Applications of Explosive Compaction for Tailings Volume Reduction

3- Part (2): Applications of Explosive Compaction for Tailings Volume Reduction :

3.1-introduction :

The method is expected to be effective where the dominant tailings components comprise granular, low to non-plastic silt/sand mixtures. The paper discusses the mechanisms associated with explosive compaction, factors influencing design, and typical ground response (vibrations, settlements, pore water pressures) observed.

The sequential detonation of below ground explosives placed in cased boreholes has been widely used in civil and mining engineering for over 70 years. Previous applications have included foundation compaction of predominantly granular soils for earth dams, bridges, buildings and offshore oil structures (Gohl et al, 2000). This has been largely driven by the need to increase seismic or static liquefaction resistance in granular soils. A potential new application of explosive compaction (EC) is discussed in the present paper with respect to volume reduction of previously impounded mine tailings in tailings ponds.

3.2-Who does it work :

EC involves placing single or multiple (decked) charges in a borehole drilled over the depth of soil to be densified . Several charges are fired sequentially, with delays selected to minimize offsite vibrations and also to promote cyclic loading of the subsoil. In general, the process is repeated a number of times to cause progressive soil compaction. Pore pressures generated in saturated soils following each blast sequence are allowed to dissipate before further blasts are carried out.

3.3-Effectiveness on the soil improvement

In saturated ground, the energy released by an explosive detonation causes liquefaction of the soil close to the blast point and causes cyclic shear straining of the soil. This process increases pore water pressures and provided strain amplitudes and number of strain cycles are sufficient, the soil mass liquefies. Liquefaction of the soil followed by time-dependent dissipation of the water pressures causes reconsolidation within the soil mass. This re-consolidation happens within hours to days following blasting, depending on the permeability of the subsoil and drainage boundary conditions, and is reflected by release of large volumes of water at the ground surface or up blast casings. “Short term” volume change is also caused by passage of the blast induced shock front through the soil mass. At close distances from a charge detonation, the hydro-dynamic pressures are large enough to cause compression of the soil water system even though the bulk compressibility of the system is relatively small.

The EC method is expected to be most effective where the tailings comprise granular, low to non-plastic, silt/sand mixtures. Experience has indicated that the degree of volume change obtained by blasting depends on the initial density of the subsoil, the total amount of explosive used per unit volume of soil (charge density), and the geometry of the blast pattern. The density of initially loose deposits typically increases considerably to relative densities of up to 70-80% for higher charge densities. Soils with initial relative densities of 60 to 70% can only be densified by a small amount. Thus, the initial density of the tailings controls how much settlement can be expected from the EC process.

3.4-Case study:

2 EC test programs (Sites 1 and 2) in loose mine tailings (sand/silt mixtures) at a tailings dam site in Northern Ontario. The principal author was also involved in initial design of blasting trials for the Canadian Liquefaction Experiment, referred to as Site 3. The latter was carried out at Syncrude’s J-pit outside of Fort McMurray, Alberta (Wride et al, 2000). Other EC data reported in the engineering literature concerning compaction of silt/sand mine tailings has also been reviewed (Klohn et al, 1981; Handford, 1988; Gohl et al, 2000), referred to as Sites 4, 5 and 6, respectively.

The basic geotechnical characteristics of the above mine tailings materials and the charge density used for each EC project are summarized in Table 1. Charge density is defined as the sum of charge weights used in blast holes located within the interior of the test area to be densified plus ½ the sum of charge weights for blast holes located around the perimeter of the area, all divided by the total volume of soil. This definition of charge density accounts approximately for blast energy radiated away from the zone to be densified for perimeter blast holes.

3.5-Post-EC Settlements :

Table 1 indicates that a range of loose silt/sand tailings mixtures have been successfully blast densified, achieving average settlements over the loose layer thickness of up to about 10%. Maximum settlements are achieved using higher charge densities, which implies the use of relatively close blast hole spacing, multiple decks of explosives, and repetitive blast sequences. Post-EC settlements averaging 2 to 4% of the loose layer thickness over the test area have been achieved using relatively low charge densities of around 5 gm/m. The latter have been based on larger blast hole spacing, typically in the range of 10 to 20 m

Settlements decay with distance from a blast hole, forming a bowl-shaped depression around each blast hole as pore pressures generated by blasting dissipate. Since shear strains and hydrodynamic blast pressures attenuate with distance from a charge detonation, this settlement reduction with distance should be expected. Depending on the spacing and pattern of blast holes fired, these depressions gradually level out with successive blast sequences, corresponding to the effects of adjacent blasts causing additional cyclic straining within a partially settled area.

