The collapse of WTC 7:
The NIST recommendations, Part 2
By Richard Schulte
Schulte & Associates, Evanston, Ill.
Part I of this series of articles provided an overview and general discussion of the recommendations contained in the final report on the investigation in thecollapse of the WTC 7 Building by the National Institute of Standards and Technology (NIST). This report, referred to as NCSTAR 1A, is dated November 2008. Part II of this article will provide an in-depth discussion and analysis of the NIST recommendations contained in NCSTAR 1A.
Background
Prior to beginning the discussion and analysis of the NIST recommendations, a brief background on the origins of the NIST investigations into the building collapses of the World Trade Center (WTC) towers and the WTC 7 Building on 9/11 is in order. On March 6, 2002, a little less than six months after the collapse of the World Trade Center towers on September 11th, the Congressional Science Committee held a hearing on the collapse of the towers. The opening statement of this hearing by the chairman of the Committee, Congressman Sherwood Boehlert, included the following excerpts:
“The [Congressional Science] Committee decided to move forward [with this hearing] for two fundamental reasons. . . . to ensure that such a catastrophic building failure, and the resulting loss of life, never happen again.”
“But this hearing is not so much about the past, as it is about ensuring that we protect lives in the future.”
One of the five witnesses who testified at the Science Committee hearing was Dr. Arden Bement, the director of NIST. Dr. Bement’s testimony to the Congressional Committee included the following excerpts:
“The implementation of the results of such an investigation [by NIST] would be critical to restore public confidence in the safety of tall buildings nationwide, enhance the safety of fire and emergency responders, and better protect people and property in the future.”
“Fourth, to study procedures and practices used to provide adequate structural reserve capacity to resist abnormal loads (e.g. blast, explosion, impact due to aircraft or flying debris from tornadoes, accidental fires, and faulty design and construction), especially those that can be anticipated prior to construction (e.g. impact of a Boeing 707) . . .”
“This broader program would address critically and urgently needed improvements to national building and fire standards, codes, and practices that have begun to be recognized in recent years. The events of September 11th have brought even more focus and priority to this already important issue.”
“The goal of this broader program would be to produce cost-effective retrofit and design measures and operational guidance for building owners and emergency responders.”
“In short, NIST would provide the technical basis and guidance for fire safety design and retrofit of structures, the predictive tools and test methods for fire resistance determination, and the performance criteria for fireproofing materials. In addition, NIST proposes to develop guidance and retrofit technologies to enhance building egress in emergencies, practical tools and guidance to enhance the safety and effectiveness of fire and emergency responders, and improved models of occupant behavior and response to enhance evacuation and communication in emergencies.”
“Yet, the United States has not developed standards, codes, and practices to assess and reduce this vulnerability. Adding to the problem for modern structures is their smaller margin of safety—and the reserve capacity to accommodate abnormal loads—due to increased efficiency in the use of building materials and refinements in analysis techniques . . .”
“In conclusion, I believe it is imperative for the U.S. to learn from the worst-ever building disasters in human history and take aggressive remedial action to minimize future losses.”
Based upon the testimony of Dr. Bement at the March 6, 2002 hearing, the United States Congress funded the NIST investigations into the collapse of the World Trade Center (WTC) towers and the WTC 7 Building.
The final report on the NIST investigation into the collapse of the World Trade Center towers, referred to as NCSTAR 1, is dated September 2005. This report contained a total of 30 recommendations for making buildings “safer”, specifically high rise buildings. These same recommendations have been repeated in the final report on the collapse of the WTC 7 Building in a condensed format and one new recommendation has been added to the original list.
The Executive Summary included in NCSTAR 1A report contained the following excerpt:
“WTC 7 was a 47 story office building located immediately to the north of the main WTC Complex. . . .On September 11, 2001, WTC 7 endured fires for almost seven hours, from the time of the collapse of the north WTC tower (WTC 1) at 10:28:22 a.m. until 5:20:52 p.m., when it collapsed. This was the first known instance of the total collapse of a tall building primarily due to fires.”
The NCSTAR 1A report also includes the following excerpts:
“The urgency of the prior recommendations is substantially reinforced by their pertinence to the collapse of WTC 7, a tall building that is based on a structural system design which is in widespread use.” (Page 63)
“The partial or total collapse of a building due to fires is an infrequent event. This is particularly true for buildings with a reliably operating active fire protection system such as an automatic fire sprinkler system. (Page 63)
“The intent of current practice, based upon prescriptive standards and codes, is to achieve life safety, not collapse prevention. However, the key premise of NIST’s recommendations is that buildings should not collapse in infrequent fires that may occur when active fire protection systems are rendered ineffective, e.g., when sprinklers do not exist, are not functional, or are overwhelmed by the fire.” (Pages 63 and 64)
“Fire scenarios for structural design based upon a single compartment or single floor fires are not appropriate representations of infrequent event fires. Such events have occurred in several tall buildings resulting in unexpected substantial losses. Instead, historical data suggests that infrequent fires which should be considered in structural design involve: ordinary combustibles and combustible load levels, local fire origin on any given floor, no widespread use of accelerants, consecutive fire spread from combustible to combustible, fire-induced window breakage providing ventilation for continued fire spread and accelerated fire growth, concurrent fires on multiple floors, and active fire protection systems rendered ineffective. The fires in WTC 7 involved all of these.” (Page 64)
With the promises and assurances of the director of NIST on the expected results of the NIST investigations, as well as the Congressional Science Committee’s understanding and expectations for the investigations, as background, an analysis of the NIST recommendations contained in the WTC 7 follows.
