For Three Mile Island, there is no problem of moderating graphite rods that go down too slowly. And no problem of incompetent officials chosen by the central power, or so-called technicians malformed. So, what they could invent this time to justify the melting of the heart?
Since the problem can not come bars moderation (as in the case of Chernobyl, it is said to want such a problem could not happen with the Western central) can already be almost sure that he comes from cooling water. And this may also indicate a priori that there has been a long series of incidents that it has come to a partial melting of the heart. And actually, when you analyze what happened, it does.
1) The conduct official "accident"
The official version is roughly as follows. Indeed, the cooling circuit has a problem. There was an incredible series of failures that led to that there is enough water in the heart, that without the operators do not notice the problem, which led to a partial melting of the latter. Fortunately, operators have come to understand what was happening, and it manages to avoid a complete melting of the heart.
Here's how things went this March 28, 1979. Everything happened in about 3h20.
A) The first moments of the accident
(Note: there is a small summary at end of subchapter, if you want to zap this passage)
Shortly before 4 am, a problem appeared with one of the 8 demineralisers water secondary circuit. These water tanks in which water flowing in the secondary circuit is demineralized. It seems that this is done using beads made of a resin to which minerals cling. Regularly so, remove the resin beads used, and replace them with new ones. This operation is done very regularly. And until then it had been a problem twice, but never reach such a disaster.
Only this time, after isolating that tank of water flow generally, the flow of air under pressure, however, used several times (and a stream of water or steam) has failed to release the accumulated resin beads. They had indeed collapsed into a compact mass. William Zewe, the supervisor, then called Fred Sheimann, a foreman, to help two operators assigned to this task to solve the problem.
Such a thing was definitely not unusual. It happened regularly. It even happened so often that it was installed on each of the 8 scrubber tanks of pressurized air hoses connected to the ventilation system generally. Short bursts of air under pressure generally sufficient to dislodge the agglomerate of resin beads.
In addition to the main ventilation system, there is a second system of ducts of pressurized air, called air instrument system, which was used to control pneumatic valves in some of the plant. This system was essential for the safety of the plant.
The official, that the operators did not know is that someone during the night, had a plastic tube connected between this system of instruments for air and water pipeline. According to some assumptions, he might have tried to pressurize the water line, or connect the two systems air units. This connection of the two systems was made possible by the fact that the instrument system, air ventilation system and main water pipes of the plant using all of the same pneumatic Chicago. The area was quite dark, and the equipment was not properly labeled. However, the water line had more pressure than the air instrument system. So, according to the official theory, water began pouring into the air ducts. However, this instrument system air pneumatically controlled valves connecting demineralisers the secondary circuit. The closing of the valves had to cause the blockage of water flow in the secondary circuit, and soon after, the automatic shutdown of pumps in the secondary circuit.
Also at 3:57, while Scheimann was on top of a pipe feeding the resin beads peeping through a window, the water finally reaches the pneumatic valve demineralisers. Carried by the air pressure, he had traveled along the instrument system to air, eventually entering the control valves.
However, shortly after the plant has entered into business, 5 years earlier, in 1974, someone was concerned the possible risk of a passage of water in the air instrument system. It was understood that it would lead to close the valves. So a plan was discussed to change the system of control valves. If such a thing happens, the valves are stuck in their current position. Nobody knows why, but the wiring for such amendment had never been connected.
B) the first in the 8th minute of the accident: the problem of spray and the secondary circuit
Note: you must keep in mind that the events described in this subpart have place in only 8 minutes. This part is perhaps the most complicated, but at the same time the least important, since in fact, most problems were solved quickly (in 8 minutes so).
When the water came from valves, or in any case, the pressure was large enough, instantly, they have closed violently. The valves of the other 7 demineralisers that remained open were closed in an instant. A water hammer occurs, driven by the sudden stop of water flow. Faust and Frederick felt the ground shake control room while the violent shock snatched control valves, crack the shell of a water pump broke and the pipes. Scheimann jumped on the side just as the pipe on which he was heaved violently. Within seconds, the whole building was filled with auxiliary steam.
In the control room, the automatic control of the plant were operating as intended. With the system of water circulation off and the control system off valves, water could not reach the steam generators. So, they might dry in seconds and get in a boil. Such a thing should certainly not happen. Because too much heat causes the rupture of the tubes. Suddenly, the water of primary circuit water flowed into the secondary water circuit.
To prevent this, automatic systems were activated. 5 seconds after the pumps stop, the turbine power generator was stopped. In general, it causes the opening of bypass valves, which discharge directly while the steam from the steam generator to the condenser, bypassing the turbine.
But the condenser at Three Mile Island was a slight flaw in its design. The steam from the bypass valve was oriented in such a way that if a jet of steam suddenly happened, it would blow the water from the condenser in the vacuum pump of the condenser. That's exactly what happened. And the vacuum pump suffocating under water stopped. When the condenser has lost vacuum due to vacuum, he could not accept steam longer. Suddenly, the pass system has stopped.
With the condenser off-duty, tons of steam were still evacuated. Also, a set of external nozzles, called atmospheric unloaders, were opened up, sending the steam flow to the outside of the plant, with a roar deafening was heard for miles. Residents around Middletown and Royalton were awakened by the first external sign of problems on the island.
The loss of water supply meant that the reactor's heat had nowhere to go. Also, temperature and pressure began to rise. Detecting this, the control system has shut down the reactor (again 5 seconds after stopping the pumps), and within seconds, all control rods were lowered.
However, the reactor core does not cool instantaneously once the bars control lowered. Residual heat can still be a few megawatts. The steam generator thus continuing to threaten to boil dry because of lack of water. Also, 3 water pumps relief, 2 electric and operated via the steam flow was automatically set in motion to provide the necessary water flow to the steam generators that were being emptied.
