What Collapsed Structures Can Teach Us

By Stephen T. Spall and Alexander J. Streichenwein

On January 12, 2010, at approximately 4:53 p.m. EST, Haiti's ground shook for the first time in more than 240 years with a magnitude 7.0 earthquake. What followed over the next hours, days, and weeks can be described as predictable and miraculous. Shortly after the earthquake, New York Task Force 1 (NY-TF1) became part of a multinational response that combed the piles of rubble searching for any signs of life. The 82-member team included two structural specialists whose responsibility it was to evaluate building stability. In previous deployments in the United States, we were accustomed to dealing with a well-defined building code. The building construction we encountered in Haiti was unlike anything you might find here. Haiti’s lack of a stringent building code allowed many of the country’s buildings to collapse.

In Haiti, as in every country, there is a building code from which all construction must be standardized and construction practices monitored. Monitoring the enforcement or use of the code is as important as the code itself. The greatest code, if left unenforced, is the same as having no code at all.

The first true building code in the United States was created as a result of the Great Baltimore Fire of 1904. It was a “prescriptive code” that laid out penalties for poor construction and described “good construction.” Many publications that discuss building construction often use the term “shoddy workmanship,” and that is exactly what the enforcement phase of a building code is meant to address, usually in the form of inspections.

According to the director of the Organisation of American States Department of Sustainable Development in Washington, Cletus Springs, Haiti uses the Caribbean Uniform Building Code, as do most of its Caribbean neighbors.1 Unfortunately, existence does not mean use or enforcement. Haiti is ranked as one of the world's poorest nations2 with an average annual income of $750 US; few Haitians can afford compliance with a code of any type.

Instead, what we found did not readily conform to the needs of a country with an active seismic zone. Even though Haiti is on the point of convergence of three large tectonic plates, the last earthquake in Haiti was in 1770. As a comparison, California’s infamous San Andreas Fault lies on only two such plates but has remained active on an almost routine basis.

Perhaps a more relevant case in point could be Chile’s recent earthquake. Following many years of recurring earthquakes, Chile instituted a comprehensive earthquake code. It is notable that the resulting substantially reduced death toll (< 100 vs. Haiti’s > 200,000) is largely attributable to the use and enforcement of this code.

Despite professional estimates that seismic precautions instituted during construction add only five to 10 percent to construction costs, little was done to make Haiti’s buildings more resistant to earthquake damage. What we found were building practices that were not part of a code but rather a “prevailing form of construction” of lightly reinforced masonry frames consisting of undersized, underreinforced columns and floor slabs.


The frame was comprised of concrete columns roughly the length and width of a concrete block (about eight × 16 inches), but some were just six to eight inches square. Each column uniformly had four reinforcement bars, commonly referred to as “rebar.” Unfortunately, not all rebar is alike. A majority of what we found in Haiti’s damaged structures was smooth. Experts in the American Concrete Institute (ACI) Concrete Code, used here in the United States, estimate that the rebar used in Haiti was at least half the diameter used by the California earthquake compliant code. It also had only half the strength of a similar diameter rebar used here.

Plain concrete has no tensile (pulling) strength, but it has significant compressive (pressing) strength. When loads are applied to a reinforced concrete member (column, floor, or beam), it is transmitted (transferred) to the steel by the bond stress between concrete and steel.3 Rebar that is not smooth but instead has “ribs” is called “deformed.” Deformed rebar transfers the forces (both compression and tension) from the concrete to the steel more efficiently. In its simplest form, the concrete carries the compressive forces, and the steel (in the form of rebar) carries the tensile (pulling) forces.

Deformed bars (steel with raised patterns) can develop a higher bond stress without slipping than a smooth bar can. (3)This may seem like a small detail, but good concrete construction is made up of such detail, and you must pay attention to it.

Each column was constructed eight to 10 feet apart, in both the length and width of the residence, starting generally at the corners. The floor slab was laid on top of these columns. The connection of the column was also a point of concern. Quality modern concrete construction requires that the rebar bend 90° and “splice” or tie-in the column directly into the floor. This will make the floor and columns act as one in what is referred to as a “rigid” connection.

