The Greater Vancouver Water District (GVWD) provides a reliable source of drinking water to over 2M people in its member municipalities. The Seymour Falls dam is a key element in the GVWD’s network of three watersheds. This 45-year old composite earthfill embankment and concrete slab and buttress dam is in excellent condition but does not meet current seismic design standards, and upgrades to meet the requirements of the Maximum Credible Earthquake (MCE) are necessary. In part one of this article, published in the August 2006 edition of International Water Power and DamConstruction, a description of the dam and its geologic setting was provided. The paper discussed the dam safety and seismic hazard assessment, and also described the earthfill design criteria as well as the foundation and liquefaction assessment. Part two describes the earthfill section remedial measures and implementation of ground improvement. All relevant figures are republished from part one.

Earthfill section remedial measures

Site investigations and liquefaction assessment at Seymour Falls dam confirmed that ground improvement was required as part of the seismic upgrade. Several techniques were reviewed including vibro-replacement, compaction piles, dynamic compaction, soil replacement and explosive compaction. The selection of the best option was based on considerations of ground conditions, operational impacts, degree of improvement required and environmental considerations.

Upgrade of the existing earthfill structure was considered but rejected due to the requirement of keeping the existing dam and reservoir operational during the work, as the Seymour reservoir plays a critical role in maintaining the GVWD water supply. The preferred upgrade option selected was to build a new dam downstream on an improved foundation. The new dam, stable under the design seismic conditions, will also provide a downstream stabilising buttress to the existing structure, mitigating future deformation of the existing dam, although the new dam is designed to stand alone even with full breach of the existing dam core. Included in the design is a connection of the new dam core to the existing core and reservoir blankets (Figure 5).

Improvement of the new dam foundation required compaction of the ground downstream of the existing dam to prevent earthquake induced liquefaction in zones critical to the new dam stability. Explosive compaction (EC) and dynamic compaction (DC) were considered techniques capable of providing the required improvement but it was necessary to assess the relative effectiveness of each technique given the unique ground conditions (Cougar Creek fan) combined with the restrictive operating conditions of working in the shadow of an operating reservoir and dam.

Based on review of options for foundation ground improvement, a conceptual design to densify the new dam foundation was prepared combining EC and DC. The next step was to assess the feasibility of these measures in the very bouldery ground of the Cougar Creek fan.

Explosive compaction method

Compaction by the EC method involves detonating explosive charges in controlled sequences to precipitate liquefaction in limited sections of the dam foundation, essentially simulating the effects of an earthquake in liquefaction susceptible ground, combined with effects of compression, shearing and volumetric strains. Following blast induced liquefaction, the soil mass re-consolidates as pore water pressures dissipate. By sequencing detonations using microdelays (50 to 100ms) between adjacent holes, blast pressures propagate across a given elevation (deck), so that, effectively, the whole deck elevation is liquefied at one time, which should theoretically enhance the reconsolidation and settlement of that deck by creating a larger settlement/collapse zone thus limiting the amount of arching above. For Seymour, timing delays were set at greater than 25ms so that the blast induced vibrations, which could adversely affect nearby structures, did not stack as they propagated away from the detonating charges. As well, the sequence of charge detonations began closest to the dam, proliferating in a direction away from the dam, providing greater distance to critical structures from subsequent detonations.

After a set of charges at a given deck elevation are detonated, sufficient delay time (200 to 600ms) is then provided to allow pore pressures to begin to dissipate (and hence ground to strain harden) before starting a second detonation sequence at a higher elevation. In this manner, decks within a given panel are liquefied and compacted. Typically, depending on the depth of the blast hole, two to three decks of main charges, plus one or two higher elevation decks of lighter pilot charges may be detonated within a blast panel, with three or more passes in a given panel (a pass is a set of blast holes with multiple charges and decks).

