Abstract Relict glacial and periglacial environments are widespread, and the deposits that they are associated with mean it is inevitable that the design and construction of many projects will be influenced by their presence and nature. Tills and other glaciogenic deposits prove to be particularly challenging in this context for reasons that include: the spatial variability of the nature of the deposits; the wide range of particle sizes often included within a given soil, including large-sized particles; spatial variation in soil type and properties; variation in depth to rockhead and variable degrees of weathering and alteration; the presence of groundwater, that is misinterpreted as perched water, as well as sub-artesian and artesian conditions; the presence of solution features and fissures, partly or completely infilled with soft or loose material; and the presence of (often shallow) shear surfaces at residual strength. In this chapter, some of the more common problems and associated solutions associated with earthworks and man-made slopes, tunnels and underground structures, dams and reservoirs, foundations, and offshore engineering and installations are reviewed. It is important that great care is taken in addressing the influences of variability, complexity and uncertainty inherent in glacial/periglacial soil formations at all stages of the construction process, from feasibility to end-of-project activities, such as preparation of the as-built drawings.

Abstract Up to 9 × 10 6 l of light non-aqueous-phase liquid (LNAPL) is present in a 400 m-wide zone in the Permo-Triassic sandstone 30 m below a working industrial site. Remediation by skimmer wells failed to meet the expectations of the regulatory authorities. A detailed study has concluded that this form of remediation is not possible in this formation. Initial estimates of the volume of LNAPL in the sandstone had been made by applying the concept of correcting for an ‘exaggerated thickness’ in the wells and multiplying by the porosity. Regulatory requirements were to remove most, preferably all, of the LNAPL. Site observations and measurements from high-quality core samples showed that the fine pores of the rock obey the capillary characteristic model. This was derived some 50 years ago in the oil production industry but has, in the opinion of the author, not been fully appreciated by the contaminated groundwater remediation sector in the UK. The model has been used to make a more reliable estimate of LNAPL volume and to demonstrate to the regulatory authorities the processes involved and how these control the ability to clean the site. The concept of ‘apparent volume’ is introduced as an overall measure of apparent thickness in the site-wide monitoring borehole network. Apparent volume is shown to be broadly negatively correlated with groundwater-level fluctuations and so the measure of changes in apparent thickness is not a sound basis for regulatory compliance. It was found that LNAPL thickness in wells is less than the thickness of the contaminated zone in the sandstone and, importantly, the degree of saturation is no more than 30%. This leads to a reduction in effective permeability of two orders of magnitude (from the fully saturated condition), which has serious implications for remediation. Most of the LNAPL is immobile and so pumping is ineffective; it does not reduce the volume of aquifer impacted nor does it have any significant effect on reducing the source zone for dissolved-phase generation, the size of any dissolved phase plume or the longevity of such a plume.