Faculty of Engineering and Information Technology
http://hdl.handle.net/11462/1
2023-12-28T10:07:29ZSustainable Groundwater Management Through Modelling Of Selected Hydrological Parameters: A Case Study Of The Modder River Catchment Of South Africa
http://hdl.handle.net/11462/2514
Sustainable Groundwater Management Through Modelling Of Selected Hydrological Parameters: A Case Study Of The Modder River Catchment Of South Africa
Alowo, Rebecca
Farmers and water users in arid and semi-arid zones of South Africa have a limited number of water resources. With 98% of South African surface water already allocated and committed, there is an increasing demand for use of groundwater through the use of wells and boreholes.
This study was carried out in the Modder River catchment, which is a sub-catchment under the larger Orange River system. increased population growth and increasing urbanisation have negatively affected water resources in the Modder River catchment since 2000. The population of the Modder River catchment increased from 618 566 in 2001 to an estimated 1 083 886 in 2016. This resulted in increasing water demand and excessive pumping of groundwater, particularly for agriculture and drinking water. At 351 million litres, the Bloemfontein area and Modder River catchment has the highest demand for water in the Upper Orange River Catchment.
Other negative effects are increased degradation of the environment, high levels of human-induced climate changes, and high variability of the natural climate. The net effects have been depletion of aquifers and prolonged periods of drought, which points to a lack of effective groundwater resources management tools and laxity in monitoring groundwater as a resource. Globally, the general trend is to adapt models and artificial intelligence to manage water resources sustainably. Groundwater sustainability models that incorporate artificial intelligence, may be used to enhance the governance of aquifers. Lack of such models is what motivated this research. This was achieved through the development of groundwater sustainability models/indices and information communications technology-based framework for groundwater monitoring that makes use of artificial intelligence to reach informed decisions for improved groundwater management.
The research was guided by a theoretical framework design, which included sustainability modelling of the interactions of the elements and physical processes linking the hydrological cycle with human instigated factors in connection with groundwater. These interactions were climate and utilisation of groundwater, surface water-groundwater water linkages and climate–groundwater interaction connections in hydrological parameters. The parameters were weighted and scored and included precipitation, slope, vegetation cover, evapotranspiration, climatic zones, water quality, aquifer yields, soil composition, aquifer sustainability, population, water use categories, the permit framework and groundwater rights. These parameters were grouped into the following; Factor A: climate, Factor B: rights and equity versus resources, Factor C: socio-economic matters and Factor D: aquifer sustainability. Delphi techniques of scoring and rating were applied. Scores were assigned with the lowest scores of one relating to poor practices and the highest score of five relating to adequate practices. Climate (Factor A) and aquifer sustainability (Factor D) were summed up to an aggregate score of 30 each, while rights (Factor B) and socio-economic aspects (Factor C) had separate scores of 20 each. The groundwater sustainability range had 19 (slightly reasonable) and 100 (profoundly feasible) scores.
The developed sustainability indices were applied to 51 boreholes in the Modder River catchment of South Africa, for which consent was obtained for the study. These boreholes were both government and private boreholes. ArcView GIS was used to demonstrate the results for each factor for the 51 boreholes towards the groundwater sustainability concept. The sustainability class in the Modder River Basin ranged from low to moderate sustainability. The results showed that only 11 boreholes were sustainable, implying that they were reasonably maintainable. The other 40 boreholes had low sustainability scores, which indicated that they were not sustainable. The moderate sustainability class was typical of areas with extensive dolerite, low slopes, and low recharge.
In order to validate the results, plots of over 70 years of rainfall data and evapotranspiration data of the Modder River Basin catchment were generated. This was to indirectly account for a mini-water balance and how much was available for groundwater recharge. It was found that rainfall was significantly lower than evapotranspiration, which meant there was more evapotranspiration than recharge for long periods of time. The Modder River catchment received its highest bout of rain in 1971/1972. The sustainability maps showed areas with low sustainability scores correlating with areas with a low aquifer sustainability factor since low recharge related to the low sustainability scores.
A further validation analysis was carried out on the final sustainability map, the river network map and recharge map of the study area, as extracted from the countrywide recharge map of South Africa. High groundwater recharge areas (over 34 mm) were found in the eastern and south-eastern parts of the study area. These areas corresponded with high river network presence. The central and western parts of the study area were marked by lower groundwater recharge values (below 12 mm). This was consistent with the derived sustainability final index maps. There was a high correlation between the sustainability index map, the drainage basin map and the recharge map. The final sustainability map value was plotted against measurable parameters (pumping rate, aquifer yield). In this thesis, a calculation of the correlation coefficient values was low at 1.04%, indicating no positive correlation or reinforcement. In addition, the plots of the comparison of the calculated indices with the measured physical values on one hand, and an analysis of the actual end users’ validation done to assess the ability of the sustainability index on the other hand, were presented as the final sustainability map value against a measurable parameter storativity (aquifer yield). For this, significant similarity between the storativity plot and the sustainability index map was observed.
