Under-deposit corrosion doesn't announce itself with a uniform wall-thickness loss an inspector can trend easily, it hides underneath sand, scale, and biofilm sitting at the six o'clock position of a pipeline, corroding a small localized area far faster than the pipe around it. Low flow velocity is exactly what lets those deposits settle and stay in place, which means the operating conditions that feel the least eventful, a steady, low-rate production line, are often the ones building the worst localized corrosion risk. Sulfate-reducing bacteria frequently colonize underneath these same deposits, compounding metal loss through a biological mechanism that standard corrosion coupons and general inspection intervals were never designed to catch. Most pipeline integrity programs still rely on fixed-interval pigging and inspection instead of tracking the flow and deposit conditions that actually predict where under-deposit corrosion is forming. AI software built around pipeline flow modeling closes that gap by identifying which segments are accumulating risk before a scheduled inspection finds the damage already done. Operators can see how that maps to their own pipeline network by contacting iFactory support.
The Slowest, Steadiest Segment of Your Pipeline Might Be the One Corroding Fastest.
iFactory's AI models flow velocity, solids transport, and deposit accumulation across your pipeline network to flag exactly which segments are building under-deposit corrosion risk, before a scheduled dig or in-line inspection finds the pit already formed.
Why Under-Deposit Corrosion Hides From Standard Inspection Programs
Under-deposit corrosion isn't a single corrosion type, it's a set of conditions, solids settling out of the flow stream, forming a deposit, and creating a localized environment underneath it where different corrosion and microbial mechanisms can accelerate metal loss well beyond the general corrosion rate happening everywhere else in the pipe.
Sand and Solids Settling
Production fluids carry sand, corrosion products, and reservoir solids that settle out whenever flow velocity drops low enough, building deposits fastest at the pipe's low point.
Localized Galvanic Cells
Partially covered deposit edges create anodic and cathodic zones side by side, driving preferential corrosion right at the boundary of the deposited area.
Microbiologically Influenced Corrosion
Sulfate-reducing bacteria thrive in the low-oxygen environment underneath settled deposits, and studies have found SRB can sharply accelerate both uniform and localized metal loss.
Dead Legs and Intermittent Flow
Bypass segments and low-flow branches are especially vulnerable, since stagnant conditions let solids and water settle undisturbed for extended periods.
Flow Velocity Is the Variable That Decides Whether a Deposit Forms
The same pipeline segment can behave very differently depending on flow rate, and knowing where a given segment sits on the velocity spectrum is the single clearest predictor of whether it's building deposit-driven corrosion risk.
| Flow Regime | What Happens to Solids | UDC Risk Implication |
|---|---|---|
| Low Velocity | Sand and solids settle and accumulate at the pipe's low point | Highest risk, deposits form and stay undisturbed |
| Critical Transport Velocity | Flow is just fast enough to keep most solids moving in suspension | Moderate risk, intermittent deposition during flow variation |
| High Velocity | Solids stay entrained and existing deposits, including protective scale, can erode | Different risk profile, erosion and scale removal rather than settling |
See Where Your Network's Flow Profile Predicts UDC Risk
iFactory can model your pipeline's flow and solids transport data against known deposit-formation conditions before your team commits to a rollout.
Fixed-Interval Pigging vs. Flow-Data-Driven Targeting
| Capability | Fixed-Interval Pigging Program | AI-Driven Flow-Based Targeting |
|---|---|---|
| Which segments get prioritized | Same schedule applied across the network regardless of flow history | Highest-risk segments identified from actual flow and solids data |
| Dead leg and low-flow branch coverage | Often deprioritized since they're harder to access for pigging | Flagged specifically because low or intermittent flow is the primary risk signal |
| Inspection timing | Calendar-driven, independent of actual deposit conditions | Timed to when deposit accumulation is predicted to reach a risk threshold |
| MIC risk visibility | Typically assessed only after a failure or targeted investigation | Tracked continuously alongside flow and deposit conditions |
What the AI Model Actually Tracks Across Your Network
Segment-Level Flow Velocity Modeling
Live and historical flow rate data is compared against known critical transport velocities to flag which segments are operating in deposit-forming conditions.
Solids Loading and Deposit Accumulation
Sand production data and flow history are combined to estimate how much solid material is likely settling in each segment over time.
Dead Leg and Low-Flow Segment Identification
Network topology and flow patterns identify bypass segments and intermittent-flow branches that carry elevated UDC and MIC risk by design.
Targeted Pigging and Inspection Scheduling
Predicted risk levels feed directly into inspection and cleaning schedules, so pigging runs and ILI tools get pointed at the segments that actually need them.
