University of Saskatchewan Research Archive
Not a member yet
14369 research outputs found
Sort by
Neural Network Fault Recognition in Power Systems with High Penetrations of Inverter-Based Resources
Full text linkThe growing demand for renewable energy resources (RER) has led to increased integration of inverter-based resources (IBRs), into existing power distribution and transmission networks. However, RER locations are often not ideally suited for direct integration, necessitating a restructuring of the grid from a traditional radial network to a more complex mesh network topology. This transition presents challenges in terms of protection and coordination, as IBRs exhibit atypical responses to power system anomalies compared to conventional synchronous generation.
To address these challenges and support existing power system protection infrastructure, this work explores the incorporation of machine learning algorithms. Specifically, an optimized convolutional neural network (CNN) is developed for real-time application in power system protection schemes. The focus is on prioritizing key performance metrics such as recall, specificity, speed, and the reduction of computational resources required for effective protection.
The machine learning model is trained to differentiate between healthy system dynamics and hazardous conditions, such as faults, in the presence of IBRs. By analyzing data retrieved from an IEEE 34-bus 24kV distribution network, the model's application is demonstrated and its performance is evaluated.
A photovoltaic (PV) source was incorporated into the IEEE 34-bus distribution feeder model at the end of the feeder. By adding a PV source at the end of the feeder, IBR characteristics, such as its response to system anomalies can be monitored through the model.
Once the modified IEEE 34-bus distribution feeder model with the PV source was set up, various system anomalies were simulated to create a diverse dataset for training the machine learning (ML) model. These anomalies included; load rejection - a sudden and complete removal of load from the distribution network, simulating a scenario where a significant portion of the load disconnects from the grid, load addition - a sudden and significant increase in load demand, representing a scenario where new loads are connected to the grid, islanding - a scenario where the distribution feeder becomes electrically isolated from the main grid, with the PV source acting as a microgrid and supplying power to the local loads, and various types of faults, such as short-circuits or ground faults, occurring at different locations along the distribution line.
To create a diverse dataset, model parameters were varied through 50 different iterations of each simulated anomaly scenario. These parameters included the PV system's capacity, the location of the anomaly on the feeder, the severity and duration of the anomaly, and other relevant grid parameters.
For each iteration and anomaly scenario, the responses of the system were recorded, including voltage levels, current flows, and other relevant synchorphasors at the PV source's point of common coupling (PCC). These responses formed the dataset for training the ML model.
The accumulated dataset was then used to train the various ML models, including the optimized convolutional neural network (CNN), to identify patterns and hidden characteristics in the data corresponding to different system anomalies. The training process involved feeding the model with input data from the various iterations and scenarios, along with corresponding labels indicating the type of anomaly present.
By exposing the ML model to diverse scenarios and varying parameters, the model learns to generalize its understanding of system dynamics and accurately distinguish between healthy system states and hazardous conditions. The models in this work were specifically trained to recognize the various fault characteristics on the system. The trained model's ability to process time-series data and recognize anomalies from the accumulated dataset enhances power system protection infrastructure's capability to respond rapidly and accurately to various grid disturbances, ensuring the reliable and stable operation of the distribution network, especially in the presence of PV and other IBRs.
The results show that the optimized CNN outperforms traditional machine learning models used in time-series data analysis. The model's speed and reliability make it an effective tool for identifying hidden characteristics in power system data without the need for extensive manual analysis or rigid programming of existing protection relays. This capability is particularly valuable as power grids integrate a higher penetration of IBRs, where traditional protection infrastructure may not fully account for their unique responses.
The successful integration of the optimized CNN into power system protection infrastructure enhances the grid's ability to detect and respond to anomalies, such as faults, in a more efficient and accurate manner. By leveraging machine learning techniques, power system operators can better adapt to the challenges posed by the increasing presence of IBRs and ensure the continued stability and reliability of the distribution network
From Powder Turns to Snow-cloaked Trees : Learning from guides about snow accumulation in mountain forests
No full textCanada First Research Excellence FundNon-Peer ReviewedA young scientist learns about his research environment from local practitioners
PhD
No full textAnthropogenic activities, particularly coal-fired plants and coal and metal mining can increase aqueous selenium (Se) concentrations in aquatic ecosystems. The historical and present aquatic Se contamination in North America has been linked to adverse reproductive and developmental effects on fish and aquatic birds. Consequently, the environmental monitoring of Se in the receiving water bodies often focuses on potential adverse effects on aquatic vertebrates. But the magnitude to which Se bioaccumulates in the environment, potentially causing toxic effects in higher trophic levels, is dictated by Se dynamics at the base of aquatic food chains. Yet, limited understanding of the processes involved in Se uptake by primary producers, microorganisms and particulate matter and trophic transfer to different invertebrate taxa remains, generating great uncertainty in the modelling of Se in higher trophic levels. The potential influences of season on Se bioaccumulation, particularly by overwintering aquatic biota, represents an additional challenge to Se modelling and risk assessment in boreal ecosystems.
