two, a six-person incident support team (IST) sent by FEMA to coordinate the federal response met with the Oklahoma City Fire Department (OCFD) leadership to establish the initial strategy for what became the 16-day incident. The elements related to engineering were as follows:

The IST would have an engineering component to aid, coordinate, and document the task force`s hazard-mitigation efforts.

The initial two task forces (Phoenix and Sacramento) would be reinforced with four others to have the three full task forces working in 12-hour shifts around the clock. (It was later decided that fresh task forces would relieve the initial six; 11 were eventually deployed.)

Local volunteer contractors would continue to play a vital role in hazard mitigation and removal of debris.

(Oscar J. Boldt Construction provided most of the pipe bracing of columns in addition to some cribbing. Flintco, Inc. provided cabling; the initial epoxy mortaring of damaged columns; and, later, the steel collars with high-strength grout. Allied Steel Construction Co., along with Mid West Wrecking Co., provided for the cranes and removal of concrete.)

Collapse Hazards

During the incident, the following collapse hazards were discovered in the remaining structure:

Column G12 was cracked at the third floor but was still supporting eight levels.

Columns F20 and F22 were supporting seven levels–Level 4 through the roof. Levels 2 and 3 had been stripped away.

Columns F16 and F18 were poorly connected to second-floor beams; they were also supporting eight levels above.

The east end of the structure was disconnected from the stair/elevator bracing walls and was only marginally stable.

Falling Hazards

The following falling hazards were also present:

A large, 35,000-pound section of roof beam/slab was hanging from the top of Column E24 by a few rebars.

Numerous sections of concrete floor slabs and beams were hanging from the edges of the remaining structure.

Many building contents were precariously perched near the edge of the remaining structure.

The supporting beam connection of precast concrete panels suspended at the third floor over the south entry were partly broken and in danger of falling.

The granite veneer panels on the east wall were badly broken. Some had fallen on the floor of the adjacent one-story concrete structure.

The blast left numerous unreinforced concrete interior wall partitions in leaning configurations.

The mitigation measures available were the same as those for any other building collapse:

Remove the hazard.

Vertically and laterally shore or brace.

Monitor the hazard and develop a warning and safe-haven/escape scheme.

Avoid the hazard.

All of these methods were used in this incident. Almost all of the engineers who participated contributed to identifying or mitigating the many structural hazards.

The relative risks of further collapse and falling hazards were particularly difficult to assess in this incident. Task force engineer training places heavy emphasis on evaluating earthquake damage, with the near certainty that aftershocks will continue to subject the damaged structure to lateral forces. Many structures experience secondary collapse after the initial seismic event, and lethal falling debris is commonplace. Examples of wood, unreinforced masonry, reinforced concrete, and precast concrete buildings that have collapsed during aftershocks are featured in the Structural Specialist Training Course.

In the case of the Murrah Building, significant secondary loading (in addition to gravity) was limited to wind pressures and forces generated by removal of large concrete debris. The structure had experienced severe damage and had become dynamic, but it came to rest. The rebar that now served to suspend large pieces of concrete probably received significantly higher stress by the actions taken to stop those sections from falling.

On the other hand, as the north half of the floor slabs separated at the building center (Line F), the nominal rebar (temperature steel) in the north-south direction, placed (as in the normal case) just above the slab`s principal bottom reinforcing, caused the slab bottom steel to be ripped out of the remaining slab for a distance of more than five feet in some cases. This area of the remaining six-inch slab, therefore, had reinforcing from its discontinuous top bars but no bottom-reinforcing steel. In addition, the floor beams that remained were now spanning between the center and south wall of the building without the benefit of the balancing effect of their north spans. This situation could cause a significant increase in stress in the remaining beams` bottom-reinforcing steel, but the stress could be monitored by observing crack patterns. These two conditions did not prove to be significant problems.

INITIAL HAZARD ASSESSMENT AND MITIGATION

The Phoenix Task Force, with S. Spec. Fred Nelson and Dan Cook, arrived on the evening of Day One and performed the initial hazard assessment. During the second day, they directed the installation of pipe braces and confinement angles at the surviving two-story column (G12) at the north wall. They started shoring operations in the first story, south of Line F, in support of the rescue operations. They were then joined by the Sacramento Task Force, which also began shoring and rescue in a nearby area of the first floor. Under the direction of the original building engineer, Ed Kirkpatrick, P.E., Boldt Construction previously had installed pipe braces at the third-floor level of the two columns (F20 and F22) in most serious danger of collapsing. Boldt also had installed cabling around the damaged precast panels suspended above the south entry.

