No abstract

This brief technical note describes a case history of pyritic shale heave discovered in an old underfloor fill of quarry waste composed of Lower Lias shales, at Barry, Glamorgan.

The Paper was first presented in 1984 at the Centenary Celebrations of the Glasgow and West of Scotland Association and was published in the centenary volume. The paper is based on Mr Burbidge's Masters thesis' in which he assembled an enormous number of case records, each of them meticulously presented and analysed.122. As mentioned in 116 no attempt was made to compare the results of the correlations derived in the Paper with predictions of other methods as this would have required a case-by-case comparison. One of my students carried out comparisons of nine of the case histories for which the soils data were well established (Maail 1985).90 The observed settlements p, were compared with the calculated settlements p, using six methods including the one given in the Paper. The results are given in Fig. 34 where the observed settlements are plotted against pJp, and pJp, for each method. A point lying above the horizontal axis represents an overestimate of settlement (conservative) whereas a point lying below the axis represents an underestimate of settlement (unconservative). The number against each point refers to the case number given in Appendix-1.123. The predictions given by Terzaghi and Peck7 are given in Fig. 34(a) and can be seen to be generally very conservative. Meyerh~f'.'~ increased the allowable bearing pressure given by Terzaghi and Peck by 50% and dispensed with the correction for the water table. As can be seen from Fig. 34(b) some of the predictions have become very unconservative. Peck, Hanson and Thornburn (1974) proposed a revised set of charts with an overburden correction applied to the SPT results.'' Tomlinson*' uses the original Terzaghi and Peck charts with an overburden correction applied to the SPT results. These two methods give results which are plotted in Figs 34(c) and (d) respectively. It can be seen that in both cases the agreement between observed and calculated settlements is reasonable. Fig. 34(e) shows the comparison with Parry's method.33 For most of the cases the calculations are unconservative. As a check Fig. 34(f) shows the comparison with the method proposed in the Paper. It is encouraging that most of the observations lie within one standard deviation of the best estimate of the settlement. Mr E. J. Wilson, Consultant Engineering GeologistThe Terzaghi and Peck7 chart for estimating settlement is arguably the most used and almost certainly the most misused chart in soil mechanics; but it was produced in 1948 by Terzaghi and Peck on the basis of a study of plate bearing tests, some standard and non-standard penetration tests, and probably no actual settlement measurements on structures. They made it clear that this was a very conser-

E. J. Wilson writes: Goulty & Al-Rawahy conclude that, since different mechanical processes are involved in active and residual subsidence, one should not attempt to predict the duration of residual subsidence from observations of active subsidence. Whilst this is obviously true in a literal sense, the possibility that active subsidence may result in delayed and possibly prolonged residual subsidence should be considered. The authors quote two cases by Orchard & Allen (1975) in the Durham Coalfield where the presence of thick competent strata prolonged residual subsidence to six years (Peterlee) and delayed the onset of residual subsidence by four years (North Durham). It is significant that both cases showed broadened subsidence profiles. They go on to quote Pine & Randon (1986) who recorded bedding separation below the thick Lower Magnesian limestone in the Selby Coalfield, where active subsidence at the Wistow Mine had produced less than half the predicted settlement, and concluded that bridging in the Permo-Trias can persist until the adjoining panels are worked out. Reference is made to further examples presented by Orchard & Allen (1975) and Collins (1978). In one of the cases quoted by Collins at Pengam, residual subsidence came to a stop after 2.5 years, and was noted by Collins to form a large proportion of the active subsidence. In that case however the active subsidence had only been about half of that predicted. At Pengam, as also at Bedwas, prolonged residual subsidence was ascribed to the presence of very thick Pennant Sandstone in the

E. J. Wilson write: Figure 1 of Edmonds' paper suggests a geographical distribution of solution features which is highly distinctive yet apparently defies geological or climatological explanation. One is reluctant to accept such distribution as a starting point for the analysis of a problem, unless either it can be explained, or it can be shown that it is not due to any other spurious factor. It is immediately obvious that the lowest density of solution features recorded is for Salisbury Plain, a large proportion of which, as a military training ground, has a land use which would be expected to have a very low sensitivity to solution features. It is also obvious that most areas of high density lie within the Home Counties. This suggests the hypothesis that solution features are more likely to be discovered, and hence to become the subject of available data, in more highly-developed areas. To put this hypothesis to the test, a comparison was made between population density as shown in the Ordnance Survey Atlas of Great Britain , and the density of solution features as shown by Edmonds. A direct and reliable comparison would be difficult, however, without much more detailed analysis, since there is considerable overlap between areas from the two sources. Furthermore, there are many small areas of high population density, but these are mostly concentrated in valley bottoms, which Edmonds' Fig. 2 shows are not a preferred location for solution features; therefore they have been ignored on the basis that they would