summaries of ongoing research
Claire Beveridge- University of Washington
Research- Improved Hydrologic and Sedimentation Modeling for Hydropower Systems: A Case Study on the Elwha River Dams
The removal of the two massive hydropower dams along the Elwha River in Washington State’s Olympic National Park completed in 2013 marked the largest dam removal in history. The interplay among biophysical, societal, and infrastructure aspects in the Elwha hydropower system reached a critical stress point leading to dam removal primarily for aquatic habitat restoration purposes. Retrospective analysis of this monumental event in hydropower history can enhance knowledge on the life-cycle of hydropower dams with respect to environmental, ecological, and societal factors.
Construction of the dams, Glines Canyon and Elwha, was completed in 1913 and 1937, respectively, and by the time of their removal in 2013 approximately 21 million cubic meters of sediment was impounded behind them. During their lifetime, upstream sediment deposition impacted hydropower capacity and operations while downstream sediment starvation led to net river scour of gravel, loss of fish habitat, and reduced flow variability. Historical data on the system hydrology and sediment collected in the reservoirs can provide a unique opportunity to test sediment supply and transport models.
Existing sediment supply and transport models used for reservoir operations are based on simplifying assumptions that poorly represent extremely complex conditions in river and reservoir systems. Models are especially deficient in steep mountainous regions. The goal of my research is to develop a novel continuous modeling approach to sediment supply and transport using state-of-the-art tools that allow for process-based representations of relevant and dominant geomorphic phenomenon driven by hydrology. This approach can be used for sustainable planning and management of reservoir sedimentation in consideration of social and ecological factors along with changing climatic and environmental conditions.
Historic Elwha River landscape sediment supply and transport will be simulated using a new modeling framework for earth surface dynamics called Landlab (http://landlab.readthedocs.org) driven by hydrology (surface runoff, streamflow and subsurface flow pore-pressure) from the Distributed Hydrology-Soils-Vegetation Model (DHSVHM; http://www.hydro.washington.edu/Lettenmaier/Models/DHSVM/). Historical climate, soil, vegetation, and hydrologic data are required to calibrate and run the models and will be obtained from US Geological Survey, US Department of Agriculture, and National Climatic Data Center. Historical simulations will be conducted to calibrate the model to observations following model spin-up runs. Then, a shuffle-deck approach will be conducted by changing the order of the seasonal storms while keeping annual total precipitation unchanged to examine the frequency and magnitude behavior of sediment input in the Elwha River. The role of uncertainty in biophysical model parameters (e.g., vegetation root cohesion) on sediment yields will be examined using ranges of select parameters. Model code and documentation will be provided on the Landlab model webpage.
Research- Improved Hydrologic and Sedimentation Modeling for Hydropower Systems: A Case Study on the Elwha River Dams
The removal of the two massive hydropower dams along the Elwha River in Washington State’s Olympic National Park completed in 2013 marked the largest dam removal in history. The interplay among biophysical, societal, and infrastructure aspects in the Elwha hydropower system reached a critical stress point leading to dam removal primarily for aquatic habitat restoration purposes. Retrospective analysis of this monumental event in hydropower history can enhance knowledge on the life-cycle of hydropower dams with respect to environmental, ecological, and societal factors.
Construction of the dams, Glines Canyon and Elwha, was completed in 1913 and 1937, respectively, and by the time of their removal in 2013 approximately 21 million cubic meters of sediment was impounded behind them. During their lifetime, upstream sediment deposition impacted hydropower capacity and operations while downstream sediment starvation led to net river scour of gravel, loss of fish habitat, and reduced flow variability. Historical data on the system hydrology and sediment collected in the reservoirs can provide a unique opportunity to test sediment supply and transport models.
Existing sediment supply and transport models used for reservoir operations are based on simplifying assumptions that poorly represent extremely complex conditions in river and reservoir systems. Models are especially deficient in steep mountainous regions. The goal of my research is to develop a novel continuous modeling approach to sediment supply and transport using state-of-the-art tools that allow for process-based representations of relevant and dominant geomorphic phenomenon driven by hydrology. This approach can be used for sustainable planning and management of reservoir sedimentation in consideration of social and ecological factors along with changing climatic and environmental conditions.
Historic Elwha River landscape sediment supply and transport will be simulated using a new modeling framework for earth surface dynamics called Landlab (http://landlab.readthedocs.org) driven by hydrology (surface runoff, streamflow and subsurface flow pore-pressure) from the Distributed Hydrology-Soils-Vegetation Model (DHSVHM; http://www.hydro.washington.edu/Lettenmaier/Models/DHSVM/). Historical climate, soil, vegetation, and hydrologic data are required to calibrate and run the models and will be obtained from US Geological Survey, US Department of Agriculture, and National Climatic Data Center. Historical simulations will be conducted to calibrate the model to observations following model spin-up runs. Then, a shuffle-deck approach will be conducted by changing the order of the seasonal storms while keeping annual total precipitation unchanged to examine the frequency and magnitude behavior of sediment input in the Elwha River. The role of uncertainty in biophysical model parameters (e.g., vegetation root cohesion) on sediment yields will be examined using ranges of select parameters. Model code and documentation will be provided on the Landlab model webpage.
Fred Carter- Lehigh University
Research- Pump-storage Hydropower Design in a Wastewater Treatment Facility with an Aerating Runner
Pumped water storage as an energy solution is an increasingly popular topic for research. Pumping water in off-peak consumption hours to help generate power during peak hours has been shown to be a viable and even profitable solution. In order to improve on the current technology, the proposed research investigates a specific pumped storage application for wastewater treatment plants, focusing on a micro-hydro pump storage system. This project involves making the technology readily available to water treatment plants, evaluating a novel aerating runner’s ability to treat wastewater, and optimizing the system design. This project also serves as an extension of a previous Hydro Research Foundation study.
The research group extensively studies micro-hydro systems and will provide the knowledge and tools to make this research a success. The project will entail speaking with wastewater treatment plant management regarding process variables, constraints, and requirements to formulate an appropriate and innovative micro-hydro pump-storage system, which will be designed and produced at our facilities. The project’s breadth is to evaluate a micro-hydro pump-storage system’s implementation in a wastewater treatment facility, and explore the advantages such a system entails. This project would involve CFD and FEA tools to create an optimized system. The project’s ultimate achievement is to derive an aerating runner capable of both the turbine and pump process. Substantial attention will be paid to the system cost analysis, including a break-even analysis and a feasibility assessment based on potential customer feedback for actual deployment.
Research- Pump-storage Hydropower Design in a Wastewater Treatment Facility with an Aerating Runner
Pumped water storage as an energy solution is an increasingly popular topic for research. Pumping water in off-peak consumption hours to help generate power during peak hours has been shown to be a viable and even profitable solution. In order to improve on the current technology, the proposed research investigates a specific pumped storage application for wastewater treatment plants, focusing on a micro-hydro pump storage system. This project involves making the technology readily available to water treatment plants, evaluating a novel aerating runner’s ability to treat wastewater, and optimizing the system design. This project also serves as an extension of a previous Hydro Research Foundation study.
