Fast, easy-to-use, machine learning-developed models of prediction of flash point, heat of combustion, and lower and upper flammability limits for inherently safer design

Abstract This study proposes easy-to-apply machine learning-developed models, which predict four flammability properties of pure organic compounds: the flash point, heat of combustion, lower flammability limit (LFL), and upper flammability limit (UFL). These flammability properties pose a strong impact on the inherently safer design of industrial processes. Similar to quantitative structure-property relationship (QSPR) or group contribution models, machine learning algorithms are utilized in this study to establish predictive models. Compared to previous models, this study uses readily available variables (i.e., the numbers of atomic elements, molecular weights, and normal boiling points) as default variables without the analysis of detailed molecular structures or the in-depth knowledge of chemistry. This study consists of two steps: Step 1) building multiple linear regression (MLR) models by incorporating default input variables and Step 2) building MLR models by incorporating interaction and transformed variables to improve the predictions from the models in Step 1. In Step 1, an optimal subset of predictors is identified by constructing an MLR model via the sequential floating backward selection (SFBS) algorithm. As a result of Step 1, the two constructed models of the flash point and heat of combustion are found to be adequate, while the predictability of LFL and UFL are insufficient. In Step 2, MLR models incorporating nonlinearity and interaction terms are constructed via the sequential floating forward selection (SFFS) algorithm by selecting the optimal subset of default variables. The results show that all the constructed models in Step 2 are adequate as predictive models; the mean absolute errors (MAEs) of the flash point, heat of combustion, LFL, and UFL are 7.31 (5.67 via the SFBS) [K], 60.6 (61.87 via the SFBS) [kJ/mol], 0.21 (0.19 via the SFBS) [vol.%], and 2.44 (2.33 via the SFBS) [vol.%], respectively. Compared to previous models, the approved models in this study provide highly competitive performance with enhanced simplicity and interpretability.

[1]  Mahmoud M. El-Halwagi,et al.  Incorporating inherent safety during the conceptual process design stage: A literature review , 2020 .

[2]  Test Method for Flash Point by Tag Closed Cup Tester , 2022 .

[3]  Suhani J. Patel,et al.  QSPR Flash Point Prediction of Solvents Using Topological Indices for Application in Computer Aided Molecular Design , 2009 .

[4]  J. Lim,et al.  QSPR prediction of the hydroxyl radical rate constant of water contaminants. , 2016, Water research.

[5]  Jiajia Jiang,et al.  Predicting the net heat of combustion of organic compounds from molecular structures based on ant colony optimization , 2011 .

[6]  Yong Pan,et al.  Prediction of flammability characteristics of pure hydrocarbons from molecular structures , 2009 .

[7]  Yong Pan,et al.  Prediction of the Upper Flammability Limits of Organic Compounds from Molecular Structures , 2009 .

[8]  M. Keshavarz,et al.  A simple and reliable method for prediction of flash point of alcohols based on their elemental composition and structural parameters , 2016 .

[9]  K. Ohtaguchi,et al.  A method for estimating flash points of organic compounds from molecular structures. , 1991 .

[10]  Mehdi Bagheri,et al.  Nonlinear molecular based modeling of the flash point for application in inherently safer design , 2012 .

[11]  R. L. Rowley,et al.  Prediction of pure‐component flash points for organic compounds , 2011 .

[12]  Richard L. Rowley,et al.  Estimation of the flash point of pure organic chemicals from structural contributions , 2010 .

[13]  T. A. Albahri.  Flammability characteristics of pure hydrocarbons , 2003 .

[14]  Regina Barzilay,et al.  Convolutional Embedding of Attributed Molecular Graphs for Physical Property Prediction , 2017, J. Chem. Inf. Model..

[15]  Hongfei Xu,et al.  Machine Learning and Deep Learning in Chemical Health and Safety: A Systematic Review of Techniques and Applications , 2020, ACS Chemical Health & Safety.

[16]  Laurent Catoire,et al.  A Unique Equation to Estimate Flash Points of Selected Pure Liquids Application to the Correction of Probably Erroneous Flash Point Values , 2004 .

