The International Information Center for Geotechnical Engineers

Ground Freezing




The following sections provide brief summaries of projects which have successfully implemented ground freezing, including the specific considerations and obstacles which were overcome by implementing AGF.  




The Boston CA/T project is perhaps the most well-known application of massive soil freezing to date.  The project involved the construction of  underground expressway tunnels to replace an aging elevated highway system.  Three tunnels were to be constructed using tunnel jacking.  The subsurface profile consisted of miscellaneous fill materials (boulders, cobbles, concrete and steel fragments, wood, brick, granite blocks, among others) for 6-8 meters, overlying 3-5 meters of organic silts, clays, and peat, overlying 1.5 meters of dense silty sand, overlying 5 meters of marine clays.  To stabilize the tunnel face, the initial design called for a combination of ground improvement techniques to handle the extreme soil variability, including chemical grouting, de-watering, horizontal jet grouting, and soil nailing.  


Figure 10. Vertical section of tunnel during operation (Dijk and Bouwmeester-van den Bos 2001)


A value engineering study showed that AGF could provide the stability required through each soil layer at the site at a lower cost than implementing four different ground improvement methods, therefore it was chosen to provide stability of the tunnel face for excavation and support of the tunnel jacking system.


The Boston CA/T project required multiple site-specific design considerations.  The tunnel sections were constructed beneath a railway (shown in Figure 9), therefore the freeze pipes were insulated at the top to keep the ground surface thawed for railway operations and maintenance.  Ground temperature and heave were monitored throughout the project and any damages to the railway system were corrected with routine maintenance.  Additionally, the freeze pipes were terminated approximately 1 meter above the tunnel invert, to prevent the tunnel from experiencing thaw settlements after construction.  Finally, heat pipes were installed at the edges of the frozen cutoff walls to prevent heave pressure from the expansion of the frozen wall from burdening the tunnel jacking system during excavations.  


Figure 11.  Railway operation around freeze pipes (FHWA 2013)


The project employed the chilled brine cooling method of AGF successfully, and varied pipe spacing to control the freeze time.  The Ramp D tunnel required a faster freeze time than the other two tunnels, therefore freeze pipe spacing was smaller (2.1 meters compared to 2.4 meters).  Freeze times were on the order of 3-4 months, depending on freeze pipe spacing.  


One system failure occurred during freezing.  The freeze pipes were damaged due to frost heave, and leaks were detected.  These leaks were repaired and all pipes were outfitted with redundant closed-end steel sleeves to prevent future leaks.  


The project was completed successfully without any further delays from failures of the ground freezing system.  The Boston CA/T project is an example of the successful implementation of ground freezing under extremely variable soil conditions over a massive volume of soil (van Dijk and Bouwmeester-van den Bos 2001).




The purpose of the project at Sophiaspoortunnel was to construct fourteen cross passages between parallel railroad tunnels.  


Figure 12. Cross section of a service shaft and cross passages (Crippa and Manassero 2006)



The subsurface consisted of a 15 m thick clay layer above a 10 m thick layer of loose sand (through which most of the tunnels exist) above another layer of clay.  The groundwater table was at a depth of 25 m.


For each of the fourteen cross passages, freeze pipes were installed horizontally between the tunnels to create a horizontal frozen soil column for excavation and support of the tunnel.  Both chilled brine (10 cross passages) and liquid nitrogen (4 cross passages) were used successfully.  Pipe spacing was on average 1 meter.  For each cross passage, 25 to 29 freeze pipes were installed to perform the freezing, with a design wall thickness of 1.8 to 2.3 m.  The total volume of frozen soil for all cross passages was 4400 cubic meters.


Per the design, the brine method took longer to reach full freezing and closure than the liquid nitrogen method.  Specifically, for the brine method, shell closure was reached in 8-15 days, with the minimum design thickness reached in 34-67 days.  For the liquid nitrogen method, closure was reached in 4-7 days, with the minimum design thickness reached in 9-14 days, much quicker than the brine method (Crippa and Manassero 2006).  


The Netherlands Sophiaspoortunnel study is a good example of the application of both brine and liquid nitrogen methods of AGF, and provides a comparison of timescale for each method as implemented in field conditions.



Add comment

NOTE: The symbol < is not allowed in comments. If you use it, the comment will not be published correctly.

Security code
*Please insert the above-shown characters in the field below.