Analysis of the NIST recommendations
The recommendations contained in final report on NIST’s investigation into the collapse of the WTC 7 Building are denoted by letters, A through M. The text of each recommendation will be followed by an analysis of the recommendation.
“Recommendation A (NIST NCSTAR Recommendation 1). NIST recommends that: (1) progressive collapse be prevented in buildings through the development and nationwide adoption of consensus standards and code provisions, along with the tools and guidelines needed for their use in practice; and (2) a standard methodology be developed - supported by analytical design tools and practical design guidance - to reliably predict the potential for complex failures in structural systems subjected to multiple hazards.
Relevance to WTC 7: Had contemporaneous standards and practices been available to expressly design WTC 7 for prevention of fire-induced progressive collapse, it would have been sufficiently robust to withstand local failure due to fires without suffering total collapse.”
Analysis-recommendation A: NIST’s recommendation regarding progressive collapse states that building collapse can be “prevented”. In the real world, engineers can only minimize the probability of progressive collapse, but it is simply not possible to reduce this probability to zero. If buildings are designed to prevent progressive collapse in the event of the failure of a single column, progressive collapse could still occur if two or more adjacent columns fail. Hence, the NIST recommendation regarding progressive collapse prevention contains an unstated assumption regarding the magnitude of the initiating event which would cause a failure of the vertical supports provided for a building. Obviously, any building structure can be made to collapse given a large enough initiating event.
Unfortunately, the NIST recommendation does not address the types of initiating failure events for which a building structure should be designed. Dr. Bement’s testimony at the Congressional Science Committee hearing on March 6, 2002 suggested that tall buildings should be designed for the impact of aircraft. It should be noted, however, that Dr. Bement did not provide any further elaboration on this subject. Obviously, the size of the aircraft, the speed of the aircraft at impact and the amount of fuel carried by the aircraft would play a major role in the damage a building would sustain and have a significant impact on the design requirements for tall buildings.
Fortunately, even NIST realized that designing a building to prevent collapse due to the structural damage caused by the impact of a large commercial aircraft at the speeds that the terrorists flew the planes into the WTC towers is simply not economically feasible. Talk of designing tall buildings for aircraft impact appears to have ceased as soon as NIST received funding for its investigation.
It should have been obvious to the structural engineers at NIST that Dr. Bement’s comments at the Congressional hearings were only intended to persuade the Science Committee into thinking that engineers can prevent the collapse of buildings under assault by aircraft used by terrorists as missiles. This tactic is commonly referred to by engineers and researchers specializing in sales as “bait and switch”. The assumption was that by the time the NIST investigations were completed, members of the Science Committee would have long forgotten the promises made in order to secure the funding.
There is no evidence in the records published on the NIST website that NIST ever went back to the Congressional Science Committee and corrected the Science Committee’s understanding that NIST would address the collapse of buildings being assaulted by aircraft intentionally flown into the building. Given the wording of Recommendation A, however, it is obvious that it is NIST’s intent not to suggest that tall buildings be designed to prevent collapse subsequent to an intentional high-speed impact by one, or perhaps more than one, large commercial aircraft flying at high speed. Still, it would have been helpful for the public’s understanding of the collapse of the towers and the intent of NIST’s recommendations if NIST had specifically included a section in the investigation report addressing the issue of building collapse due to aircraft impact.
“Recommendation B (New), NIST recommends that buildings be explicitly evaluated to ensure the adequate performance of the structural system under maximum credible (infrequent) design fires with any active fire protection system rendered ineffective. Of particular concern are the effects of thermal expansion in buildings with one or more of the following features: (1) long-span floor systems which experience significant thermal expansion and sagging effects, (2) connection designs (especially shear connections) that cannot accommodate thermal effects, (3) floor framing that introduces asymmetric thermally-induced (i.e., net lateral) forces on girders, (4) shear studs that could fail due to differential thermal expansion in composite floor systems, and (5) lack of shear studs on girders. Careful consideration should also be given to the possibility of other design features that may adversely affect the performance of the structural systems under fire conditions.”
“Building owners, operators, and designers are strongly urged to act upon this recommendation. Engineers should be able to design cost-effective fixes to address any areas of concern that are identified by these evaluations. Several existing, emerging or even anticipated capabilities could have helped prevent the collapse of WTC 7. The degree to which these capabilities improve performance remains to be evaluated. Possible options for developing cost-effective fixes include:
• More robust connections and framing systems to better resist the effects of thermal expansion on the structural system.
• Structural systems expressly designed to prevent progressive collapse. The current model building codes do not require that buildings be designed to resist progressive collapse.
• Better thermal insulation (i.e., reduced conductivity and/or increased thickness) to limit heating of structural steel and to minimize both thermal expansion and weakening effects. Currently, insulation is used to protect steel strength, but it could be used to maintain a lower temperature in steel framing to limit thermal expansion.
• Improved compartmentation in tenant areas to limit the spread of fires.
• Thermally resistant window assemblies which limit breakage, reduce air supply, and retard fire growth.
Industry should partner with the research community to fill critical gaps in knowledge about how structures perform in real fires, particularly considering: the effects of fire on the entire structural system; the interactions between subsystems, elements, and connections; and scaling of fire test results to full-scale structures, especially structures with long span floor systems.
Relevance to WTC 7: The effects of restraint of free thermal expansion on the steel framing systems, especially for the long spans on the east side of WTC 7, were not considered in the structural design and led to the initiation of the building collapse.”
