Mastergradsoppgaver i teknologi - energi, klima og miljø
https://hdl.handle.net/10037/213
2024-03-29T08:56:25ZEfficient Fuel Consumption Minimization for Green Vehicle Routing Problems using a Hybrid Neural Network-Optimization Algorithm
https://hdl.handle.net/10037/30446
Fossum, Astrid<br />
Efficient routing optimization yields benefits that extend beyond mere financial
gains. In this thesis, we present a methodology that utilizes a graph convolutional neural network to facilitate the development of energy-efficient waste
collection routes. Our approach focuses on a Waste company in Tromsø, Remiks,
and uses real-life datasets, ensuring practicability and ease of implementation.
In particular, we extend the dpdp algorithm introduced by Kool et al. (2021) [1]
to minimize fuel consumption and devise routes that account for the impact of
elevation and real road distance traveled. Our findings shed light on the potential advantages and enhancements these optimized routes can offer Remiks,
including improved effectiveness and cost savings. Additionally, we identify
key areas for future research and development.<br />
2023-06-01T00:00:00ZEfficient Fuel Consumption Minimization for Green Vehicle Routing Problems using a Hybrid Neural Network-Optimization AlgorithmFossum, AstridEfficient routing optimization yields benefits that extend beyond mere financial
gains. In this thesis, we present a methodology that utilizes a graph convolutional neural network to facilitate the development of energy-efficient waste
collection routes. Our approach focuses on a Waste company in Tromsø, Remiks,
and uses real-life datasets, ensuring practicability and ease of implementation.
In particular, we extend the dpdp algorithm introduced by Kool et al. (2021) [1]
to minimize fuel consumption and devise routes that account for the impact of
elevation and real road distance traveled. Our findings shed light on the potential advantages and enhancements these optimized routes can offer Remiks,
including improved effectiveness and cost savings. Additionally, we identify
key areas for future research and development.UiT The Arctic University of NorwayUiT Norges arktiske universitetBordin, ChiaraKampffmeyer, MichaelMaster thesisMastergradsoppgaveWind resource assessment and siting for the expansion of Fakken wind park
https://hdl.handle.net/10037/30435
Bjugg, Martine<br />
Norway is anticipated to experience an energy deficit in the coming years, and one solution to address this issue is the implementation of new renewable energy sources. To mitigate the projected energy deficit, increased energy production is required. With its considerable wind potential, the country offers ample opportunities for wind energy development. Consequently, there is a growing demand for accurate methods to identify suitable areas for establishing new wind farms.
The study examines the wind conditions at Vannøya in Northern Norway and evaluates the performance of WRF and WAsP models in accurately estimating wind conditions and energy production at the existing Fakken 1 wind park.
The study found that the WRF model's resolution was insufficient to capture terrain features that influence the wind conditions, and the model tended to underestimate the high wind speeds. Initially, the WRF model provided the closest estimation to the observed production at Fakken 1, with a 9% deviation on average. However, when considering the production losses that the models cannot capture, the WAsP model provided more accurate production estimates and better captured the variations in production within the wind park.
For the proposed expansion, Fakken 2, the WAsP models estimated a yearly production of 173 GWh, while the WRF model estimated a production of 155 GWh. Additionally, the study proposed improvements in turbine positions for Fakken 2, with the potential for increased production.
The WAsP model features and computational speed made it more suitable for wind resource assessment. LiDAR proved valuable in preliminary evaluations but unreliable as a sole measurement tool.
Overall, the WAsP model was deemed more reliable and efficient for wind resource assessment and siting for the Fakken wind power plant.<br />
2023-05-31T00:00:00ZWind resource assessment and siting for the expansion of Fakken wind parkBjugg, MartineNorway is anticipated to experience an energy deficit in the coming years, and one solution to address this issue is the implementation of new renewable energy sources. To mitigate the projected energy deficit, increased energy production is required. With its considerable wind potential, the country offers ample opportunities for wind energy development. Consequently, there is a growing demand for accurate methods to identify suitable areas for establishing new wind farms.
The study examines the wind conditions at Vannøya in Northern Norway and evaluates the performance of WRF and WAsP models in accurately estimating wind conditions and energy production at the existing Fakken 1 wind park.
