After the IAEA DG report on Iran noted that the Iranians were unloading the fuel from Bushehr, Bill Broad and David Sanger provided several column inches of speculation about whether STUXNET had struck again.
Now, Bill Broad reports that ROSATOM has offered a more “prosaic” explanation — a failed cooling pump that is about my age. (And, boy, do I feel like anything my age could fail at any moment!)
The Rosatom statement is in Russian:
28.02.2011 17:21 | Департамент коммуникаций Госкорпорации “Росатом”В одном из четырех насосных агрегатов расхолаживания АЭС “Бушер” (Иран) были обнаружены повреждения внутренних элементов. В этой связи возникло предположение, что металлические частицы (преимущественно стружка размером менее 3 мм) могли вместе с водой проникнуть в корпус реактора и, пройдя через внутрикорпусные устройства, попасть на тепловыделяющие сборки (ТВС). Планируется, что, в случае обнаружения металлических частиц на ТВС, все сборки будут промыты, корпус реактора очищен, после чего топливо будет вновь загружено в реактор энергоблока.
Причиной выхода из строя насосного агрегата стали особенности конструктивного исполнения, в частности недостаточная надежность узла крепления внутренних устройств насоса. В результате узел пришел в негодность в условиях повышенной вибрации при пульсации давления, что характерно для центробежных насосов на низких подачах. Аналогичные несоответствия на других трех насосных агрегатах в результате ревизии устранены. Данные насосные агрегаты являются частью оборудования, поставленного на площадку АЭС “Бушер” в семидесятые годы прошлого века, которое, по условиям контракта, российская сторона была обязана интегрировать в проект.
С учетом того, что ядерное топливо еще не было активировано, работы по его выгрузке, осмотру, возможной промывке и загрузке с технической точки зрения являются штатной задачей, не требующей привлечения дополнительного оборудования и специалистов.
As Broad reports, this is Russian for “damage to one of the reactor’s four main cooling pumps … necessitated removal of the fuel core and an inspection of the reactor and its fuel assemblies …”
As explanations go, this is sort of boring — a 1970s-era German (or should I say West German) cooling pump didn’t age so well. Whereas the original story conjured terrifying images of men with beards and turbans on the verge of powering up CHERNOBYL: THE SEQUEL, today’s explanation only reminds us that parts get old and German engineers are, after all, human.
The Rosatom explanation seems plausible enough, not least because the day before Iran notified the IAEA it would unload fuel assemblies from the core, Sergey Kirienko “made a one-day working visit to Iran” to discuss “topical issues of preparation … including operation of the equipment supplied more than 30 years ago and integrated in the design.”
Rosatom announced that visit in English, as well as Russian, the day after Iran informed the IAEA it would unload the fuel.
In retrospect, that sounds like a cryptic reference to the cooling pumps, which the Russians had complained about before. In 2006, ITAR-TASS reported that “the cooling machines were installed at the nuclear power plant after equipment set up by the German company Siemens at Bushehr 20 years ago was integrated into the Russian technology.” Anton Khlopkov and Anna Lutkova, have written an interesting article about how difficult it is to transform the beginnings of a German nuclear reactor into a completed Russian one.
There is apparently another sordid story about Russian and Iranian disputes over delays caused by a third party (reportedly South Korean) that delivered the cooling equipment for the safety systems. (The Russians were miffed that the Iranians shopped elsewhere; the Iranians accused the Russians of using delays to slow down construction over a payment dispute). But that is another story entirely, and it is late.
***
By the way, how frickin’ awesome is Google Translate? If you type “АЭС ‘Бушер’ (Иран),” it generates “Bushehr nuclear power plant (Iran).” That’s impressive — the algorithm not only knows that the Russian acronym атомная электростанция (АЭС) — literally “atomic electro station” — is rendered “nuclear power plant (NPP)” in English, but also understands that the Russian style of “NPP Bushehr” should be translated into “Bushehr nuclear power plant.” Try it!
BBC Monitoring Former Soviet Union – Political
Supplied by BBC Worldwide Monitoring
December 29, 2006 Friday
Russian nuclear engineers complete further stage at Iran’s Bushehr plant
LENGTH: 290 words
Text of report by Russian news agency ITAR-TASS
Moscow, 29 December: Russian experts have completed the next stage of precommissioning work on the first power unit at Iran’s Bushehr nuclear plant, the Atomstroyeksport [nuclear construction export] contractor company press service told ITAR-TASS today. The company added that “a 5.6MW cooling machine has been tested in idle mode”.
The agency source said that “this could be done after precommissioning work was completed on the modular pumping station and after the pumping system for cooling water for the power unit users was switched on”. “This pump was the first nuclear power plant mechanism operating on sea water,” Atomstroyeksport added.
The Russian company pointed out that “the cooling machines were installed at the nuclear power plant after equipment set up by the German company Siemens
Enhanced Coverage Linking
German company Siemens -Search using:
* Company Profile
* News, Most Recent 60 Days
* Company Dossier
at Bushehr 20 years ago was integrated into the Russian technology”. “New cooling equipment for the nuclear power plant’s reserve diesel engines is to be supplied from third countries in line with Iranian companies’ contracts,” Atomstroyeksport said.
Atomstroyeksport also said that “more than 95 per cent of the construction of the first nuclear power unit at Bushehr has been completed”. The physical start-up of the power unit is scheduled for September 2007 in line with the timetable for precommissioning work agreed with the Iranian side.
Earlier Rosatom head Sergey Kiriyenko pointed out that “the start-up of the first power unit will take place as scheduled, if at the final stage of the work the Iranian side provides sufficient finance and if the remaining equipment is delivered to the site from third countries”.