The effect of several blast sequences on local settlement within the center of a test EC area (Sites 1,2 and 4) is shown in Figure 1, showing the progressive tailings shakedown with each blast series. For Sites 1 and 2, each blast sequence involved the sequential detonation of several charges. For Site 4, only 1 charge detonation was used for each blast series. The data indicate that initial soil liquefaction following the first blast series typically caused vertical settlements equal to 5 to 6% of the loose layer thick-ness within a few meters of a blast hole. This corresponds to zones of largest shear strain in close proximity to the blast hole. Several blast series result in large vertical settlements around the blast hole, in the range of 12 to 14% of the loose layer thickness. These localized settlements typically exceed the average settlement over the test area.

Ishihara and Yoshimine (1992) suggest that post-liquefaction settlements in loose sands (initial relative densities of 40 to 50%) equal to 3.5 to 4.5% of the loose layer thickness should be expected following earthquake shaking. Larger earthquake settlements should be expected for looser sands. For blast loading, the effects of large shear strain amplitude, number of strain cycles, and hydrodynamic blast pressures apparently leads to larger post-liquefaction

settlements locally around a blast hole compared to that caused by earthquake loading.

Experience with blasting in loose sands and silts suggests that the zone of significant settlement is approximately ½ of the radius of liquefaction. The looser the soil, the broader the radius of liquefaction and radius of significant settlement. Field trials are typically carried out to confirm the amount of actual settlement achieved since this will depend on the initial densities and other geotechnical properties of the tailings, the size of charge detonation per delay, the number of charges detonated sequentially, and the depths of charge burial.

3.6-Pore Pressure Response :

Charge detonations lead to stress wave propagation away from a blast hole, producing dynamic changes in mean stress and shearing stress within the soil medium. The changes in mean stress are typically very high (MPa range) within a few meters of a blast hole (for the typical charge sizes used by the authors in EC projects), leading to transient hydrodynamic pore pressures. The shearing stresses developed are limited by the undrained strength of the tailings materials, leading to permanent shear strain offsets in a soil element following passage of the shock front. The shear strain pulses are considered primarily responsible for build-up in residual pore water pres-sure. In close proximity to a blast hole, high amplitudes of dynamic mean stress may also lead to residual pore pressure build-up if the soil-water compressibility is such that soil skeleton volume change occurs. This would also lead to residual pore pressure buildup. High pressure, load-unload hydro-static compression tests on saturated soil elements would be required to confirm this effect.

He typical pattern of transient pore pressure pulses and the gradual buildup of residual pore pressures are seen in Figure 2 for a multiple hole detonation sequence at Site 2. The rate of data acquisition used was not rapid enough to accurately capture all the hydrodynamic pressure pulses, but the general trend of gradual buildup in residual pore pressure is evident. The distance of the nearest blast hole to the point of pore pressure measurement is considered far enough that the effects of dynamic mean stress on residual pore pressure change is minor. Pore pres-sure build-up resulting from shear straining is considered to be the dominant factor.

The amount of residual pore pressure buildup at a reasonable distance from a blast hole (neglecting hydrodynamic pressure effects) depends primarily on shear strain amplitude and number of strain cycles (Dobry et al, 1982). These in turn will depend on charge weight per delay, distance from a charge detonation, and number of charges detonated sequentially. In the extreme, if the residual pore pressures over and above the pre-blast static water pressures in the ground equal the initial vertical effective stresses, then a condition of soil liquefaction results. The residual excess pore pressures divided by the initial vertical effective stress is defined as a pore pressure ratio, PPR. Figure 2. Pore pressures versus time during sequential blasting at Site 2, showing peak hydrodynamic pressures and residual pore pressure buildup. The rate of data acquisition was not high enough to accurately capture all the hydrodynamic pres-sure –pulses. train pulses.

3.7-Parameters :

A plot of PPR versus scaled hypo central distance (R) and average charge weight (kg) per delay (W) for Sites 1 to 4 is shown in Figure 3. Reliable pore pressure data were not available for Sites 5 & 6.