The research group extensively studies micro-hydro systems and will provide the knowledge and tools to make this research a success. The project will entail speaking with wastewater treatment plant management regarding process variables, constraints, and requirements to formulate an appropriate and innovative micro-hydro pump-storage system, which will be designed and produced at our facilities. The project’s breadth is to evaluate a micro-hydro pump-storage system’s implementation in a wastewater treatment facility, and explore the advantages such a system entails. This project would involve CFD and FEA tools to create an optimized system. The project’s ultimate achievement is to derive an aerating runner capable of both the turbine and pump process. Substantial attention will be paid to the system cost analysis, including a break-even analysis and a feasibility assessment based on potential customer feedback for actual deployment.
Joy Foley- Colorado School of Mines
Research- Detecting lnternal Erosion in Earthen Dams
Hydropower depends on the reliability and integrity of water-retention infrastructure, making the understanding of common dam failure mechanisms crucial to the longevity of any hydropower project. Internal erosion is one of the most significant failure modes in earthen dams and can result in loss of life and property. lnternal erosion occurs over four distinct phases: lnitiation, Continuation, Progression, and Breach. Although there exists external evidence of the lnitiation and Continuation phases, the progression phase is not externally apparent, making it difficult to accurately assess the risk of breach. This research project will help reduce the uncertainty of the stability of hydropower infrastructure and reduce the risk of failure due to internal erosion.
The goal of this proposed research is to investigate the initiation and progression of internal erosion (piping) within materials representative of a typical earthen dam and to investigate methods for detecting internal erosion. This will involve answering three questions: (1) What is the general behavior of internal erosion within a homogeneous sand, and how do different material properties (e.9. gradation and compaction) influence its initiation and progression? (2) How do pore water pressures at fixed locations change temporally as internal erosion initiates and propagates upstream? Can this behavior be used to detect internal erosion? (3) What is the general behavior of internal erosion within a representative layered soil sample, consisting of granular material with an overlying cohesive soil layer?
All tests will be carried out in a 60-cm tall 30.5-cm diameter rigid permeameter with five equally-spaced ports extending vertically along one side designed for measuring pore water pressures. The permeameter will be filled with saturated soil and will be placed in a load frame, which will provide stability and will record vertical displacement of the soil. Water will be flowed upwards through the sand sample, and pore pressure transducers mounted on the side of the permeameter will measure pore water pressure continuously. Upstream water head will be gradually increased while downstream head will be held constant until internal erosion initiates and progresses backwards towards the upstream face of the soil, which can visually be observed by the occurrence of sand boils. Total applied head differences, pore pressures, locally measured hydraulic gradients, soil displacement, and visual observations will be combined in order to understand the internal erosion mechanisms and determine if pore water pressure measurements can be used to detect erosion.
The permeameter will initially be filled with homogeneous sands and flow tests will be conducted to induce internal erosion. Multiple tests will be performed for sands of varying gradations and compactions. The results will be used to assess the relationship between gradation and the critical gradient at which initiation and progression occur, as well as the effects of compaction. The permeameter will then be filled partly with homogeneous sand. A clay cover will be placed on top of the sand on the downstream face (the top) of the permeameter. Multiple tests will be performed with clay caps of varying thicknesses and material properties.
The test result data and analysis will be written up in the form of a journal paper along with recommendations for next steps for non-invasive testing of internal erosion in earthen dams, including Direct Current and harmonic resistivity testing for which the permeameter is already equipped.
Research- Detecting lnternal Erosion in Earthen Dams
Hydropower depends on the reliability and integrity of water-retention infrastructure, making the understanding of common dam failure mechanisms crucial to the longevity of any hydropower project. Internal erosion is one of the most significant failure modes in earthen dams and can result in loss of life and property. lnternal erosion occurs over four distinct phases: lnitiation, Continuation, Progression, and Breach. Although there exists external evidence of the lnitiation and Continuation phases, the progression phase is not externally apparent, making it difficult to accurately assess the risk of breach. This research project will help reduce the uncertainty of the stability of hydropower infrastructure and reduce the risk of failure due to internal erosion.
The goal of this proposed research is to investigate the initiation and progression of internal erosion (piping) within materials representative of a typical earthen dam and to investigate methods for detecting internal erosion. This will involve answering three questions: (1) What is the general behavior of internal erosion within a homogeneous sand, and how do different material properties (e.9. gradation and compaction) influence its initiation and progression? (2) How do pore water pressures at fixed locations change temporally as internal erosion initiates and propagates upstream? Can this behavior be used to detect internal erosion? (3) What is the general behavior of internal erosion within a representative layered soil sample, consisting of granular material with an overlying cohesive soil layer?
All tests will be carried out in a 60-cm tall 30.5-cm diameter rigid permeameter with five equally-spaced ports extending vertically along one side designed for measuring pore water pressures. The permeameter will be filled with saturated soil and will be placed in a load frame, which will provide stability and will record vertical displacement of the soil. Water will be flowed upwards through the sand sample, and pore pressure transducers mounted on the side of the permeameter will measure pore water pressure continuously. Upstream water head will be gradually increased while downstream head will be held constant until internal erosion initiates and progresses backwards towards the upstream face of the soil, which can visually be observed by the occurrence of sand boils. Total applied head differences, pore pressures, locally measured hydraulic gradients, soil displacement, and visual observations will be combined in order to understand the internal erosion mechanisms and determine if pore water pressure measurements can be used to detect erosion.
The permeameter will initially be filled with homogeneous sands and flow tests will be conducted to induce internal erosion. Multiple tests will be performed for sands of varying gradations and compactions. The results will be used to assess the relationship between gradation and the critical gradient at which initiation and progression occur, as well as the effects of compaction. The permeameter will then be filled partly with homogeneous sand. A clay cover will be placed on top of the sand on the downstream face (the top) of the permeameter. Multiple tests will be performed with clay caps of varying thicknesses and material properties.
The test result data and analysis will be written up in the form of a journal paper along with recommendations for next steps for non-invasive testing of internal erosion in earthen dams, including Direct Current and harmonic resistivity testing for which the permeameter is already equipped.
Sudershan Gangrade - University of Tennessee
Research - Potential risks on critical energy-water infrastructure, specifically major dams and nuclear power plants
Critical infrastructures require highly-conservative design criteria that may withstand the most severe hydro-climatic extreme events. These design criteria are based on probable maximum precipitation (PMP) and probable maximum flood (PMF) estimates. Theoretically, PMF is upper bound of flood that can occur over a given region and occurs under an adverse set of hydro-meteorological conditions. Since the magnitude and frequency of extreme precipitation and flood are likely to increase in a warming climate, it enhances the need to more accurately quantify the risks from PMP and PMF under climate change.