The study found that the WRF model's resolution was insufficient to capture terrain features that influence the wind conditions, and the model tended to underestimate the high wind speeds. Initially, the WRF model provided the closest estimation to the observed production at Fakken 1, with a 9% deviation on average. However, when considering the production losses that the models cannot capture, the WAsP model provided more accurate production estimates and better captured the variations in production within the wind park.
For the proposed expansion, Fakken 2, the WAsP models estimated a yearly production of 173 GWh, while the WRF model estimated a production of 155 GWh. Additionally, the study proposed improvements in turbine positions for Fakken 2, with the potential for increased production.
The WAsP model features and computational speed made it more suitable for wind resource assessment. LiDAR proved valuable in preliminary evaluations but unreliable as a sole measurement tool.
Overall, the WAsP model was deemed more reliable and efficient for wind resource assessment and siting for the Fakken wind power plant.UiT The Arctic University of NorwayUiT Norges arktiske universitetBirkelund, YngveMaster thesisMastergradsoppgaveOptimal sizing and placement of Electrical Vehicle charging stations to serve Battery Electric Trucks
https://hdl.handle.net/10037/30230
Isaksen, Ole-André<br />
For Norway to reach the emission limits in the Paris Agreement, a substantial amount of CO2 must be reduced. Road traffic alone accounts for a high percentage of the total emissions during 2021. This thesis will focus on electrifying the transport sector and analyzing charging infrastructure for heavy-duty electric vehicles. New charging infrastructure for heavy-duty Electric Vehicles (EVs) provides issues regarding profitability due to the currently low adaption rates. However, heavy-duty EVs use the same charging sockets as EVs. As a result, EVs may finance the charging infrastructure needed to increase the adaption of heavy-duty EVs. Projections from Norwegian grid operators suggest that the total electricity surplus is diminishing during the next years and will be negative by 2027. This highlights the importance of modeling the power system in combination with finding optimal locations for charging stations. This study uses prescriptive analytics to suggest optimal locations for charging infrastructure to maximize returned profits to motivate station builders to implement more charging stations. A soft-linking will be done with PyPSA-eur to model the power system, where the new infrastructure is added as an additional load. Analyzing the results, it is possible to see that charging infrastructure has the potential to become profitable as the adaption rate for heavy-duty EVs rise. The collaboration between the models offers an open-source tool for scholars, researchers, and planners to study how new charging infrastructure affects key components in the Norwegian power system and could be useful in modeling state-of-the-art technologies.<br />
2023-06-01T00:00:00ZOptimal sizing and placement of Electrical Vehicle charging stations to serve Battery Electric TrucksIsaksen, Ole-AndréFor Norway to reach the emission limits in the Paris Agreement, a substantial amount of CO2 must be reduced. Road traffic alone accounts for a high percentage of the total emissions during 2021. This thesis will focus on electrifying the transport sector and analyzing charging infrastructure for heavy-duty electric vehicles. New charging infrastructure for heavy-duty Electric Vehicles (EVs) provides issues regarding profitability due to the currently low adaption rates. However, heavy-duty EVs use the same charging sockets as EVs. As a result, EVs may finance the charging infrastructure needed to increase the adaption of heavy-duty EVs. Projections from Norwegian grid operators suggest that the total electricity surplus is diminishing during the next years and will be negative by 2027. This highlights the importance of modeling the power system in combination with finding optimal locations for charging stations. This study uses prescriptive analytics to suggest optimal locations for charging infrastructure to maximize returned profits to motivate station builders to implement more charging stations. A soft-linking will be done with PyPSA-eur to model the power system, where the new infrastructure is added as an additional load. Analyzing the results, it is possible to see that charging infrastructure has the potential to become profitable as the adaption rate for heavy-duty EVs rise. The collaboration between the models offers an open-source tool for scholars, researchers, and planners to study how new charging infrastructure affects key components in the Norwegian power system and could be useful in modeling state-of-the-art technologies.UiT The Arctic University of NorwayUiT Norges arktiske universitetBordin, ChiaraMishra, SambeetMaster thesisMastergradsoppgaveSmart Senja electrical network expansion modeling
https://hdl.handle.net/10037/30229
Marthinussen, Oskar<br />
