Slashdot Mirror
How To Monitor Leaky Radioactive Water Tanks
freaklabs writes "The radioactive water leaks are getting worse at Fukushima Dai-Ichi. In a recent New York Times article, it was mentioned that TEPCO didn't have a reliable way to monitor the water storage tanks for leaks. I decided to write a tutorial on how to wirelessly monitor water levels in storage tanks."
I guarantee that there is frustrated engineer with a workable solution who spends half of all
his days trying to argue for the installation of monitoring equipment, but the organization
- doesn't really want to monitor the tanks
- is too incompetent to execute anything
- has a turf war over who is supposed to be monitoring the tanks
- is hung up on acquisition/budget issues
- is hung up on safety protocols
So now we'll never know whether they remembered to take into account radiation hardening.
So this is fine when it concerns non-radioactive water, but this solution wasn't tested in an environment where the radioactive levels are higher than usual, there was no test case in the story for that. Will the electronics live long enough? Also what about humidity, how long before this stops working because of higher humidity levels?
You can't handle the truth.
Because your website needs more hits and the experts in Japan certainly never thought of some of the most obvious ideas, yes?
You may not be familiar with japanese culture. I am, at least a bit.
In the US, this admition would translate to "we can't be arsed to give it some attention".
In Japan, this is a major loss of face and could well mean the end of someone's career.
This face thing is a major problem in many cases in Japan, because people won't admit to mistakes until they can't hide them anymore. Yes, even more so than in the West.
It would be fantastic if someone from the japanese geeks involved in the whole thing would read /. and rip your blog-wiseassing to shreds. Unfortunately, that's unlikely and so your ego can feed on a false sense of superiority.
Assorted stuff I do sometimes: Lemuria.org
http://en.wikipedia.org/wiki/Radiation_hardening
Maybe the submitter wrote about it, but the site is unavailable right now and his summary certainly does not reveal he knows a thing about the special considerations of electronics in radioactive environments. There is a reason we (in the East German army) had big tube-powered big bulky radios instead of smaller transistor-based ones.
The radiation in these tanks is easily stopped by the tank wall. (Its almost solely Beta radiation). So climbing the tank is not particularly a problem.
Water is always being pumped into and out of these tanks (they are used to circulate cooling water for shut down reactors and the separation plant where radioactive elements are separated). As such, water level in the tank is not static, there are surges as pumps start and stop, etc. Think of the tank as a buffer in a continuous flowing circuit. There are systems to make sure there is always sufficient water in the circuit, and water may be added at locations far removed from the actual tank. Its vitally important to make sure there is adequate cooling water, it can never be allowed to run dry.
When you view it this way, missing a couple hundred gallons over the course of a month is not something you can count on detecting by monitoring water lever in a tank, because it fluctuates naturally, loss will be automatically compensated by new water additions.
So thanks for playing along, but I believe this issue is best left to the big boys,(even the ones you might, in your make-believe environment, consider to be incompetent). The problem is much more complex than you know, and won't be solved with your cute little lash-up toys.
Sig Battery depleted. Reverting to safe mode.
I don't get it; If your water heater was leaky, you'd notice it right away. Same if your gas tank, radiator, or brake fluid reservoir on your car was. In fact, the main way people figure out something is leaking is because something is present where it shouldn't be, and a supposedly-contained source is nearby.
You don't need complex wizz-bang devices to figure this out. You need the Mark I Human Eyeball. TEPCO knew, okay? They didn't want to know, so they ignored evidence that it was leaking. "Well, the tanks emptied out... but it must have just been normal evaporation", or "we expected a certain amount of leakage", or "we were understaffed," or "it was the contractor's fault," or any other rationalization you can think of. The problem here is not technology and it won't be solved by technology.
The problem is management didn't want to know it had a problem, because plausible deniability means no responsibility. So they will go to incredible lengths to avoid noticing the problem. You can't slashvertise your way into a solution here... "wifi sensors! That'll fix it!" Okay... who's going to monitor the sensors? What are the sensors actually sensing? Mind you... sensors being improperly read is what led to the Three Mile Island disaster. Do you trust your $12 wizz-bang to do the job of a trained nuclear engineer? This is what it all comes down to: The tanks were leaking, and somebody noticed. I don't know who that somebody is, but somebody knew enough to look. Whether they did or not...
