The influence of earthworks construction on porewater pressures in clays and mudstones of the Lias Group

The trial cutting and embankment were located at Boddington near Wormleighton, 14 km north of Banbury, England (52° 11′ 14″ N, 1° 20′ 21″ W). Boddington is within a c. 18.2 km-wide outcrop of the Charmouth Mudstone Formation that is crossed by the HS2 railway between London and Birmingham (Fig. 1). Boddington is located in the East Midlands Shelf basin of the Lias Group, where the Charmouth Mudstone Formation is c. 100–150 m thick (Hobbs et al. 2012). It was formed c. 183–199 Myr ago in a shallow-marine environment and was subsequently overconsolidated. The formation is principally mudstone, with alternating sequences of mudstone, limestone and sandstone that result from cyclical patterns of deposition. Boddington is located to the north of a region of lowland periglaciated terrain in central and southern England (Booth et al. 2015), and close to the southern limit (c. 147 kyr ago) of the Moreton Stadial glaciation of the Late Wolstonian Substage (Gibson et al. 2022). It was therefore subjected to periglacial and contemporary weathering during the Quaternary Period. The mudstone has weathered to clay at shallower depths (0–11 m below ground level (bgl)), forming a gradational weathering profile from the ground surface (Briggs et al. 2022).

The trial cutting was constructed between 23 July 2019 and 11 December 2019 to a final depth of 13 m bgl at the centre. Aerial drone surveys were undertaken to measure the progress of the excavation at approximately weekly intervals. When finished, the cutting was c. 290 m long and 130 m wide at the original ground level (Fig. 2). The base of the cutting was 15 m wide, with slopes excavated to c. 1 (V) in 4 (H). Directly to the north and adjacent to the trial cutting, a trial embankment was constructed in stages between 7 November 2020 and 9 December 2020 using fill excavated from the trial cutting. It was c. 150 m long and 95 m wide at the base, with a crest width of 55 m and a slope angle of c. 1 (V) in 2.3 (H) (Fig. 2). The final height of the embankment was 8.2 m. It was demolished almost 2 years later in the autumn of 2022. The Oxford Canal passes 100 m to the north of the trial embankment at a low point in the local topography at c. 116 m above Ordnance Datum (AOD). Surface water ponding occurred within the trial cutting at Boddington in the autumn of 2019, following a wet summer and prolonged autumn rainfall in central England (Meteorological Office 2024a). Surface water ponding also occurred in the spring months of both 2020 and 2021, immediately after construction of the trial cutting and trial embankment, respectively. This correlates with instances of high water levels in the River Cherwell at Cropredy Bridge, 5 km south of Boddington (Environment Agency 2024).

A ground investigation was undertaken by commercial contractors along the full length of the HS2 alignment through the Charmouth Mudstone Formation outcrop, including at Boddington. Instruments including weather stations, piezometers, inclinometers and extensometers were installed beneath the trial cutting and the trial embankment prior to construction, and monitored during and after the construction period. The monitoring data were used to observe the heave and settlement of the trial cutting and trial embankment, respectively, to inform the design of earthworks for the HS2 railway. This study analyses only the piezometer and weather station data.

More than 40 boreholes were drilled and logged at Boddington prior to the construction of the earthworks. Samples were recovered for laboratory index testing, oedometer testing and triaxial testing. In situ testing was undertaken to measure the saturated hydraulic conductivity and stiffness of the ground profile.