[17]  F. Gharagheizi.  Quantitative Structure−Property Relationship for Prediction of the Lower Flammability Limit of Pure Compounds , 2008 .

[18]  Group contribution method for predicting the lower and the upper flammable limits of vapors in air , 1991 .

[19]  C. Yaws.  Chemical properties handbook , 1999 .

[20]  A. Murty Kanury.  A Relationship between the Flash Point, Boiling Point and the Lean Limit of Flammability of Liquid Fuels , 1983 .

[21]  Noor Quddus,et al.  Developing Quantitative Structure–Property Relationship Models To Predict the Upper Flammability Limit Using Machine Learning , 2019, Industrial & Engineering Chemistry Research.

[22]  Atsushi Fujii,et al.  Correlation between flash points and vapor pressures of organic compounds , 1982 .

[23]  George W. McLaren,et al.  Flammability properties of hydrocarbon solutions in air , 1972 .

[24]  G. Zahedi,et al.  Estimation of flash point and autoignition temperature of organic sulfur chemicals , 2012 .

[25]  E. Castro,et al.  QSPR Modeling of Heats of Combustion for Carboxylic Acids , 2007 .

[26]  Mike Chapple,et al.  Practical Machine Learning in R , 2020 .

[27]  Mahmoud M. El-Halwagi,et al.  What can the trove of CSB incident investigations teach us? A detailed analysis of information characteristics among chemical process incidents investigated by the CSB , 2021 .

[28]  Takahiro Suzuki,et al.  Quantitative Structure-Property Relationships for the Estimation of Boiling Point and Flash Point Using a Radial Basis Function Neural Network , 1999, J. Chem. Inf. Comput. Sci..

[29]  Qiang Wang,et al.  Prediction of the Flash Point Temperature of Organic Compounds with the Positional Distributive Contribution Method , 2012 .

[30]  G. Sin,et al.  Group-contribution based property estimation and uncertainty analysis for flammability-related properties. , 2016, Journal of hazardous materials.

[31]  Zhirong Wang,et al.  Quantitative structure-property relationship studies for predicting flash points of alkanes using group bond contribution method with back-propagation neural network. , 2007, Journal of hazardous materials.

[32]  Mohammad Amin Sobati,et al.  New structure-based models for the prediction of flash point of multi-component organic mixtures , 2019, Thermochimica Acta.

[33]  Harold U. Escobar-Hernandez,et al.  Review of recent developments of quantitative structure-property relationship models on fire and explosion-related properties , 2019, Process Safety and Environmental Protection.

[34]  G. S. Patil.  Estimation of flash point , 1988 .

[35]  G. Cooke,et al.  Prediction of Flash Points of Middle Distillates , 1956 .

[36]  K. Satyanarayana,et al.  Note: Correlation of flash points , 1991 .

[37]  Daniela M. Witten,et al.  An Introduction to Statistical Learning: with Applications in R , 2013 .

[38]  Amin Alibakshi,et al.  Strategies to develop robust neural network models: Prediction of flash point as a case study. , 2018, Analytica chimica acta.

[39]  F. Quina,et al.  A simple method to evaluate, correlate and predict boiling and flash points of alkynes , 2012 .

[40]  Zeren Jiao,et al.  Machine learning prediction of hydrocarbon mixture lower flammability limits using quantitative structure‐property relationship models , 2020, Process Safety Progress.

[41]  Farhad Gharagheizi,et al.  A New Neural Network−Group Contribution Method for Estimation of Flash Point Temperature of Pure Components , 2008 .

[42]  Bernd Bischl,et al.  mlr: Machine Learning in R , 2016, J. Mach. Learn. Res..

[43]  Josef Kittler,et al.  Floating search methods in feature selection , 1994, Pattern Recognit. Lett..

[44]  Ritu Jain,et al.  QSPR Analysis of Flash Points , 2001, J. Chem. Inf. Comput. Sci..

[45]  John H. Maindonald,et al.  Data Analysis and Graphics Using R: An Example-Based Approach , 2010 .

[46]  Takahiro Suzuki.  Note: Empirical relationship between lower flammability limits and standard enthalpies of combustion of organic compounds , 1994 .