My research proposal for Hydro Research Foundation Research Awards 2017 involves utilization of an integrated high-resolution process-based hydrometeorologic modeling framework to 1) develop a multi-ensemble future PMF projection based on best-available historic observations and future climate change simulations, 2) quantify the effects of climate change on PMF, and 3) assess the potential flood hazards on major energy-water facilities using probabilistic flood inundation maps. The end product will be multi-ensemble, high resolution probabilistic flood maps that illustrate uncertainties associated with model inputs, parameterization and hydro-meteorologic factors which can better inform decision-making for future emergency preparation.
Research - Potential risks on critical energy-water infrastructure, specifically major dams and nuclear power plants
Critical infrastructures require highly-conservative design criteria that may withstand the most severe hydro-climatic extreme events. These design criteria are based on probable maximum precipitation (PMP) and probable maximum flood (PMF) estimates. Theoretically, PMF is upper bound of flood that can occur over a given region and occurs under an adverse set of hydro-meteorological conditions. Since the magnitude and frequency of extreme precipitation and flood are likely to increase in a warming climate, it enhances the need to more accurately quantify the risks from PMP and PMF under climate change.
My research proposal for Hydro Research Foundation Research Awards 2017 involves utilization of an integrated high-resolution process-based hydrometeorologic modeling framework to 1) develop a multi-ensemble future PMF projection based on best-available historic observations and future climate change simulations, 2) quantify the effects of climate change on PMF, and 3) assess the potential flood hazards on major energy-water facilities using probabilistic flood inundation maps. The end product will be multi-ensemble, high resolution probabilistic flood maps that illustrate uncertainties associated with model inputs, parameterization and hydro-meteorologic factors which can better inform decision-making for future emergency preparation.
Seth Gregg- Colorado School of Mines
Research- Advanced Analysis of Hydroelectric Turbine Condition Monitoring Data for Early Damage Detection
Condition monitoring is an important part of reducing operations and maintenance costs in hydroelectric plants. By monitoring the turbines running parameters in real time, both short and long term operating decisions can be made to prevent catastrophic failure and lengthen the run time between maintenance. Many hydroelectric turbines are instrumented to collect condition monitoring data, however, analysis of the data can be a lengthy and complex process. In order to make real time operational decisions, analysis of the condition monitoring data must be done in an ongoing way using monitoring software and damage detection algorithms that look at real time health of the turbine. A key part of understanding the health of a turbine is to understand when it is undergoing cavitation.
Cavitation damage on turbines is caused when vapor bubbles implode on the surface of the guide vanes, wicket gates, turbine runners, or draft tubes. Cavitation can be relatively harmless or quite destructive and leads to efficiency losses, severe vibration and lengthy repair outages. Differentiating between types of cavitation and predicting the health of the turbine remains a challenge and this research aims to aid this effort by developing an algorithm for detecting the different types of cavitation and classifying its destructiveness. The algorithm will be developed by applying statistical learning (machine learning) techniques to existing condition monitoring and running data collected on turbines undergoing cavitation.
The first task for creating an effective algorithm will be to correlate the many different types of condition monitoring data – from turbine vibration and flow to generator power and speed – with cavitation damage and turbine health. This process is used to features of the data that correlate well with damaging cavitation and turbine health. Once the correct data features are chosen, we will apply statistical methods that learn how to identify damaging cavitation from condition monitoring data. Since operating characteristics are different for every turbine, we will develop an algorithm that automates the learning process and creates cavitation and health parameters specific to each turbine being monitored.
Research- Advanced Analysis of Hydroelectric Turbine Condition Monitoring Data for Early Damage Detection
Condition monitoring is an important part of reducing operations and maintenance costs in hydroelectric plants. By monitoring the turbines running parameters in real time, both short and long term operating decisions can be made to prevent catastrophic failure and lengthen the run time between maintenance. Many hydroelectric turbines are instrumented to collect condition monitoring data, however, analysis of the data can be a lengthy and complex process. In order to make real time operational decisions, analysis of the condition monitoring data must be done in an ongoing way using monitoring software and damage detection algorithms that look at real time health of the turbine. A key part of understanding the health of a turbine is to understand when it is undergoing cavitation.
Cavitation damage on turbines is caused when vapor bubbles implode on the surface of the guide vanes, wicket gates, turbine runners, or draft tubes. Cavitation can be relatively harmless or quite destructive and leads to efficiency losses, severe vibration and lengthy repair outages. Differentiating between types of cavitation and predicting the health of the turbine remains a challenge and this research aims to aid this effort by developing an algorithm for detecting the different types of cavitation and classifying its destructiveness. The algorithm will be developed by applying statistical learning (machine learning) techniques to existing condition monitoring and running data collected on turbines undergoing cavitation.
The first task for creating an effective algorithm will be to correlate the many different types of condition monitoring data – from turbine vibration and flow to generator power and speed – with cavitation damage and turbine health. This process is used to features of the data that correlate well with damaging cavitation and turbine health. Once the correct data features are chosen, we will apply statistical methods that learn how to identify damaging cavitation from condition monitoring data. Since operating characteristics are different for every turbine, we will develop an algorithm that automates the learning process and creates cavitation and health parameters specific to each turbine being monitored.
Surabhi Karambelkar - University of Arizona
Research - Hydropower Operations on the Colorado River Basin: Institutional Analysis of Opportunities and Constraints
40 million people, 4.5 million acres of farmland, 22 federally recognized tribes, and 4 million U.S. homes depend on the Colorado River for their water and energy needs. Dams, hydropower stations, canals, aqueducts, and laws, were put in place to control this vast plumbing system. However, climate change is questioning the power of these man-made controls as reservoirs on this river—built to make the desert bloom—are predicted to go dry by 2050 in the absence of effective drought management.
Given this imminent risk, the government, policy and scholarly community are focusing efforts on identifying the impacts of the drought and strategies to mitigate them. However, not all water uses have received equal attention in these efforts. Agricultural and municipal water needs have dominated the research agenda, while other uses such as hydropower generation have received limited attention. Moreover, the studies on hydropower that do exist focus on the technical calculations of percent reduction in mega-watts generated under sustained drought conditions in the Colorado River.
Hydropower, however, is more than just the mega-watts of power generated; on the Colorado River, it is also the revenue that pays for the operation and maintenance of dams, and subsidizes water for the 40 million people and 4.5 million acres of farmland. This relationship between hydropower and other water users, is an outcome of the particular way in which water laws are structured on the Colorado River. Moreover, while drought is an increasingly obvious factor influencing hydropower operations, changes in energy and environmental law and policy—manifesting in the form of rapid changes in electricity markets, increasing penetration of renewable energy sources (especially distributed generation), and growing recognition of environmental water needs— are playing an equally important role in influencing hydropower generation.
Given this background, then, this project aims to understand how water, energy, and environmental laws and policies influence hydropower generation on the Colorado River Basin. The intent of this project is to conduct a legal and policy analysis of the historic and ongoing changes in key water, energy, and environmental laws, to understand how these laws create constraints and opportunities for hydropower generation.