Five engineers from the Army Corps of Engineers and I arrived during the early hours of Day Three (April 21). We became the IST engineering component and divided into two shifts of three engineers each. I became the daytime lead engineer; Tom Neidernhofer, of the Corps, became the night lead engineer.

As dawn broke on Day Three, my task was to review the initial mitigation and become familiar with the building. After an extensive review of the building plans (several complete sets were made available) and a short meeting with Kirkpatrick and Bill White and Lex Paine of Boldt, I agreed with most of what had been initially accomplished. I found, however, that much more work was needed, in the following priority:

Columns F20 and F22 needed more substantial bracing to avoid collapse.

The east end of the structure was only marginally stable and had to be monitored (no operations were ongoing within this area).

Many falling hazards were hanging from the north edge of the remaining structure. The risk of their falling was less than in the case of seismic disaster, but eventually they would need to be addressed.

During Day Three, the engineers from the four newly arrived task forces were briefed regarding the building`s construction, using the existing plans; on previously accomplished mitigation; and on the current mitigation effort. Boldt arranged to locate larger-diameter pipes to more adequately brace Columns F20 and F22. A theodolite, for monitoring the east end of the structure, was borrowed from the Maryland-Task Force 1 cache.

On Day Four, rain slowed operations. With the aid of Los Angeles County Task Force Engineers Keith Martin and Rod Spears and Firefighters Smith and Cosey, the IST engineering staff marked (numbered) all the structure`s remaining columns and Prepared Mylar(TM) overlay drawings, color-coded to indicate the areas of initial collapse, current hazards, and completed shoring. These drawings then were available for future briefings and were updated as the incident progressed. During the following day`s initial briefing, I used the overlay drawings to more clearly define the important, remaining structural issues and suggest ways to mitigate them. This proved to be very timely, since, during the previous night, an ill-advised attempt had been made to cut down the large roof slab hanging from Column E24.

After the briefing and subsequent day-long effort to deal with the large roof slab, OCFD Chief Gary Marrs called an evening meeting of all S. Spec and task force leaders, IST leadership, and lead engineers to establish and prioritize an acceptable plan of action for dealing with the hazards in concert with the very active victim-recovery mode of the operation. The S. Spec. were instructed to coordinate their efforts through twice-daily meetings (at shift change) with the IST engineering staff. Inputs to modify the plan of action were to be coordinated through the IST lead engineer so that engineering could “speak with one voice.”

Having a coordinated engineering opinion became even more important as five additional task forces were rotated in to replace the original six. The new S. Specs., of course, had their own opinions; and as the incident moved into its second week, the risk-reward ratio shifted. Safety issues became more important as the hope for live victims faded. The new opinions needed to be filtered through engineering meetings and then discussed with operations leadership. The process worked well for the most part; but some engineers had trouble accepting the process when their own opinions did not prevail.

DISCUSSION OF SPECIFIC HAZARDS

Columns F20 and F22

Due to the collapse of all floor levels to the north and the second and third floors to the south, these two columns were left carrying seven floors without any lateral support at the second- and third-floor levels. Instead of these concrete columns being confined by the second- and third-floor beams, those floors had been ripped away, leaving cracked, uneven surfaces. The columns were being somewhat restrained by the debris pile that surrounded them and extended about 25 feet high on their north side and less than 15 feet on the south side (within the Pit area). If the columns failed, however, they would do so suddenly by buckling and cause all seven currently supported slabs to be compressed together in a heap.

