Towards a Digital Twin for the Quantitative Risk Assessment of Domino Effects in Fuel Storage Facilities

Fuel storage facilities concentrate large inventories of flammable substances, so that a single primary accident may escalate, through thermal radiation, into a cascading sequence involving several tanks (domino effect). This work presents the probabilistic core of a digital twin for such facilities, understood as a probabilistic replica of the facility with a fast query interface (live-data assimilation is not yet implemented): a permanently available, interactively queryable model that characterises, in probabilistic terms, the thermal radiation field generated by accident scenarios in a storage facility and derives from it an integrated assessment of its safety and economic risk. The framework combines (i) a Monte Carlo simulation of 108 facility-years, in which primary accidents and their fire-driven escalation are sampled using validated models of pool fires, thermal radiation and equipment failure; (ii) a point-wise condensation of the resulting radiation intensity into Gamma distributions fitted by maximum likelihood, whose parameter maps embody the frequency of every fire configuration; and (iii) a machine-learning surrogate, based on Generalized Regression Neural Networks, that renders the probabilistic field continuous over the terrain and evaluable in a fraction of a millisecond. On this backbone, the model delivers mutually consistent, receptor-level indicators: annual exceedance frequencies of harm thresholds (contour maps), the expected number of fatalities, distinguishing sheltered and outdoor population, and the Expected Annual Loss of every asset. In a demonstration on a five-tank facility storing light naphtha, jet fuel and diesel within a mixed industrial–residential complex, 16.5% of the accidents escalated to multiple tanks, the individual risk at the most exposed third-party location lay more than one order of magnitude below the 10−6 per year land-use criterion, and 97.5% of the expected losses were borne by the facility itself. The scope is limited to pool fires in fixed-roof atmospheric tanks, in wind-free conditions and without credited safety barriers.

Authors

Institutions

Publication Details

Journal
Fire
Published
2026-10-15
DOI
https://doi.org/10.3390/fire9100457
Primary Topic
Risk and Safety Analysis
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Towards a Digital Twin for the Quantitative Risk Assessment of Domino Effects in Fuel Storage Facilities

Ramón Quiza, Felipe San Luis Gutiérrez, Deivis Ávila, Yanelys Cuba Arana et al.
Fire
Risk and Safety Analysis
article

Towards a Digital Twin for the Quantitative Risk Assessment of Domino Effects in Fuel Storage Facilities

Ramón Quiza, Felipe San Luis Gutiérrez, Deivis Ávila, Yanelys Cuba Arana, Jesús Luis-Orozco
article en

Abstract

Fuel storage facilities concentrate large inventories of flammable substances, so that a single primary accident may escalate, through thermal radiation, into a cascading sequence involving several tanks (domino effect). This work presents the probabilistic core of a digital twin for such facilities, understood as a probabilistic replica of the facility with a fast query interface (live-data assimilation is not yet implemented): a permanently available, interactively queryable model that characterises, in probabilistic terms, the thermal radiation field generated by accident scenarios in a storage facility and derives from it an integrated assessment of its safety and economic risk. The framework combines (i) a Monte Carlo simulation of 108 facility-years, in which primary accidents and their fire-driven escalation are sampled using validated models of pool fires, thermal radiation and equipment failure; (ii) a point-wise condensation of the resulting radiation intensity into Gamma distributions fitted by maximum likelihood, whose parameter maps embody the frequency of every fire configuration; and (iii) a machine-learning surrogate, based on Generalized Regression Neural Networks, that renders the probabilistic field continuous over the terrain and evaluable in a fraction of a millisecond. On this backbone, the model delivers mutually consistent, receptor-level indicators: annual exceedance frequencies of harm thresholds (contour maps), the expected number of fatalities, distinguishing sheltered and outdoor population, and the Expected Annual Loss of every asset. In a demonstration on a five-tank facility storing light naphtha, jet fuel and diesel within a mixed industrial–residential complex, 16.5% of the accidents escalated to multiple tanks, the individual risk at the most exposed third-party location lay more than one order of magnitude below the 10−6 per year land-use criterion, and 97.5% of the expected losses were borne by the facility itself. The scope is limited to pool fires in fixed-roof atmospheric tanks, in wind-free conditions and without credited safety barriers.

FireVol. 9(10)
Universidad de La Laguna (ES), University of Matanzas (CU)
Affordable and clean energy, Climate action, Industry, innovation and infrastructure
Openalex Percentile: Top 11%
Risk and Safety Analysis
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.