Chapter 15 — Case Study: Oil Sands Tailings Dams of Northern Alberta¶

About this case study

This chapter is an original case study written for this reference. It synthesises publicly available practice guidance for the oil sands industry of northern Alberta — principally the CDA Bulletin article "Application of CDA Dam Safety Guidelines to the Geotechnical Design of Oil Sands Tailings Dams in Northern Alberta" (Eshraghian & Becker, Golder Associates, Canadian Dam Association Bulletin, Winter 2016) and the companion conference paper "Application of Geotechnical Instrumentation in Monitoring In-Pit Dykes Performance" (Soe Moe & Biggar, 2013). It is a teaching case study, not a reproduction of those sources and not a design document. The oil sands region is regulated by the Alberta Energy Regulator (AER) under the Dam Safety Technical Guideline; always work from the current AER/CEA guidance.

Oil sands tailings dams are, in the words of the source paper, "among the world's third largest proven oil reserves" — a scale that makes their containment structures some of the most demanding tailings dams anywhere. Yet they are not water-retention dams, and the distinction is the whole point of this case study: they are built over decades around an active mining operation, they contain a material that can liquefy, and they are held to a dam-safety regime written with water dams in mind. Applying that regime intelligently — rather than mechanically — is the engineering story.

15.1 The setting: why northern Alberta is a special case¶

15.1.1 Location and scale¶

The oil sands deposits (bituminous sands) lie in northern Alberta, within the Athabasca region primarily, with smaller areas in Cold Lake and Peace River. Current economic conditions limit most active mining to deposits within about 75 m of the surface; roughly 20 % of the resource is accessible to surface mining, the rest by in-situ methods. The figure below (from Alberta Geological Survey data, after Eshraghian & Becker) shows the three deposit areas.

Figure reference. The source article's Figure 1 maps the Athabasca, Cold Lake and Peace River oil sands areas across Alberta. (Not reproduced here.)

15.1.2 The mining and extraction process¶

The oil sands operations include mining, extraction, upgrading and supporting infrastructure. The mining sequence is a chain of disturbances:

Overburden (sands, till, clay) is stripped and stockpiled separately.
The oil sands are mined by truck-and-shovel, dragline or bucket-wheel excavator.
Extraction uses hot water, caustic and/or diluent, followed by hydrotransport and primary and secondary separation.
Tailings — sand, fines, water and residual bitumen — are piped as slurry to disposal areas.

The mining schedule is dictated by the extraction plant, not by the dam engineer: sand, fines, overburden and coarse tailings are stored in different facilities at different rates, and the sequence continues for decades.

15.1.3 Two climates, two histories (the geological context)¶

The geotechnical behaviour that matters is dominated by glaciation. The Athabasca deposit is mantled by glacial deposits: low-plasticity till of the Murray Formation (glaciolacustrine, younger than the Clearwater Formation) and the Clearwater Formation clays beneath it, which are understood to be glacially over-consolidated and pre-sheared under successive ice advances. The clays' closely spaced fissures can control stability even though the clays are otherwise strong — a lesson from glacial geology that recurs in every slope in the region.

Two further geological facts drive design:

Karst — the Devonian Waterways Formation below includes limestone and anhydrite; dissolution (karst) can control the geotechnical stability of tailings dams. (The Fort McMurray area is the site of an active collapse feature, the "3 km subsidence" documented in regional seismic-hazard work.)
Seismicity — the Athabasca region is as far as ~600 km from the Canadian Rocky Mountain seismic belt, so ground motions are regional, not near-field. The Krinitzsky–Chang (1996) McMurray Region Seismic Hazard Study provides the seismic parameters on which modern design is based, and the Geological Survey of Canada assessment places the Fort McMurray region among the lowest-probability seismic hazard areas of Canada (National Building Code of Canada, Open File 4459).

15.2 Two dam traditions: water-retention vs oil sands¶

The organising insight of the source paper is that CDA guidelines were written for water-retention dams, and oil sands tailings dams differ in ways that make direct application inappropriate. Four contrasts matter:

Dimension	Water-retention dam	Oil sands tailings dam
Construction period	Short (weeks to ~3 years)	Decades (20–50+ years of operation)
Operating life	Given, tens of years	Continues as long as the operation lives (design life 25–40 y, often longer)
Design basis	Uses reservoir water level as the load case	Has no normal reservoir level; the phreatic surface moves as the beach and pond change
Failure consequence	Flood from stored water	Movement of tailings (with or without free water) — environmental and reclamation consequence

These differences cascade into how the design flood, freeboard and loading conditions are treated. A tailings dam does not have a "normal water level" in the water-dam sense; the design must instead follow the operational life of the sand dam, with the phreatic surface maintained throughout that life.