This project will use semi-structured interviews in conjunction with legal and policy analysis to develop a more nuanced understanding of constraints and opportunities. Stakeholders that will be interviewed will include the Bureau of Reclamation, Western Area Power Administration, and the Colorado Energy Distributors Association. The expected project output will be a policy report that provides an overview of the legal and policy opportunities and constraints for hydropower generation on the Colorado River.
This project, will not only help address a scholarly research gap on the Colorado River Basin, but also provide policy-relevant information that can be used in decision-making by actors in government agencies, civil society, and industry.
Research - Hydropower Operations on the Colorado River Basin: Institutional Analysis of Opportunities and Constraints
40 million people, 4.5 million acres of farmland, 22 federally recognized tribes, and 4 million U.S. homes depend on the Colorado River for their water and energy needs. Dams, hydropower stations, canals, aqueducts, and laws, were put in place to control this vast plumbing system. However, climate change is questioning the power of these man-made controls as reservoirs on this river—built to make the desert bloom—are predicted to go dry by 2050 in the absence of effective drought management.
Given this imminent risk, the government, policy and scholarly community are focusing efforts on identifying the impacts of the drought and strategies to mitigate them. However, not all water uses have received equal attention in these efforts. Agricultural and municipal water needs have dominated the research agenda, while other uses such as hydropower generation have received limited attention. Moreover, the studies on hydropower that do exist focus on the technical calculations of percent reduction in mega-watts generated under sustained drought conditions in the Colorado River.
Hydropower, however, is more than just the mega-watts of power generated; on the Colorado River, it is also the revenue that pays for the operation and maintenance of dams, and subsidizes water for the 40 million people and 4.5 million acres of farmland. This relationship between hydropower and other water users, is an outcome of the particular way in which water laws are structured on the Colorado River. Moreover, while drought is an increasingly obvious factor influencing hydropower operations, changes in energy and environmental law and policy—manifesting in the form of rapid changes in electricity markets, increasing penetration of renewable energy sources (especially distributed generation), and growing recognition of environmental water needs— are playing an equally important role in influencing hydropower generation.
Given this background, then, this project aims to understand how water, energy, and environmental laws and policies influence hydropower generation on the Colorado River Basin. The intent of this project is to conduct a legal and policy analysis of the historic and ongoing changes in key water, energy, and environmental laws, to understand how these laws create constraints and opportunities for hydropower generation.
This project will use semi-structured interviews in conjunction with legal and policy analysis to develop a more nuanced understanding of constraints and opportunities. Stakeholders that will be interviewed will include the Bureau of Reclamation, Western Area Power Administration, and the Colorado Energy Distributors Association. The expected project output will be a policy report that provides an overview of the legal and policy opportunities and constraints for hydropower generation on the Colorado River.
This project, will not only help address a scholarly research gap on the Colorado River Basin, but also provide policy-relevant information that can be used in decision-making by actors in government agencies, civil society, and industry.
Kevin Kircher- Cornell University
Research- Coupling Hydropower, PV and Flexible Loads in a Carbon P Free Microgrid
The State of Hawaii currently imports 85% of its food and relies on fossil fuels for 95% of its energy. Most of these fuels are imported, making Hawai'ian electricity costs (about 42 ¢/kWh) the most expensive in the nation. These facts, coupled with the state's excellent water and solar resources, make sustainable energy and agriculture projects in Hawai'i extremely attractive.
The state's regulatory environment, however, often makes grid-connected renewable projects politically infeasible: the utilities place hard caps on the amount of distributed generation capacity allowable in their networks. Many communities have already reached these caps. Additionally, much arable land is located far from existing electrical distribution infrastructure, making the cost of interconnection significant. Since agricultural activities such as irrigation, refrigeration and processing are often energy-intensive, the lack of affordable access to electricity is a challenge.
As a step toward addressing these issues, we conducted an initial hydropower site assessment along the Lower Hamakua Ditch on the Big Island. This irrigation canal was constructed for a sugarcane plantation that has long since left the island. The State now maintains the canal in service of several farms and homes, but the revenues it generates are insufficient to offset the operation and maintenance costs. A small, autonomous hydropower plant could bring the ditch into the black while adding significant value to the community by powering a proposed agricultural processing facility. Access to affordable processing facilities is a major bottleneck for local farmers.
In the initial site assessment, we found two attractive locations: a low-head, high-flow site near the top of the ditch, and a low-flow, high-head site with a large penstock at the lower end. After techno-economic analysis of both sites, we suspect that a Pelton turbine below the terminal reservoir is the most attractive option.
In this work, we will extend the initial site assessment to design a microgrid including a lower reservoir and a pumped-storage system, a PV array, and direct control of flexible electric loads. This design will include control algorithms to optimally balance seasonal variations in water flow rates; seasonal, diurnal and real-time fluctuations in PV output; and the time-varying costs and constraints of a diverse portfolio of flexible loads. Estimation and prediction algorithms will be required in order to anticipate changes in resource availability and in electricity demand. This work will be conducted in collaboration with Hawai'ian partners, who have taken steps toward implementing such a system.
We hope that the development of this microgrid will highlight the potential of small-scale hydropower and pumped storage to 'firm up' the less reliable (but more scalable) renewables.
Research- Coupling Hydropower, PV and Flexible Loads in a Carbon P Free Microgrid
The State of Hawaii currently imports 85% of its food and relies on fossil fuels for 95% of its energy. Most of these fuels are imported, making Hawai'ian electricity costs (about 42 ¢/kWh) the most expensive in the nation. These facts, coupled with the state's excellent water and solar resources, make sustainable energy and agriculture projects in Hawai'i extremely attractive.
The state's regulatory environment, however, often makes grid-connected renewable projects politically infeasible: the utilities place hard caps on the amount of distributed generation capacity allowable in their networks. Many communities have already reached these caps. Additionally, much arable land is located far from existing electrical distribution infrastructure, making the cost of interconnection significant. Since agricultural activities such as irrigation, refrigeration and processing are often energy-intensive, the lack of affordable access to electricity is a challenge.
As a step toward addressing these issues, we conducted an initial hydropower site assessment along the Lower Hamakua Ditch on the Big Island. This irrigation canal was constructed for a sugarcane plantation that has long since left the island. The State now maintains the canal in service of several farms and homes, but the revenues it generates are insufficient to offset the operation and maintenance costs. A small, autonomous hydropower plant could bring the ditch into the black while adding significant value to the community by powering a proposed agricultural processing facility. Access to affordable processing facilities is a major bottleneck for local farmers.
In the initial site assessment, we found two attractive locations: a low-head, high-flow site near the top of the ditch, and a low-flow, high-head site with a large penstock at the lower end. After techno-economic analysis of both sites, we suspect that a Pelton turbine below the terminal reservoir is the most attractive option.
