'''プラズマアーク切断機'''の概念図を'''図２'''に示す。この装置は、'''炉心下部構造物'''（LCSA: Lower Core Support Assembly）の切断/解体に用いられた。プラズマアーク切断法は、様々な切断方法（ウォータージェット、シャーリング、アークソー、酸素燃焼、超音波破断、レーザー切断など）の中から選定され、デブリ取り出し初期に、炉心上部格子や端栓の解体に利用されていた。LCSAは、約１年かけて約50個のパーツに解体された。解体作業自体は短時間であったが、システムの不具合や設計変更に時間を要した。
# 回収したサンプルをFuel収納缶に装荷できること（RPV内でのデブリ取り扱いシステムを使用）
# 炉心中央から、2.4m径の範囲でボーリング可能なこと（'''<span style="color:blue">SWP''': Shielded Working Platformの上に設置したドリルマシンからのドリルビットの到達範囲）
One account indicated that the reactor vessel defueling water cleanup system could have been sized to restore water clarity more rapidly. Operation of underwater equipment with limited visibility greatly reduced productivity while increasing the risk of equipment damage and lost time. The design basis for water clarity should have been 0.1 nephelometric turbidity unit (ntu) (about 0.1 parts per million suspended solids) instead of 1.0 ntu for manual tool operation at depths of 6 to 10 meters. (260) Working inside the reactor vessel often resulted in the resuspension of very fine particulates, the mass of which precluded settling. This resulted in a smoky effect that reduced visibility from 12 meters of water shielding to less than 1 meter, and often much poorer.
Analysts had not predicted the extent to which the molten material had formed into a large monolith within the reactor vessel or understood the material’s resistance to fracture. (In fact, fuel melt was first reported in 1985 after careful examination of the first “grab” samples from the damaged core, 6 years following the accident.) The molten material that formed in the TMI-2 core was a complex mixture of the major core constituents (uranium, zirconium, and iron) with lesser amounts of control rod material (silver, indium, and cadmium) and other alloy constituents (chromium and nickel). The molten material also contained a significant quantity of oxygen, making the solidified melt a highly refractory ceramic. A portion of the molten material solidified into a heterogeneous, funnel-shaped disk over a meter thick. The mass was located in the center of the reactor vessel below the plenum. Elsewhere in the reactor vessel, molten material flowed down between the baffle plates and the core barrel and solidified in the reactor vessel’s lower head region. Both solidified masses weighed many tons. Heavy-duty defueling tools (impact chisels and wedges) were unable to break apart the monolith. A specialized drilling apparatus (core bore) was finally used to pulverize the material.
. Estimating the number of canisters to be used in defueling evolved into a high art form and one with significant potential cost. The original estimate was 243. The first half of the INEL pool could store 288 canisters. The project’s total estimate grew as defueling progressed because the weight loaded into each canister was usually less than anticipated. As important, the relative numbers of fuel, filter, and knockout canisters changed in response to the techniques used and unexpected conditions in the reactor vessel. The final estimate was between 349 and 360 canisters. (266) At the conclusion of the defueling effort, a total of 342 canisters of core debris, in 22 rail shipments, were transported to INEL. The total number of canisters included 286 fuel canisters, which contained partially intact fuel assemblies and large debris picked up from the reactor vessel; 12 knockout canisters, which contained core debris vacuumed from the reactor vessel and reactor coolant system; and 62 filter canisters, which contained fine debris that had passed through the knockout canisters
Extraction of Standing Fuel Assemblies. The outer ring of the core contained standing fuel assemblies of various lengths. Once the first assembly was removed by the lasso tool, the others came out easily. The tools that were developed for removing standing assemblies included hydraulic side grippers, spears, and long L-shaped assembly lifters. (268) • Cutting Incore Guide Tubes. The core boring machine proved very adept at cutting out in-core guide tubes and support posts, which was a weakness of the automated cutting equipment system. However, cutting was only effective when enough space existed for the chips to fall away. (269) • Baffle Plate Handling. (270, 271) Vertical baffle plates formed the peripheral boundary of the core. Roughly 4,000 kilograms of fuel debris relocated behind the baffle through one large hole in the baffle plates caused by the flow of molten core material from the core region. Several options were considered for removing thehighly irradiated baffle plates (with an estimated peak contact gamma radiation field of approximately 3,000 roentgen per hour in air) to gain access to this debris. o Options. The removal options included the following: (●) Cut the plates into small pieces and load them into fuel canisters or other specially designed containers. (●) Remove the small pieces of plates from the vessel and store them either in a remote location (e.g., the containment building basement) or in a modified core flood tank (sections of the lower core support assembly were stored in these tanks). (●) Cut the plates into large sections, rotate each section, and remove debris one plate at a time. The latter option, which was selected, required no plant modifications; resulted in lower radiation exposures; required minimum tool development; and did not introduce any new safety concerns (e.g., lifting and handling out of the vessel) or difficult failure scenarios. o Selection. The plates were cut into eight sections using the plasma arc torch. A total of 864 bolts that held the baffle plates to the core barrel were removed using an untorquing tool and a drill tool. The 4-meter-long pieces of highly irradiated stainless steel were highly radioactive with an estimated peak of 3,000 roentgen per hour gamma. Baffle plate handling exposed the upper core support assembly for defueling of the core former area. Two of the eight baffle plate sections were removed and hung from vent valve seats. The exposed area was defueled before removal of the next plate section. A computer model of the vessel was used to help select how to shuffle the baffle plates from hanger to hanger in a manner that least affected defueling and required the leasthandling. Each plate was essentially rotated 90 degrees from its original location to its final location.
[6] P.L. Winston, Management of the Three Mile Island Unit 2 Accident Corium and Severely Damaged Fuel Debris, Contribution to International Atomic Energy Agency Coordinated Research Proposal T13015, INL-EXT-21-61607, rev. 2, 2022.
[11] M.S. McGough et al., Performance of the Automated Cutting Equipment System During the Plasma Cutting of the Three Mile Island Unit 2 Lower Core Support Assembly, Nucl. Technol., 87 (1989) 648-659.
[2] A.W. Marley, D.W. Akers and C.V. Mclsaac, Sampling and Examination Methods Used for Three Mile Island Unit 2, Nucl. Technol. 87 (1989) 845-856.
[2] A.W. Marley, D.W. Akers and C.V. Mclsaac, Sampling and Examination Methods Used for Three Mile Island Unit 2, Nucl. Technol. 87 (1989) 845-856.
[6] P.L. Winston, Management of the Three Mile Island Unit 2 Accident Corium and Severely Damaged Fuel Debris, Contribution to International Atomic Energy Agency Coordinated Research Proposal T13015, INL-EXT-21-61607, rev. 2, 2022.
[11] M.S. McGough et al., Performance of the Automated Cutting Equipment System During the Plasma Cutting of the Three Mile Island Unit 2 Lower Core Support Assembly, Nucl. Technol., 87 (1989) 648-659.
[2] A.W. Marley, D.W. Akers and C.V. Mclsaac, Sampling and Examination Methods Used for Three Mile Island Unit 2, Nucl. Technol. 87 (1989) 845-856.
