Investigation and prediction of co-pyrolysis between oily sludge and high-density polyethylene via in-situ DRIFTS, TGA, and artificial neural network

Abstract

Oily sludge (OS), a hazardous waste produced in a large amount in petroleum industries, has raised great interest in research, and thermal treatment method such as pyrolysis is promising to achieve the goal. This research studied the pyrolysis behavior during co-pyrolysis of OS and high-density polyethylene (HDPE, plastic waste model component) with different blend ratios. The changes of functional groups against pyrolysis temperature were detected via in-situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS). The interactive effect between the two feedstocks during co-pyrolysis was analyzed via thermogravimetric analysis (TGA). The DRIFTS showed that the pyrolysis of HDPE can provide abundant •CH3 free radicals which may accelerate the pyrolysis of OS. The disparities between theoretical and experimental thermogravimetric/derivative thermogravimetry (TG/DTG) curves of mixtures indicated a synergistic interaction during co-pyrolysis. Besides, the activation energy of co-pyrolysis was obtained based on kinetic analysis, which showed that the average activation energy followed the trend of O3H1 < OS < O1H1 < HDPE < O1H3. In addition, two artificial neural network (ANN) models, namely ANN model Ⅰ with training and testing R2 of 0.99 (for prediction of interactive effect) and ANN model Ⅱ with training and testing R2 of 0.92 (for prediction of activation energy), were established and validated by comparing with the experimental data, which showed satisfied credibility. Ultimately, the two ANN models were used to reversely guide and optimize experiment design for the highest synergistic effect and the minimum activation energy. This study offers a new insight and strategy to aid pyrolysis experimental studies.