Analysis-Recommendation B: As was pointed out by NIST in the Executive Summary of NCSTAR 1A, the collapse of the WTC 7 Building “was the first known instance of the total collapse of a tall building primarily due to fires.” It should be further noted that fires in the WTC 7 Building were apparently caused by the collapse of the WTC 1 tower and that the collapse of the towers damaged the municipal water supply at the site. Further, the fire department (FDNY) made a decision not to fight the fires in the WTC 7 Building following the loss of a large number of fire fighters in the collapse of both towers. The failure of the water supply at the WTC site and the decision by the fire department not to fight the fires are also the first known instances of these occurrences at a major fire in a high rise building in the United States. Would the collapse of the WTC 7 Building have occurred if the water supply at the site had not been impaired and if the fire department had attempted to manually control the fires? NIST has avoided addressing this question in their investigation report, but the answer to this question seems obvious. Of course, a manual attack on the fires in the WTC 7 Building supported by an adequate water supply would have prevented the collapse of the building. Given the above, the foundation for Recommendation B in NIST’s report is certainly questionable.
A careful review of Recommendation B indicates that this recommendation does not really address the issue of manual fire fighting operations and whether the capabilities of the fire department should be ignored in the design of high rise buildings. The most recent study on the reliability of sprinkler systems published by the National Fire Protection Association (NFPA) in January 2009 indicates that wet pipe sprinkler systems protecting office, residential, assembly and mercantile occupancies operate and are effective in excess of 97 percent of fires largest enough to activate sprinklers. A study of the historical records of fires in high rise buildings indicates that a major fire has never occurred in a high rise building protected throughout by a sprinkler system in the United States at any time other than on September 11th. While the reliability of sprinkler systems protecting high rise buildings in the United States is not 100 percent, the reliability of sprinkler systems in combination with structural fire protection presently required by model building codes and manual fire fighting operations can be considered to be 100 percent effective, except in the event of a terrorist attack. Hence, the need to further enhance the structural fire protection required for high rise buildings and to upgrade the structural fire protection presently provided for existing high rise buildings should be questioned.
Perhaps, the capital spent on enhancing the structural fire protection of new high rise buildings and upgrading the structural fire protection of existing buildings as recommended by NIST would be more productively spent on maintaining the sprinkler system in an operating condition and on maintaining the structural fire protection which is presently required. It seems to me that enhancing or upgrading the structural fire protection without maintaining the structural fire protection is a rather foolish approach to building fire protection. If both the sprinkler protection and structural fire protection are both properly maintained, then the probability of a collapse is so minute that it is essentially nil. Given this, NIST’s recommendation seems to be a misdirected approach to fire safety in high rise buildings. Certainly, there are other more probable life safety concerns which need to be addressed than the issue of the collapse of high rise buildings protected by properly maintained sprinkler protection and structural fire protection.
“Recommendation C (NIST NCSTAR 1 Recommendation 4). NIST recommends evaluating, and where needed improving, the technical basis for determining appropriate construction classification and fire rating requirements (especially for tall buildings) - and making related code changes now as much as possible - by explicitly considering factors including:3
• timely access by emergency responders and full evacuation of occupants, or the time required for burnout without partial collapse;
• the extent to which redundancy in active fire protection (sprinkler and standpipe, fire alarm, and smoke management) systems should be credited for occupant life safety;
• the need for redundancy in fire protection systems that are critical to structural integrity;
• the ability of the structure and local floor systems to withstand a maximum credible fire scenario without collapse, recognizing that sprinklers could be compromised, not operational, or non-existent;
• compartmentation requirements (e.g., 12,000 ft2) to protect the structure, including fire rated doors and automatic enclosures, and limiting the air supply (e.g., thermally resistant window assemblies) to retard fire spread in buildings with large, open floor plans;
• the effect of spaces containing unusually large fuel concentrations for the expected occupancy of the building; and
• the extent to which fire control systems, including suppression by automatic or manual means, should be credited as part of the prevention of fire spread.
“Footnote 3. The construction classification and fire rating requirements should be risk-consistent with respect to the design-basis hazards and the consequences of those hazards. The fire rating requirements, which were originally developed based on experience with buildings fewer than 20 stories in height, have generally decreased over the past 80 years since historical fire data for buildings suggests considerable conservatism in those requirements. For tall buildings, the likely consequences of a given threat to an occupant on the upper floors are more severe than the consequences to an occupant on the first floor or the lower floors. For example, with non-functioning elevators, both the time requirements are much greater for full building evacuation from upper floors and emergency responder access to those floors. The current height and area tables in building codes do not provide the technical basis for the progressively increasing risk to an occupant on the upper floors of tall buildings that are much greater than 20 stories in height.”
“Relevance to WTC 7: The floor systems in WTC 7 failed because thermal effects within the structural system, especially thermal expansion, were not considered in setting the fire rating requirements in the construction classification, which are determined using the ASTM E119 or equivalent testing standard. Such evaluation is not required under current codes and standards.”
Analysis-Recommendation C: This recommendation is based, in least in part, on the two assumptions. The first assumption is that life safety risk to occupants is greater on higher floors than on lower floors. The second assumption is that NIST’s hypothesis on cause of the collapse of the WTC 7 Building is correct. Both of these assumptions are questionable at best.
While a case can be made that the “risk” to occupants on higher floors is greater than the “risk” to occupants on lower floors of high rise buildings, the case supporting this conclusion is actually rather superficial. Are occupants on 60th floor of a high rise building really at greater “risk” than occupants on the 50th floor of a building? I suppose that the answer to that question is technically yes (assuming that all other things are equal), however, the difference in the level of risk is so minimal that the level of risk can be considered to be essentially the same. In other words, the difference in the level of risk would be similar to the difference between driving the same car at a speed of 66 miles per hour, rather than 65. Certainly, the difference in risk is not the same as driving at 85 miles an hour, rather than 65 miles per hour.
