Comments about technological history, system fractures, and human resilience from James R. Chiles, the author of Inviting Disaster: Lessons from the Edge of Technology (HarperBusiness 2001; paperback 2002) and The God Machine: From Boomerangs to Black Hawks, the Story of the Helicopter (Random House, 2007, paperback 2008)

Sunday, March 27, 2011

Cooling Fukushima: If there's a plan it's hard to see

In my last post I pointed to the Union of Concerned Scientists' strong recommendation that evacuation should begin immediately out to wider distance than 12 miles. On Friday the Japanese government did begin "encouraging" people to start leaving a zone extending to 18 miles. One reason may have been this airborne study from the Department of Energy's National Nuclear Security Administration indicating a radioactive plume extending out to the northwest and beyond the mandatory evacuation radius, released March 25. Here's slide 3:

I think the reading public is close to giving up on trying to follow this remarkably opaque disaster, which is compounded by a company apparently dead-set against sharing information with the government or any other non-nuclear outsiders, multiple reactors of different makes, and faulty measurements or no measurements.

Just to add to the confusion, we have international radioactivity standards that inexplicably shifted from old and familiar terms like rads, rems, and curies to grays, sieverts, and becquerels. Here's a glossary that compares old and new terms. Example: "One gray is equivalent to 100 rads." Good luck figuring out the dosages that justify evacuation!

Meanwhile, we have our daily dose of confusion over which reactor did what, and why. One day a unit is declared stable and the next day there's a cloud of smoke or steam or dust that nobody can explain. Things are supposedly well in hand one day, and the next we hear of radioactive iodine briefly showing up in Tokyo's water supply. (I haven't see any indication that Tokyo residents are at significant risk from this, but the event is odd given that the nearest point of the watershed of the Tone River -- which supplies most of Tokyo's water -- is more than a hundred miles southwest of the reactors.)

Yes, there are tables and reports showing the status of all six reactors and these are of help, but none that I've seen so far are fully authoritative, prompt, and thorough. That's because Tepco is holding back the information that would make such updates possible.

To illustrate how frustrating it is to keep track, there were reports yesterday that radioactivity had reached astonishing levels in the basement of Unit 2's turbine building, according to this update from the Bulletin of Atomic Scientists. Citing official sources, the BAS said the level of Iodine-134 in Fukushima Dai-ichi Unit 2 had reached 2.9 billion becquerels/cubic meter. Then this item from NHK said the level was 2.9 billion per cubic centimeter, which would be much higher. 

Then the authorities retracted all those numbers today as too darned high, according to this piece in the LATimes.  

Op-ed writers have been busy opining on the future of the nuclear power industry, but the information quality out of Fukushima is so poor that I'm refraining from conclusory statements such as "the existing fleet is pretty safe, except for outliers like these," or "let's drop the whole idea of electricity from fission," or "the new generation of reactors relying on natural-circulation cooling are the really safe ones." 

It will be at least a year until objective information comes out on the course of events at Fukushima. Right now, it's enough of a challenge just to figure out what Tepco is trying to do.

Setting aside the latest flap over Unit 2's basement numbers, Tepco says that radioactivity levels went up shortly before three workers were exposed to radiation while wading through a basement. Tepco said that it had assumed the area was still safe for workers. 

Obviously that was a flawed approach to safety, but it also suggests either that a new leak allowing water from the reactor core into the turbine-building basement has opened or an accidental criticality is underway somewhere, adding to the iodine and cesium inventory (less likely but troubling if so).

So, Question 1: BWR control panels are supposed to have an instrument called the neutron flux indicator (here's a link to a patent document for one), showing the level of fission in the core. What are the readings from the neutron flux indicators in the two control rooms for Units 1-4? Set aside the mysterious "neutron beams" detected in the vicinity last week: the neutron flux indicators are supposed to be definitive. In this post, the Union of Concerned Scientists studied official photos of the re-opened control room of Unit 2 and compared critical instruments to those in a BWR simulator, pointing out that the Unit 2 control room didn't seem to be functional at the time of the photography.

A credible case has yet to be made that the cooling circuit "repairs" to the cores of the most hazardous units are going to work in any manner other than dumping water in one end (via spray heads or feedwater pipes) and having it gush out somewhere else. Injected water seems to be going into various basements, either under the reactor buildings or the turbine buildings, and then finding its way out to the sea.

Question 2: Has Tepco been using something like a trash pump to dump water from the basements directly to the sea? 

