Why Do Construction Projects Fail? Inspections of Boston's Big Dig reveal hundreds of bolts that need to be reinforced. Guests talk about the materials and techniques used in major construction projects. What keeps houses, bridges and tunnels from falling apart? And are there ways to easily detect damage before it becomes severe?
NPR logo

Why Do Construction Projects Fail?

  • Download
  • <iframe src="https://www.npr.org/player/embed/5589627/5589628" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
Why Do Construction Projects Fail?

Why Do Construction Projects Fail?

  • Download
  • <iframe src="https://www.npr.org/player/embed/5589627/5589628" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

JOE PALCA, host:

This is TALK OF THE NATION: SCIENCE FRIDAY from NPR News. I'm Joe Palca. Ira Flatow is away.

This week, Matthew Amorello, the Massachusetts Turnpike Authority chairman, resigned after facing increasing criticism of his management of the Big Dig, a huge tunnel project through Boston that's been called one of the biggest feats of engineering in the country.

Earlier this month concrete panels weighing several tons each, fell from the ceiling of part of the tunnel, killing a woman. The panels were held up by a system of bolts and epoxy glue embedded in the roof of the tunnel. Inspectors now say that over 1,400 of the bolt hangars will need to be reinforced for safety.

This hour we're going to talk about engineering structures and some of the basic building materials that make up the world around us - materials like concrete, steel and wood.

If you'd like to join the conversation, give us a call. Our number is 1-800-989-8255, that's 1-800-989-TALK. And if you want more information about what we're talking about this hour, go to our Web site at www.sciencefriday.com.

Now, I'd like to introduce my guests. Steven Cramer is a professor in the Department of Civil and Environmental Engineering and is associate dean of academic affairs in the College of Engineering at the University of Wisconsin - Madison. He joins us from the studios of Wisconsin Public Radio. Welcome to the program.

Mr. STEVEN CRAMER (Civil and Environmental Engineering; University of Wisconsin-Madison): Hello, Joe. Delighted to be here.

PALCA: Great!

And we also have with us Christopher Earls. He's the editor-in-chief of the Americas - of the International journal Steel and Composite Structures and is an associate professor in the School of Civil and Environmental Engineering at Cornell University and is - I'm sorry - in New York. And he joins us from the studio on the Cornell campus. Welcome.

Mr. CHRISTOPHER EARLS (Editor-in-Chief, Steel and Composite Structures): Thank you, Joe. Pleasure to be here.

PALCA: Great! So let's - I know I'm not going to - I know that neither of you gentleman have been directly associated, or I don't believe you're directly associated with the investigation into this problem they've had in Boston. But I'm just wondering in a theoretical sense if you - is there a way in advance to say uh-oh, we have a problem here. Maybe I can put that question to you Christopher Earls.

Mr. EARLS: So the possibility exists that there may, in a circumstance like this, be some early warning sign in the form of a gross deformation that may present itself and be measurable. And I say gross from the standpoint that it's a large-scale deflection even though a single anchor bolt may be what the initiator of the failure was. That could be detected, for instance, using some sort of a technique such as related to surveying or something like this.

PALCA: So how would that work? How would you use a surveying technique to see whether there was any kind of deformation?

Mr. EARLS: So there are technologies that exist where you can map large-scale components to very high accuracies and seek or be able to measure what your current condition is - compared to some baseline condition - and then be able to gauge a change and how much change has occurred. And this can be as small as fractions of a millimeter using these types of, in this case, laser scanning.

PALCA: Huh. And how, I mean, what would you do, for example, I mean, what would be the next step if you saw one of these deformations and, you know, what would you do to prevent something bad happening?

Mr. EARLS: So, again, I'm not terribly familiar with this particular failure or the Big Dig itself.

PALCA: Right.

Mr. EARLS: But if we were to, let's say, scan the tunnel lining or the ceiling of this particular piece that fell, and were to observe that this concrete slab that was about 3,000 pounds - quite large - had moved a millimeter or so in advance then, perhaps, one might want to discern why it was that that motion occurred or that movement occurred and go seek the cause.

PALCA: I see.

