On March 31, Chubu Electric Power Company (Chuden) submitted a \u201cReport on Inappropriateness in the Formulation of Design Basis Seismic Ground Motion in the Application for Compliance with the New Regulatory Standards for Hamaoka Nuclear Power Plant\u201d to the Nuclear Regulation Authority (NRA) regarding the data fabrication, disclosed on January 5 of this year (2026), of simulated seismic ground motions for regulatory review at the Hamaoka Nuclear Power Plant. The main content of the report concerns the method and history of the data fabrication, based on materials still remaining at Chuden and interviews with some of the related personnel.<\/p>\n
Two methods are commonly used to create simulated seismic ground motions: a method based on response spectra and a method based on fault models. Response spectra are used to divide the ground motion caused by an earthquake into periods and thus measure the strength of the impact. Assuming a point source, the response spectra method uses an empirical formula (distance attenuation formula) created from earthquake observation records to determine the ground motion at the destination based on the scale of the earthquake and the distance from the hypocenter. (The time history waveform is determined after determining the control point of the response spectrum.) Assuming the hypocenter as a plane (seismic fault plane), the method based on the fault model calculates the ground motion at the destination by modeling the destruction process of the fault plane. (The response spectrum is created after obtaining the time history waveform.)<\/p>\n
\u00a0<\/p>\n
Data fabrication methods<\/strong><\/p>\n What was discovered at Hamaoka this time was that the fabrication was based on the fault model. (More specifically, it is described, among fault-model-based methods, as being a \u201chybrid synthesis of Stochastic Green\u2019s Function Method and wavenumber integration.\u201d) At the review meeting, Chuden explained that it would select a representative seismic ground motion wave by randomly creating 20 sets of seismic ground motion waveforms, converting them into a response spectrum, then determining the average value, finally selecting the wave closest to the average value as the representative wave. (This approach itself is quite problematic, being only a rough assessment of the seismic capacity of nuclear power plants.) In fact, however, this was not what was done. Instead, the company rigged the seismic wave so that the representative wave selected would be exactly what the company wanted, or close to it.<\/p>\n Chuden has given two explanations of how it fabricated the representative seismic ground motion wave. The first, Method 1, was to create (20 sets of) seismic waves over and over again, not randomly, until a wave that was close to what it wanted appeared. According to Chuden, Method 1 had been in use since at least around 2012, soon after the Fukushima Daiichi nuclear accident.<\/p>\n The second, more blatant, Method 2, was to first determine the desired seismic wave and then create or select another 19 sets so that it would be selected as the representative wave. According to Chuden, Method 2 was prompted by a request from the NRA at the 570th Review Meeting on May 11, 2018 to change the position of the upper edge of the seismogenic layer from 8 km to a more severe 5 km in order to recognize the A-17 fault, located just to the west of the site of the Hamaoka Nuclear Power Plant, as an active fault and assess possible earthquakes. From this, it is natural to assume that the desired seismic wave would have a small value and only a minor impact on the Hamaoka Nuclear Power Plant.<\/p>\n \u00a0<\/p>\n Meaning of the fabrications<\/strong><\/p>\n According to the commissioned report (neither the report nor the contractor have been disclosed), Method 1 was used in at least 105 of the 225 cases selected by Chuden, and in 77 cases it was not possible to determine whether or not Method 1 or Method 2 was used. According to interviews with related personnel, Method 2 was used in 80 cases, only three cases being confirmed from descriptions in the commissioned report.<\/p>\n In any case, there is no doubt that fabrication was conducted not only on a part of the data, but over a wide range of fundamental data, and thus Chuden is not qualified to undergo deliberations for compliance. Furthermore, this places into question the raison d\u2019\u00eatre of the NRA, which failed to perform checks after making its request to Chuden. At the very least, NRA should have checked the supporting data and its formulation process.<\/p>\n \u00a0<\/p>\n Situation at the Hamaoka Nuclear Power Plant<\/strong><\/p>\n
![\"\"](\"https:\/\/cnic.jp\/english\/wordpress\/wp-content\/uploads\/2026\/08\/233_Hamaoka1-e1785732999507.jpg\")
<\/a>