Thursday, December 28, 2006

Electric Power: Was It Broke? Did We Fix It?

Like any other profession, engineering has its particular proverbs and sayings. One of my favorites is, "If it ain't broke, don't fix it." As with most proverbs, this one captures only part of the whole picture of a complex situation. But when I look at the potential and actual problems we have these days with the U. S. electric power system, I wish more people in authority had paid attention to that particular proverb.

Electricity is an unusual commodity in that it must be produced exactly as fast as it is sold. If a million people suddenly turn on their lights all at once, somebody somewhere has to supply that much more electricity in milliseconds, or else there is big trouble for everybody on the power distribution network. For lights to come on reliably and stay on all across the country, the systems of generating plants, transmission lines, distribution lines, and monitoring and control equipment have to work in a smooth, coordinated way. And, somebody has to pay for it all.

From an economic point of view, approaches to electric utility management and financing lie somewhere between two extremes. At one extreme is completely centralized control, billing, and coordination, often performed in many countries by the national government. France is an example of this approach. Large, complex electric systems are a natural fit to large, complex government bureaucracies, and in the hands of competent, dedicated civil servants, government-owned and -operated utilities can be a model of efficiency and advanced technology. Government control and ownership can provide the stability for long-term research and development. This is one reason that France leads the world in the development of safe, reliable nuclear power, which provides most of the electricity in that country.

The other extreme can be found in third-world countries where there is little or no effective government regulation of utilities, either through incompetence, war, or other causes. In this type of situation, private enterprise rushes in to fill the gap and you have private "utilities"—often nothing more than a few guys with a generator and some wire—selling electricity for whatever the market will bear, in an uncoordinated and inefficient way. This approach leads to a spotty, inefficient market in which the availability and reliability of electricity depends on where you live, and typically large portions of the market (in rural or dangerous areas) are not served at all.

In the U. S., we have historically swung from near one extreme to the other. As electric utilities began to grow in the late 1800s and early 1900s, they began as independent companies. But the technical economies of scale quickly became apparent, and the Great Depression brought on tremendous consolidation of companies into a few large firms, which were then taken under the regulatory wing of federal and state governments. What we had then was a kind of benevolent dictatorship of the industry by government, in which private investors ceded much control to the various regulatory commissions, but received in turn a reliable but relatively small return on their investment.

This state of affairs prevailed through the 1970s, whereupon various political forces began a move toward deregulation. The record of deregulation is spotty at best, probably because it represents an attempt to have our regulatory cake and eat it too. No one wants the electricity market here to devolve to the haphazard free-for-all that it is in places like Iraq, or even India, where electricity theft is as common as beggary. So rightly, some regulations must be left in place in order to protect the interests of those who cannot protect themselves, which in the case of electric utilities means most of us.

The most noteworthy recent disasters having to do with deregulation were the disruptions and price explosions in California of a few years ago, caused in large part by Enron and other trading companies who manipulated the market during hot summers of high demand. Even if the loopholes allowing such abuses are closed and inadequate generating capacity is addressed with more power plants, however, many problems remain. A recent New York Times article points out that because the existing rules provide disincentives for power companies to spend money on transmission and distribution equipment (power lines), certain parts of the country have to pay exorbitant rates in the form of "congestion charges."

The basic problem is, there are not lines enough to carry cheap power from where it is available to where it is needed. Somebody would have to pay to build them, and somebody else would have to approve the construction. In these days of "not in my back yard" attitudes, it is increasingly hard to construct new power lines anywhere, even in rural areas. The net result of these complications is that as time goes on and demand for power increases, more and more areas may find themselves starved for power, and will have to pay rates that might be as high as twice the prevailing rate of surrounding regions.

My personal bias is that we have gone way too far in attempts to privatize the electric utility industry. It is a business which technologically fits better with a centralized authority and center of coordination. But in today's political climate, the chances of going back to a more centralized way of doing things are small. It looks like the best we can do is to continue to tinker with what regulations remain, fixing problems where pernicious disincentives appear, and keeping an eye out for Grandma and her electric heater that she needs to get through the winter. But in my opinion, the whole thing wasn't broke to begin with, and the fix of deregulation didn't need to be applied the way it was.

