Inverse and forward uncertainty quantification of models for foam–assisted enhanced oil recovery

Like many other engineering applications, oil recovery and enhanced oil recovery are sensitive to the correct administration of economic resources. Pilot tests and core flood experiments are crucial elements to design an enhanced oil recovery (EOR) project. In this direction, numerical simulators are accessible alternatives for evaluating different engineering configurations at many diverse scales (pore, laboratory, and field scales). Despite the advantages that numerical simulators possess over laboratory experiences, they are not fully protected against uncertainties. In this thesis, we show advances in analyzing uncertainties in two-–phase reservoir simulations, focusing on foam–based EOR. The methods employed in this thesis analyze how experimental uncertainties affect reservoir simulator’s responses. Our framework for model calibration and uncertainty quantification uses the Markov Chain Monte Carlo method. The parametric uncertainty is tested against identifiability studies revealing situations where posterior density distributions with high variability are related to high uncertainties and practical non–identifiability issues. The model’s reliability was evaluated by adopting surrogate models based on polynomial chaos expansion when the computational cost was an issue for the analysis. Once we quantified the model’s output variability, we performed a global sensitivity analysis to map the model’s uncertainty to the input parameters distributions. Main and total Sobol indices were used to investigate the model’s uncertainty and highlight how key parameters and their interactions influence the simulation’s output. As a consequence of the results presented in this thesis, we show a technique for parameter and uncertainty estimation that can be explored to reduce the uncertainty in foam–assisted oil recovery models, which in turn can provide reliable computational simulations. Such conclusions are of utmost interest and relevance for the design of adequate techniques for enhanced oil recovery.

Technical note: using an automated head chamber system to administer an external marker to estimate fecal output by grazing beef cattle

The objective of this experiment was to determine if titanium dioxide (TiO2) dosed through an automated head chamber system (GreenFeed; C-Lock Inc., Rapid City, SD USA) is an acceptable method to measure fecal output. The GreenFeed used on this experiment had a 2hopper bait dispensing system where hopper 1 contained alfalfa pellets marked with 1% titanium dioxide (TiO2) and hopper 2 contained unmarked alfalfa pellets. Eleven heifers (BW = 394 ± 18.7 kg) grazing a common pasture were stratified by BW and then randomized to either 1) dosed with TiO2-marked pellets by hand feeding (HFD; n = 6) or 2) dosed with TiO2-marked pellets by the GreenFeed (GFFD; n = 5) for 19 d. During the morning (0800), all heifers were offered a pelleted, high-CP supplement at 0.25% of BW in individual feeding stanchions. The HFD heifers also received 32 g of TiO2-marked pellets at morning feeding, whereas the GFFD heifers received 32 g of unmarked pellets. The GFFD heifers received a single aliquot (32 ± 1.6 g; mean ± SD) of marked pellets at their first visit to the GreenFeed each day with all subsequent 32-g aliquots providing unmarked pellets; HFD heifers received only unmarked pellets. Starting on d 15, fecal samples were collected via rectal grab at feeding and every 12 h for 5 d. A two-one sided t-test method was used to determine agreement and it was determined that the fecal output estimates by HFD and GFFD methods were similar (P = 0.04). There was a difference (P < 0.01; Bartlett’s test for homogenous variances) in variability between the dosing methods for HFD and GFFD (SD = 0.1 and 0.7, respectively). This difference in fecal output variability may have been due to variability of dosing times-of-day for the GFFD heifers (0615 ± 6.2 h) relative to the constant dosing time-of-day for HFD and constant 0800 and 2000 sampling times-of-day for all animals. This research has highlighted the potential for dosing cattle with an external marker through a GreenFeed configured with two (or more) feed hoppers because estimated fecal output means were similar; however, consideration of the increased variability of the fecal output estimates is needed for future experimental designs.