The goal of this research was to investigate Se distribution, uptake, and bioaccumulation in the periphyton-benthic macroinvertebrates food chain of a boreal lake (McClean Lake) receiving continuous low-level (< 1 µg/L) Se input from a uranium mill in northern Saskatchewan, and to refine site-specific Se risk assessment at this site. In the summers 2018 and 2019, ten sampling stations representing different degrees of uranium mill effluent exposure were selected to assess aqueous Se bioaccumulation potential in periphyton and benthic macroinvertebrates (BMI): two stations were placed at a high effluent exposure site upstream from Mclean Lake (Vulture Lake inlet and outlet), seven stations were placed on McClean Lake east basin (EB), including one near the effluent diffuser site, and another one towards Collins Creek outlet. The tenth station was a reference site located in McClean Lake west basin (WB) for comparison to background results. In winter 2021, sampling efforts were reduced, and Se bioaccumulation potential was investigated only at McClean Lake EB (sites 4, 5, 8 and 9). In summer 2019 and winter 2021, periphyton and BMI were sampled using sediment grab samples. Selenium concentrations measured at the base of the food web were subsequently incorporated into biodynamic models to estimate Se body burdens in resident fish species (northern pike, white sucker, lake whitefish and nine spine stickleback).
Higher Se concentrations in surface water (2.3 ± 0.3 µg/L), sediment (9.4 ± 1.4 µg/g d.w.), periphyton (13.0 ± 2.4 µg/g d.w.) and BMI (10.3 ± 4.2 µg/g d.w.) were observed at high-effluent exposure sites in Vulture Lake. However, Se concentrations in periphyton (16.7 ± 4.4 µg/g d.w) and some benthic detritivore taxa (6.0 ± 0.4 µg/g d.w) sampled near the effluent diffuser site on McClean Lake (Sites 4 and 9) reached levels comparable to those sampled at the high effluent exposure site (TSe > 2µg/L). Moreover, significant differences in Se bioaccumulation among benthic primary consumer taxa within sites were observed; the bivalve genus Sphaerium was considered a poor Se accumulator (1.25 ± 0.1 µg/g d.w), whereas the caddisfly genus Agrypnia was a potential Se hyper-accumulator (12.0 ± 4.7 µg/g d.w). Furthermore, the shorter deployment duration used for the artificial substrates in 2018 (5 vs. 7 weeks) seemed to have influenced Se bioaccumulation levels in BMI. Based on effluent distribution in McClean Lake traced using electrical conductivity (EC) measured with autonomous sensors (Libelium) and Se bioaccumulation in periphyton and benthic primary consumers sampled in 2018 and 2019 using artificial substrates (Hester-Dendys), a theoretical Se threshold for protection of fish populations of 0.7 ± 0.2 µg/L was derived for McClean Lake.
Additional field and laboratory studies were conducted in 2019 and 2021 to clarify Se results from periphyton and BMI sampled with the artificial substrates and the derived total aqueous Se thresholds for protection of resident fish of McClean Lake. Selenium bioconcentration in biofilms was substrate-dependent, with significantly higher Se concentrations in biofilms growing on artificial substrates (2.4 ± 1.5 µg/g d.w) relative to those collected from top-layer sediment (1.1 ± 1.3 µg/g d.w). Moreover, significantly higher Se bioconcentrations were observed in winter (3.5 ± 1.0 µg/g d.w) relative to summer (1.1 ± 1.3 µg/g d.w). Differences in Se concentrations in periphyton and BMI influenced by sampling method and season translated into differences in estimated fish body burdens and derived tissue-based aqueous Se benchmarks for McClean Lake EB. Overall, estimated and observed Se body burdens were below national and site-specific benchmarks for total Se in fish tissue and overall risks of adverse effects of Se on fish populations were considered low. Significantly lower tissue-based aqueous Se benchmarks were calculated for McClean Lake EB in summer (artificial substrates: 0.7 µg/L, grab samples: 2.5 µg/L) relative to winter (12 µg/L), suggesting a higher Se bioaccumulation potential at the former. This is the first study to date to document the direct effects of sampling strategies, low water temperature, and prolonged ice-cover period on Se bioaccumulation at the base of an aquatic food web. Collectively, the findings presented herein provide practical considerations for future Se study designs and will help inform decision-making processes related to Se risk assessment and management in boreal ecosystems receiving continuous, low-level Se input
DELIVERY OF BUTYRATE PRECURSORS IN THE GASTROINTESTINAL TRACT OF RUMINANTS AS A STRATEGY TO SUPPORT GASTROINTESTINAL FUNCTION AND DEVELOPMENT