Sources: The New York Times article on congestion charges appeared in the Dec. 13, 2006 online edition at http://www.nytimes.com/2006/12/13/business/13power.html?hp&ex=1166072400&en=dcfbff42cc8f19d4&ei=5094&partner=homepage.

Tuesday, December 19, 2006

America's Chernobyl Waiting to Happen

"Dallas, Texas, Mar. 30, 2005 (AP) --- An apparent nuclear explosion in Amarillo, Texas has cut off all communications with the West Texas city and regions in a fifteen-mile radius around the blast. Eyewitness accounts by airline pilots in the vicinity report an 'incredible flash' followed by a mushroom cloud reaching at least 35,000 feet. Speculation on the source of the explosion has centered on Amarillo's Pantex plant, the nation's only facility for construction and disassembly of nuclear weapons."

In case you think you missed something a year ago last March, the news item above is fiction. But according to some sources, it is plausible. It could have happened. And there is reason to believe that unless some serious housecleaning takes place in Amarillo, the chances that something like this might happen in the future are higher than any of us would like.

The end of the Cold War brought hopes that instead of piling up megaton after megaton
of mutually assured destructive power in the shape of thermonuclear weapons, the U. S. and the Soviet Union (or what was left of it) would begin to disassemble their nuclear stockpiles to make the world a safer place. Over the past fifteen years, international agreements have been reached to do exactly that. From a peak of over 30,000 nuclear warheads in 1965, the U. S. stockpile has declined to just a little over 10,000 as of 2002. And here is where the engineering issues come in, because for every downtick of that number, somebody somewhere has to disassemble a nuclear warhead.

A nuclear bomb or missile is not something that you just throw on the surplus market to dispose of. First it has to be rendered incapable of exploding. Then the plutonium and other dangerous explosive materials have to be removed in a way that is both safe to the technicians doing the work, and also to the surrounding countryside and population. As you might imagine, these operations are difficult, dangerous, and require secret specialized knowledge. For more than thirty years, the only facility in the U. S. where nuclear weapons were made or disassembled has been the Pantex plant outside Amarillo, Texas. It is currently operated by a consortium of private contractors including BWXT, Honeywell, and Bechtel, and works exclusively for the federal government, specifically the Department of Energy. If you want a nuclear weapon taken apart, you go to Pantex, period. And therein lies a potential problem.

Where I teach engineering, the job of nuclear weapon disassembler is not one that comes up a lot when students tell me what they'd like to be when they graduate. I imagine that it is hard to recruit and retain people who are both willing and qualified to do such work. But at the same time, it is not the kind of growth industry that attracts a lot of investment. So it is plausible to me that as the demand for disassembly increases, the corporate bosses in charge of the operation might tend to skimp on things like deferred maintenance, safety training and execution, and hiring of additional staff. That is the picture which emerges from an anonymous letter made public recently by the Project on Government Oversight, a government watchdog group.

Anonymous letters can contain exaggerations, but what is not in dispute is the fact that on three occasions beginning Mar. 30, 2005, someone at Pantex tried to disassemble a nuclear weapon in a way that set off all kinds of alarms in the minds of experts who know the details. I'm speculating at this point, but as I read between the lines and use my knowledge of 1965-era technology, something like this may have happened.

A nuclear weapon built in 1965 probably contained no computers, relatively few transistors, and a good many vacuum tubes. Any safety interlocks to prevent accidental detonation were probably mechanical as well as electronic, and consisted of switches, relays, and possibly some rudimentary transistor circuits. But somewhere physically inside the long cylindrical structure lies a terminal which, if contacted by a grounded piece of metal, will probably set the whole thing off and vaporize Amarillo and the surrounding area.