Global Sensitivity Analysis of a Homogenized Constrained Mixture Model of Arterial Growth and Remodeling

Growth and remodeling in arterial tissue have attracted considerable attention over the last decade. Mathematical models have been proposed, and computational studies with these have helped to understand the role of the different model parameters. So far it remains, however, poorly understood how much of the model output variability can be attributed to the individual input parameters and their interactions. To clarify this, we propose herein a global sensitivity analysis, based on Sobol indices, for a homogenized constrained mixture model of aortic growth and remodeling. In two representative examples, we found that 54–80% of the long term output variability resulted from only three model parameters. In our study, the two most influential parameters were the one characterizing the ability of the tissue to increase collagen production under increased stress and the one characterizing the collagen half-life time. The third most influential parameter was the one characterizing the strain-stiffening of collagen under large deformation. Our results suggest that in future computational studies it may - at least in scenarios similar to the ones studied herein - suffice to use population average values for the other parameters. Moreover, our results suggest that developing methods to measure the said three most influential parameters may be an important step towards reliable patient-specific predictions of the enlargement of abdominal aortic aneurysms in clinical practice.

Improving flood hazard datasets using a low-complexity, probabilistic floodplain mapping approach

As runoff patterns shift with a changing climate, it is critical to effectively communicate current and future flood risks, yet existing flood hazard maps are insufficient. Modifying, extending, or updating flood inundation extents is difficult, especially over large scales, because traditional floodplain mapping approaches are data and resource intensive. Low-complexity floodplain mapping techniques are promising alternatives, but their simplistic representation of process falls short of capturing inundation patterns in all situations or settings. To address these needs and deficiencies, we formalize and extend the functionality of the Height Above Nearest Drainage (i.e., HAND) floodplain mapping approach into the probHAND model by incorporating an uncertainty analysis. With publicly available datasets, the probHAND model can produce probabilistic floodplain maps for large areas relatively rapidly. We describe the modeling approach and then provide an example application in the Lake Champlain Basin, Vermont, USA. Uncertainties translate to on-the-ground changes to inundated areas, or floodplain widths, in the study area by an average of 40%. We found that the spatial extent of probable inundation captured the distribution of observed and modeled flood extents well, suggesting that low-complexity models may be sufficient for representing inundation extents in support of flood risk and conservation mapping applications, especially when uncertainties in parameter inputs and process simplifications are accounted for. To improve the accuracy of flood hazard datasets, we recommend investing limited resources in accurate topographic datasets and improved flood frequency analyses. Such investments will have the greatest impact on decreasing model output variability, therefore increasing the certainty of flood inundation extents.

Characterizing Color Quality, Damage to Artwork, and Light Intensity of Multi-Primary LEDs for Museums

Light causes damage when it is absorbed by sensitive artwork, such as oil paintings. However, light is needed to initiate vision and display artwork. The dilemma between visibility and damage, coupled with the inverse relationship between color quality and energy efficiency, poses a challenge for curators, conservators, and lighting designers in identifying optimal light sources. Multi-primary LEDs can provide great flexibility in terms of color quality, damage reduction, and energy efficiency for artwork illumination. However, there are no established metrics that quantify the output variability or highlight the trade-offs between different metrics. Here, various metrics related to museum lighting (damage, the color quality of paintings, illuminance, luminous efficacy of radiation) are analyzed using a voxelated 3-D volume. The continuous data in each dimension of the 3-D volume are converted to discrete data by identifying a significant minimum value (unit voxel). Resulting discretized 3-D volumes display the trade-offs between selected measures. It is possible to quantify the volume of the graph by summing unique voxels, which enables comparison of the performance of different light sources. The proposed representation model can be used for individual pigments or paintings with numerous pigments. The proposed method can be the foundation of a damage appearance model (DAM).

Power Generation Optimization of the Combined Cycle Power-Plant System Comprising Turbo Expander Generator and Trigen in Conjunction with the Reinforcement Learning Technique

In this paper, a method that utilizes the reinforcement learning (RL) technique is proposed to establish an optimal operation plan to obtain maximum power output from a trigen generator. Trigen is a type of combined heat and power system (CHP) that provides chilling, heating, and power generation, and the turbo expander generator (TEG) is a generator that uses the decompression energy of gas to generate electricity. If the two are combined to form a power source, a power generation system with higher efficiency can be created. However, it is very difficult to control the heat and power generation amount of TEG and trigen according to the flow rate of natural gas that changes every moment. Accordingly, a method is proposed to utilize the RL technique to determine the operation process to attain an even higher efficiency. When the TEG and trigen are configured using the RL technique, the power output can be maximized, and the power output variability can be reduced to obtain high-quality power. When using the RL technique, it was confirmed that the overall efficiency was improved by an additional 3%.