Environmental problems facing the livestock

Transcription

1 se 1818 ms 7/9/01 10:47 AM Page 981 MANURE SOLIDS SEPARATION BY FILTRATION WITH FOUR CROP RESIDUES M. Zhang, J. C. Lorimor ABSTRACT. Laboratory research was conducted to determine the effectiveness of using five crop residues as filters to separate solids from liquid swine manure. Solids removal efficiencies and plugging characteristics were examined. Oat straw, soybean stubble, and corn stover were found to be effective filter materials; corncobs and ground corncobs were not. As expected, solids removal efficiencies were higher for manure with 6.0% solids initially than for manure with 1.4% solids, but plugging occurred sooner. Most of the separation occurred within the top portion of the filters. Keywords. Biomaterial, Biofilter, Animal waste, Water quality, Liquid-solids separation. Environmental problems facing the livestock industry, especially concerns for odors and water pollution, have increased the pressure on livestock operators to provide better treatment of swine manure. Solid-liquid separation is one manure treatment process that can be of benefit. Solid separation produces a solid fraction and a liquid fraction. The solid fraction can be recycled as livestock feed or bedding, or transported (due to more concentrated nutrients) to be applied to fields as fertilizer or a soil conditioner. If composted, the solid fraction can provide excellent mulch for nursery and landscape use or serve as bedding. The liquid fraction may be irrigated, and is easier to handle with standard pumping and piping systems. The reduced organic loading of the liquid fraction helps minimize odors and increase the service life of lagoons and storage basins. Most separation methods are based on particle size and particle density differences. While advanced filtration methods have been developed for water treatment, and municipal and industrial waste treatment systems, less has been done with animal manure. The two most common methods of solids separation in animal operations are settling basins and mechanical separators such as screens and screw presses. Mechanical reliability and cost are major issues with mechanical systems. Settling basins work best in open feedlot applications. Removal efficiency is a critical parameter for the evaluation of separation systems. Moore et al. (1975) measured settling efficiency with time. They reported that 52% of total solids (TS) were removed from 1% solids Article was submitted for publication in January 2000; reviewed and approved for publication by the Structures & Environment Division of ASAE in June Presented as ASAE Paper No This is Journal Paper No of the Iowa Agriculture and Home Economics Experiments Station, Iowa State University, Project No Funding for this study was provided by the Iowa Pork Producers Association. The authors are Meian Zhang, Graduate Student, and Jeffery C. Lorimor, ASAE Member Engineer, Assistant Professor, Agricultural and Biosystems Engineering Department, Iowa State University, Ames, Iowa. Corresponding author: Jeffery Lorimor, Iowa State University, 200A Davidson Hall, Ames, IA 50011, phone: , fax: , address: <jclorimo@iastate.edu>. swine manure in the first minute of settling, 62% of TS settled after 10 min, and 66% of TS were removed at 100 m Huijsmans et al. (1984) evaluated the separation efficiencies for four different mechanical separators, ranging from 12 to 25%. Zhang and Westerman (1997) reported a range of efficiencies from 3 to 67% for various separators; most were well under 50% efficient. Separation efficiencies for swine manure are generally much lower than for beef or dairy manure. Satisfactory separation of swine manure is hard to achieve partly due to the manure characteristics: small particle size, nonhomogeneous, slimy, corrosive, and abrasive. Although removing solids from swine manure is desirable in many situations, most of the existing separation techniques are either inefficient, costly, or both. This study was designed as a preliminary investigation of filtering swine manure using readily available crop residues as filter media. MATERIALS AND METHODS A 25-cm-diameter 58-cm-tall PVC cylinder was used in a laboratory study to hold the biofilter materials. A bottom drain allowed collection of filtered liquid manure. Two expanded metal screens inside the cylinder, one on the bottom and one on the top of the filtration materials, established the filter volume (the position of the top screen was adjustable). The screens were coarse enough (> 0.6 cm) to not act as filters themselves. The bottom screen prevented possible loss of filtration material through the bottom opening. Biomaterials