Initially, four-inch-diameter pipe braces were extended horizontally at the third-floor level from the undamaged columns at Line E and connected to F20 and F22. This work was suggested by Kirkpatrick, but the 34-foot-long braces installed by Boldt had a slenderness ratio (length/radius) well beyond normally acceptable limits. My judgment was that these columns were our highest priority collapse issue, and I concentrated my initial efforts on designing and constructing a more reliable bracing system. Boldt located some six-inch-diameter pipe in 40-foot lengths, which had a much greater compression capacity. Also, these columns needed bracing in the east-west direction (as well as in the north-south direction) at the second- and third-floor levels. The pipes could be prefabricated on the adjacent entry plaza leading to the second-floor entry and cause only minimal delays in the rescue efforts in the Pit while being erected. (Note that a 40-foot-long, six-inch pipe weighs 800 pounds; a four-inch pipe weighs only 450 pounds.) The construction proceeded as follows:

Four-inch pipes were placed in the north-south direction at the second floor on Lines 20 and 22, extending from E to F. This size pipe was used again, but the pipes were trussed to the previously installed north-south pipes at the third floor to cut the effective length of both in half. Then, the four-inch pipes were also braced in the horizontal direction.

Since the collapsed-slab rubble pile projected south of Line F, it was not possible to place an east-west brace along Line F. The six-inch pipes–configured on a diagonal to provide the east-west component of the needed bracing–therefore, were extended to F20 and F22 at each floor. Since the angle was 30 degrees, the available east-west force was only 50 percent of the pipe`s capacity, but it was enough to provide sufficient bracing (two percent of the column`s vertical load normally is considered an adequate column brace). The diagonal configuration proved desirable in that it kept the braces from being bumped during the removal of the collapsed slabs.

The pipes were anchored to the columns at each end with four 34-inch drilled-in anchors. Unfortunately, some of the anchors did not hold during a subsequent slab-removal operation. Boldt then installed half-inch cables with a loop around each column at each floor to guard against further pull-out.

To patch and better confine the concrete at each column where the floors had been torn away, Flintco summoned an epoxy specialist, who attempted to place epoxy mortar over wraps of small wire rope at Column F20. Because of ongoing rescue operations and the inability to reach the north side of the column due to obstructions, this effort was only marginally effective. This contractor effectively injected epoxy in the anchorage at Line E to better guard against anchor pull-out.

On Day 10 (April 28), a large slab adjacent to Line 22 slipped onto the north-south second-floor brace near Column F22, causing the brace to bend and some of its anchors to be partly pulled out. Boldt straightened the brace with steel-pipe shoring. A vertical pipe bracing wall was then built under it and the adjacent diagonal pipe. Diagonal angles were added to the bracing wall to provide greater resistance from future shifting slabs.

The final enhancements to this column-bracing project occurred on Day 11, when the continuing slab removal allowed access to all sides of each column at the third floor. Boldt extended the existing pipe-brace connections so that they fully surrounded each column, providing a more rigid tension connection than the cables. Flintco fabricated two-foot-long sleeves and placed them over the cracked and poorly confined concrete at the third-floor joints of F20 and F22. The sleeves were then filled with high-strength, quick-setting grout to form a structural bandage. These columns were monitored using Smart Levels® provided by the Oklahoma office of Macklandburg-Duncan. The levels were placed on each of two surfaces between the second and third floors and set to zero. Any change in angle could then be measured with an accuracy of about one-tenth of a degree (lateral movement of one-quarter inch at each column joint). During the removal of all large adjacent slabs, the levels were read using 10 3 50 binoculars from the safe area adjacent to Line E; no significant movement was detected for the remainder of the incident.

Columns F14, F16, and F18

The remaining second-floor beams and slabs between Lines F and E were cracked and nearly severed from the columns at Line F. This area was shored, using mostly vertical wood post shoring. The shoring was installed by several task forces at the first story in this area (which became known as “the Forest”) during the first several days of the incident. This work was done not only to support the floor but also to allow the floor to provide the columns on Line F with whatever lateral support it could.

It became clear that the collapsed rubble–slabs leaning on these columns from the north–would have to be moved eventually. Therefore, more positive lateral bracing for the column had to be designed. Knee braces made from four-inch-diameter pipe were extended down from the third-floor beams to provide a tension or compression force on two perpendicular faces of each column. Vertical shoring was added between Lines 16 and 18 to support part of the third floor when it was observed that some of its cracks increased in width.

On Day Seven, one of the knee braces along Line F was knocked off by a large slab being removed by the crane. The force pulled the drilled-in anchors out of the column even though they appeared to have been installed properly. A new brace–made from higher-tension-capacity anchors, supplied by the drill manufacturer representative–was installed and carefully tightened (a torque wrench, which is strongly recommended for this task, was not immediately available; neither were compressed-air cans or other air-jet devices to clean out the holes after drilling). In addition to the enhanced anchorage, a cable stay was added to Column F14 at the time the large slabs leaning against it were scheduled to be removed. Shortly thereafter, victim-location data indicated that all potential victims had been removed from this area and no additional nearby slabs needed to be moved. Therefore, no additional cable stays were installed.