[47]  Juan A. Lazzús,et al.  Prediction of Flash Point Temperature of Organic Compounds Using a Hybrid Method of Group Contribution + Neural Network + Particle Swarm Optimization , 2010 .

[48]  D. Mathieu,et al.  Insight into the contribution of individual functional groups to the flash point of organic compounds. , 2014, Journal of hazardous materials.

[49]  M. Keshavarz,et al.  Simple method for reliable predicting flash points of unsaturated hydrocarbons. , 2011, Journal of hazardous materials.

[50]  Fu‐Yu Hshieh.  Correlation of closed‐cup flash points with normal boiling points for silicone and general organic compounds , 1997 .

[51]  L. G. Britton.  Using heats of oxidation to evaluate flammability hazards , 2002 .

[52]  A simple equation for prediction of net heat of combustion of pure chemicals , 2008 .

[53]  Robert Gould,et al.  A Modern Approach to Regression with R , 2010 .

[54]  Qingsheng Wang,et al.  Prediction of Minimum Ignition Energy from Molecular Structure Using Quantitative Structure–Property Relationship (QSPR) Models , 2017 .

[55]  Michail Tsagris,et al.  Multicollinearity. , 2021, American journal of orthodontics and dentofacial orthopedics : official publication of the American Association of Orthodontists, its constituent societies, and the American Board of Orthodontics.

[56]  Tareq A. Albahri,et al.  MNLR and ANN structural group contribution methods for predicting the flash point temperature of pure compounds in the transportation fuels range , 2015 .

[57]  Neural network techniques applied to predict flammability limits of organic compounds , 1995 .

[58]  G. Marshall,et al.  An introduction to descriptive statistics: a review and practical guide , 2010 .

[59]  Kaili Xu,et al.  Prediction of upper flammability limits for fuel mixtures using quantitative structure–property relationship models , 2018, Chemical Engineering Communications.

[60]  Frank H. Quina,et al.  Development of a Simple Method to Predict Boiling Points and Flash Points of Acyclic Alkenes , 2011 .

[61]  Mehdi Bagheri,et al.  BPSO-MLR and ANFIS based modeling of lower flammability limit , 2012 .

[62]  W. Affens.  Flammability Properties of Hydrocarbon Fuels. Interrelations of Flammability Properties of n-Alkanes in Air. , 1966 .

[63]  A novel model for predicting lower flammability limits using Quantitative Structure Activity Relationship approach , 2017 .

[64]  Fu‐Yu Hshieh.  Predicting heats of combustion and lower flammability limits of organosilicon compounds , 1999 .

[65]  Iva B. Stoyanova-Slavova,et al.  QSPR modeling of flash points: an update. , 2007, Journal of Molecular Graphics and Modelling.

[66]  Qingsheng Wang,et al.  Development of machine learning based prediction models for hazardous properties of chemical mixtures , 2020, Journal of Loss Prevention in the Process Industries.

[67]  F. Quina,et al.  Simple Method to Evaluate and to Predict Flash Points of Organic Compounds , 2011 .

[68]  M. S. High,et al.  Prediction of upper flammability limit by a group contribution method , 1987 .

[69]  J. J. Montaño Moreno,et al.  Using the R-MAPE index as a resistant measure of forecast accuracy. , 2013, Psicothema.

[70]  Yong Pan,et al.  A novel QSPR model for prediction of lower flammability limits of organic compounds based on support vector machine. , 2009, Journal of hazardous materials.

[71]  Gintaras V. Reklaitis,et al.  Process systems engineering - The generation next? , 2021, Comput. Chem. Eng..

[72]  Robert Tibshirani,et al.  The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd Edition , 2001, Springer Series in Statistics.

[73]  W. R..  The Elements of Physical Chemistry , 1902, Nature.

[74]  F. Quina,et al.  Improved Prediction of Hydrocarbon Flash Points from Boiling Point Data , 2010 .

[75]  Kaili Xu,et al.  Prediction of lower flammability limits of blended gases based on quantitative structure–property relationship , 2018, Journal of Thermal Analysis and Calorimetry.

[76]  F. Gharagheizi,et al.  Prediction of Flash Point Temperature of Pure Components Using a Quantitative Structure–Property Relationship Model , 2008 .