Pore Pressure
This section presents the main results from pore pressure measurements. All piezometers were logged continuously over a total period of approximately 8 months. To establish representative reference values, the piezometers were installed approximately 4 weeks before the drilling in Area A commenced. Data were logged at 1-h intervals during the whole test period and changed to one per day when all drilling was completed.
Monitoring data show that all the drilling methods caused excess pore pressures in the surrounding clay. The observed response on the pore pressure was generally much the same with all methods. However, Method 2 in Area B (DTH air hammer) and Method 5 in Area E (top hammer and concentric drill bit) resulted in significantly higher excess pore pressures than the other methods. Table
3 gives a summary of maximum excess pore pressures registered at each test area. The largest observed excess pore pressure was 70 kPa in PZ E9 at a depth of 10 m while drilling of anchor (casing) E107 in Area E with a minimum distance of approximately 1.1 m to the piezometer. The main reason for the large excess pressure was likely the high penetration rate, approximately
, combined with pressurized water flushing (Table
2). This resulted in the highest ratio of maximum excess pore pressure (
) to the effective overburden stress (
) of all the piezometers with a value of 1.14 (PZ E9). The minor changes in PZ E9-2 and PZ E10 (
between 0 and 5 kPa) was most likely due to the much greater distance between the casings and the piezometers as well as the relatively small dimension of the anchors (
).
Measurements in Area A showed relative moderate changes in pore pressures with a maximum value, in PZ A9. The results were likely affected by the relatively small dimensions of the drill bit () and the change in inclination from 45° to 56° for the last six anchors, thereby increasing the theoretical minimum distance between the piezometers and the anchors.
Fig.
6 shows changes in pore pressure (
) with respect to time during drilling in Area B [Fig.
6(a)], Area C Fig.
6(b)], and Area D [Fig.
6(c)], respectively. Time of drilling for each individual anchor is indicated with gray bars in the figures. Fig.
6(a) show that drilling of Anchor B104 caused an immediate excess pore pressure of approximately 60 kPa in PZ B10 at a depth of 17 m, while PZ B9-2 at a depth of 4.5 m showed only minor change (
). No data were available from Piezometer B9, which was out of function during the field tests. Drilling of Anchor B104 also caused an increase of approximately 13 kPa in PZ A9 (10 m depth) and 4 kPa in PZ A10 (17 m depth) in Area A, at a distance of around 30 m from Anchor B104. The excess pressures were most likely caused by flushing with compressed air [1,200–1,500 kPa (12–15 bar)] when drilling through the sand/moraine layer above bedrock. Small outbursts of air, water, and remolded clay were observed up along the outside of the casing as well as along the previously installed Anchor Rods A104 and A103 in Area A. This shows that drilling with compressed air caused pneumatic fracturing, not only along the casing wall but through the moraine layer.
Despite the closer proximity to PZ B10, drilling of Anchor B103 and B102 had less impact on the excess pressures than Anchor B104. The results indicate that some of the flushing air evacuated through the moraine and joints/fissures in bedrock and up into the casing for Anchor B104, rather than building pressures in the ground as with Anchor B104. This mechanism was also observed for some of the other anchors in Area B.
Drilling of Anchors B101, B108, and B107 reduced the excess pore pressure in PZ B10 with approximately 15 kPa in total. This reduction could be caused by groundwater that was sucked into the casings with the backflow (drill cuttings) when drilling into bedrock. Based on visual observations, the amount of water is roughly estimated to be between 20 and .
Piezometer B9-2 at a depth of 4.5 m showed insignificant changes, with a maximum accumulated excess pore pressure of approximately 4 kPa during drilling in Area B. The longer distance between PZ B9-2 and the anchors compared to PZ B10, combined with lower soil stress at shallow depth, may explain this difference in response.
Fig.
6(b) shows that drilling with the DTH water hammer in Area C resulted in considerable lower excess pore pressures in the surrounding clay compared to Areas B and E. The major difference is reasonable, considering the lower penetration rate when drilling through the soft clay in Area C (Table.