A significant concern in using EC at Seymour Falls dam was the bouldery nature of the ground. The ground conditions, through several investigation phases, proved difficult and harsh on equipment. Drilling blast holes for an extensive EC programme in ground with many boulders greater than 5m diameter was potentially problematic. One of the working tenets for using EC is that it is a cheap method, in terms of densification per volume of soil. This assumption is based on reasonable drilling cost; however, this was not the case at Seymour Falls dam, where the difficult bouldery ground had the potential to make the EC method prohibitively expensive.

Dynamic compaction method

Ground improvement using DC is a well recognised and widely used technique that involves the repetitive dropping of large weights in a grid pattern, with the direct application of high-energy causing the compaction and strengthening of the ground. The design of the drop pattern and average energy application depends on the soil types to be improved and the desired depth of improvement. Typically, a large steel tamper is dropped several times from heights of up to 30m at a given grid or drop point, creating craters of up to 3m depth, with one to three offset grid patterns in the ‘high-energy’ phase, followed by infilling of the craters, and an ironing phase with a lighter tamper dropped from a lower height. The ironing phase is designed to compact the crater backfill and apply a final compaction of near surface ground. The ironing phase differs from the high-energy drops as it is applied across the entire surface of the panel as opposed to discrete grid drop points.

EC and DC trial programme

To assist in the final design of the work, a trial compaction programme was conducted in 1998, using EC from depths of about 10-20m and 20-30m, and DC with an input energy of 420 tonne meters on a 30m by 30m panel within the footprint of the new dam (Figure 3a and 3b). The trial DC programme achieved (N1)60-CS > 25 blows/0.3m down to at least 10m. The effectiveness of the DC dropped off significantly below 10m depth. The EC trial achieved (N1)60-CS 1) > 20 blows per 0.3m in zones in the range 10m to 20m depth below ground level but gave little or no improvement below about 25m depth. Test hole drilling between the application of the EC and DC trials indicated that some arching had occurred in the bouldery ground above the EC zone. Post trial FLAC analyses showed that the overall effect of the EC/DC programme was sufficient to achieve a stable dam post MCE.

The following site-specific equations were developed during the trial programme. The equations, in general, are empirical modifications to published forms of equations by Narin Van Court.

Peak Particle Velocity PPV (mm)(RH/W1/2)-1.5 x 2500

Regional Pore Pressure Ratio (PPR)1.25 – 0.2 (Rp/W1/3)

PPR in Blast Pattern1.35 – 0.15 (Rp/W1/3)

where: RH is hypocentral distance (m), Rp is plan distance (m), W is maximum charge weight per delay, Kg, PPR is the ratio of increase in pore water pressure to pre-blast in-situ vertical effective stress. PPR = 1.0 represents complete liquefaction.

The EC trial also demonstrated that Barber downhole hammer drilling techniques could be effective for advancing blast holes in these ground conditions. However, the trial further demonstrated the general difficulties of advancing and interpreting becker penetration test results (BPT), but provided evidence that carefully advanced and interpreted (standard penetration test) SPT tests could be conducted in mud-rotary drillholes using a method of calculating the blow count as 4x the lowest three consecutive blows recorded over 25mm increments, in the penetration interval 0.15 to 0.45m. This value was then corrected to (N1)60-CS, and SPT testing was adopted as a potential means of verification of EC and DC performance.

Shear wave testing (downhole and seismic refraction surveys) were conducted during the field trial as a check for improvement. Seismic refraction, a relatively inexpensive and quick method for assessment of large areas, was limited by an inherent inability to detect slow velocity layers beneath high velocity layers. Nevertheless, seismic refraction was able to detect near-surface improvements below the thin slow surface layer. However the inability to measure deeper slower layers precluded use of seismic refraction as a verification method. Downhole seismic and crosshole seismic tests were more effective in producing seismic profiles with depth; however, results can be suspect where large quantities of grout are used during drilling, and in the difficult ground conditions, downhole or crosshole seismic was deemed uneconomic as a means of verification of improvement.