The addition of variable renewable energy sources into the electrical energy systems of the world has been increasing in recent years. This form of distributed energy production with high production volatility can introduce massive challenges in operating a lower voltage distribution network. One of these affected networks is on the island of Senja in northern Norway, with an eldering radial electrical network with a single connection to the national transmission grid. In this study, prescriptive analysis of the network through mathematical optimization is implemented to investigate if there are more effective solutions to this problem other than building more electrical lines. In selected parts of the island, the electrical network experiences electrical faults of different magnitude and concern affecting 1500 hours a year. In this thesis, the model GenX is presented which prescribes solutions reducing these faults to zero while also cutting costs compared to the baseline scenario of today’s system. Results from the model indicate that simple installments of distributed power generation in conjunction with electrical energy storage drastically improve network capacity and industrial expansion opportunities. Also investigated is the feasibility of operating the electrical network on the island without any connection to the external grid. Meant as a proof of concept for the application of mathematical optimization on electrical grids in other more remote parts of the world. The model proves that investments in local electricity production positively impact the system at a fraction of the cost of building new regional distribution infrastructure. Finally, some drawbacks of the chosen analytical tool used to construct the mathematical optimization model are presented alongside selected methods applicable to apprehend or circumvent these limitations.<br />
2023-05-30T00:00:00ZSmart Senja electrical network expansion modelingMarthinussen, OskarThe addition of variable renewable energy sources into the electrical energy systems of the world has been increasing in recent years. This form of distributed energy production with high production volatility can introduce massive challenges in operating a lower voltage distribution network. One of these affected networks is on the island of Senja in northern Norway, with an eldering radial electrical network with a single connection to the national transmission grid. In this study, prescriptive analysis of the network through mathematical optimization is implemented to investigate if there are more effective solutions to this problem other than building more electrical lines. In selected parts of the island, the electrical network experiences electrical faults of different magnitude and concern affecting 1500 hours a year. In this thesis, the model GenX is presented which prescribes solutions reducing these faults to zero while also cutting costs compared to the baseline scenario of today’s system. Results from the model indicate that simple installments of distributed power generation in conjunction with electrical energy storage drastically improve network capacity and industrial expansion opportunities. Also investigated is the feasibility of operating the electrical network on the island without any connection to the external grid. Meant as a proof of concept for the application of mathematical optimization on electrical grids in other more remote parts of the world. The model proves that investments in local electricity production positively impact the system at a fraction of the cost of building new regional distribution infrastructure. Finally, some drawbacks of the chosen analytical tool used to construct the mathematical optimization model are presented alongside selected methods applicable to apprehend or circumvent these limitations.UiT The Arctic University of NorwayUiT Norges arktiske universitetChiesa, MatteoMaster thesisMastergradsoppgaveProbabilistic Wind Power Forecasting with Deep Neural Sequence Models
https://hdl.handle.net/10037/30228
Svenøe, Sofie<br />
As the world strives to fulfill the goal of zero-emission established during the Paris agreement, an increasing amount of wind power is integrated into the liberalized electricity markets. With this escalation comes the need for wind power forecasting (WPF) due to the intermittent nature of wind, and WPF is therefore becoming an important field of study to successfully incorporate wind power to the electricity market. Given the rapid growth of machine learning, deep learning and probabilistic forecasting has emerged as good alternatives for WPF due to their non-linear processing methods and their ability to model uncertainties.
In this study, two probabilistic deep learning networks and a statistical model are tested as WPF models for a 54 MW wind power park. The models are trained to predict for the day-ahead and intraday electricity market, which respectively has 12-26~h and 1-24~h as associated forecasting horizons. Historical wind power production and Numerical weather predictions (NWP) are used as input to the WPF models. NWPs are modeled from the MEPS model, operated by the Norwegian Meteorology Institute (MET Norway).
The tests show that the two neural network models Temporal Fusion Transformer, and DeepAR, produces better predictions than the statistical model, SARIMAX, for the day-ahead market. The neural networks achieved P50/P90-Risk respectively of 0.153/0.081, and 0.175/0.091. While, for the intraday market, the models DeepAR, and SARIMAX performed substantially better than the Temporal Fusion Transformer, with P50/P90-Risk of respectively, 0.111/0.056, and 0.184/0.099. This implies that Transformer sequence models perform best on long-term forecasting, whereas autoregressive models still perform best on short-term forecasting.<br />
2023-05-27T00:00:00ZProbabilistic Wind Power Forecasting with Deep Neural Sequence ModelsSvenøe, SofieAs the world strives to fulfill the goal of zero-emission established during the Paris agreement, an increasing amount of wind power is integrated into the liberalized electricity markets. With this escalation comes the need for wind power forecasting (WPF) due to the intermittent nature of wind, and WPF is therefore becoming an important field of study to successfully incorporate wind power to the electricity market. Given the rapid growth of machine learning, deep learning and probabilistic forecasting has emerged as good alternatives for WPF due to their non-linear processing methods and their ability to model uncertainties.