When the difference in risk to occupants of a 60 story building is compared to the risk of occupants of a 10 story building, one of the factors which is usually not considered is the management and security staffs of the buildings. Buildings which are 60 stories, 70 stories or even taller will have far larger and more profession management and security staffs than a 10 story building. Given this, the attention to maintenance of safety systems will be far greater in tall buildings than in buildings of lesser heights. Given this, a good case can be made that the “risk” to occupants of tall buildings is actually less than the “risk” in buildings of lesser heights.
Of course, fire safety statistics support this conclusion. Where do most of the fire fatalities in the United States occur? Why, of course, it’s 1- and 2-family dwellings and these dwellings are typically one and two stories in height. High rise buildings are actually one of the “safest” types of buildings which we construct, regardless of whether the building is 8 stories in height or 100 stories in height. If you don’t believe it, take a look at the fire fatality statistics for high rise buildings. The fire safety record for high rise buildings in the United States is simply magnificent.
With respect to NIST’s assumption that their hypothesis regarding the cause of the collapse of the WTC 7 Building is correct, another potential cause for the collapse could have been that the structural fire protection for the building was either improperly applied when the building was constructed, or that the structural fire protection was not maintained. It should be pointed out that NIST did not explain how it determined the condition of the structural fire protection immediately prior to the terrorist attack, however, NIST spent considerable time discussing this issue in its report on the collapse of the WTC towers. Could missing structural fire protection actually been the cause of the collapse of the WTC 7 Building? We will never know the answer to that question, but, from an experience and probability standpoint, there is little doubt that it was highly unlikely that all of the structural fire protection required for the WTC 7 Building was actually in place immediately prior to the fires.
With respect to collapse prevention, if the structural fire resistance is equal to the fire exposure which will be generated by the contents (assuming noncombustible building construction), and the structural fire protection is properly installed and maintained, collapse of the building should still be prevented in the event of the failure of the sprinkler system and failure of the fire department to initiate manual fire fighting operations. In other words, both the sprinkler system and manual fire fighting operations can be considered to be factors of safety. Given the historical record of high rise building fires in the United States, these factors of safety have been more than adequate to prevent the collapse of a high rise building due to fire. There is simply no need to increase the structural fire resistance provided as the height of high rise buildings increase because the probability of collapse does not increase with the height of the building. As indicated above, it is my opinion that the probability of structural collapse actually decreases (although only marginally) with the increasing height of a building. Hence, it’s my opinion that NIST’s Recommendation C is based upon fundamentally flawed thinking both about risk and probability.
“Recommendation D (NIST NCSTAR 1 Recommendation 5). NIST recommends that the technical basis for the century-old standard for fire resistance testing of components, assemblies, and systems be improved through a national effort. Necessary guidance also should be developed for extrapolating the results of tested assemblies to prototypical building systems. A key step in fulfilling this recommendation is to establish a capability for studying and testing the components, assemblies and systems under realistic fire and load conditions.
Of particular concern is that the Standard Fire Resistance Test does not adequately capture important thermally-induced interactions between structural sub-systems, elements, and connections that are critical to structural integrity. System-level interactions, especially due to thermal expansion, are not considered in the standard test method since columns, girders, and floor assemblies are tested separately. Also, the performance of connections under both gravity and thermal effects is not considered. The United States currently does not have the capability for studying and testing these important fire-induced phenomena critical to structural safety.
Relevance to WTC 7: The floor systems failed in WTC 7 at shorter fire exposure times than the specified fire rating (two hours) and at temperatures lower than the endpoint temperature (593oC, 1100oF) because thermal effects within the structural system, especially thermal expansion, were not considered in setting the endpoint criteria when using the ASTM E119 or equivalent testing standard. The structural breakdowns that led to the initiating event and the eventual collapse of WTC 7 occurred at temperatures that were hundreds of degrees below the criteria that determines structural fire resistance ratings.”
Analysis-Recommendation D: Once again, the basis for making this recommendation is NIST’s hypotheses on the cause of the collapse of the WTC 7 Building. The certainty that this hypotheses is actually correct has already been addressed above. NIST statements regarding ASTM E119, the fire test used for establishing fire resistance ratings in the United States, are correct, however, when the protection of the building structure provided by sprinkler protection and manual fire fighting operations are also considered, ASTM E119 is actually a very conservative measure of the structural fire protection provided. The only way that ASTM E119 cannot be considered to be conservative is if the protection provided by sprinkler systems and by manual fire fighting operations are ignored. NIST has simply not made a case on why the fire protection design for tall buildings should be based upon the assumption that both sprinkler protection and manual fire fighting operations will both fail. There is a far higher probability that the structural fire protection provided for tall buildings will fail (when considered by itself) than the probability of failure of either sprinkler protection or manual fire fighting operations considered separately.
“Recommendation E (NIST NCSTAR 1 Recommendation 7). NIST recommends the adoption and use of the “structural frame” approach to fire resistance ratings. This approach requires all members that comprise the primary structural frame (such as columns, girders, beams, trusses, and spandrels) be fire protected to the higher fire resistance rating required for the columns. The definition of the primary structural frame should be expanded to include bracing members that are essential to the vertical stability of the primary structural frame under gravity loading (e.g. girders, diagonal bracing, composite floor systems that provide lateral bracing to the girders) whether or not the bracing members carry gravity loads. Some of these bracing members may not have direct connections to the columns, but provide stability to those members directly connected to the columns. . . .This recommendation ensures consistency in the fire protection provided to all structural elements that contribute to overall stability. State and local jurisdictions should adopt and enforce this requirement.
