Monitoring for pollutants in potable water is an area of interest and concern for water supply companies. Supply of sub-standard water can draw complaints from public and industrial customers. Water and water tainted with pollutants were used to assess the application of a non-specific chemical sensor array (eNose) to monitor for changes in the headspace generated from a flow-cell by means of statistically designed experiments. 2-chlorophenol and diesel were used to further assess trends in headspace generation during trials where different combinations of sampling parameters were applied. Field trials were conducted in accordance with the most suitable methodology determined during initial studies under laboratory conditions. The headspace is generated by bubbling nitrogen through the flow-cell containing a water sample. The liquid sample is flushed and regenerated after each sensor acquisition cycle. The resultant headspace sample is transferred to the sensor array module where the resistance of the conducting polymer sensors is monitored as they are exposed to each respective headspace sample. The change in each sensor resistance after 60 seconds of exposure is used to represent the headspace character. Subsequent acquisitions are added to a data set and then presented graphically. Sudden changes in the sensor resistance plots represent changes in water quality. The results showed that the developed apparatus and sampling methodology can determine the presence or absence of pollution in a water matrix. Laboratory analysis showed that detection levels for 2-chlorophenol and diesel were both <5 ppm in the mixed stream. Future developments should focus on increasing the sensitivity of the system by concentrating the pollutants in either the liquid or gas phase or by modifying the sampling protocol to enable sensor recognition at lower concentration levels. The sensor array could act as a screening technique to support quantitative and characterising analytical equipment at the abstraction point. Establishing a pollution alarm limit, within the bounds of acceptable system variation, would enable conventional analytical techniques to remain on standby until activated by a statistically significant change in water quality. Once established continued testing would enable alarm levels to be incorporated into a contaminant database for additional pollutant compounds and combinations of known taste and odour causing compounds.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:484642 |
| Date | January 2003 |
| Creators | Hogben, Peter James |
| Contributors | Stuetz, Richard |
| Publisher | Cranfield University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://dspace.lib.cranfield.ac.uk/handle/1826/121 |
Page generated in 0.0018 seconds