In this study, two probabilistic deep learning networks and a statistical model are tested as WPF models for a 54 MW wind power park. The models are trained to predict for the day-ahead and intraday electricity market, which respectively has 12-26~h and 1-24~h as associated forecasting horizons. Historical wind power production and Numerical weather predictions (NWP) are used as input to the WPF models. NWPs are modeled from the MEPS model, operated by the Norwegian Meteorology Institute (MET Norway).
The tests show that the two neural network models Temporal Fusion Transformer, and DeepAR, produces better predictions than the statistical model, SARIMAX, for the day-ahead market. The neural networks achieved P50/P90-Risk respectively of 0.153/0.081, and 0.175/0.091. While, for the intraday market, the models DeepAR, and SARIMAX performed substantially better than the Temporal Fusion Transformer, with P50/P90-Risk of respectively, 0.111/0.056, and 0.184/0.099. This implies that Transformer sequence models perform best on long-term forecasting, whereas autoregressive models still perform best on short-term forecasting.UiT The Arctic University of NorwayUiT Norges arktiske universitetBianchi, Filippo MariaChiesa, MatteoMaster thesisMastergradsoppgaveAttention-Based Neural Network for Solving the Green Vehicle Routing Problem in Waste Management
https://hdl.handle.net/10037/30103
Breimo, Marit Utheim<br />
The transport sector is a major contributor to the emission of greenhouse gases and air pollution. As urbanization and population growth continue to increase, the demand for transportation services grows, emphasizing the need for sustainable practices. Therefore, incorporating sustainability into the transport sector can effectively reduce its negative impacts on the environment and optimize the utilization of resources.
This thesis aims to address this issue by proposing a novel method that integrates neural networks into the development of a green vehicle routing model. By incorporating environmental considerations, particularly fuel consumption, into the optimization process, the model seeks to generate more sustainable route solutions. The integration of machine learning techniques, specifically an attention-based neural network, demonstrates the potential of combining machine learning with operations research for effective route optimization.
While the effectiveness of the green vehicle routing problem (GVRP) has been demonstrated in providing sustainable routes, its practical applications in real-world scenarios are still limited. Therefore, this thesis proposes the implementation of the GVRP model in a real-world waste collection routing problem. The study utilizes data obtained from Remiks, a waste management company responsible for waste collection and handling in Tromsø and Karlsøy.
The findings of this study highlight the promising synergy between machine learning and operations research for further advancements and real-world applications. Specifically, the application of the GVRP approach to waste management issues has been shown to reduce emissions during the waste collection process compared to routes optimized solely for distance minimization. The attention-based neural network approach successfully generates routes that minimize fuel consumption, outperforming distance-optimized routes. These results underscore the importance of leveraging the GVRP to address environmental challenges while enhancing decision-making efficiency and effectiveness. Overall, this thesis provides insights for developing sustainable and optimized routes for real-world problems.<br />
23.08.23: Trekkes tilbake fra visning som løsning på at oppgaven ble ferdigstilt fra studieadministrasjonen litt for fort/IHTI<br />
2023-06-04T00:00:00ZAttention-Based Neural Network for Solving the Green Vehicle Routing Problem in Waste ManagementBreimo, Marit UtheimThe transport sector is a major contributor to the emission of greenhouse gases and air pollution. As urbanization and population growth continue to increase, the demand for transportation services grows, emphasizing the need for sustainable practices. Therefore, incorporating sustainability into the transport sector can effectively reduce its negative impacts on the environment and optimize the utilization of resources.
This thesis aims to address this issue by proposing a novel method that integrates neural networks into the development of a green vehicle routing model. By incorporating environmental considerations, particularly fuel consumption, into the optimization process, the model seeks to generate more sustainable route solutions. The integration of machine learning techniques, specifically an attention-based neural network, demonstrates the potential of combining machine learning with operations research for effective route optimization.
While the effectiveness of the green vehicle routing problem (GVRP) has been demonstrated in providing sustainable routes, its practical applications in real-world scenarios are still limited. Therefore, this thesis proposes the implementation of the GVRP model in a real-world waste collection routing problem. The study utilizes data obtained from Remiks, a waste management company responsible for waste collection and handling in Tromsø and Karlsøy.