Figure 3 shows simplified stratigraphic ground profiles beneath the trial cutting and trial embankment, as informed by the borehole strata descriptions. The ground surface at the trial cutting was located at 136 m AOD and consisted of 13 m of weathered clay overlying weathered mudstone and a c. 2 m-thick band of limestone (described as calcareous siltstone) at 104 m AOD (32 m bgl). Unweathered mudstone was present below 102 m AOD. The trial embankment was located downslope and to the north of the trial cutting at an elevation of 122 m AOD. At the trial embankment site, the ground profile comprised c. 12 m of weathered clay overlying weathered mudstone and a c. 2 m-thick band of limestone at 104 m AOD (18 m bgl). Again, the mudstone below 102 m AOD was unweathered. At both the trial cutting and trial embankment sites, an additional limestone stratum was identified at c. 84 m AOD in a limited number of boreholes. Examination of borehole logs from across the wider Charmouth Mudstone Formation outcrop showed that the limestone strata beneath Boddington dips gently (<1°) to the SE. The transition from clay to mudstone and limestone (Fig. 3) was discernible in optical borehole images from Boddington (not shown). Figure 4 shows that the moisture content (%) and plasticity index (%) of the clay and mudstone at Boddington are depth dependent. They are greatest and most variable near the surface (<5 m bgl) but more uniform at depth. These data are consistent with measurements in the Charmouth Mudstone Formation outcrop north of Banbury, England (Briggs et al. 2022).

Figure 5 shows the saturated hydraulic conductivity (k_sat) profile at Boddington, derived from tests on triaxial specimens 75–100 mm in diameter (BS EN ISO 17892-11:2019: BSI 2019), oedometer tests (BS EN ISO 17892-5:2017: BSI 2017) and in situ packer permeability tests (BS EN ISO 22282-3:2012: BSI 2012) using the Preene (2019) interpretation. The laboratory tests show that the saturated vertical hydraulic conductivity (k_v,sat) of the samples was between 10⁻¹³ and 10⁻¹⁰ m s⁻¹. This is within the general range for weathered clays and mudstones (Smethurst et al. 2006; Hobbs et al. 2012), although laboratory test data from larger samples (at least 250 mm in diameter) would have given a better indication of the influence of soil fabric (Rowe 1972). The in situ tests gave values of saturated hydraulic conductivity many orders of magnitude greater (10⁻⁷–10⁻⁶ m s⁻¹) than the laboratory tests. This is because the in situ tests involve substantially horizontal flow and hence are most representative of the saturated horizontal hydraulic conductivity (k_h,sat). The horizontal hydraulic conductivity of a natural deposit is usually greater than the vertical hydraulic conductivity, often by orders of magnitude if there are preferential flow pathways controlled by bedding (Powrie and Roberts 1990) or other discontinuities (Rowe 1972; Preene 2019). Neither the laboratory nor the in situ measurements showed any significant decrease in saturated hydraulic conductivity with depth; this is consistent with the highly overconsolidated nature of the deposit.

Figure 6 shows shear-wave velocity (m s⁻¹) measurements made in four boreholes (DHGEO) at Boddington by a specialist contractor for the HS2 ground investigation. Shear-wave velocities were generated by a sledgehammer striking the end of a timber sleeper at the ground surface. They were detected by a BGK-7 multi-element geophone with one vertical and six horizontal sensors, with azimuthal axes at 30° intervals. The BGK-7 multi-element geophone was pneumatically clamped in the boreholes at each successive test depth, which were at 1 m intervals. Broadly, the measurements show transitions in the shear-wave velocity profile and, by implication, in stiffness (Poulos 2022). These have been plotted by elevation so that the measurements in the deeper, less-weathered clays and mudstones from both the trial cutting and trial embankment can be compared. Figure 6 shows four zones: (i) 128–136 m AOD; (ii) 116–128 m AOD; (iii) 102–116 m AOD; and (iv) below 102 m AOD. There was a bi-linear increase in shear-wave velocity (and therefore stiffness) with depth in the clay (>123 m AOD) and weathered mudstone (116–123 m AOD) in zones (i) and (ii). This corresponds to a stiffness profile dependent on in situ stress and the void ratio, as is typical for a clay (Vardanega and Bolton 2013; Briggs et al. 2024). Below 116 m AOD in zones (iii) and (iv) the measurements are more scattered and increase at a lesser rate with depth or in situ stress than above. This is typical of a transition from clay to rock (i.e. an unweathered, cemented mudstone). The shear-wave velocity profile did not vary at the transition between the mudstone and the limestone (102–104 m AOD). The measurements at DHGEO_2 differ from those at other locations in that they reduce to low values below 116 m AOD. This could not be explained with reference to the borehole strata descriptions or the optical borehole log, and the reason for it remains unclear.

