On November 4, 2006, Europe suffered one of the largest disturbances in its electrical power grid of the past half century. 15 million households were disconnected from the grid for 20-40 minutes. A trivial error that should not have meant any significant disturbances cascaded and spanned much of continental Europe. This nearly threw the continent into a blackout that would have dwarfed the North American power-out in 2003. While the direct cause of the 2006 disturbance was operator error, a root cause of the problem and a significant factor in delaying getting the grids back online, was wind power and combined heat and power (CHP) plants.
Background
The European power grid has been constructed, basically from scratch, since the end of World War II. The grid in itself is divided between a number of different Transmission System Operators (TSOs), like for instance E.ON (Germany), RTE (France) and TenneT (The Netherlands). Each country's domestic power grid is serviced by one or more TSOs. In order to provide stability and safety of the power supply TSOs are connected to each other. The basic idea was that if an area suffers a major disruption, neighbouring TSOs can help by pushing power from their grids into the affected area.
However over time things have changed. With the increased use of supposedly more environment friendly power plants, such as wind turbines and CHP plants, coupled with requirements to cut down on emissions, TSOs today are required to accept power from neighbours if the latter has a surplus of low-emission power. While this sounds good in theory as it forces TSOs to use "greener" power, instead of for instance coal power, this means that the power grid is used in a way which it was not designed for. Shunting power from one area to another puts high loads on the power transmission lines. In effect the European power grid is often operated close to maximum of what it can handle. To rectify this problem, more power lines should be built, but this is becoming increasingly more difficult and thus we are stuck with the problem.
Due to environmental reasons, the development of the transmission system is more and more affected by stricter constraints and limitations in terms of licensing procedures and construction times. The reality today is that many UCTE TSOs face significant difficulties to build new overhead lines due to long authorization procedures and regulatory regimes.
Final Report
System Disturbance on 4 November 2006
union for the co-ordination of transmission of electricity
The windy weekend
November 4, 2006, was a Saturday. The event took place in the evening, starting at 22:10 CET. While power consumption is low in evenings and even more so during weekends, the load on the grid was still high. The reason for this was that TSOs use these low-periods to disconnect some power lines to perform maintenance on them. This means that remaining lines still operate at near full capacity even if demand is low.
Such conditions also meant that the fraction of intermittent and distributed power production such as wind and CHPs was higher than normal, something that was further compounded by the fact that Germany was experiencing windy conditions that evening. Large amounts of power was therefore being routed into The Netherlands and Poland, straining the power lines.
Problems begin
E.ON Netz was to disconnect a power line. Calculations had predicted that this should have been alright without compromising the safety of the power grid. However due to changes in the conditions - Germany's unforeseen windy weather, and a rescheduling that was communicated late to the other TSOs - the disconnect and rerouting of power overloaded the remaining lines. These power lines tripped (automatically disconnected) in order to protect them from becoming damaged. This caused more overloads in other power lines, causing further trips. This is what is know as a cascade. The cascade had within seconds divided the European power grid into three isolated "islands": west, north east and south east.
The severe disturbance of nov. 4 2006 divided Europe
The west area, having lost the connection to eastern Germany where much power was coming from, suffered a large deficit in power production. In order to protect the power grid and equipment connected to it, TSOs started load shedding, that is to say they started disconnecting customers from the grid in order to lighten their load. For each TSO 3-20% of all customers were disconnected and suffered power outs. In total approximately 15 million European households were affected by this.
In the north east area, the problem was the opposite. With the consumers in the west disconnected, the power generators surged as there was no-one that could accept the power they produced. Windmills are particularly sensitive to this and automatically disconnected from the power grid. Within a minute the north east area stabilized.
The south east area suffered a rather small power deficit in comparison and the impact was therefore smaller.
Trying to get back online
The west area started up power generation reserves to counter the production deficit. Within fifteen minutes the west area had stabilized to nominal production, although they were still disconnected from the north east.
In the north east however the different TSO operators had big problems. This was because the windmills and CHP plants that had tripped, automatically reconnected themselves, again causing a production surplus. There was no way to keep the windmills from doing this. To counter it the TSO operators had to manually prevent overloading the grid by instructing other power plants to cut down on production or shut down completely, and engaging power buffers such as pump storage units. This was while they at the same time tried to diagnose what had happened and why they had a disturbance in the first place. There was much confusion and which meant that the north east area was delayed in getting things back to normal.
In order to help counter the production surplus from the uncontrollable windmills, the control block leader for CENTREL (TSOs in Poland, the Czeck Republic, Hungary and Slovakia) somewhat unconventionally agreed to accept much of the surplus. However this meant that huge amounts of power was suddenly being routed eastwards, overloading some power lines up to as much as 140% of normal capacity, severely risking splitting the power grid again. Luckily this did not happen.