East End of Structure

This end of the structure had become structurally separated from the stair/bracing walls and, therefore, was judged to be only marginally stable. This section of the structure would have to resist the lateral forces generated by wind. It survived the blast with significant damage to the granite infill wall veneer (mounted on metal studs placed vertically between floor edge beams). The edge beams were also badly cracked at their connections to the northeast concrete vertical duct shaft. The shaft also had large duct openings at each floor and was badly cracked between the first and second floors. It was decided this section of the building would be monitored by sighting with a theodolite from the top to the bottom of the northeast concrete duct shaft to determine if it remained vertical.

During the following three days, the maximum observed east-west movement at the top of this 120-foot structure was only one-half inch, which could probably be attributed to temperature change from one side of the cylindrical section to the other. Fortunately, the Oklahoma winds remained relatively light (35 mph maximum) during the 16-day incident and no larger than one-half-inch movement was observed at any time. (Continuous direct contact with the local weather service kept us informed on wind conditions.)

Since the missing floors in the Column E24 bay left the column with lateral support only at the first floor and roof, the column`s north-south movement was monitored by another theodolite. Many pieces of beams with attached slabs were hanging from this column; debris up against it in the first two stories was leaning to the south; and the column was cracked at about the fifth floor. The theodolite revealed the column bowed to the south, in the middle, about one-half inch (the line of sight remained within the outline of the three-quarter-inch corner chamfer). This condition remained constant throughout the incident. Since this section of the building was only accessible by crane or rope, it was initially searched and then avoided until the monitoring demonstrated that the probability of its collapse was relatively remote.

Large Section of Roof Beam/Slab (“Slab from Hell” or “Mother Slab”)

This 35,000-pound section of concrete was hanging from the top of Column E24 by two one-inch-diameter beam bottom bars. Several small slab bars were still attached to adjacent slabs, but the primary forces were being carried by the two bottom bars. Since these bars were configured at approximately a 45-degree angle, forces of about 25,000 pounds were acting on the column vertically and horizontally, pulling to the north. The horizontal force was being resisted by the remaining roof slab and transferred into the nearby stairwell bracing walls. The force also may have been helping to resist the collapsed concrete slabs leaning against the lower two stories of Column E24. In any case, abruptly cutting the bars would cause the 25,000-pound horizontal force to change to zero immediately, possibly with catastrophic results to the adjacent, marginally stable east end of the building. On Day Four, an attempt was made to remove this hazard by cutting the rebar and allowing the slab to fall to the rubble pile just north of Column E24. The idea of removing this significant hazard was good, but the method could have caused a more serious problem than had been anticipated. The building had been evacuated prior to cutting, so a secondary collapse would not have injured rescue teams. However, if parts of the adjacent areas had collapsed, rescue efforts would have been seriously affected.

After several of the smaller bars and one of the two beam bottom bars were cut, the concrete slab shifted abruptly so that the lower edge of the 20-inch-thick beam section swung in to the west and came to rest on the eighth floor near Line 22 and about 10 feet from Line E. The cutting operation was discontinued at this point to reassess the situation.

At about 0800 hours on Day Five, I accompanied a group that included OCFD Operations Chief Mike Shannon, FEMA Operations Chief Ray Downey, Phoenix S. Spec. Fred Nelson, and others to the ninth floor to assess the slab`s current condition and devise methods for further mitigation. We found the slab was still suspended from a single, one-inch-diameter rebar extending from one edge of the bottom of the concrete beam to the top of Column E24 at roof level. In addition, the lower end of the 35,000-pound slab was firmly bearing on the third floor. Nelson and I agreed it was probable that the single rebar, due to its angle, was exerting a horizontal and vertical force on the column and it would be unwise to proceed with cutting the bar. Cutting the bar might have resulted in the following.

The 35,000-pound slab section could fall eight stories and severely damage anything in its path.

The release of the potential 25,000-pound horizontal force at the column top would change the current force system in the roof and lead to unpredictable consequences, the most dire of which would be the collapse of the east end of the building.