Furthermore, the results showed that the ideal pumping rate should be 138 080 cubic metres per year per borehole. This means that only this amount may be pumped, without disturbing the dynamic equilibrium of the aquifer. This pumping rate would support the capture principle approach for the C52 catchment. Currently, the average pumping rates is higher than the aquifer yield. The majority of the points on the major plotted graphs were below the best fit line. This meant that the boreholes were abstracted at a rate that was higher than the aquifer yield. With this finding, the researcher calculated and recommended pumping rates/volume of water for each borehole. This meant that only that amount may be pumped without disturbing the dynamic equilibrium of the aquifer. This pumping rate would support the capture principle approach for the C52 catchment. The recommended pumping rates were based on applying the corrected formula: Pumping rate = 157 397x + 138 080, which was derived after the pumping rate analysis. The recommended pumping rates were proposed because the moderate to low classes represented areas that were the opposite of a favourable scenario: overly high abstraction activity, unfavourable climatic conditions, and slow to little groundwater processes and interactions (high or steep slopes and low rainfall).
From the graph of storativity versus licensed volume that was plotted, this research revealed that higher volumes were licensed, compared to the storage volume of the aquifer. The findings on the final sustainability map provided in this thesis further correlated to this. There was no correlation between the pumping rate/licensed volume and the storativity of C52.
The findings above confirmed the widely documented challenges of groundwater monitoring in the southern African region that most of the water administrative authorities continue to face. One of the reasons is that many water clients do not submit the required compulsory critical data for effective monitoring of water use. This, combined with the absence of limits by water authorities regarding boreholes dug by permit holders, has led to difficulties in decision-making and groundwater conservation. To address this, the potential of artificial intelligence in groundwater management is explored in this research. A framework that makes use of artificial intelligence in groundwater management is proposed. The aim of this framework is to increase the success chances of groundwater management, conservation, and sustainability by means of a monitoring system through real-time updates and early warning signals to water management, and to, thus, reinforce positive behavioural practices on water abstraction and effective water usage. With this, groundwater consumers will have the means to gauge when they are using too much water and municipalities will have the means to gauge if a catchment is over-abstracted. The net gain or impact will be water savings, cost savings and preferable performance of provinces, municipalities and departments that manage water. Besides, once implemented, this framework would ensure automation of the sustainability modelling of groundwater system discussed in the thesis. This would result in an efficient catchment monitoring, allocation, licensing, and characterisation for an effective groundwater management system.
Thesis
2020-11-01T00:00:00ZShort-Term Load Demand Forecasting For Transnet Port Terminal (Tpt) In East London Using Artificial Neural Network
http://hdl.handle.net/11462/2513
Short-Term Load Demand Forecasting For Transnet Port Terminal (Tpt) In East London Using Artificial Neural Network
Figlan, Mncedisi Sityebi
The daily and weekly energy consumption patterns at the Transnet Port Terminal (TPT) in East London varies stochastically. This is as a result of the transient weather patterns that exist at the harbor. It has therefore become imperative to wisely manage this load in order to save electricity costs and for future infrastructure development. Hence the ongoing supply of electricity to port consumers requires an accurate and adequate short-term load forecast (STLF) for quality, quantity, and efficient management.
Many researchers have recently proposed Artificial Neural Networks for short-term load prediction. However, most of the studies have not considered the quickly changing weather patterns that exist at the port. Therefore, the objective of this study is to establish a supervised short-term load prediction using ANN models, and to verify the effectiveness of such predictions by using the real load data from the TPT. The suggested system architecture uses open- loop training with real load and weather information, and then a closed-loop network is used to produce a prediction with the predicted load as its feedback data.
Data collection points were set up in the ring network of the port by installing new power measuring meters, and weather data obtained from local meteorology offices in order to build a suitable alternative of localised data management (data base) for saving all data gathered. Hence, profiling of the load in the TPT was done and load forecasting was carried out, leading to improved load management strategies for the harbor terminal. ANN short-term load prediction (STLP) models were developed utilising its own performance to improve precision by essentially implementing a load feedback loop that is less reliant on external data. To ensure that the timeseries data recorded at the port were well modeled, the Nonlinear autoregressive exogenous model (NARX) for load prediction were developed using mean squared error (MSE) as a performance metric.
Furthermore, to show the efficacy of the proposed model for STLP, the adaptive neuro-fuzzy inference system (ANFIS) was used with the same data for short-term predictions. The minimum mean squared errors obtained for both NARX and ANFIS models were 0.0010939 and 0.0032 respectively, indicating that the NARX model is more accurate during the forecast of departmental loads. The results of the predictions using the hourly timeseries indicated a close match between the forecasted and actual load demand at the port terminal. The effects of the load forecast could be used as a guide for implementing management plans for internal load, such as the generation of urgent electricity and the programme of implementation for demand-side management policies.