But there was another problem. Valves that allow the connection of those pumps back to the secondary circuit were closed. A test was conducted on the emergency circuit 42 hours before, causing the closure of two valves (called EFW-12 A and B). And while the procedure requires an imperative that they be immediately reopened, the operators had forgotten to do so. And no operator knew they were closed. Operators should have appreciated, as there is a light that shows whether the valve is open or closed. But one of the two lights were hidden by a label maintenance and operators simply do not pay attention to the second. Indeed, they did not expect that the valves are closed since they were always open usual. Moreover, one of the first things checked in the checklist operators control room, is that the backup circuit works well. But Faust did not see the indicator light remained visible, and alleged that the valves were open, as they should be and had always been before.
So in fact, the emergency pumps walked. But the standby circuit was not open. So, the water still did not circulate in the secondary circuit.
Because the blocking of the secondary circuit, immediately, heat and pressure increased in the primary circuit. Also, very quickly, at T = 3 seconds, the discharge valve Automatic, on top of the pressurizer opened, releasing steam from the primary circuit in a containment tank. She should close automatically when the pressure had dropped. But despite an order of automatic closure, it was not the case. She is stuck in open position.
Since the reactor was stopped, the heat in the primary system began to decline. The water started to cool and thus also the pressure (the pressure decreases also due to the discharge valve open). Also, the water level in the pressurizer began to decline. It was a phenomenon expected. Faust and Frederick attended the automatic start at T = 2 min, two pumps for supplying water to offset the volume decrease (pumps off). Then, when the level continued to decline, special pumps for the injection system at high pressure have been working to bring more water into the reactor vessel. At that time, the relief valve should have closed but, as we have seen, this was not the case.
In general relief, the water level began to stabilize. Then, general anxiety, he began to climb again. Fearing that the cushioning effect of the pressurizer is soon lost as he was finishing his fill, Faust stopped the injection system with high pressure coolant system. But the level continued to rise. So he stopped the pumps off. It still continued to rise. Frederick looked sweating big drops. He recalled the numbers of water level, while the water was rising in the pressurizer until it almost overflowed through the relief valve.
Faust, Frederick Zewe and tried to understand the nature of the problem. Nothing stuck.
Regarding the secondary circuit, the water level continued to decline in the steam generators. In fact, one of them was completely dry and hot. It is a very dangerous situation, because the generators are not designed to reach such temperatures. If a pipe from the steam generator were to crack or break due to heat, radioactive water from the primary circuit could be mixed with water from the secondary circuit and leave the containment building, which would be catastrophic. There was considerable confusion regarding the original reason for the steam generator (OTSG) was drying up and boil. Indeed, from what I understand the operators, the emergency pumps water supply were in the process to provide all water necessary for the steam generators. But in reality, with the valves closed, the pumps were useless.
The amount of water available in the primary circuit is usually measured by measuring the water level in the pressurizer. More water means that the vapor bubble decreases and the water level rises, and vice versa. However, while the primary system was losing water very quickly, operators have seen an increase in water level. The official explanation given a posteriori is that the flow of steam from the discharge valve to the disrupted water level indicator. One can imagine that the flow of steam upwardly traced the indicator. Therefore, operators have been mistaken in their estimate of the pressure due to water level indicator.
The temperature continued to rise in the primary circuit, despite the fact that the reactor had been shut down. This was the result of the lack of provision of emergency water to steam generators, for removing heat from the primary circuit.
also lowered the pressure in the primary circuit. Insofar as the temperature and the pressure going in the same direction in a closed system, all three were completely lost face this paradox. Of course, had they known that the discharge valve was opened, they would have known they did not do in a closed system.
One thing was certain, if pressure fell too low, or if the temperature rose too, the primary circuit water start to boil. If such things happen, and if the temperature rose enough in the heart of the steam begin to form in the reactor vessel. If enough steam was produced, it would lower the water level inside the reactor below the level of uranium rods, leaving them exposed. The steam does not cool as well as water, and uranium rods would be severely damaged by the accumulation of heat. They would break quickly. After a while, the uncovering of the core implies that the uranium could begin to melt, see ignite, or even possibly rearrange in a more compact, which would entail the production of even more heat. It would obviously be disastrous. So, the heart must never be overdrawn.
Too much water in the primary circuit would also be a problem. If the pressurizer, the steam bubble allowed in the primary circuit was completely filled, any sudden shock or transient could cause the pipes break or damage the primary circuit water pumps. This is something to avoid at all costs, because a broken pipe in the primary circuit is the worst nightmare of Engineers reactor. The operators are regularly repeat never, ever "complete system". So, being on the brink of this type of event carries with it a palpable fear in the control room.
Deceived by an indication of incorrect water level, operators decided to open the floodgates ejection and start pumps to drain water from the primary circuit. Now the water coming out of the primary circuit by either one, but two ways out: the relief valve and the ejection system. Reacting to the loss of pressure injection pumps activated at a low pressure situation, have automatically begun to pour water into the primary circuit. Operators, unaware of the real situation, have adopted (at T = 4 min 38s). While water coming out of the reactor through the discharge valve, they were simply neutralized the only system able to replace the discharged water, and they evacuated more.
Suddenly, water began to boil out of the heart (at T = 5 min 30s)
At that time, Faust has come quickly once again the power of the checklist emergency water from the secondary circuit. This time checking each valve in the system, he finally removed the label and saw the red lights indicating that the water supply valves 12A and 12B were closed, blocking the flow.
He then yelled at Zewe: "The 12 are closed!". AT = 8min 18s, Faust and then put in the open position. The cold water rushed boiling in the pipes of the steam generator. The heat of the primary circuit has finally been cooled by the secondary circuit. The indicator of water level in the pressurizer was then slowly stabilized, and the temperature rise began to slow, but just slow.
Abstract: Many human errors led assumed that all valves are closed demineralisers, so that the movement of water in the secondary circuit is blocked. The automatic safety system operated normally. They have arrested the nuclear reaction in the heart. But there were two failures. The first, which was resolved in 8 minutes, is that the supply valves relief secondary circuit water had been shut down following a recent technical inspection. As a result, the secondary circuit was emptied of its water. This, together with the primary circuit were no longer cooled and the heat was increasing rapidly in both circuits. Fortunately, in the 8th minute, operators have understood the problem and have opened the floodgates in question. The secondary circuit has found when supplied with water, and thus was re-cooled properly. The second failure is that the valve venting steam from the primary circuit in case of pressure had been opened by the automatic safety, but had not closed. Moreover, operators have cut the automatic water injection in the primary circuit (the first high pressure, then the low pressure). The water is consequently of primary circuit without being replaced. And it could not be cooled. It is this problem which has not been resolved for over two hours, which resulted in melting the heart.
