Characterization and Dynamics of Residual Organics in Oil Sands Fluid Fine Tailings

Sasar, Mohammadhasan; Johnston, Cliff T.; Santagata, Marika

doi:10.1021/acs.energyfuels.2c01315

Cited by 4 publications

(1 citation statement)

References 101 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In particular, it is possible that this bioturbation process can partly destabilise the sediment bed in Lake Miwasin and cause a temporary remobilisation of suspended particles and particulate organic matter that can yield to overall fluctuation in turbidity levels. Fluid tailings is a source of particulate organic matter (Sasar et al, 2022) and may increase turbidity values during periods of water column stratification. Overall, the influential input parameters (EC, WT, AT, DO, Turb) are very well captured by tree-based models in our SA for prediction of NH4 + .…”

Section: What New Insights Do the ML Models Provide Into The Underlyi...mentioning

confidence: 99%

On Robustness of the Explanatory Power of Machine Learning Models

Panigrahi,

Razavi,

Doig

et al. 2024

Preprint

View full text Add to dashboard Cite

Machine learning (ML) is increasingly considered the solution to environmental problems where only limited or no physico-chemical process understanding is available. But when there is a need to provide support for high-stake decisions, where the ability to explain possible solutions is key to their acceptability and legitimacy, ML can come short. Here, we develop a method, rooted in formal sensitivity analysis (SA), that can detect the primary controls on the outputs of ML models. Unlike many common methods for explainable artificial intelligence (XAI), this method can account for complex multi-variate distributional properties of the input-output data, commonly observed with environmental systems. We apply this approach to a suite of ML models that are developed to predict various water quality variables in a pilot-scale experimental pit lake. A critical finding is that subtle alterations in the design of an ML model (such as variations in random seed for initialization, functional class, hyperparameters, or data splitting) can lead to entirely different representational interpretations of the dependence of the outputs on explanatory inputs. Further, models based on different ML families (decision trees, connectionists, or kernels) seem to focus on different aspects of the information provided by data, although displaying similar levels of predictive power. Overall, this underscores the importance of employing ensembles of ML models when explanatory power is sought. Not doing so may compromise the ability of the analysis to deliver robust and reliable predictions, especially when generalizing to conditions beyond the training data.

show abstract

Section: What New Insights Do the ML Models Provide Into The Underlyi...mentioning

confidence: 99%

On Robustness of the Explanatory Power of Machine Learning Models

Panigrahi,

Razavi,

Doig

et al. 2024

Preprint

View full text Add to dashboard Cite

show abstract

Polymeric flocculation of oil sands tailings: Unveiling rheological indicators and dynamics of bitumen release

Sasar,

Johnston,

Santagata

2024

Can J Chem Eng

View full text Add to dashboard Cite

Polymeric flocculation is widely used to accelerate the dewatering and consolidation of oil sands fluid fine tailings (FFTs). Optimizing flocculation requires a fundamental understanding of the changes to the internal structure of the material with polymer addition. Key challenges include sensitivity of flocculation to polymer dose, mixing conditions, and composition of individual FFTs. Moreover, despite the environmental implications, little is known of the effects of flocculation on the mobility of the residual organics present in the tailings. In this paper, advanced rheological tests are used to probe the formation and development of the polymer–clay structure during flocculation in near‐real time. This is achieved through a novel setup for controlled delivery of the polymer directly into the measuring cell of a rheometer. This enables continuous monitoring of the rheological parameters during the flocculation process and yields consistent and reproducible samples. An optical monitoring system is used to relate rheological measurements to water release and to changes in the surface accumulation of the residual bitumen. The water‐release polymer dose is found to be associated with a distinct rheological response highlighting the potential use of rheometry for polymer dosing/mixing optimization in real time and on a continuous basis in thickeners and inline flocculation systems.

show abstract