WATER RESOURCES BULLETIN VOL. 29, NO. 1 AMERICANWATERRESOURCES ASSOCIATION FEBRUARY 1993 SLUDGE APPLICATION EFFECTS ON RUNOFF, INFILTRATION, ANDWATERQUALITY' A. C. Bruggeman and S. Mostaghimi2 ABSTRACT: Land application of sewage sludge requires careful sludge have been reviewed by Kelley et al. (1984). monitoring because of its potential for contamination of surface These problems include accelerated eutrophication of water and ground water. A rainfall simulator was used th investi- surface waters by phosphorus, ground water contami- gate the effects of freshly applied sludge on infiltration, and on runoff of sediment and nutrients from agricultural crop lands. Rain nation by leaching of nitrate, and health risks from was applied to 16 experimental field plots. A three-run sequence transfer of pathogenic organisms to humans or ani- was used to simulate different initial moisture conditions. Runoff, mals and from the uptake of trace elements (heavy sediment, and nutrient losses were monitored at the base of each plot during the simulated rainfall events. Sludge was surface metals) by plants. applied and incorporated at conventionally-tilled plots and surface Several authors have studied the effect of sludge applied at no-till plots, at rates of 0, 75, 150 kg-N/ha. Steady-state application on the quality of runoff water from agri- infiltrability increased as a result of sludge application, although cultural lands. Kelling et al. (1977) found significantly the no-till practice was more effective in increasing the infiltrability lower runoff and sediment losses from sludge treated than the sludge application. No-till practices greatly reduced runoff, sediment, and nutrient losses from the sludge treated plots, areas, relative to untreated areas. However, relative to the conventional tillage practices. Incorporation of orthophosphorus (P04-P), total phosphorus (Pt), and the sludge was effective in reducing nutrient yields at the nitrite plus nitrate (N02+N03) loads in runoff water conventionally-tilled plots. This effect was more pronounced during the third rainstorm, with wet initial conditions. Peak loadings of from the sludge treated plots increased, compared to nutrients appeared during the rainstorm with wet initial condi- the control plots. Dunnigan and Dick (1980) found tions. that surface application of sewage sludge resulted in (KEY TERMS: sludge; land application; infiltration; runoff; nutri- higher loadings of nitrogen (N) and phosphorus (P) in ents; water quality.) runoff water, as compared to loadings from plots in which sludge was incorporated into the soil. The beneficial impacts of sludge applications to reduce surface runoff and soil erosion can be related INTRODUCTION to reduction of raindrop impact by the surface protec- tive sludge layer and increased infiltration as a result Application to agricultural lands is a cost-effective of the improvement of the physical condition of the alternative for disposal of sewage sludge. Secondary soil (Mostaghimi et al., 1988). Significant increases in benefits of application of sludge to agricultural lands aggregation and large pore space as a result of sludge include nutrient supply to crops, increase of soil application were observed during a one-year field organic matter content, and improvement of soil prop- study by Kladivko and Nelson (1979). Otis (1985), erties. However, the potential hazards of surface however, reported that application of waste water water and groundwater contamination from land dis- effluent significantly reduced the hydraulic conductiv- posal of sewage sludge is a major environmental con- ity and infiltration rate of soils due to pore clogging. cern. Potential problems of land application of sewage He attributed pore clogging to the suspended solids in 'Paper No. 92037 of the Water Resources Bulletin. Discussions areopenuntil October 1, 1993. 2Respectively, Graduate Research Assistant, 200 Seitz Hall, Agricultural Engineering Department, VirginiaPolytechnic Institute and State University, Blacksburg, Virginia 24061; and Associate Professor, 308 Seitz Hall, Agricultural Engineering Department,Virginia Polytechnic Institute and State University, Blacksburg, Virginia 24061. 15 WATER RESOURCES BULLETIN Bruggeman and Mostaghinii the waste water effluent, which mechanically sealed spread, as uniformly as possible within the subareas, the entrances to the larger pores. In addition, the using rakes. Within each tillage treatment two sludge nutrients in the waste water, which stimulated the application rates, representing 150 kg-N/ha and 75 biological activity in the soil, causing breakdown of kg-N/ha, and an untreated control plot, were estab- the structural units of the soil and, in turn, increased lished. Loading rates were determined using the pore clogging. Chang et al. (1983) indicated that the methodology suggested by Simpson et al. (1985). The amount of sludge required to cause significant sludge was black in color and fairly crumbly when changes in soil properties was much greater than the applied. The