C) In the 8th minute to 3:20, an almost total inaction despite the possession of the items to solve the problem
So there we were in the 8th minute of the accident. The incident lasted 3:20, it means that everything that is told now lasts 3:12, so much longer than the first part. Operators have had much more time to react. And yet for 1:52 (to T = 2h), they did nothing, and for another 1:20, despite some actions, they still have not understood what was happening.
a) From the 8th to the 15th minute: no detection of the opening of the valve of the primary circuit
Getting back to the beginning of the accident, as we have seen, when the pumps Main the secondary cooling circuit broke down, 3 seconds later to prevent the pressure increases too much in the primary circuit, the pressurizer relief valve of the primary circuit opened automatically. The steam began to flow into the tight containment building.
Especially this valve has been a problem throughout the duration of the accident. She should have closed down again once the pressure in the primary circuit. But despite the automatic order closure, it was not the case.
The explanation given is that the valve was designed to close after have removed a certain amount of pressure. But it was not reliable. In fact, the electromagnetic valve for discharge, made by Dresser Industries was known to have problems of failure to close. It was estimated that the average number of maneuvers to open / close before there was a failure of only 40 laborers. We thought it was sufficient, because the valve was supposed to open only rarely (the same valve on Unit 1 at Three Mile Island had never been open, except during tests). But the flaw in the design of the condenser unit 2 training she opened at each stop of the turbine generator.
operators have been very close on this circuit discharge steam manually. But the problem is that the lights showed the control valve closed. Indeed, the light was poorly designed, and unresponsive to the closure order, not the actual closing. It was enough to give the closure order for the light indicates a closed position, even if it was true. Therefore, operators, believing the problem resolved, have not responded, and the valve still open, the pressure continued to decline in the primary circuit.
operators should have known there was a problem with the valve, because there is a good indicator of water pressure in the primary circuit (called the pressurizer). So as the pressure decreases as the valve was still open, the problem with the valve should have been quickly detected. But again, no luck, we are told (on Wikipedia):
"The drop in pressure in the primary circuit led the automatic startup of the circuit safety injection (t = 2min 01s), to bring water in the primary circuit. But at the same time as the pressure decreased, the "empty (Of water vapor actually) were formed in the primary circuit. These voids générèrent complex movements of water which, paradoxically, filled the pressurizer water (although the top of the circuit). The operator, having information that the pressurizer was full, erroneously concludes that while the primary circuit and was also stopped manually circuit safety injection (t = 4min 38s). "
So the primary circuit was emptied, but we could not know that because of complex movements of water continued to fill the pressurizer. Really decidedly unlucky.
And operators were unaware of the problem of the relief valve for 2 hours, until the arrival of the morning shift. And they were unaware of the problem of draining the primary circuit during 3h20.
Suddenly, the water injection system of an automatic safety was off, and the amount of water decreases more in the primary circuit, water has begun to boil over at the outlet of the heart (t = 5min 30s). A note that the safety injection system was turned off, not only stopped occasionally.
b) In the 15th minute to second time of the accident
The tank in which water flowed from the primary circuit through the valve quickly regained full. An alarm was triggered. A temperature above normal was also detected in the pipe connected to the wastegate, and a higher temperature and pressure in the containment building. This indicated clearly that there was a problem. But, initially, these instructions have been ignored by the operators (note: in fact, they were ignored for 3 hours).
So, since the reservoir was full and that he continued to fill, the safety disc ruptured tank (at T = 15 minutes). Quickly, the containment building was flooded with radioactive water, part of it flowed into the drains of the ground and was sucked into the auxiliary building tanks. Frederick thought to check the tank level (on an instrument panel located behind the main, not visible from where he sat), but only after breaking the record of safety. And at that moment, as the tank is emptied, the water had returned to normal levels. Decidedly dogged by bad luck. The pressure in the containment building began to rise.
Following this, for another 1 hour, while full of indicators already allow to fully understand the situation, the operators do not understand what is happening.
Regarding the time the first alarm to radioactivity began to ring, the sources are contradictory. Normally, it's a start when the tank was punctured containment, therefore, towards the 15th minute of the accident. This seems to say the site "Engineering.com", which says that soon after the safety disc had been broken, the alarm to the radioactivity began to ring.
But in "TMI step by step", it says that the alarm radioactivity simply does not ring. Other sources (Everything) say that the operators have realized that the water was radioactive in the 45th minute. And they put it on the fact that the water was pumped from the containment building into the auxiliary building. Specifically, they say that when the tank was punctured at its base in the 15th minute, spilling 250,000 gallons (about 950 tons) of water in the containment, pump has started to evacuate automatically the water to the auxiliary building (in fact to other reservoirs in the auxiliary building). 29 minutes later, the operators realized that the cooling water was being transferred to the auxiliary building, was radioactive. They immediately cut off the pump.
On the English version of Wikipedia, it says that only T = 2:45, the alarm goes off to radioactivity.
Finally, the French version of Wikipedia, is at T = 3:12 the radioactivity warning is triggered, because the injected water to time in the containment building there would have been highly radioactive.
Engineering.com contradicts himself apparently himself, it is said that just when Porter arrived in the control room (so around 6 am, or at T = 2h), the radioactive water located in the containment building happens in the annex building through drains water and the radioactivity alarms begin to ring and then move into position maximum. So the alarm would have sounded much later.
For the problem of when the alarm went off radioactivity, things are really not clear.
To return to what was happening also in the 20th minute, the temperature and pressure began to increase sharply in the containment building, due to the heat of the steam. Operators would have overview and then activating the cooling system of the containment building. It is said that they did not understand that these conditions resulted from the fact that the primary circuit lost its coolant indicates a significant deficiency in their training to analyze the symptoms of such an accident.