As the power levels were restored they tried to get the different areas connected again. Starting at about 25 minutes after the problems began, they tried to resynchronize the different networks. However for another quarter hour, the networks would not connect to each-other because they were fluctuating too much or not lining up to each-other. Either they wouldn't even attempt a connection or they would trip out again after a few seconds. At 22:47 did the networks begin to connect properly, and not until over an hour later, at 23:57, were things back to normal.
Conclusion
Opponents of nuclear power and proponents of "green" power such as wind and CHPs often argue that nuclear power leads to poor safety and reliability of the power production. The argument is that "putting all eggs in one basket" puts us at a risk of a serious power shortage. They argue that distributed power such as wind is better because it spreads the risk.
However the events of November 4, 2006, point to the exact opposite. While the grid was always made to handle large single-point outages, such as a nuclear power plant going offline, with fairly local load balancing, it was not made to being operated the way we are forced to do with intermittent power. With wind power being essentially uncontrollable and fluctuating there is a need to shunt power long distances through power grids that were never built to handle it. This in turn puts strain on the grids, lessening the margins and risking cascading chaotic failure of an entire continent.
With increased use of wind and other intermittent power sources, this risk can only be increased unless we basically tear out our entire transmission grid and build a new one. This is an investment that will not come cheap. Those arguing for large scale power production using intermittent sources, such as wind, must seek an answer to the following question:
Is it prudent, economical, or even feasible to replace the entire power grid of a continent, just to accommodate a notoriously troublesome source of power?
/Michael Karnerfors, for Nuclear Power? Yes Please
Source: UCTE - Final Report, System Disturbance on 4 November 2006
.png)
The Forsmark incident was not Chernobyl
Wednesday, January 28th, 2009This is the second blog response to a blog entry made by The King of the country Lagom. The previous entry dealt with his claims that opinions are sacred and how one must not speak up against them. This entry will deal with the purely factual errors of his claims about nuclear power.
The King of Lagom claims that an incident that took place in 2006 at the Forsmark nuclear power plant could have escalated into a Chernobyl-type accident. Well... first he says that, and then he says it could have become something entirely different. If this sounds confusing it is because the King of Lagom probably doesn't quite know what he's talking about but rather builds this statement on misconceptions about what actually happened at Chernobyl and Forsmark respectively. So let's examine the incidents and compare.
The 1986 Chernobyl accident
April 26, 1986. The night shift at reactor 4 at the V.I Lenin Nuclear power plant, 20 km north west of the town of Chernobyl, Ukraine, has been ordered to do a test. Due to operator error, they accidentally poison the RBMK-type reactor which makes it almost grind to a halt. They don't know why the reactor is giving so little power though because they were mostly coal plant workers, inexperienced with nuclear power, and oblivious to things such as nuclear poisoning. The shift boss, determined to finish the test, gives orders to proceed, telling the operators to perform actions that go against several operating rules of the reactor. This puts the reactor in an unstable state.
When the test is finished and they shut down the reactor, a fatal flaw in the reactor's control system causes the reactivity to spiral out of control, making it output between ten to onehundred times normal thermal effect. The water in the reactor flash boils and the enormous steam pressure blows the building apart. A few seconds later a chemical explosion, when water that has been split into hydrogen and oxygen burns, rocks the complex again. The reactor is on fire for ten days, resulting in a large plume of radioactive fallout.
There are several factors that allowed this accident to happen.
First it was operated by poorly educated personnel, in a political system where safety came second. In the Soviet Union, you did not rise to attractive jobs like this one by being good at your craft but by kissing up to the communist party. Also you did not stay at jobs like this by speaking up against safety issues, because such things made the party look bad. For instance this particular test was supposed to have been run years ago when the plant was commissioned. But since it failed back then, it had to be done again, this time in secret from the Soviet nuclear regulatory authorities.
This shouldn't have been a problem. But the second reason the accident could take place was the deliberate violations of the operating rules of the reactor. The test was to have taken place when the reactor was outputting at least 700 MW; they started when it was at 200 MW. They were not allowed to withdraw more than a certain number of control rods; they withdrew almost all of them. They were not allowed to increase water flow in the reactor past a certain amount when operating at low power; they did. They were not allowed to disengage the safety systems that would have shut down the reactor when they did any of the aforementioned; but they did indeed disable them.
