Unlocking Large Tonnage Filtered Tailings Stacks

Design philosophy and paradigm shift

Asked whether filtered tailings stacks call for a different design philosophy than conventional embankments, Joe Scalia, Jeronimo Covacevich, and Josh Rogers agreed that they do. Because a filtered stack is not a containment structure, it can be allowed to deform without releasing material, which opens the door to a performance-based, risk-informed design approach rather than a single fixed factor of safety. Jeronimo compared this to how waste rock is already managed on site, and both he and Josh emphasized defining the deformation a given site can tolerate and managing risk around that, rather than defaulting to the static-structure mindset used for slurry impoundments.

Long-term characterization testing

Chris Bareither described CSU's current work compacting tailings to anticipated field states and tracking how density changes under the loads expected at different positions in a stack over its life, in both saturated and unsaturated conditions, benchmarked against the critical state line (the dividing line between dense/dilative and loose/contractive behavior). He noted geochemical change is harder to test at lab scale and recommended involving a geochemist to identify which changes could most plausibly push material toward a brittle, contractive fabric.

Field trial plans for high-energy compaction

Josh Rogers outlined the consortium's approach to an upcoming field trial: get a data point on what's achievable under good conditions using impact rolling, rapid impact compaction, and possibly heavy vibrating plate (HVP) technology, benchmarked against conventional thin-lift roller compaction. Success criteria are achieving target densities, verifying that reliably through QA/QC, and confirming meaningful cost savings. The specific test plan will depend on the site chosen; the consortium is still evaluating candidate sites and welcomes suggestions. Joe Scalia added that preliminary cost-savings estimates (previously cited in the main talk as roughly 73% for impact rolling and 83% for rapid impact compaction) are directional, not universal — actual results will depend on particle size, plasticity, and lift thickness at a given site.

Warning signs before failure

Joe Scalia explained that a dense, dilative stack gives measurable warning before failure — deformation that shows up in InSAR, LiDAR, inclinometers, and extensometers — in contrast to a brittle material that loses strength rapidly with little warning. This supports an observational approach: defining in advance what deformation scenarios look like and what response they trigger.

In-pit disposal vs. filtered tailings stacks

Jeronimo Covacevich called the comparison somewhat unfair, since in-pit disposal has its own large capital costs, but said that on balance, a well-managed large filtered stack is expected to carry lower geotechnical risk than slurry tailings disposal, noting the answer is highly site- and tailings-specific.

Regulatory treatment and GISTM

Jeronimo noted that most jurisdictions don’t yet have specific requirements for non-conventional (filtered) tailings management, and cautioned that if filtered stacks end up regulated identically to slurry dams, that could be an “overkill” mismatch. GISTM currently requires consideration of alternative technologies during tailings facility design but does not set specific requirements for non-conventional management — he sees room for the industry to advocate for more tailored treatment.

Managing rainy seasons / monsoon climates

Chris Bareither pointed to the Greens Creek, Alaska facility as an example of managing high rainfall through surface smooth-rolling and aggressive surface-water management, plus options like temporarily placing tailings in off-spec areas to dry before final placement, tarping, and reducing active working area during wet periods. He framed this primarily as a surface-water and erosion-management problem rather than an infiltration problem. Joe Scalia added that this circles back to density: if saturation during part of the year is expected, achieving adequate density before that point becomes critical, along with strategies like adjusting placement moisture, dewatering harder, or shifting where in the stack work happens seasonally.

Design guidance for filtered tailings systems

Joe Scalia said no dedicated design guidance document exists yet beyond the consortium’s Study Manager’s Guide, and connected this to Jeronimo’s regulatory point — there’s a governance question of what legally constitutes a “dam” versus a filtered stack, referencing recent CDA language distinguishing the two, but no separate detailed design guidance currently exists for filtered tailings specifically.

Waste rock containment / co-disposal

Joe Scalia noted that waste rock containing tailings that are still liquefiable (insufficiently dense/dilative) essentially still functions as a dam. However, using rock in the lowest-deformation zone to manage erosion, drainage, or trafficability is a legitimate direction already used at some (lower-tonnage) operating sites. Co-disposal and commingling with waste rock were explicitly excluded from the white paper’s scope but were flagged as an adjacent, important area — large-tonnage operations will likely combine filtered tailings with waste rock and stream separation rather than run pure filtered tailings alone.

