人眼变成电子眼——《改进的过程监测杜绝了浑浊度超标的根本原因》

原文出处：J Opflow

翻译：阮辰旼

Improved Process Monitoring Isolates Root Cause of Turbidity Events

改进的过程监测杜绝了浑浊度超标的根本原因

Public health organizations throughout the world have recognized the importance of measuring drinking water quality via turbidity. Turbidity can be measured with online, laboratory, or field instrumentation, but online measurement allows drinking water producers to constantly monitor their operations and ensure proper filtration. In addition, process turbidity measurement should respond quickly to turbidity changes to ensure a prompt reaction to potential filter breakthrough and other unwanted events.

世界各地的公共卫生组织已经认识到通过浑浊度来表征饮用水水质的重要性。浑浊度可以通过在线监测、实验室检测或现场仪器测量获得，其中在线监测的方式可以允许运营企业不断监测他们的运行操作效果并确保采用适当的过滤操作。此外，浑浊度的过程监测能快速响应浑浊度变化，以确保对潜在的过滤设备穿透和其他不利的事件发生时能够及时作出反应。

A recently commissioned drinking water facility in Parker, Colo., which is equipped with ceramic membranes for microfiltration, experienced daily occurrences of turbidity events. The plant treats surface water from a reservoir and employs preoxidation, high-rate sedimentation, and organics removal with ballasted sand and activated carbon before filtration. The ceramic membranes, a relatively new filtration process, were organized to alternate the work of individual membrane skids as well as clean-in-place (CIP) and backwash processes.

科罗拉多州帕克市对一个最近投入使用的饮用水设施配备了微滤的陶瓷膜，但每天都会发生浑浊度超标。该水厂处理的原水是来自水库的地表水，并在过滤前采用预氧化、高速沉淀，并用压载砂和活性碳去除有机物。作为一个相对较新的过滤工艺，陶瓷膜组件通过分组，交替运行，以便原位开展清洗（CIP）和反冲洗的过程。

The plant faced intermittent operational challenges dictated by water demand. Frequent production disruptions resulted in high turbidity levels that violated the utility’s self-imposed strict standards and, at times, state requirements for the quality of finished drinking water released into the distribution system (Figure 1).

该水厂间歇性运行操作的现状正面临着水量的需求带来的挑战。频繁的生产中断导致较高的浑浊度水平，违反了供水企业自我设定的严格标准，有时也违反了国家对进入输配系统中的成品饮用水水质的要求（图1）。

The spikes in turbidity readings on combined filter effluent (CFE) occurred at plant startup daily and repeatedly caused automatic shutdown. At first, plant personnel suspected malfunctioning analytical instrumentation and reached out to the original equipment manufacturer to help determine the root cause. Based on the plant’s treatment process and routinely low turbidity readings, the manufacturer suggested using a more sensitive, faster-responding turbidimeter than the instrumentation used by the facility. Implementing a different turbidimeter would also eliminate potential variables and verify the cause of the troubles— the instrumentation or the process.

所有膜组滤后水混合后的过滤出水（CFE）浑浊度读数的峰值在水厂每天启动时出现，并多次导致水厂运行的自动中断。起初，水厂人员怀疑是分析仪器出现故障，并联系了原始设备供应商以帮助确定故障发生的根本原因。根据该水厂的处理过程和常规的低浑浊度读数情况，供应商建议使用比该厂在役的仪器更敏感、反应更迅速的浊度仪。使用不同的浊度仪可以消除潜在的不确定性变量，并帮助验证是仪器的故障还是工艺的原因。

A new Hach TU5400sc turbidimeter was installed at the existing sampling location for CFE beside two older Hach FT660sc analyzers, and the logged data from all three instruments were collected and analyzed. Figure 2 shows the analysis process and results.

在现有的混合后的滤后水取样地点安装了一台新的Hach TU5400sc浊度仪，边上是两台旧的Hach FT660sc分析仪，收集和分析了所有三台仪器的记录数据，图2显示了分析的过程和结果。

The data collected at the CFE clearly correlated with the spikes found in the production flow and didn’t reveal any instrumentation error, but the results were inconclusive. Next, the operators installed another TU5400sc on an individual membrane skid—analogous to individual filter effluent (IFE)—and on the same sample line as an existing FT660sc turbidimeter.