2). Drilling of the first four anchors in Area C (C104 to C101) had minor influence on the piezometers except PZ C10, which showed an accumulated increase to a maximum value of 18 kPa after drilling of Anchors C102 and C101. The excess pressure then decreased and was almost unaffected during drilling of Anchors C108 to C105 because of the greater distance to the casing. PZ C9 showed, however, excess pressure of approximately 18 kPa while drilling of Anchors C107 and C106 with a minimum distance of approximately 1.1 m from the casings. Piezometer C9-2 at a depth of 4.5 m showed an approximately 10-kPa increase in pore pressure during drilling of Anchor C105, even with a minimum distance of approximately 5 m to the casing. This was—two to three times higher compared to the piezometers at a depth of 4.5 m in the other test areas. The difference from the other drilling methods could be related to the significantly higher water pressures and flow rates used during drilling in clay [
at 6,000–8,000 kPa (60–80 bar) from the water pump]. The flushing might have caused some hydraulic fractures in the upper part of the clay, extending the influence zone.
The measurements in Area D (Method 1) are not directly comparable with those obtained using the other methods since the test was aborted after drilling of the first two anchors (casings). The results are, however, interesting with respect to the installation effects from drilling. Fig.
6(c) shows that drilling of Anchor D104 resulted in a pore pressure reduction of approximately 3 kPa in PZ D10 (17 m depth) and PZ E10 (14.8 m depth), which decreased to approximately 5 kPa during the following 24 h. The pore pressure in PZ D10 reduced further to a minimum value of approximately 17 kPa right after drilling of Anchor D103, still being approximately 10 kPa below the reference pressure 4 days later. Piezometer D9 (10 m depth) showed a temporary pressure reduction of approximately 2 kPa during drilling of Anchor D103 before it increased evenly to a maximum excess pressure of approximately 8 kPa in the following 4 days. Some minor temporary increase in pressure between 2 and 4 kPa was also observed in PZ E10 and E11 during drilling.
The pore pressure reductions observed in both Area D and E were likely caused by some minutes of air flushing during drilling in Area D when trying to improve the transport of drill cuttings and penetrate through a layer of dense sand/moraine encountered at a soil depth of approximately 18 m. Flushing with air probably caused a so-called air-lift pump effect (
Behringer 1930;
Kato et al. 1975) in front of the drill bit when the water and drill cuttings inside the casing was flushed up to the surface by pressurized air. This caused a lower pressure inside the casing compared to the pore pressure in the surrounding sand/moraine, creating a gradient, i.e., flow of groundwater, toward the drill bit like a pumping well. Owing to the higher permeability (hydraulic conductivity) in the sand/moraine layer compared to the clay, the effects of air flushing were noticeable in Area E over 20 m from the anchors. It is reasonable to assume that the recovery time for the pore pressure was increased since the two casings in Area D were not filled manually with water again after the drilling was aborted. The amount of drill cuttings generated from Anchors D104 and D103 indicates that the drilling formed a cavity, i.e., volume loss, around the casings in the moraine layer. The volume loss is likely the main reason for the pressure reduction in PZ D10, caused by suction in the clay above the cavity in the sand/moraine layer.
Fig.
7 shows dissipation of excess pore pressures against time for some piezometers in each test area. Most of the excess pore pressures dissipated rapidly after drilling was completed. This behavior coincides well with results from pile driving in clay reported by, among others, Li et al. (
2019), Karlsrud (
2012), and Edstam and Küllingsjö (
2010). A similar dissipation trend is also observed with pile drilling (e.g.,
Ahlund and Ögren 2016;
Veslegard et al. 2015). In Area E only 10 kPa of the maximum excess pressure of 70 kPa remained 2 days after drilling of Anchor E107, and approximately 30 days later it had completely dissipated. Pulling of the casings in Area E might have made it easier for the excess pressure to dissipate to the grouted borehole, thereby speeding up the process compared to the other test areas. PZ B10 showed a slower trend with approximately 8 kPa excess pressure remaining 150 days after drilling was completed in Area B. The longer dissipation time indicate a more severe influence from drilling on the surrounding clay.