Vibrating wire piezometers and GeO flushable piezometers were installed using the fully grouted method and maintained at Boddington by a contractor to inform the design of earthworks for the HS2 railway (Fig. 2). Piezometers installed below the excavation level of the trial cutting (i.e. below 123 m AOD) and beneath the centre of the trial embankment (i.e. below the original ground surface at 122 m AOD) were selected for analysis because these locations were at the centroids of the zones of uniform unloading or loading. Piezometers located in regions of non-uniform loading, such as below the slopes of the trial cutting, above the excavated level of the cutting or close to the edge of the trial embankment, were not considered. A small number of piezometers malfunctioned and required resetting or reflushing during the monitoring period. These were also excluded from the analyses. During the installation of piezometers in borehole P1, a loss of grout through a preferential flow pathway occurred, which was accompanied by a water-level rise at a borehole downslope. The borehole then collapsed at 12 m bgl overnight. The measurements from the shallower piezometers at P1 (at 10 and 20 m bgl) were therefore considered unreliable, and those from the adjacent borehole P1A (Table 1) were used in the analyses in preference.

Details of the fully grouted piezometers installed at Boddington and selected for analysis are shown in Table 1. Geotechnical Observations Ltd GeO flushable piezometers (GeO FP) were installed above 104 m AOD at the trial cutting to allow measurement of both positive and negative porewater pressures. Geosense VWP-3000 series vibrating wire piezometers (VWP3000) were installed below 104 m AOD at the trial cutting. RST Instruments VW2100 vibrating wire piezometers (VW2100) were installed beneath the trial embankment. The GeO flushable piezometers have Geokon 4500C sensors and high air-entry (3 bar) porous filters so they can measure negative porewater pressure (suction) to −100 kPa (Ridley et al. 2003). The porous filters were saturated in a vacuum chamber and transported to the site in de-aired water. Each piezometer consists of a 50 mm-diameter UPVC pipe with a high air-entry porous filter at the bottom of the pipe and a sensor/valve that is screwed into place at the bottom of the pipe and sealed with an O-ring. The pipe is grouted in place and the grout acts as a secondary filter for transmitting negative porewater pressures from the soil to the piezometer (Tarantino et al. 2008). According to BS EN ISO 18674-4:2020 (BSI 2020), piezometers that do not include a method of removing air (e.g. flushing tubes) should have low air-entry filters. The Geosense and RST Instruments vibrating wire piezometers were therefore fitted with low air-entry porous filters, thus making them suitable for the fully grouted installation method. Both the Geosense vibrating wire piezometers and the RST Instruments vibrating wire piezometers were assembled by joining the bodies of the piezometers and the porous filters in de-aired water before they were attached to UPVC installation pipes (facing upwards). They were then lowered into the boreholes and grouted in place using a water–cement–bentonite grout ratio of 2.0 : 1.0 : 0.3 by weight. The piezometers beneath the trial cutting were installed between January and March 2019, before excavation began in July 2019. The piezometers were installed beneath the trial embankment in October 2020, before construction began in November 2020.

Two weather stations were installed at Boddington (Fig. 2). The weather stations recorded hourly measurements from: (i) a Campbell Scientific CS215 sensor to measure temperature (°C) and relative humidity (RH: %); (ii) a Campbell Scientific ARG100 tipping-bucket rain gauge to measure rainfall (mm h⁻¹); (iii) a Campbell Scientific 03002 Wind Sentry to measure wind speed (m s⁻¹) and direction (°); and (iv) a Campbell Scientific CS106 barometer to measure the barometric pressure (kPa). The two weather stations were located within 300 m of each other (Fig. 2) and their measurements were in close agreement.

