Show simple item record

dc.contributor.advisorBezuidenhout, C.C.
dc.contributor.advisorBezuidenhout, J.J.
dc.contributor.authorO’Reilly, Guzéne
dc.date.accessioned2020-07-17T10:42:26Z
dc.date.available2020-07-17T10:42:26Z
dc.date.issued2020
dc.identifier.urihttps://orcid.org/0000-0002-7576-1723
dc.identifier.urihttp://hdl.handle.net/10394/35180
dc.descriptionPhD (Environmental Sciences), North-West University, Potchefstroom Campusen_US
dc.description.abstractSource water is becoming a scarce resource in South Africa, especially with the extreme periods of drought that the country has faced the last few years. Continuous pollution of rivers and dams is drastically deteriorating the water quality. This puts more pressure on water purification plants to ensure the water is adequately treated and safe drinking water is produced. However, some water purification plants may not have the infrastructure or financial resources to produce drinking water of acceptable quality. Therefore, there is a need for inexpensive and rapid solutions to produce safe drinking water. In this study, three decision-making tools (Hazard Analysis Critical Control Point (HACCP) concept, Artificial Neural Networks, Evolutionary Algorithms and isolation of VBNC bacteria) were evaluated for the improved monitoring of treatment processes. The HACCP concept was evaluated at three water purification plants (Plant A, Plant B and Plant C). This study demonstrated that monitoring certain parameters after each step in the treatment process was useful to identify process failures. Results for Plant B indicated that due to a failure in the filtration process, unacceptable turbidity levels were present in the drinking water. The second decision-making tool investigated in this study was the application of ANNs and evolutionary algorithms (EAs) at Plant A, Plant B and Plant C. Results from this study indicated that the combination of ANNs and EAs resulted in the accurate prediction of electrical conductivity (EC) in the drinking water of Plant A and Plant C. Additionally, this study indicated the importance of consistent monitoring. A prediction model for Plant B could not be generated due to a lack of historical data. The application of EAs resulted in the formation of accurate predictive rule-sets for Plant A and Plant C. The last decision-making tool was the recovery of Escherichia coli (E. coli) in drinking water. Results of this study indicated that viable-but-non-culturable E. coli was recovered from drinking water. Two of the isolates were identified as E. coli O177 and O157. This is concerning as O177 and O157 are Shiga toxin-producing strains of E. coli and could pose a serious health risk. In addition, this study indicated the development of a new resuscitation method adapted from a current method. Based on all the results obtained, decision-making tools for the improved monitoring of water treatment processes was demonstrated.en_US
dc.language.isoenen_US
dc.publisherNorth-West University (South Africa)en_US
dc.subjectHazard Analysis Critical Control Pointen_US
dc.subjectArtificial neural networksen_US
dc.subjectEvolutionary algorithmsen_US
dc.subjectViable-but-nonculturable Escherichia colien_US
dc.titleDecision-making tools for establishment of improved monitoring of water purification processesen_US
dc.typeThesisen_US
dc.description.thesistypeDoctoralen_US
dc.contributor.researchID12540110 - Bezuidenhout, Cornelius Carlos (Supervisor)
dc.contributor.researchID10926542 - Bezuidenhout, Johannes Jacobus (Supervisor)


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record