In this work, we will extend the initial site assessment to design a microgrid including a lower reservoir and a pumped-storage system, a PV array, and direct control of flexible electric loads. This design will include control algorithms to optimally balance seasonal variations in water flow rates; seasonal, diurnal and real-time fluctuations in PV output; and the time-varying costs and constraints of a diverse portfolio of flexible loads. Estimation and prediction algorithms will be required in order to anticipate changes in resource availability and in electricity demand. This work will be conducted in collaboration with Hawai'ian partners, who have taken steps toward implementing such a system.
We hope that the development of this microgrid will highlight the potential of small-scale hydropower and pumped storage to 'firm up' the less reliable (but more scalable) renewables.
Megan Kenworthy- University of Idaho
Research- Maximizing the Habitat Restoration Potential of Controlled Releases at Hydropower Dams; Understanding Impacts of Hydrograph Form on Sediment Transport
Regulation of streamflow for hydropower often alters the pre-regulation hydrograph (characterized by the timing, frequency of occurrence, duration, flow magnitudes, and rate of change between discharges). Combined with a reduction in upstream sediment supply, this typically coarsens the sediment on the stream bed downstream of a dam, impacting the availability of habitat for macroinvertibrates, spawning fish and even vegetation. To mitigate for this and other types of habitat impacts, a growing number of restoration projects have included controlled releases designed to mimic key characteristics of the pre-regulation hydrograph. Examples include the use of high flows to aid rebuilding of sand bars and high flows coupled with mitigation gravels to improve spawning habitat for endangered fish. Future projects are likely because the Federal Energy Regulatory Commission (FERC) requires habitat mitigation efforts in the relicensing process for hydropower facilities. However, the outcomes of these costly restoration projects are often hard to predict because little is known about how hydrograph form actually impacts bedload transport in rivers.
This research uses a comprehensive set of flume experiments to gain improved conceptual understanding and quantification of hydrograph influence on bedload transport processes. The work will be limited to armored, gravel bed channels, which is common downstream of dams. (1) How does hydrograph form impact bedload transport rates for a given discharge and the total bedload transported during a high flow event? (2) Does hydrograph form impact what grain sizes are mobile during a high flow event and thus the resulting grain size distribution of the channel bed? (3) Can the impacts of hydrograph form on bedload transport processes be included in and improve upon predictions of bedload transport?
Flume experiments were run at the Center for Ecohydraulics Research (CER) Stream Lab at the University of Idaho in Boise, ID. Experiments included four runs with steady flow and nine hydrograph runs that tested four different hydrograph shapes. The hydrographs were scaled and experiments designed to isolate the impact of more rapid vs gradual changes in discharge. In addition, impacts of repeated hydrographs, sediment supply, and peak flow were explored. Data collection included bedload transport rates and total transport volumes; grain sizes of the bedload material; grain sizes of the bed before, during, and after runs; applied bed shear stress; changes in bed slope and topography; armor persistence; and entrainment/deposition rates for grain size fractions.
The data will be used to examine and quantify the impacts of the various hydrograph shapes on bedload transport processes and the resulting bed conditions and incorporate these effects into a predictive equation for bedload transport. This will provide restoration scientists with a better understanding and new predictive tool to examine the ways in which possible “design floods” will transport sediment and alter instream habitat. This information can be used to reduce costs but also maximize beneficial outcomes of projects that use controlled releases.
Research- Maximizing the Habitat Restoration Potential of Controlled Releases at Hydropower Dams; Understanding Impacts of Hydrograph Form on Sediment Transport
Regulation of streamflow for hydropower often alters the pre-regulation hydrograph (characterized by the timing, frequency of occurrence, duration, flow magnitudes, and rate of change between discharges). Combined with a reduction in upstream sediment supply, this typically coarsens the sediment on the stream bed downstream of a dam, impacting the availability of habitat for macroinvertibrates, spawning fish and even vegetation. To mitigate for this and other types of habitat impacts, a growing number of restoration projects have included controlled releases designed to mimic key characteristics of the pre-regulation hydrograph. Examples include the use of high flows to aid rebuilding of sand bars and high flows coupled with mitigation gravels to improve spawning habitat for endangered fish. Future projects are likely because the Federal Energy Regulatory Commission (FERC) requires habitat mitigation efforts in the relicensing process for hydropower facilities. However, the outcomes of these costly restoration projects are often hard to predict because little is known about how hydrograph form actually impacts bedload transport in rivers.
This research uses a comprehensive set of flume experiments to gain improved conceptual understanding and quantification of hydrograph influence on bedload transport processes. The work will be limited to armored, gravel bed channels, which is common downstream of dams. (1) How does hydrograph form impact bedload transport rates for a given discharge and the total bedload transported during a high flow event? (2) Does hydrograph form impact what grain sizes are mobile during a high flow event and thus the resulting grain size distribution of the channel bed? (3) Can the impacts of hydrograph form on bedload transport processes be included in and improve upon predictions of bedload transport?
Flume experiments were run at the Center for Ecohydraulics Research (CER) Stream Lab at the University of Idaho in Boise, ID. Experiments included four runs with steady flow and nine hydrograph runs that tested four different hydrograph shapes. The hydrographs were scaled and experiments designed to isolate the impact of more rapid vs gradual changes in discharge. In addition, impacts of repeated hydrographs, sediment supply, and peak flow were explored. Data collection included bedload transport rates and total transport volumes; grain sizes of the bedload material; grain sizes of the bed before, during, and after runs; applied bed shear stress; changes in bed slope and topography; armor persistence; and entrainment/deposition rates for grain size fractions.
The data will be used to examine and quantify the impacts of the various hydrograph shapes on bedload transport processes and the resulting bed conditions and incorporate these effects into a predictive equation for bedload transport. This will provide restoration scientists with a better understanding and new predictive tool to examine the ways in which possible “design floods” will transport sediment and alter instream habitat. This information can be used to reduce costs but also maximize beneficial outcomes of projects that use controlled releases.
Kang Li - University of North Carolina at Charlotte
Research- Investigation of the performance and cost benefits of using a magnetically geared generator (MGG) for a hydropower generation.
Magnetic Gearbox - Unlike a mechanical gearbox a magnetic gearbox (MG) creates speed change without any physical contact, eliminates gear lubrication and can operate at high efficiency. A MG also has inherent overload protection as excessive torque will simply cause the poles to slip, whereas a mechanical gearbox would fail. As the MG operates using magnetic fields the MG rotors can be coupled with a generator stator winding thereby enabling a very compact integrated hydropower generator to be realized. Recently Kang Li, and his co-Advisor, Dr. Bird, have demonstrated that a coaxial MG can operate with a torque density far in excess of prior-art designs. This level of torque density is comparable to state-of-the-art Nabtesco and Sumitomo cycloidal mechanical gearbox. The coaxial MGs studied by the Kang Li for his Ph.D. are able to achieve much higher torque densities than prior-art designs because they utilize a flux focusing typology and make insightful use of the pole combinations and outer radii.
The single-stage coaxial MG designs have so far only been able to achieve a high torque density at a low gear ratio [5, 6], typically less than 10:1, such a low gear ratio is not suitable for most hydropower applications because the input speed for the generator will be very low. One approach to increasing the gear ratio is to utilize
multiple MGs in series thereby forming a multistage MG.
