Linköping Electronic Conference Proceedings
Not a member yet
1113 research outputs found
Sort by
Numerically Efficient Degradation Model of Catalyst Layers in PEM Fuel Cells using Modelica
Get PDFDegradation of the catalyst layer is a major challenge for the commercialization of polymer electrolyte membrane fuel cells (PEMFCs). Numerical modeling helps to understand and analyse the degradation phenomena, to transfer results from accelerated stress tests (ASTs) to real applications and to optimize operating conditions regarding degradation. We implemented a typical catalyst degradation model for platinum used in literature in Modelica. A numerical analysis shows the problem of “stiffness” for these models, meaning the tremendous difference in time constants. Assuming the platinum ion concentration in the ionomer to be in quasi-equilibrium helps to reduce the “stiffness”, increases simulation speed and numerical robustness without any relevant inaccuracy. For a typical AST, the simulation speed can be more than doubled ending in a real-time factor of over 1,000. Thus, 500 hours of AST can be simulated within less than 30 minutes, which gives room for extensive analysis with the model
Calibration Workflow for Mechanical and Thermal Applications
Get PDFThe calibration of models against measurement data is important to ensure model dynamics that are close to its real-world system. Derivative-free minimizing methods can be used for any model calibration regardless of continuous differentiability requirements, and find a (local) minimum in a reasonable number of iteration steps. A user-friendly, python-based calibration Dash app to use with the cloud-based Modelica platform Modelon Impact is introduced. Basic calibration setup is done through the GUI of the app and graphical feedback (i.e. plots) is provided. Two example calibrations are shown: A mechanical Furuta pendulum that only uses Modelica Standard Library components is calibrated against real-world measurement data, and a low-fidelity heat exchanger testbench model that uses Modelon’s Air Conditioning Library is calibrated against a corresponding high-fidelity model
Manual and Automatic Identification of Similar Arguments in EFL Learner Essays
Get PDFArgument mining typically focuses on identifying argumentative units such as claim, position, evidence etc. in texts. In an educational setting, e.g. when teachers grade students’ essays, they may in addition benefit from information about the content of the arguments being used. We thus present a pilot study on the identification of similar arguments in a set of essays written by English-as-a-foreignlanguage (EFL) students. In a manual annotation study, we show that human annotators are able to assign sentences to a set of 26 reference arguments with a rather high agreement of κ > .70. In a set of experiments based on (a) unsupervised clustering and (b) supervised machine learning, we find that both approaches perform rather poorly on this task, but can be moderately improved by using a set of six meta classes instead of the more finegrained argument distinction
Process Simulation, Dimensioning and Automated Cost Optimization of CO2 Capture
Get PDFA standard process for CO2 capture has been simulated with an equilibrium-based model in Aspen HYSYS. The simulation has been combined with equipment dimensioning and cost calculation in an integrated spreadsheet facility. New in this work is that Murphree efficiencies are varied to obtain automatic optimization of absorber height and inlet temperature. The optimum process was found as the process with minimum calculated sum of capital and operational cost over 25 years. The cost optimum process parameters for the standard process were calculated to 15 m absorber packing height, 13 K minimum approach temperature and 34 °C in inlet gas temperature. This study demonstrates that it is possible to calculate the optimum packing height and inlet temperature automatically by varying the Murphree efficiency in a case study function
Multimodal sensor suite for identification of flow regimes and estimation of phase fractions and velocities – Machine Learning Algorithms in Multiphase flow metering and Control