Drainage systems and saturation vs. stability

Joe Scalia reiterated that stability is governed by density, not simply by whether the material is saturated. Drainage design still matters — informed by consolidation, triaxial, and compression testing plus soil-water characteristic curves — to manage the phreatic surface and rate of rise, and drains should sit above any liner (not below) so the liner is actually effective. Heterogeneity in a filtered stack differs from the layered “varving” seen in slurry-deposited tailings, so drainage design needs to account for that difference.

Sheet piling and intelligent compaction

A question on sheet piling was deferred — Rob asked the questioner to email the consortium directly. On intelligent compaction, Rob explained the general concept (real-time feedback from the compactor indicating when target density has been reached) but noted the consortium, and Josh Rogers specifically, is still actively investigating this.

Dilative behavior, maximum stack height, and binders

Joe Scalia explained the risk that as a stack grows taller, increasing stress can push initially dense/dilative material across the critical state line into contractive behavior, depending on compressibility and fabric — so there may be a limiting height beyond which further densification is needed. Geochemical changes (e.g., dissolution increasing void ratio over time) pose a similar long-term risk. On binders, Chris Bareither cautioned that while they increase strength, they “lock in” the existing fabric — if that fabric later transitions toward contractive behavior under added stress, the bound material could still fail, and binders will fail at some threshold stress/condition, so they should be used carefully rather than as a shortcut.

Transportation

A tailings-transport question was noted as outside the white paper’s geotechnical scope. Rob pointed to a separate consortium paper on filtered tailings transportation presented at the Tailings Conference in Santiago as a better resource.

Carrara case study (uncompacted thick lifts, arid climate)

Responding to two related questions about the Carrara operation (Western Australia, arid, low seismicity, ~15 m lifts placed without mechanical compaction), Joe Scalia said the site has shown adequate density from placement/depositional energy plus climate for its first lift, though the story is incomplete since they are moving into subsequent lifts. He emphasized the risk of internal resaturation or an emerging water table if lifts are built too fast, too thick, or without enough dewatering, but noted Carrara’s own climate, degree of dewatering, and specific tailings characteristics have made the approach workable so far — including an early-stage episode where material was too wet and the site temporarily reverted to more conventional management before returning to large-tonnage filtered tailings.

Terminology: "filtered" vs. other dewatering methods

A participant challenged the “filtered” terminology given that belt presses, screw presses, and centrifuges also dewater tailings. Rob responded that these are also considered filtration technologies in this context, and the panel will keep using “filtered tailings” as the working term.

Compaction uniformity in thick lifts

On whether compaction of tailings placed in thick lifts (e.g., 3 m) has been evaluated, Rob said no — this is exactly what the planned field trials aim to test, particularly whether uniform density can be achieved through the full lift thickness rather than a denser crust over looser material below (“layer cake” effect). Josh Rogers added that planned field methods include in-situ testing (DCP/CPT) correlated to actual densities and void ratios during a test-fill program, feeding into critical-state data to assess liquefaction susceptibility.

Microbial-induced calcite precipitation (bio-cementation)

Neither Rob nor Joe had direct background on this. Chris Bareither said he wasn't aware of MICP being applied via well injection specifically to filtered tailings deposits, and flagged the practical challenge of achieving uniform microbial distribution through a dense material after placement; he suggested Arizona State's Center for Bio Geotechnics as a resource. Joe Scalia raised the same “locking in a loose fabric” concern discussed for binders, plus uncertainty about the long-term geochemical stability of microbially precipitated cement, while noting bio-inspired approaches may still have value for adjacent problems like dust and erosion control. Both suggested this would be a good topic to submit to the consortium by email.

White paper scope on transport/stacking assumptions

Rob confirmed the white paper made no assumptions about tailings transport and placement operations — it was deliberately scoped to the geotechnical question only.

Heterogeneity and keeping the stack unsaturated

Asked whether a dewatering/drainage system could reliably keep a stack unsaturated instead of relying on achieving dilative density, Chris Bareither said he doesn’t believe full desaturation can be guaranteed anywhere in a real stack — heterogeneity can produce perched water even with drainage in place. He recommended site-specific hydrogeologic modeling of the stack and surrounding natural soils to anticipate where a phreatic surface might develop over the long term.

Other references mentioned

A participant referenced an AECOM report on tailings and overburden pile guidelines prepared for a public prosecutor's office in Minas Gerais, Brazil; Rob was not familiar with it but noted it should be searchable. A submitted question from an attendee (Corinne) did not come through audibly and was not answered live.

Closing comments