15.3 Tailings deposits and dam types¶

Tailings are discharged as slurry into a disposal area. The coarse fraction (sand) settles out quickly, building a sand beach; fines settle out slowly and accumulate in the tailings pond. Depending on the depositional method (spigotting, sub-aerial, sub-aqueous), four deposit types are recognised:

Sand tailings — the coarse beach and dykes.
Fine tailings (fluid fine tailings, FFT — see Chapter 2).
Composite tailings — sand and fines blended for strength and fast reclamation.
Soft tailings — the fine, soft, low-density deposits in the pond centre.

The tailings dams themselves are built by the conventional earthfill techniques of Chapter 2 — spigotting the sand and constructing dykes by sequential raising — but with one constraint foreign to most tailings practice: the fill is often placed as a slurry and beach sand is often used as placed rather than dried and compacted, so the "fill" is much more like a deposited, saturated material than a compacted embankment.

Cyclic deposition. The largest downstream-consequence structures are typically the external tailings dykes — larger, higher upstream-construction dams containing total tailings (fine tailings and sand). During early years, an upstream tailings dyke cannot sustain the full load, so the fines-to-sand ratio and deposition are managed.

15.4 What fails, and why: the failure modes¶

The paper classifies oil sands tailings dam failure modes into two families:

Internal dam failure — failure surface from inside the dam and continuing through the dam and its foundation, into the tailings.
Upstream-slope instability — failure surface from the dam crest towards the pond, with the deposit acting as a resisting mass; during early construction, a high phreatic surface close to the dam crest can cause this.

To this, add the modes common to all tailings dams (Chapter 2):

Overtopping, external erosion, static liquefaction of loose beach and sand layers, seismic liquefaction (though the region is of low seismicity), and internal erosion and piping.
Karst-driven instability where the underlying Devonian limestone/anhydrite dissolves.

The paper's key design implication: the deposit itself must be included in the stability analyses, because the tailings pond and beach impose a load and a pore-pressure field on the dam. A water-retention dam is analysed with water on one side; a tailings dam must be analysed with a heterogeneous, changing, potentially liquefiable material on one side.

15.5 Applying the CDA guidelines: what transfers, what must change¶

The paper systematically maps the CDA (2007) Guidelines loading cases against what oil sands dams actually experience. The result is the practical heart of the case study.

CDA (2007) case	Water dam	Oil sands dam	Comment
During and of construction	Static loading + operating-basis static during dam raising	1.3 / 1.3	The dam must withstand static loads during construction and operation due to gravity applied to the dam
Long-term	Piezometric surface at maximum (long-term) operation	1.5 / 1.5	Long-term design assumes the phreatic surface reverts to a natural (no-free-water) level — there is no permanent reservoir
Full or partial rapid drawdown	1.2–1.3 / 1.2–1.3	Not applicable	Oil sands dams have no mechanism for a controlled reservoir drawdown, so this case does not apply
Seismic (pseudo-static / post-earthquake)	1.0 / 1.0	1.2–1.3 / 1.2–1.3	Lower probability but high consequence; design assumes tailings that are susceptible to liquefaction

Seepage considerations (the paper's §4.3). Because the deposit and dykes are saturated, the paper treats seepage as second in importance only to stability:

Steady-state seepage through the dam, dykes and foundation is analysed with flow nets or numerical methods.
Under-drainage, drains and collection systems are provided and maintained.
A high phreatic surface can develop active seepage and reduce stability; a low phreatic surface reduces that risk.
The pond-filling method (sub-aqueous deposition) itself reduces seepage.

Seismic considerations. Although the Athabasca region has low seismicity, the high consequence of failure means pseudostatic and post-earthquake analyses are warranted. The paper notes that a limit-equilibrium stability analysis is adequate for most dam design, and that the target minimum factor of safety for each case is set accordingly — but that more sophisticated methods (and in-situ testing to estimate susceptibility) may be needed for liquefaction-prone deposits.

Acceptable factors of safety. The paper's Table 1 (reproduced in concept above) gives the minimum target factors of safety for water and oil sands tailings dams side by side. Note the explicit statement that oil sands dams built using fill placement methods are not analysed using the same factor of safety as water-retention dams, and that the factors are target minimums, not guarantees.

15.6 Operational challenges: the observational approach¶

The paper's §5 is about operational practice, and it is where a design case study becomes a monitoring case study.

15.6.1 The case for the observational approach¶

Because the operating life and raising rates of oil sands dams change throughout the long construction period, the paper argues that the observational approach (the "learn-as-you-go" method formalised by Peck, 1969) is more appropriate to oil sands dams than to water-retention dams. The approach maximises design efficiency by updating analyses from measured performance; unlike most water-retention dams, whose conditions change little after construction, oil sands conditions change continuously.

The method's pre-conditions are demanding: the engineer must identify in advance what will happen and why, predefine the plan of action for every credible departure, and continuously review the monitoring against that plan.

15.6.2 Performance monitoring¶

The paper stresses that monitoring must be "in good detail and frequently consistent with the design" — and flags the trap that observational monitoring for dams during initial fill has not always been well done. The monitoring plan should:

Confirm the design — verify each safety assumption.
Supplement — detect departures early.
Fit the timeline — be appropriate to the planned pace of the dam (initial fill, early operation, long-term).
Relate to the observational approach — linked to a risk assessment methodology so that management can act.

