Turbidity is a water quality monitoring parameter that is key to determining the amount of particulate matter present in a water sample. Lower filter effluent turbidity values have been shown to correlate with a reduced risk of pathogens, such as Cryptosporidium and Giardia , passing through the filters to the finished water. Turbidity is also the primary water quality optimization parameter for the Partnership for Safe Water’s treatment plant optimization program, and is a critical monitoring and process control parameter for plants at all phases of the program. Turbidity values can drive plant operational activities, such as coagulation and filter backwash that are essential to producing a safe, high quality water. Because of the critical nature of turbidity measurements and the use of turbidity as a key process control parameter, it is important to ensure that that turbidity data collected by plant instrumentation and staff is as accurate as possible. This can be achieved in a variety of manners. This article focuses on the use of calibration, verification, and sampling techniques to help ensure the accuracy of turbidity data. Calibration The USEPA requires that turbidimeters are calibrated on a quarterly basis. The regulatory requirements for turbidimeter calibration may vary in other areas, but it is important to note that turbidimeters may be calibrated more frequently than required by regulation. They may not be calibrated on a less frequent basis. Turbidimeters are typically calibrated with a primary turbidity standard, such as formazin or a stabilized form of the formazin standard. Both types of primary standards are commercially available. Different instruments and manufacturers may recommend different ranges of standards and techniques for turbidimeter calibration, so it is important to follow the calibration instructions for a given instrument, as provided by the manufacturer. Typically, online turbidimeters with a measurement range of 0-100 NTU will be calibrated with a 20 NTU standard solution. Although 20 NTU is a high value in comparison with typical filter effluent turbidity measurements, calibration at this level minimizes the potential impact of any error on the accuracy of the calibration process and subsequent measurements. More sensitive laser nephelometers are typically calibrated with lower level standards, according to manufacturer instructions. Benchtop turbidimeters are typically calibrated with a range of standards reflecting the measurement range of the instrument. In addition to quarterly calibration, instruments should also typically be calibrated after performing significant maintenance, such as changing the light source or cleaning the instrument body. These activities can impact the amount of light reaching the instrument’s detector, and therefore impact the accuracy of turbidity measurements. Calibrations may be performed by plant operators, instrumentation technicians, or the manufacturer’s service representative – the most important thing is that the instruments are calibrated properly at the frequency required by the regulatory agency. Calibration records should also be maintained according to regulatory requirements. A standard operating procedure (SOP) for instrument calibration is highly recommended so that all staff calibrate instruments in a consistent manner. Verification The accuracy of turbidimeter readings should also be periodically verified. Many utilities will verify online and benchtop turbidity measurements on a weekly or monthly basis. Turbidity readings can be verified in a variety of manners including use of a primary turbidity standard, use of a secondary turbidity standard, or by comparison with another instrument. To verify turbidimeter readings using a primary standard, a volume of primary standard with a known turbidity value may be poured into a sample cell and read on a benchtop instrument or poured into a calibration cylinder/chamber to be read using an online instrument. The instrument reading is then compared with the standard’s value. Typically, if the instrument reading matches the value of the standard within a given percent or NTU value (for example +/- 10% or 0.05 NTU – this may vary depending regulatory requirements), the instrument can be considered to be reading accurately. Although calibration may be performed at higher turbidity levels, verification may be performed closer to the range of sample measurements, and stabilized low-level turbidity standards are generally commercially available. Benchtop and online instruments may also be verified using secondary standards. Secondary standards for benchtop turbidimeters typically consist of gel-based standards. Secondary standards for the verification of online instruments may vary in design depending on the manufacturer but are typically consist of a set of filters designed to represent various turbidity readings. It is important to follow the manufacturer’s instructions for the type of secondary standard and instrument being used. Most secondary standards will be read and assigned a value directly after completion of a primary standard calibration. The secondary standard is then measured periodically, between calibrations, and its value compared with the assigned value. A significant increase in deviation may indicate the need to recalibrate the instrument. Secondary standard verification can be a convenient way to ensure accuracy since it does not require the use of a chemical or liquid reagent. Online turbidimeter performance may also be verified by comparison with a grab sample that is measured in the lab. If a grab sample comparison is used to verify turbidimeter performance, it is important that the sample collected for lab measurement is representative of the sample measured by the online turbidimeter. The optimal place to collect this sample is from the turbidimeter drain. Use caution when collecting samples from laboratory sinks, particularly if they are located a long distance from the online turbidimeter, as a significant lag time can mean that the sample collected in the lab is not representative of that being measured by the online turbidimeter. When verifying filter effluent turbidity samples at very low NTU values, it is essential to use good lab technique for low-level turbidity samples. For example; ensure sample cells are excruciatingly clean, use silicone oil on the outside of the cell to mask minor scratches, avoid introducing air bubbles to the sample, and minimize the impact of condensation on the cell. Standard should be set by the plant to determine the level of agreement between lab and online turbidity values that is considered to be acceptable. If an instrument’s 4-20mA output is used to communicate with the plant’s SCADA system, the accuracy of the instrument output and SCADA value should also be periodically verified. This ensures that an accurate reading is relayed to the plant’s SCADA system - which is important because it may be used for process control. Verifying this can be achieved by comparing the instrument local turbidity reading with that displayed on the SCADA computer. If the values do not agree, further troubleshooting may be required. No matter how turbidimeter verification is achieved, SOPs are recommended to ensure that verification procedures are being carried out in a consistent manner among all plant staff. Turbidity Sampling The accuracy of turbidity readings is also dependent on sampling technique. To obtain an accurate measurement, the sample must be representative of the water being measured. For online instruments, this equates to installing turbidimeters as close as possible to the sample being measured. For example, consider locating turbidimeters directly in the filter gallery, rather than in a laboratory location that may be located a significant distance from the sample. Ensure that samples are collected from an area of the pipe free from sediment buildup, and be sure that the flow rate provided to online instruments is consistent with the manufacturer’s specifications (an air gap on the drain can be helpful for providing a quick visual confirmation of flow). Although many online turbidimeters incorporate internal bubble traps in their design, a separate external bubble trap may be required for samples that contain an excessive amount of entrained air, in order to avoid air bubbles causing a positive interference in the turbidity measurements. It is also important to ensure that the turbidimeter itself is suitable for the application and expected turbidity levels. For example, a filter effluent turbidimeter may not be suitable for monitoring raw water turbidity, due to limitations in range and the potential for maintenance issues caused by high levels of suspended solids introduced into the turbidimeter body. Sampling is also important for benchtop turbidity measurements. As described previously, when collecting grab samples or online instrument calibration, be sure that they samples are representative of the water being measured by the online turbidimeter. Also, when collecting samples for laboratory analysis, consider collecting them in a clean sample cell to minimize any potential sources of contamination as well as the introduction of air bubbles caused by sample transfer. Finally, as with many other parameters, it is recommended to measure turbidity sample as soon as possible after sample collection. If samples are to be stored for short durations of time, be sure to follow the guidance provided in Standard Methods for sample storage techniques to minimize the impact of microbial activity on the turbidity value. Establishing consistent and accurate calibration, verification, and sampling techniques can help ensure the accuracy of plant turbidity measurements. At the heart of the Partnership for Safe Water’s treatment plant optimization program, turbidity is an essential part of determining a plant’s optimization status, quantifying plant performance, and the plant’s process control system. Ensuring accurate turbidity measurements helps to provide better data for driving plant decision making and – ultimately – achieving optimization.