Building codes in the U.S. are at a crossroads. Until recently, we took the safety of our buildings for granted, confident that designs meeting code requirements would perform well. However, the collapse of 60% of the Murrah Federal Building in Oklahoma City in 1995 and destruction of the World Trade Center towers in 2001 shook that confidence, not only with the general public, but with engineers as well.

I propose a new path for building codes–performance design within a unified U.S. code. Instead of a ‘one-size-fits-all' code mandate, performance design requirements can vary, based on the hazards being considered, the owner's risk tolerance and the user's need for building functionality after an event.

Performance design already is being used for sensitive facilities in seismic regions. Extending it to address other rare but severe hazards will require three resources sometimes in scarce supply: money (for research), patience (for assimilating findings) and cooperation among construction industry stakeholders (architects, engineers, officials, owners, contractors, etc.). The goal should be a widely applicable national building code that incorporates technically sound, practical design guidance while reflecting a range of performance levels depending on hazards, risk tolerances, building types, uses and regional differences.

We are not facing a simple problem with a simple solution. The Murrah Building in Oklahoma City was designed and constructed in 1974 to meet all prevailing codes, but the bomb detonation at one column caused 60% of the building to collapse.

The shattering of the column could not have been prevented, but ductile details, similar to those used in earthquake areas (so members can bend without breaking), could have helped two other, less-impacted columns survive. Disproportionate collapse, also called progressive collapse, followed the failure of key structural elements. Ductility and redundancy (alternate load paths) probably would have significantly reduced the extent of collapse.

The World Trade Center towers, designed in the late 1960s and early 1970s, also met prevailing codes. In addition, they had significant resistance to progressive collapse. The airplane impact instantly destroyed many exterior columns in each tower, but the remaining structural members were able to redirect loads and save many lives. However, gradual weakening of structural components by simultaneous fast-developing fires on multiple floors eventually compromised the load-bearing capacity and led to total collapse.

The Murrah Building and World Trade Center events illustrate that further research into blast, progressive collapse and fire performance–leading to development of enhanced building code provisions–could provide tangible benefits against a variety of extreme hazards.

Some will say that collapses are so rare that our present approach is fine. But failures teach us the limits of building performance. They must be studied, not discounted as exceptions. Others believe they already know the problem and how to fix it. But structural framing arrangements, fire resistance systems and interactions under extreme events are complex.

Any fix must have solid technical underpinning. Therefore, research on performance is a vital first step in improving building codes.

Solid data. U.S. codes were developed from solid technical data to address both the properties of construction materials and systems, and a wide range of environmental conditions and natural hazards, such as earthquakes and hurricanes. They were written and adopted on a consensus basis and have been regularly reevaluated and improved. For reliability and efficiency, any code provisions addressing new or expanded concerns such as blast, progressive collapse, large conflagrations, malfunctioning fire suppression systems and loss of fire protection coverings should be based on the same level of scientific rigor and be widely applied.

No society has unlimited resources, so we must consider cost-effectiveness. The most effective way to protect building occupants against malevolent acts is to eliminate the threats, but hazards will still exist. We must study the relationships between building performance, types of hazards, risks posed by those hazards and costs to reduce risk.

From past experience, the costs need not be great. They can be affordable if incorporated into the original design. Retrofitting or renovating existing buildings after the fact is, for the most part, cost prohibitive. Still, for iconic or high-profile properties, retrofit costs might be justified.

What philosophy should apply to newly considered hazards? In my view, the code should establish a set of performance enhancements, such as stronger or more ductile elements and connections, higher load factors at key structural elements and greater fire resistance at selected locations. Then the level of enhancements should be varied to reflect the risks that different buildings face from malevolent acts and extreme man-made events, much as acceleration maps and code "importance factors" do for seismic hazards. For maximum effectiveness, I urge that the code be applicable nationally.

Charles H. Thornton, Ph.D., P.E., is the Washington, D.C.-based co-chairman of The Thornton-Tomasetti Group Inc., an engineering firm that helped manage World Trade Center response for New York City's Dept. of Design and Construction. He is also chairman of the Building Seismic Safety Council. He may be e-mailed at