A section of the eighth floor would collapse when all of the weight of the 35,000-pound slab was released.

Since the goal was to remove this large falling hazard, it was suggested that the slab be broken in place using normal cutting tools and small explosives and catching the pieces in a dumpster suspended by crane. The OCFD agreed to summon Jim Redyke, president of Dkyon Inc. in Tulsa and a trusted local explosives expert. Redyke arrived at 1600 hours the same day and advised it would require from 24 to 30 hours to properly remove the slab using explosives. He further suggested that instead of the explosive-removal scheme, we cable-tie the slab to the stair wall on Line 22. This could be accomplished much sooner. The OCFD, sensitive regarding the use of explosives in this bomb-caused collapse and in view of the length of time rescue operations would be restricted in a major portion of the building during the removal process, allowed the slab to be cable-tied and remain in place. I was disappointed with the time estimate for the explosive removal of the slab but could not argue with the OCFD decision, based on the alternatives presented.

During the night of Day Five, Flintco tied the slab with two separate loops of cable. The loops passed around the slab and a section of stairwell wall between its north end and a window opening. Several things about the cabling system were less desirable than I had expected, such as the following:

The cables were bent around 4 3 4 wood strong backs on the slab instead of steel pipe, which is not subject to cable cut-through and would have provided a smoother cable corner bend.

The cable`s configuration was essentially horizontal and did not provide any vertical lift to the slab. (S. Spec. Rick Barrett and Rigging Spec. Bob Simmons arrived days later with the Menlo Park Task Force. They devised a more competent cable scheme that included running additional cables through new holes drilled in the roof slab and then extending them around the roof beam on Line 22 near Line E. Their scheme was never implemented, since shortly thereafter the incident changed from rescue to recovery mode. (At that time, rescuers were excluded from working under this slab, and an excavator with a long reach was used to uncover the remaining victims in that area.)

One of the cables, in addition to being looped through the stair window opening, was run over a precast wall spandrel panel. The change in direction of the stressed cable caused a downward force in the panel and its supporting steel beam. On observing this at the beginning of my shift on Day Six, I asked Jay Harris, of Flintco, to provide some solid wood shoring between the bottom of the north edge of the precast panel and the top of the concrete floor edge beam at Line E.

The cable clamps used to make the loops for attachment to the ratchet pullers were configured such that the middle of the three clamps had its U-bolt against the live side of the loop instead of all three clamps being aligned with their grippers (saddles) on the live side (never saddle a dead horse). The local contractor explained that reversing the middle clamp was better since it gripped both sides of the loop; I subsequently learned it is not the safest way and is not recommended by the clamp manufacturer (another lesson learned).

The slab was monitored with a theodolite and personal checks by me, my replacement John Osteraas (of Failure Analysis Associates, Menlo Park, California), and others at least once a day for the remainder of the incident. No significant movement was observed, but the slab became a constant source of controversy regarding the risk-reward ratio of working under it. Regardless of how well an object as large as this is secured, it preys on engineers` (and rescuers`) minds until it is removed.

Numerous Sections of Hanging Floor Slabs

During windy spells that followed Day Five, small pieces of concrete and other debris fell from the slabs hanging from the north edges of the remaining floors. Much of the easily removable material was taken off and pulled back beginning on Day Two. The roofing membrane, free hanging rebar, and building contents were all reached at their individual floor/roof levels and removed or pulled back.

During Day Six, Los Angeles County Task Force Leader Mike Idol suggested the team`s S. Specs. Keith Martin and Rod Spears carefully inspect and record the relative hazards involved with each of these concrete slabs (which by now had been given names such as “Australia”). I agreed with their criteria for removal, which were the following:

Remove large (baseball size and larger) pieces.

Remove slabs where the bar from which they hung them could not be seen to pass through them.

Remove slabs suspended by bottom north-south rebar, since there was no assurance of where they were spliced.

After a survey of most of the hanging slabs was completed on Day Seven, a short meeting was held to obtain IST Operations and OCFD concurrence in removing the slabs. As a result, a group consisting of S. Specs. Martins, Spears, Jim Lambert of the OCFD, and me was assigned to review the survey. We quickly agreed on the slabs that had to be removed. The meeting occurred after our shift ended; we four remained through part of the night and were raised in a crane-lifted man bucket to physically mark each slab. Night rescue operations on the north side of the building were then suspended so that both cranes could be used to rig, cut, and lift slabs or cut and drop them into a suspended dumpster. The work was nearly completed by sunrise.

