A BOF vessel campaign is a race against refractory wear from the first heat to the last, and every plant manager knows the uncomfortable trade-off: push the lining too far and risk a breakout, relining too early and you sacrifice heats that were still available. Lance condition and slag splashing effectiveness are the two biggest levers on how far a lining actually stretches, yet most plants still manage both with periodic inspection rather than continuous data. Plant managers evaluating a more precise approach can start at ifactoryapp.com/support.
A Vessel Campaign Has Three Distinct Risk Phases
Refractory wear is not linear across a campaign, and treating early, mid, and late campaign heats the same way is where most plants lose either lining life or safety margin. The lifecycle below reflects how wear risk actually shifts, and where AI monitoring adds the most value at each stage.
Lining wear is slow and predictable. Focus is on establishing baseline slag splashing coverage patterns and lance position calibration that the model will compare against for the rest of the campaign.
Wear rate accelerates at hot spots — typically the trunnion and charge pad areas. Slag splashing effectiveness monitoring becomes critical here to reinforce these zones before they become thin points.
Breakout risk rises sharply. Shell temperature monitoring and lining thickness estimates drive the relining decision, replacing a fixed heat-count target with an actual wear-based endpoint.
Lance Condition Is the Other Half of the Equation
Oxygen lance position accuracy and nozzle wear directly affect both blowing efficiency and how evenly slag splashing coats the vessel lining — a worn lance changes the jet penetration depth, which changes where slag is deposited during splashing and leaves uneven refractory protection. Monitoring lance tip wear alongside vessel lining data lets a plant manager see cause and effect across both systems instead of treating them as separate maintenance items.
| System | Key Wear Indicator | Effect on Campaign Life |
|---|---|---|
| Oxygen Lance | Nozzle erosion, tip geometry drift | Uneven slag splash coverage, accelerated lining hot spots |
| Vessel Lining | Shell temperature, thickness estimate | Direct determinant of safe heats remaining |
| Slag Splashing | Coating thickness uniformity | Protective layer effectiveness across trunnion and charge pad zones |
| Trunnion Ring | Localized thinning rate | Common first failure point in extended campaigns |
What Precise Campaign Management Delivers
Extending a campaign by even one week of additional heats changes the plant's relining and refractory procurement schedule, and it does so predictably rather than as a surprise outcome. Book a Demo to see the wear model applied against your own campaign history.
Getting Started Without Disrupting Active Campaigns
Deployment is designed to begin monitoring mid-campaign if needed, rather than requiring a plant to wait for a fresh reline before starting data collection. Shell temperature sensors and lance position tracking can be added during a normal maintenance window, and the model begins building a wear baseline from that point forward while incorporating whatever historical relining records exist for calibration.
Frequently Asked Questions
The model provides a confidence-scored range rather than a single exact number, since refractory wear depends on multiple interacting factors including heat chemistry, slag basicity, and lance performance that vary heat to heat. That range narrows as the campaign progresses and more live wear data accumulates, giving the plant manager a much tighter planning window by mid to late campaign than a fixed heat-count assumption would provide. The goal is to replace a guess with a data-backed estimate, not to promise false precision. Every estimate is shown alongside the underlying shell temperature and thickness trend so the team can verify it against physical inspection.
No, the core monitoring approach relies on external shell temperature sensors and existing lance position and pressure data rather than sensors mounted inside the vessel refractory. This avoids introducing new failure points inside a harsh, high-temperature environment. Shell temperature is a well-established proxy for lining thickness because heat conduction through a thinning refractory layer produces a measurable and consistent shell temperature signature. Installation is completed externally during a normal maintenance window.
A worn lance nozzle changes the oxygen jet's penetration angle and splash pattern during slag splashing, which means the protective slag coating applied to the lining shifts away from its intended coverage areas — often leaving the trunnion or charge pad zones under-protected. The model correlates lance wear trend data against the specific zones showing accelerated shell temperature rise, which frequently reveals lance condition as a contributing factor the maintenance team had not connected to the lining wear pattern. This connection is one of the more actionable insights plants report after adoption, since lance maintenance is often scheduled independently of vessel campaign planning.
The most useful starting data is 2-3 years of relining records with heat counts, along with any available shell temperature logs, lance change history, and slag splashing operating parameters such as splash duration and nitrogen flow rate. If detailed historical data is limited, the model can still begin with a baseline calibration period during live operation, though the confidence range will start wider and narrow more slowly. Heat chemistry and slag basicity records, if available, further improve the model's ability to distinguish process-driven wear variation from equipment-driven wear. A data readiness review during onboarding will confirm exactly what is usable from your existing records.
Yes, this platform is designed as a complementary early-warning layer and does not replace or interfere with an existing breakout detection or emergency shutdown system. Breakout detection systems are typically tuned for near-term, high-confidence alarms close to the point of failure, while this wear model is designed to give weeks of advance notice well before that threshold is reached. Running both together gives a plant manager both the long lead time for planning and the safety-critical near-term protection already in place. ifactoryapp.com/support can confirm compatibility with your specific breakout detection vendor.