Final design of ground improvements

Using experience from the trial programme, the ground improvement for the new dam was designed as follows:

• Pre-excavate the upper 10m of the site.

• Conduct EC from 10m to 20m depth below the excavation base.

• After EC, conduct DC on the base of excavation from 0 to 10m depth.

The sequence provides for the mechanical compaction of the DC to follow behind the primarily liquefaction induced settlement of the EC, such that any arching induced by the EC could be broken down by the subsequent DC.

The EC/DC was designed to achieve the following;

• 2% to 5% settlement of the EC compacted zone, i.e. an average of about 200mm to 500mm. An additional 2% to 5% settlement of the DC zone was expected for an overall settlement of about 500mm to 1000mm. These values were estimated from the settlement achieved during the trial programme.

• Liquefy the EC treated zone by achieving a pore pressure rise ratio of 0.9 or higher in the active blast panel.

• Stay within site peak particle velocity (PPV) limits which vary from 13mm/sec at the sensitive fish hatchery to 25mm/sec at the crest of the concrete dam, 75mm/sec at the chlorination plant and concrete gravity wall, 120mm/sec at the nearby 90 inch (228.6cm) water main (Main #2) to 150mm/sec for the toe of the existing earth dam.

• Keep within settlement limits at structures; generally settlement of less than 50mm was specified.

• Keep within pore pressure rise and water body overpressure restrictions of key points; pore pressure rise was controlled to maintain slope stability; overpressure was to be less than 50kPa at spawning pools in the Seymour river to protect the fish.

• Meet restrictions for chemical residuals at surface discharge to meet regulatory requirements for aquatic life (e.g., <25.7mg/l ammonia, <200mg/l nitrate).

The nearest EC blast panel is offset about 300m from the fish hatchery, 18m from the Main #2 and the chlorination building, and about 8m from the earth dam toe. The nearest DC drop points were at similar distances, except DC extended to within 8m of Main #2 and the chlorination building.

The footprint of the improved ground is located within 8m of the existing embankment, gravity wall, chlorination building and Main #2, all critical structures. To maintain settlement within safe limits, predictions of settlement were prepared, based on measured performance during the trial EC programme. The horizontal distance to 50mm settlement was calculated by projecting a line 2V:1H from the base of the closest blast hole and adding 4m. The distance to 25mm settlement is calculated the same way but an additional 6m is added. The EC design provides for three passes over a series of nominally 30m by 30m panels over the depth interval 10m to 20m below excavated ground level and using a charge density (powder factor) of 0.10 to 0.15kg/m3 of treated ground.

Peak over pressure limits are assumed to be met if PPV limits are not exceeded. Nevertheless, real-time checking was conducted for water pressures at the hatchery pools. Chemical residuals concentrations were checked by field measurements.

DC was designed using similar settlement and PPV relationships but without the need for overpressure or chemical residual checks. For full-scale field compaction a minimum energy of 575 tonne/m per drop and minimum applied total heavy energy application of 550 tonne m/m2 were selected based primarily on Mayne 1984. The ironing pass energy was specified as an additional 90 tonne m/m2.

Based on the trial EC/DC programme, achievable SPT (N1)60-CS values were expected to be:

• 0 to 10mDC zone25 blows/0.3m

• 10 to 14mEC Zone18 blows/0.3m

• 14 to 20mEC Zone13 blows/0.3m

These values were specified as performance criteria for the DC work in the upper 10m of the improved ground. The FLAC and stability analyses indicated these values were generally required in the upper 10m for stability reasons. The settlement targets of 2% to 5% in the DC zone were considered secondary checks.

In the EC treatment zone (below 10m depth), the stability analyses indicated the improvement in (N1)60-CS values could fall below these values in some areas yet still satisfy the overall stability criteria. Combined with the difficulty and expense of obtaining SPT results in the bouldery ground of the Cougar Creek fan, the performance criteria for EC effectiveness was primarily evaluated by achieved settlements, which were expected to range between 2% and 5% of the vertical thickness of EC treated ground. SPT test results in the EC zone were specified as a secondary check but not as the primary performance criteria for the EC contractor.