The findings of this study highlight the promising synergy between machine learning and operations research for further advancements and real-world applications. Specifically, the application of the GVRP approach to waste management issues has been shown to reduce emissions during the waste collection process compared to routes optimized solely for distance minimization. The attention-based neural network approach successfully generates routes that minimize fuel consumption, outperforming distance-optimized routes. These results underscore the importance of leveraging the GVRP to address environmental challenges while enhancing decision-making efficiency and effectiveness. Overall, this thesis provides insights for developing sustainable and optimized routes for real-world problems.UiT The Arctic University of NorwayUiT Norges arktiske universitetKampffmeyer, MichaelBordin, ChiaraMaster thesisMastergradsoppgaveUtilizing batteries in the Norwegian distribution grid
https://hdl.handle.net/10037/30102
Eidissen, Magnus<br />
Recent research indicates that photovoltaic (PV) induced overvoltage can occur in high PVpenetration low voltage distribution networks, due to reverse power flow from power injected
to the grid. Since January 2020, the number of PV installations in Norway has seen a 2.7-fold
increase in a rising trend. Simultaneously there have been several reports of grid-overvoltage
and PV curtailment in the relation of grid-connected PV systems. This study aims to
investigate the effects of batteries on peak injected power to the grid in Norwegian conditions.
Further, an economic evaluation is done for different battery usage scenarios, including PV
power self-consumption, peak shaving for reduced grid fee cost and arbitrage trading. Finally,
the study investigates what regulation measures that must be in place, to make PV battery
energy storage systems more profitable than PV-only systems. The study confirms that
batteries can be used to reduce overvoltage, also in Norwegian conditions. Additionally, the
findings indicates that for all scenarios investigated, batteries can only be considered
profitable using electricity prices from 2022 averaging at 3.88 NOK/kWh, including taxes, in
the NO1 price area. Conclusively, battery subsidies of 2666 NOK/kWh capped at 47500 NOK
in combination with a fixed feed-in tariff of 30% is suggested, and the findings shows that for
most scenarios, such a change in legislation would make batteries a more favourable
investment than PV, if a PV-system already is installed.<br />
2023-06-04T00:00:00ZUtilizing batteries in the Norwegian distribution gridEidissen, MagnusRecent research indicates that photovoltaic (PV) induced overvoltage can occur in high PVpenetration low voltage distribution networks, due to reverse power flow from power injected
to the grid. Since January 2020, the number of PV installations in Norway has seen a 2.7-fold
increase in a rising trend. Simultaneously there have been several reports of grid-overvoltage
and PV curtailment in the relation of grid-connected PV systems. This study aims to
investigate the effects of batteries on peak injected power to the grid in Norwegian conditions.
Further, an economic evaluation is done for different battery usage scenarios, including PV
power self-consumption, peak shaving for reduced grid fee cost and arbitrage trading. Finally,
the study investigates what regulation measures that must be in place, to make PV battery
energy storage systems more profitable than PV-only systems. The study confirms that
batteries can be used to reduce overvoltage, also in Norwegian conditions. Additionally, the
findings indicates that for all scenarios investigated, batteries can only be considered
profitable using electricity prices from 2022 averaging at 3.88 NOK/kWh, including taxes, in
the NO1 price area. Conclusively, battery subsidies of 2666 NOK/kWh capped at 47500 NOK
in combination with a fixed feed-in tariff of 30% is suggested, and the findings shows that for
most scenarios, such a change in legislation would make batteries a more favourable
investment than PV, if a PV-system already is installed.UiT The Arctic University of NorwayUiT Norges arktiske universitetChiesa, MatteoMaster thesisMastergradsoppgaveComparison of numerical wind modelling on hectometric and km scales at Kvitfjell wind power plant
https://hdl.handle.net/10037/30101
Rein, Kristine<br />
Evaluating the performance of numerical weather models (NWMs) is crucial to identify the
best choice of model for wind resource assessments. This thesis has investigated how well
NWMs with different horizontal resolutions reproduced the wind directions, the wind speeds
and the associated power production at a wind power plant located in complex terrain. The
evaluated models included NORA3, WRF1km, and AROME Troms and Finnmark (ATF300m)
throughout the 9 months from January to September 2022, as well as WRF111m for some case
studies of shorter duration. The models were evaluated by comparing the simulations’ results
to hub-height wind measurements. ATF300m outperformed the other models in terms of lower
errors when considering the time period as a whole, and also for a case with wind directions
similar to the main wind direction registered for the park. This could be as a result of the better
topographic representation by the model, thereby including terrain induced effects on the wind
more accurately such as orographic blocking. However, NORA3 outperformed the other
models in terms of lower errors for two cases with wind directions coming from the NW sector.