TEPCO management needs to be dragged to Geneva and held for crimes against humanity. No, I'm dead-serious about this... Japan has a long an inglorious history of allowing epic amounts of corporate failure because it's not in their culture to admit wrongdoing. Trains fly off tracks and crash into apartment complexes and outside investigators conclude that a punitive and stress-inducing corporate culture was what primed the young train conductor to race around the bend too fast to stay on the tight schedule... and the corporation, faced with dozens of fatalities... says everything is fine and keeps the policy as-is: It was the conductor's fault. He couldn't handle our "high standards". This is a prime example of Japanese culture people. It's toxic and it needs international attention and condemnation.
It does not need a wiz-bang sensor monitor. It needs a gun to the head and a "come with us, we're taking a long flight to your trial, asshole".
#fuckbeta #iamslashdot #dicemustdie
The reality is TEPCO doesn't want the radiation monitored. For the same reasons the beef industry doesn't want cows tested for Mad Cow. The absence of testing allows for plausible deniability.
Wow, what an amazing totally thought out plan!
I love how you thought about radiation causing random flipped bits, the need for adaquete power and shielding, etc.
Amazing, I didn't know the Arudino was rad-hardened and certified for use in safety critical industrial applications. I will instruct everyone I know to use the Arduino for everything. Why pay hundreds of thousands when I can get all the rad-safety and life-safety cerified components for pennies on the dollar!
Sorry mate...
I used to hack for a company that would do this for about $25 a month + installation fees and hardware costs. Although if someone had used the words "nuclear" we'd probably have consuled a lawyer and said hell no.
Complete with certified electricians, low power, cellular, satellite, and 802.x plans, off site data storage, backups, recovery, control, and a pretty little website that could get you the data a dozen different ways including read into your ear over the phone. For the big boys some of our stuff was even in escrow.
But you see, you didn't use radiation hardened electronics.
You're using an arduino which even in its deep sleep is consuming at least 4x as much power as we did while awake nearly a decade ago. And you've got no backup or emergency cry. Or thermistor, or cut off, or sensor stabilization routine...
Look, it's a noble effort for a home hacker, and the proof of concept if you didn't believe it was possible is cute.... and to somebody who can't go corporate I can see why you'd use this at home ...
But you really haven't addressed any of the core engineering issues, and have added an additional maintenance problem.
Now frankly, this case is so straightforward it's nothing a competent engineer and developer can't solve -- you've got the mostly right sensors, you just need to package it up properly. And that is the problem. You're going to pay a small fortune for the certifications you'll need to touch that industry.
Also, as others have said -- the problem isn't that they can't fix it. It's that TEPCO is culturally poisoned. But maybe the VP of engineering will save some face by falling on his sword.
When you have a constant inflow and outflow, if there is any calibration error on the flow meters that error will integrate over time to look like a gain or loss in level. You could imagine having flow and level sensors everywhere, but you still have issues with temperature variations (the thermal expansion of water isn't all that small). Still you could do it with arrays of temperature sensors in each tank, level sensors and calibrated flow-meters. Starts to get pretty complex.
There are lots of commercial solutions out there for level monitoring with distributed data systems but when you are trying to find leaks that represent a small fraction of the total volume it gets difficult.
It is still possible that a simple level monitoring system would help - but if so, they might as well use a commercial system .
Hi. Actually these are storage tanks and are designed to hold water. Once they are filled, its unlikely water is pumped out unless there's a suitable place to dump it. That doesn't seem to be the case.
Providing these are storage tanks whose content is waiting until it can be cleaned up I would suggest to freeze them.
"The likes of Facebook and WhatsApp are free to those whose privacy is of zero value."
No what this needs is a proper industrial solution fit for purpose, not some home made solar gear with an arduino strapped to it. We use wireless guided wave radars from Emerson at work and they last in excess of 10 years on a single battery charge.