Unfortunately, in Haiti, this was not the method used. Often at this connection, the rebar simply continued vertically up through the floor, into the next column, resulting in a weak connection. A majority of the structures we searched failed catastrophically at this location because of the resultant heavy mass (the floor) atop “thin” columns—“thin” because the ratio of height (length) to thickness (or diameter), referred to as L/D (for length to diameter), was too high. It was evident that the failure in Haiti occurred at the column-slab connection (photo 1).

(1) Standard Haitian cast-in-place columns. Notice the rough surfaces. The earthquake did not cause them; they are actually air gaps or “pockets” (right two columns) that occurred during the pour of the column. In the United States, we use mechanical vibration tools to compact and remove these air gaps that, if left, weaken the column. Although this building appears old and weather-beaten, it was relatively new—less than 10 years old. (Photos courtesy of NY-TF1.)


The floor slabs in Haiti were both a marvel of economy and a disaster waiting to happen.

To construct the floor, a system of jacks and plywood panels—the standard for cast-in-place concrete construction—is assembled, and it forms the horizontal surface that is the underside of “the pour.” In the United States, the workers who construct this formwork are called “lathers.” We could tell by the smooth surface left behind in Haiti that the formwork for the floor systems was either thin layers of tin or plywood. (We are assuming that tin was used because no one had seen even a single sheet of plywood.) The poles or “jacks” used to support the slab during construction were all steel screw jacks rather than the 4 × 4s that are more prevalent in the United States, because of the very limited supply of lumber in Haiti.

Once the jacks and formwork were in place, three-celled, hollow-core cinder blocks were laid flat on their side, end to end, in rows. Some had all the cells pointing in one direction and in line with each other, creating a hollow tile floor; some had been placed at random. Small gaps of less than two inches separated each tile in a row and acted to encase the blocks in the pour. Every fourth to sixth row was omitted, about one every three feet, leaving a six-inch gap for the entire length. The pattern was then repeated with another four to six rows, and another gap, until the entire area was covered. Later, as the concrete was poured, two pieces of rebar were laid the length of these gaps that created small girders encased in the floors. At the ends of these girders, they met similar, but larger, girders at right angles, which spanned from column to column, creating a grid. Because the surface of the block was not smooth, the blocks became bonded to the concrete poured over and around them and remained in place. After curing, the formwork was removed, and the entire ceiling received a three-quarter-inch layer of plaster from the floor below. The floor created was then leveled with a two-inch chalky layer of vermiculite and smooth enough to accept tile. The result was a remarkable product that easily supported the requirements of the average residential occupant of 40 to 60 psf. Unfortunately, this clever design also proved insufficient under earthquake loading (photo 2).

(2) A view of the floors as seen from below. You can clearly see individual blocks. This area would receive three-quarters of an inch of plaster as a finished ceiling. Also notice that the columns are no longer straight (plumb). You can see rebar sticking out of each cast-in-place row. You can also see (upper left) that when the builder ran out of full block he used pieces to fill in the area.


Haiti’s cinder blocks are not like ours. (Note: “Cinder blocks” have not been made since the 1940s. Now they are called “concrete blocks,” but we still refer to them as cinder blocks.) Some estimates claim that Haiti’s standard cinder block was half the weight of those manufactured in the United States. We found them to be somewhere around three-quarters the weight, or about 40 pounds, but that was not the true problem.

In New York City, we use a variety of materials for what we call nonbearing, curtain, or “in-fill” walls. They are just what the name implies: They “fill in” the space between the columns, which actually support the structure, and provide no more strength than hanging a curtain over the opening. The difference is that the New York City Building Code, like all construction codes worldwide, requires masonry walls to be “tied in” or made more substantial than a curtain. This is usually done with small lengths of rebar that connect the wall to the column on each end as well as where they meet the floor and ceiling.

