Technological Realism Should Replace Optimism
As a “served engineer” on a nuclear powered submarine, I learned a long time ago that things go wrong, even with the very best technology. The recognition of inevitable “problems” should not deter technical development and should not make people afraid to develop new products and services, but it should add a healthy dose of humility backed up by continuous efforts to prepare for the worst.
My experiences have taught me to be uncomfortable with any proclamation of inevitable progress. I have worked on IT projects, been a full participant in the digital revolution, operated a custom plastics manufacturing company, and watched the nuclear industry work to regain respectability after some serious missteps in its early development history. Progress is hard work and there are often failures that reset the development cycle just as it seems ready to take off.
Too many technology observers and pundits point to Moore’s Law as some kind of a general rule for technical developments. Moore’s Law is a very particular pronouncement – in 1965, Gordon Moore recognized that there was a recognizable path forward that would allow manufacturers to double the number of transistors that could be inexpensively placed on a chip every year for the next ten years and he recognized that he could apply that law to the 15-20 years of chip development that had already happened. He modified his prediction in 1975 to increase the doubling time to two years instead of one. He predicted that the implementation of that path would allow an increasing quantity of processing power, assuming that it would be possible to keep all of the transistors firing at the same rate as before.
Moore’s Law does not apply to software development, to steel making, to underwater sensors, to remote manipulators, to wind energy collection systems, or to the rate of IP data transmission using satellite networks. It is not even infinitely applicable to semiconductor based processors – there are physical limits to the size of transistors and connecting wires that will eventually provide an asymptote that levels out the growth of processing power.
I have never had much “faith” in technology. I like technology. I use lots of technology; my children have occasionally called me “Inspector Gadget” because of all of the tools (my wife and children sometimes call them “toys”) I have accumulated over the years. However, I understand the limits of the technology that I use. I read the manuals, heed the warnings, plan for failure, and worry about the potential consequences of inappropriately using technical devices. I know that no technology can overcome physical barriers; nothing I or anyone else can do will provide power from the wind when it is not blowing and nothing that I or anyone else can invent will enable chemical combustion to provide reliable heat energy without both a source of oxygen and a place to dump the waste products. Nothing that I or anyone else can invent will enable oil extraction from a dry well.
I also know that not everything that breaks can be fixed, even if there is an unlimited amount of time and money. Some breaks and fissures can never be welded shut or forced to heal.
This is where I believe that humble engineers and technicians who are not driven by sales numbers have a huge role to play. Their (our) natural pessimism can help to reduce the consequences of always listening to the optimists, the people who say “damn the torpedoes”, “failure is not an option”, or “whatever it takes”. Failure is always possible. Before stretching limits it is important to recognize the consequences of the failure to determine if they are acceptable. If the reasonably predictable “worst possible event” results in consequences that cannot be accepted, the prudent course of action is to avoid the action in the first place.
I place deepwater drilling for oil and gas into that category. It is pretty obvious that the possible consequences are unacceptable and that technological development has not yet found a way to mitigate those consequences. I am not sure what the limits of “deepwater” should be, but it is apparent that 5,000 feet is beyond the limit. I do not place operating nuclear energy production facilities in that category. However, there are very definitely some kinds of nuclear plants – like very large graphite-moderated, water-cooled reactors operated by people who override safety systems and ignore warning indications – that have proven that they can cause consequences that are not acceptable.
The real value comes in determining what the reasonably predictable consequences might be and what failure modes are reasonable to assume. For people who have no firm foundation in real world mechanics, chemistry and physics, it is possible to spin all kinds of scary scenarios that depend on a series of impossible events. (Note: Just because I believe that there is always something that can go wrong, I do not believe that all things are possible.)
My prescription for progress is not “faith” in engineers or technologists. It is for people to approach challenges with knowledge, a questioning attitude, humility and a willingness to expend the resources necessary to operate safely. A thirst for maximizing short term profits or an attitude of blind optimism are both incompatible with performing difficult tasks in potentially dangerous environments.
Rod. I completely agree with you. I worked in research for power production: geothermal, some coal and gas pollution control, and nuclear. Later, after the strain of running my own company and meeting a payroll, I was a tech writer for several years at Silicon Valley companies.
Both experiences were “high tech” but they were very very different. In power production situations, if you didn’t understand what was happening, you had to do experiments to find out. You had to measure your ideas against nature’s results. You want to decrease NOx in combustion? Will the solid state catalyst help? Do an experiment!
On the other hand, in software development, if something seemed weird, you could ask someone. The software architect and developers knew how it was supposed to work, the QA people knew how it was failing and what work-arounds were in place. Somebody KNEW. It wasn’t people figuring it out about nature and trying to design the next experiment to check their hypothesis. Software was much more civilized and predictable. Software is pretty much entirely the product of human intelligence, and someone can explain it to you.
EPRI and the middleware firm I worked at about fifteen years after leaving EPRI (TIBCO Software) were both located on Hillview Avenue in Palo Alto. They might as well have been at opposite ends of the earth.
The emphasis on advances in software have mislead people badly.
I really like the opposition of “failure is not an option” and “failure is always a possibility”. Well said!
To me it seems that wind development is unlikely to double in efficiency from the current state of the art.
Solar on the other hand does have some possibility of doubling efficiency (maybe even more). I hope for a breakthrough but it may not happen.