Embankment dam design

The seismic upgrade, following the completion of ground improvement, includes construction of an approximately 30m high by 100m long extension to the concrete gravity retaining wall (GWE). The GWE will form the left abutment of the expanded earthfill dam, retaining the earthfill and forming a transition from the earthfill embankment founded on sediments to the concrete slab and buttress dam founded on bedrock. The GWE is to be built immediately downstream of the existing gravity retaining wall (GRW) and will include curtain grouting in the foundation bedrock as well as provisions for drainage measures from the earthfill/concrete interface and from the earthfill crest.

The zoned earthfill embankment will then be constructed on the newly densified ground. The dam section is a conventional impervious core embankment with wide granular shells and wide filters and drains. The impervious core will be extended to tie into the existing dam core thus providing a continuous low permeability barrier, connecting to the existing lake and land blankets. Zoned fills and the core tie-in are shown in Figure 5.

Embankment fills will include a granular shell wraparound at the new GWE, as well as regrading and landscaping of the work area back to original ground levels.

The new embankment centreline will be approximately 75m downstream of the existing dam centreline, and the new crest (including remnant of existing embankment) will be about 75m wide (measured down-valley). The toe varies from about 80 to 90m downstream of the existing embankment toe. The new dam is approximately 180m long (parallel to the existing dam), and was about 20m maximum height.

Dam performance will be monitored by a series of electric and pneumatic piezometers terminated in an instrument house with capability for datalogging and telemetry. A series of pressure relief wells are included along with a network of surface and foundation drains to quickly relieve earthquake induced pore pressures and to deal with any long-term increase in seepage due to earthquake damage to the impervious lake blanket.

Ground improvement implementation

Construction of this project began in February 2004, and is scheduled for completion in late 2006 or early 2007. As the work is still in progress, performance results are not fully available at this time. The following section therefore, presents discussion on the general works to date, construction methods and specifications implemented and an overview of the design compliance monitoring carried out to date.

The construction contract was awarded to Peter Kiewit Sons (PKS) in January 2004, with primary sub-contractors, Explosive Compaction Inc. (ECI) for the EC work, and Geopac, with Lampson, for the DC work. klohn-crippen supplied a full-time Resident Engineer with additional staff provided by Acres and GVWD for technical review and support as required to monitor or review construction activities and to provide quality assurance checks and testing. Project management was awarded to Pacific Liaicon Associates.

Construction timeline

The EC work began with site preparation, demolition and excavation in February 2004, with the first blast in May 2004, and the last blast in January 2005. The DC work began in December 2004, with excavation and panel preparation. DC and EC work continued concurrently in January 2005 until the EC work was completed. The DC was completed in April 2005 followed by construction of the concrete GWE completed in March 2006. The other miscellaneous concrete works began in May 2004 and will be complete by the end of August 2006.

Fill placement of the earthfill embankment began in May 2005 with the drains and drainage blanket placed concurrent with the GWE concrete. In April 2006, construction of the main embankment fills began and is expected to be completed by late 2006. The clay borrow deposit for the impervious core has a high in-situ moisture content, and therefore conditioning and placement of the core material has been challenging. Final landscaping and completion is expected in 2007.

Site preparation and monitoring

The site preparation included re-alignment of utilities around the work area, security fencing, clearing and grubbing, decommissioning of several piezometers in the new dam footprint, and demolition or moving of structures including a guardhouse, caretaker’s house, filtration pilot plant and miscellaneous roadways, drains and abandoned utilities.

Selected pre-existing piezometers (primarily standpipes), were decommissioned by grouting to surface. Although most of these piezometers were subsequently destroyed by the EC or DC work, it was deemed sound practice to first grout them to surface where feasible.