In my opinion, this was a good example of how S. Specs. who are respected and well-integrated within their task forces can work through the prescribed chain of command to resolve the conflicting issues that often accompany hazard mitigation.

As a follow-up comment, it was ironic that, after these slabs were removed, a piece of six-inch-thick insulating concrete from the roof fell near one firefighter during a subsequent wind squall. Weighing about 10 pounds, it fell approximately 100 feet and could have caused injury. We engineers had inspected it along with the other slabs; however, since the wire-mesh reinforcing appeared to be restraining it, we judged it to be a reasonable risk.

The remaining hanging slabs, those below the “Mother Slab” at Column E24 and those hanging from the east section of the building, later were mitigated by the Menlo Park Task Force. The task force devised some unique cable stays and enclosures (diaper) of the larger slabs under Mother Slab that could not be reached by the crane.

ENGINEERING LESSONS LEARNED

Thirty-seven Fema engineers participated in this incident (22 task force and 15 IST). This was a group effort. All worked as well as they could to help achieve positive results. We all attempted to act responsibly in minimizing risk without hampering rescue efforts. Most engineers became more uncomfortable with the uncertainties of some hazards as the days passed and the chance to experience the reward of finding live victims diminished. We were all better at identifying hazards than assessing their probability and devising creative, efficient solutions.

The Oklahoma City engineering experience brought to light the following:

The IST needs an engineering component. A dedicated S. Spec. staff within the IST is needed to evaluate, coordinate, and implement structural engineering issues affecting USAR resources. Also, IST should have an engineering cache that includes monitoring and testing tools.

The Structural Specialist Training Course has validity. The three S. Specs. who did not participate in this training were not as well prepared to deal with the chaos of the Murrah Building structure. Those who attended benefited from the experience; they had seen examples and worked tabletop problems. They had a better idea of what to expect.

IST and task force engineering functions must be supported by the Army Corps of Engineers. Task force S. Specs. need assistance when building monitoring and contractor supervision are required. The Corps appears willing and is able to offer many of its more than 70 trained S. Specs.

Private contractors assigned to an incident must be adequately supervised and monitored. We must be reminded that quality construction requires quality control.

Task forces should integrate their engineers through local training and exercises. The mutual trust between engineers and the task force leads to better communication of engineering issues and more effective work.

Define the role of the Geographical Information System (GIS) in USAR operations and in ensuring rapid deployment; use the system.

Conflict must be addressed. This can be done through better communication, using persuasion instead of confrontation, adequately defining the level of acceptable risk, and developing creative alternatives.

Lack of continuous reinforcement of horizontal and vertical structural members in concrete construction makes these types of buildings particularly vulnerable to upward pressure caused by a blast.

We must strike a balance between “building” shoring and “acceptably engineered” shoring. This is an issue of training and currently is being addressed by the FEMA USAR teams.

Actions already have been taken to address some of these issues. S. Spec. training and Rescue Spec. training will include new opportunities to practice better communication. Keith Martin, Los Angeles County Task Force, is developing ideas on how to better confine hanging concrete and on the design of a portable victim shelter. Information on cable capacity and hardware, new metal detection devices, more sensitive tilt meters, and the testing of anchors is being collected to be shared in the future. In the following months, we will continue to develop new, more timely methods.

* * *

We must not assume that the next incident we face will be similar to the Oklahoma incident. This incident has added much to our base of structural information, but we must be better prepared to more aggressively deal with the new situations of the future. We performed well, and we will get better. I am very proud of our response. n

Day Three: North side, including shoring of Column G12. (Photos courtesy of author.)

(Top) Pipe bracing of Column F22 at the second floor level. (Bottom) The author and Division Chief Jim Hone, Santa Monica (CA) Fire Department, checking soundness of grout pour in steel sleeve “bandage” on third-floor Column F20. Note the Smart Levels® on two sides of the column below the third floor.

(Left) Anchorage of pipe bracing to strong Column E20 and (background) vertical shoring along Line 18, start of “the Forest.” (Right) Day Three: Cables on precast panels over south entry.