After a time not specified but must be close to 45 minutes (thus T = 1 hour), the primary circuit pumps begin to tremble because they are pumping more steam than water. Indeed, early, pressure and temperature were such that the water began to boil. Des bulles de vapeur ont commencé à se former, à voyager à travers le circuit et à atteindre les pompes. These huge machines, as wide as a cement truck, and 20 times more powerful, began to vibrate and to be dangerously strained, as they struggled to pump the mixture of steam and water. Flow velocity began to decrease, raising the temperature even further. Vibration as they could blow the seals of the rotors of the pump, and spewing water from the primary circuit, and making the pump unusable. Knowing he had no choice, Zewe ordered that the pumps are stopped before they destroy themselves and the pipes to which they were connected.
The problem is that the vibration of pumps also indicated that the primary circuit had more water and therefore the valve was left open. So if Zewe saw that the pumps were shaking, it would necessarily do what the reasoning. But no, Zewe does not understand.
So, at T = 1:13, operators decide to stop the pump 1 of the primary circuit. They are still nearly 30 minutes without doing anything, then at T = 1:40, they cut the pump 2 of the primary circuit. It does them no problem, because normally, with the engine stopped, the convection of heat in the primary circuit water should be sufficient to cool the heart. In short, the reactor stopped, normally, there is no need for pumps.
Now, the only movement of water remaining in the circuit is the movement of natural convection. That operators do not realize is that because of steam formation, portions of the primary system are now blocked by steam. Suddenly, the water can not flow by convection alone. A large vapor bubble, began to develop in the upper part of the reactor vessel, and rapidly increased in size. Soon, the upper heart began to emerge from the water and began to overheat.
At one point falling in this period of time, Zewe, in a sudden flash of lucidity, began to suspect that the relief valve might well have remained stuck open. He asked a technician to read the temperature at the outlet of the valve. High temperature would indicated that the steam was going through the valve, but the technician had mistakenly read the temperature of another outlet valve, which was low and normal. As a result, the valve remained open.
c) The arrival of the morning shift at T = 2h
At 6 o'clock in the morning is to say, to T = 2h, the morning shift arrives. The engineer in the morning a man named Ivan Porter, remarks at T = 2:20 minutes the temperature in the containment tank is very high (thus, there was also a temperature sensor in the tank containment, in addition to detector water level). He also sees that the primary circuit pressure is low (thus, the pressure indicator showed very good pressure for a while if the engineer could see that the pressure was low), and the pressure of the containment building is high . He made the connection between these different information. He then suggested closing a valve blockage in the flow stream, immediately behind the discharge valve was blocked. Once this is done, the pressure starts to rise again in the primary circuit.
is at T = 2:22, 40 minutes after the last action, then there is no more than a meter of water in the heart, the engineer in the morning decided to close a valve downstream of the valve of the pressurizer. It stops the draining of the primary circuit.
In fact, Porter and others do not seem to understand the situation, because then they would inject water into the primary circuit, they do nothing like that. Which would however be logical, since if Porter has decided to close the valve in question, he understood that the primary circuit was emptied. And given the heat and various other problems, it should be obvious to them that the primary circuit is almost empty. So, this is only an hour after they decide to activate the emergency water injection.
The English version of Wikipedia, at T = 2:45, that is to say at 6:45, the alarm goes off to radioactivity. Information found on "TMI step by step". But they stipulate that operators receive the first indications that the radiation level is increasing.
We do not know why, but operators decide to resume at T = 2:54 (30 minutes after closing the valve), a primary circuit pump. Suddenly, he saw no more than a meter of water in the heart, and that it was damaged due to lack of water, it rocks heavily contaminated water (And given what has already been mentioned, it does not stir much. It should run virtually empty).
AT = 3h, due to the high temperatures observed in the heart, operators are beginning to wonder if it is emerged or if it is still submerged, and therefore begins to say that the measured temperatures may be false. In addition, Gary Miller, the director, who had already had a conference call at 6 am with the leaders present in the plant comes into it.
20 minutes after (T = 3:12), operators decide (again, we do not know why) to stop the pump. Immediately after, they decided to reopen for 5 min isolation valve that closed the valve of the pressurizer. The water then flows from the reactor back into the containment building. The French version of Wikipedia, this is the moment that the radioactivity warning is triggered, because the injected water at this point in the containment building there would have been highly radioactive.
Operators then leave the building in haste and close tightly. The alarms go off around the plant. The site is declared Emergency and evacuation of areas near the plant begins.
Only this time, hearing the alert to radioactivity, as operators begin to understand the situation. They say the radioactivity warning means that the heart was severely damaged and he must run out of water. At 7:20 am, either at T = 3:20, they put into operation the safety injection. That is to say, they take water in the primary circuit and thus drown the heart again under water. According to the English Wikipedia is in fact at 11 am (ie at T = 7h). By
that, they took the risk of creating a steam explosion or cause a vessel rupture due to thermal shock. But none of this has happened, and at T = 3:45, the tank is again under water.
In quite incredible, it was not yet clear to operators that they were facing a LOCA (loss of coolant accident), that is to say, loss of water in the primary circuit .
They then tried to determine the temperature of the heart. There were a number of instruments for measuring temperature in the core, computer controlled. But the computer had been calibrated for measure temperatures below 700 degrees. Above, the computer did that printing of question marks. The software designers never imagined that a higher temperature than it would be reached.
Porter then used a multimeter to read the thermocouples directly. The measures he had corresponded to a temperature of about 10.000 degrees (note: a priori, considering the source, these are Fahrenheit. So, approximately 5,500 degrees Celsius) on a number of thermocouples of the heart. The technician simply could not believe his eyes. Porter himself was about to reject measures by considering thermocouples that were defective. But he then noted that temperatures near the center were higher than those on the sides. That's when they were sure that the heart was severely damaged and there was therefore a loss of fluid in the primary circuit.
They then alternated for several hours, water injection at high pressure and opening the valve of the primary circuit to gradually from the vapor bubble hole in it. The situation has stabilized and the coolant pumps were able to restore service to T = 3:49 p.m..