Disposal and laydown areas were grubbed and stripped and local/temporary access roads were constructed. Topsoil and organic debris from the dam site was stockpiled for future restoration. Site drainage was controlled by a series of ditches, directed to a large settling pond.

A new water supply and miscellaneous upgrades were provided to the Seymour salmon hatchery. Four monitoring wells were installed to allow monitoring of water quality and chemistry during the EC blasting, to ensure ammonia and nitrate levels in the groundwater remained within acceptable limits, as even with the new water supply, the hatchery still relied partially on groundwater for fish rearing ponds.

Six piezometer bundles were installed at the toe of the existing embankment including pneumatic, electric strain-gauge, and electric vibrating wire piezometers at two depths in each of three holes. The piezometers were used to monitor pore pressure changes during and porewater dissipation after EC blasting. In addition the piezometers provided data on groundwater levels during dewatering for the GWE excavation and the DC trench excavation.

A set of temporary monitoring points (TMPs) were established on critical structures, including the toe and crest of the existing embankment, the existing gravity wall, the concrete transition block, the concrete spillway piers and deck, the chlorination building, and on Main #2. These survey points were monitored before and after each blast to check that settlements were within acceptable tolerances. The TMPs are shown in Figure 7.

Eleven deep settlement posts (DSPs) were installed in the foundation footprint, with the base of the posts just within the top of the EC improvement zone. The purpose of the DSPs was to allow monitoring of settlement at depth, and comparison against settlement at surface. Greater settlement at depth than surface might be indicative of arching of soils above the EC zone, a real possibility considering the massive boulders present in the Cougar Creek fan.

Earthworks

Excavation of about 115,000m3 of primarily cobbly, bouldery, sand and gravel was conducted in advance of the EC work, with excavation and panel preparation generally proceeding from the highest elevation on the west abutment to the lowest elevation near the existing gravity wall (Figure 7). Excavated materials were stockpiled at one of several disposal or stockpile sites, separated into topsoil and organics; sand, gravel, cobbles and boulders less than approximately 600mm diameter, and boulders greater than about 900mm diameter. In addition, a stockpile of about 300 to 900mm diameter boulders was prepared.

The intent was to re-use the ‘intermediate’ boulders by crushing for riprap and rockfill and to re-process by crushing, screening, and possibly washing the 600mm minus bouldery granular fraction for dam shell and structural fills.

The excavation encountered several hundred boulders greater than 5m3 in volume, and dozens of boulders of greater than 20m. The largest boulder encountered was approximately 80m3. Boulders larger than 5mwere split by blasting, to allow handling and moving by the site fleet of Cat 345 excavators and Volvo A35 off-road trucks (photograph in Figure 6.

Additional excavation was conducted for DC panel preparation and for the GWE extension bringing the total excavation volume to approximately 190,000m3.

Explosive compaction

EC was conducted over approximately two thirds of the footprint of the new dam in fifteen panels. EC proceeded from the highest bench (furthest west) on the right abutment down to the lowest bench, closest to the gravity wall. The staging was partially selected to allow the first few blasts to occur in the least critical area, allowing the project team the opportunity to fine-tune the blasting programme.

Panels were about 800 to 1000m2 with slight variations in size but significant shape variations as shown in Figure 7. The panels delineated areas with similar proximity to structures or depth to bedrock to allow panel-specific blast patterns and charge densities to be specified. The specifications therefore prescribed the EC blast patterns for each panel, based on the predicted PPV and PPR results using the formulae derived during the trial programme.

The contract specified hole spacing, depth, charge weight and sequence of charge detonations were provided. By specifying the blast patterns, it was possible to predict and control vibrations and pore pressure response within and adjacent to panels. The contractor was required to perform the work such that the prescribed EC patterns were constructed and safely detonated. The contractor was required to demonstrate the specified loading and that all detonations had occurred in the specified sequence. This method shared the risk such that the owner bore the risk of meeting ground settlement limits while the contractor bore the risk of drilling and explosive handling. This proved to be successful in achieving the end result with minimal disputes.