These results may suggest that 3 km grid spacing gives a sufficient representation of the wind
fields when the wind is coming from the open ocean and is less influenced by the terrain before
entering the park. Another suggestion may be that the coarser spatial resolution of NORA3
provides lower simulated wind speeds which counteracts the overestimation related to the
absence of a wind farm parameterization.<br />
2023-06-01T00:00:00ZComparison of numerical wind modelling on hectometric and km scales at Kvitfjell wind power plantRein, KristineEvaluating the performance of numerical weather models (NWMs) is crucial to identify the
best choice of model for wind resource assessments. This thesis has investigated how well
NWMs with different horizontal resolutions reproduced the wind directions, the wind speeds
and the associated power production at a wind power plant located in complex terrain. The
evaluated models included NORA3, WRF1km, and AROME Troms and Finnmark (ATF300m)
throughout the 9 months from January to September 2022, as well as WRF111m for some case
studies of shorter duration. The models were evaluated by comparing the simulations’ results
to hub-height wind measurements. ATF300m outperformed the other models in terms of lower
errors when considering the time period as a whole, and also for a case with wind directions
similar to the main wind direction registered for the park. This could be as a result of the better
topographic representation by the model, thereby including terrain induced effects on the wind
more accurately such as orographic blocking. However, NORA3 outperformed the other
models in terms of lower errors for two cases with wind directions coming from the NW sector.
These results may suggest that 3 km grid spacing gives a sufficient representation of the wind
fields when the wind is coming from the open ocean and is less influenced by the terrain before
entering the park. Another suggestion may be that the coarser spatial resolution of NORA3
provides lower simulated wind speeds which counteracts the overestimation related to the
absence of a wind farm parameterization.UiT The Arctic University of NorwayUiT Norges arktiske universitetBirkelund, YngveSamuelsen, EirikMaster thesisMastergradsoppgaveFeasibility analysis of capacity expansion in Skjerka power station based on production simulation in ProdRisk.
https://hdl.handle.net/10037/30099
Ingolfsson Dragset, Eivind<br />
The Norwegian energy system has traditionally had an energy surplus with a large share of hydro power. Due to increasing demand of power from large scale electrification, the power system is estimated to experience hours of national power deficient in 2030 even with moderate increase of consumption.
Extensive increase of variable production renewable power from wind and solar in Northern Europe has led to increased volatility in power prices and a need for larger amounts of balancing power. This thesis will research, through a socioeconomic perspective, the feasibility of two expansion alternatives with the net present value method: a 100 MW Francis turbine expansion or 100 MW reversible pump turbine expansion. Results are obtained through simulations by the optimization program ProdRisk, given three price scenarios with varying volatility and fixed average price.
Simulations results indicates increased revenue when volatility increases. Pump usage of the reversible pump turbine also increases in line with volatility and leads to larger gross energy production and revenue compared to a Francis turbine expansion of the same installed capacity.
The economic analysis utilizes the revenue and energy production difference compared to a reference simulation of todays installed capacity at Skjerka power station, of 200 MW. Due to the project investment cost, the only net present values that proved to be feasible where the ones obtained from the price scenario with largest volatility.
The reversible pump turbine expansion proved to be the most feasible option using the results obtained in simulations, despite having a higher investment cost compared to a Francis expansion. In addition, it has the ability to be used in pump mode, thus providing valuable balancing power for an improved transition to a power system with larger share of variable renewables.<br />
2023-05-31T00:00:00ZFeasibility analysis of capacity expansion in Skjerka power station based on production simulation in ProdRisk.Ingolfsson Dragset, EivindThe Norwegian energy system has traditionally had an energy surplus with a large share of hydro power. Due to increasing demand of power from large scale electrification, the power system is estimated to experience hours of national power deficient in 2030 even with moderate increase of consumption.