Mr. CRAMER: Of course.

PALCA: Go ahead. Steven Cramer, sure, go ahead.

Mr. CRAMER: Yeah, of course, the challenge is knowing where to look. And the displacements or deformations that Chris described are usually so small that you wouldn't notice them unless you thought you had a problem there in the first place.

PALCA: Right. So how, I mean, if you're - when you teach engineers what to look for, what are you teaching them to look for? Steven Cramer, maybe I can put that to you.

Mr. CRAMER: Sure. Well, what we have here - clearly in the case of the Big Dig something happened there that wasn't supposed to happen. And when it occurred and where it occurred, you know, at this point I don't think we know.

PALCA: Mm-hmm.

Mr. CRAMER: So, you know, we train engineers to handle risk - this is basically a risk problem - to insure that the load, in this case, the weight of the panels, would be less than the material resistance, in this case, the epoxy and the bolts. And if everyone does their job correctly, then the panels don't fall down.

The trick here is that during construction and the design process, of course, things can happen that aren't anticipated. And very few of us, including the public, want to pay for a structure that I would call too safe.

PALCA: Mm-hmm.

Mr. CRAMER: They want to pay for a structure which is safe, but not one that's too safe.

PALCA: I got it...

Mr. CRAMER: And therein lies the challenge.

PALCA: Yeah, so it's navigating the reasonable amount of safety without making it so, I mean, so safe that it's got 10-foot thick walls. Actually, I heard that necessarily, people tend to associate really thick walls with safety, but in some cases that's not the case. So, anyway, I guess it's a series of engineering decisions that have to be made.

Mr. CRAMER: Exactly.

PALCA: Okay.

Mr. CRAMER: And we train our engineers and educate our engineers to be able to look at the numbers and weigh these risks and safeties. Clearly, though, some of this is codified. There are building codes that are legal documents and engineers, at a minimum, have to meet those requirements.

PALCA: Got it. Sorry, did you want to finish that thought.

Mr. EARLS: So I was going to reiterate what Steve was saying...

PALCA: Oh, Mr. Earls, go ahead, yeah.

Mr. EARLS: ...so I was going to reiterate what Steve had said about what we, you know - there's an expectation on the part of the designer, the engineer, that that which is designed and specified will be constructed in the fashion and manner in which it was intended.

And so it's possible, for instance in this case, that the hangar bolts, that they may have been, on paper, shown to have adequate capacity, but as Steve had pointed out, perhaps there were site conditions that prevented the proper installation of the bolts.

I had heard of one instance where there may have been some problems related to the tunnel lining and so there could have been water infiltrating into the areas where the epoxy should be curing and potentially, you know, that could affect its final strength and so forth. But these are important differences that do occur in the real world that would violate assumptions made by the designer, and that could lead to a condition where there may be a compromise in capacity over what was intended.

PALCA: I'd like to invite one other person to join our conversation. Paul Monteiro is a professor of civil and environmental engineering at the University of California in Berkeley. And he joins me today by phone from Sao Paolo, Brazil. And perhaps we'll find out what he's doing in Brazil, but thanks for joining us today.

Mr. PAUL MONTEIRO (Civil and Environmental Engineering, University of California at Berkeley): Oh, my pleasure. Thanks for inviting me.

PALCA: So, you've just written an article in the proceedings of the National Academy of Sciences, trying to evaluate the longevity, I guess, of concrete. And I'm just wondering. I mean given what we just heard about the possibility that water would've gotten into concrete. Is that the kind of thing you have to try to anticipate when you're building a structure with material like concrete that can be deformed if water's dripping into it?

Mr. MONTEIRO: Oh, precisely. The problem with concrete is that if not properly done, it can very porous. And if it is very porous, water can percolate. And that can be even worse because this water could contain some aggressive elements like chloride sulfates, which can crack the concrete and expand it.

PALCA: I see. You know, and what surprises me is that concrete has been around since the Greeks and the Romans were building structures. Is there really more to learn about concrete? It seems like that would've been a really well studied piece of building material.