Fig.
8 presents the maximum change in pore pressure (
) at different depths in each test area against the distance from the anchors (casings). Fig.
8(b) shows results against the ratio of radial distance from casing
to the radius of casing
, and Fig.
8(b) shows results against metric radial distance. The data represent the maximum values obtained from the drilling of single anchors with each method. For comparison, typical excess pore pressure curves are also shown for a driven closed-ended pile, based on the strain path method (SPM) theory (
Baligh 1985) and the procedure described by Karlsrud (
2012). The estimated excess pore pressure is for a single pile with a diameter equal to the casings in Areas B and C (
), at depths of 10 and 17 m below ground level. Driven closed-ended piles are here considered to represent a worst-case scenario in terms of generating soil displacements and excess pore pressures in the surrounding ground. Drilling should ideally represent the opposite, i.e., a method by which the soil volume of the casing being installed is removed, thereby limiting soil displacements. Fig.
8 includes some results from a recent field test with the jacking of open-ended concrete piles (
,
) in soft organic clay (
Li et al. 2019). It also include results reported by Ahlund and Ögren (
2016) from a field test comparing air- and water-driven DTH hammers for the drilling of piles (
). The effect of air flushing was much higher than that of water flushing, but both methods are in the lower range compared to the results presented in this paper.
Results from this field trial show that drilling may cause much higher excess pressures than previously reported for driven closed-ended piles. The main reason for the excess pressures is likely related to the high penetration rate used for all drilling methods except in Area C (Table
2). The results indicate that the flushing process enhances the effect on soil displacements.
Ground Settlements
All 40 settlement anchors at the test site were monitored over a period of approximately 9 months. The first measurements (baseline) were made September 9, 2013, 10 days before the field tests started in Area A. Settlements were measured more frequently during the field test (between one and three times a week) so that any immediate effects from drilling could be documented. The frequency was reduced to every 4–5 weeks after drilling was completed in Area C. Accuracy of the measurements was specified as to 2 mm by the surveying company.
Fig.
9 presents the vertical ground settlements (
) measured on Settlement Anchor 4 (Fig.
4) in each test area from September 9, 2013, to January 7, 2014. The gray bars show the time when drilling took place in each area. The monitoring data show that drilling in Area B (Method 2) and, likely, in Area D (Method 4) caused almost immediate settlements between 2 and 7 mm of all settlement anchors over the entire test site. Results of the pore pressure measurements and observations during drilling clearly indicate that these settlements were caused by air flushing used with Methods 2 and 4. The settlements in Areas B and D can be explained by local erosion and volume loss around the casings, but such volume loss can hardly explain the influence on the other test areas. None of the other drilling methods caused similar short-term settlements or effects in other areas.
After drilling in Area B was completed, subsequent measurements until June 2014 showed 2- to 6-mm settlements over a period of 3 months (between January 7 and April 4, 2014). During this period some of the remaining excess pore pressure (approximately 5–10 kPa) in PZ B10 dissipated, indicating reconsolidation of remolded clay. There were no significant further settlements in Area A, C, D, or E during this period, but there were indications of heave (1–3 mm) in some points in Areas C, D, and E. Freezing of the topsoil during the winter may have led to undesired uplift on the settlement anchors. Mitigating measures in terms of insulating the outer pipe (casing) above the ground surface and filling frost inhibiting liquid between the outer pipe and the settlement anchors were carried out. Despite these efforts, the anchors seem to have experienced some frost-induced uplift.
Fig.
10 shows the resulting ground settlement profiles for Settlement Anchors 1–6 in each test area. The settlements in Area B clearly stand out compared to the other areas with a maximum value of 12 mm (Anchor B3). The results clearly show larger settlements in the area above where the anchors hit bedrock, i.e., Settlement Anchors 2–5.