Construction of the trial cutting lowered the ground level by up to 13 m. This is considerably below the depth of 0.8 m bgl, below which porewater pressures were found to be hydrostatic prior to construction (Briggs et al. 2022). Thus, once the excavation was complete, it provided a local sink for groundwater flow through the base and sides of the cutting, which was removed by surface drainage. This resulted in a long-term lowering of the equilibrium porewater pressures in the ground beneath the earthworks at Boddington, with a maximum reduction of 127.5 kPa at the centre of the trial cutting.

Figure 8 shows annual cumulative totals of rainfall and PET at Boddington, compared with Met Office (2024b) annual long-term average (LTA) rainfall at Wellesbourne and for the Midlands. Wellesbourne is at a lower elevation and drier than Boddington. The PET is relatively constant each year, in agreement with observations at other UK earthwork sites (Smethurst et al. 2006; Briggs et al. 2016). The total annual rainfall reduced each year from 2020 to 2023 and the earthworks monitoring period was generally drier than the LTA, with the exception of the wet summer and autumn of 2019 (Met Office 2024a). The atmospheric water balance (B_atm) at Boddington (given by the rainfall minus the PET) increased during the winter months (September–April) – that is, the atmosphere became wetter – and reduced during the summer months (April–September) (Fig. 9).

Figure 9 shows changes in porewater pressure and the atmospheric water balance (B_atm) beneath the trial cutting (Fig. 9a, b) and the trial embankment (Fig. 9c) following the start of construction. The piezometers located beneath the cutting and above 104 m AOD showed rapid reductions in porewater pressures during construction, followed by gradual long-term reductions in porewater pressures of up to about 125 kPa (Fig. 9a). The piezometers beneath the cutting and below 104 m AOD also showed rapid, but more limited, reductions in porewater pressures during construction, followed by more gradual long-term reductions in porewater pressures (c. 25–75 kPa) than at higher elevations. There were small, seasonal variations in these porewater pressures (<5 kPa), showing a partial recovery of excess porewater pressures towards the end of the excavation period (Fig. 9b). The measurements beneath the trial embankment (Fig. 9c) show rapid increases in porewater pressures during construction, rapid decreases during removal of the embankment and long-term trends of gradual porewater pressure reductions (c. 25 kPa over the period), similar to those observed at greater depths beneath the cutting. There are small, seasonal variations in these porewater pressure measurements (<10 kPa) that are greatest at depths below 104 m AOD.

Figure 10 shows profiles of porewater pressure beneath the trial cutting with elevation for dates at the start of construction, at the end of construction and in the longer term. A limited number of piezometer measurements (Briggs et al. 2022) indicated a hydrostatic profile of porewater pressures below 0.8 m bgl in the clay prior to construction (not shown). Figure 10 indicates an initial under-drained porewater pressure profile within the shallower mudstone (104–123 m AOD) and hydrostatic pressures below a zero-pressure line at c. 120 m AOD in the deeper mudstone (below 104 m AOD). After excavation of the trial cutting, porewater pressures reduced most significantly at the middle and base of the shallower mudstone. This corresponds to zone (iii) in the shear-wave velocity profile in Figure 6. Smaller reductions of porewater pressure were measured near the excavated surface (zone (ii) in Fig. 6) and in the deeper mudstone (<104 m AOD). The porewater pressures measured near the excavated surface (115–123 m AOD) show greater scatter than at depth. They may have been influenced by ground disturbance during the excavation or the infiltration of ponded surface water during the wet autumn of 2019.

Figure 11 shows profiles of porewater pressure beneath the trial embankment with elevation for dates at the start of construction, the end of construction and following removal of the embankment. It shows hydrostatic porewater pressures throughout the ground profile that are in agreement with pre-construction measurements in the clay (Briggs et al. 2022). There were no piezometer measurements within the shallower mudstone (104–110 m AOD), so it was not possible to assess whether it was under-drained by the limestone, as it was below the trial cutting site. The porewater pressure profile was hydrostatic below 0.8 m bgl (121.2 m AOD). This is similar to the porewater pressure profile below 120 m AOD in the deeper mudstone beneath the trial cutting (Fig. 10). Porewater pressures increased during construction of the trial embankment at all elevations, and then reduced during removal of the embankment.

