In 1977, I toured an under construction nuclear power plant being built by my employer. Outside the containment building, there was a row of diesel generators – it was explained they were needed for emergency power – which creates a “What? Why would a power plant need power?”.
The basic way a nuclear plant works (there are variations) is that the uranium gets hot, pipes carrying a liquid (water/steam is most common in the U.S.) is heated and sent to a turbine which then spins and drives a generator, which then sends its power to transformers that regulate the power and put it onto high tension lines to connect to the grid. The water is cooled in a cooling tower and then reinjected back into the containment vessel. Unless a pipe leaks, the liquid used to generate power never is contact with the radiation.
If the power plant “falls off the grid” for more than a few seconds – the circuit is tripped offline, a transmission line fails and some or all of the load goes away – the plant must go into immediate emergency shutdown (or the plant might shut itself down because of an earthquake or a detected problem which is what happened in Japan). The emergency shutdown process begins by lowering the control rods into the core to slow down the nculear reaction process – but the process and the related heat take a long time to completely stop. If the control rods fail to drop, you are in one big heap o trouble very fast.
When the output of the generators has nowhere to go, the turbine generators would quickly spin up and explode – which would be a bad thing. The energy being generated in the containment building needs somewhere to go – it can take days for the uranium to cool. So the plant has a way to internally “use” the massive amount of energy it needs to get rid of – one way is to have a great big vat of oil and “heat” the oil while the plant is winding down. That vat of oil will begin to billow out black smoke which is alarming, but that’s basically what it is supposed to do.
But a nuclear power plant needs -stable- electric power at the right voltages to run the lights, the computers, the pumps, the valves, the safety systems. In addition to the outgoing high voltage power lines, the plant itself will be connected to the grid to use power, the same as any industrial site. But there is a high probability that if the nuclear power plant goes offline, that it will at the same time lose its own external power source.
At that point, the on-site diesel generators detect the failure, and their batteries start the generator, then that power is used to control the plant while it shuts down (or until regular reliable power supplies are back online).
It’s critical that those generators be tested regularly – if the batteries aren’t charged, they won’t be able to start the generator. Diesel fuel deteriorates over time – you need to burn it off and replace it over time, or when you do need the generator to work, you’ll find it is nothing but a gooey sludge. (My last employer did this type of testing every month)
Back around 10 years ago, the East Coast went dark for a couple days and nuclear plants fell offline for this exact reason, and some of them had similar issues. Once the nuclear plant goes into emergency shutdown, you can’t just say “Never mind” and tell it to resume generating power – it has to be systematically brought under control and slowly restarted.
The plant I visited was finished around 1980 – the design life was 50 years. Under the current regulatory environment, it will take close to 20 years to build replacements. The clock is ticking.
Reading up some more…. these reactors in Japan were built by GE. They’re currently on battery power.
If the cooling loop fails (the injection pumps stop, the pipes get clogged, the pipes rupture), then the inside of the containment vessel (the big metal pot inside the containment building) has no way to dissipate the heat – the water inside the vessel starts to increase in pressure – to keep the vessel from exploding, the steam inside the vessel (which might be midly radioactive) has to be vented – that’s what’s going on now. The steam is vented inside the containment building and will condense back to water and collects (and possibly pumped out eventually)…. but if there is enough steam and the process isn’t brought under control, the containment building will build up pressure.
But venting steam means you’re losing water inside the containment vessel – in order to deal with that, the plant has an emegency water supply to keep pumping new water into the containment vessel – the emergency water supply needs a pump to push the water in – it’s pushing against the pressure of the steam inside the containment vessel. The higher the pressure gets, the less likely the pumps will be able to inject the water to cool it down.
In Three Mile Island, what happened was the power generators got clogged (they were cleaning the pipes). The power generating loop shut down, forcing the plant into shutdown. So far, so good. The containment vessel built up pressure, the overpressure valve opened and started venting steam. When the pressure dropped, the operators told the valve to shut (but it didn’t)…. so the containment vessel continued to vent steam. The valves to inject the emergency water into the core did not work, and they didn’t know that. So the water continued to boil off from the core – as the water boiled off, the sensors to measure the water temperature were uncovered, so they stopped reporting a water temperature. The printer telling the operator what was happening was backed up half an hour with bogus useless messages obscuring what was really going on.
The water level in the containment vessel continued to drop, unknown to the operators. Without water, eventually the fuel rods melt. About this time, they called the president of the company I had worked for and asked him “What the hell is going on here?”…. Duke Power (In North Carolina) was run by nuclear engineers, not accountants – they had designed their own plants, not bought them “off the shelf” from Babcox/Wilcox or GE.
So very quickly someone asked “Is the vent at the top of containment vessel open or closed? “Oh, Crap”… Once they realized the core was exposed, then the question was – do we now inject water into the superhot core, or will that then cause the entire containment building to explode? The engineers were sure it was safe, but the politicians were not – and wanted no radiation released outside the containment building. Eventually, the scientists won, and the water was put into the plant – the resulting pressure required releasing a very small amount of radiation, the core cooled and the fuel rods didn’t melt down (although they were seriously damaged)
Clarifications/corrections- I was working as a nuclear transient engineer in 1979 for the nuclear plant vendor who built TMI:
“In Three Mile Island, what happened was the power generators got clogged (they were cleaning the pipes).”