tested in this study were oat straw, soybean stubble, corn stover, and corncobs. Each test was replicated four times. Before filtration, the selected filtration biomaterial was packed in its separation cylinder with specific orientation, depth, and density. The freshly collected liquid swine manure used in this study had a TS content of 10 to 12%. It was diluted with clean water to approximately 4% TS before use. For filtration, 3.78 L (one gallon) of swine manure was applied on the top of the filter. The liquid fraction was collected in a 3.78 L container below the outlet of the cylinder. The duration for the filtration process was recorded. After measuring the volume, the filtrate was completely mixed and samples were taken immediately for TS analysis. This Transactions of the ASAE VOL. 43(4): American Society of Agricultural Engineers / 00 /

2 se 1818 ms 7/9/01 10:47 AM Page 982 process was repeated (3.78 L/time) until the filter was eventually plugged. Total solids were determined using Standard Methods for Water and Wastewater Analysis (APHA,1992). Removal efficiency (percentage of solids removed) was calculated based on the normalized difference between concentrations of TS in applied manure and filtrate. Removal efficiencies based on mass differences would increase since some liquid was retained within the filters. Initial tests were run with a 30 cm depth of oat straw. Oat straw was tested oriented horizontally and vertically at densities of and g/cm 3, respectively. Densities were established by placing known weights of filter material in a preset length of cylinder. The TS concentration of applied manure was 4.3%. The horizontally oriented medium was observed to be more likely to plug quickly, and showed no improvement in separation efficiency, compared to the vertically oriented medium. Therefore, further testing (at four different densities) was performed only with the vertical orientation. Each biomaterial was tested for removal efficiency, filtration duration, and total volume of liquid manure that could be filtered before plugging. Plugging was determined to have occurred when liquid would no longer infiltrate the biofilter. Soybean stubble and corn stover filter density was tested using 4.13% TS manure. For corncobs, two orientations (random and vertical) and two filter depths (4 and 8 cm) were tested. In addition, ground corncobs (average size was about 1 cm) were tested at filter depths of 1, 1.5, and 3 cm. Because most of the solids were trapped in the top portion of the filters (as would be expected) with very little being retained in the bottom, 10-cm oat straw filter depth was tested for comparison to the 30 cm depth. To test the effect of initial manure TS on removal efficiency, manure with initial solids concentrations of 1.4, 2.6, 3.9, 4.8, and 6.0%, was tested using the 10-cm oat straw filter with a density of g/cm 3. RESULTS AND DISCUSSION The initial TS of swine manure tested ranged from 3.6 to 4.3%. Among the media tested, oat straw, soybean stubble, and corn stover were found to be effective, with average removal efficiencies from 24.5 to 42.2% based on the difference in TS concentrations. Corncobs and ground corncobs were basically ineffective, due to either low removal efficiency or plugging problems. The oat straw filter had the higher ratio of manure filtered/unit biomass, compared to the other media. Manure filtered/unit of biomass was higher for the 10-cm oat straw filter than for the 30-cm depth filter because of similar removal efficiencies resulting from the shallower biomass depth. OAT STRAW The removal efficiency and the ratio of manure filtered/unit biomass for the preliminary horizontally and vertically oriented oat straw tests are given in table 1. The filter depth was 30 cm. The initial TS of the applied liquid manure was 4.3%. The lower ratio of manure filtered/unit biomass with the horizontal straw resulted from faster plugging. Since the removal efficiencies were similar, but Table 1. Solids removal for 30-cm-deep oat straw tested horizontally and vertically Initial Manure Bio- Filter Manure Filtered Removal material Density TS to Biomass Efficiency* Type (g/cm 3 ) (%) (L/kg) (%) Horizontal Vertical * 100 (solids concentration in concentration out)/solids concentration the horizontal straw plugged much faster, we selected the vertical orientation for further study. The average removal efficiency for the vertical orientation of 30-cm oat straw at various densities ranged from 37.8% to 42.2% as shown in table 2. Most of the solids were trapped within the top 10 cm of the filter. Therefore, the filtration test was repeated on the 10-cm straw filter with different densities. Average