No full textThe general hypothesis for the four experiments reported in this thesis was that increasing the intestinal supply of butyrate will stimulate gastrointestinal tract development and functionality, leading to greater growth rates and feed efficiency in ruminants. The first study evaluated the dose response of hydrogenated fat-embedded calcium gluconate (HFCG), a butyrate precursor. Growing lambs were assigned to dietary treatments containing HFCG: 0.0% (CON), 0.075% (LOW), 0.3% (MED), or 0.6% (HIGH) of the diet (DM basis). Increasing dose of HFCG linearly decreased dry matter intake (DMI), average daily gain (ADG), and final body weight (BW). In the cecum, increased dose of HFCG linearly increased the molar proportion of acetate and linearly decreased the molar proportion of propionate, while the total short-chain fatty acid (SCFA) concentration was cubically affected, where LOW had the greatest concentration. Increased butyrate molar proportion was not observed in any of the of the intestinal regions analyzed. The second study aimed to compare the post-ruminal provision of calcium gluconate (CaG) and calcium butyrate (CaB) to explore the mode of action of CaG relative to butyrate. Four treatments were used, aiming to provide the same amount of butyrate post-ruminally (excluding CON) for beef heifers: 1) negative control (CON; ruminal infusion of double-distilled water); 2) abomasal infusion of CaB (AB; 0.029% of BW); 3) abomasal infusion of CaG (AG; 0.0077% of BW); and 4) ruminal infusion of HFCG (RG; 0.0192% of BW). Relative to CON, both AB and CaG treatments (AG + RG) reduced the molar proportion of acetate and increased the molar proportion of propionate in the colon. The ruminal disappearance of total SCFA, acetate, propionate, and butyrate were increased for AB but not for CaG, relative to CON. Barrier function was not affected by treatments when measured in vivo using urinary excretion of Cr-EDTA; however, when measured using ex vivo conditions, AB tended to decrease mannitol flux across ruminal tissue relative to CON. Neither AB nor CaG increased the intestinal molar proportion of butyrate. In the third experiment, HFCG was fed to finishing heifers at 0.0% (CON), 0.09% (CG09), or 0.18% (CG18) of the diet (DM basis). Increasing the dose of HFCG did not affect feed intake or growth; however, feed efficiency and ruminal stratum corneum thickness tended to be quadratically affected by HFCG dose. Quadratic effects were observed for jejunal pH, molar proportion of acetate, valerate, and isovalerate in the colon, and meat cook loss; however, changes in the molar proportion of butyrate were not observed in the intestinal regions analyzed. Treatments did not affect carcass characteristics and descriptive sensory analysis of the meat. Finally, in the fourth experiments, tributyrin (TB) was used as a butyrate source and was provided in two forms: unprotected (UTB) and rumen protected (RTB). Eighty-four lambs were divided in two groups that were fed with different diets containing moderate (MF) or low forage (LF) inclusion, and within each group lambs were assigned to 1 of 6 treatments including a negative control (CON; no dietary inclusion of TB), dietary inclusion of RTB at 0.1% (RTB1), 0.2% (RTB2), and 0.3% (RTB3), or dietary inclusion of UTB at 0.1% (UTB1) and 0.3% (UTB3). Relative to RTB3, lambs fed UTB3 tended to finish heavier and tended to have an overall higher ADG when fed the MF diet. When lambs were fed the LF diet, RTB tended to increase DMI, and decrease feed efficiency and blood BHBA concentration relative to dietary inclusion of UTB. The use of CaG as a butyrate precursor did not increase the butyrate concentration in the intestine. When compared to the abomasal supply of CaB, post-ruminal supply of CaG did not yield the same results on ruminal SCFA disappearance, leading to the speculation that the site of exposure of butyrate in the gastrointestinal tract could yield different outcomes. The first three studies indicate that post-ruminal supply of CaG when provided at different doses, in differing diets, and ruminant species did not yield positive responses in growth and feed efficiency as the literature reports on the butyrate supply. As well, TB delivery both ruminally and post-ruminally did not increase carcass characteristics in lambs fed either with a MF or a LF diet
Make Sure we Hear them First : Integrating users and science through collaborative research projects
No full textCanada First Research Excellence FundNon-Peer ReviewedExperience of a Global Water Futures researcher in working collaboratively with local communities, designing research to meet societal needs
Cholesterol Ester Metabolism Governs Intracellular Cholesterol Crystal Formation
