A Think-tank Solution For Monitoring Radioactive Water Storge Tanks


When we hear reports of radioactive water leaking into the ocean from the [Fukushima Dai-Ichi] plant in Japan we literally have to keep ourselves from grinding our teeth. Surly the world contains enough brain power to overcome these hazards. Instead of letting it gnaw at him, [Akiba] is directing his skills at one solution that could help with the issue. There are a number of storage tanks on site which hold radioactive water and are prone to leaking. After hearing that they are checked manually each day, with no automated level monitoring, he got to work. Above is the wireless non-contact tank level sensor rig he built to test out his idea.

A couple of things made this a quick project for him. First off, he just happened to have a MaxSonar MB7389 waterproof sonar sensor on hand. Think of this as a really fancy PING sensor that is water tight and can measure distance up to five meters. [Akiba’s] assumption is that the tanks have a hatch at the top into which this sensor would be positioned. The box next to it contains a Freakduino of his own design which includes hardware for wireless communications at 900 MHz. This is the same hardware he used for that wireless toilet monitor.

We really like seeing hacker solutions to environmental problems. A prime example is some of the cleanup hacks we saw around the time of the BP Gulf of Mexico oil spill.


46 thoughts on “A Think-tank Solution For Monitoring Radioactive Water Storge Tanks

    1. Plastic bucket with water-level detector is not exactly new. Suggesting it to Fukushima, as you say, without testing it can take radioactivity, is a bit useless really. I’m pretty sure there’s lots of regulations around what you can store radioactive waste in.

      For the idea that perhaps the nuke guys should use containers that don’t leak, well, yeah. They probably know that.

  1. It’s a nice, subtle advertisement for the hardware they sell, but not a very useful solution.

    There seems to be no consideration given to the fact that it’s dealing with a radioactive environment other than in the statement of the problem.

  2. Wow! I didn’t know the all powerful arudino was radiation hardened and guarenteed for life-safety and radiation safety environments. Why pay thousands for these components when I can get them for pennies on the dollar!

    I love it how we can simply disregard conditions like the different density of radioactive water, radiation causing random bit errors, and all the other fail-safe systems that are [not] discussed in this article.

    The moron engineers running the nuclear plants don’t know anything. We should hire a bunch of arduino programmers!

    1. Today’s news that the level of radiation is 1800 millisieverts an hour, not 100, suggests that the engineers running the nuclear plants are morons…

      BBC article: “the company said the equipment used to make that recording could only read measurements of up to 100 millisieverts. The new recording, using a more sensitive device, showed a level of 1,800 millisieverts an hour.”

      I suggest you *think* a little next time you’re planning to post something dumb.

  3. I work with radar/sonar level sensors. They suck. I’d rather have a float switch any day. I get too many false readings with radar/sonar level sensors when stuff gets stuck to it or gets caught in the way.

    1. I’ve spent the last 5 years of my life trying to eliminate industrial float switches from the site. Replaced them with radars (I agree with you on sonar).

      The biggest complain I get is they are unreliable and they cause alarms. My reply is always the same, based on the data they are far more reliable than anything you had and the alarm is to tell you they don’t work, which is a luxury that is not afforded by a dumb on/off float switch.

      If I had my way discrete signals would not exist in an industrial environment. I like to know when my stuff is failing.

  4. Want to stop the leaks? Haul in enough rotationally molded polyethylene tanks to hold the water from one leaking tank. (In sufficient thickness polyethylene is a radiation barrier.) Second step, dry out the inside of the tank. Third step, spray the inside of the tank with Rhino Lining. Some sort of analog tele-operated robot arm to do the spraying would be a good idea. Could be built using air cylinders and solenoid valves for movement, with the most sophisticated part a video camera to observe the operation. The wires to run the valves would go to a control system outside the tank so radiation wouldn’t bother it.

    Once the lining solidifies, pump the water back in, repeat with the next tank.

  5. It’s actually true that I didn’t design in a rad-hard part. I’ve already designed geotagging geiger counters being used in Fukushima based on the Arduino platform and they haven’t had bit-flipping issues in the 1.5 years they were deployed. This doesn’t mean that the water storage monitors don’t need to be rad-hard, though. My point is that a solution can be found quickly and should at least be attempted. If it works, then fine. If not and rad-hard is required, then at least we have more information to build a solid solution. I’d prefer not to sit on my hands like TEPCO and just say that a solution doesn’t exist though.

    1. Just because you don’t “see” a flipped bit, doesn’t mean it hasn’t happened.

      Where is the error checking? Are you using ECC, a CRC, or even parity?

      A flipped bit in a simple Arduino sketch might not cause any problems, but in life-safety critical applications, disaster.

      Thats where the more difficult and robust algorithms come in, with majority detection circuits, and ECC, CRCs and other codes to ensure data integrity everywhere in the chain.

    2. Of course solutions exist. If they wanted measurements on this thank they’d pick up the phone, call Yokogawa, send them a PO and be done with it.