Since the problem can not come bars moderation (as in the case of Chernobyl, it is said to want such a problem could not happen with the Western central) can already be almost sure that he comes from cooling water. And this may also indicate a priori that there has been a long series of incidents that it has come to a partial melting of the heart. And actually, when you analyze what happened, it does.
1) The conduct official "accident"
The official version is roughly as follows. Indeed, the cooling circuit has a problem. There was an incredible series of failures that led to that there is enough water in the heart, that without the operators do not notice the problem, which led to a partial melting of the latter. Fortunately, operators have come to understand what was happening, and it manages to avoid a complete melting of the heart.
Here's how things went this March 28, 1979. Everything happened in about 3h20.
A) The first moments of the accident
(Note: there is a small summary at end of subchapter, if you want to zap this passage)
Shortly before 4 am, a problem appeared with one of the 8 demineralisers water secondary circuit. These water tanks in which water flowing in the secondary circuit is demineralized. It seems that this is done using beads made of a resin to which minerals cling. Regularly so, remove the resin beads used, and replace them with new ones. This operation is done very regularly. And until then it had been a problem twice, but never reach such a disaster.
Only this time, after isolating that tank of water flow generally, the flow of air under pressure, however, used several times (and a stream of water or steam) has failed to release the accumulated resin beads. They had indeed collapsed into a compact mass. William Zewe, the supervisor, then called Fred Sheimann, a foreman, to help two operators assigned to this task to solve the problem.
Such a thing was definitely not unusual. It happened regularly. It even happened so often that it was installed on each of the 8 scrubber tanks of pressurized air hoses connected to the ventilation system generally. Short bursts of air under pressure generally sufficient to dislodge the agglomerate of resin beads.
In addition to the main ventilation system, there is a second system of ducts of pressurized air, called air instrument system, which was used to control pneumatic valves in some of the plant. This system was essential for the safety of the plant.
The official, that the operators did not know is that someone during the night, had a plastic tube connected between this system of instruments for air and water pipeline. According to some assumptions, he might have tried to pressurize the water line, or connect the two systems air units. This connection of the two systems was made possible by the fact that the instrument system, air ventilation system and main water pipes of the plant using all of the same pneumatic Chicago. The area was quite dark, and the equipment was not properly labeled. However, the water line had more pressure than the air instrument system. So, according to the official theory, water began pouring into the air ducts. However, this instrument system air pneumatically controlled valves connecting demineralisers the secondary circuit. The closing of the valves had to cause the blockage of water flow in the secondary circuit, and soon after, the automatic shutdown of pumps in the secondary circuit.
Also at 3:57, while Scheimann was on top of a pipe feeding the resin beads peeping through a window, the water finally reaches the pneumatic valve demineralisers. Carried by the air pressure, he had traveled along the instrument system to air, eventually entering the control valves.
However, shortly after the plant has entered into business, 5 years earlier, in 1974, someone was concerned the possible risk of a passage of water in the air instrument system. It was understood that it would lead to close the valves. So a plan was discussed to change the system of control valves. If such a thing happens, the valves are stuck in their current position. Nobody knows why, but the wiring for such amendment had never been connected.
B) the first in the 8th minute of the accident: the problem of spray and the secondary circuit
Note: you must keep in mind that the events described in this subpart have place in only 8 minutes. This part is perhaps the most complicated, but at the same time the least important, since in fact, most problems were solved quickly (in 8 minutes so).
When the water came from valves, or in any case, the pressure was large enough, instantly, they have closed violently. The valves of the other 7 demineralisers that remained open were closed in an instant. A water hammer occurs, driven by the sudden stop of water flow. Faust and Frederick felt the ground shake control room while the violent shock snatched control valves, crack the shell of a water pump broke and the pipes. Scheimann jumped on the side just as the pipe on which he was heaved violently. Within seconds, the whole building was filled with auxiliary steam.
In the control room, the automatic control of the plant were operating as intended. With the system of water circulation off and the control system off valves, water could not reach the steam generators. So, they might dry in seconds and get in a boil. Such a thing should certainly not happen. Because too much heat causes the rupture of the tubes. Suddenly, the water of primary circuit water flowed into the secondary water circuit.
To prevent this, automatic systems were activated. 5 seconds after the pumps stop, the turbine power generator was stopped. In general, it causes the opening of bypass valves, which discharge directly while the steam from the steam generator to the condenser, bypassing the turbine.
But the condenser at Three Mile Island was a slight flaw in its design. The steam from the bypass valve was oriented in such a way that if a jet of steam suddenly happened, it would blow the water from the condenser in the vacuum pump of the condenser. That's exactly what happened. And the vacuum pump suffocating under water stopped. When the condenser has lost vacuum due to vacuum, he could not accept steam longer. Suddenly, the pass system has stopped.
With the condenser off-duty, tons of steam were still evacuated. Also, a set of external nozzles, called atmospheric unloaders, were opened up, sending the steam flow to the outside of the plant, with a roar deafening was heard for miles. Residents around Middletown and Royalton were awakened by the first external sign of problems on the island.
The loss of water supply meant that the reactor's heat had nowhere to go. Also, temperature and pressure began to rise. Detecting this, the control system has shut down the reactor (again 5 seconds after stopping the pumps), and within seconds, all control rods were lowered.
However, the reactor core does not cool instantaneously once the bars control lowered. Residual heat can still be a few megawatts. The steam generator thus continuing to threaten to boil dry because of lack of water. Also, 3 water pumps relief, 2 electric and operated via the steam flow was automatically set in motion to provide the necessary water flow to the steam generators that were being emptied.
But there was another problem. Valves that allow the connection of those pumps back to the secondary circuit were closed. A test was conducted on the emergency circuit 42 hours before, causing the closure of two valves (called EFW-12 A and B). And while the procedure requires an imperative that they be immediately reopened, the operators had forgotten to do so. And no operator knew they were closed. Operators should have appreciated, as there is a light that shows whether the valve is open or closed. But one of the two lights were hidden by a label maintenance and operators simply do not pay attention to the second. Indeed, they did not expect that the valves are closed since they were always open usual. Moreover, one of the first things checked in the checklist operators control room, is that the backup circuit works well. But Faust did not see the indicator light remained visible, and alleged that the valves were open, as they should be and had always been before.
