To implement Digital Transformation of a Petrochemical Plant in Africa to avoid Process Upsets/ Incidents and Production Loss.
A petrochemical plant in Africa needed to equip itself with necessary technology that would help avert process upsets/incidents and production losses. The business objective included:
- To develop an AI and first principle hybrid solution that would predict and detect anomalies and also identify the root cause of these anomalies well in advance
- To prevent high solvent ppm from going to the downstream system, and thereby, avoiding any potential upset/ incident there
Due to an upset in upstream system or around the solvent stripper of petrochemical plant, the solvent ppm in outlet product stream increases. High solvent ppm in pellets along with high conveying air temperature can ignite pellets and cause fire in the downstream system.
- Considers all tags and system around the stripper and develops the model using most critical parameters and cause(s) of the problem
- Detects the anomaly, the contributing factors of anomaly and identifies the cause of the problem for safe operation of plant
- Predicts high anomaly score well in advance so that a corrective action can be taken to avoid upset and fire incident
Business Impacts & Outcomes:
- Process excursion detection to prevent long term damage and enhances reliability
- Increases plant availability and uptime; thus, avoids production and revenue loss
- Minimizes cleaning/ re-startups and legal costs due to process upsets/ fire/ accidents