Runoff from construction sites has received increased interest because high levels of turbidity can adversely impact aquatic life in receiving streams. The current practice is to use polyacrylamide (PAM) to flocculate and settle suspended particles prior to release of the storm water into the environment. Not much, however, is understood about factors that control the interactions between PAM and soil particles. The goal of this study was to both identify the factors that lead to optimal turbidity reductions and determine the best screening method that can be applied on construction sites.
Soil from 22 counties in North Carolina were collected and tested for flocculation with 13 PAMs. These had charge densities from 0.0 to 30% and molecular weight in the ranges standard (STD), medium (SH), and high (VHM). During the preliminary screening, soil suspensions were prepared at 10 g/L and tested with PAM concentrations ranging from 1.0 to 250 mg/L. Upon hand shaking for 10 seconds and sedimentation for 30 seconds, the supernatant water turbidity was measured. Nonionic polymers were more effective in reducing turbidity than their anionic counterparts. PAM concentrations of 1.0 and 5.0 mg/L led to the lowest turbidities in all soils tested, and increasing PAM concentrations gradually resulted in increased turbidities. The effect of PAM molecular weight was found to be dependent on the charge density of the PAM in use. Larger turbidity reductions were observed in soils with higher clay and silt content relative to the sandy soils.
Using a jar tester, the soil suspensions were also mixed with PAM at different intensities (G = 48, 130, and 640 s-1) for periods of 20 to 600 seconds. The results indicated that mixing intensity plays a key role in the flocculation of sediments. Only G values of 130 and 640 s-1 resulted in measurable turbidity reduction compared to the hand shaking test. The highest turbidity reductions were achieved at G = 130 s-1. In contrast to the hand-shaking results, anionic PAMs were more effective at reducing turbidity than the nonionic one. This suggests that the choice of the most effective PAM also is dependent upon the screening method used. Increasing mixing time using the hand-shake test negatively affected the performance of the nonionic PAM with the soils with substantial clay and silt content. However, on the jar tester, an increase in mixing time resulted in reduced turbidity for all soils, regardless of the PAM used.
To evaluate the effectiveness of PAM on the field relative to the laboratory experiments, two 17-meter-long model ditches were constructed at the Sediment and Erosion Control Research and Education Facility (SECREF) of the Crop and Soil Sciences Department of North Carolina State University. Four PAMs having charge density 0, 3, 10, and 30%, respectively, were used. Following the flocculation tests, conducted with 0, 1, and 3 check dams installed across the channels, the anionic PAM with 3% charge density consistently achieved the highest turbidity reductions in all soils tested. This suggests that the jar tests may better predict PAM performances on construction sites, compared to the hand-shake method. Furthermore, no significant difference was found between the effects of 1 and 3 check dams on turbidity reduction in all soils used for the tests.