对混合后的滤后水浑浊度采集的数据与生产流程中发现的峰值情况有明显相关，暂没有发现任何仪器故障，但结果并不确定。接下来，操作人员将另一台TU5400sc安装在一个单独的膜组上，检测单一模组单独的过滤出水（IFE）的浑浊度，这个新的浊度仪与现有的FT660sc浊度仪安装在同一条生产线上。

As shown in Figure 3, the sample flow through the turbidimeters was intermittent, unlike the test at the CFE, and reflected the skid operation schedule with regular backwash and CIP procedures. Both instruments showed low levels of turbidity when the skid was in operation, which was expected. However, the detailed analysis of the data revealed different implications about the root cause.

如图3所示，通过浊度仪的样品流态是间歇性的，与混合后的滤后水的检测不同，它反映了有定期反冲洗和自清洗程序的周期性的运行规律。两台仪器都显示了膜组运行时的低浑浊度水平，这是符合预期的。然而，再对数据进行详细的分析可以揭示关于根本原因的不同理解。

The FT660sc mostly reacted with small turbidity spikes at the beginning of each operation cycle, but the TU5400sc displayed disturbance mostly at the end of each cycle. The % relative standard deviation (RSD) data generated by each instrument were analyzed, showing that the particulate matter probably changed properties, such as particle size and distribution, during the skid operation, including the backwash and CIP cycle. The new turbidimeter consistently showed a warning about vial clarity, which was difficult to understand until some sediment collected in the vial was discovered during a visual inspection (Photo 1).

FT660sc大多在每个运行周期开始时出现小的浊度峰值，但TU5400sc大多在每个周期结束时出现干扰。对每台仪器产生的相对标准偏差（RSD）的数据进行了分析，显示颗粒物在包括反冲洗和自清洗循环在内的操作过程中可能改变了特性，如颗粒大小和分布。新的浊度仪一直显示出关于样品瓶透明度的警告，这很难理解，直到在目测中发现样品瓶中收集的一些沉淀物（照片1）。

To further discern whether the analyzer or the process was the main problem, the plant’s operators decided to implement an automatic cleaning module (ACM) and test it at the CFE location to see if it would help to remove the particles and avoid the turbidity spikes to at least prevent automatic shutdown at the plant. But even if the ACM could help, it wouldn’t eliminate the root cause, so the operators continued analyzing the plant’s piping design and processes. Meanwhile, a TU5400sc with ACM was placed at the same CFE sampling point as in the original test.

为了进一步辨别是浊度仪的问题还是工艺的问题，该水厂的操作人员决定实施一个自动清洁模块（ACM），并在混合后的滤后水的采样点进行分析，看看它是否有助于清除颗粒物质，起到避免浑浊度飙升，防止水厂运行的自动中断的作用。但是，即使自动清洁模块可以提供帮助，它也不能从根本上解决问题，所以操作人员需要继续分析工厂的管道设计和流程。同时，在最初测试的同一混合后的滤后水的取样点上放置了一个装有自动清洁模块的TU5400sc。

The ACM slightly improved the turbidimeter’s performance, but it didn’t eliminate the problem entirely. This confirmed the notion that the root cause of the problem was in the process, not in the analyzer.

自动清洁模块稍微改善了浊度仪的性能，但并没有完全消除这个问题。这证实了一个观点，即问题的根源在于工艺，而不在于分析仪。

The test confirmed the following facts:

这个试验确认了以下事实：

●The pipes tapped for turbidity sampling were collecting sediment, which was then transferring to the turbidimeters.

用于为浊度仪采样的管子内不断沉积颗粒物，并将这些沉积的颗粒物其转移到浊度仪中。

●The sediment was washing off at every startup and caused the turbidity spikes appearing in the readings.

沉积的颗粒物在每次仪器启动时都会被冲刷走，这就导致了浊度仪的读数中出现的峰值。

●Although the intermittent presence of particulates in the CFE wasn’t representative of the overall treated water quality, it displayed as the turbidity compliance events.

