Some myths die hard because so many people mistakenly believe them. Other myths persist, not because they are well known, but because they are obstinately repeated by a handful of people who strongly want to believe them.
In this diary, I'm going to address the myth that there was a scram failure during the accident at Unit 2 of the Three Mile Island (TMI) nuclear power plant in March 1979. That is, the myth claims that the nuclear chain reaction was not stopped within seconds of the initiation of the accident.
This is going to be a quick diary. I have already addressed the false claims several times in the comments of this website. Nevertheless, since this myth keeps reappearing over and over—most recently in the past week—I felt that the solid facts that debunk this particular lie deserve their own diary.
As far as I know, the "Failed Scram" Myth first appeared on DailyKos in a diary that was published a little over two years ago. This diary appears to draw from material in a small pamphlet published in 1985 by an organization calling itself the "Coalition for Independent Investigation" (scanned PDF available here). It came to my attention because I was quoted briefly in the diary and was called various unflattering names by the diarist.
Both the diary and the pamphlet built their case for a scram failure through conspiracy-theory reasoning about discrepancies between the details provided in the Report of The President's Commission on the Accident at Three Mile Island (which is available online), reports of the accident in the media, and details provided in later reports.
In particular, the early reports, including the President's Commission report, mention the reactor's 69 control rods and emphasize that the "control rods automatically dropped down into the reactor core to halt its nuclear fission ... as it was designed to do." Later reports included more detail, such that only 61 rods dropped into the core to shut down the reactor.
The eight-rod difference was taken by the Coalition to mean that "the reactor experienced significant scram failure" (and thus, the reactor was not shut down), even though they did discover and acknowledge that these additional eight rods were Axial Power Shaping Rods (APSR's), "which provide no safety or criticality control function." (However, a handwritten footnote in the scanned copy even claims that there was no such thing as an APSR before 1983!)
Although the Coalition did discover a significant source of confusion in the documented accounts of the accident, it is a shame that nobody in the Coalition had the first clue about how the control rods in the TMI-2 design operate. Furthermore, it is a shame that none of them bothered to check either the publicly accessible documents available from the Nuclear Regulatory Commission (NRC) or to check with the reactor vendor, since they could have obtained an accurate description of the design from either source.
The Truth
So what happened? Well, everyone agrees that 61 rods successfully fell into the core beginning eight seconds into the accident. The only controversy is over the eight additional rods, the APSR's.
Like the other unit at the Three Mile Island plant, Unit 2, which experienced the accident in 1979, was designed and built by the Babcock & Wilcox (B&W) company. Many of the components—such as the steam generators, the fuel assemblies, and the control rods—are specific to this company's designs.
Fortunately, the NRC provides some useful information on B&W reactors. For example, it has produced a "Crosstraining Course Manual" on B&W Technology. The chapter on the Control Rod Drive Control Mechanism is available online (PDF). On page 6.2-6, it describes how the rods are divided into eight groups that are symmetrically arranged about the core.
The first four groups consist of the safety rods. These rods remain completely out of the core while the reactor is in operation, and their sole purpose is to drop in the event that the chain reaction needs to be rapidly stopped, called a "trip" or a "scram."
Groups 5, 6, and 7 are used to control the power in the core. They are withdrawn in sequence to start up the reactor. As the reactor runs, the rods are inserted or withdrawn by the operators to control the neutron flux and can be used to change the reactor power output. Like the safety rods, these rods drop into the core when the reactor is tripped.
The last group, number 8, consists of eight APSR's. In the reactor, the neutron flux (and hence the power) is substantially less at the top and bottom of the core, due to neutron leakage and other factors. Thus, the APSR's, which have a short active region of neutron-absorbing material, can be positioned with the active region near the center of the core to reduce the neutron flux there and control the shape of the power along the length of the core (hence the name "Axial Power Shaping Rod").
Note that TMI-2 had only 69 rods (not 76 as mentioned in the manual). Their locations are diagrammed below. The eight APSR's are denoted by little x's.
So are the rods in Group 8 supposed to fall into the core like the other rods during a reactor trip?
More information on the APSR's is available in NUREG-1430 (PDF), Standard Technical Specifications for Babcock and Wilcox Plants. This document describes the NRC's requirements for safely operating a B&W reactor like TMI-2. It is a long, boring document full of technical details that is guaranteed to cure insomnia. Nevertheless, Section B-3.1.6 covers APSR's, and on page B-3.1.6-1 (page 59 of the PDF), it says (emphasis mine):
The APSRs, which control the axial power distribution, are positioned manually and do not trip.
Section B-3.3.4 explains how control rods are used in a trip. Page B-3.3.4-3 (page 163 of the PDF) reads (again, emphasis mine):
Accident analyses rely on a reactor trip for protection of reactor core integrity, reactor coolant pressure boundary integrity, and reactor building OPERABILITY. A reactor trip must occur when needed to prevent accident consequences from exceeding those calculated in the accident analyses. The control rod insertion limits ensure that adequate rod worth is available upon reactor trip to shut down the reactor to the required SDM. Further, OPERABILITY of the CRD trip devices ensures that all CONTROL RODS (except Group 8) will trip when required.
Thus, the rods in Group 8 are not supposed to fall during a reactor trip. They never were. They are not designed to.
Thus, if 61 rods dropped into the core at the beginning of the accident, then the reactor trip proceeded as planned, and the "Failed Scram" Myth is busted. Anyone repeating this myth either is uninformed or is spreading disinformation.
By the way, as the website of the Smithsonian National Museum of American History describes, the APSR's were one of the first tools used to test the conditions in the reactor's core after the accident. Unfortunately, the information that they provided turned out not to be useful.