Relevance to WTC 7: Thermally-induced breakdown of the floor system in WTC 7 was a determining step in causing collapse initiation and progression. Therefore, the floor system should be considered as an integral part of the primary structural frame.”
Analysis-Recommendation E: The scope of NIST’s investigation was limited to the collapse of the WTC 7 Building. It would have been of interest if NIST had also looked into the major fires which occurred at the First Interstate Bank Building in Los Angeles in May 1988 and the One Meridian Plaza Building in Philadelphia in February 1991. A comparison of the original design of the First Interstate Bank Building and the One Meridian Plaza Building to the design of the WTC 7 Building could have determined why both of these buildings were able to continue to support their structural loads while the WTC 7 Building structure failed. (All three buildings were steel-framed buildings.) Both the First Interstate Bank Building and the One Meridian Plaza Building were originally constructed without sprinkler protection and were Type I construction as defined by the model codes which were in use in Los Angeles and Philadelphia. (The City of Los Angeles used an amended version of the Uniform Building Code when the First Interstate Bank Building was being designed while the City of Philadelphia used an amended version of the BOCA Basic Building Code when the One Meridian Plaza Building was being designed. It should be noted that both of these buildings were being retrofitted with sprinkler protection at the time the fires occurred, however the sprinkler system installations in both buildings had not been completed at the time of the fire.) Both of these buildings were constructed with 2 hour floor assemblies based upon the code requirements which were applicable.
Although the fire in the First Interstate Bank Building spread to a total of 5 floors and the fire in the One Meridian Plaza Building spread to a total of 16 floors, neither building collapsed. In the case of the First Interstate Bank Building, the building was re-occupied after a few months of cleaning, while the fire damage to the One Meridian Plaza Building made the building structurally unstable and the building had to be demolished. If the First Interstate Bank Building and the One Meridian Plaza Building could survive multi-floor fires with only two hour floor construction, the fires in these two buildings appear to demonstrate that there is no need to provide three hour floor construction as recommended by NIST in Recommendation E.
“Recommendation F ([NIST] NCSTAR [1] Recommendation 8). NIST recommends that the fire resistance of structures be enhanced by requiring a performance objective that uncontrolled building fires result in burnout without partial or global (total) collapse. Such a provision should recognize that sprinklers could be compromised, non-operational, or nonexistent. Current methods for determining the fire resistance of structural assemblies do not explicitly specify a performance objective. The rating resulting from current test methods indicates that the assembly (component or subsystem) continued to support its superimposed load (simulating a maximum load condition) during the test exposure without collapse.
Model Building Codes: This recommendation should be included in the national model codes as an objective and adopted as an integral part of fire resistance design for structures. The issue of non-operational sprinklers could be addressed using the existing concept of Design Scenario 8 of NFPA 5000, where such compromise is assumed and the result is required to be acceptable to the Authority Having Jurisdiction.
Analysis-Recommendation F: Building codes used in the United States mandate that sprinkler protection be provided in all high rise buildings. In addition, codes mandate that a standpipe system also be provided in high rise buildings in event that the sprinkler system fails to control the fire. Beyond these requirements, codes also mandate that high rise buildings be provided with two hour floor construction to prevent the rapid spread of fire between floors and further require that the structural frame of a high rise building develop either a two or three hour fire resistance rating. To further prevent the spread of fire between floors, codes mandate that floor openings be enclosed in fire resistive enclosures. NIST’s Recommendation F essentially assumes that all of the mandated protection listed above will fail with no discussion regarding the failure rates of these various means of preventing fire spread between floors. In effect, NIST’s Recommendation F simply adds another layer of protection without an analysis of the need for this new layer of protection and whether or not this new layer of protection will provide any measurable increase in the level of safety already provided.
The more common sense alternative to implementing Recommendation F would be recommend that the required means of protection already included in codes actually be maintained so that a new layer of protection is not necessary. Based upon our 35 years of experience with providing sprinkler protection in high rise buildings, we know that sprinkler protection is a highly reliable form of fire protection for high rise buildings. Sprinkler protection in high rise buildings can be made even more effective by requiring duplicate pumps and requiring a secondary water supply for each sprinkler system zone in a high rise building. Implementing these changes in the design of sprinkler systems for buildings should increase the reliability of sprinkler protection in high rise buildings to close to 100 percent. Implementing changes to the design requirements outlined in this paragraph would essentially eliminate the need to implement NIST Recommendation F and would be far less costly than the implementation of this recommendation.
Based upon our experience with the fires at the First Interstate Bank Building and at the One Meridian Plaza Building, we now know how to prevent the spread of fire from floor-to-floor in a tall building in the event of a sprinkler system failure. By simply “pre-wetting” the combustibles on floors above the fire, manual fire fighting operations can prevent the spread of fire to multiple floors in tall buildings. Hence, there should never be a need to address the spread of fire to more than two floors in a tall building, unless we assume simultaneous sprinkler system failure and failure of fire fighters to show up at a fire. Based upon the performance of fire fighters in the WTC towers on September 11th, it is my opinion that fire fighters are going to make every effort to battle a fire in a tall building. If the collapse of the WTC towers and the death of hundreds of fire fighters had not preceded the fires in the WTC 7 Building, the WTC 7 Building would still be standing today.
