Premium
Sensitivity Analysis Applied to Unsaturated Flow Modeling of a Retorted Oil Shale Pile
Author(s) -
Freshley Mark D.,
Reisenauer A. E.,
Gee G. W.
Publication year - 1985
Publication title -
soil science society of america journal
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.836
H-Index - 168
eISSN - 1435-0661
pISSN - 0361-5995
DOI - 10.2136/sssaj1985.03615995004900010005x
Subject(s) - oil shale , hydraulic conductivity , sensitivity (control systems) , environmental science , petroleum engineering , infiltration (hvac) , permeability (electromagnetism) , flow (mathematics) , pile , geology , shale oil , evapotranspiration , geotechnical engineering , soil science , soil water , engineering , mechanics , biology , thermodynamics , paleontology , ecology , physics , genetics , electronic engineering , membrane
Commercial recovery of oil from oil shale will require disposal of large quantities of retorted or spent shale. Because no commercial disposal piles have been constructed and monitored, uncertainties exist in predicting water movement through large spent‐shale disposal piles. To better understand these uncertainties, we investigated water movement through spent shale by numerical modeling. A first‐order sensitivity analysis was used as a part of the numerical investigation of water movement through a hypothetical pile of spent oil shale. The sensitivity analysis was useful in identifying the most important components of the flow model. Results indicate that the flow model is sensitive to parameters that directly supply water to or control extraction of water from the profile, namely the initial condition of spent shale, precipitation, and estimates of potential evapotranspiration. The flow model is less sensitive to parameters that control water movement, specifically hydraulic conductivity of the spent shale and soil cover. The sensitivity analysis demonstrates that the flow model is also sensitive to parameters describing the sink term, namely rooting depth and rooting density. To improve predictions of the flow model, the input parameters contributing the most to model sensitivity should be measured carefully; parameters that produce less model sensitivity can be estimated without as much effort.