For each panel, drill holes were drilled and cased in three passes, with number of holes per pass ranging from 6 to 23 (averaging about 18 holes per pass), with each hole between 10 and 23m depth (average about 20m). A typical blast hole layout is shown in Figure 9(a). Approximately 800 holes, totalling over 16,000m were drilled. Holes were drilled using a Rotex Oy Symmetrix downhole hammer system that proved very effective in the coarse bouldery ground. In general, drilling progressed at an average rate of about 18 lineal metres per rig per 12 hour shift, with two rigs working 6 days/week, 24 hours/day.

Drill holes were advanced using 168mm OD steel casing, then 100mm ID PVC casing was installed, and the steel casing removed. Casings were filled with water to facilitate installation, but were pumped dry prior to loading with explosives to minimise potential hydrodynamic shock between charges.

Holes were loaded with Iremite TX, a cap-sensitive emulsion manufactured using high strength microballoons for use in applications with high transient over pressures. Up to 30kg were placed in each deck with a minimum 2m vertical separation between the top and bottom of subsequent decks. Electronic detonators capable of 1ms delay precision were used. In total, over 33,000kg of explosive was detonated in 56 blasts.

Significant settlement was achieved during the EC work, meeting the project requirements, while staying within the restrictive limits of PPV and PPR. Detailed surveillance monitoring was conducted prior and following each blast including review of piezometric data, survey of monitoring points, mapping of tension/settlement cracks around blast panels, vibration monitoring at critical installations, water overpressure measurements in the salmon rearing ponds, and assessment of pore pressure response at the toe of the existing embankment.

The site-specific formulae for PPV and PPR developed during the trial programme were revised as the work proceeded and greater understanding of the site ground response was obtained. These revised formulae were extremely valuable in optimising the blast programme.

During the EC work, some problems with casing jacking and charge movement in the blast holes was encountered, with blast casing sometimes rocketing over 30m. The contractor together with the project team derived a unique means of anchoring the blast casing to prevent excessive movements while still allowing pressure release in the blast holes. A spring-loaded binder was affixed to the top of each casing, tied down by two 226.8kg mini lock blocks (see photograph in Figure 8. Following implementation, movement of the blast casings was limited.

All blasts were carefully instrumented and monitored to ensure detonation of all charges, including the use of Nonel tubing or coaxial cable monitoring to ensure detonations. A few misfires were suspected, but each was dealt with by an established protocol to investigate and, if necessary, detonate the charge. One sympathetic detonation occurred in adjacent charges and was attributed to faulty detonators; the suspect lot was removed from use.

Extensive monitoring of groundwater in the four monitoring wells installed downstream of the EC blast area mitigated water quality concerns. Trace quantities of nitrate and ammonia were occasionally noted in the well water, but were within specified limits.

Surface settlement at the EC panels was compared with that measured at the deep settlement posts. By the third pass in each panel, little difference was noted, leading to the conclusion that the effects of multiple passes in blast panels were beneficial in minimising the effect of arching in the bouldery ground above the blast zone. Average settlements met or exceeded the upper target of 5% in the EC zone.

Dynamic compaction

The dynamic compaction work was conducted over an area of about 22,000m2, covering most of the new dam footprint, as well as an area about 25 to 75m south of the new dam. A small margin of 5 to 8m width was maintained between the edge of the DC zone and the existing embankment and facilities.

The crane was a 350-ton DC-purpose built Lampson crane (LDC-350) capable of dropping the weight from a maximum of 30m. Three steel tampers were used, a 25 tonne tamper, an 18 tonne tamper, and a 16 tonne ironing tamper. The majority of the heavy energy application was done with the 25 tonne tamper, but the 18 tonne tamper was used in two locations where proximity to Main #2 or the chlorination building caused some concern over PPV levels, requiring a reduction in the compactive energy. The majority of the production drops were conducted from 26m height, resulting in a drop energy of 625 tonne metres/drop when using the primary tamper. The average energy application was approximately 550 tonne metres/m2 plus 90 tonne metres/m2 for ironing, as per design.