A piece of equipment that has been sitting around since 1965 in a cold, drafty missile silo is probably a little corroded here and there. Screws and plugs that used to come apart easily are now stubborn or even frozen in place. The technician in charge of beginning disassembly of this baby probably tried all the standard approaches to unscrewing a vital part in order to disable it, without success. At that point, desperation overcame judgment. The official news release from the National Nuclear Safety Agency puts it in bureaucratese thus: "This includes the failures to adhere to limits in the force applied to the weapon assembly and a Technical Safety Requirement violation associated with the use of a tool that was explicitly forbidden from use as stated in a Justification for Continued Operation." Maybe he whammed at it with a big hammer. Maybe he tried drilling out a stuck bolt with an electric drill. We may never know. But what we do know is, the reasons for all these Technical Safety Requirements is that if you violate them, you edge closer to setting off an explosion of some kind.

Not every explosion that could happen at Pantex would be The Big One with the mushroom cloud and a megaton of energy. The way nuclear weapons work is by using cleverly designed pieces of conventional high explosive to create configurations that favor the initiation of the nuclear chain reactions that produce the big boom. A lot of things have to go right (or wrong, depending on your point of view) in order for a full-scale nuclear explosion to happen. Kim Jong Il of North Korea found this out not too long ago when his nuclear test fizzled rather than boomed. But even if nothing nuclear happens when the conventional explosives go off, you've got a fine mess on your hands: probably a few people killed, expensive secret equipment destroyed, and worst from an environmental viewpoint, plutonium or other hazardous nuclear material spread all over the place, including the atmosphere.

This general sort of thing was what happened at Chernobyl, Ukraine in 1986, when some technicians experimenting late at night with a badly designed nuclear power plant managed to blow it up. The bald-faced coverup that the USSR tried to mount in the disaster's aftermath may have contributed to its ultimate downfall. So even if the worst-case scenario of a nuclear explosion doesn't ever happen at Pantex, a "small" explosion of conventional weapons could cause a release of nuclear material that could harm thousands or millions of people downwind. Where I happen to live, incidentally.

I hope the concerns pointed out by the Pantex employees who apparently wrote the anonymous letter are exaggerated. I hope that the statement from Pantex's official website that "[t]here is no credible scenario at Pantex in which an accident can result in a nuclear detonation" is true. But incredible things do happen from time to time. Let's just hope they don't happen at Pantex any time soon.

Sources: The Project on Government Oversight webpage citing the Pantex employees' anonymous letter is at http://www.pogo.org/p/homeland/hl-061201-bodman.html. The official Pantex website statement about a nuclear explosion not being a credible scenario is at http://www.pantex.com/currentnews/factSheets.html. Statistics on the U. S. nuclear weapons stockpile are from Wikipedia's article on "United States and weapons of mass destruction."

Tuesday, December 12, 2006

Hacker Psych 101

Well, it's happened again. The Los Angeles Times reports that for more than a year prior to Nov. 21, 2006, somebody was siphoning personal information such as Social Security numbers from a database of more than 800,000 students and faculty at UCLA. Eventually, the system administrators noticed some unusual activity and suppressed the hack, but by the time they closed the door, a great many horses had escaped the barn.
This is one of the biggest recent breaches of data security at a university, but it is by no means the only one. The same article reports that 29 security breaches at other universities during the first six months of this year affected about 845,000 people.

Why is hacking so common? This is a profound question that goes to the heart of the nature of evil. It's good to start with the principle that, no matter how twisted, perverse, or just plain stupid a wrong action looks to observers, the person doing it sees something good about it.

For example, it's not a big mystery why people rob banks. In the famous words of 1930's gangster Willie Sutton, "Because that's where the money is." To a bank robber, simply going in and taking money by force is a way to obtain what they view as good, namely, money.