If the promise from Tepco is that it plans to restore a closed circuit to recirculate the primary cooling water that comes into contact with the reactor core, so that it isn't dumped into the basements or the ocean, this will be a truly formidable job at Units 2 and 3. 

Recently I looked into the subject of "the world's toughest industrial problems" for a television producer, and nothing I came across in that research approached the difficulty of this one. A workmanlike job couldn't even start until an expert assessment of the equipment status from reactor building to turbine building to control rooms, which would need to assess all damage due to the earthquake shocks; overheating; overpressure; stress corrosion cracking; salt deposits around the fuel rods; the operation of gate, check and pressure-relief valves; and the unintended effects of attempted workarounds. I suspect something like this has been distributed to experts at GE, Hitachi, and JAIF

Question 3: If so, can we get a look at the damage assessment?

Let's assume that experts have mapped out what equipment is broken in each reactor, and what can be relied on; and they've been able to procure suitable replacement parts from check valves to instruments to elastomeric seals to piping. Finding spare parts is not easy for way-old equipment like that at Fukushima, but given that a number of similar reactors are still in service, it should be possible, like finding parts for 1970s muscle cars.

Which leads to Question 4: Can we get a copy of the new primary cooling circuit diagram being used to guide work at Units 2 and 3, assuming that something more is planned than a once-through use of cooling water? Given that the upper walls and roof of Unit 3 has collapsed on the upper levels of the main reactor structure, I'm curious how repairs will proceed. Here's a link to a useful diagram of a BWR, drawn in the style of a subway network.
All in all, the time for a "fix" of the primary cooling circuit in any conventional sense seems to have expired. One reason is that radiation levels in some areas is so high that Tepco is running out of contract workers to handle high-dosage jobs. 

Question 5: Exactly which contractor companies are providing temporary workers for Tepco, and which neutral health agency is monitoring the health of every worker, not just the ones sent to the hospital? 

Assuming for the sake of argument that the contract labor supply dries up, and that the government doesn't turn to ordering troops in, that leaves teleoperated machinery, robots .... or abandonment in place to let the radioactivity die down. I think the latter scenario is increasingly likely at Units 2 and 3.

Sidebar: Temporary, unskilled workers brought in for high-dosage reactor-maintenance jobs are called "jumpers" in this country. They're outside non-technical people like farm workers, clerks, or auto mechanics paid to learn a simple job like replacing a nut in a confined space. Then they suit up, put on a dosimeter, and go in. The allowable time for any given jumper to do his job may be only minutes, if the radioactivity level is high. So it can take several jumpers to remove a nut and put a new one on.

Wednesday, March 23, 2011

Salt-Encrusted Fuel Rods: A growing concern

As I noted in a post four days after the Japan earthquake, salt accumulation due to emergency use of seawater to cool the damaged Dai-ichi reactors might seriously interfere with the cooling of fuel-assemblies in the primary containment vessels. The reason is that as the seawater boils off, it leaves mineral deposits at the metal-water interface. This crust accumulates with time and reduces the heat transfer efficiency. It's possible that uranium metal inside that crust will melt with the trapped heat of decay regardless of how much seawater is pumped in or dumped in. 

Now worries about salt's side effects are on the rise. This item notes concerns by the French authorities.

Here's an excerpt from the NYTimes published March 23, "Optimism on Hold at Japan Plant as New Problems Arise:"

Richard T. Lahey Jr., who was General Electric’s chief of safety research for boiling-water reactors when the company installed them at the Fukushima Daiichi plant, said that as seawater was pumped into the reactors and boiled away, it left more and more salt behind.

He estimates that 57,000 pounds of salt have accumulated in Reactor No. 1 and 99,000 pounds apiece in Reactors No. 2 and 3, which are larger.

The big question is how much of that salt is still mixed with water and how much now forms a crust on the reactors’ uranium fuel rods. Chemical crusts on uranium fuel rods have been a problem for years at nuclear plants.
The NYTimes article continues:
A Japanese nuclear safety regulator said on Wednesday that plans were under way to fix a piece of equipment that would allow freshwater instead of seawater to be pumped into at least one of the reactors.

