Observations and prediction of frost heave of an experimental pipeline

Observations are reported of the frost heave beneath an experimental chilled pipeline (-5°C) buried in a controlled environment facility in Caen, France. Half of the length of the pipe is in a frost susceptible silt and the other half is in a sand. Over a 380-day period the pipe has heaved about 190 mm in the silt and only 40 mm in the sand. About 90% of the heave in the silt is due to segregated ice formation, with 14% of the total heave occurring within frozen ground (secondary heave). Predictions of the observed heave have been made based on laboratory frost heave test data and the thermal data collected from the main experiment. The laboratory heaving behaviour was found to depend on the method of sample preparation and on cyclic freezing. The test data were summarised by relating the velocity of water migration to the temperature gradient. Based on the data from compacted samples, the predicted heave is significantly low over a 380-day period but the prediction based on consolidated samples is generally within 10% of the observed heave. However, some divergence in the results is apparent after 250 days, because the approach is based on the temperature gradient rather than the net heat flow per se .