There are hackers whose motivation is essentially no different than Willie Sutton's. Identity theft turns out to be one of the easiest types of crime for them to commit, and so they turn to hacking, not because they especially enjoy it, but because it will lead to a result they want: data they can use to masquerade as somebody else in order to obtain money and goods by fraud. This motivation, although deplorable, is understandable, and fits into our historical understanding of the criminal mind, such as it is. As technology has advanced, so must the technical abilities of criminals. At this point it isn't clear whether money was the motive behind the UCLA breach or not. Because the breach had gone on so long without notable evidence of identity theft, it's possible that this was a hack for the heck of it.

Many, if not most, hacks fall into this second category. For an insight into why people do these things if they're not making money or profiting in some other way, the insights of Sarah Gordon, a senior research fellow at Symantec Security Response, shed some light on the matter.

Gordon's specialty is the ethics and psychology of hacking. In her job at Symantec, she has encountered just about every kind of hack and hacker there is. In an interview published in 2003, she says that the reason many people feel little or no guilt (at least not enough for them to stop) when they write viruses and do hacks is that they don't consider computers to be part of the real world. Speaking about school-age children learning to use computers for the first time, she said, "They don't have the same morality in the virtual world as they have in the real world because they don't think computers are part of the real world."

Gordon says that parents and teachers should share part of the blame. When a child steals someone's password and uses it, for example, a teacher could ask, "Would you steal Johnny's house key and use it to poke around in his bedroom?" Presumably not. But the analogy may be a difficult one for children to make—and many adults, for that matter.

Gordon thinks it may take a generation or two for our culture's prevailing morality to catch up with the hyper-speed advances in computer technology. She sees some progress in the U. S., noting that there is a new reluctance to post viruses online, whereas a few years ago no one thought there was anything wrong with the practice. Still, she thinks that hacking and virus-writing is an act of rebellion that remains popular in countries where young people are experiencing computers and networks for the first time, and rebellion is just part of human nature. A boy who grew up in a thatched hut with no running water, moves to a city, and finds that he can disrupt the operations of thousands of computers halfway across the world with a few keystrokes can receive a power buzz that he can get nowhere else in his life.

It seems to me that the anonymity provided by the technical nature of computer networks also contributes to the problem. Some say that a test of true morality is to ask yourself whether you would do a bad thing if you were sure you'd never get caught. The nature of computer networks ensures that very few hackers and virus writers do get caught, at least not without a lot of trouble. And it looks like lots of people fail that kind of test.

Well, I'm a teacher, so if there are any students reading this, I'm here to tell you that just because you can hide behind a computer screen, you shouldn't abandon the Golden Rule. But it may take a few years for the message to soak in. At the same time, I recognize a broader generalization of Sarah Gordon's notion that rebellion is part of human nature: evil and sin are part of human nature. I think this was a feature of humanity that many computer scientists neglected to take into consideration way back when they were establishing the foundations of some very pervasive systems and protocols that would cost billions of dollars to change today. Eventually things will get better, but it may take a generation or more before password theft and bicycle theft are viewed as the same kind of thing by most people.

Sources: The Dec. 12 L. A. Times story on the UCLA security breach is at http://www.latimes.com/news/local/la-me-ucla12dec12,0,7111141.story?coll=la-home-headlines. The interview with Sarah Gordon is at http://news.com.com/2008-1082-829812.html.

Tuesday, December 05, 2006

Superman Works for Airport Security Now

I've had occasion to mention Superman before ("Sniffing Through Your Wallet with RFID", Oct. 25, 2006), but my reference then to his X-ray vision was in jest. Well, a news item from the U. S. Transportation Security Administration says that in effect, they've hired Superman (at least, the mechanical equivalent of his X-ray vision ability) to watch passengers at Phoenix's Sky Harbor International Airport. The effect is to allow strip searches without stripping.

According to the Dec. 1 Associated Press news item, in the initial tests of the system, which uses a type of X-ray technology called "backscatter," security officials will examine only people who fail the primary screening. These passengers will be offered the choice of either a pat-down search or examination by the backscatter machine. The images, which reportedly are blurred by software in "certain areas," are nevertheless detailed enough to show items as small as jewelry next to the body. The technology is already in use in prisons, and the intensity of X-rays is much lower than a typical medical X-ray.