The DC was conducted in seven panels, several of which were subdivided further by the contractor for convenience. The DC energy was applied in three high or heavy-energy passes. The first and second passes were conducted on 10m by 10m grids offset by 5m from each other, with the third pass conducted on the intervening 5m grid points. Typically, 40 drops were conducted at each Phase 1 drop point, 22 drops at each Phase 2, and 10 drops at each Phase 3 location, roughly a reduction by half phase-to-phase. Craters were backfilled between phases and an overall pre- and post-DC survey was conducted to calculate settlement from DC, with correction made for the crater backfill. The main picture on p26 shows the DC purpose-built crane on a typical DC panel with a grid of drop-point craters.

Dewatering was required, through ditches, sumps or pumping wells, where groundwater was closer than 1m to surface.

Penetration testing during construction

During the EC and DC phases, penetration testing was conducted using conventional ASTM testing techniques using a mud rotary rig as follows:

• Tests were instrumented with an energy analyser to measure hammer efficiency.

• The drill bit has specially designed jets to direct mudflow upward and not at the base of the hole.

• SPT blows per 25mm were counted and summed where possible over the penetration range 150mm to 450mm in the normal manner.

• Where full penetration was not achieved an SPT value is calculated as four times the lowest blow count over consecutive 75mm in the 150mm to 450mm penetration range.

• Drilling was carefully advanced through cobbles and boulders and SPT started immediately when finer material was encountered.

In this way, typically 20m deep holes, with continuous SPT testing in the upper 10m and 1 to 1.5m spacing in the lower 10m, were conducted in three to five days. A total of 24 SPT drill holes were conducted, five during the EC programme and 19 during the DC programme.

SPT data were reduced in the normal manner (Youd 2000) to derive (N1)60-CS values for use in liquefaction assessment. An important component of the (N1)60-CS estimate is obtaining fines content. Good sample recovery was achieved by wrapping the split spoon core catcher in cling film. Using the noted method and discarding blows impacted by coarse particles, the SPT testing indicated that the EC and DC work have generally met performance criteria.

Conclusion

The seismic upgrade of the Seymour Falls dam is progressing as per design. The ground improvement phase has been successfully completed using the complementary techniques of explosive and dynamic compaction in very difficult ground conditions, including very bouldery ground. Strict operational constraints were implemented and met in order to maintain the integrity of nearby facilities and the ground improvement was completed with no service outages of the Seymour reservoir, and no measured impacts to the nearby downstream salmon hatchery.

The use of a trial programme to define the effectiveness and provide predictive formula for application in construction proved to be a critical and valuable tool. Innovative monitoring and verification methods including detailed analysis of SPT test results were developed.

By combining these two techniques potential limitations of both methods were overcome. The EC provided improvements at depths beyond the limits of conventional DC, and the DC effectively removed any arching of materials in the upper bouldery ground, a potential result of the EC programme.

The upgrades to the concrete works and the construction of the Embankment dam remain to be completed by 2007. The constructed new dam will bring the Seymour Falls dam facility to full compliance with the current Canadian Dam Association earthquake safety guidelines and meet Provincial standards.


Author Info:

The authors are Len Murray, P.E. P.Eng and Neil K. Singh, P.Eng of Klohn Crippen Berger Ltd., Vancouver, BC, Canada; and Frank Huber P.Eng and David Siu, P.Eng of Greater Vancouver Water District, Vancouver, BC, Canada.

Parts 1 and 2 of this article were printed with kind permission of the Association of State Dam Safety Officals. For further information, please visit www.damsafety.org.

Related Articles
Seismic upgrade of Seymour Falls dam