Archives

An Investigation Of The World Trade Center Collapse

by John J. Fanning

On August 25th at 11:00 am I gathered with about 30 other people in a conference room at Underwriters Laboratories (UL) in Northbrook, Illinois. What brought us here was an invitation by UL and the National Institute of Standards and Technology (NIST), to witness a fire resistance test of the floor system used in the World Trade Center towers in New York.

Michael Newman, the Director of Media Relations and Spokesman for the WTC Investigation for NIST, opened the meeting by welcoming us to the UL Lab before introducing S. Shyam Sunder, Deputy Director and the Lead Investigator for the NIST, who offered us an overhead presentation, detailing why we were there and what we were about to witness.

$16 million has been allocated for an investigation into what happened at the World Trade Center and about $5.5 million has been awarded in contracts so far. As Sunder explained, the objectives of the investigation are to:

• Determine why and how the WTC Towers collapsed following the initial impact of the aircraft and why and how the 47-story WTC 7 building collapsed.

• Determine why the numbers of injuries and fatalities were so low or high, depending on location, fire protection, occupant behavior, evacuation and emergency response.

• Determine what procedures and practices were used in the design, construction, operation and maintenance of the WTC buildings, and,

• Identify what, if any, fire codes, standards and practices should be revised as a result of the investigation.

Sunder detailed that this segment of the investigation is trying to determine the validity of assumptions and procedures used by the builders of the WTC as they relate to compliance with the New York City building and fire codes. To determine this, UL constructed scale models of the structural flooring used in the WTC, duplicating as closely as possible the exact methods employed in the late 1960’s construction.

To illustrate the attention to detail NIST and UL paid to this test, Sunder explained how the steel floor trusses used in constructing the WTC, employed a welding technique that is no longer in use today. UL engineers researched this outdated welding method and duplicated it on the truss models that we used for testing. The concrete mix and placement of the floor trusses also replicated the “as specified” construction as closely as possible, keeping in mind that the tested specimens were scale models of the actual floors in the WTC.

The model being tested this day was 17 feet long and had a spray-on fireproofing thickness of 1/2 inch. UL had already completed a test at their Northbrook lab on a similar model using a fireproofing thickness of 3/4 inch. Earlier, UL had conducted tests on two, 35-foot long models in another of their laboratories, in Canada.

At the time the towers were built, a half-inch of spray-on fireproofing was specified for the towers. How this half-inch specification came into the final construction plans remains unclear. No full-scale fire resistance test of the floor system was ever made. In 1966 the original architect, and in 1975, following a fire in the WTC, a Structural Engineer, expressed the need for such testing, each stating that the fire rating of the floor system could not be determined without such a test. Despite their recommendations, the flooring systems were never tested. It remains unclear who actually specified the half-inch fireproofing thickness. A letter dated October 30th, 1969 from Robert J. Linn, the Manager of Project Planning and sent to Mr. Louis DiBono, the fireproofing contractor, stated that “all tower beams, spandrels, and bar joists requiring spray-on fireproofing are to have 1/2 inch covering of Cafco (a fireprofing material)”. But it is unclear if Linn would have the sole authority or the technical expertise to unilaterally make such a specification.

In 1995, two years after the first terrorist bombing of the WTC, the Port Authority decided to upgrade the fireproofing in the towers. Tower 1 was upgraded with a specified coating of 1-1/2 inches of fireproofing. When I asked if asbestos-containing material (ACM) was used, Sunder informed me that ACM had been originally used on approximately 39 of the floors in the tower, but he was uncertain as to what the status of this ACM was and whether it was removed. But according to the information provided by NIST, although a 1-1/2 inch thickness of fireproofing was specified as the finished depth for the renovation, the Port Authority reported a finished depth of 2-1/2 inches, or one-half inch greater than what was specified.

Doing the math, I could not help but wonder if the Port Authority had left the ACM fireproofing intact and applied the 1-1/2 inches of new fireproofing directly over it as an encapsulation measure. If such was the case, it would explain the discrepancy between the specified depth and the reported finished depth. I understood that questions about ACM were not relevant to the UL or NIST tests that were being conducted, but I felt it important to ask because of the illnesses being reported by rescuers and workers who labored at the Ground Zero site. I also wanted to satisfy myself that all of the information relating to the construction of the WTC has been obtained prior to these tests.

Significant, too, for determining the exact state of the existing fireproofing in the WTC is the desire to answer the question as to why Tower 1 stood nearly twice as long as Tower 2, despite the fact that they were struck by nearly identical aircraft.