No full textFree cholesterol (FC) accumulation underlies cholesterol crystallization, yet the mechanistic factors contributing to this process are unknown. The presence of cholesterol crystals (CC) in atherosclerotic lesions has been recognized to increase risk of thrombosis leading to major adverse cardiovascular events. Likewise, recent studies on patients with non-alcoholic steatohepatitis (NASH) indicate that CC in hepatocyte lipid droplets distinguish NASH from benign steatosis. In vitro investigations attempting to recapitulate the effects of CC on macrophages in the context of atherosclerosis demonstrate that CC activate inflammatory signaling and remodel macrophage metabolism. A wealth of literature has reported the production of intracellular CC by incubating macrophages, smooth muscle cells, and endothelial cells with modified lipoproteins. In addition, electron microscopy has provided evidence that CC nucleate intracellularly on macrophage lipid droplet surfaces. However, few studies have explored intracellular CC formation in hepatocytes, despite a strong link to NASH. Whether hepatocyte crystals actively contribute to liver disease pathogenesis and progression remains to be investigated. Due to the ability of lipid droplets to concentrate cholesterol in the form of esters, we reasoned cholesterol esterification, subsequent lipid droplet localization, and then de-esterification will concentrate FC at the lipid droplet surface to drive cholesterol nucleation and crystal formation. Hence, we hypothesized that enzymes that mediate cholesterol esterification and enzymes that de-esterify lipid droplet cholesterol ester (CE) are required for intracellular CC formation. To investigate this, human hepatocytes and mouse macrophages were loaded with cholesterol or acetylated low-density lipoprotein (acLDL), respectively, and were imaged under polarized light to detect CC. Crystals formed reliably in a time- and dose-dependent manner in both cell types. Fluorescence imaging and polarized light microscopy reveal that these crystals are associated with lipid droplets, but not lysosomes. Inhibiting acyl-coenzyme A: cholesterol acyltransferase 1 (ACAT1) using pharmacologic and genetic methods diminished cholesterol crystallization and CE production. Remarkably, nonselective ACAT inhibition prevented plate-like CC cleft formation in hepatocytes. Also, blocking the de-esterification of lipid droplet CE with neutral CE hydrolase 1 (nCEH1) inhibitor JW480 blunted CC formation, but was more effective in hepatocytes than in macrophages. Collectively, we identify ACAT1 and nCEH1 as critical factors that function in series to drive intracellular CC formation. Our results provide needed mechanistic insight into cholesterol crystallization and highlight potential targets to prevent CC formation in human disease
Variation in resource acquisition in a food-caching mammal, the North American red squirrel (Tamiasciurus hudsonicus)
No full textLife history theory predicts that organisms will allocate their limited energy budgets towards growth, survival, and reproduction in such a way to maximize their fitness. As organisms use energy to fuel biological processes, they must also replenish their energy levels. Given the benefits of increased energetic resources such as enhanced survival and reproductive success, why there remains a high degree of variation observed in resource acquisition remains an important question in ecology and evolution. North American red squirrels (Tamiasciurus hudsonicus) in the southwest Yukon harvest and cache white spruce (Picea glauca) cones in conspicuous larders within exclusively defended individual territories, allowing individual resource acquisition to be quantified and linked to a single individual. The seasonal nature of cone availability, compounded with high interannual variability in cone crop size makes for a fluctuating resource environment for squirrels to manage through foraging and caching efforts. Previous work shows sex-specific selection acting on total cached resources accumulated in autumn, but it was unknown as to what caused variation in cache size. Using field, laboratory, and analytical tools, I investigated hypothesized behavioural, physiological, ecological, and evolutionary drivers of observed variation in resource acquisition. Despite evidence of selection acting on caching effort, I found little evidence that phenotypic variation is attributable to heritable genetic differences, leaving a large environmental component of caching success to explore. Contributing to this remaining variation were age, sex, local resource availability in the environment, days since the birth of the most recent litter (for females, specifically in years of high resource abundance), and how much food an individual has stored off-body prior to the caching season. Interestingly, neither body fat nor body mass were associated with variation in caching success, but correlations from long-term historic data revealed males with greater body mass gain between summer and autumn cached more cones. The research I present in this thesis demonstrates the roles that sex, age, environment, and interactions thereof can play in the acquisition (or not) of a key component of the life history equation: energy
Land use in the Prairie Pothole Region influences the soil bacterial community composition and relative abundance of nitrogen cycling genes