Mr. MONTEIRO: Of course, this is the natural question that every time when it goes to a party, people will always wonder why somebody could spend a lifetime studying concrete. And the fact of the matter it is, concrete's a completely complex material that porous structures ranges from the microbe size - excuse me - continue to the nanometers. And the big problems in the construction is that sometimes people don't do a good job and the imperfection can be in the sizes of inches.

So we have a scale of seven times, which is tremendous. So if you want to do some things scientific in modeling it's quite of a challenge. And if I may, basically the modern - well, modern is a strong word - but certainly the Roman concrete had two advantages. One of this is the safety factor was tremendous. You see whatever is left, you see the sizes of everything.

The second part is that they did not use reinforcing steel. That is to say they could not put concrete through tremendous amount of tension, otherwise it would crack. So we have to do something very safe, very economical, and also ecologically correct.

PALCA: And what would be ecologic - I mean what's the ecological component?

Mr. MONTEIRO: Certainly. Just think in terms of the volumes of concrete that's being used in the world now, which is about 11.5 billons tons per year - in the world. So if you think in terms of the population of the world, we are talking about, what, six billion people? In average, we have just about two tons per person per year.

Now, the question is to produce cement, we have to generate CO2 in the atmosphere because of the burning process. So for each ton of cement, it generates one ton of CO2. So the production of cement, which is used in concrete, is responsible for about 7 percent of the CO2 emission.

PALCA: So that's something I hadn't really contemplated. But by building a concrete structure, you're contributing to global warming in a sense, or at least to the buildup of greenhouse gases in the atmosphere if you believe that to contribute to global warming.

Mr. MONTEIRO: Absolutely. On the other hand, concrete is really a - if properly done, I want to emphasize that - is a very ecological process because we can use waste products for other industries. For instance, we can use a huge amount of fly ash, which is a byproduct from the process of burning coal...

PALCA: Dr. Monteiro, I'm going to have to interrupt you there for - we have to take a quick break. But we're talking about the structures that contribute to the things that we live in, work in, build, drive through, what have you. And we'll be taking your calls and talking more about this when we come back after a short break, so stay with us.

This is TALK OF THE NATION from NPR News.

(Soundbite of music)

PALCA: From NPR News, this is TALK OF THE NATION: SCIENCE FRIDAY. I'm Joe Palca. We're talking this hour about engineering and materials.

My guests are Steven Cramer. He's a professor in the department of civil and environmental engineering at the University of Wisconsin-Madison. Christopher Earls is an associate professor in the School of Civil and Environmental Engineering at Cornell University in Ithaca, N.Y. And Paul Monteiro is a professor of civil and environmental engineering at the University of California at Berkeley, although today he's in Brazil.

And you're welcome to the conversation. Our number is 1-800-989-8255. That's 1-800-989-TALK. And let's take a call right now. How about Guntrum(ph) - I think I've got that right - in Newton, Mass. Welcome to the program.

GUNTRUM (Caller): Well, thank you. Yes. I have two questions actually. The first question is why do they put these both in vertically. That seems the worse possible way of doing it when you could put it at an angle or even use a pair of bolts at the two opposite angles. That's my first question. The second question is: they keep talking about supporting the weight of the slab. It seems to me there's forces much larger to consider. I mean, ten slabs fell down, not one. So when the first one fell down, I assumed what happened is that it sort of wedged itself against the next one causing it to fall down also and down the line.

PALCA: Well, Guntrum, I have to point out that none of my guests are explicitly working on this particular problem, but perhaps one you gentlemen can at least make some comment on that. Yes, go ahead.

Mr. EARLS: This is Chris.

PALCA: Chris, go ahead.

Mr. EARLS: I could probably point something out that is indicative of the anchor bolts, itself is typically you want it to be loaded it in withdrawal(ph), which is a uniaxial(ph) state of stress. You have pure tension. And so if you were to put a bolt at an angle, the bolt then would be experiencing some bending and beam action, and that actually makes the anchor bolt weaker.

And so you try to avoid that in a situation such as this. You would like the bolt to be as strong as it can be and that would be in a tension-only condition. And so I think that's why it's not angled in.