So in fact, the emergency pumps walked. But the standby circuit was not open. So, the water still did not circulate in the secondary circuit.
Because the blocking of the secondary circuit, immediately, heat and pressure increased in the primary circuit. Also, very quickly, at T = 3 seconds, the discharge valve Automatic, on top of the pressurizer opened, releasing steam from the primary circuit in a containment tank. She should close automatically when the pressure had dropped. But despite an order of automatic closure, it was not the case. She is stuck in open position.
Since the reactor was stopped, the heat in the primary system began to decline. The water started to cool and thus also the pressure (the pressure decreases also due to the discharge valve open). Also, the water level in the pressurizer began to decline. It was a phenomenon expected. Faust and Frederick attended the automatic start at T = 2 min, two pumps for supplying water to offset the volume decrease (pumps off). Then, when the level continued to decline, special pumps for the injection system at high pressure have been working to bring more water into the reactor vessel. At that time, the relief valve should have closed but, as we have seen, this was not the case.
In general relief, the water level began to stabilize. Then, general anxiety, he began to climb again. Fearing that the cushioning effect of the pressurizer is soon lost as he was finishing his fill, Faust stopped the injection system with high pressure coolant system. But the level continued to rise. So he stopped the pumps off. It still continued to rise. Frederick looked sweating big drops. He recalled the numbers of water level, while the water was rising in the pressurizer until it almost overflowed through the relief valve.
Faust, Frederick Zewe and tried to understand the nature of the problem. Nothing stuck.
Regarding the secondary circuit, the water level continued to decline in the steam generators. In fact, one of them was completely dry and hot. It is a very dangerous situation, because the generators are not designed to reach such temperatures. If a pipe from the steam generator were to crack or break due to heat, radioactive water from the primary circuit could be mixed with water from the secondary circuit and leave the containment building, which would be catastrophic. There was considerable confusion regarding the original reason for the steam generator (OTSG) was drying up and boil. Indeed, from what I understand the operators, the emergency pumps water supply were in the process to provide all water necessary for the steam generators. But in reality, with the valves closed, the pumps were useless.
The amount of water available in the primary circuit is usually measured by measuring the water level in the pressurizer. More water means that the vapor bubble decreases and the water level rises, and vice versa. However, while the primary system was losing water very quickly, operators have seen an increase in water level. The official explanation given a posteriori is that the flow of steam from the discharge valve to the disrupted water level indicator. One can imagine that the flow of steam upwardly traced the indicator. Therefore, operators have been mistaken in their estimate of the pressure due to water level indicator.
The temperature continued to rise in the primary circuit, despite the fact that the reactor had been shut down. This was the result of the lack of provision of emergency water to steam generators, for removing heat from the primary circuit.
also lowered the pressure in the primary circuit. Insofar as the temperature and the pressure going in the same direction in a closed system, all three were completely lost face this paradox. Of course, had they known that the discharge valve was opened, they would have known they did not do in a closed system.
One thing was certain, if pressure fell too low, or if the temperature rose too, the primary circuit water start to boil. If such things happen, and if the temperature rose enough in the heart of the steam begin to form in the reactor vessel. If enough steam was produced, it would lower the water level inside the reactor below the level of uranium rods, leaving them exposed. The steam does not cool as well as water, and uranium rods would be severely damaged by the accumulation of heat. They would break quickly. After a while, the uncovering of the core implies that the uranium could begin to melt, see ignite, or even possibly rearrange in a more compact, which would entail the production of even more heat. It would obviously be disastrous. So, the heart must never be overdrawn.
Too much water in the primary circuit would also be a problem. If the pressurizer, the steam bubble allowed in the primary circuit was completely filled, any sudden shock or transient could cause the pipes break or damage the primary circuit water pumps. This is something to avoid at all costs, because a broken pipe in the primary circuit is the worst nightmare of Engineers reactor. The operators are regularly repeat never, ever "complete system". So, being on the brink of this type of event carries with it a palpable fear in the control room.
Deceived by an indication of incorrect water level, operators decided to open the floodgates ejection and start pumps to drain water from the primary circuit. Now the water coming out of the primary circuit by either one, but two ways out: the relief valve and the ejection system. Reacting to the loss of pressure injection pumps activated at a low pressure situation, have automatically begun to pour water into the primary circuit. Operators, unaware of the real situation, have adopted (at T = 4 min 38s). While water coming out of the reactor through the discharge valve, they were simply neutralized the only system able to replace the discharged water, and they evacuated more.
Suddenly, water began to boil out of the heart (at T = 5 min 30s)
At that time, Faust has come quickly once again the power of the checklist emergency water from the secondary circuit. This time checking each valve in the system, he finally removed the label and saw the red lights indicating that the water supply valves 12A and 12B were closed, blocking the flow.
He then yelled at Zewe: "The 12 are closed!". AT = 8min 18s, Faust and then put in the open position. The cold water rushed boiling in the pipes of the steam generator. The heat of the primary circuit has finally been cooled by the secondary circuit. The indicator of water level in the pressurizer was then slowly stabilized, and the temperature rise began to slow, but just slow.
Abstract: Many human errors led assumed that all valves are closed demineralisers, so that the movement of water in the secondary circuit is blocked. The automatic safety system operated normally. They have arrested the nuclear reaction in the heart. But there were two failures. The first, which was resolved in 8 minutes, is that the supply valves relief secondary circuit water had been shut down following a recent technical inspection. As a result, the secondary circuit was emptied of its water. This, together with the primary circuit were no longer cooled and the heat was increasing rapidly in both circuits. Fortunately, in the 8th minute, operators have understood the problem and have opened the floodgates in question. The secondary circuit has found when supplied with water, and thus was re-cooled properly. The second failure is that the valve venting steam from the primary circuit in case of pressure had been opened by the automatic safety, but had not closed. Moreover, operators have cut the automatic water injection in the primary circuit (the first high pressure, then the low pressure). The water is consequently of primary circuit without being replaced. And it could not be cooled. It is this problem which has not been resolved for over two hours, which resulted in melting the heart.
