Adaptive Neuro-Fuzzy Inference System Prediction Method for Percentage Fatalities of Jet Fire Incident in Methanol Production Plant
Mohd Aizad Ahmad1, Zulkifli Abdul Rashid2

1Mohd Aizad Ahmad, Faculty of Chemical Engineering, Universiti Teknologi MARA, Shah Alam, Selangor, Malaysia.
2Dr Zulkifli Abdul Rashid, Faculty of Chemical Engineering, Universiti Teknologi MARA, Shah Alam, Selangor, Malaysia,.
Manuscript received on September 23, 2019. | Revised Manuscript received on October 15, 2019. | Manuscript published on October 30, 2019. | PP: 5765-5772 | Volume-9 Issue-1, October 2019 | Retrieval Number: A3060109119/2019©BEIESP | DOI: 10.35940/ijeat.A3060.109119
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: This research highlights on development of simpler prediction method for percentage fatalities occurred during jet fire incident in methanol production plant. A lot of parameters involved before fatalities can be determined using consequence model analysis. The parameters involved needed to calculate surface emitting power, view factor, transmissivity and area affected footprint to determine estimated fatalities. HYSYS software used to simulate density of mixture, mass and volume fraction of each component resulting from carbon dioxide and hydrogen reaction. These values used as input in ALOHA simulation to estimate area footprint. Affected area footprint then calculated in MARPLOT, which in turn used for estimating percentage fatalities. These resulting fatalities used as ANFIS prediction analysis to predict percentage fatalities, then compared to the simulation data from ALOHA and MARPLOT. The selected input data was operating pressure, volume, mass, size of leakage and wind speed. The predicted data from ANFIS attained R2 at 0.9998 for both membership function used, triangular and Gaussian while for capabilities test, R2 of 0.998 was achieved using Gaussian. Therefore, simpler method to predict percentage fatalities for the event of jet fire in methanol plant was successful.
Keywords: ANFIS, jet fire, methanol, prediction method.