No full text© 2023 Author Helgason, and His Majesty the King in Right of Canada, as represented by the Minister of Agriculture and Agri-Food. This work is licensed under a Creative Commons
Attribution 4.0 International License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.Agriculture and Agri-Food Canada Growing Forward II program.Peer ReviewedThe undulating topography of Prairie Pothole Region of North America creates spatial and temporal variability in soil moisture and nutrient levels, affecting microbial community processes and greenhouse gas emissions. By identifying differences in soil bacterial and archaeal community composition and the abundance of nitrogen cycling genes in permanent cover versus annual crop land over two growing seasons (2017 and 2018), we were able to assess the effects of topography and land use on the functional capacity of the soil microbiome. Permanent grassland cover was associated with higher bacterial diversity in upland positions and lower diversity in low-lying depressions. Bacterial community composition was also significantly different between cultivated and permanent cover at all points along the topographic slope, with the largest effects seen in the footslope and backslope positions. Compared to permanent cover, soil from annual cropland had consistently more abundant nitrifiers, including Nitrospira in the toeslope and backslope, and Nitrososphaeraceae in the shoulder and knoll samples while soils from permanent cover had a greater abundance of several Alphaproteobacteria from Rhodospirillales and Hyphomicrobiaceae across multiple upland positions. Upland soils from annual cropland also had consistently higher abundance of both bacterial and archaeal ammonia oxidizing (amoA) genes and a higher ratio of nirK:nirS genes compared to those from permanent cover. These differences in microbial community composition were associated with higher N2O and CO2 emissions in upland soils in annual cropland; however, there were no differences in GHG emissions between the two systems in low-lying positions
Iodine Radiolabeled Mesenchymal Stem Cell (MSC)-Exosomes and Their CD73 Enzymatic Activities
Full text linkIODINE RADIOLABELLED MESENCHYMAL STEM CELL (MSC)-EXOSOMES AND THEIR CD73 ENZYMATIC ACTIVITIES Chang-Tong Yang a,b*,Ruenn Chai Laic, Sai Kiang Limc, David Chee Eng Ng a,b a Department of Nuclear Medicine and Molecular Imaging, Radiological Sciences Division, Singapore General Hospital, Outram Road, 169608 Singapore; b Duke-NUS Medical School, 8 College Road, 169857 Singapore; cInstitute of Molecular & Cell Biology, 8A Biomedical Grove #05-16 Immunos, 138648, Singapore; *[email protected] Introduction MSC-derived exosomes have shown therapeutic potential in the areas of cardiovascular, orthopaedic, ophthalmologic, immune, dermatologic diseases and radiation sickness. Efficient radioisotope-labelling of exosomes remains as a challenging process. We demonstrate iodine-131 radiolabeled exosomes using both chloramine-T and Pierce Iodination methods, and characterized I-labelled exosomes via their CD73 enzymatic activities. Experimental methodology and results Two classic radio-iodination methods have been used to label exosomes due to several advantages: relatively long half-life of I-131 (half-life 8 days) and I-124 (half-life 4.2 days, 25.6% positron emission) could enable a desired tracking kinetics of exosomes in vitro and in vivo; radiolabeling of iodine to peptides and antibodies is a well-established chemistry; the unlabeled free iodine after radio-labelling can be easily removed to reach high radiochemical purity. By using chloramine-T, the radiolabeling yield of 131I-labeled-exosomes achieved ~30-40% with a radiochemical purity > 90% after running through PD10 column purification. Using Pierce Iodination, the radiolabeling yield drops to ~15-20%, radiochemical purity achieved >90% after the same purification process. The integrity of I-labeled-exosomes is important in the reproducibility and development of exosome clinical therapeutics. No radioactive iodine was labelled to exosome for characterization of their integrity. The results showed that chloramine T radiolabeling affected the structures of I-labeled-exosomes as the CD73 enzymatic activity of I-labeled exosomes was destroyed, the particle size became much larger and caused broader exosome size distribution. While with Pierce iodination the CD73 activity drops by 50 % when compared to that of the unlabeled exosomes, and the particles kept the same size. Conclusions Using chloramine T method showed that the CD73 enzymatic activity of I-labelled exosomes was destroyed, suggesting the labeling process damaged the structure of exosomes. By comparison, using Pierce Iodination method preserved the CD73 enzymatic activity, indicating that exosomes can be radiolabeled using Pierce Iodination for in vitro and in vivo tracking and pharmacokinetic studies