C) In the 8th minute to 3:20, an almost total inaction despite the possession of the items to solve the problem
So there we were in the 8th minute of the accident. The incident lasted 3:20, it means that everything that is told now lasts 3:12, so much longer than the first part. Operators have had much more time to react. And yet for 1:52 (to T = 2h), they did nothing, and for another 1:20, despite some actions, they still have not understood what was happening.
a) From the 8th to the 15th minute: no detection of the opening of the valve of the primary circuit
Getting back to the beginning of the accident, as we have seen, when the pumps Main the secondary cooling circuit broke down, 3 seconds later to prevent the pressure increases too much in the primary circuit, the pressurizer relief valve of the primary circuit opened automatically. The steam began to flow into the tight containment building.
Especially this valve has been a problem throughout the duration of the accident. She should have closed down again once the pressure in the primary circuit. But despite the automatic order closure, it was not the case.
The explanation given is that the valve was designed to close after have removed a certain amount of pressure. But it was not reliable. In fact, the electromagnetic valve for discharge, made by Dresser Industries was known to have problems of failure to close. It was estimated that the average number of maneuvers to open / close before there was a failure of only 40 laborers. We thought it was sufficient, because the valve was supposed to open only rarely (the same valve on Unit 1 at Three Mile Island had never been open, except during tests). But the flaw in the design of the condenser unit 2 training she opened at each stop of the turbine generator.
operators have been very close on this circuit discharge steam manually. But the problem is that the lights showed the control valve closed. Indeed, the light was poorly designed, and unresponsive to the closure order, not the actual closing. It was enough to give the closure order for the light indicates a closed position, even if it was true. Therefore, operators, believing the problem resolved, have not responded, and the valve still open, the pressure continued to decline in the primary circuit.
operators should have known there was a problem with the valve, because there is a good indicator of water pressure in the primary circuit (called the pressurizer). So as the pressure decreases as the valve was still open, the problem with the valve should have been quickly detected. But again, no luck, we are told (on Wikipedia):
"The drop in pressure in the primary circuit led the automatic startup of the circuit safety injection (t = 2min 01s), to bring water in the primary circuit. But at the same time as the pressure decreased, the "empty (Of water vapor actually) were formed in the primary circuit. These voids générèrent complex movements of water which, paradoxically, filled the pressurizer water (although the top of the circuit). The operator, having information that the pressurizer was full, erroneously concludes that while the primary circuit and was also stopped manually circuit safety injection (t = 4min 38s). "
So the primary circuit was emptied, but we could not know that because of complex movements of water continued to fill the pressurizer. Really decidedly unlucky.
And operators were unaware of the problem of the relief valve for 2 hours, until the arrival of the morning shift. And they were unaware of the problem of draining the primary circuit during 3h20.
Suddenly, the water injection system of an automatic safety was off, and the amount of water decreases more in the primary circuit, water has begun to boil over at the outlet of the heart (t = 5min 30s). A note that the safety injection system was turned off, not only stopped occasionally.
b) In the 15th minute to second time of the accident
The tank in which water flowed from the primary circuit through the valve quickly regained full. An alarm was triggered. A temperature above normal was also detected in the pipe connected to the wastegate, and a higher temperature and pressure in the containment building. This indicated clearly that there was a problem. But, initially, these instructions have been ignored by the operators (note: in fact, they were ignored for 3 hours).
So, since the reservoir was full and that he continued to fill, the safety disc ruptured tank (at T = 15 minutes). Quickly, the containment building was flooded with radioactive water, part of it flowed into the drains of the ground and was sucked into the auxiliary building tanks. Frederick thought to check the tank level (on an instrument panel located behind the main, not visible from where he sat), but only after breaking the record of safety. And at that moment, as the tank is emptied, the water had returned to normal levels. Decidedly dogged by bad luck. The pressure in the containment building began to rise.
Following this, for another 1 hour, while full of indicators already allow to fully understand the situation, the operators do not understand what is happening.
Regarding the time the first alarm to radioactivity began to ring, the sources are contradictory. Normally, it's a start when the tank was punctured containment, therefore, towards the 15th minute of the accident. This seems to say the site "Engineering.com", which says that soon after the safety disc had been broken, the alarm to the radioactivity began to ring.
But in "TMI step by step", it says that the alarm radioactivity simply does not ring. Other sources (Everything) say that the operators have realized that the water was radioactive in the 45th minute. And they put it on the fact that the water was pumped from the containment building into the auxiliary building. Specifically, they say that when the tank was punctured at its base in the 15th minute, spilling 250,000 gallons (about 950 tons) of water in the containment, pump has started to evacuate automatically the water to the auxiliary building (in fact to other reservoirs in the auxiliary building). 29 minutes later, the operators realized that the cooling water was being transferred to the auxiliary building, was radioactive. They immediately cut off the pump.
On the English version of Wikipedia, it says that only T = 2:45, the alarm goes off to radioactivity.
Finally, the French version of Wikipedia, is at T = 3:12 the radioactivity warning is triggered, because the injected water to time in the containment building there would have been highly radioactive.
Engineering.com contradicts himself apparently himself, it is said that just when Porter arrived in the control room (so around 6 am, or at T = 2h), the radioactive water located in the containment building happens in the annex building through drains water and the radioactivity alarms begin to ring and then move into position maximum. So the alarm would have sounded much later.
For the problem of when the alarm went off radioactivity, things are really not clear.
To return to what was happening also in the 20th minute, the temperature and pressure began to increase sharply in the containment building, due to the heat of the steam. Operators would have overview and then activating the cooling system of the containment building. It is said that they did not understand that these conditions resulted from the fact that the primary circuit lost its coolant indicates a significant deficiency in their training to analyze the symptoms of such an accident.
