BKP Mill Stabilizes Causticizing Efficiency and Improves White Liquor Quality Using Mt. Fuji AI
Haber deployed Mt. Fuji®, its predictive industrial AI platform, to optimize the causticizing process at a kraft pulp mill. By continuously predicting causticizing efficiency and prescribing optimal operating conditions across the slaker, causticizers, and clarifiers, Mt. Fuji stabilized white liquor quality, reduced chemical consumption, and improved downstream digester performance.
The Challenge
Causticizing is a critical step in the kraft recovery cycle, directly influencing white liquor quality, digester stability, pulp yield, and chemical consumption. Variability across the slaker, causticizers, and clarifiers often leads to inconsistent causticizing efficiency, forcing mills to increase lime and caustic usage to maintain cooking performance.
The mill experienced:
- Wide efficiency fluctuations (76–84%)
- Inconsistent white liquor strength
- Higher lime and caustic consumption
- Variable digester yield and Kappa number
- Increased bleach chemical demand downstream
Traditional control strategies relied on manual adjustments and delayed laboratory feedback, making the process reactive rather than predictive.
Haber's Approach
Mt. Fuji continuously collected live process data including green liquor density, slaking temperature, causticizer retention time, white liquor density, slaker retention time, and clarifier performance.
Machine learning models identified the critical relationships between slaker operation, lime chemistry, liquor properties, and downstream causticizing efficiency.
Mt. Fuji predicted efficiency drift before it occurred and prescribed the optimal operating conditions to stabilize white liquor strength while minimizing process variability.
Self-learning AI models continuously adapted to feed variability, lime quality, production changes, and seasonal operating conditions to sustain optimization over time.
Evidence-Backed Results
| Metric | Before | After | Change |
|---|---|---|---|
| Causticizing efficiency | 76–84% | 80–81% | Stable operating window |
| Operating variation | 8 percentage points | ~2 percentage points | 75% reduction |
Business Impact
Mt. Fuji transformed causticizing from a manually controlled operation into a predictive optimization system that continuously stabilized white liquor quality and improved chemical recovery performance.
For a 350,000 ADT/year kraft pulp mill, the deployment generated an estimated $2.0 million in annual run-rate value
Conclusion
Mt. Fuji enabled the mill to shift from reactive causticizing control to predictive process optimization. By continuously stabilizing causticizing efficiency and white liquor quality, the platform reduced chemical consumption, improved digester performance, and generated approximately $2 million in annual economic value.