By using a hydropower MGG the generator could have high reliability, compact size and high efficiency. The MGG offers the advantages that come with using a mechanical gearbox while also having the operational reliability that comes with using a direct-drive generator.
Current State of the Art - The majority of hydro power generators utilize either a direct-drive synchronous generator or a mechanical gearbox coupled to a generator. If a mechanical gearbox is used, then the generator is typically either a PM generator or an induction generator. The use a gearboxes introduces significant reliability
issues and therefore this can severely affect the levelized cost of energy (LCOE). Direct-drive synchronous generators are often utilized however as the torque
density of a direct-drive synchronous generator is thermally limited (by current) they do not normally achieve torque densities greater than 50Nm/L.
Therefore, direct-drive generators become very large when scaled-up in size and require a large amounts of material. In order to increase the use of hydro power and pumped storage the generator should be low-cost while at the same time it must be exceptionally reliable, and highly efficient. A MGG could potentially offer the ability to increase efficiency and increase reliability for a hydropower generator.
Research- Investigation of the performance and cost benefits of using a magnetically geared generator (MGG) for a hydropower generation.
Magnetic Gearbox - Unlike a mechanical gearbox a magnetic gearbox (MG) creates speed change without any physical contact, eliminates gear lubrication and can operate at high efficiency. A MG also has inherent overload protection as excessive torque will simply cause the poles to slip, whereas a mechanical gearbox would fail. As the MG operates using magnetic fields the MG rotors can be coupled with a generator stator winding thereby enabling a very compact integrated hydropower generator to be realized. Recently Kang Li, and his co-Advisor, Dr. Bird, have demonstrated that a coaxial MG can operate with a torque density far in excess of prior-art designs. This level of torque density is comparable to state-of-the-art Nabtesco and Sumitomo cycloidal mechanical gearbox. The coaxial MGs studied by the Kang Li for his Ph.D. are able to achieve much higher torque densities than prior-art designs because they utilize a flux focusing typology and make insightful use of the pole combinations and outer radii.
The single-stage coaxial MG designs have so far only been able to achieve a high torque density at a low gear ratio [5, 6], typically less than 10:1, such a low gear ratio is not suitable for most hydropower applications because the input speed for the generator will be very low. One approach to increasing the gear ratio is to utilize
multiple MGs in series thereby forming a multistage MG.
By using a hydropower MGG the generator could have high reliability, compact size and high efficiency. The MGG offers the advantages that come with using a mechanical gearbox while also having the operational reliability that comes with using a direct-drive generator.
Current State of the Art - The majority of hydro power generators utilize either a direct-drive synchronous generator or a mechanical gearbox coupled to a generator. If a mechanical gearbox is used, then the generator is typically either a PM generator or an induction generator. The use a gearboxes introduces significant reliability
issues and therefore this can severely affect the levelized cost of energy (LCOE). Direct-drive synchronous generators are often utilized however as the torque
density of a direct-drive synchronous generator is thermally limited (by current) they do not normally achieve torque densities greater than 50Nm/L.
Therefore, direct-drive generators become very large when scaled-up in size and require a large amounts of material. In order to increase the use of hydro power and pumped storage the generator should be low-cost while at the same time it must be exceptionally reliable, and highly efficient. A MGG could potentially offer the ability to increase efficiency and increase reliability for a hydropower generator.
Robert Queen – Colorado State University
Research - Investigation of flow and sediment transport over RoR dams, and potential effects on upstream and downstream geomorphology, sedimentation, and aquatic habitat.
The construction of new small hydropower projects is likely to increase in the coming years. On August 9, 2013, President Obama signed into law the “Hydropower Regulatory Efficiency Act of 2013,” which promotes small hydroelectric projects by streamlining the permitting process for the construction and operation of small and low-impact projects. Many of these projects likely use so-called low-head “run-of-river” (“ROR”) dams, which span the width of the stream channel and do not inhibit discharge of water over the dam. The hydraulics, sedimentation, and upstream and downstream effects of RoR dams are largely unstudied, and the few studies that have been done have reported considerable variability in sedimentation upstream of RoR dams. The extent to which RoR dams constitute sediment traps, and the consequent environmental effects from stream discontinuities that may emerge, is not fully known. The primary objective of this proposed research is to use numerical morphodynamic modeling to investigate the mechanisms of flow and sediment transport over RoR dams, and their potential effects on upstream and downstream geomorphology, sedimentation, and aquatic habitat.
Morphodynamic models, which simulate river channel evolution by solving for fluid flow and sediment transport rates, and conserving sediment mass, are useful tools in both understanding fluvial processes and making predictions of channel response to potential restoration or management actions. Morphodynamic models are extremely promising tools to better understand the impacts of RoR development because they allow researchers to investigate a very broad range of controlling variables such as flow, sediment supply, grain size, and channel slope. The research will use one- and twodimensional morphodynamic models to investigate the effects of RoR dams on sedimentation and channel morphology. The 1D model has been developed in-house at Colorado State University and will be applied over relatively long spatial and temporal scales. The 2D model (Nays2DH) is a freely-available code distributed with the iRIC (international River Interface Cooperative; www.i-ric.org) interface and it will be used to investigate sediment-water interactions in the vicinity of RoR dams. Together, the model simulations will be used to test the following three hypotheses. The amount of sediment stored above a RoR dam will depend on the upstream sediment supply, dam height, grain size, and discharge. The amount of sediment stored above RoR dams will dynamically shift under unsteady flow conditions, scouring at high flow and filling at low flow. The upstream and downstream effects of RoR dams will depend upon the sediment storage efficacy of the dam. Dams that store more sediment will have a
greater effect upstream and downstream than dams that store less sediment.
To test these hypotheses, I will conduct a series of modeling runs where the discharge, sediment supply, sediment grain size distribution, channel slope, and dam height will be systematically varied and equilibrium channel morphologies will be predicted. This ensemble of model runs will provide a general understanding of the important controls on sediment storage behind RoR dams, and the upstream and downstream effects of RoR dams on channel geomorphology. These simulations will allow me to explore a wide parameter space of potentially important controls, enabling me to develop scaling relations between the passage of sediment in terms of grain size, dam height, and flow parameters, which (along with the models themselves) could be used as management tools. The one-dimensional morphodynamic model, as well as any changes to the two-dimensional morphodynamic model source code, will be incorporated into the iRIC interface and distribution. In this manner, these modeling tools and associated tutorials and user guides will be available to anyone interested in applying them to specific RoR projects or other river management situations.