After a time not specified but must be close to 45 minutes (thus T = 1 hour), the primary circuit pumps begin to tremble because they are pumping more steam than water. Indeed, early, pressure and temperature were such that the water began to boil. Des bulles de vapeur ont commencé à se former, à voyager à travers le circuit et à atteindre les pompes. These huge machines, as wide as a cement truck, and 20 times more powerful, began to vibrate and to be dangerously strained, as they struggled to pump the mixture of steam and water. Flow velocity began to decrease, raising the temperature even further. Vibration as they could blow the seals of the rotors of the pump, and spewing water from the primary circuit, and making the pump unusable. Knowing he had no choice, Zewe ordered that the pumps are stopped before they destroy themselves and the pipes to which they were connected.
The problem is that the vibration of pumps also indicated that the primary circuit had more water and therefore the valve was left open. So if Zewe saw that the pumps were shaking, it would necessarily do what the reasoning. But no, Zewe does not understand.
So, at T = 1:13, operators decide to stop the pump 1 of the primary circuit. They are still nearly 30 minutes without doing anything, then at T = 1:40, they cut the pump 2 of the primary circuit. It does them no problem, because normally, with the engine stopped, the convection of heat in the primary circuit water should be sufficient to cool the heart. In short, the reactor stopped, normally, there is no need for pumps.
Now, the only movement of water remaining in the circuit is the movement of natural convection. That operators do not realize is that because of steam formation, portions of the primary system are now blocked by steam. Suddenly, the water can not flow by convection alone. A large vapor bubble, began to develop in the upper part of the reactor vessel, and rapidly increased in size. Soon, the upper heart began to emerge from the water and began to overheat.
At one point falling in this period of time, Zewe, in a sudden flash of lucidity, began to suspect that the relief valve might well have remained stuck open. He asked a technician to read the temperature at the outlet of the valve. High temperature would indicated that the steam was going through the valve, but the technician had mistakenly read the temperature of another outlet valve, which was low and normal. As a result, the valve remained open.
c) The arrival of the morning shift at T = 2h
At 6 o'clock in the morning is to say, to T = 2h, the morning shift arrives. The engineer in the morning a man named Ivan Porter, remarks at T = 2:20 minutes the temperature in the containment tank is very high (thus, there was also a temperature sensor in the tank containment, in addition to detector water level). He also sees that the primary circuit pressure is low (thus, the pressure indicator showed very good pressure for a while if the engineer could see that the pressure was low), and the pressure of the containment building is high . He made the connection between these different information. He then suggested closing a valve blockage in the flow stream, immediately behind the discharge valve was blocked. Once this is done, the pressure starts to rise again in the primary circuit.
is at T = 2:22, 40 minutes after the last action, then there is no more than a meter of water in the heart, the engineer in the morning decided to close a valve downstream of the valve of the pressurizer. It stops the draining of the primary circuit.
In fact, Porter and others do not seem to understand the situation, because then they would inject water into the primary circuit, they do nothing like that. Which would however be logical, since if Porter has decided to close the valve in question, he understood that the primary circuit was emptied. And given the heat and various other problems, it should be obvious to them that the primary circuit is almost empty. So, this is only an hour after they decide to activate the emergency water injection.
The English version of Wikipedia, at T = 2:45, that is to say at 6:45, the alarm goes off to radioactivity. Information found on "TMI step by step". But they stipulate that operators receive the first indications that the radiation level is increasing.
We do not know why, but operators decide to resume at T = 2:54 (30 minutes after closing the valve), a primary circuit pump. Suddenly, he saw no more than a meter of water in the heart, and that it was damaged due to lack of water, it rocks heavily contaminated water (And given what has already been mentioned, it does not stir much. It should run virtually empty).
AT = 3h, due to the high temperatures observed in the heart, operators are beginning to wonder if it is emerged or if it is still submerged, and therefore begins to say that the measured temperatures may be false. In addition, Gary Miller, the director, who had already had a conference call at 6 am with the leaders present in the plant comes into it.
20 minutes after (T = 3:12), operators decide (again, we do not know why) to stop the pump. Immediately after, they decided to reopen for 5 min isolation valve that closed the valve of the pressurizer. The water then flows from the reactor back into the containment building. The French version of Wikipedia, this is the moment that the radioactivity warning is triggered, because the injected water at this point in the containment building there would have been highly radioactive.
Operators then leave the building in haste and close tightly. The alarms go off around the plant. The site is declared Emergency and evacuation of areas near the plant begins.
Only this time, hearing the alert to radioactivity, as operators begin to understand the situation. They say the radioactivity warning means that the heart was severely damaged and he must run out of water. At 7:20 am, either at T = 3:20, they put into operation the safety injection. That is to say, they take water in the primary circuit and thus drown the heart again under water. According to the English Wikipedia is in fact at 11 am (ie at T = 7h). By
that, they took the risk of creating a steam explosion or cause a vessel rupture due to thermal shock. But none of this has happened, and at T = 3:45, the tank is again under water.
In quite incredible, it was not yet clear to operators that they were facing a LOCA (loss of coolant accident), that is to say, loss of water in the primary circuit .
They then tried to determine the temperature of the heart. There were a number of instruments for measuring temperature in the core, computer controlled. But the computer had been calibrated for measure temperatures below 700 degrees. Above, the computer did that printing of question marks. The software designers never imagined that a higher temperature than it would be reached.
Porter then used a multimeter to read the thermocouples directly. The measures he had corresponded to a temperature of about 10.000 degrees (note: a priori, considering the source, these are Fahrenheit. So, approximately 5,500 degrees Celsius) on a number of thermocouples of the heart. The technician simply could not believe his eyes. Porter himself was about to reject measures by considering thermocouples that were defective. But he then noted that temperatures near the center were higher than those on the sides. That's when they were sure that the heart was severely damaged and there was therefore a loss of fluid in the primary circuit.
They then alternated for several hours, water injection at high pressure and opening the valve of the primary circuit to gradually from the vapor bubble hole in it. The situation has stabilized and the coolant pumps were able to restore service to T = 3:49 p.m..
0 comments:
Post a Comment