He said that an informal international group of experts on boiling-water reactors was increasingly worried about salt accumulation and was inclined to recommend that the Japanese try to flood each reactor vessel’s containment building with cold water in an effort to prevent the uranium from melting down. That approach might make it a harder to release steam from the reactors as part of the “feed-and-bleed” process that was being used to cool them down, but that was a risk worth taking, he said.
Why is that particular risk worth taking? I read the regulator's comment as suggesting there are concerns that the fuel may overheat, then slump to the bottom of the primary containment. If the fuel rearranges itself into a glob that's not under the influence of control rods and boron neutron absorbers, and has some water remaining to act as a neutron moderator, it could lead to a criticality accident.

That means an unwanted, uncontrolled fission reaction. In this case fission would restart in fuel that had been undergoing fission before the March 11 earthquake, but that had stopped reacting soon after the earthquake, as each operating reactor's fission-damping control rods were inserted automatically. (Unfortunately the fission byproducts in the fuel rods didn't stop decaying, which produces the unwanted heat the reactor operators have been struggling with.)

Uncontrolled fission would add to the inventory of radioactive cesium and iodine. It would heat up the vicinity even more than decay heat. It would be detectable by a sharp, sudden rise in hard radiation, as in the 1999 criticality accident at the Tokai fuel facility. If it's of any reassurance, there have been at least two dozen accidental criticalities in Atomic-Age history and none led to an atomic blast. Some stayed critical only for an instant, but in other cases criticality lasted for minutes, irradiating people in the line of sight. 

Out in the open, a critical mass that produces a lot of energy tends to be self-extinguishing; it gets so hot that it melts or otherwise takes itself apart, which eliminates the conditions necessary for a critical mass. So it stops reacting. But if a mass of melted uranium or uranium/plutonium is trapped inside a containment vessel and can't blow itself apart in a pressure wave, the stopping point for the uncontrolled criticality is harder to predict.

Right now there are just two principal barriers against this happening: (1) that the overheated fuel assemblies stay intact and don't crumble or melt into a fissionable shape, and (2) the boron added to the water being injected by emergency pumps. 

That's not many barriers left, given the deleterious effect of salt crust on cooling. It suggests to me that experts should be thinking about the physics of a possible criticality, and gathering remotely-operated equipment to have on hand given the high radiation levels that would be encountered if it happens. It may already be a subject of quiet discussion in Japan nuclear circles. 

Maybe it won't be a problem: maybe a criticality will just boil away the coolant water that moderates neutrons and somehow settle down without something really bad happening, as with the ancient African natural reactor I mentioned in an earlier post. For now, all we amateurs can do is speculate. 

Perhaps coincidentally, Ed Lyman the Union of Concerned Scientists posted a statement on March 23 that is rather critical of the Japanese government for not extending the evacuation zone around the reactor now, rather than waiting to see how things turn out. Excerpt:
Despite the US advisory [that American citizens keep a 50-mile distance from the reactors], the Japanese government is still maintaining its current order, which is evacuation only to a distance of 12 miles, and “shelter in place” for those between 12 and about 18 miles from the reactor site. “Shelter in place” means that people are directed to stay indoors and seal their windows and doors. Our assessment is that the Japanese government is squandering the opportunity to initiate an orderly evacuation from larger areas around the site–especially of sensitive populations, like children and pregnant women. It is potentially wasting valuable time by not undertaking a larger scale evacuation at this time.
(Note: There is a news report in the Japan Times that a low-intensity neutron flux was detected 13 times in the general vicinity of the damaged reactors. That's not necessarily a sign that criticality has happened, because fissile materials -- particularly plutonium -- release neutrons spontaneously.)

Thursday, March 17, 2011

Techno-catastrophes: Nature's new force multiplier

This article in the Washington Post by Steven Pearlstein ponders why major catastrophes seem to be rising in frequency.
Here's another perspective across two centuries of technological trouble, from the concluding chapter of Inviting Disaster (2002):
“Some of the worst accidents in the field had a flawed technology meeting an unexpectedly strong force of nature.” 
In addition to the series of catastrophic breakdowns triggered by the 9.0-magnitude Sendai Earthquake at the Fukushima Dai-ichi reactor station (see this link for updated, astute analysis), nature-weakened technology played a starring role in:
  • Titanic loss, 1912
  • Tay Bridge disaster in a storm, 1879
  • Banquio-Shimantan dam failures in a typhoon, 1975
  • Hartford Civic Center roof collapse following snowstorm, 1978
  • Alexander Keilland and Ocean Ranger offshore platform disasters, during storms
  • New Orleans levee failures following Hurricane Katrina, 2005
  • Deepwater Horizon blowout at the Macondo 252 high-pressure, weak-formation prospect, 2010
I see four risk factors behind the worst of these:
  • A group of people are in a remote and potentially dangerous place. Their safety depends entirely on a machine working properly;
  • Problems have been showing up in this machine, long before the crisis, even in benign conditions;
  • Leaders did not follow up on these danger signals when they had plenty of time to prepare;
  • Which left an opening for a natural force to tear away the facade of safety.
Once the system fracture is fully underway, any improvised reactions by operators in such extreme conditions tend to be ineffective or even counterproductive, at least in the short term. But in the long term we have the opportunity to learn and do better.