Figure 12 compares the changes in porewater pressures due to construction and removal of the earthworks with changes in the average total stresses calculated using equation (5). This uses porewater pressures measured at the end of construction of the trial cutting (c. 141 days) and at the end of construction of the trial embankment (c. 32 days). A line for a B ratio of unity is also shown. Figure 12 shows that the results from construction (loading) and removal (unloading) of the trial embankment are almost equal and opposite. They lie closer to the line of equality than the results from the construction of the trial cutting. The seven results for the trial cutting at p ≤ −250 kPa are those from the shallower mudstone within the weathered clay and mudstone in zone (ii), shown in Figure 6. Consolidation analysis based on likely drainage path lengths (discussed later) shows that a significant dissipation of excess porewater pressure would have occurred before the end of construction at both the trial cutting and trial embankment (Fig. 9). Therefore, B values of less than unity would be expected, even in a fully saturated and relatively compressible soil. This is especially the case for the trial cutting, with its construction time of 141 days, and for piezometers at both sites that were located at the shallowest depths.

Figure 13 plots the calculated B values (equation 6) immediately after the construction and removal of the earthworks against the elevations of the piezometers. Beneath the trial cutting, the B values were generally less than 0.3 in the weathered clay and mudstone in zone (ii). In the mudstone (zones (iii) and (iv)) below, they reduced with elevation from a maximum value of 0.43 to 0.04. The B values directly beneath the trial embankment were 0.57 (at 113 m AOD). This was described as weathered clay in the borehole logs at the location of the embankment (Fig. 3) but corresponds to the elevation of zone (iii) in the shear-wave velocity profile from beneath the nearby trial cutting (Fig. 6). Lower B values (B < 0.5) are shown in the mudstone at lower elevations. At each elevation, the B values beneath the trial embankment were similar during construction (i.e. loading) and removal (i.e. unloading).

The porewater pressures in the clay and mudstone above 104 m AOD responded immediately to the construction of the earthworks (i.e. increased or decreased) but the magnitude of the change was less than the estimated change in average total stress. The B values were approximately 0.1–0.4 and 0.1–0.6 at the trial cutting and trial embankment, respectively. Porewater pressures in the mudstones located below the limestone (at 102–104 m AOD) were hydrostatic below a zero pressure line at c. 120–121 m AOD. They were relatively unresponsive to loading from the construction of earthworks at the ground surface.

At the trial cutting, the shallower mudstone above the limestone was initially under-drained by the strata below. Corresponding data were not available from beneath the trial embankment. Following construction, the porewater pressures at Boddington gradually reduced towards a new, lower equilibrium level. This new equilibrium level may reflect permanent changes to the ground and groundwater levels at the trial cutting (i.e. a 13 m lower elevation), which may also have influenced the trial embankment located directly downslope to the north of the cutting.

The observations at Boddington may be compared with those from a deep, staged, multi-propped excavation (Richards et al. 2007) in Upper Atherfield Clay and Lower Atherfield Clay (Greensand Group) strata under-drained by the more permeable Weald Clay (Wealden Group). Richards et al. (2007) showed that porewater pressures immediately reduced at each stage of construction and following reduction of the total stresses, with B values of between 0.67 and 0.77. New, reduced porewater pressure equilibrium conditions were reached in 4–15 weeks. This was the result of a short construction time and the result of passive, permanent drainage through a semi-permeable, contiguous piled wall, which created shorter drainage paths than at Boddington. Similar measurements in a deep, staged, multi-propped excavation in Gault Clay (Selbourne Group) showed immediate porewater pressure changes in the clay. The initial B values ranged between 0.5 and 0.75, increasing towards 0.9 as the level of the excavation approached the elevation of the piezometers (Ng 1998).