When I read this story, it brought back memories of my days as a junior terrorist. Before you get up from your computer to call the FBI, let me explain. In the 1990's, I did some consulting work for a company that was developing a contraband detection system using short radio waves called millimeter waves. It turns out that the human body emits these waves just because it's warm. With a detector that is sensitive enough, you can detect the waves coming through clothing, and if you are wearing something like plastic explosive under your shirt, the shadow of it will show up in the image.

We built a system, and to test it, several of us took turns playing terrorist by wearing lumps of modelling clay and plastic pistols taped to our shirts underneath a windbreaker. It was a tedious task, because the machine took 15 minutes or more to make a decent picture and you had to hold still the whole time. The results looked like blurry photographic negatives, but you could see the outlines of the contraband clearly. You could also see the main features of the body underneath the clothing, and that led to some privacy concerns, as you might imagine. The wife of the company president volunteered to be our female subject. I never saw the resulting picture—apparently it was detailed enough to be censored. For a number of reasons, both technical and social, that particular machine never made it to market, but all this was before 9/11 and the sea change in our attitudes toward airport security that resulted.

This change in attitudes has done funny things to some people, notably Susan Hallowell, who is the Transportation Security Administration's security laboratory director. A picture accompanying the article shows Ms. Hallowell in the X-ray altogether, and shows about the same detail as a department-store mannequin from the 1950s, or a Barbie doll. I suppose Ms. Hallowell's willingness to pose was motivated by a sincere desire to increase the quality of airport security with less discomfort to passengers, but it wouldn't surprise me if her strategy backfires. If I put myself in the mindset of a middle-aged woman who faces the choice of either letting another woman do a pat-down search, or knowing that somewhere out of sight, somebody—possibly another woman but possibly not—is going to see every single bulge, sag, and fold underneath my clothes, I would choose the pat-down search every time. In fact, I'd go screaming to my Congressman to stop implementation of the backscatter system before my naked profile showed up on MySpace. Yes, the TSA says the images won't be stored or transmitted. And maybe they will be able to keep that promise. But if there's a leak somewhere—say Madonna goes through one of these things and a paparazzi manages to bribe an inspector—the whole plan could go up in political flames.

Besides which, there is a principle that is largely neglected today, but still deserves some attention: the Constitutional prohibition against unreasonable searches and seizures. The Fourth Amendment says in full, "The right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures, shall not be violated, and no warrants shall issue, but upon probable cause, supported by oath or affirmation, and particularly describing the place to be searched, and the persons or things to be seized." I'm no Constitutional or legal scholar, and obviously some legal means has been found to get around constitutional challenges to airport security inspections. Probably the argument is, if you don't want to be searched, take the bus. But letting somebody I don't know see me without clothes, simply on the slight chance that I'm carrying a gun or a bomb, seems to cross a line that we as a nation have hesitated to cross before.

When George Orwell portrayed the ever-present unblinking eye of Big Brother in his dystopia 1984, the idea of being spied on constantly had the power to shock, because it was so novel. But today there are places in England where you can walk for many blocks and never be out of sight of security cameras. This has not destroyed England, and it has actually helped track down terrorists such as those who committed the London subway bombings. The thing we lose when one more privacy barrier comes down is so hard to describe because it's silent, has no public relations agent promoting it, and doesn't show up in the compilation of gross national products. But it's the kind of thing that you notice mainly after it's gone. And once it goes, it can be very hard to recover.

Sources: The AP article describing the Phoenix tests was carried by many media outlets, among them MyWay (http://apnews.myway.com/article/20061201/D8LO1JLO2.html). The paper describing my foray into contraband detection was entitled “Contraband detection through clothing by means of millimeter-wave imaging,” by G. R. Huguenin et al., SPIE Proc. 1942 Underground and Obscured Object Imaging and Detection, Orlando, FL, pp. 117-128, 15-16 April 1993.