Tuesday, March 15, 2011

Quake-driven nuclear problems: Seawater plus uranium equals _?_

Adding to my list of questions on the effects of the Sendai Earthquake on the Fukashima Dai-ichi plants.

What's the effect of injecting seawater in the Fukushima BWRs, other than its obvious corrosivity?

In typical boilers, any time mineral-laden water is used, a mineral scale develops at the interface between steam and water. It's most noticeable at the waterline, but also happens below the surface where bubbles form. Thousands of gallons of seawater have been boiling into vapor the last couple of days inside the primary containment vessels of Units 1 and 3, leaving hundreds of pounds chlorides (about three percent by weight of seawater) behind.

If mineral scale is building up in the primary containment vessels, given that this hard substance is not a good conductor of heat, it would make the cooling of the wrecked fuel rods more problematic. I know the melting point of salt is well below the melting point of uranium but mineral scale could aggregate onto the now-oxidized zirconium alloy of the ruptured cladding.

So: is mineral buildup playing into recent information about what is happening at Units 1 and 3, the first to have their secondary containment structures blow open? According to this report the cores inside the primary containment structures at Units 1 and 3 aren't cooling as they should be, given the seawater being injected by fire hose:
"However, by Monday night there were reports that efforts to continue cooling Units 1 and 3 might be running into problems." 
 I understand that core cooling at 1 and 3 are not of the same urgency as dealing with the latest crises: Unit 2 (possible breach in primary containment, maybe around the wetwell) and Unit 4 (possible fuel fire in a spent fuel storage area, due to uncovering of stored fuel assemblies as water leaked out of the pool and was not replaced.)

Link to Google satellite imagery: here

Sunday, March 13, 2011

Japan Temblor: Reactor links and questions

A few of my questions about the Fukushima I cooling crisis at Unit 1 and possibly Unit 3:

What's the cloud visible in videos? It looks like concrete dust to me.

How much of the crisis was really due to lack of emergency generation capacity, and how much to difficulty of getting access to controls and instrumentation?

How did radioactivity get into the control room?

Is the primary containment intact?

What's the status of the spent fuel in the pool next to the drywell? Once the secondary containment shell blew out, it would appear that a fair amount of structural steel and concrete fell in there.

I'll be adding some links, but in the meantime here's a good one from David Lochbaum on the workings and layout of the troubled reactor.

Thursday, March 10, 2011

Texas Power Grid's Freeze-Up, Five Weeks On

We're waiting for an authoritative report explaining why dozens of generators were unavailable, or tripped, on February 1-4. And we're waiting for a complete list of units involved. The only list on the ERCOT site is a partial one dated February 16. ERCOT might stand on its rights under the law of deregulation and wait until April before releasing a final list. 

I count roughly 60 distinct power stations; most of the entries are multiple units within a single station. An example of a single station having multiple units is Midlothian Energy LP in Ellis County, which according to the ERCOT list had three gas turbines unavailable sometime during the declared emergency. 

See this 2007 power-plant list from the Texas Commission on Environmental Quality for names of power plants along with information about the prime movers, their manufacturers, and the nameplate power output.This makes it a little easier to understand the ERCOT list. 

The majority on ERCOT's list are gas turbines. Many are the more efficient combined-cycle units in which exhaust gases are routed through an HRSG (short for heat recovery steam generator), which uses the heat to boil water to run a steam turbine. 

HRSGs are an important component in today's plants that raise the thermal efficiency of gas-fired units but they can be finicky during ramp up and shutdown, and positively troublesome if not protected from freezing. Among the HRSG components known to be vulnerable if not kept warm heated: condensate drains, water-level sensing lines between steam drums and SCADA transmitters, instrument-air lines, and end-caps of headers.  