The capital spent on complying with NIST Recommendation F could better be spent on highway safety. In all probability, the implementation of this recommendation will likely not save even a single life, while investing the capital which would be spent implementing this recommendation on highway safety could save thousands, perhaps hundreds of thousand of lives. The thousands of lives which could be saved by investing capital wisely is a high price for the American public to pay simply to implement Recommendation F.
“Recommendation G (NIST NCSTAR 1 Recommendation 9). NIST recommends the development of: (1) performance-based standards and code provisions, as an alternative to current prescriptive design methods, to enable the design and retrofit of structures to resist real building fire conditions, including their ability to achieve the performance objective of burnout without structural or local floor collapse:[;] and (2) the tools, guidelines, and test methods necessary to evaluate the fire performance of the structure as a whole system. . . .
This performance-based capability should include the development of, but not limited to:
a. Standard methodology, supported by performance criteria, analytical design tools, and practical design guidance; related building standards and codes for fire resistance design and retrofit of structures, working through the consensus process for nationwide adoption; comprehensive design rules and guidelines; methodology for evaluating thermostructural performance of structures; and computational models and analysis procedures for use in routine design practice.
b. Standard methodology for specifying multi-compartment, multi-floor fire scenarios for use in design and analysis of structures to resist fires, accounting for building-specific conditions such as geometry, compartmentation, fuel load (e.g., building contents and any flammable fuels such as oil and gas), fire spread, and ventilation; and methodology for rating the fire resistance of structural systems and barriers under realistic design-basis scenarios.
c. Publicly available computational software to predict the effects of fires in buildings-developed, validated, and maintained through a national effort-for use in the design of fire protection systems and the analysis of building response to fires. Improvements should include the fire behavior and contribution of real combustibles; the performance of openings, including door openings and window breakage, that controls the amount of oxygen available to support the growth and spread of fires and whether the fire is fuel-controlled or ventilation-controlled; the floor-to-floor flame spread; the temperature rise in both insulated and uninsulated structural members and fire barriers; and the structural response of components, subsystems, and the total building system due to fire.
d. Temperature-dependent thermal and mechanical property data for conventional and innovative construction materials[.]
e. New test methods, together with associated conformance assessment criteria, to support the performance-based methods for fire resistance design and retrofit of structures. The performance objective of burnout without collapse will require the development of standard fire exposures that differ from those currently used.
Relevance to WTC 7: A performance-based assessment of the effects of fire on WTC 7, had it considered all of the relevant thermal effects (e.g., thermal expansion effects that occur at lower temperatures), would have identified the vulnerability of the building to fire-induced collapse and allowed alternative designs for the structural system.”
Analysis-Recommendation G: This recommendation proposes that owners of existing tall buildings have the structural systems of these building analyzed to determine whether or not the buildings are vulnerable to a collapse mechanism similar to that which NIST hypothesizes occurred in the WTC 7 Building. Given that the probability of a collapse mechanism similar to NIST hypothesis occurring in other buildings is practically zero and the aggregate cost of upgrading existing building can be conservatively estimated to be in the billions of dollars, the implementation of this recommendation simply cannot be justified based upon a rational cost-benefit analysis.
The cost to implement this recommendation will far exceed the potential for loss in the event the WTC 7 fire scenario should ever occur again. Upgrading existing buildings to address the WTC 7 Building fire-induced collapse scenario is simply an irrational response to the collapse of this building. As previously stated, the capital invested in implementing this recommendation would be better spent on highway safety in the United States.
“Recommendation H (NIST NCSTAR 1 Recommendation 12). NIST recommends that the performance and possibly the redundancy and reliability of active fire protection systems (sprinklers, standpipe/hoses, fire alarms, and smoke management systems) in buildings be enhanced to accommodate the greater risks associated with increasing building height and population, increased use of open spaces, high-risk building activities, fire department response limits, transient fuel loads, and higher threat profile.
Reliability is affected by (a) the redundancy such that when one water supply is out of service (usually for maintenance) the other interconnected water supply can continue to protect the building and its occupants, (b) automatic operation of water systems (not only for starting fire pumps but also for testing and tank replenishment with appropriate remote alarms to the fire department and local alarms for notifying emergency personnel), © the use of suitable equipment and techniques to regulate unusual pressure considerations. Relevance to WTC 7: No water was available for the automatic suppression system on the lower 20 stories of WTC 7 once water from street-level was disrupted. This lack of reliability in the source of the primary and secondary water supplies allowed the growth and spread of fires that ultimately resulted in the collapse of the building.
Analysis-Recommendation H: Of all of the recommendation made by NIST, the implementation of this recommendation can be accomplished relatively easily and at minimal cost. Given this, the implementation of this recommendation can be rationally justified. If this recommendation is implemented, it would seem that the implementation of many of the other recommendations made by NIST is simply unnecessary.
“Recommendation I (NIST NCSTAR 1 Recommendation 24). NIST recommends the establishment and implementation of codes and protocols for ensuring effective and uninterrupted operation of the command and control system for large-scale building emergencies.
a. State, local, and federal jurisdictions should implement the National Incident Management System (NIMS). The jurisdictions should work with the Department of Homeland Security to review, test, evaluate, and implement an effective unified command and control system. NIMS addresses interagency coordination and establishes a response matrix - assigning lead agency responsibilities for different types of emergencies and functions. At a minimum, each supporting agency should assign an individual to provide coordination with the lead agency at each incident command post.
b. State, local, and federal emergency operation centers (EOCs) should be located, designed, built, and operated with security and operational integrity as a key consideration.
c. Command posts should be established outside the potential collapse footprint of any building which shows evidence of large multi-floor fires or has serious structural damage. A continual assessment of building stability and safety should be made in such emergencies to guide ongoing operations and enhance emergency responder safety. The information necessary to make these assessments should be made available to those assigned the responsibility. . . .
d. An effective command system should be established and operating before a large number of emergency responders and apparatus are dispatched and deployed. Through training and drills, emergency responders and ambulances should be required to await dispatch requests from the incident command system and not to self-dispatch in large-scale emergencies.
e. Actions should be taken via training and drills to ensure a coordinated and effective emergency response at all levels of the incident command chain by requiring all emergency responders that are given an assignment to immediately adopt and execute the assignment objectives.
f. Command post information and incident operations data should be managed and broadcast to command and control centers at remote locations so that information is secure and accessible by all personnel needing the information. Methods should be developed and implemented so that any information that is available at an interior information center is transmitted to an emergency responder vehicle or command post outside the building.