Nam Song - University of Washington
Research - Forecasting Tools for Run-of-River Hydroelectric Systems
Run-of-river (ROR) hydroelectricity is an increasingly popular form of hydroelectric generation in the state of Washington because it is perceived as being environmentally friendly. By constructing penstocks that divert some water flow from rivers and creeks, electricity can be generated without the need to construct reservoirs. However, since these ROR stations have minimal capacity for energy storage, their power output cannot be controlled as it depends on seasonal river flows and precipitation levels. Therefore, accurate prediction models are needed to integrate them more effectively with the economic operation of the larger power system and avoid causing reliability problems. The results of this research will enhance the utilization of ROR resources, and hence make the power system more sustainable while maintaining its reliability.
The goal of this research project is to develop accurate forecasting tools for run-of-river hydroelectric projects. These tools should provide long term (7-day ahead), medium term (day-ahead), and short term (90-minute ahead) estimates of power output that can be used by dispatchers to refine their position in the bulk and energy imbalance markets, as well as to ensure that operating constraints on the power network are not violated. In addition, these tools should help maximize the energy produced by these plants while maintaining a sufficient water flow through the creeks to preserve the local ecosystem. Providing these forecasts is also a requirement imposed by the Federal Energy Regulatory Commission (FERC).
The forecast models developed by this tool will be tested on 3 small capacity ROR hydroelectric projects under development by the Snohomish County Public Utility District (Snohomish PUD). These plants are located on the Western slope of the Cascades in NW Washington. They consist of a 7.5MW plant along Youngs Creek, which is fed by snowmelt from Bessemer Mountain, a 6MW plant along Hancock Creek sourced from Hancock Lake, and a 6MW plant along Calligan Creek sourced from Lake Calligan. Some factors that would affect the hydrological model forecasts for each of the creeks include precipitation levels in the region, water levels of the lakes, and snowpack and temperature forecasts at Bessemer Mountain as they both influence the snowmelt.
To develop the desired forecast tools, several potential models will be developed using available data from the PUD and the public domain. Firstly, a conceptual hydrological model will be selected and calibrated at the 3 water sources using public domain weather forecasts as well as real-time precipitation data. Using this predictive model, power scheduling models will be developed.
Research - Forecasting Tools for Run-of-River Hydroelectric Systems
Run-of-river (ROR) hydroelectricity is an increasingly popular form of hydroelectric generation in the state of Washington because it is perceived as being environmentally friendly. By constructing penstocks that divert some water flow from rivers and creeks, electricity can be generated without the need to construct reservoirs. However, since these ROR stations have minimal capacity for energy storage, their power output cannot be controlled as it depends on seasonal river flows and precipitation levels. Therefore, accurate prediction models are needed to integrate them more effectively with the economic operation of the larger power system and avoid causing reliability problems. The results of this research will enhance the utilization of ROR resources, and hence make the power system more sustainable while maintaining its reliability.
The goal of this research project is to develop accurate forecasting tools for run-of-river hydroelectric projects. These tools should provide long term (7-day ahead), medium term (day-ahead), and short term (90-minute ahead) estimates of power output that can be used by dispatchers to refine their position in the bulk and energy imbalance markets, as well as to ensure that operating constraints on the power network are not violated. In addition, these tools should help maximize the energy produced by these plants while maintaining a sufficient water flow through the creeks to preserve the local ecosystem. Providing these forecasts is also a requirement imposed by the Federal Energy Regulatory Commission (FERC).
The forecast models developed by this tool will be tested on 3 small capacity ROR hydroelectric projects under development by the Snohomish County Public Utility District (Snohomish PUD). These plants are located on the Western slope of the Cascades in NW Washington. They consist of a 7.5MW plant along Youngs Creek, which is fed by snowmelt from Bessemer Mountain, a 6MW plant along Hancock Creek sourced from Hancock Lake, and a 6MW plant along Calligan Creek sourced from Lake Calligan. Some factors that would affect the hydrological model forecasts for each of the creeks include precipitation levels in the region, water levels of the lakes, and snowpack and temperature forecasts at Bessemer Mountain as they both influence the snowmelt.
To develop the desired forecast tools, several potential models will be developed using available data from the PUD and the public domain. Firstly, a conceptual hydrological model will be selected and calibrated at the 3 water sources using public domain weather forecasts as well as real-time precipitation data. Using this predictive model, power scheduling models will be developed.
Laurel Stratton- Oregon State University
Research- Controls on the Burial Efficiency of Reservoir Sediments: Toward Understanding the Net Carbon Balance of Hydroelectric Reservoirs
As the effects of climate change become a reality across the globe and societal pressures build toward ‘green’ energy, the exigency of understanding the carbon footprint of dams and reservoirs is undeniable. Reservoirs both emit and sequester huge volumes of carbon, but the in-reservoir processes and controlling factors favoring emission vs. sequestration are poorly understood. Systematic uncertainties in quantification leads to the question: what impact does a large water reservoir have on the carbon budget of a temperate ecosystem? What is the carbon footprint of water reservoirs? By understanding the fundamental processes controlling carbon processing, emission, and sequestration in reservoirs, this research will help the hydropower industry prepare itself to meet the societal pressure for ‘green’ energy head-on and be able to proactively address and manage greenhouse gas emissions.
To begin addressing the storage component of carbon processing in reservoir sediments, I performed a study of the stratigraphy of carbon sediments in former Lakes Aldwell and Mills, Olympic, WA (Stratton, OSU Hydrophiles Water Research Symposium, April 26-28 2015). This work, the first detailed examination of reservoir sediment in situ worldwide, shows that extensive beds of organic carbon (primarily composed of forest detritus) are broadly deposited in prograding delta sediments and topset beds, but generally do not extend into the lacustrine reservoir pool. This work suggests that the Elwha River reservoirs are endmembers for carbon storage, representing a high-sediment, low-nutrient load system, and that the patterns of carbon burial vs. mineralization will be significantly different in a less pristine watershed.
To test this hypothesis, the research is investigating the burial efficiency of the three major arms in Lake Billy Chinook, central Oregon. Burial efficiency is the mass balance between the total input of organic carbon to the sediment/water interface, losses of organic carbon due to respiration in the biologically active surface sediments, and sequestration of organic carbon by burial. Many processes control burial efficiency in reservoirs, including oxygen exposure time, carbon source and supply, depth, distance from the dam, seasonal variation, nutrient input, depositional patterns, and reservoir operations, among others. Because Lake Billy Chinook inundates the confluence of three distinct river systems sourced in different climates and terrains, it provides a natural laboratory to investigate differences in sediment input, nutrient loading, and water temperature.
Research- Controls on the Burial Efficiency of Reservoir Sediments: Toward Understanding the Net Carbon Balance of Hydroelectric Reservoirs
As the effects of climate change become a reality across the globe and societal pressures build toward ‘green’ energy, the exigency of understanding the carbon footprint of dams and reservoirs is undeniable. Reservoirs both emit and sequester huge volumes of carbon, but the in-reservoir processes and controlling factors favoring emission vs. sequestration are poorly understood. Systematic uncertainties in quantification leads to the question: what impact does a large water reservoir have on the carbon budget of a temperate ecosystem? What is the carbon footprint of water reservoirs? By understanding the fundamental processes controlling carbon processing, emission, and sequestration in reservoirs, this research will help the hydropower industry prepare itself to meet the societal pressure for ‘green’ energy head-on and be able to proactively address and manage greenhouse gas emissions.