Dissertation
2021-01-01T00:00:00ZEfficient Wireless Power Transfer For Low Power Wide Area Networks
http://hdl.handle.net/11462/2511
Efficient Wireless Power Transfer For Low Power Wide Area Networks
Makhetha, Molefi, Johannes
Wireless power transfer (WPT) technologies for small devices and low power sensors have
drawn substantial research attention in recent years. Traditional near and far- eld WPT
systems cannot provide e cient-high power transfer while at the same time maintaining
long range power transfer. A possible candidate to overcome these challenges is the
strongly coupled magnetic resonance (SCMR) WPT technique which can transfer power
at higher transmission e ciency in the medium range. Heretofore, the focus has been to
improve the e ciency and range of the SCMR system. On the other hand, the study to
develop optimal coils or loops of the WPT system utilising less computational resources
as well as using co-simulations between less and high intense software has been limited.
More so, the existing WPT systems are complex and bulky in size making it a challenge
to use these technologies for small footprint applications. Therefore, innovative SCMR
systems that are designed to be easy to fabricate and with low losses and of small footprint
will notably improve various technologies in a variety of applications.
The optimal and small footprint SCMR WPT systems are studied in this work. The
analytical models of the Conformal-SCMR (CSCMR) system are presented rst through
design methodology and analysis. The designed CSCMR systems' performance is envisaged
from the identi ed optimal design parameters through this analysis. Furthermore,
the derived optimal parameters are fabricated, analysed and compared in a 3D simulator,
a conventional CSCMR model and a 2-layer self-resonant resonator model. It was
noted that the 2-layer self-resonant model performed better than the conventional model
and this was veri ed by mathematical formulae and equivalent circuit models. The two
models were then optimised using their derived physical parameters. This was done
through a co-simulation. The results showed that the co-simulation increased the simulation
speeds, therefore saving computational resources. In conclusion, the two optimised
model's transmission e ciency was improved by 30% and 4% for the conventional derived
and the 2-layer self-resonant CSCMR-WPT systems. This was achieved while the
footprint of these systems was reduced.
Dissertation
2021-01-01T00:00:00ZPractical Implementation Of Hybrid Energy Systems For Small Loads In Rural South Africa
http://hdl.handle.net/11462/2509
Practical Implementation Of Hybrid Energy Systems For Small Loads In Rural South Africa
Meje, Kelebogile, Confidence
Hybrid renewable energy systems (HRESs), are alternative off-grid methods of generating power to remote rural areas, where power lines are not economically viable. Most of the research studies on renewable hybrid systems or microgrids (MGs) in South Africa, focus mainly on the optimal sizing and optimal control of different systems, by making use of renewable energy simulation softwares, however, there is a lack of research carried out on the implementation of these hybrid systems in real time.
The aim is to develop a real time control method for an isolated hybrid system submitted to a variable load, as well as resources. The first step towards achieving this aim, was to critically review available published research works, to describe recent developments in improving the optimum operating concept of microgrid controllers for stand-alone or grid-connected systems. Secondly, to investigate any real-time implementation established by either hierarchical or distributed control. Then to, analyze their reliability and functionality in practical set up of the controller, in managing power in the system to the variable load.
The study provided a brief overview of microgrid prototype systems, microgrid controls, operating modes and multi-DER microgrid types built into a hybrid system, which introduces a number of strategies or techniques for managing remote rural application prototypes in an isolated or grid-connected system. However, hierarchical control was found to be more appropriate for large microgrids with multiple types of distributed energy resources (DERs), compared to distributed control, particularly when combined with energy storage systems (ESSs), in isolated mode.
The rising of hybrid system controllers in real-time renewable energy for the optimum energy management system (EMS), required the design of a real-time controller to operate the entire system in real time. Increasing popularity of renewable energy (RE) has a control strategy that determined the overall efficiency of the hybrid system (HS), although the energy management system of these systems is particularly complex to be managed.
The study's main contribution is to investigate the feasible controller and, later, to present an advanced control strategy for managing and controlling the flow of hybrid renewable energy with a diesel generator (DG) and battery (BT) as a backup in a rural application of SA. EMS would be implemented, using a fuzzy logic controller (FLC) in MATLAB / SIMULINK. This study analysed input and output variables for the design of a controller, with a set of rules and a three-dimension (3D) surface. Simulation results of related studies with different objectives were analysed, with the aim of sussing out an appropriate controller for the current study.
Arduino Mega was used for coding and uploaded to the implementation of practical implementation of the study. The system operated successfully by supplying the load. This study finally answered the question of the feasibility of the controller in real-time applications.
Dissertation
2021-05-01T00:00:00Z