removal efficiencies for the 10 cm depth ranged from 30.6 to 39.4% on a concentration basis, similar to removals obtained from 30-cm straw filter. The two higher densities at 10 cm depth yielded separation efficiencies not significantly different from the 30 cm straw. The 10 cm filter resulted in higher manure filtered/unit biomass which implied that the similar separation efficiency can be achieved using less filtration biomaterial. Lower removal efficiencies occurred at the beginning of the process due to large initial pore space within the filter, which resulted in less opportunity for the interception and adsorption of solids. As solids accumulated on the inner surface of the filter, a layer of solid cake was formed in the top portion of the filter. The cake layer became deeper and less porous with time until it eventually plugged the filter. This is a well known phenomenon in filtration work (Purchas, 1971, 1996; Dickenson, 1997; Rushton, 1996). The porosity of the filter decreased with the increasing volume of the applied manure. As a result, the removal efficiency generally increased exponentially with the volume of manure filtered according to the equation: Y = ae bx (1) Table 2. Solids removal for 10- and 30-cm-deep oat straw at different densities Manure Removal Removal Initial Filtered Efficiency* Efficiency Filter Manure to (conc. (mass Density TS Biomass basis) basis) Biomaterial Type (g/cm 3 ) (%) (L/kg) (%) (%) Oat straw (30-cm) b 45.8 (vertical orientation) bc b b 52.1 Oat straw (10-cm) a 40.8 (vertical orientation) b b 49.2 * Letters indicate significant differences at the 0.05 level with four replications. 100 (solids concentration in concentration out)/solids concentration 100 (solids mass in-solids mass out)/solids mass 982 TRANSACTIONS OF THE ASAE

3 se 1818 ms 7/9/01 10:47 AM Page 983 where Y represents the removal efficiency on a percent concentration basis, a and b are empirically derived constants, and x is the volume of applied manure. Table 3 shows the empirical constants a and b, and r 2 for the regressions of removal efficiency on volume filtered at different filter densities for oat straw and soybean stubble. Figure 1 shows results for the relationship for 10-cm-deep oat straw at three densities. Equation one also describes the relationship between incremental filtration time (the time it takes to filter an additional 3.78 L of manure, as a function of the cumulative volume filtered (up to L for the 10 cm oat straw and L for the 30 cm oat straw), as the filter progressively plugs. The coefficients a and b are shown in table 4. Table 3. Empirical constants for equation 1 describing solids removal efficiency (%) as a function of liters filtered Filter Density Biomaterial Type (g/cm 3 ) a b r 2 Oat straw (10-cm) (vertical orientation) Oat straw (30-cm) (vertical orientation) Soybean stubble (vertical orientation) Most solids were retained within the straw filter, rather than on its surface. This phenomenon and the related separation mechanism make this filtering method different from most of the existing animal manure separation techniques, such as settling basins and mechanical separators whose mechanism are based on differences of particle density and particle size, respectively. Since most of the solids were trapped within the straw filter (instead of on the surface), this filtration method could be categorized as a type of deep-bed filtration (Chen, 1997). Though categorized as deep-bed filtration, the cake formation starts from the top of the filter and grows into the deeper biomass. Cake formation increases the efficiency of trapping solids in the top portion. And that, in return, speeds up the cake formation near the top. As a result, the top portion of the filter plugs before the rest of the filter becomes effective. This explains why most of the solids were retained primarily in the upper part of the 30-cm depth straw filter, and why similar removal efficiencies could be achieved using a 10-cm depth filter. Unlike municipal or industrial wastes, animal manure solids concentrations can be highly variable. Swine pit TS concentrations can vary from 1 to 10%, depending on the feed and water system, and management. To determine whether deep bed filtration would work for manure with varying solids concentrations, five initial solids concentrations were tested (1.4, 2.6, 3.9, 4.8, and 6.0%) using 10 cm, vertical oat straw. The removal efficiency and plugging rate increased with increasing initial solids concentration of the applied manure. Results are shown in figures 2 and 3. SOYBEAN STUBBLE The average removal efficiency for three different soybean stubble densities were 29.9, 24.5, and 