加拿大艾伯塔省埃德蒙顿市的EPCOR Water市中心水处理厂位于86个雨水管排水口的下游。由于我们担心雨水管排水口可能会产生污染物，因此需要寻找一种在线污染报警监测器，该监测器可以及时进行快速分析，以便采取措施处理污水。分析时间为一分钟的Kontron在线紫外线（UV）监测器似乎能够充当污染监测器。它测量了240 nm至290 nm范围内水源水1000次紫外线扫描基线的标准偏差（SD）。 它会将任何新扫描的标准偏差与上次扫描的标准偏差进行比较，并生成一个称为上次扫描差异（LSD）的值。当LSD与其基线正常标准偏差显著不同时，会触发污染警报。该系统于1998年10月至1999年4月在河水中运行，7月至10月在明矾澄清河水中运行。从这两个运行期收集的结果表明，无论季节和水源水的差异如何，基线标准偏差都是相同的。 发生了一些“报警”事件，但在报警事件期间通过测试目标化合物来识别污染物的尝试没有成功。报警事件期间的LSD数据与颜色、浊度、pH值、挥发性有机化合物、UV254吸光度和氨的比较未显示任何相关性。主要问题是提供足够低浊度且无气泡的河水，以使仪器按照设计工作，这是许多设置挑战。当使用明矾澄清河水时，当警报和污染警报分别设置为平均LSD+3 SD和平均LSD+4 SD时，仪器工作良好。 然而，由于加拿大的服务支持不佳，我们已于2001年6月停止使用该系统。包括参考资料、表格、图表。

EPCOR Water downtown water treatment plant in Edmonton, Alberta, Canada, is downstream from 86 storm sewer outfalls. Our concern of possible contaminants coming from the storm sewer outfalls have resulted in a search for a on-line pollution alarm monitor which can do rapid analysis in time to take action on the treatment of the water. The Kontron On-line ultraviolet (UV) monitor with an analysis time of a minute appeared to be capable of acting as a pollution monitor. It measures the standard deviation (SD) of the baseline of 1000 UV scans of the source water from 240 nm to 290 nm. It will compare the standard deviations of any new scan with that of the last scan and generate a value known as the Last Scan Difference (LSD). When the LSD is significantly different from its baseline normal standard deviations, a pollution alarm is triggered. The system was operated between October 98 and April 99 on river water and from Jul 00 to Oct 00 on alum clarified river water. The results gathered from these two periods of operation indicated that the baseline standard deviation was identical regardless of season and differences in the source water. A few "alarm" events occurred, but attempts to identify pollutants during an alarm event by testing for target compounds were unsuccessful. Comparison of the LSD data during alarm events with color, turbidity, pH, volatile organic compounds, UV254 absorbance and ammonia did not show any correlations. There were many set-up challenges with the major problem of supplying enough river water that is low in turbidity and free of bubbles for the instrument to work as per designed. The instrument worked well when alum clarified river water was used and when the Warning and Pollution Alarm was set at Average LSD + 3 SD and Average LSD + 4 SD respectively. However we have discontinued the use of the system in June 2001 because of poor service support in Canada. Includes reference, tables, figures.