Source: ITAR-TASS news agency, Moscow, in Russian 1102 gmt 29 Dec 06
OT,
I have been occasionaly impressed by google translates apparant ability to handle idiomatic expressions, as well as stuff like that acronym. Someone told me that apparantly partly it works by finding rosetta stone style documents that have been translated rather than trying to follow a rules based appraoch, so it can be quite good at handling expressions and acronyms. It doesn’t know the rules when it does this, it just recognises. Or so I’m told.
Hmm. The ITAR-TASS piece does make it seem that this was the result of a design deficiency, but “Siemens” and “узел пришел в негодность в условиях повышенной вибрации при пульсации давления, что характерно для центробежных насосов на низких подачах”(*) make me think that Stuxnet shouldn’t be taken completely out of consideration.
(*)the unit failed in conditions of increased vibration in the presence of pressure pulsations that are characteristic for centrifugal pumps at low output.
BTW, has anybody examined the possibility of an ancestral connection between Stuxnet and the possibly apocryphal story of sabotaged pumps in Soviet pipelines back in the 1980s?
Perhaps it’s just a coincidence, but both of Brazil’s nuclear reactors at Angra went offline recently. Apparently the cause was equipment failure. It should be mentioned that Angra-2, built by Siemens, is same reactor type as at Bushehr.
http://www.bloomberg.com/news/2011-02-23/brazil-nuclear-plants-angra-1-and-2-disconnected-from-grid-1-.html
Re: Google
Do not underestimate the power of the little mouse-over which opens asking “do you have a better translation for this?”. People, and in particular technically-minded people, often offer “better translations” which Google then retains.
I don’t see a connection between a failed coolant pump and having to inspect the reactor’s fuel rods.
I can think of two reasons for testing a fuel rod.
First, is the explanation given, to inspect it for defects. This is not the preferred method for doing this, however. In the event of a defect, samples of the primary’s water would have an elevated radiation level that with a unique signature that would identify exactly what type of defect occurred. Physical removal of the rods would only be required to repair one once a defect is detected.
But how likely is a defect from a failed pump? The only reason a fuel rod might fail is if the temperature in the core reached to high of a level. This can happen if there is no cooling from all pumps failing (e.g. 3 mile island), but not from only one pump failing. This can also happen if there is a “reactivity spike” (e.g. Chernobyl). In a PWR, a reactivity spike happens when cold water flows into the core, which happens when a pump’s speed is increased without following proper procedures. If, in fact, there was a fuel rod failure, this seems the most likely cause to me. Stuxnet, or a similar virus, could still be responsible for a failure like this. The question, however, is why did it wait until now?
There is a second reason to inspect the fuel rods: to directly measure the U-235/U-238/Pu-XXX/transuranic ratios present inside them to create a history of the reactor’s power level. Iran should have a clear idea of what its power levels have been, so there should be no reason for them to want this information. But the IAEA and other foreign agencies could want this information to verify Iran’s reported records. This information could be used to calculate the amount of Pu produced and potentially removed by the plant. Doing this in conjunction with a “failed” coolant pump allows Iran to provide this information in a discrete way.
Caveat: I’ve spent a few years operating a PWR (the same basic idea, but not exact design as Bushehr), but don’t have any direct experience with the design issues.
The last paragraph/sentence in the Russian piece Jeffrey quotes said the reactor hadn’t yet gone critical. (С учетом того, что ядерное топливо еще не было активировано…)
Ha! I knew that too… serves me right for posting late at night.
Jack, if a coolant pump fails catastrophically — say the thrust bearing spalls badly — you can spew parts all over the primary coolant loop. Shards of metal, solid lubricant, packing grease, and who knows what else. It could be difficult or impossible to predict where all those pieces would end up.
I bet the Russian engineers are concerned about the potential for fretting between the fuel rods and the spacers, and having junk in there would make things much worse. (Analogy: pebble in your shoe.) Since the core hasn’t gone critical yet, it’s relatively easy to pop the top and assure yourself that there are no shards in delicate places.
But that’s just speculation on my part.
The worry on foreign objects is likely the case. There are several examples (from Western plants, too) where a repair work during an outage was botched badly enough that large amounts (in NPP terms) of solid “stuff”, typical something like metal shavings or loose screws, ended up in cooling water and in core. If this kind of stuff remains there when the reactor is started and the coolant flow increases, it has big potential to cause fuel leaks.
If the amount of stuff that is in the primary loop, and especially if it has migrated into core and lodged itself into the water channels in fuel assemblies, unloading the fuel for check and cleaning, is the best course.
Hmm, I wonder if perhaps the failed cooling pump was driven through a variable frequency drive ? Perhaps in turn controlled by say a Siemens PLC ?
Or perhaps it was indeed just old. I would also imagine the configuration of any drives for the cooling pumps at Bushehr would be a far more common configuration that the drives targeted at Natanz.
.
A variable speed drive would have to be huge to power this presumably very large pump .
it’s much simpler to run the motor as a constant speed wiring and control the flow/pressure with instrumentation
I think it’s unlikely that Stuxnet was involved in the pump failure. From what I’ve read Stuxnet attacks Siemens S7 PLCs, and these are far faster (and more expensive) than would ever be warranted for reactor cooling pump control.
I’m inclined to agree with Thermopile: a failed pump may release material into the primary circuit despite the best efforts of the designers to prevent this. Often during commissioning there will be an additional strainer downstream of the pump to prevent the escape of debris in the event of a failure, but very small particles can still pass through a strainer and if such particles lodge in a fuel assembly’s fins or channels they may cause a local hotspot.
I’d imagine it would have to be a pretty big failure to warrant the removal of the whole core, but in the event that there WAS a big failure of the pump then the commissioning engineers really have no option but core removal if they are to sign off the next stage of commissioning with a clear conscience.