Extensive increase of variable production renewable power from wind and solar in Northern Europe has led to increased volatility in power prices and a need for larger amounts of balancing power. This thesis will research, through a socioeconomic perspective, the feasibility of two expansion alternatives with the net present value method: a 100 MW Francis turbine expansion or 100 MW reversible pump turbine expansion. Results are obtained through simulations by the optimization program ProdRisk, given three price scenarios with varying volatility and fixed average price.
Simulations results indicates increased revenue when volatility increases. Pump usage of the reversible pump turbine also increases in line with volatility and leads to larger gross energy production and revenue compared to a Francis turbine expansion of the same installed capacity.
The economic analysis utilizes the revenue and energy production difference compared to a reference simulation of todays installed capacity at Skjerka power station, of 200 MW. Due to the project investment cost, the only net present values that proved to be feasible where the ones obtained from the price scenario with largest volatility.
The reversible pump turbine expansion proved to be the most feasible option using the results obtained in simulations, despite having a higher investment cost compared to a Francis expansion. In addition, it has the ability to be used in pump mode, thus providing valuable balancing power for an improved transition to a power system with larger share of variable renewables.UiT The Arctic University of NorwayUiT Norges arktiske universitetMatteo, ChiesaCarl Andreas, VeieArild, HelsethMaster thesisMastergradsoppgaveForecasting Wind Turbine Production Losses due to Icing
https://hdl.handle.net/10037/30098
Andersen, Rikke Bjarnesen<br />
In order for wind energy producers to avoid the extra costs of inaccurate production estimates for the day-ahead energy market, precise forecasts of power production during the next day have to be made. For renewables such as wind energy, power production is particularly difficult to predict as it depends on the fluctuating wind speed. In addition, forecasting the power production in cold climates such as Norway is further complicated by the icing on wind turbine blades. The extra load on turbine blades due to icing leads to a decrease in power production.
Kjeller Vindteknikk has a state of the art model that estimates the production loss due to icing for new wind farms. This model uses historical weather data from the numerical weather prediction model WRF. To meet the interest from customers, Kjeller Vindteknikk wishes to further develop the icing model into IceLossForecast - An operational model providing production forecasts with icing loss for wind energy producers in cold climates. For this to work, the first step is to implement forecasting data into the IceLoss2.0 model and see if icing loss forecasts is comparable to historical IceLoss2.0 data. This is the objective of this thesis. Forecasting data from WRF forecast and the MetCoOp Ensemble Prediction model (MEPS) is implemented, and their results are compared to each other and to the current IceLoss2.0 estimates.
In this thesis MEPS and WRF forecasts has been implemented to IceLoss2.0 successfully. Results from one turbine at Kvitfjell Wind Farm has shown promising results both for WRF and MEPS that are comparable to the traditional IceLoss2.0 estimates with historical WRF data.<br />
2023-05-31T00:00:00ZForecasting Wind Turbine Production Losses due to IcingAndersen, Rikke BjarnesenIn order for wind energy producers to avoid the extra costs of inaccurate production estimates for the day-ahead energy market, precise forecasts of power production during the next day have to be made. For renewables such as wind energy, power production is particularly difficult to predict as it depends on the fluctuating wind speed. In addition, forecasting the power production in cold climates such as Norway is further complicated by the icing on wind turbine blades. The extra load on turbine blades due to icing leads to a decrease in power production.
Kjeller Vindteknikk has a state of the art model that estimates the production loss due to icing for new wind farms. This model uses historical weather data from the numerical weather prediction model WRF. To meet the interest from customers, Kjeller Vindteknikk wishes to further develop the icing model into IceLossForecast - An operational model providing production forecasts with icing loss for wind energy producers in cold climates. For this to work, the first step is to implement forecasting data into the IceLoss2.0 model and see if icing loss forecasts is comparable to historical IceLoss2.0 data. This is the objective of this thesis. Forecasting data from WRF forecast and the MetCoOp Ensemble Prediction model (MEPS) is implemented, and their results are compared to each other and to the current IceLoss2.0 estimates.
In this thesis MEPS and WRF forecasts has been implemented to IceLoss2.0 successfully. Results from one turbine at Kvitfjell Wind Farm has shown promising results both for WRF and MEPS that are comparable to the traditional IceLoss2.0 estimates with historical WRF data.UiT The Arctic University of NorwayUiT Norges arktiske universitetGrini, SigbjørnBirkelund, YngveMaster thesisMastergradsoppgave