Final picture of the 35,000-pound “Mother Slab,” with two cable stays, lower end on eighth floor, upper end hanging from top of Column E24 by one large rebar.

Hanging concrete slabs cabled and diapered to Column E24.

FEMA STRUCTURE/HAZARDS EVALUATION MARKING

The structures specialist (or other task force member as appropriate) outlines a two-foot by two-foot square box at any entrance accessible for entry into any compromised structure. Aerosol cans or spray paint (orange color only) are used for this marking system. It is important that an effort be made to mark all normal entry points to a building under evaluation to ensure that task force personnel approaching the building can identify that it has been evaluated and discern its condition.

Specific markings are clearly made inside the box to indicate the condition of the structure and any hazards at the time of this assessment. Normally, the square box marking is made immediately adjacent to the entry point identified as safe. An arrow is placed next to the box indicating the direction of the safe entrance if the structure hazards evaluation marking must be made somewhat remote from the safe entrance.

The depictions of the various markings are as follows:

Structure is accessible and safe for search and rescue operations. Damage is minor with little danger of further collapse.

Structure is significantly damaged. Some areas are relatively safe, but other areas may need shoring, bracing, or removal of falling debris and collapse hazards. The structure may be completely pancaked.

Structure is not safe for search and rescue operations and may be subject to sudden additional collapse. Remote search operations may proceed at significant risk. If rescue operations are undertaken, safe haven areas and rapid evacuation routes should be created.

Arrow located next to a marking box indicates the direction to the safe entrance to the structure, should the marking box need to be made remote from the indicated entrance.

Indicates that a hazardous-materials condition exists in or adjacent to the structure. Personnel may be in jeopardy. Consideration for operations should be made in conjunction with the hazardous materials specialist. Type of hazard may also be noted.

The time, date, and specialist ID also are noted outside the box at the upper right-hand side. This information is made with pieces of carpenter`s chalk or lumber crayon. An optional method may be to apply duct tape to the exterior of the structure and write the detailed information on the tape with a grease pencil or black marker.

The depiction above indicates that a safe point of entry exists above the marking (possibly a window, upper floor, and so on). The single slash across the box indicates the structure may require some shoring or bracing before continuing operations. The assessment was made on July 15, 1991, at 1:10 p.m. There is an apparent indication of natural gas in the structure. This evaluation was made by Oregon-Task Force 1. n

STATUS REPORT–STRUCTURAL SPECIALISTS, FEMA

TUES 4/25/95 (0700-1900)

0730 Structural briefing between oncoming and offgoing structural specialists with Task Force structural specialists.

0900 Checked cable connections to Column E22 on 9th floor.

0945 Transit observation of NE airshaft (Col. E28) for East/West deflection. (No additional deflection.)

1000 Received 2 complete sets of full-sized drawings for Bldg.

1000 Agreed with Boldt Const. Co. to add additional lateral support bracing on the 2nd floor level in the North/South direction between Columns F22 & E22.

1000 – 1200 Discussed lifting of slabs on N. side of Bldg. at Column Line F.

1200 Transit observation of NE airshaft (Col. E28) for East/West deflection (No additional deflection.)

1345 -1515 R. Tillman escorted COE bomb damage expert from Preventive Design Center to show damages and shoring.

1400 Visited Column F18 to assess need for bracing and shoring identified by L.A. Co. Task Force #2.

Notified Boldt Const. of need for pipe braces from Column F18 to F16 between 2nd & 3rd floors.

Briefed Puget Sound task force struc. specialists and started them on shoring at mid slab between F16/F18 and E16/E18.

All pipe bracing completed by 1800 hrs. n

DAVID J. HAMMOND, a structural engineer, was the leader of U.S. Search Dog Team 3 in the Mexico City 1985 Earthquake and actively involved in numerous other disasters as a support member of the California Rescue Dog Association. He is a member of the FEMA Urban Search and Rescue Advisory Committee and lead instructor for the structural specialists as well as other training courses. As a director of USAR Inc., an information-coordinating nonprofit group, he has prepared and taught courses regarding structural aspects of urban search and rescue for heavy rescue firefighters since 1987. He was a lecturer in the architecture and civil engineering departments at Stanford University from 1965 to 1976 and chaired the Disaster Emergency Services Committee, Structural Engineers Association of Northern California.