Relevance to the WTC 7: (1) The New York City Office of Emergency Management (OEM) was located in WTC 7 and was evacuated early in the day before key fire ground decisions had to be made. The location of OEM in WTC 7 contributed to the loss of robust interagency command and control on September 11, 2001. (2) Due to the collapse of the WTC towers and the loss of responders and fire control resources, there was an evolving site leadership during the morning and the afternoon. Key decisions (e.g., decisions not to fight the fire in WTC 7 and to turn off the power to the Con Edison substation) were reasonable and would not have changed the outcome on September 11, 2001, but were not made promptly. Under different circumstances (e.g., if WTC 7 had collapsed sooner and fire fighters were still evaluating the building condition), the outcome could have been very different.”
Analysis-Recommendation I: This NIST recommendation essentially addresses our emergency response infrastructure and addresses the question as to how large of an infrastructure is necessary to protect the American public. Should our emergency response infrastructure in the United States be large enough to handle any conceivable disaster which could potentially occur or should we limit the size of the disaster for which we are prepared to respond? Of course, the size of our emergency response infrastructure is directly related to the cost of providing this infrastructure.
Each community, and every person in the community, essentially addresses this issue on a daily basis. The number of police officers and fire fighters on duty, the apparatus and equipment available to emergency responders, the size of roads (necessary for full evacuation of a city or town), etc. all play a part in emergency response. As with everything else in life, cost is also a factor in emergency preparedness and the more prepared we are to handle any emergency or contingency, the more probable that this preparedness will be never be utilized (e.g. nuclear “fall-out” shelters stocked with emergency food and water supplies in the 1950's and 1960's). The issue of emergency response preparedness deserves a frank and public discussion. Unfortunately, NIST’s recommendation has not “sparked” this discussion and given the state of the American economy at this point in time, it is doubtful that this discussion will likely occur anytime soon. It’s rather interesting that the discussions caused by the events on 9/11 have been drowned out by the economic “disaster” which has occurred recently. Perhaps, it is unfortunate that the economic disaster didn’t occur simultaneous with the release of the NIST recommendations in the Fall of 2005. Our perspective on the importance of implementing the NIST recommendations would certainly been different if that had occurred.
“Recommendation J (NIST NCSTAR 1 Recommendation 27). NIST recommends that building codes incorporate a provision that requires building owners retain documents, including supporting calculations and test data, related to building design, construction, maintenance and modifications over the life of the building. Means should be developed for offsite storage and maintenance of the documents. In addition, NIST recommends that relevant building information be made available in suitably designed hard copy or electronic format for use by emergency responders. Such information should be easily accessible by responders during emergencies.
Relevance to WTC 7: The efforts required in locating and acquiring drawings, specifications, tenant layouts, and material certifications, and especially shop fabrication drawings, significantly lengthened the investigation into the collapse of WTC 7.”
Analysis-Recommendation J: In theory, this recommendation is easily implemented and compliance with this requirement is relatively easy, however, simply including a requirement in building codes that records must be retained by building owners does not necessarily mean that building owners will comply. Hence, there is a need for NIST to indicate how NIST envisions that code provisions relating to the implementation of this recommendation will actually be enforced.
From the standpoint of the use of this information by emergency responders, providing reams of information on a building may be of little assistance in the event of an emergency. Imagine a battalion chief in the lobby of one of the WTC towers on September 11 attempting to sort through giga-bytes of information on the towers. Our capability to store information exceeds our capabilities to retrieve necessary information in a timely manner. Given this, careful consideration on how this recommendation should best be implemented is required.
With respect to future investigations, requiring building owners to keep every scrap of information that could be useful to NIST in an investigation will be a waste of time for close to one hundred percent of building owners since close to 100 percent of buildings will likely never be involved in a disaster. The aggregate cost of all building owners complying with this recommendation will be in the billions. It is likely that NIST would not have made this recommendation if NIST was required to compensate building owners for the cost of implementing this recommendation. It is likely that the cost associated with implementing this recommendation would far exceed NIST’s entire budget each year.
It’s far easier to spend someone else’s money than it is to spend your own. It is unlikely that this recommendation would have been made if the researchers involved in this investigation were required to compensate building owners for the costs involved directly from their own paychecks. Perhaps, NIST (and other government regulators) should establish this concept as a new standard for making recommendations and developing new regulations.
“Recommendation K (NIST NCSTAR 1 Recommendation 28). NIST recommends that the role of the “Design Professional in Responsible Charge” be clarified to ensure that: (1) all appropriate design professionals (including, e.g., the fire protection engineer) are part of the design team providing the standard of care when designing buildings employing innovative or unusual fire safety systems, and (2) all appropriate design professionals (including, e.g., the structural engineer and the fire protection engineer) are part of the design team providing the standard of care when designing the structure to resist fires, in buildings that employ innovative or unusual structural and fire safety systems.