If you reinforce the wall vertically and horizontally, you can create a “shear wall.” Shear walls are what resist lateral (horizontal) loads on a building such as wind or the side-to-side motion of an earthquake. Generally, masonry shear walls have small welded wires, called “trusses,” laid horizontally in the mortar joint and rebar up through the hollow sections of the block, where they are spliced into rebar imbedded in the floor and ceiling. These details help the wall and framework to act in unison to resist the horizontal forces that could cause a building to collapse.

Most infill walls we found in residences in Haiti did not have any such connections or reinforcement. Remember, almost all of the reinforcement we found was in the frame’s columns, not the walls. Instead, we found mortar joints that did little to hold the wall in place during the earthquake, generally resulting in the entire panel’s hinging at the floor and falling in one piece.

Once again, the lack of proper connections between the building’s walls and columns resulted in no (or very little) lateral stability. This caused the buildings to collapse in a “lean over” or “pancake” collapse pattern (photo 3).

(3) This is typical wall construction in Haiti. Notice the thin cast-in-place columns and the blocks on the end (on the right side of the wall) that are not interlaced, just stacked straight up. This is a weak connection.


Mixing quality concrete is a complex and precise process that is constantly undergoing revision and testing and is too complex a process to adequately define here. Generally, it combines water, cement, and an aggregate such as stone or sand. There are different cements and different mixtures, based on application. The water-to-cement ratio is key. If there is too much water, the strength is lowered. When water is added, the workability is increased and more aggregate is added. This means fewer bags of cement per cubic yard, and the cost is also lowered. But beware: Too much sand will form a mixture that will have greater shrinkage with more resulting cracks. All concrete has cracks, but the properties of the resulting cracks and the placement of rebar unite to make the dependable structures we trust.

The most plentiful commodity the people of Haiti have for construction is sand, one of the main components of concrete. Unfortunately, because of cost, the prevalent practice is to use the readily available beach sand. This brings with it moisture and sea salt that continue to attack the rebar throughout the life of the concrete, rusting and weakening the bond that is so necessary for the required strength.

There were no concrete delivery trucks in Haiti, at least not in the remote residential neighborhoods in which we often operated. Indeed, many of the homes were not reachable by car but rather by a series of small, winding, turning alleys. The result is that all concrete was mixed on site, with the resulting lack of quality control or consistency.


Another thing that we noticed was the similarity of Haiti’s construction practices to our wood-frame construction. We found that the size of the columns on a building’s first floor were often exactly the same size as those used on the top floor, such as is done across the United States in residential construction. There was no allowance made for the larger weight bearing down on the first floor by the subsequent levels of concrete constructed above. Throughout the United States, a similar concrete building would have noticeably larger column dimensions on the lower floors to support the accumulated load.

Although understandable because of the lack of a recent history of such events, ultimately Haiti’s failure to pay attention to detail, in both design and enforcement, led to structures that failed to adequately protect the occupants during the devastating effects of an earthquake.


1. “Engineers Urge Overhaul of Haiti’s Archaic Building Practices,” Jacqueline Charles and Curtis Morgan, The Miami Herald, posted January 25, 2010.

2. Central Intelligence Agency (2009). “Haiti.” The World Factbook. https://www.cia.gov/library/publications/the-world-factbook/geos/ha.html. Retrieved 2010-01-28.

3. Merritt, Frederick S. Building Construction Handbook (first edition), 5-9.

STEPHEN T. SPALL is a 24-year veteran of the Fire Department of New York (FDNY) and has been the captain of SQ-61 since 2003. He is the lead structural specialist on NY-TF1 and has been a member since the team’s inception in 1992. He is also an instructor at the FDNY Technical Rescue School and has a bachelor of science degree in civil engineering from SUNY at Buffalo.

ALEXANDER J. STREICHENWEIN is a 10-year veteran of the Fire Department of New York (FDNY), assigned to Squad 61. He has a bachelor of science degree in civil engineering from Manhattan College. He is a structural specialist on NY-TF1 and an instructor at the FDNY Technical Rescue School.

More Fire Engineering Issue Articles
Fire Engineering Archives

No posts to display