Journal of Cold Regions Engineering

Over the last decade, wind turbines have been constructed in several villages throughout western Alaska because of the abundance of wind sources. Because of the extensive existence of warm permafrost and the high cost of concrete foundations, many wind turbine towers are founded on steel-pipe piles connected by a prefabricated RC cap and thermosiphons. This type of foundation is softer than concrete foundations, and its stiffness is subject to variation due to seasonal and long-term temperature changes in permafrost and seasonal freezing and thawing in the active layer. Field performance data of wind turbine towers supported by such foundations are scarce. This paper describes a wind turbine founded on a steel-pipe pile group embedded in warm permafrost; presents climate data, site conditions and foundation performance of a wind turbine in western Alaska; and presents the dynamic response data during the 2013–2014 winter season. It identifies the seasonal variation of the fundamental frequency of the turbine tower–foundation system and potential operational issues for wind turbines founded in warm permafrost. In particular, this paper discusses in detail the resonance between the rotor and the tower–foundation system observed in the recorded data, and possible solutions to this issue. In addition, it presents the dynamic properties of the wind turbine foundation system, including modal frequencies and damping ratios identified by the Hilbert–Huang transform (HHT) using the resonance data. These findings and the dynamic properties could be useful in design and dynamic response analyses of wind turbines in cold regions.

One of the most important lessons learned from Alaska’s major earthquakes is that lateral spread of frozen ground crust overlying liquefiable soils induced extensive bridge foundation damage. A shake table experiment was performed to study the interaction mechanisms of frozen soil-pile foundation in this scenario. Data collected from the shake table experiment were presented and analyzed. Three plastic hinges formed on the model pile: one at the ground crust surface, one at the ground crust-loose sand interface induced by strong confinement of the stiff ground crust, and one within the partially liquefied medium dense sand layer induced by laterally spreading ground crust. The failure mechanisms of single piles revealed in this experiment corresponded well with pile-pile cap connection failures observed in previous earthquakes and the frozen ground crust can be thought of as acting like a natural pile cap that restricts the rotation of the single pile head. This study provided data for computer model calibration and helped gain insight into the frozen ground crust impact on the seismic performance of piles under lateral spread induced by liquefaction.

The in situ strength of polar snow is not easily measured; most assessments rely on indices derived from snow density and/or are limited in depth. To overcome these deficiencies, modified cone penetration testing (CPT) equipment is used to obtain rate-variable tip-resistance and sleeve-friction data in Antarctic firn to depths of 10 m. A physical model of snow deformation ahead of the penetrating cone is considered to allow generation of a continuous profile of snow strength. Dynamic profiling of strength to depth in polar snow is possible using the cone penetration test. However, more in situ strength data are needed to comprehensively validate the simple method presented herein.

Long-term thermal effects of air convection embankments (ACEs) over 550-km-long permafrost zones along the Qinghai–Tibet railway were analyzed on the basis of 14-year records (2002–2016) of ground temperature. The results showed that, after embankment construction, permafrost tables beneath the ACEs moved upward quickly in the first 3 years and then remained stable over the next 10 years. The magnitude of this upward movement showed a positive correlation with embankment thickness. Shallow permafrost temperature beneath the ACEs decreased over a 5-year period after embankment construction in cold permafrost zones, but increased sharply concurrent with permafrost table upward movement in warm permafrost zones. Deep permafrost beneath all the ACEs showed a slow warming trend due to climate warming. Overall, the thermal effects of ACEs significantly uplifted underlying permafrost tables after embankment construction and then maintained them well in a warming climate. The different thermal effects of ACEs in cold and warm permafrost zones related to the working principle of the ACEs and natural ground thermal regime in the two zones.