The B values measured by both Richards et al. (2007) and Ng (1998) were greater than those measured in the weathered clays and mudstones in zone (iii), above 104 m AOD, at Boddington. This could be partly due to the longer construction times at Boddington (and the longer period available for the dissipation of excess pore water pressures) and partly due to the cementation and greater stiffness of the mudstone, relative to the more compressible Atherfield Clay and Gault Clay. For example, the small-strain shear modulus (G₀) of the mudstone at Boddington is c. 350 MPa at 16 m bgl (Briggs et al. 2024). This is more than three times greater than the Gault Clay at the equivalent depth (c. 100 MPa: Ng et al. 1995) at the site described by Ng (1998). The high stiffness of the mudstone at Boddington makes the B values, even in truly undrained conditions, very sensitive to small reductions in the saturation ratio and the presence of pore air (Black and Lee 1973). Pore air could arise from the exsolution of dissolved gas during groundwater lowering (Powrie and Roberts 1990). Black and Lee (1973) showed that a reduction in the saturation ratio of just 0.5% from 100% (fully saturated) to 99.5% causes a reduction in the B value from 1.0 to 0.2 in very stiff soils such as those at Boddington. This is confirmed by the calculation shown in Appendix  A (equation A1). Thus, the low B values at Boddington may be attributable to the high stiffness of the clays and mudstones in two respects: first, the direct impact of the stiffness in increasing the consolidation coefficient and reducing the timescale required for excess pore pressure dissipation; and, secondly, the resulting sensitivity to very small degrees of desaturation. The effects are seen especially in the B values at shallow depth (c. 120 m AOD) beneath the trial cutting, where the reductions in mean total stress are greatest, the porewater pressures are lowest and the drainage pathway for excess porewater pressure dissipation during construction was shortest.

Beneath the trial cutting, the 19 m-thick layer of clays and mudstones above the limestone (>104 m AOD) can drain upwards into the ground surface and downwards into the limestone below (d_v of c. 9.5 m), or horizontally downslope and to the north (d_h of c. 300 m). The small-strain shear modulus (G₀) of the clays and mudstones above the limestone at Boddington is c. 350 MPa, reducing to 30% of this value, G_0.1, at 0.1% strain (Briggs et al. 2024). Table 2 shows estimates of in situ saturated hydraulic conductivity in the vertical (k_v,sat) and horizontal (k_h,sat) directions based on these assumptions. Table 2 shows an estimated in situ k_v,sat value of 2 × 10⁻¹¹ m s⁻¹. This is approximately one order of magnitude lower than was measured in the triaxial permeability tests but one order of magnitude greater than was measured in the oedometer compression tests at depth (Fig. 5). Therefore, the estimate of in situ k_v,sat is between those measured in the two types of laboratory test. This is reasonable given the difficulty of accurately assessing the hydraulic conductivity of low-permeability (Preene and Powrie 1993) or fissured soils (Rowe 1972), and that this is likely to be a lower-bound estimate for k_v,sat below the trial cutting due to the assumed drainage path length. Table 2 shows that the estimated in situ k_h,sat value of 2 × 10⁻⁸ m s⁻¹ is approximately two orders of magnitude lower than was measured in the in situ packer tests (Fig. 5), even assuming d_h is c. 300 m. If the assumed drainage path length was reduced to a d_h value of c. 30 m, the estimated in situ k_h,sat value would reduce further to 2 × 10⁻¹⁰ m s⁻¹. Therefore, the estimated in situ k_h,sat values are likely to be upper bounds that are still far less than those measured in the in situ packer tests. The estimates of k_v,sat and k_h,sat suggest that the dissipation of excess porewater pressures due to construction of the trial cutting, and the embankment, are most consistent with vertical drainage paths and the k_v,sat values measured in laboratory tests.