Get PDFMultiphase flow metering is a challenging task because of the complexity of multiphase flow. In this paper, nonintrusive multiphase flow metering techniques, including machine learning (ML) / artificial intelligence models for the identification of flow regimes and estimation of flow parameters of a two-phase flow in a horizontal pipe are proposed that use data from Electrical Capacitance Tomography (ECT) and conventional measurements such as differential pressure in the pipe. The flow regimes are classified into five types, namely plug, slug, annular, wavy and stratified. Two-phase air/water flow experimental data from ECT are collected by running extensive experiments using the horizontal section of the multiphase flow rig at the University of South-Eastern Norway (USN). Exploratory data analysis (EDA) is performed on these data to extract features for use in classification and regression algorithms. Time series of normalized capacitance data from ECT sensors are used to classify flow regimes and identify flow parameters. ML techniques of Artificial Neural Network, Support Vector Machine (SVM), K-Nearest Neighbors (KNN) and Decision Tree (DT) are used to classify flow regimes by using features extracted from ECT data. The cross-correlation technique is used to estimate flow velocity using data from a twinplane ECT module. ML regression techniques are used to estimate phase fractions. Fusing data from differential pressure sensors enhances the flow regime classification. An overall system performance is given with suggestions for designing dedicated control algorithms for actuators used in multiphase flow control
Hydrodynamic study of a CO2 desorption column using computational fluid dynamics
Get PDFDesorption of CO2 from the rich amine solvent is one of the main operations in the amine-based CO2 capture process. Proper vapour and liquid flow through the packing materials would enhance the heat transfer that is needed for stripping CO2 from solvent. This is achieved by increasing the surface area of the flowing solvent by using the packing material. In this study, the created CFD (Computational Fluid dynamics) model in OpenFOAMTM was able to simulate the factors influencing TCM (Technology Centre Mongstad) desorption performance, including liquid distribution, wettability and film thickness within the packing material. Three scenarios were considered including a base case for a better understanding of the hydrodynamics in the desorption column. Two of these are to compare the influence of mass flow rates, while one is used to investigating potential improvement. Simulation revealed that introducing a deflector plate and CO2 bypass tube has a positive hydrodynamic effect in the desorption column
Dynamic Modelling and Part-Load Behavior of a Brayton Heat Pump
Get PDFAmong the environmental-friendly technologies recently proposed in the literature, high-temperature heat pumps represent a promising solution to foster the complete penetration of renewables within the power grid. Such systems may be based on closed Brayton cycles and leverage many existing components. As they are meant to provide high-temperature heat while using renewable electricity, their potential field of application ranges from industrial heating to energy storage. Several variants are currently under development to assess the feasibility of such systems in providing flexibility to the electricity grid. To do so, they need to operate in part-load conditions and quickly react when the load must be adjusted. In this regard, this study investigates the transient capabilities of Brayton heat pump technology. To this extent, a detailed transient model of a novel prototype proposed in the literature is presented, accounting for controls, thermal inertia and volume dynamics related to heat exchangers and piping. Furthermore, the model is used to assess the transient performance of the system in response to sudden load variations, which is achieved by adapting the turbomachinery operating velocities. Results show that the system can safely operate in part-load conditions with regulation times compatible with industrial needs
Future Potential Impact of Wind Energy in Sweden’s bidding area SE3
Get PDFThis research addresses the potential for increasing wind power in Sweden’s bidding area SE3. Sweden currently faces an energy imbalance, with larger production in the north and high demand in the south. Four bidding areas were introduced to incentivize energy production in the south. SE3, the largest bidding area, represents 60% of total demand. Using Seasonal Auto-Regressive Integrated Moving Average (SARIMA), historic data analysis from 2007 to 2022 is forecasted to a medium long-term future of 2035. Forecasting the observed trends reveals a potential supply deficit even under minimum demand growth scenarios made in literature. Closure of nuclear plants contributes to the shortfall, and the increasing trend in solar and wind power falls short. To study the impact wind power can have, the monthly wind patterns are analyzed, and used to calculate the power potential of different turbine capacities. Offshore areas show the highest potential for increasing wind power capacity in SE3. Economic factors, like payback time, are considered. The research concludes that there is technically and economically viable potential for wind power capacity to address the demand-supply gap by 2035. However, it depends on permitted areas, excluding built areas, UNESCO sites, and fishing routes. Future research should further explore these restrictions and address the seasonal variability in wind power to improve the understanding of the potential for wind power in the SE3 bidding area