PALCA: And this other question, Guntrum says about panels falling down serially. Does one thing lead to another, as they say?

Mr. EARLS: That I'm not familiar with the configuration of the system in general. And so it would be - I could only guess what would cause multiple panels to fall. But sometimes that's not uncommon where you would have a - it's still a small scale failure. One, two, three, four panels come down, hundreds of others remain in place. So it still would, in my mind, qualify as a small-scale failure even though it involved more than one panel, perhaps emanating from single bolt.

PALCA: You know, what I'm wondering about in this particular circumstance is people have been building tunnels for hundreds of years certainly, building bridges for that long and much longer, thousands of years. How is it possible, given those thousands of years of experience, to build something today that doesn't even last a decade? I mean is that a fair question for someone to ask? Maybe I can put that to you Steven or Christopher. Go ahead.

Mr. CRAMER: Sure. This is Steve.

PALCA: Alright. Steve, go ahead.

Mr. CRAMER: Yeah, that's a fair question. I think you have to remember is that our materials and our techniques are changing over time. And as Paul pointed out we're using different additives in concrete than we've used in the past. And with wooden materials that I work with, we're combining the wood, raw material, in different ways in the structural numbers.

So our techniques and our materials are changing and they're changing fairly dramatically. And surprisingly, yes, there is a lot to learn yet about these materials. I think sometimes in our excitement over biotechnology and some of the other so-called hot areas of research, we sometimes to forget that these everyday things we don't understand completely yet either.

PALCA: So, maybe I could ask you this. Then there are new materials, if they're being introduced they obviously have some advantages. Maybe you can talk about the advantages of new materials and the disadvantages. For example, drywall verses plaster.

Mr. CRAMER: Sure. Not too long ago I was doing a study on drywall, much to the amusement of my children at the dinner table. Wondering why anyone would study something so unexciting. But it turns out that drywall is ubiquitous in structures. It's used as the primary fire protection mechanism in light frame structures, including houses and apartment buildings, and smaller buildings.

And we don't know very much about how it actually behaves from an engineering point of view. And we know it works. We know it slows the progression of fire. It's clearly saved lives in many instances. But yet it's never been really researched and fully understood and designed from an engineering point of view.

And that's just an example of several holes we have in structures and materials that we need to fill.

PALCA: Let's take another call now and go to David in Portsmouth, Virginia. David, welcome to Science Friday.

DAVID (Caller): Hey, good morning, or good afternoon. Listen, I'm an engineer. I've worked with bolts for about the last 35 years exclusively. And I had a couple of questions. Number one, exactly what did fail? Was it the epoxy, was it the concrete, or was it the bolt itself?

PALCA: Well, David, I want to reinforce - reinforce - I want to remind everybody that none of my guests on this show are explicitly involved in this investigation, so I'm not sure that they can answer that question. But just in general, what are the - maybe we can make it a more general question - what are the possible failure points when a hanger bolt fails. Maybe, I don't know, Christopher Earls, is that something...

Mr. EARLS: Sure...

DAVID: Well, there are some that are obvious. The epoxy - when mixing epoxy, the manufacturer has certain rules and regulations. It's been my experience that the engineers say, here's what we want and here what happens. And it finally goes down to a mill (unintelligible) - it's some place n the (unintelligible) field - and they mix it up sort of the way they think they're going to mix it up. Quality control of mixing that epoxy must've, just has to have been very, very exquisite. It has to be right exactly on the mark.

And I also wondered, what physical traits would keep that bolt from falling out? Was there a head in there? Was there a crosspiece? Was there a cross... Was there any... What would've kept that bolt up in there?

PALCA: Okay. Let me see if we have any responses to that. I'm sorry, Christopher Earls, can you address that at all?

Mr. EARLS: Well, so Dave makes an excellent observation from the standpoint that clearly for the epoxy to be properly functioning and to carry its design intents it has to be mixed properly and cured properly.

And again, there's, you know, it just in what's available in media coverage up to this point, there's some indication that perhaps maybe the inside conditions created a circumstance where the curing may not have been able to cure properly I suppose.