25.2%, respectively, for filter densities of 0.024, 0.027, and gm/cm 3 as shown in table 5. Similar to the oat straw filter, both filtration duration and removal efficiency increased as a function of the applied volume. The higher density filter plugged significantly earlier than the lower density filter. The filtration duration of each 3.78 L unit was generally longer for the filter with the higher density, Figure 1 Solids removal efficiency using 10 cm depth of vertically oriented oat straw. Table 4. Empirical constants for equation 1 describing filtration duration (min) per unit of manure as a function of liters filtered Filter Density Biomaterial Type (g/cm 3 ) a b r 2 Oat straw (10-cm) (vertical orientation) Oat straw (30-cm) (vertical orientation) Soybean stubble (vertical orientation) Figure 2 Solids removal efficiency of 10 cm, vertically oriented oat straw with varying initial manure solids concentrations. VOL. 43(4):

4 se 1818 ms 7/9/01 10:47 AM Page 984 Table 6. Solids removal for 30-cm-deep corn stover at different densities Manure Removal Removal Initial Filtered Efficiency* Efficiency Filter Manure to (conc. (mass Density TS Biomass basis) basis) Biomaterial Type (g/cm 3 ) (%) (L/kg) (%) (%) Corn Stover (30 cm) a 41.0 (vertical orientation) b b 50.1 * Letters indicate significant differences at the 0.05 level with four replications. 100 (solids concentration in concentration out)/solids concentration 100 (solids mass in-solids mass out)/solids mass Figure 3 Filtration duration of manure with varying initial solids concentrations through 10 cm, vertically oriented oat straw. Table 5. Solids removal for 30 cm deep soybean stubble at different densities Manure Removal Removal Initial Filtered Efficiency* Efficiency Filter Manure to (conc. (mass Density TS Biomass basis) basis) Biomaterial Type (g/cm 3 ) (%) (L/kg) (%) (%) Sbn stubble (30 cm) a 41.9 (vertical orientation) b b 36.8 * Letters indicate significant differences at the 0.05 level with four replications. 100 (solids concentration in concentration out)/solids concentration 100 (solids mass in-solids mass out)/solids mass which was another indication of rapid plugging for the high-density filter. The removal efficiency of soybean stubble is somewhat lower than that of oat straw. Soybean stubble has some leaves and pods which should enhance removal efficiency by increasing the opportunities for solids interception. More surface roughness should also be an advantage. The soybean stubble had much thicker stalks than oat straw. Consequently the total number of oat straws in the filter was greater than the number of soybean stalks for equal filter volumes densities. Due to the greater number of individual oat straws, the average size of the open flow channels inside the straw filter was reduced compared to soybean stubble filter. CORN STOVER The average removal efficiencies were 28.9, 37.2, and 40.9% for corn stover filter densities of 0.028, 0.037, and g/cm 3, respectively, as shown in table 6. Results were very similar to results obtained from oat straw. In general, both filtration duration and removal efficiency increased exponentially as a function of the applied volume. Cake formation was observed on the top part of the filter for all densities. The leaf portion of corn stover is large compared to soybean stubble. This may increase the potential for corn stover to intercept solids if the filtration material is tightly packed inside the cylinder. Once packed tightly, the leaves and stalks are squeezed together, and may block some open channels between the corn stalks. Vertical orientation helps to reduce the plugging problem. The removal efficiency of the corn stover filter appeared to be slightly better than that of the soybean stubble filter, but manure filtered/unit biomass was less. CORNCOBS AND GROUND CORNCOBS Corncobs were selected for testing primarily due to their surface roughness, which was suspected of potentially helping to retain solids. The results for solid separation with corncobs and ground corncobs are given in table 7. The average solid removal efficiency was only 2.6% for the 4 cm filter. The corresponding value for 8 cm filter was 6.2%. Corncobs placed randomly were basically ineffective for swine manure solids separation. The removal efficiency was improved by orienting the corncobs vertically. As shown in table 7, the efficiency was increased to 14.3% for the 4 cm filter, greater than the corresponding value obtained from random orientation with the same depth. However, the filter plugged much faster for the vertical orientation than the random orientation, as indicated by the ratio of manure filtered to biomass (11.7 vs 91.5). Because