a sequence of physical events - Three Mile Island

[The low steam pressure in steam generator Band the increase in reactor building pressure werebelieved to be caused by a leak from the steamgenerator.] At 1 hour 27 minutes, steam generatorB was isolated (taken out of service). [With hindsight,it can be seen that the low pressure was simplycaused by steam bubbles and a reduction ofheat transfer in the B loop following stoppage of thepumps. A small change in building pressure wasnoted when the steam generator was isolated. Theoccurrence of this change at this time was probablycoincidental.]March 28,1979-5:30 a.m.At 1 hour 30 minutes, the apparent neutron levelincreased again. An RCS sample showed evenlower boron concentration and increased radioactivity.[The activity was probably due to crud.]The temperature of the RCS coolant in all primarysystem piping had been slowly increasing. Eventually,the primary side of steam generator A got hotenough so that more steam was produced on thesecondary side, and the steam pressure began torise. The increased steam production had two sideeffects: (1) the water level on the secondary sidedropped and the steam generator boiled dry for thesecond time, and (2) the increased heat removalbrought the RCS temperature down again.[The efficiency of the reactor coolant pumps wasstill decreasing, and at 1 hour 37 minutes, the frothymixture became too light to circulate. Separation ofthe froth would have sent the steam to the highparts of the system, while water collected in the lowparts. An analogy is a kitchen blender with the bowlhalf full of water. With the blender at high speed,enough air bubbles are whipped into the water sothat the bowl is full. If the speed drops, the air bubblesare lost and the lower half of the bowl is solidlyfilled with liquid water. This was reflected in thebehavior of the neutron instrumentation. Apparentlythe downcomer, which had been previously filledwith froth, now filled with water. The increasedshielding stopped neutrons from reaching the detectorand the apparent neutron level dropped by afactor of 30.]Operators recognized that steam generator Awas dry, and in an attempt to regain water level,they increased feedwater flow.March 28,1979-5:41 a.m.At 1 hour 41 minutes, both remaining reactorcoolant pumps (RC-P1A and 2A) were stopped becauseof increasing vibration and erratic flow. [Theonly heat transfer through the steam generatorswas now achieved by reflux flow (Figure 11-5). Thiswas inadequate for core cooling. It is now believedthat the core was drying out. The operators werehoping to establish natural circulation in the primarysystem. Natural circulation was blocked by steam,and refluxing would be ineffective because thesecondary temperature was nearly as high as theprimary temperature.][The pressurizer is at a higher level than thereactor. It was assumed that the presence of waterin the pressurizer meant that the core must becovered. Actually, because the PORV was open,pressure in the upper part of the pressurizer wasreduced. The strong boiling that was occurring inthe core, however, caused more steam to go intothe upper part of the reactor vessel, and the pressurethere was increased. The difference of pressureforced the water level higher in the pressurizerthan in the reactor vessel.][Previous reports have alluded to a "loop seal,"thus giving the false impression that the piping configurationalone somehow created this difference ofl evel. Even with the loop configuration, to maintain ahigher level in the pressurizer when the water in thepressurizer is saturated, a higher pressure is requiredin the reactor than in the pressurizer. If thepressures are equalized with the hot leg voided, thesaturated pressurizer water level would drop to thelevel of the connection of the pressurizer surge lineinto the hot leg. Subcooled water could be maintainedat a higher level. During most of the accident,the water in the pressurizer was slightly subcooledor saturated. During the time that the surgeline was uncovered, the water in the pressurizerwas subcooled. It was the combination of loop sealand temperature that kept the level high, rather thanthe loop seal alone.]March 28,1979-5:42 a.m.At 1 hour 42 minutes, the decreasing level in thereactor vessel again reduced the shielding of theneutron instrumentation, and the apparent neutroncount increased by about a factor of 100. Emergencyboration was commenced to avert a restart.[Actually, a restart was impossible because of thepartial emptying of the core, but no one recognizedthis. A further discussion of this topic is given inSection II.C.2.b.]The hot-leg temperature now became decidedlyhigher than the cold-leg temperature. Superheatedsteam was present in the hot leg. [The superheatingof the hot leg showed that a fair amount of the323