All of this made reactor come into an unstable state that let its most critical design flaw come into play: the positive void coefficient. The void coefficient is a quality in a nuclear reactor that tells us what happens when it gets too hot. When coolant boils in a reactor that has a positive void coefficient, the nuclear reaction increases. This makes the reactor hotter, which makes more water boil. This speeds up the reaction more, making it even hotter... and so forth. And not only was the void coefficient in the RBMK-reactors of the Chernobyl plant positive, it was also dangerously high.
Finally, because the reactor had no real core vessel, nor any concrete containment, the force of the explosion wrecked the building completely. A fire started in the hundreds of tons of graphite that was in the reactor. Also the building itself that was supposed to have been made from fireproof material, was not, and the debris caught fire as well.
This is what is known as a criticality accident, when the nuclear reaction goes out of control. In this case it produced so much heat that the entire reactor blew up from all the thermal energy. This accident was not a nuclear meltdown.
The 2006 Forsmark incident
July 27, 2006. At the switch-yard for Forsmark-1, an electrical arc causes a short circuit which leads to the unit being disconnected from the power grid. This is serious as the plant relies on power to keep all pumps going.
If a nuclear reactor does not have working pumps, eventually the cooling water in the reactor will boil away. If that starts to happen you must engage the emergency core cooling, reserves of water kept for this very purpose. If this too fails and the reactor boils dry, the heat can be such that the reactor core becomes damaged, popularly called a meltdown. This can happen even when the nuclear reaction has been stopped because decay heat continues to be produced a few hours after a reactor is shut down as very short-lived nuclear waste falls apart. This is what happened at Three Mile Island in 1979.
So when a nuclear plant becomes disconnected from the power grid, the reactor is shut down and on-site diesel generators start to provide power for the pumps to deal with the decay heat, and this was what happened at Forsmark 1. However in this case, two out of the four diesel generators did not start, disabling two safety trains out off four. But the two remaining diesel generators were more than enough to drive the pumps. Hence the reactor was cooled and emergency core cooling was not necessary. The reactor shutdown proceded normally.
Similarities?
No, there are no similarities between these two incidents. The Chernobyl disaster was the case of a criticality accident that caused an extremely violent explosion that completely wrecked the reactor core; the building it operated in; burned for days. The Forsmark incident was the case of slight degradation of safety features while the reactor and its cooling operated normally. The cooling system was operational the whole time; the emergency cooling did not need to be engaged; the reactor core was not damaged; the reactor tank was in no way threatened; the over one meter thick reactor contaiment remained perfectly safe. And fire? Naw... there is no graphite in Forsmark-1. Water handles that job instead.
So when the King of Lagom says that the Forsmark incident could have become another Chernobyl, he is wrong. There is no way that Forsmark-1 or any of the other Swedish nuclear reactor could undergo the process that led to the explosion in Ukraine in 1986. And this is not just because we employ people that know what they are doing; care about safety first; follow procedure; don't do things behind the back of the nuclear regulatory authorities. No, the most important reason why a Chernobyl-type criticality accident cannot happen in Sweden is the reactors themselves. Because unlike the RBMK-reactors of the Soviet Union, our boiler- and pressurized water reactors do not have a positive void coefficient. We did it the opposite way, so that when water starts boiling in the reactor, the nuclear reaction slows down because of inescapable laws of physics. It's nature's own choke collar on nuclear reactions.
Conclusion
The RBMK-type of reactor was employed only in the Soviet Union. The international community is working hard to get the twelve RBMK's that are still in operation closed. Even though I'm a nuclear friend I'm not an idiot, and as such I am very glad that one of the remaining RBMK's. Ignalina-2, will be shut down in 2009, meaning that Lithuania no longer operates them. Now we just need to get Russia to shut down theirs and we'll finally be rid of this blight.
When discussing nuclear safety, anyone that uses Chernobyl as an example of what could go wrong in nuclear reactors is ignoring reality. The BWR/PWR reactors of the world hold about as much in common with the RBMK-design of the Soviet Union as does slavery to common work; as does forced child soldiers to commissioned adults. There just is no comparing them as they operate differently down to subatomic level.
The Forsmark incident was not, and could not have become, another Chernobyl. This is not an opinion, it is physical reality.
/Michael
ADDENDUM: As I posted a link to this entry in his blog, and called him on his Ad hominem attacks, he first approved the entry, then he cencored it and claimed that I was violating his right to have "free opinions", i.e. he doesn't want anyone telling him he's wrong.
Tags: BWR, Chernobyl, containment, criticality accident, Forsmark, kärnkraft, meltdown, nuclear, nuclear power, nuclear safety, olycka, PWR, RBMK, Reactor, safety
Posted in Bad Science, Commentary, English, Swedish | 2 Comments »