Analys av dricksvattenrening med metoderna Mikrobiologisk riskanalys (MRA) och God desinfeksjonspraksis (GDP)

Drinking water is produced from raw water and is either from groundwater or surface water. This thesis aims to find out if the cleaning process of raw water is sufficiently effective. This is important because consumers are otherwise at risk of waterborne infection caused by pathogens. There are three groups of pathogens; bacteria, virus and parasite. These have different characteristics which mean that they require different water treatment to be separated. In addition to normal operation, a number of scenarios were examined. This is to investigate how water treatment would do if they became a reality. The thesis has examined Borg´s waterworks operated by Norrköping Vatten AB. It was defined to cover the distance from water source to the consumer. In the work, the model Quantitative Microbial Risk Assessment (QMRA) was used to perform risk analysis by simulating the normal operation and different scenarios of the water purification process. Thus, knowledge can be obtained about the effectiveness of separation by bacteria, viruses and parasites. However, the QMRA-model is considered to contain some flaws and for that reason the Norwegian model called Good Disinfection Practice (GDP) was also used. GDP is a theoretical model which is based on formulas and tables. The model takes into account the raw water quality and also provides deductions for various measures that the water plant possesses to ensure a good supply of water. The results obtained with both models were similar and showed that the water treatment is sufficient for the bacteria, but not viruses and parasites. Both models were considered to be reliable but viruses and parasites are very difficult to analyze, which has resulted in uncertain literature values and hence in the results. The result also showed that neither viruses nor parasites exceeded the limit by so much that more hygienic barriers to the reduction of them are necessary. The conclusion which may be drawn from the fact that no parasites have been detected in the raw water is that the water treatment still might be sufficient. To determine the effects that an exclusion of various barriers may give, the normal operation was simulated and a purification step at a time was excluded. The result showed that the purification steps which are most important to maintain the treatment process are chemical precipitation followed by rapid filtration, slow filtration and disinfection with chlorine. If any of these cleaning steps were to fail, this introduces a large increase in the risk of waterborne disease. The results showed that the chemical precipitation step gave the greatest separation effect on the virus but also on the parasites. However, the slow filtration gave the largest separation of the parasites. Free chlorine had the greatest effect on bacteria. The investigated scenarios were assumed to be wastewater discharges, sewage discharges in relation to flood the nearby pastures, and sewage overflows due to heavy rainfall. The results of the simulated scenarios were the same when it was only bacteria that in all cases produced a result within the limits of the daily infection probability. Both viruses and parasites exceeded both values. However, there were few studies on these and thus literature values needed to be implemented in the QMRA-model. Hence, the uncertainty of the results was great. The QMRA-model also contained deficiencies in the simulation of the discharge of effluents, where the amount of virus was about 1000-10000 times too much. If this problem as well as more specific data for the investigated area, and more Swedish studies were available, a more credible simulation of the scenarios could be implemented