尽管在混合后的滤后水中间歇性出现的颗粒物并不能代表工艺处理后的真实水质，但它却被记录为浊度超标事件。

●The new turbidimeter helped to resolve yet another issue, which appeared after filter backwash operations. Along with rising levels of turbidity after backwash, invisible to the previous turbidity instrumentation, iron oxides and other sediments accumulated in the TU5 vials, which led the operators to realize that the filter-to-waste system wasn’t properly implemented. As a result, after each skid finished the backwash cycle, the backwash discharge would blend with finished water, causing rising turbidity in the CFE and the distribution system.

新的浊度仪在过滤设备的反冲洗操作之后帮助解决了另一个问题。随着反洗后浑浊度水平的上升，以前的浊度仪观察不到氧化铁和其他沉积物在TU5样品瓶中积累，这使操作人员意识到膜的废水排放系统没有被正确安装。因此，导致了在每个周期膜组完成反冲洗后，反冲洗废水会与成品水混合，导致混合的滤后水和进入输配系统的水的浑浊度上升。

Figures 4a and 4b show the latest confirmation of the backwash influence. The filter backwash involves water from the distribution system pressurizing a tank that, in conjunction with a separate pressurized air tank, enters the filter element on the filtrate side. When the process is finished, the water that remains on the filtrate side is the water from the distribution system, preventing a pressure differential across the membrane. The water from the distribution system is a blend of the plant’s production water and water from an external source, which is, in turn, a blend of treated ground and surface water and may contain some amounts of iron and manganese. In the instance depicted in Figures 4a and 4b, the external source was experiencing a higher concentration of iron, which entered the backwash tank and consequently the filter during a backwash. When that filter returned to service, the water that remained on the filtrate side entered the CFE, resulting in higher turbidity readings. This prompted a modification of the program and introduced an operator-adjustable filter-to-waste function to flush that remaining water before that individual filter rack would begin producing water after backwash.

图4a和4b显示了对反冲洗的影响的最新确认。膜过滤器的反冲洗过程是来自输配系统的水通过水箱进行加压，该水箱的水与一个单独的加压空气箱的加压空气混合，进入滤液一侧的过滤单元。当这个反冲洗过程结束后，留在滤液一侧的水是来自输配系统的水，防止了膜上的压力差。来自输配系统的水是水厂出厂水和外部水源的混合水，而外部水源又是经过处理的地下水和地表水的混合水，可能含有一些铁和锰的量。在图4a和4b中，外部水源的铁浓度较高，在反洗过程中进入反洗池，从而进入过滤器单元。当该过滤器恢复使用时，留在滤液一侧的水进入混合的滤后水，导致浑浊度读数升高。这个情况促使我们对程序进行了修改，并增加了一个将过滤废水水进行排放的功能，在膜组反洗后并开始生产水之前，对膜组内剩余的水进行冲洗。

In the case of the Parker plant, filter-to-waste function was added, chemically enhanced backwash and CIP were reconfigured for additional chemical contact leading to more efficient flushing, and the turbidimeter taps were rerouted to better accommodate the changes. Since the improvements were implemented, the facility has minimized turbidity events to ensure and prove consistent water quality. The findings drove gradual replacement of trusted older turbidimeters with the new technology. The fast response and accuracy of the new instruments on individual filter racks, CFE, and backup lines help utilities compare real-time process changes with turbidity readings, better understand the root cause of issues, and implement solutions.

就帕克市的水厂而言，增加了膜过滤器废水排放的功能，重新配置了化学强化反洗和自清洗程序，增加了化学接触，提高了冲洗效率，并重新调整了浊度仪的安装点位，以更好地适应这些变化。自从这些改进措施实施以来，该设施已将浑浊度异常事件降至最低，以确保并实现了水质的一致性。研究结果推动了用新的浊度仪技术逐步取代以往值得信赖的旧的浊度仪。新仪器在单个膜组、整体的混合滤后水和备用管线上的快速反应和准确性帮助公用事业部门将实时工艺变化与浑浊度读数进行比较，更好地了解问题的根本原因，并实施解决方案。