Implementation and Calibration of an IoT Light Attenuation Turbidity Sensor

Abstract

Turbidity is an important characteristic of water quality that can indicate the presence of undesirable suspended particulate matter. Having access to an inexpensive and effective turbidity sensor unlocks numerous Internet of Things (IoT) possibilities for remote environmental monitoring. Optical light attenuation turbidity sensors operate on the premise of detecting signal degradation from a light source due to scattering from particles in a solution. This approach is technologically unpretentious and only requires a handful of inexpensive electronic components to construct. However, while this method is touted as “simple”, a significant challenge lies in sensor calibration. That is, converting an analogue signal into a meaningful and accurate digital reading in a known turbidity measurement standard (e.g., Nephelometric Turbidity Units (NTU)). This paper presents an IoT light attenuation turbidity sensor design and explores the calibration process to determine the sensor's range and accuracy. Sensor calibration is undertaken using Formazin turbidity standards and is cross-checked against a commercial turbidimeter. We provide a step-by-step procedure for determining the correct signal strength to use and a functional form for the sensor response to the Formazin standard. Finally, we specify an estimate of the accuracy of the sensor and suggest the next steps in the proposed turbidity sensor's development. Results indicate that the sensor achieves within 2%-10% of accuracy at higher ranges (100–4000 NTU), but its performance becomes significantly less reliable in low NTU ranges (< 100 NTU) where the error rate increases to 26%. The turbidity sensor is used as part of an IoT remote aquatic environmental monitoring platform.