      There’s nothing here the DIY community needs to get involved with other than going out and beating a couple of upper management drones senseless. Though that would be quite hard given their lack of sense to begin with.

      The fact of the matter is that radiation hardening is irrelevant as to why your solution shouldn’t be deployed in an industrial environment. Industrial solutions exist. They are hardened. They are certified to high safety-integrity-level, and they have a lot of internal monitoring to ensure they are working correctly.

      Above all they aren’t even very expensive either.

    3. Personally I appreciate the effort. People over here don’t realize that it is still a problem, and is basically being ignored by tepco. tepco has had what, a year? What have they done? They still send in people daily. If you were that guy you would love this too. I really hope the tiny bit of media it is getting helps get the gov involved more (or doing more I suppose tepco is gov?)

      1. They’re going to send in people daily whether or not there’s a sensor on the tanks. Why? To make sure they’re not leaking.

        They’re not going to send the guys home just because some warning light isn’t on.

      2. If you can read Japanese there are constant updates on what is going… If you can’t read Japanese try Yomiuri’s English language site; http://the-japan-news.com/. Check the twitter feed box as some stories don’t appear on the front page.
        I have a hunch that all of the foreigners in Japan claiming that there is some cover up in progress and that the media are implicit in the cover up are strangely also those foreigners that couldn’t write their name in Katakana.

    4. Do you really, really think that the idea of telemetry monitoring hasn’t occurred to the people who are risking their lives on a daily basis to collect this information manually? I do not have the domain expertise to understand why this hasn’t been automated, but I do have enough respect for my fellow engineers not to just assume they haven’t thought of the solution at all.

  6. Ok putting aside radiation hardening, lack of FMEA studies to determine failure modes, the use of simple off the shelf electronics in an industrial environment which don’t even pay basic attention to corrosion etc, this is NOT a case study in wireless.

    Wireless networks exist. Actually they are used extensively for monitor tank farms. I mean there’s an entire IEC standard dedicated to WirelessHART protocol (IEC62591) which is an industrial hardened mesh networking standard.

    I work in oil and gas, and the only thing worse than no measurement is a wrong measurement. Giving an operator a number which may or may not be right is worse than making them go and check to be certain. I’m very glad this is a purely theoretical blog post.

    If TEPCO wanted measurement on the tank they’d call the first industrial instrumentation company that comes to their mind (Yokogawa is nice and Japanese) and simply have some wireless industrial instrumentation installed.

  7. This is water coming right off a reactor. This is bad stuff. This water is very hot, radioactively speaking. I would give the silicon chips a week or less before the radiation destroys them. Just like a EPROM you can erase with UV light, radiation will do the same thing. It does funny things to silicon. Epic fail…

    1. Radiation hardening isn’t just about keeping the radiation out, it’s about ensuring components are chosen that can survive the radiation, understanding failure modes and detecting them, and above all having extensive circuitry that can catch errors that result when electronics are exposed to radiation.

      There are multiple IEEE standards which cover the design of instrumentation in nuclear services, and usually local laws / standards which specify that instrumentation MUST be certified.

  8. How about we take Akiba’s board, stick it in a 100mm block of lead and blast it for a couple of weeks with radiation (seems to me the 1.6MeV gamma of Cs-134 is the biggest issue).

    Not everything needs to be radiation hardened, you just need sufficient shielding in some cases. Anyway, if it dies at least we did another fun experiment!

    1. No the biggest issue is that we have a non-fault tolerant device, not fit for purpose, hacked together which *may* fail. This is NOT at all suitable for a critical industrial service.

      It’s not about just radiation hardening. It’s everything. No failure mode analysis. No internal diagnostics. No reporting of a failed link. Not being tolerant to external interference. Complete lack of adherence to standards.

      Combine all this will the knowledge that a) devices on the market actually exist that will do what TEPCO want, and b) the single worst thing you can do is give an operator a number that is potentially wrong. It’s far worse than not having a number at all as people tend to believe instrumentation until they are given direct indication that an instrument has failed. THAT is what leads to major incidents.

  9. I don’t know the details of the situation, but if the tank did have sensors like this, would the problem have been caught any sooner? If they were monitoring the tank manually twice daily and it wasn’t caught, then I’m not sure automated monitoring would have caught it either.

    Manual tank levels are reliable and trusted.

    You’ve also got to take into account this isn’t a bin of water. This is a holding tank of heated water. The temperature of the inflow, ambient temperature, evaporation etc. will all cause the tank levels to change. You are ultimately measuring height, inferring volume and then inferring mass from that. These extra steps can introduce problems.

    1. While a lot of what you said is true a manual tank dipping is not reliable, though it is trusted and that leads to a lot of problems.

      The second part of what you said is a direct case for electronic systems as it’s trivial to tie a tank ullage to current volume given temperature and even flex in the tank shell via a properly prepared strapping table.

      1. It would be a direct case for electronic systems if these weren’t tanks that were in-situ in an area with significant radioactivity. I don’t think it is trivial to do that when none of the sensors are in place. A single electronic depth reading doesn’t provide anything beyond a manual reading.