Relevance to WTC 7: Following typical practice, none of the design professionals in charge of the WTC 7 project (i.e., architect, structural engineer, and fire protection engineer) was assigned the responsibility to explicitly evaluate the fire performance of the structural system. Holistic consideration of thermal and structural factors during the design or review stage could have identified the potential for the failure and might have prevented the collapse of the building.”
Analysis-Recommendation K: Given the frequency of fire-induced collapses of tall buildings, it seems reasonable to ask whether the implementation of this recommendation is actually necessary. The capital expended complying with this recommendation could be better spent investing in the inspection of the fire safety systems as a tall building is being constructed and in maintaining the fire safety systems after the construction is completed. The implementation of this recommendation will not measurably increase the level of safety provided for tall buildings.
“Recommendation L (NIST NCSTAR 1 Recommendation 29). NIST recommends that continuing education curricula be developed and programs be implemented for (1) training fire protection engineers and architects in structural engineering principles and design, and (2) training structural engineers, architects, fire protection engineers, and code enforcement officials in modern fire protection principles and technologies, including fire-resistant design of structures, and (3) training building regulatory and fire service personnel to upgrade their understanding and skills to conduct the review, inspection and approval tasks for which they are responsible. The outcome would further the integration of the disciplines in effective fire-safe design of buildings.
Relevance to WTC 7: Discerning the fire-structure interactions that led to the collapse of WTC 7 required research professionals with expertise in both disciplines. Assuring the safety of future buildings will require that participants in the design and review processes possess a combined knowledge of fire science, materials science, heat transfer, and structural engineering and design.”
Analysis-Recommendation L: Upgrading the level of education of architects, engineers and code enforcement personnel is certainly important, however, just as important is the realization that maintenance of fire safety systems in buildings is essential for maintaining the safety of the building occupants. One means of increasing awareness that maintenance of building fire safety systems is important is for the fire service to get more involved in fire prevention inspection and code enforcement activities. Without support for fire prevention activities by the fire service, it is unlikely that building owners and the public will take fire safety system maintenance as seriously as is necessary for these systems to function properly in a fire.
“Recommendation M (NIST NCSTAR 1 Recommendation 30). NIST recommends that academic, professional short-course, and web-based training materials in the use of computational fire dynamics and thermostructural analysis tools be developed and delivered to strengthen the base of available technical capabilities and human resources.
Relevance to WTC 7: NIST stretched the state-of-the-art in the computational tools needed to reconstruct a fire-induced building collapse. This enabled identification of the critical processes that led to that collapse. Making these expanded tools and derivative, validated, and simplified modeling approaches usable by practitioners could prevent future disasters.”
Analysis-Recommendation M: At present, many architects, engineers and code enforcement personnel are perplexed by the complexity of code provisions. Introducing an even more complex analyses into the code compliance system will likely not improve the level of compliance with code requirements. Hence, NIST’s recommendation is simply not a realistic solution to the “problem” of fire safety in high rise buildings.
The maintenance of the fire protection systems presently mandated by building codes is both the simplest and least costly means of protecting new and existing high rise buildings. It is also the most reliable means of providing protection for high rise buildings. Rather than attempting to implement a new “exotic” form of structural analysis for high rise buildings, the simplest, and also the least costly solution is far more likely to achieve the goal of preventing fire-induced building collapse.
Conclusion
The Executive Summary of NCSTAR 1A states that the collapse of the WTC 7 Building “was the first know instance of the total collapse of a tall building primary due to fires.” This same report go on to state that “the partial or total collapse of a building due to fires is an infrequent event.” NIST is correct on both of these counts. Given this, the question which is in need of further discussion and debate is: how do we respond to such low frequency events? One response, of course, is to simply do nothing and accept the risk, while the polar opposite of this response is to address the risk, regardless of the cost. Since NIST has acknowledged that they have not done a cost/benefit analysis on any of their recommendations, it appears that NIST has adopted the second option.
Every day, Americans make risk assessments either consciously or unconsciously. We drive cars, fly in airplanes, cross streets, ride bicycles, invest in the stock market, invest in real estate and put money in banks. Most Americans consider driving, flying, bicycling and crossing the street to be “safe” activities and, up until a few months ago, most Americans considered investing in the stock market and real estate to be the smart thing to do. Today, putting money in a bank could even be considered to be a “risky” proposition.
What the NIST reports on the collapse of WTC towers and the WTC 7 Building failed to do was to put the collapses of these buildings into perspective. From a statistical standpoint, the probability of dying as a result of driving, flying, crossing the street in an urban area or riding a bicycle in heavy traffic is far greater than the probability of dying as the result of a collapse of a high rise building caused by a fire or a terrorist attack. Similarly, the aggregate cost of traffic accidents and plane crashes in the United States far exceeds the cost of the collapse of a high rise building in an urban area. In effect, what NIST has asked us to do with their recommendations is to divert capital from more important and pressing safety issues in the United States in order to prevent a “freak” event from occurring once again sometime in the future.
Should we, as a nation, divert capital which could be used to save thousands of lives, perhaps even hundreds of thousands of lives, in order to implement the NIST recommendations, which in all likelihood, will not even save a single life in the future? The common sense answer to this question is obvious.
Richard Schulte is a graduate of the Fire Protection Engineering Program at the Illinois Institute of Technology. He formed Schulte & Assoc. in 1988. His consulting experience includes work on the Sears Tower and many other notable structures. He also has acted as an expert witness in the litigation involving the fire at the New Orleans Distribution Center.