15.6.3 Development of tailings liquefaction and the fills/consequences¶

The paper connects depositional method to liquefaction. Where tailings are deposited sub-aqueously, the pond is saturated and loose, and static liquefaction is possible for the fragile, low-density material. Whatever the deposition method, monitoring of pore pressure and deposit density is the core of liquefaction control (Chapter 6).

15.6.4 In-situ testing to verify the design¶

The paper's §5.4 is a practical contribution: in-situ testing (CPT, CPTu, seismic cone, etc.) to verify the deposit's state (density, strength, liquefaction susceptibility) is essential and is done during operations, not just at design. As the paper says: "The purpose of in-situ testing is to verify the design and assumptions... and to build confidence that the design basis and assumptions can be achieved." This is the observational approach applied to the deposit as well as to the dam.

15.7 The in-pit dyke experience: monitoring in practice (Albian)¶

The companion paper — "Application of Geotechnical Instrumentation in Monitoring In-Pit Dykes Performance" (Soe Moe & Biggar, 2013) — reports how this playbook is executed at the CNRL Albian in-pit dykes in the Athabasca region. This is the closest thing to a site-level case study available in the public record, and it shows the theory in operation:

Instrumentation matched to failure modes. Piezometers (vibrating-wire, typically) to track pore pressure and phreatic surface; inclinometers and settlement gauges to track deformation; survey monuments for surface movement — all tied to the identified failure modes and reported against thresholds.
In-pit dykes are built inside the mined-out pit and raised with the operation, so their behaviour is intimately coupled to the deposition sequence and to the pit floor geology.
Monitoring as construction control. Because the fill is deposited and the deposit changes, monitoring is not merely verification — it is part of the construction method, telling the operator when to advance, pause, or modify deposition.
The observational approach requires the monitoring to be fast enough to detect a departure in time to act — a point the CDA paper makes explicitly.

15.8 The reclamation arc (what makes oil sands different)¶

Underlying all of this is a purpose a water dam does not share: reclamation. As the paper's Figure 2 shows, a tailings dam is ultimately meant to become a "reclaimed landscape" — the tailings pond surface is part of the closure design, and the dam itself becomes a landform. This is why:

The deposit's long-term behaviour (consolidation, strength gain, density) matters as much as the dam's.
Composite tailings and beach conditioning are used to build strength for fast reclamation.
The "no water-retention dam" framing is not just a legal nicety — the structure is not intended to hold a permanent reservoir at closure.

15.9 Lessons for any tailings engineer¶

Even outside northern Alberta, the case study teaches transferable lessons:

Know which guideline you are inheriting. The CDA case shows that a reputable dam-safety standard may need interpretation before it fits a tailings facility — and that not all loading cases apply.
Let the operational reality shape the design. Decades-long construction, rate-controlled deposition and a moving phreatic surface are design inputs, not details.
Include the deposit in your analysis. A tailings dam's "load" is a changing, often liquefiable material — analyse it, and verify its state by in-situ testing.
Seepage is second only to stability. Under-drainage, drains and pond management are structural, not cosmetic.
Adopt the observational approach deliberately — with predefined actions, fast enough monitoring and continuous review.
Instrumentation is construction control, not just compliance.
Design for closure from day one — the dam becomes the landform.

15.10 Key takeaways¶

Oil sands tailings dams in northern Alberta are large, decade-long, operation-driven structures — not water-retention dams.
CDA guidelines apply, but require interpretation: the design-flood, freeboard and drawdown cases do not transfer; grading is set by the dam's actual loading cases.
Tailings liquefaction, seepage, and karst are the controlling geotechnical concerns; seismicity is low but consequence is high.
The observational approach is more appropriate here than for water dams because conditions change continuously.
In-situ testing and performance monitoring verify the design and underpin the observational approach.
The Albian in-pit dykes demonstrate monitoring used as construction control.
Reclamation is the purpose the design must serve.

Sources¶

Eshraghian, A. & Becker, D.E. (Golder Associates) — Application of CDA Dam Safety Guidelines to the Geotechnical Design of Oil Sands Tailings Dams in Northern Alberta, Canadian Dam Association Bulletin, Winter 2016, pp. 26–41.
Soe Moe, K.W. & Biggar, K.W. (2013) — Application of Geotechnical Instrumentation in Monitoring In-Pit Dykes Performance, conference paper (CNRL Albian).
Canadian Dam Association (2007) — Dam Safety Guidelines and Technical Bulletin: Geotechnical Design of Dams.
Alberta Energy Regulator — Dam Safety Technical Guideline and Guide to the Management of Tailings Facilities (with MAC).
Krinitzsky & Chang (1996) — McMurray Region Seismic Hazard Study.
Geological Survey of Canada — National Building Code of Canada seismic hazard, Open File 4459.

Regulatory currency. The AER Dam Safety Technical Guideline and the CDA guidelines are revised periodically. Verify the current edition before any design or submission.