A Composite Scenario: The Low-Flow Segment That Never Made the Pigging Schedule
A midstream operator ran a fixed annual pigging schedule across its gathering network, cycling through segments in a set rotation regardless of individual flow history. One low-rate segment feeding a marginal well cluster had operated well below critical transport velocity for over a year, a condition the fixed schedule had no way to flag since the segment wasn't due for its pig run for another eight months.
After connecting flow rate and solids data to a predictive model, the segment was flagged as high UDC risk based on its sustained low-velocity operation, moving it up in the pigging queue by six months. The pig run recovered a deposit consistent with the model's predicted accumulation, and an in-line inspection of the cleaned segment found localized wall loss at the six o'clock position that had not yet reached a reportable threshold, catching the damage while a repair was still straightforward.
Rolling AI-Driven UDC Monitoring Out Across a Network
Map Flow History Across the Network
Existing flow rate data by segment establishes a baseline for which parts of the network have spent the most time below critical transport velocity.
Identify Dead Legs and Low-Flow Branches
Network topology is reviewed alongside flow data to surface bypass segments and intermittent branches that carry structurally higher UDC risk.
Validate Predictions Against a Pigging Run
Predicted high-risk segments are checked against actual deposit recovery and ILI findings before the model drives scheduling decisions.
Shift the Pigging Program to Risk-Based Scheduling
Once validated, the fixed-interval schedule transitions to a continuously updated, risk-ranked queue across the full network.
Common Mistakes Pipeline Operators Make With Under-Deposit Corrosion
Treating All Segments With the Same Pigging Interval
A flat schedule ignores that low-flow segments accumulate deposits far faster than high-velocity ones, leaving the highest-risk sections under-serviced.
Deprioritizing Dead Legs Because They're Hard to Pig
Low-flow and bypass segments carry some of the highest UDC and MIC risk precisely because they're the hardest to mechanically clean and inspect.
Assessing MIC Only After a Failure
Sulfate-reducing bacteria activity typically isn't investigated until a leak or failure occurs, even though it frequently co-develops with the same deposits driving UDC.
Relying on General Wall-Thickness Trending Alone
UDC is localized by nature, so a general corrosion rate trend can look stable even while a specific six o'clock location is losing wall thickness far faster.
Frequently Asked Questions
What exactly causes under-deposit corrosion in a pipeline?
Under-deposit corrosion happens when solids like sand, scale, or corrosion products settle out of the flow stream and form a deposit on the pipe wall, most commonly at the six o'clock position where gravity concentrates solids and water. That deposit creates a localized environment underneath it, often low in oxygen and hospitable to sulfate-reducing bacteria, where corrosion mechanisms can progress far faster than the general corrosion happening across the rest of the pipe. Operators can review how this mechanism applies to their own pipeline's flow conditions by reaching out to iFactory support.
Why does low flow velocity specifically increase UDC risk?
Flow velocity determines whether solids stay suspended in the fluid or settle out onto the pipe wall. Research on pipeline corrosion has found that lower flow velocity directly promotes the deposition of solid particles like sand, while higher velocities tend to keep solids entrained and can even damage or erode existing deposits. That means the steadiest, lowest-rate segments of a network, the ones that feel the least operationally eventful, are often building the most UDC risk over time.
How is under-deposit corrosion connected to microbiologically influenced corrosion?
The two mechanisms frequently occur together, since the same deposits that drive UDC create ideal conditions for sulfate-reducing bacteria to establish colonies underneath them. Studies on carbon steel have found that SRB activity can significantly accelerate both uniform and localized corrosion rates when present in or under a deposit. Because of that overlap, an effective monitoring approach has to account for both mechanisms rather than treating UDC as a purely chemical or mechanical process.
Do we need new flow instrumentation to model UDC risk across our network?
Most pipeline networks already collect flow rate data through existing SCADA and metering systems, and that historical and live flow data is the primary input the predictive model uses. Network topology and known solids production data are layered on top rather than requiring new sensors at every segment. Book a demo to see how a rollout maps to your existing flow data.
Can this replace pigging and in-line inspection, or does it work alongside them?
It works alongside them, and its main value is making pigging and inspection more targeted rather than replacing the physical work of removing deposits and measuring wall thickness. Flow-based risk prediction tells a team which segments are most likely accumulating deposits and where MIC risk may be developing, so pigging runs and ILI tools get scheduled around actual conditions instead of a fixed calendar rotation. The physical inspection and cleaning still confirm what the model predicts.
Find the Segment Building Corrosion Risk Before Your Next Scheduled Pig Run Does
iFactory models flow velocity, solids transport, and deposit accumulation across your pipeline network to target inspections where under-deposit corrosion is actually forming.
