Scholarly record

PREDICTIVE RISK ASSESSMENT AND CFD-INFORMED RESOURCE ALLOCATION FOR FIRE SERVICE RESPONSE IN INDUSTRIAL FACILITIES

Dennis-Paul FRENTIU, Titus-Paul Frentiu, Alexandru MINA, Ionel Barbul, Dragos Gabriel Vasilescu

Abstract

Industrial sites concentrate high-energy processes, hazardous substances, and complex building layouts that can amplify fire, explosion, and toxic-smoke events. Fire and rescue units (ISU) often arrive with incomplete or outdated information about the facility, which increases response time and exposes crews to avoidable hazards. This paper proposes a simple, patent-oriented framework for optimizing ISU interventions in industrial zones by combining four building blocks: operational response data, computational fluid dynamics (CFD) scenarios for smoke and flame propagation, industrial IoT sensing, and predictive risk software. The core concept is a continuously updated digital risk profile for each site. Historical incidents, near-misses, and dispatch logs are used to map typical delays (access routes, hydrant availability, internal traffic constraints) to facility attributes (stored chemicals, compartmentation, ventilation, and hall geometry). A compact library of CFD scenarios is pre-computed for representative ignition points and ventilation states to estimate tenability limits (visibility, temperature, toxic exposure) and to derive simple decision cues such as safe approach corridors and priority isolation zones. In parallel, low-cost IoT nodes (temperature, smoke, gas, pressure, door status) provide real-time signals that select the closest CFD scenario and update the predicted escalation timeline. For optimization, the framework uses a two-stage logic that is easy to deploy. Stage 1 supports dispatch with a risk-weighted resource recommendation (vehicles, breathing apparatus, foam agent, hazmat support) based on a severity-probability score. Stage 2 supports on-scene management with dynamic allocation of teams to objectives (search, containment, cooling, ventilation control) using simple constraints: crew safety limits, water supply, access, and critical-infrastructure protection. The output is a digital alerting and assignment module that can integrate with existing dispatch systems and generate standardized safety tasks for site personnel (SSM) before and during an incident. From an inventor's perspective, the novelty is the coupling of a CFD scenario index with live sensor-triggered selection and a resource-allocation engine that produces audit-ready recommendations. Claimable elements include the scenario-indexing method, the risk-to-resource mapping rules, and the safety constraint set. Future work targets field validation in multiple industries, improved interoperability (GIS, BIM, and emergency radio data), and machine-learning calibration to reduce false alarms while shortening time-to-first-action.

Publication details

Back to publication list