Field Observation of Construction-Induced Liquefaction during Closure of a Coal Ash Impoundment

Ponded coal combustion residual (CCR) is an unusual soil with distinct particle and bulk-scale characteristics compared with natural sediments (Bachus et al. 2019). The catastrophic accident at the Kingston coal-powered electric generating plant in 2008 revealed the destructive potential of CCR pond failures (Santamarina et al. 2019). There are over 700 operating CCR ponds in the US, many of which are currently under closure-related construction following regulations imposed by the US Environmental Protection Agency, known as the 2015 CCR Rule (USEPA 2015).

In 2019, construction loading–induced liquefaction was observed during closure of a CCR pond located in Virginia. The approach used for this project was closure-in-place, in which liquids were removed from the ponded CCR material, the top surface was graded, and a final cover was placed over the pond. Construction was closely monitored using automated instruments, including vibrating wire piezometers (VWP), in-place-inclinometers (IPI), and an automated motorized total station. Liquefaction initially was observed in an area where grading was performed. A standard penetration test (SPT) was conducted and a VWP was installed in the same borehole at a depth of 5.5 m. Fig. 1 presents the subsurface profile at this location and time histories of pore pressure ratio ($r_u$ = excess pore pressure/initial effective vertical stress) calculated based on the VWP recordings. The subsurface mainly consisted of soft fly ash (0–6 m) and fat clay (6–10 m). The average SPT $N$ value for the CCR was 1.1. Notably, $N=0$ was reported at a depth of 3.6 to 6.0 m, indicating the extremely soft CCR materials at this site.

After the start of grading, pore pressure was observed to increase due to construction loading imposed by the bulldozers. On Day 5, $r_u$ reached 1.0 and a CCR ejection cone was observed, indicating liquefaction had occurred in the very soft CCR material below. Construction was temporarily halted due to safety concerns, and the excess pore pressure started to dissipate. On Day 11, $r_u$ had decreased to 0.67 and, when construction resumed, pore pressure again increased rapidly, with $r_u=0.98$ reached in 10 h, and construction was halted. On Day 21, $r_u$ had dropped to 0.47 and grading resumed. The value of $r_u$ was observed to increase again, but no CCR ejection was observed at this time. On Day 23, grading was finished in the area. Fig. 2 shows a CCR ejection cone observed near the VWP location ($\mathrm{diameter}\approx 6\mathrm{m}$).