Accuracy of predictions made by machine learned models for biocrude yields obtained from hydrothermal liquefaction of organic wastes

Cheng, Feng; Belden, Elizabeth R.; Li, Wenjing; Shahabuddin, Muntasir; Paffenroth, Randy; Timko, Michaël T.

doi:10.1016/j.cej.2022.136013

Cited by 34 publications

(12 citation statements)

References 72 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Many different types of machine-learned models have been published in the literature for a range of different engineering applications. − Accordingly, a wide variety of model types were evaluated as the ideal model for a given data set cannot be predicted a priori. The seven models studied included three linear models (LR, ridge regression, and lasso regression), three nonlinear tree models (DT, RF, and Xtreme Gradient Boosting), and one nonlinear model (ANN).…”

Section: Resultsmentioning

confidence: 99%

See 1 more Smart Citation

Machine Learning Predictions of Oil Yields Obtained by Plastic Pyrolysis and Application to Thermodynamic Analysis

et al. 2022

Self Cite

View full text Add to dashboard Cite

Chemical recycling via thermal processes such as pyrolysis is a potentially viable way to convert mixed streams of waste plastics into usable fuels and chemicals. Unfortunately, experimentally measuring product yields for real waste streams can be time- and cost-prohibitive, and the yields are very sensitive to feed composition, especially for certain types of plastics like poly(ethylene terephthalate) (PET) and polyvinyl chloride (PVC). Models capable of predicting yields and conversion from feed composition and reaction conditions have potential as tools to prioritize resources to the most promising plastic streams and to evaluate potential preseparation strategies to improve yields. In this study, a data set consisting of 325 data points for pyrolysis of plastic feeds was collected from the open literature. The data set was divided into training and test sub data sets; the training data were used to optimize the seven different machine learning regression methods, and the testing data were used to evaluate the accuracy of the resulting models. Of the seven types of models, eXtreme Gradient Boosting (XGBoost) predicted the oil yield of the test set with the highest accuracy, corresponding to a mean absolute error (MAE) value of 9.1%. The optimized XGBoost model was then used to predict the oil yields from real waste compositions found in Municipal Recycling Facilities (MRFs) and the Rhine River. The dependence of oil yields on composition was evaluated, and strategies for removing PET and PVC were assessed as examples of how to use the model. Thermodynamic analysis of a pyrolysis system capable of achieving oil yields predicted using the machine-learned model showed that pyrolysis of Rhine River plastics should be net exergy producing under most reasonable conditions.

show abstract

Section: Resultsmentioning

confidence: 99%

“…This was repeated sequentially until all 10 groups had acted as the validation set. Cheng et al provided a visual representation of how the K-fold method works . K-fold cross validation helps to prevent overfitting or fortuitus fitting as the model must be trained on all of the data, and the final result represents an average of all 10 models.…”

Section: Methodsmentioning

confidence: 99%

Machine Learning Predictions of Oil Yields Obtained by Plastic Pyrolysis and Application to Thermodynamic Analysis

et al. 2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…10 HTL biocrude yields are highly dependent on the feedstock composition. 11 The lipid content plays a crucial role in HTL biocrude yields due to the inherent lyophilic nature of the constituent fatty acids. 12 The relationship between the lipid content and biocrude yields means that many food wastes, 5,13 some sewage sludges, 14,15 and certain types of algae cultivated on wastewater 12,16 are especially suitable for HTL.…”

Section: ■ Introductionmentioning

confidence: 99%

Emergent Chemical Behavior in Mixed Food and Lignocellulosic Green Waste Hydrothermal Liquefaction

LeClerc

Page

Tompsett

et al. 2022

ACS Sustainable Chem. Eng.

Self Cite

View full text Add to dashboard Cite

Hydrothermal liquefaction (HTL) is a promising strategy for the conversion of energy-dense waste streams to fuels. Mixed-feed HTL aggregates multiple feed streams to achieve greater scales that capitalize on local resources, hence lowering costs. The potential for new pathways and products upon feedstock blending becomes a compounding level of complexity when unlocking emergent chemistries. Food and green waste streams were evaluated under HTL conditions (300 °C, 1 h) to understand the effect of feed molecular composition on product distributions and mechanisms. Thousands of emergent chemical compounds were detected via Fourier transform ion cyclotron resonance mass spectrometry, ultimately leading to the emergence of two dominant outcomes. First, the presence of small amounts of food waste into green waste results in substantial decarboxylation and subsequent polymerization to biocrude than chars. Second, in the other limit, small amounts of green waste promote the capping of oxygenates into the biodiesel range, such as with the emergence of fatty acid methyl esters.

show abstract

“…10 HTL biocrude yields are highly dependent on feedstock compostion. 11 Lipid content plays a crucial role in HTL biocrude yields due to the inherent lyophilic nature of the constituent fatty acids. 12 The relationship between lipid content and biocrude yields means that many food wastes, 5,13 some sewage sludges, 14,15 and certain types of algae cultivated on wastewater 12,16 are especially suitable for HTL.…”

Section: Introductionmentioning

confidence: 99%

“…21 Assuming a moisture content of 50% as a reasonable value 2 implies that a mid-sized city generates sufficient waste for economical HTL, but only if multiple types of waste streams are blended together with one another. The challenge then becomes selecting the feeds to be co-fed, optimizing reaction conditions, 11,12 and predicting the corresponding biocrude yield that can be obtained using HTL.…”

Section: Introductionmentioning

confidence: 99%

Emergent chemical behavior in mixed food and lignocellulosic green waste hydrothermal liquefaction

LeClerc

Page²,

Tompsett

et al. 2022

Preprint

View full text Add to dashboard Cite

Hydrothermal liquefaction (HTL) is a promising strategy for conversion of energy-dense waste streams to fuels. Mixed-feed HTL aggregates multiple feed streams to achieve greater scales that capitalize on local resources, hence lowering costs. The potential for new pathways and products upon feedstock blending becomes a compounding level of complexity when unlocking emergent chemistries. Food and green waste streams were evaluated under HTL conditions (300 °C, 1 hr) to understand the effect of feed molecular composition on product distributions and mechanisms. Thousands of emergent chemical compounds were detected via FT-ICR MS, ultimately leading to the emergence of two dominant outcomes. First, the presence of small amounts of food waste into green waste results in substantial decarboxylation and subsequent polymerization to biocrude then chars. Second, in the other limit, small amounts of green waste promote capping of oxygenates into the biodiesel range, such as with the emergence of fatty acid methyl esters.

show abstract

Accuracy of predictions made by machine learned models for biocrude yields obtained from hydrothermal liquefaction of organic wastes

Cited by 34 publications

References 72 publications

Machine Learning Predictions of Oil Yields Obtained by Plastic Pyrolysis and Application to Thermodynamic Analysis

Machine Learning Predictions of Oil Yields Obtained by Plastic Pyrolysis and Application to Thermodynamic Analysis

Emergent Chemical Behavior in Mixed Food and Lignocellulosic Green Waste Hydrothermal Liquefaction

Emergent chemical behavior in mixed food and lignocellulosic green waste hydrothermal liquefaction

Contact Info

Product

Resources

About