corncobs have a columnlike shape, the random orientation produced some very large pores inside the filter. When all the cobs were stood vertically so they squeezed each other in the separation cylinder, increased solid separation efficiency resulted from the reduced pore volume. Premature plugging was a problem with ground corncobs. Although, as indicated in table 7, the removal Table 7. Solids removal for different depths of corncobs and ground corncobs Removal Initial Manure Efficiency* Filter Manure Filtered (conc. Density TS to Biomass basis) Bio-material Type (g/cm 3 ) (%) (L/kg) (%) Corncobs (4 cm) (random orientation) Corncobs (8 cm) (random orientation) Corncobs (4-cm) (vertical orientation) Ground corncobs (1 cm) Ground corncobs (1.5 cm) Ground corncobs (3 cm) TRANSACTIONS OF THE ASAE

5 se 1818 ms 7/9/01 10:47 AM Page 985 efficiency was 14.3, 21.4, and 23.2% from the filters filled with ground corncobs of 1 cm, 1.5 cm, and 3 cm, respectively, plugging was observed after only 7.5 L (two doses) of liquid manure were applied. The 1 cm depth was approximately equal to the average size of the ground cobs, which meant only a single layer of this material was used as the filter in this case. Even with this very thin layer, plugging occurred almost immediately. MANURE FILTERED:BIOMASS RATIO The ratio of the amount of manure filtered to the initial weight of the biomass filter (called the manure filtered/unit biomass ratio) is an important consideration for the design of systems in the field. The greater the ratio, the less biomass is required to filter a given volume of manure. The manure filtered/unit biomass increased with decreasing filter density (fig. 4) due primarily to greater volumes of manure filtered rather than lower filter weight. The 10-cmdeep oat straw had the highest ratio of manure filtered to biomass (181.4 L/kg). With a ratio of 150, approximately 11 kg of biomass would be required per finishing swine space per year. Assuming a typical large round bale of straw weighs approximately 450 Kg, one bale would filter liquid manure from approximately 41 head of finishing swine per year; or a 1000 head finisher would require 50 bales each year. The biomass would likely be a good candidate for composting after use as a filter, with the added nitrogen and moisture from the filtering process. That evaluation was not part of this investigation. Figure 4 Filter density vs manure/biomass ratio for combined data for 30-cm-deep oatstraw, soybean stubble, and corn stover. SUMMARY Oat straw, corn stover, and soybean stubble were effective as filter media for solids removal from swine manure. Corncobs, both ground and whole, were ineffective. Solids removal was more efficient, and plugging occurred sooner from initially thicker manures. Removal efficiencies and incremental filtration times increased exponentially with each additional unit of manure filtered. Removal efficiencies generally increased with increasing initial filter densities and increasing solids concentrations of the manure Manure filtered/unit biomass ratio decreased with increasing filter density due to more rapid plugging. In the best case 180 L of 3.9% TS manure were filtered/kg of oat straw. CONCLUSIONS The use of porous materials as a filter media is a new application in the area of animal manure management even though municipal and industrial wastewater treatment systems have used similar techniques for many years. This preliminary study shows that some crop residues can be used to filter solids from liquid swine manure. More work is needed to test additional materials, and to develop acceptable methods for field applications of this filtration technique. REFERENCES Chen, W Solid-liquid separation via filtration. Chemical Eng. (Feb): Dickenson, T. C Filters and Filtration Handbook, 4th Ed. Oxford, U.K.: Elsevier Advanced Technology. Huijsmans, J., and J. A. Lindley Evaluation of a solid-liquid separator. Transactions of the ASAE 27(6): Moore, J. A., R. O. Hegg, D. C. Scholz, and E. Strauman Settling solids animal waste slurries. Transactions of the ASAE 18(4): American Public Health Association Standard Methods for the Examination of Water and Wastewater, 18th Ed. Washington, D.C. Purchas, D. B Industrial Filtration of Liquids, 2nd Ed. London, U.K.: L. Hill Handbook of Filter Media. Oxford, U.K.: Elsevier Advanced Technology. Rushton, A., A. S. Ward, and R. G. Holdich Solid-Liquid Filtration and Separation Technology. New York, N.Y.: Weinheim. Zhang, R. H., and P. W. Westerman Solid-liquid separation of annual manure for odor control and nutrient management. Applied Engineering in Agriculture 13(3): VOL. 43(4):

6 se 1818 ms 7/9/01 10:47 AM Page 986