ARCADIS conducted a settling study on pulp and paper mill wastewater mixed with various dilutions of receiving river water in order to generate settling rate data used to model the fate and transport of wastewater particles after discharge. The overall objective was to evaluate the impacts and assess the spatial extent of sediment contamination caused by past wastewater treatment practices on downstream receiving water quality in support of a litigation case.
The test apparatus consisted of a receiving water pump tank and filtration system, a flocculation tank with speed controlled, power monitored, mixer, and three 10-foot-tall, 6-inch-diameter, clear PVC settling columns fitted with sampling ports and insulated to control liquid temperatures and prevent convection currents that could affect settling.
Duplicate settling tests were conducted on raw (unfiltered) river water to estimate settling rates of native particles. Triplicate settling tests (with pre-mixing/pre-flocculation) were conducted on each of three different dilutions of wastewater with raw river water.
River water was emptied into the feed tank and pumped through garden hosing and a 0.65-micron in-line filter up into the pre-mixing/flocculation tank. The premixing/flocculation station was located within a "hut" constructed over the testing lab out of two-by-four framing and blueboard insulation and containing a mechanical, shaft-mounted laboratory mixer. A multimeter was connected to the internal leads of the mixer power unit in order to monitor and record current and voltage during mixing. After flocculation, columns were filled and TSS samples were taken at prescribed intervals from sample ports of varying heights over a 12- to 72-hour period.
Flocculation (slow mixing) occurred at a target mixing intensity – specified as a G-value of 30/sec - for 20 minutes. The G-value was calculated in two ways: using power input calculated from current and voltage readings from the multimeter and subtracting out background power draw and using power input calculated using torque and rpm readouts from the mixer controller. The results of power input monitoring showed that the actual, measured G-value ranged between 25 and 32 sec-1.
Temperature was monitored during the settling test using thermocouples that were located at 1 foot, 5 feet, and 8 feet off the bottom of the settling columns and were hard-wired to a thermocouple input board and PC for temperature tracking using VisualBasic software with data points logged every 2 minutes.
Great care was taken to control the ambient temperature and thus the liquid temperature within the columns, both temporally and spatially. Generally, spatial variations (e.g., major temperature gradients across length of columns, acute "hot or cold spots" at locations along the columns, etc.) were considered most critical to control as convection currents generated by such variations could impact settling.
Insulation of the settling columns was effective and spatial variations appeared to be negligible during the four settling tests. Temporal variations were not considered to be of such a significance as to affect settling test results.
All of the settling column test samples were analyzed for TSS by a contract laboratory. In addition to these settling test samples, two samples each of raw river water and wastewater were sent for particle-size distribution testing and particle-density testing. Particle density was determined by drying samples and using a gas displacement technique (pycnometer). Particle size distribution was determined by the Electrozone method.