But in general, failure modes in bolts, again, if they're loaded in tension as they're supposed to be, you can have an excessive elongation. Like a piece of gum being stretched way too far and that's that one failure mode. And that gives you a great deal of warning because you can see the bolt elongating and detect the stress in the structural system prior to a catastrophic failure.

And the other cause would be one that's unfortunate, and that would be a brittle, sudden failure fracture on a nut section. And in that instance, there would little or no warning and the bolt would fracture and fail. But, again, it's not, I don't know whether there was a failure in the epoxy or a failure of the bolt itself.

I hear in the media that it's the epoxy, but I don't know.

PALCA: Well, there was some mention also in the media today that perhaps some construction nearby or the building that was going on causing a vibration. Maybe I could turn back to Paul Monteiro and ask, if you build a building out of concrete with a certain set of expectations - or for example, a dam with a certain set of expectations - and suddenly there's something that changes at the site. Can you make any predictions about how the material's going to behave?

Mr. MONTEIRO: Correct. This always a tough question to address. (unintelligible) all the design of all this important structures, the engineers spend a tremendous amount of scenarios of what could go wrong, because sometimes it will go wrong.

So in case of dams, that you ask, we try to take care of all the possibilities. And even so, there is always something that goes wrong - well not always - but certainly it could be very small that you could fix without any problem. But it's very rare that a large project would go without any problem. That doesn't mean to say that expect anybody to have tremendous consequence in the structure.

But the point I'd like to address is the extreme needs of development of non-destructive techniques. And we are developing completely new extenders and (unintelligible) in the last 10 years or say there's a new revolution on the way non-destructive techniques are being developed.

Of course there is a big transition from the laboratory to the field practice. And it's going a little bit slower but it's quite expectable that in a short period of time we have new equipment that can address many of these concerns.

PALCA: So what do you mean a non-destructive technique? I mean how does that work?

Mr. MONTEIRO: Right. The standard way of doing things in concrete, because it's so massive, once you cast the structure you have no clue, unless you see cracks that things really are going wrong. And unlike, say, a piece of small metallic components - say in a car or an airplane that you can send an acoustic emission process or sound waves.

Concrete is very massive so it generates these waves. So you want to put a lot of sensors inside and outside the concrete structure and it use a system approach and try to identify what's wrong and what's right and that creates a loops. Things that are going well, according to plan, we got confidence. When things do not follow exactly what we predict, we go there and check and inspect. And we do that without doing any damage to the structure.

PALCA: Now I got it. Okay. Well, you know, what's always - I mean this is something that I've always wondered about. Maybe this is my chance to find out. Christopher Earls, there are signs before you cross over a bridge frequently that say, this bridge is rated to carry a load of 16,000 pounds.

Now I saw a cartoon once. It was a Calvin and Hobbes where Calvin, who's the little kid, is asking his dad how they decide what the weight-bearing load is. And he says, well, they keep driving trucks across it until it collapses and they rebuild it to the, and then they know how much - I don't think that's the right answer.

But it makes me wonder if - I mean you're obviously, at some level, you're making an inference about, okay, this material is so strong and this structure is so strong. But you have to assume that the ground isn't shifting or the materials are the way they are. How do they make those kinds of judgments?

Mr. EARLS: Well, in the case of modern design specifications - and not all bridges are designed using most modern techniques - but as we move forward with design specifications, there's a probabilistic framework that's overlaid on top of the deterministic engineering theories that we use how much strength will this beam have and so forth and so on.

But you're absolutely right. There's a variability in terms of the loading. I can never know with any certainty what types of loads will be on the bridge deck at any given time. In addition, I won't know if those loads are occurring in conjunction with a snowstorm, or if they're occurring in conjunction with high winds.

In addition, I won't know with any absolute certainty that the steel that I specified for the bridge and my design calculations actually made it into the bridge in exactly the same way that I wanted. Or, that the alignment of the bridge is precisely the same way that I anticipated.

And so these are little variabilities, perturbations to the system that have to be accounted for. And modern design specifications adopt a probabilistic framework so that we could try to come up with what's known as a reliability index, an ability to quantify it with some degree of confidence what the probability of a failure will be at an arbitrary point in time.