The WTC was owned and built by the Port Authority of New York and New Jersey. Such an authority is a quasi-government entity and is not bound by any local, state or federal jurisdiction. Accordingly, the Port Authority was not required to follow New York building codes even though the WTC was located in the heart of New York’s financial district. But a longstanding policy of the Port Authority was to meet and where appropriate, exceed the building and fire code requirements of local municipalities. This policy seems to have been in effect in at least the fireproofing portion of construction. In the same letter written by Linn in 1969 to the fireproofing contractor Linn stated: “the above requirements (1/2 inch Cafco) must be adhered to in order to maintain the Class 1-A Fire Rating of the New York City Building Code”. It is notable that the completed buildings were eventually designated as 1-B.

In 1968, during the time of construction of the WTC, the city of New York revised its building codes for 1-B structures and lowered the required fire resistance rating of columns from the 1938 standard of a four hour fire rating to three hours. They also lowered the fire resistance rating of floors from three hours to two hours. In 2001, New York again revised the ratings lowering them to two hours for columns and 90 minutes for floors. The primary purpose of the test we were there to witness, was to determine if the 1/2 inch of fire retardant used in the construction of the WTC, met the 1968 New York Fire Code.

The test would subject the flooring model to a temperature of approximately 1,350 degrees F. during the test, numerous thermocouples to determine heat levels and sensitive instruments designed to detect bowing from the slab would be used. Determining what, if any, bowing of the flooring is considered crucial for understanding why the exterior restrained columns of the trade center towers appeared to bow inward, prior to the structure collapse. The earlier testing in UL’s Canadian lab examined two 35-foot flooring models, one restrained (held in place to inhibit expansion) and one unrestrained. Both spans contained 3/4 of an inch of fireproofing.

The results of those tests indicated that in the restrained model, the flooring would merit a fire rating of 1-1/2 hours. In the unrestrained test, the flooring would merit a fire rating of two hours. The results of this earlier test produced a surprise to engineers who traditionally believed that unrestrained tests would produce lower ratings than restrained testing.

In an earlier test at the Northbrook UL lab, a 17-foot restrained model with 3/4 inch fireproofing was tested and merited a two-hour rating. The test we were to witness was comprised of another 17-foot model with a 1/2-inch coating of fireproofing.

Standing before a battery of monitors and instruments, we watched as the temperature of the UL furnace quickly increased. We could observe the model within the furnace and soon detected a slight bowing in the slab. Such a bowing or sagging would have been experienced in the actual flooring of the WTC as extreme temperatures forced expansion of the flooring. At the conclusion of the test, only a one-hour rating was assessed to the tested span.

The four tests completed so far in this NIST investigation produced fire ratings of between one and two hours for the floor models that were used. Based solely on these findings, one might assume that the 1/2-inch of specified fireproofing would not meet the 1968 New York Fire Code for B-1 structures. In the testing of the 35-foot spans, which used the upgraded 3/4-inch fireproofing, the tests produced a 1-1/2 hour rating in one test, and a two hour rating in the other. This difference in results calls into question the validity of the testing methods used and whether restrained or unrestrained testing methods should be used for fire testing of flooring.

Another question raised by the results is whether a valid test of a large floor span such as the 35-foot and 60-foot span actually used in the WTC, can be accurately performed with the use of a scaled model.

As engineers, we can appreciate the importance of the findings obtained, despite the inconclusive results obtained. Experimentation that results in calling into question the protocol used for testing, or produces dissimilar results depending on the testing methods used, is an achievement in science and an advancement of knowledge. Understandably however, for the families of victims of 9/11 who still ask how and why, such results can prove frustrating.

As building professionals interested in what actually happened on 9/11, many of us may be left with questions as a result of this test. Questions such as:

• What, if any, impact do flooring materials play on the slab and truss fire rating?

• What, if any, impact did the higher heating value of jet fuel have on floor fire ratings?

• What, if any, impact did chimney effect and draft produce on temperature and fire spread?

But it is important to keep in mind that tests conducted by UL for the NIST were not designed to answer these questions. At best, the NIST/UL tests were just one step in creating a single piece to a massive puzzle.

Even if we knew all of the answers to these questions and others that continue to gnaw at us, it is doubtful we will ever really know exactly what happened on September 11, 2001. Too many variables and unknowns will likely confound the most astute hypothesis. Perhaps however, what can be learned is how to build and maintain safer buildings, how to put together better evacuation plans and how to establish lifesaving procedures for first responders.

And in the end perhaps - just perhaps, that may be enough.

Archives