Multipurpose reservoir operation management using evolutionary optimisation under uncertainty in water demand and supply. (2021)
Type of ContentTheses / Dissertations
Thesis DisciplineCivil Engineering
Degree NameDoctor of Philosophy
PublisherUniversity of Canterbury
Global freshwater resources are under more pressure due to high demand by industrial, recreational, municipal, and agricultural sectors. Additionally, higher standards of living, growing population, and climate variability have caused water shortages and are increasing conflict among water users.
Multi-purpose reservoirs play a crucial role in fulfilling water demands and minimising the risk of water shortages. However, multiple water users with different objectives under a variety of constraints result in water allocation challenges. Optimal water allocation from existing reservoirs has thus become a critical requirement for sustainable water resource management. Most reservoirs operate within an environment in which water demands and supplies have high levels of uncertainty. Therefore, it is crucial to recognise and analyse the impact of uncertainties on reservoir operations and the process of optimising water allocation. Although there are many optimisation techniques, genetic algorithm optimisation, using a population-based algorithm, offers a well-established and effective method for solving multi-objective problems.
This research aims to assess the water supply reliability of multi-purpose reservoirs over various operational time frames using genetic algorithm optimisation under uncertainty in climate change, land use and water demands. To achieve this aim, a thorough literature review of optimisation models, uncertainty, climate and land use change, and the application of the SWAT (Soil & Water Assessment Tool) model was conducted. A modelling framework that couples the SWAT model and the @RISK genetic algorithms optimisation tool was developed and applied to the case study of the Nuicoc watershed reservoir system in the north of Vietnam. The SWAT model, a well-known catchment model that is built to quantify and predict the effects of land management on water resources under varying climate, land use and management conditions over time, was used to simulate reservoir inflow uncertainty. Water allocation optimisation was then carried out using a probabilistic optimisation approach and genetic algorithm for various scenarios of changes in land use, climate, and water demands. The specific objectives of the case study were to (i) assess the impact of climate and land use change on water and sediment inflow into the reservoir using the SWAT model, (ii) use the probabilistic optimisation approach to compute the range of reliability, resilience and vulnerability of the reservoir under land use and climate change scenarios and (iii) suggest water allocation policies and best management practices to improve the performance of the reservoir to sustainably supply water.
Climate data and water demands were initially kept the same as historical data to consider the impact of land-use changes on the reservoir reliability. The modelling results indicated that an expansion of the urban areas by 10% and conversion of 5% of forest to agricultural areas yielded the highest water releases for downstream demands of all simulated scenarios, with 5 Mcm/year greater water releases than the baseline, thereby not considering sedimentation. However, when sedimentation was considered, it resulted in the greatest decrease in water releases, with 6.25 Mcm/year less than the baseline. Additionally, it was found that the spatial distribution of land-use significantly affected sediment inflows into the reservoir, highlighting the importance of targeted sediment management.
Furthermore, the results showed that land use and climate change combined impacted streamflow and sediment yield from the watershed, thereby negatively affecting the reliability of the reservoir. A 10% increase in urban areas and conversion of 5% of the forest to agricultural areas under the Global Climate Model (GCM) GFDL-CM3 and Representative Concentration Pathway (RCP) 8.5 produced the highest average water inflows, at 24.72 m3/s. This scenario, however, generated the lowest reliability and resilience of all scenarios (10% and 30% lower than the baseline, respectively), and highest vulnerability (4 lower than the baseline) due to a 114.8 million cubic metres increase in reservoir sedimentation, which may also cause greater downstream flood risk in the wet season. Modifying the water allocation policies and application of targeted best management practice (BMP) increased reservoir performance. The results indicated that implementing BMP’s helped to mitigate soil erosion and increased the reliability and resilience by up to 2.7% and 9.5%, respectively, compared with the cases without BMP’s. This highlights the importance of BMP’s for improving reservoir performance, as sedimentation has a major long-term influence on reservoir water supply. The proposed framework was demonstrated to be useful for decision-makers in assessing the impact of water allocation policy, BMP’s, land-use and climate change on the reservoir operation. The results obtained from the case study are valuable to decision makers for management of water demands, land use and sedimentation under a broad range of uncertainties. Furthermore, simulation of scenarios can help the government to formulate clearer policies to adapt and mitigate the impact of land use and climate change on the reliability, resilience, and vulnerability of the reservoir water supply.
RightsAll Right Reserved
Showing items related by title, author, creator and subject.
Nguyen A; Pahlow M; Cochrane, Thomas (2019)Available fresh water resources are becoming scarce due to higher demands by industrial, recreational, municipal, and agricultural sectors. Reservoirs are one of the most efficient types of structures used to manage water ...
Starkey, Stephen Robert (University of Canterbury. Management, Marketing and Entrepreneurship, 2014)This thesis first develops a range of wholesale water market design options, based on an optimisation approach to market-clearing, as in electricity markets, focusing on the extent to which uncertainty is accounted for in ...
Demand Response in the Residential Sector: A Critical Feature of Sustainable Electricity Supply in New Zealand Gyamfi, S.; Krumdieck, S.; Brackney, L. (University of Canterbury. Mechanical Engineering., 2008)The world summit on sustainable development in Johannesburg concluded that changing unsustainable patterns of energy use is a key area for global action to ensure survival of our planet. Demand response in the residential ...