high and low application rates provided amount normally used to satisfy crop nutrients an average sludge layer of approximately 1.0 and 0.5 requirement. cm, respectively, although no uniform cover was The objective of this study was to investigate the obtained. Sludge was surface applied to all no-till short-term impacts of sludge application method, plots. For the conventional tillage plots, sludge was application rate, and tillage practice on infiltration, both incorporate and surface-applied. The plots with runoff, and sediment and nutrient losses from agricul- the incorporated treatments were retilled to incorpo- tural lands. rate the sludge into the upper 15-20 cm of the soil profile. Two replications of each treatment required a total of 16 plots. Plot treatments are listed in Table 1. All treatments were randomly assigned to the experi- MATERIALS AND METHODS mental plots. Sixteen experimental field plots, located at Virginia Tech's Price's Fork Agricultural Research Farm, 10 km west of Blacksburg, were used for this study TABLE 1. Soil and Sludge Characteristics and Plot Treatments. (Mostaghimi et al., 1988). Plots are located on a Soil Characteristics Groseclose silt loam soil (clayey, mixed mesic Typic Hapludult). The soil is deep and well drained with a Type: Groseclose Silt Loam Percent* Sand 17.9 slowly permeable subsoil. The plowed surface (Ap) Percent Silt 58.9 horizon is typically 25 cm thick, with a loam texture Percent Clay 23.2 and moderately fine granular structure. Soil charac- Percent Organic Matter 3.7 teristics are presented in Table 1. Bulk Density 1.39 g/cm Each plot had a surface area of 0.01 ha (18.3m x Initial Moisture Content 19.2% 5.5m). Metal borders were installed around each plot Sludge Characteristics to prevent surface and subsurface runoff from flowing across the plot boundaries. The 30 cm high borders Type: Anaerobically Digested Sewage Sludge Percent Solids 16.0 were inserted to a depth of about 15 cm. A concrete Percent NH4-N 0.96 gutter with a pipe outlet at the base of each plot col- Percent TKN 3.02 lected and transported the runoff to an H-flume. All Percent Phosphorus 2.0 plots were planted in winter rye in the fall and pH 7.3 sprayed with paraquat, a week before the simulation Plot Treatments runs, in early spring of the following year. No-till treatments were established on the killed rye stand. Sludge Sludge Crop residue was measured by randomly locating a Tillage Application Application 0.6 x 0.6 m square in each plot and removing all Method Method Rate residue in the square for laboratory analysis. No-Till N/A 0 Conventional tillage was represented by removing No-Till Surface 75 crop residue from the plots, tilling to a depth of 15-20 No-Till Surface 150 cm with a power-take-off (PTO) driven rototiller, and Conventional N/A 0 Conventional Surface 75 disking. Conventional Incorporated 75 Anaerobically digested, polymer conditioned Conventional Surface 150 sewage sludge was obtained from the James River Conventional Incorporated 150 plant in Hampton Roads, Virginia. General character- *All percentages are on a dry weight basis. istics of the sludge are given in Table 1. Sludge was NOTE: N/A =NotApplicable. distributed over the plots by subdividing each plot into four equal-sized subareas. The sludge was WATERRESOURCES BULLETIN 16 Sludge Application Effects on Runoff, Infiltration, and Water Quality A rainfall simulator (Dillaha et al., 1987) was used RESULTS AND DISCUSSION to apply approximately 90 mm of rainfall to the plots over a two-day period. A 60-minute initial dry run Infiltration and Peak Runoff (Ri) was followed 24 hours later by a 30-minute wet run CR2), followed 30 minutes later by another 30- Infiltration and runoff rate data collected during minute very wet run (R3). The three-run sequence is a common artificial rainfall sequence used for erosion the first (dry) run (Run 1) and the third (very wet) run research in the U.S. to represent different initial (Run3) are presented in Table 2. Runoff rates record- moisture conditions (Dillaha et al., 1987). A 40-45 ed during the second (wet) run (Run2) were generally mm/hr rainfall rate was used for all simulations. A higher than those from Runl, but always lower than one-hour storm with a 40-45 mm/hr intensity has a Run3. The effect of the initial soil moisture condi- tions, as represented by the sequence of runs, is most 2-5 year return period in Southwest Virginia obvious when comparing Runi and Run3 data. Thus, (Hershfield, 1961). The sludge was applied 24 hours data on Run2 are not presented here. For most treat- before the first simulated rainstorm. Rainfall within ments, steady-state conditions were approached at the 24 to 48 hours after the application of sludge is the end of the third (30 minute) run, as can be seen in expected to result in extremely high loadings of nutri- Figures 1-3, except for the sludge treated no-till plots ents to surface runoff, since sludge had been applied and the conventionally-tilled plots, without sludge within the previous 24 to 48 hours.