Machine learning assisted adaptive heat load consumption forecasting in district heating network
No full textDistrict heating system often consists of a long, complex network of piping carrying heat from a power plant to the consumers. The supply temperature from the plant is either controlled by the operator from experience or a predefined curve based on the outdoor temperature. An optimized supply temperature which would be lower than the one obtained traditionally would lead to lower heat loss and reduced peak load on the power plant. In this paper, we investigate the machine learning models for heat load forecasting which is a crucial parameter in the optimizing process. Models are generated using supervised machine learning algorithms: Linear models (Linear Regression, Ridge and Gaussian Process Regressor), Random Forest Regressor, Support Vector Machine (SVM) and Long Short-Term Memory (LSTM) recurrent neural network (RNN). Data-driven models are used extensively in the literature to predict heat load prediction based on the weather and the time effect on a fixed training set, however, in this study, we model the heat load in the network in real-time scenarios i.e., adaptive training and forecasting. The model is adaptively updated as well as the training of the machine learning model in real time. It provides a “plug-and-play” solution for real-time prediction without significant pre-tuning requirements. The results of all the models are compared with various time horizons i.e., 6 hrs, 10 hrs, 24 hrs and 1 week, using the district heating data obtained for the city of Vasteras in Sweden. The performance of the prediction algorithms is evaluated using Mean Absolute Percentage Error (MAPE) and Root Mean Squared Error (RMSE). An algorithm with the best accuracy is selected based on the performance comparison. Also, models suitable for short-term and long-term forecasting are discussed towards the end of the articl
Plattform för analys av förmågan att hantera händelser med avseende på ett förändrat klimat (PAKT)
No full textFör att kunna hantera framtida klimatrelaterade händelser behöver samhällets beredskap anpassas och sannolikt utökas. Kunskapsläget över vilken förmåga som behövs för att hantera risker i ett förändrat klimat behöver sammanställas och framtida behov behöver jämföras mot nuvarande beredskapsnivåer för att identifiera lämpliga åtgärder. En analys behöver göras kring eventuella brister, utifrån vilken åtgärder sedan kan prioriteras och vidtas avseende olika delar i förmågan att hantera händelser, bland annat avseende tillhandahållande av statliga förstärkningsresurser. Denna studie analyserar vilken systematik som kan vara aktuell för att möjliggöra sådana analyser och utredningar.
Studiens övergripande syfte är att ge förslag kring hur bedömningar kan göras om framtidens behov av statliga förstärkningsresurser i syfte att möta framtidens klimatrelaterade risker. Detta inkluderar en sammanställning av kunskapsläget kring risker och scenarion avseende framtida naturhändelser utifrån ett klimat-förändringsperspektiv, med information om var mer kunskap kan hittas. Det inkluderar också att titta på metoder för att dels analysera behov av förmåga på olika nivåer att hantera de klimatrelaterade riskerna och dels hur dimensionering av förmåga kan göras för att möta behovet.
Studien består av två delar. I del 1 ges en sammanställning av kunskapsläget om riskbilder och scenarion i Sverige vad gäller framtida naturhändelser utifrån ett klimatförändringsperspektiv och tidsperioden fram till ca år 2100. I del 2 presenteras initialt ett konceptuellt ramverk för kartläggning av beredskap uppdelat i fyra huvud-domäner: Behov, Resurs, Styrning och Beroenden. Baserat på ramverket presenteras därefter en metodik för att analysera samhällets förmåga att hantera framtida händelser i ett förändrat klimat. För att kunna identifiera en lämplig framtida beredskapsnivå, behöver samhällets förmåga att hantera händelser gentemot en framtida riskbild analyseras. Del 2 fokuserar på hur relevanta myndigheter ska kunna gå tillväga för att göra en sådan analys som slutligen kan utgöra beslutsunderlag avseende framtida beredskap. Avslutningsvis ges ett exempel på hur metoden kan användas där händelsen skogsbrand, av omfattningen att nationella förstärkningsresurser i form av MSBs upphandlade helikoptrar och flygplan för skogsbrandbekämpning kan bli aktuella att nyttja, utgör själva händelsen