SAGD Produced Water: Silica Management and the Economics of OTSG Feedwater

In Steam-Assisted Gravity Drainage (SAGD) operations, the once-through steam generator (OTSG) consumes the largest share of produced water treatment capacity on site — and silica, more than any other single contaminant, determines how far that water can be recycled. Silica chemistry shapes pretreatment selection, caps boiler recovery, and quietly drives a significant share of SAGD operating cost.

Why Silica Controls OTSG Feedwater Quality

OTSGs operate at subcritical pressure (typically 80–90% steam quality, with the residual 10–20% discharged as blowdown). Unlike utility drum boilers, they have no internal steam separation stage capable of handling high dissolved solids — so feedwater quality limits are tight. Silica is the binding constraint because it is the only common constituent that volatilizes with steam at these pressures. Calcium and magnesium scale stays on the tubes; silica travels.

Silica carryover into the steam phase deposits on turbine blades and superheater surfaces as a hard, glassy scale that is extremely difficult to remove and progressively degrades turbine efficiency and mechanical balance. This is why OTSG feedwater silica limits are set at 5–10 mg/L (and often lower for facilities with steam turbines) even though other dissolved solids may be tolerated at far higher concentrations. Exceeding the silica limit is not a gradual performance issue — it is a production and equipment-integrity event.

Hot Lime Softening vs. RO: Two Treatment Paths

Alberta SAGD facilities historically relied on warm or hot lime softening (WLS/HLS) for produced water treatment. Hot lime softening at 90–120°C precipitates hardness, magnesium, and a substantial fraction of silica — silica removal of 60–80% is achievable with proper magnesium-to-silica ratios — producing OTSG feedwater that meets silica limits with a mature, well-understood process. Its drawbacks are significant at today’s recycle rates: large chemical consumption (lime, soda ash, magnesium salts), a substantial sludge handling burden, and a recovery ceiling driven by the process itself.

Reverse osmosis-based treatment offers a fundamentally different profile. Microfiltration pretreatment operating at 90–98% water recovery removes suspended solids, oil, and hardness-scaling precursors ahead of the membranes, and advanced RO systems treat the high-TDS produced water stream at 80–90% recovery with 99%+ TDS rejection. Silica is rejected along with everything else — RO typically achieves 97–99% silica rejection — so the RO permeate comfortably meets OTSG silica limits without chemical softening.

• Chemical consumption: HLS consumes lime and coagulants continuously; MF-RO pretreatment chemistry is limited to antiscalant and periodic CIP chemicals
• Sludge volume: HLS generates tonnes per day of chemical sludge requiring dewatering and disposal; MF-RO concentrates the rejected mass into a much smaller blowdown or reject stream
• Recovery: HLS-based trains typically recycle 70–80% of produced water; MF-RO trains push overall recovery higher, reducing fresh water makeup and disposal volumes
• Footprint and operability: MF-RO is compact and modular; HLS requires large clarifier vessels, solids handling, and significant operator attention

For greenfield or debottlenecking projects, the practical question is which path the site’s specific chemistry favours. Produced waters with high silica-to-hardness ratios tend to favour RO-based trains, because HLS silica removal depends on precipitating magnesium silicate — chemistry that becomes chemical-hungry exactly when silica is highest. Waters with moderate silica and high hardness can be excellent HLS candidates, particularly where the sludge has an existing handling pathway. This is one of the few process decisions in SAGD where a few months of treatability work changes the capital answer by tens of millions of dollars.

Silica and the Blowdown Recycle Ceiling

The economic significance of silica extends beyond the OTSG itself. Every SAGD facility wants to recycle boiler blowdown — the concentrated stream rejected from the OTSG — back through the treatment train, because blowdown volume directly sets fresh water makeup and disposal requirements. But blowdown is where silica concentrates. If feedwater contains 10 mg/L silica at 80% boiler steam quality, blowdown contains roughly 50 mg/L. Treating that stream means pushing RO recovery against amorphous silica saturation, the least forgiving scaling chemistry in the entire train.

Amorphous silica solubility is strongly temperature- and pH-dependent, and once it polymerizes and deposits on membrane elements, conventional CIP is often ineffective. Managing silica-limited recovery requires careful pH control, antiscalant selection validated against the specific water chemistry, and conservative recovery staging. In practice this is the point where pilot testing earns its cost: silica kinetics at high recovery are water-specific enough that projections from generic indices routinely misjudge the operable limit by 5–10 percentage points of recovery.

The Economics: Reliability First, Recovery Second

OTSG feedwater economics are dominated by three terms: steam reliability (unplanned OTSG outages in SAGD directly defer bitumen production), chemical and consumables cost, and water balance (makeup plus disposal). Silica management touches all three. Facilities running HLS with marginal silica performance pay in turbine washing, acid cleaning cycles, and constrained blowdown recycle; facilities running MF-RO pay in membrane replacement and CIP frequency. Neither is universally cheaper — the correct answer depends on produced water chemistry, disposal access, and the facility’s water stress over the asset’s remaining life.

GWTS evaluates SAGD produced water treatment through pilot-validated process design rather than vendor defaults: confirming silica rejection and operable recovery on the actual water, then engineering the train — MF, advanced RO, and any polishing stage — around the measured limits. For facilities weighing a retrofit or expansion of OTSG feedwater capacity, that data is the difference between a defensible capital decision and an expensive assumption.

The near-term outlook favours membrane-based trains. As Alberta SAGD facilities push recycle rates higher under water stress and produced water chemistry continues to evolve, the flexibility of MF-RO — adjusting recovery and staging as the water changes, without rebuilding a lime plant — is increasingly the deciding factor in debottlenecking studies.