Porewater pressures in weathered clays and mudstones of the Charmouth Mudstone Formation (Lias Group) at Boddington, central England, were measured during the excavation of a 13 m-deep trial cutting and construction of an 8.2 m-high trial embankment. These were compared with ground investigation data and weather station measurements. The following conclusions can be drawn:

1. Ground investigation data show a gradational weathering profile with depth in the clays and mudstones of the Charmouth Mudstone Formation. The clay near the ground surface is weathered, with weathered mudstone and unweathered mudstone below. A layer of calcareous siltstone (limestone) is located at c. 104 m AOD, below both the trial cutting and the trial embankment. The porewater pressures in the mudstone below the limestone were hydrostatic below c. 120 m AOD, at both the trial cutting and the trial embankment. The initial porewater pressure measurements (prior to construction) show that where the shallower mudstone is of significant thickness (e.g. it is 19 m thick beneath the cutting), it is under-drained by the underlying limestone. In terms of seasonal variations, the mudstone at elevations below 104 m AOD is isolated from the mudstone and clay layers above.

2. Construction of the earthworks immediately affected the porewater pressures beneath both the trial cutting and the trial embankment. A consistent B ratio was obtained for both the loading and unloading of the weathered clays and mudstones beneath the trial embankment. However, the B ratios were lower than those reported from back-analyses of deep excavations in younger, more compressible overconsolidated clays and mudstones (Ng 1998; Richards et al. 2007). This may be explained by the construction times (30–140 days) being significant in comparison with the timescale of excess pore pressure dissipation, so that the construction and excavation processes were partly drained. The inferred higher values of consolidation coefficient are partly a result of the much greater stiffness of the clays and mudstones forming the Lias Group, compared with younger and more clay-like geological deposits. This high stiffness also results in an increased sensitivity of the B value, even in genuinely undrained conditions, to very small degrees of desaturation (e.g. from 100% to 99.5% saturated) in stiff clays, as shown in laboratory tests (Black and Lee 1973) and in a calculation presented in Appendix  A.

3. The piezometer data show a reduction in long-term equilibrium porewater pressures beneath both the trial cutting and the trial embankment. This is the result of reductions in the surface load and groundwater lowering associated with excavation of the cutting. Long-term porewater pressure lowering is most evident in the weathered clay and mudstone strata above the limestone layer at 104 m AOD, at both the trial cutting and the trial embankment. An estimate of the in situ saturated hydraulic conductivity in the vertical (k_v,sat) and horizontal (k_h,sat) directions was compared with results from in situ and laboratory tests. This showed that excess porewater pressures generated beneath the trial cutting and the trial embankment during construction were likely to have dissipated along vertical drainage paths at a rate that was consistent with k_v,sat values derived from laboratory tests.

The data were provided by HS2 Ltd. The authors are grateful to Dr Nicola Bicocchi, Mr Daniele Fornelli, the staff at Geotechnical Observations Ltd, COWI and EKFB; the late Prof. Lee Barbour and his colleagues at the University of Saskatchewan; and to the reviewers for their thoughtful comments and suggestions.

KMB: data curation (equal) formal analysis (lead), methodology (lead), visualization (lead), writing – original draft (lead); YTG: data curation (equal), investigation (equal), writing – review & editing (supporting); GJM: formal analysis (supporting), methodology (supporting), writing – review & editing (supporting); AR: writing – review & editing (supporting); WP: formal analysis (supporting), supervision (equal), writing – review & editing (supporting); SB: writing – review & editing (supporting); NS: supervision (equal), writing – review & editing (supporting).

This work was funded by the Engineering and Physical Sciences Research Council (grant No. EP/R034575/1 awarded to K.M. Briggs and W. Powrie) and the Royal Academy of Engineering (grant No. RCSRF1920\10\65 awarded to K.M. Briggs).

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

The data presented in this paper are available online via the University of Bath institutional repository (Briggs 2024) and may be accessed at https://doi.org/10.15125/BATH-01380