Here the hypo central distance R refers to the distance (meters) between the pore pressure measurement point and the nearest charge detonation. The data include both single and multiple charge detonations. The data indicate that PPR increases with in-creasing charge weight per delay, decreasing distance between a blast point, and increasing number of charge detonations. The data indicate that a radius of liquefaction (where PPR ≥ 0.9) can be estimated as equal to αW^0.33 in loose tailings material with α values in the range of 3 to 9. For design, the available data suggest that an average radius of liquefaction equal to 6W^0.33 may be presumed, assuming multiple charge detonations are employed.

Data from cyclic, strain controlled triaxial tests on loose sands (relative densities of about 45%) indicate that it is necessary to achieve peak shear strains during one strain cycle of about 0.3% to produce soil liquefaction after 4 to 10 strain pulses (Dobry et al (1982). Thus, design estimates of maximum shear strain versus distance from a charge detonation may be used in advance of blasting field tests to estimate maximum radii of liquefaction around blast holes. The procedures used are described in Section 3 on Blast Design, involving the use of nonlinear blast analysis.

3.8-Vibration control:

Induced vibrations on nearby structures need to be controlled where blast densication is carried out in developed areas. Blasting within 30±40 m of existing structures requires a reduction in the charge weights per deck (involving a reduction in blast hole spacing), and in the number of holes detonated at any one time. Also, when blasting is carried out on or adjacent to slopes, blast patterns are adjusted to restrict the zone of residual pore water pressure build-up and minimize the risk of slope instability. For the above reasons, the number of charges detonated .

Sequentially is often restricted to minimize the duration of shaking .

Longer-duration shaking causes more damage to structures and increases residual pore water pressure build-up. Individual charge delays are also selected so that destructive interference in the frequency range of interest occurs between ground waves from the sequence of blasts. Design of the appropriate charge delays between adjacent decks in each borehole and between adjacent boreholes is carried out using the following process:

(a) Ground vibration patterns (peak particle velocities and frequency content are determined at a particular location of concern remote from the blast point due to a single charge.

This is best done using old measurements, but can also be carried out theoretically.

(b) The frequency range of potentially damaging vibrations is selected based on structural vibration theory or other considerations.

(c) The effects of sequential charge detonation from a decked array of boreholes are assessed by a simple linear combination of the single charge wave trains in which time delays between decks and between adjacent boreholes are varied. Optimum blast delays are then determined to minimize the peak particle velocity or, alternatively, the vibrational energy content in the frequency range of interest.

Based on vibration measurements recorded during detonation of a single charge at various distances from a blast source at an alluvial site in Japan, the linear wave superposition model described above was applied to compute the likely peak particle velocities (PPV) at the ground surface for a particular direction resulting from multiple detonations. The computed and measured peak velocities are plotted against each other in Fig. 3 and indicate that the use of linear combination of wave motions (incorporating appropriate time-shifts in the waveforms based

on the prescribed detonation sequence and waveform scaling depending on charge weight ± distance relationships determined for the site) generally leads to a conservative over-prediction of PPV. This is particularly true for small source ± site distances, where non-linear effects caused by soil liquefaction around a blast point would be expected to reduce near old motions.

Blast hole layout and detonation sequencing Blast patterns generally use a staggered rectangular grid of boreholes at spacing of 4±9 m. Staggering is used to provide a pattern of two (or more) passes within a uniform grid. Boreholes are drilled over the full depth of soil deposit to be denied, and 75±100 mm diameter plastic casing installed (this casing size is convenient with the drills normally used). The casing is then loaded with explosive at one or more levels in the borehole (decks). A series of boreholes, each containing one or more decks, is then sequentially detonated. The number of blast holes detonated in any shot depends on vibration control considerations and on concerns about the effect of liquefaction and settlement on adjacent slopes and structures.

3.9-CONCLUSIONS

Explosive compaction has been used effectively for a wide variety of civil and mining engineering projects, primarily with respect to improving foundation soil resistance to static and seismic liquefaction. A new application of the EC process is proposed to re-duce the volume of previously impounded mine tailings, thereby increasing storage capacity within the tailings pond.

Previous experience with EC in non-plastic silt/sand tailings materials indicates that significant volume change can be induced by blasting in saturated materials. The amount of volume change largely depends on charge density, which is governed by blast hole spacing and the charge weights used in each hole. The geometry of the blast pattern further influences the uniformity of the compaction process. Data collected from previous EC field trials in tailings msign of the EC process. Estimates of post-EC settlement as a function of tailings depth and soil properties, blast hole layouts and charge densities are facilitated by application of nonlinear blast analysis.