Now nuclear. What is your opinion of the cost of an AP1000 (built in the United States) once past the first of a kind issues? Could the cost come down 10%, 50%, or even 100%?
believing in Moore’s law is like being in soviet russia and believing in the 10-year plan of the central planned economy. It’s the madness of our time that people tend to confuse statistics with planning. Just because something developed historically in a certain way (e.g. economic growth, computer speeds, oil consumption) doesn’t mean there is a relationship. Central planning has replaced sponaneous individualism in many places now. Everytime some future researcher declares “we’re running out of oil by the year 20XX” this is just a case of central planning: it ignores the power that lies in individual decisions made by millions of people in a free market of supply and demand.
This was recently mailed to me by the Canadian Nuclear List:
From: Andrew Daley
Date: Mon, May 17, 2010 at 2:51 PM
Subject: BP Oil Spill: What are the Lessons Learned?
To: Nuclear
All,
I was recently “volun-told” to give a Safety Meeting to my department. While the focus of these meetings is usually “Conventional Safety” I couldn’t help throw in a bonus topic dealing with Nuclear Safety.
That talk (attached, with corporate logos removed) was not “lessons learned” focused but was rather intended merely as “something to think about”.
A lot of good talk was generated. One disturbing development… my section manager subsequently asked some of his networked colleagues the question “what can the nuclear industry learn from the BP oil disaster?”
One of the answers was: “Nothing. They should be learning lessons from us”.
To me, that answer reeks of over-confidence.
Are we that far removed from TMI (which was before I was born!) and Chernobyl that we have developed a “can’t happen here” attitude all over again?
(Note – comment was edited and reposted to remove email addresses.)
The second sentence in that overconfidant answer was ok. BP could learn things from the nuclear industry, and almost certainly has adopted some systems in part (risk assessment based asset managment for example).
But to declare that the explosion and massive pollution at Deepwater Horison holds no lessons is a foolish waste of an opportunity to reinforce the need to pay attention to signs of problems early on and to have contingencies for all circumstances. And there are lessons still happening for those who would learn – for example, your first attempts to resolve the problem may fail, so work on the options.
Rod, just for shits and giggles, would you care to share your opinion of the singularitarians(e.g. Ray Kurzweil)l?
@Soylent – I have tried and failed to read Kurzweil. His work is not well written enough to be good SF. I do not happen to believe that it has any real value as prediction for the future.
I have been wrong before, but I am pretty sure on this one.
Speaking of technological realism replacing optimism, (or in this case the other way around) I see some are calling for the Gulf oil blowout to be sealed with a nuclear explosive as the Russians did with a few runaway gas wells:
http://howestreet.com/articles/index.php?article_id=13568
As illustrated here:
http://www.youtube.com/watch?v=CpPNQoTlacU
Personally I think this was a boy-with-a-hammer fix on the Russian’s part.
I think the concept is worthy of investigation. There are many variables that have to be considered, many variables, including how deep the deposit is, how much pressure it’s under, the amount of oil in the deposit, the rate of decrease of the pressure of the outgas plume, which is an indirect sign of how large the deposit is; if the pressure has dropped by very little, then I would be scared, very scared; I read that the pressure behind the blowout preventer was somewhere in the realm of 20000 PSI!?!?! (this is a very high number), at least when the accident occurred; this is versus around 2100 psi from water pressure at that depth; 5000 ft deep divided by 33 ft/atm = 151 atm * 14.7 psi/atm = 2100 psi, the progress of the relief wells, the possibility of flow induced cavitation to the borehole due to the rapid flow of the oil widening it the further that you go down, what the maximum feasible risk is.
Some have said “Yellowstone Caldera” levels of risk. (http://hardware.slashdot.org/story/10/05/13/1953208/Gulf-Gusher-Venting-a-Million-BarrelsDay ) ( http://www.examiner.com/examiner/x-8199-Breakthrough-Energy-Examiner~y2010m5d13-A-volcano-of-oil-erupting) In that case, forget about it. You don’t want to turn a disaster into a cataclysm. Though, if these estimates are right, it may already be a cataclysm. But the person who said “Yellowstone Caldera” didn’t appear to be an expert in the field of petroleum geology: in fact, in this article ( http://pesn.com/2010/05/23/9501654_Gulf_gusher_size/ ) he talks about how we need to pursue “magnetic energy” and “free energy” (and there ain’t any such thing as a free lunch) as the solution to all our problems. He also talks about how the “New World Order” is trying to cover things up – possible, but not likely.
Just because this man has some unique or faulty assumptions does not make his argument fallacious, though. It does not make it true, either.
There is a time and a place for everything. If the well can be rapidly contained or relieved, then continuing on the path that we are following may be acceptable, e.g. wait for the relief well to come in. If the situation is degrading, then perhaps we must consider other options, if they hold the potential to contain the situation, or at the least, to not make it worse. That means all credible options that stand to work and don’t make the situation worse. Using the only thing that has sufficient energy to fuse a plug in this rock in a matter of milliseconds might be our only credible option if it’s safe.
I don’t work for BP. I’m not a petroleum geologist. I’m not a nuclear weapons specialist. I don’t know what we’ve gotten ourselves into. But what I do know is that no option should be ignored – all options should be considered. Or a substantial part of the Gulf of Mexico is going to die.
Well it’s true the test ban treaties have clauses allowing so called peaceful uses of nuclear explosives. And I believe the US had at one time a contingency plan to do something like this is this sort of emergency. But I understand those plans were scrapped at one point because it was thought that there was too big a risk that matters could be made much worse if the shot went wrong.