  10. Maybe they were too busy monitoring the water levels in the tanks to notice the water puddling on the floor?

    The root of the problem is that the cooling systems are still leaking, so if they can’t stop or reduce those leaks, they’ll have a continuously expanding volume of water to store (anyone know what the half-life of reactor cooling water is?). Presumeably this water is also contaminated with random crud from wherever it leaked to, so they don’t want to just put it back in the cooling system (although if they could find a way to filter it adequately, that might be the best solution).

    Secondly, they were in a hurry so they used tanks made from sheet metal panels bolted together, with plastic seals between the panels. To me, this would call for a backup plan for what to do when (not if) the tanks started to leak. Maybe all of their best engineers work for Toyota rather than TEPCO?

    1. As a former nuclear operator and maintenance person in the u.s.and now a field engineer in advanced water treatment worldwide I feel compelled to comment. all of this isnt necessarily a reply to the above comment.

      1. Leakage from underground and above ground tanks happens regularly. Technology has brought about an inflatable heat-set plastic liner for pipelines and theres a chance this could be adapted to this use. No need for a robot to line tanks SCBA or air hood wearing humans do this regularly.

      2. Alot of the instrumentation available to the commercial market isnt much better than that arduino, IMHO, due to the random firmware issues, bad design, bad wiring by electrical workers, problems with interference, grounding, etc.

      But, Rosemount, Yokogawa, Itt, and others make nuclear grade instrumentation that works just fine.

      3. Using a tape measure with a plumb bob on the end “sounding tape” is the worst way ever to maintain logs of a tank level, We screwed it up regularly on our ship when I was in the navy. At Yorktown Naval Weapons Station I visited a site where 212 wells were drilled and they were pumping steam in and number 4 fuel oil out of the ground due to a tank leak that went unchecked for 80 years or so.

      4. Probably the only way to fix this problem both at Fukishima and others is to never build an undergound tank again. Like an aove poster said “didnt they notice the water on the floor”. Alot of times the answer is no, it leaks undergound or in a vault where no one goes due to the radioactivity. Workers arent allowed to get “unnecessary” dose due to a fundamental misunderstanding of radiations effects. See Hormesis and understanding radiation. – Luckey, Boyar, et al.

      5. Actually I like the essence of this hackers point-of-view. If we ask ourselves, “isn’t this as simple as figuring out when we lost some of the content of this tank?” Then the asnwer is yes. Let’s use positive displacement flowmeters on both sides of the tank and a direct reading level indicator and compare it with a sightglass regularly. Lets develop a control philosophy that says hey I put in more than the sum of the current level and whats been pumped out somethings wrong.

      What this whole thing boils down to though is sensational reporting. Everything is about your point of reference. Ive worked in nuclear plants. Their threshold for personal contamination is very low for instance. A “hot particle” (a small amount of radioactive material that is on you, wont come off, and you cant find using the handheld meter) will get you held up in “nuclear limbo” for quite a while. A “tungsten” used for tig welding and available at harbor freight cant be brought anywhere close to a monitor.

      How radioactive is this leaking material compared to whats already been released?

  11. Political problem, indeed… How is more information, such as this hack provides, going to help with the political coverup reportedly going on?
    Japan’s nuclear watchdog has now declared the leak of radioactive water from Fukushima a “state of emergency.” Each day, 300 tons of radioactive water seeps into the ocean, and it’s now clear that TEPCO has engage in a two-and-a-half-year cover-up of immense magnitude.

    Just how out of control is the situation at Fukushima? It’s so out of control that TEPCO recently had to admit 10 of its workers were somehow — yeah, see if you can figure this out — sprayed with highly radioactive water while waiting for a bus.

    “The workers’ exposure above the neck was found to be as much as 10 becquerels per square centimeter,” reports Bloomberg.com

    How exactly did highly radioactive water manage to find its way to a bus stop in the first place? TEPCO isn’t sure. It’s confusing with all those radiation alarms going off all the time. In order to concentrate, the company has found it’s easier to just disable all the alarms and pretend nothing’s wrong.”

    “At this current time in July of 2013, Fukushima is 80 to 100x more expansive and more intense — letting out about 100x more of the radiation of Chernobyl,” reports Dr. Simon Atkins Phoenix Rising Radio on a BlogTalkRadio interview.

    “The problem with Fukushima is that it’s not only continuing for 865 days… I mean, let’s wrap our minds around that for a second — it has been leaking out radiation in increasing volumes for 865 days.”

    The U.S. government, of course, plays along with the charade because its own top weapons manufacturer — General Electric — designed and built the Fukushima Daiichi power plant in the first place. And the design decisions made by GE, such as storing spent fuel rods in large pools high above the ground, now look not just incompetent but downright idiotic. It turns out there was never any long-term plan to dispose of the spent fuel rods. The idea was to just let them build up over time until someone else inherited the problem.
    Read more here:

    So, the real question with so much misinformation being provided by both sides in this issue: How can we find out what “reality” we are allowed to know about this issue?

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.