PALCA: We're talking about structures, materials that are used to make buildings that we live in. And we're talking about ways of assessing how safe they are and how likely they are to collapse, which I'm happy to say most of the time is very unlikely.

I'm Joe Palca and this is TALK OF THE NATION from NPR News.

Although I also have to add, you know, in terms of this probabilistic model, that my wife is a bit of a fatalist when it comes to these structures and she's certain that every parking structure that she drives into is going to collapse while she's parking her car.

I'm glad you're laughing, because I laugh and she thinks I'm just making fun of her, but I'm not.

Mr. EARLS: How many people do you know that have died in a structural collapse?

PALCA: It's irrelevant. I've tried that argument.

Mr. EARLS: I know, but that's one possible way to...

PALCA: Go ahead, Steven Cramer (unintelligible)

Mr. CRAMER: Well, I think there's a new sensitivity. Ever since 9/11 and then following up with Katrina and Rita last year, I think there's a new sensitivity on the part of the public that buildings are not invulnerable and that they can collapse. And I'm not sure people were thinking about that every day prior to 9/11. I think there's a new sensitivity in the public in that regard.

PALCA: And what is reasonable for the public to expect given a new building? I mean, obviously, it would be very difficult to design a building anticipating that an airplane would be flown into it. But what's reasonable?

Mr. CRAMER: Exactly. 9/11 was clearly a very extreme event for which the building was not designed and it still survived for some time after that impact. But we do have load data, both due to human movement, structure movements, car movements in the case of your parking ramp, we have weather data for snow loads and wind loads. And we can regionalize that data as well.

So, for example, in Paul's area in California, we have earthquakes that we have to consider, and yet in Madison, Wisconsin, that's a relatively minor risk. So we have those historical records by which we can characterize loads. And similarly we have reasonable information about materials and how much they can resist.

And that's where the safety index comes in and ratio-ing those two to ensure that the chances of that parking garage collapsing as your wife drives into it is extremely small.

PALCA: I think we have time for one more quick question. So let's try to get that in. Bruce in Berkeley, Massachusetts. Bruce, welcome to the program.

BRUCE (Caller): Hi.


BRUCE: I find it hard to believe that you could have sound engineering practices that puts 10-ton pieces of concrete on top of people driving unless the concrete is absolutely necessary for the bridge. It didn't seem like it was a structural element. How could it be sound practice to put those things, no matter how they are supported, on top of people if they don't have to be there. Can someone explain that to me?

PALCA: Okay. Well, we'll take a crack at that but we're going to have to keep it short. Paul Monteiro, maybe you can give a brief answer to that complex question, I'm sure.

Mr. MONTEIRO: I'll be very brief. I'm not familiar with the design at all. The accident happened while I was abroad so I'm getting very small piece of information. So if anybody has any insight...

PALCA: Okay. We'll throw it open to others. Do any of your gentlemen have a...

Mr. CRAMER: Well, sometimes in a critical application, you try to build on redundancy. And by that I mean that you don't rely on any single bolt or any single component to keep the structure safe. That's a strategy that sometimes engineers employ in a life critical situation.

I do not know what kind of redundancies existed or why the panels were needed in the Big Dig.

PALCA: Okay. Well, I'm afraid that's going to have to be where we leave it. I'd like to thank my guest, Steven Cramer, who's a professor in the department of civil and environmental engineering at the University of Wisconsin, Madison. Christopher Earls is the editor-in-chief for the Americas of the International Journal Steel and Composite Structures, and an associate professor in the School of Civil and Environmental Engineering at Cornel University in Ithaca. And Paul Monteiro is a professor of civil and environmental engineering at the University of California at Berkeley.

Thanks to all of you.

Mr. CRAMER: Thank you very much, Joe.

PALCA: Okay. And when we come back, some places you might consider going for your summer vacation, although you'll need a lot more than frequent flyer miles to get there. Stay with us.

This is TALK OF THE NATION from NPR News.

Copyright © 2006 NPR. All rights reserved. Visit our website terms of use and permissions pages at www.npr.org for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.