Not exactly. There was a turbine trip and the plant separated from the grid. Turbine trips are pretty normal.
“The containment vessel built up pressure, the overpressure valve opened and started venting steam.”
Following the turbine trip there was an automatic reactor trip (scram), also normal and expected. Pressure increased in the reactor vessel and primary coolant loop as designed, and a relief valve on the pressurizer opened to relieve the temporary increase in pressure. The steam was NOT released into containment. It went to a water filled quench tank that is designed to capture and condense released steam.
“When the pressure dropped, the operators told the valve to shut (but it didn’t)…. so the containment vessel continued to vent steam.”
The pressurizer relief valve stuck open and did not automatically shut as designed when primary pressure began to fall. Indicated water level in the pressurizer began to INCREASE as the steam an the top of the pressurizer was vented to the quench tank. Eventually the quench tank was overcome and a rupture disk on the tank opened as designed, allowing the steam to vent to containment, where the containment spray would automatically actuate and condense the steam.
“The valves to inject the emergency water into the core did not work, and they didn’t know that.”
NOT TRUE. The emergency core cooling systems (ECCS) actuated and worked perfectly as designed. The plant operators TURNED OFF the ECCS for almost four hours, trying to get the pressurizer water level to decrease. As additional safety systems actuated (such as containment spray), they turned those off as well!
“So the water continued to boil off from the core – as the water boiled off, the sensors to measure the water temperature were uncovered, so they stopped reporting a water temperature.”
Partially true. Coolant began boiling in the core. The in-core thermocouples (temperature sensors) correctly reported high temperatures that indicated the core cooling problem. The operators ignored that information, telling themselves that the sensors must be bad, and again went back to trying to get the pressurizer water level to come down.
“The printer telling the operator what was happening was backed up half an hour with bogus useless messages obscuring what was really going on.”
TRUE, but in addition to this the operators had silenced key initial alarms that told them what the real problem was. The accident happened during 3rd shift with the least experienced staff in the control room. When the morning shift reported early, the senior reactor operator walked into the control room, glanced at the panels, quickly deduced the problem and and immediately ordered closure of the Pressurizer PORV block valve, which stopped the steam flow from the pressurizer. But by then the damage was already done. The core had already partially melted.
TMI was very unusual and the lessons learned from that accident revolutionized the western nuclear power industry. Had the plant operators simply stood by and done NOTHING there would have been no “accident” at all. As plant designers we never envisioned that the operators would actively disable the safety systems in the middle of an event. But they did! And that changed our design philosophy forever- now “human factors engineering” is hugely important both in aviation and nuclear power.
At one of the plants, the backup diesel generators were destroyed by the tsunami wave. That’s really really bad news.
When Chicago had the “underground flood”, they brought in portable emergency generators on trucks that are for exactly this type of purpose [who knew such a thing existed?] – but if the area was hit by the tsunami, then the roads are impassible.
Kiss nuclear power goodbye for the next 100 years….
Even if the plants are brought under control, Japan now has a serious power crisis. They were already running oil powered generators before the quake because of problems at the nuclear plants. They probably don’t have the generating capacity to keep the country going on just oil. Good thing Libyian oil is available….. oops….
Japan gets 1/3 of its electric power from nuclear [Source]
The Tokyo Electric Power company is warning that most of the country may soon lose electric power.
Oh, no worries, Art. When I lived in Japan (almost 15 years), chief among Japan’s oil suppliers was Iran. See? No worries, like I said.
Side note: We are having big trouble trying to locate relatives in Japan. Phones are not responding and internet communication is questionable. Wife is highly concerned.
After the quake, I went to the USGS “Did you feel it” web site, and noticed that there were no reports at all from anywhere beyond about 100 km north of Tokyo, which I interpreted to mean there was a huge power, phone and internet outage. I almost pointed that out, but decided against it – not wanting to be creating alarm without a reason (more than I already was)….
I’m reading today that they have revised the quake estimate up to 9.1 – it’s a exponential scale, not a linear one
Fire up those Ham Radios!
And not to belabor the point, but I start from the assumption that many readers know less about history than I do / did….
Japan’s dependence on oil was a big factor in why World War 2 happened. Japan has no internal oil supply. Its leaders were painfully aware of that vunerability. Japan was involved in a protracted war with / invasion of mainland China…. without digressing into why there was a China/Japan war…. FDR decided that the US *must* get involved and stop Liby… errr. Japan… So FDR imposed a blockade on imports of oil to China. The Pacific fleet stationed in Pearl Harbor Hawaii, was responsisble for enforcing that blockade.
Despite being very careful about limiting oil consumption, Japan had a very short time period after which they would be out of oil. The Japanese military had only about enough oil left to do one last attack to break the blockade, so they expended what they had left to try to sink the U.S. Pacific fleet.
Not arguing the justification of what Japan did in China or Hawaii – once you know that background, one thing is very clear. When FDR talked to the country and said that Japan’s attack on Pearl Harbor as “unprovoked”, FDR was lying.
If the US destroys Libya’s air defense capabilities and bombs its air fields, and Libya starts blowing up things around the world, those are not “unprovoked” acts.
FDR lied about a lot of things. :-)
But to be fair, Japan was also going after Indonesia, the Philippines, and possibly even Australia in their quest for oil. The U.S. pretty much told Japan to go ahead and fight with China and Russia all you want, but leave the rest of the Pacific out of it.