To begin addressing the storage component of carbon processing in reservoir sediments, I performed a study of the stratigraphy of carbon sediments in former Lakes Aldwell and Mills, Olympic, WA (Stratton, OSU Hydrophiles Water Research Symposium, April 26-28 2015). This work, the first detailed examination of reservoir sediment in situ worldwide, shows that extensive beds of organic carbon (primarily composed of forest detritus) are broadly deposited in prograding delta sediments and topset beds, but generally do not extend into the lacustrine reservoir pool. This work suggests that the Elwha River reservoirs are endmembers for carbon storage, representing a high-sediment, low-nutrient load system, and that the patterns of carbon burial vs. mineralization will be significantly different in a less pristine watershed.
To test this hypothesis, the research is investigating the burial efficiency of the three major arms in Lake Billy Chinook, central Oregon. Burial efficiency is the mass balance between the total input of organic carbon to the sediment/water interface, losses of organic carbon due to respiration in the biologically active surface sediments, and sequestration of organic carbon by burial. Many processes control burial efficiency in reservoirs, including oxygen exposure time, carbon source and supply, depth, distance from the dam, seasonal variation, nutrient input, depositional patterns, and reservoir operations, among others. Because Lake Billy Chinook inundates the confluence of three distinct river systems sourced in different climates and terrains, it provides a natural laboratory to investigate differences in sediment input, nutrient loading, and water temperature.
Jesse Thornburg - Carnegie Mellon University
Research - Energy Storage & International Development
I’m developing a model of conventional hydropower with a reservoir for storage, programming in MATLAB and Simulink to build multi-tiered control diagrams for different subsystems. The hydro components fit into larger work with my Kigali-based advisor, Professor Bruce Krogh, on a microgrid simulator that uses probability distributions to characterize each generation technology and customer load. This approach accounts for variability in both supply and demand. With this method we are characterizing the system's load curve and reducing the likelihood of power outages through sizing and dynamically controlling the supplies. The tool covers several of HRF's focus areas in providing a market tool and reducing environmental impact through intelligent control. We are specifically testing how prioritizing hydro and effectively managing it can reduce the need to burn fuel at peak demand hours. Hydro also stabilizes microgrid systems with a reliable baseload and some dispatchable bandwidth to account for the sporadic nature of other renewables like wind and PV. My hydro research thus far has been focused on conventional hydropower, with potential energy storage in reservoirs I’ve modeled with Simulink. This reservoir model has been tested and is now running, while I plan to expand and improve the hydro model with detailed simulations of the electromechanical components. I recently modeled and performed initial testing on these subsystems, again using Simulink.
In addition to energy storage in a hydro reservoir (as potential energy to be converted into hydropower), my microgrid model includes a battery storage system. Professor Krogh and I plan to add pumped storage into the grid model, which would incorporate a similar reservoir model as conventional hydro but with the new option to pump water into the reservoir on command. With this addition we plan to do side-by-side comparisons between battery systems and pumped storage hydro to contrast their economic returns, responsiveness, and sensitivity to outside factors.
While the majority of my work focuses on simulations and computational models, we are also taking steps to test these models and operating strategies with physical hydro plants. Since managing a 440kW run-of-river plant in Rwanda from 2013-2014, I have stayed connected with several running hydro plants and their project developers in Rwanda. I've collected their data (working with Rwandan MS students) to use actual hydro plant outputs for my modeling work. I am also working with Stanford professor Tim Chou and Professor Krogh on developing a smart-meter-enabled microgrid in northwest Rwanda. This project is called the Renewable Energy System Innovation Center (RESIC) and will operate in conjunction with CMU’s Rwanda campus. I’ve been talking with Canyon Hydro about installing a 50 kW Pelton or reverse pump system similar to one I previously worked on in Tanzania. Together with wind and PV, the RESIC hydro installation will power a hospital and small village which agree to provide demand data for research purposes. The hydro plant will serve as a training platform for local technicians, and RESIC as a whole will provide training and research opportunities for engineers from Rwanda, East Africa, and the wider world. My proposed research combines the theoretical with the practical to build hydro capacity and engineering expertise in Sub-Saharan Africa.
Research - Energy Storage & International Development
I’m developing a model of conventional hydropower with a reservoir for storage, programming in MATLAB and Simulink to build multi-tiered control diagrams for different subsystems. The hydro components fit into larger work with my Kigali-based advisor, Professor Bruce Krogh, on a microgrid simulator that uses probability distributions to characterize each generation technology and customer load. This approach accounts for variability in both supply and demand. With this method we are characterizing the system's load curve and reducing the likelihood of power outages through sizing and dynamically controlling the supplies. The tool covers several of HRF's focus areas in providing a market tool and reducing environmental impact through intelligent control. We are specifically testing how prioritizing hydro and effectively managing it can reduce the need to burn fuel at peak demand hours. Hydro also stabilizes microgrid systems with a reliable baseload and some dispatchable bandwidth to account for the sporadic nature of other renewables like wind and PV. My hydro research thus far has been focused on conventional hydropower, with potential energy storage in reservoirs I’ve modeled with Simulink. This reservoir model has been tested and is now running, while I plan to expand and improve the hydro model with detailed simulations of the electromechanical components. I recently modeled and performed initial testing on these subsystems, again using Simulink.
In addition to energy storage in a hydro reservoir (as potential energy to be converted into hydropower), my microgrid model includes a battery storage system. Professor Krogh and I plan to add pumped storage into the grid model, which would incorporate a similar reservoir model as conventional hydro but with the new option to pump water into the reservoir on command. With this addition we plan to do side-by-side comparisons between battery systems and pumped storage hydro to contrast their economic returns, responsiveness, and sensitivity to outside factors.
While the majority of my work focuses on simulations and computational models, we are also taking steps to test these models and operating strategies with physical hydro plants. Since managing a 440kW run-of-river plant in Rwanda from 2013-2014, I have stayed connected with several running hydro plants and their project developers in Rwanda. I've collected their data (working with Rwandan MS students) to use actual hydro plant outputs for my modeling work. I am also working with Stanford professor Tim Chou and Professor Krogh on developing a smart-meter-enabled microgrid in northwest Rwanda. This project is called the Renewable Energy System Innovation Center (RESIC) and will operate in conjunction with CMU’s Rwanda campus. I’ve been talking with Canyon Hydro about installing a 50 kW Pelton or reverse pump system similar to one I previously worked on in Tanzania. Together with wind and PV, the RESIC hydro installation will power a hospital and small village which agree to provide demand data for research purposes. The hydro plant will serve as a training platform for local technicians, and RESIC as a whole will provide training and research opportunities for engineers from Rwanda, East Africa, and the wider world. My proposed research combines the theoretical with the practical to build hydro capacity and engineering expertise in Sub-Saharan Africa.