Reverse engineering of biochar

• Starting biomass and peak pyrolysis temperature jointly affect biochar properties.
• 19 different physico-chemical properties of biochar were properly modeled by GLM.
• Models reveal complex relationships between biochar properties and predictors.
• Ubiquitous non-Gaussian and non-linear attributes were accounted for in GLMs.
• Proposed correlation networks, models and web-tool can be used to engineer biochar.

This study underpins quantitative relationships that account for the combined effects that starting biomass and peak pyrolysis temperature have on physico-chemical properties of biochar. Meta-data was assembled from published data of diverse biochar samples (n = 102) to (i) obtain networks of intercorrelated properties and (ii) derive models that predict biochar properties. Assembled correlation networks provide a qualitative overview of the combinations of biochar properties likely to occur in a sample. Generalized Linear Models are constructed to account for situations of varying complexity, including: dependence of biochar properties on single or multiple predictor variables, where dependence on multiple variables can have additive and/or interactive effects; non-linear relation between the response and predictors; and non-Gaussian data distributions. The web-tool Biochar Engineering implements the derived models to maximize their utility and distribution. Provided examples illustrate the practical use of the networks, models and web-tool to engineer biochars with prescribed properties desirable for hypothetical scenarios.

Figure optionsDownload as PowerPoint slide