Some geek-friendly information about the hard freeze of February is in this document filed with the state Public Utility Commission by El Paso Electric, "Report on Weather Event: February 2-4, 2011."

El Paso Electric (which isn't within the zone covered by ERCOT, by the way) said there was no shortage of fuel gas to its plants. The problem was intense cold that overwhelmed the usual half-measures like electric heat traces. 

Power plants can run fine at much colder temperatures (like ours in Minnesota) but only if they are fully prepared.

On p. 13 of the El Paso report there's a bone-chilling description of how employees and contract workers took up positions on catwalks or ladders 10 to 15 feet apart, wielding blowtorches to heat up a water line between the steam drum and the sensing transmitter. A detailed chronology in Exhibit C describes how units would trip; come on line after much effort; then trip shortly afterward.

Texas households who lost power but are facing higher bills even so are not happy. Said North American Electric Reliability Council (NERC) Chairman Gerry Cauley to a Senate committee's field hearing three weeks ago:
"The events of February 2011 give me cause for significant concern. These are not new issues. We’ve had severe weather before.  We must continue to ensure industry is learning from the past, and must not allow institutional knowledge to fade.  These issues must be kept at the forefront."

Saturday, March 5, 2011

Space-mystery-plane X-37B: New and old together

The unmanned X-37B spacecraft is scheduled for a lift into orbit later today, after a weather scrub yesteday. It will fly itself back to the runway at Edwards AFB sometime in the next nine months. This news item is breathless about the mystery plane. There's no great mystery, though: it's a technology demonstrator for military work to come ... if the US can afford it.

The Union of Concerned Scientists expresses anxiety in this statement, which warns that the X-37B is a toe in the door for the USAF's long-term designs on space domination.

Or space-based defense, at any rate. This was a theme in speculative military literature a decade ago. See this 2001 paper from an officer-student at the Air University, X-37 Space Vehicle: Starting a New Age in Space Control? on how such reusable vehicles could help the USAF establish a presence in space, or at least give it a mobile, fast-response launch and recovery capability that could be usable for anti-satellite work or to hurl kinetic projectiles before an enemy's land forces can reach its major objective or, better yet, before the attack starts. See this RAND Corporation paper on what's known to tacticians as pre-emptive attack during the halt phase. More context on the X-37B is at MilSat magazine.

Shaped like a small space shuttle, the X-37B is to ride into space inside the nose fairing of an Atlas expendable launch vehicle. This the second X-37B to be readied for launch. The first one (denoted OTV-1) went into orbit last year, using its own navigational prowess to return nine months later. 

One item of historical interest is the rocket engine used for maneuvering the X-37B once in orbit, the venerable Rocketdyne AR 2/3. It runs on jet fuel and 90% hydrogen peroxide, also known as HTP for high-test peroxide. HTP is dangerous stuff if it combines with contaminants in a storage tank, but highly valued as an oxidizer for torpedoes and rocket engines.

Remember the movie The Right Stuff? The end of the story showed test pilot Chuck Yeager trying to break an altitude record in an F-104. The film used a standard model F-104G, but in reality Yeager flew a high-altitude modified version called the NF-104A, which in addition to the jet engine used for lower altitudes had a rocket engine very similar to that used on the X-37B. 

The AR 2/3 rocket engine on the NF-104A, which took over at high altitude after the pilot turned off the J79 jet engine, was mounted at the base of the airplane's vertical stabilizer. A photo of the Rocketdyne in action is at this Wikipedia page.

Thursday, March 3, 2011

Updates on Previous Posts: San Bruno and Grand Chancellor

San Bruno, CA, pipeline blast, Sept. 9: NTSB posted a major release of investigative documents on the docket page on Tuesday, timed to a public hearing this week. A NYTimes wrapup of the hearing and highlights of the information is here. Most news accounts are about the faulty welds and poor recordkeeping on Pipeline 132, and the failure of PG&E to install automatic shutoff valves in case of catastrophic pipe failure. Still going through the material myself. Here's a link to an interesting forum about the emergency response, at Firehouse World.

Hotel Grand Chancellor, Christchurch, NZ: Still standing after earthquake on Feb. 22. The 26-story building is stable enough that US&R teams have been able to enter and check portions of the building for remains. Here's a summary of work by international US&R Teams in the building. Stairwell damage is making passage between floors difficult. 

No decision by authorities yet on whether to bring the tower down. Still the most likely option, but it's a troublesome prospect for owners of buildings in the rubble-range that survived the temblor in good order.