7/28/2019 11560.pdf

1/10

.

THE EFFECT OF SYNTHETIC BULKING AGENTS ON PRESSURE DROP THROUGH FOODWASTE COMPOST

Robyn L. McGuckin,Dr.Mark A. Eiteman,Dr. .C.DasDepartment of Biological and Agricultural EngineeringUniversity of GeorgiaAthens, Georgia

SUMMARYPressure drop w as measured through a column of lettuce and onion food processing w astecontaining synthetic bulking agents from chemical separations processes. Six different polypropylenebulking agents were compared against the more traditional bulking agent pine bark. The b ulking agentswere used at v arious different fractions, or bulking agent densities. The results consistently sho wed that thesynthetic bulking agents caused a pressure drop three to five times low er than pine bark. Pressure drop

models suggested by Shedd, Hukill and Ergun w ere used to fi t these data. A parameter in &guns equationwas used to quantify the fraction of bulking agent necessary to implemen t 50% viscous energy pressuredrop losses.INTRODUCTION .

Vegetab le wastes have several characteristics which have m ade comp osting problem atic, theseinclude high m oisture content, lo w carbon to nitrogen ratio, and often high concentration s of odor causingchemicals such as sulfur. Bulking agents may mitigate problems associated with vegetable wastecomp osting. An ideal bulking agent would contribute pore space througho ut the compost pile and process(Haug. 1993).Low -cost synthetic bulking agents might indeed be designed to provid e optimal sup port andvoid space integrity. Similar criteria have been used for the design of numerous packing materials for gasabsorption used in th e chemical manufacturing industries. Indeed, low pressure drop is a principal concernin the selection of these inert packing m aterials (Coke r. 19911.The high m oisture content of vegetable waste results in considerable deformation and compactionduring the degradation process, causing significant reduction of pore space through w hich air may flow(Haug, 1993). As pore space is reduced, pressure drop across the compos t pile will increase, necessitatinghigher energy input to the blowers to m aintain the air flow (K eener, 1993). These problems may bepartially circumven ted with bulking m aterials that provide a relatively constant pore matrix in the compost.The low C:N ratio of lettuce and onion w astes may caus e incomplete microbial digestion.Therefo re. these wastes should be augmented with a substrate containing a high C:N ratio. In add ition, thehigh sulfur content of the veg etables (particularly onions) increases the likelihood of odo r form ation,particularly if an aerobic environment is not maintained (Haug, 1993). Sulfur caused odors may be reducedby augmenting vegetable waste with l o w sulfur substrates to provide a chemical balance, or by insuringadequate aeration during the entire compost process. Adequate aeration is also facilitated with bulkingagents and high pore space.CORRELATIONS FOR PRESSURE DROP

Num erous correlations are available for predicting pressure drop en countered by fluids flowing-- through a matrix of po rous material such as comp ost. I n general, pressure drop per depth of material ( hp )IS a function of superficial fluid velocity (U). Shedd (1951, 1953) proposed the follow ing equation foradlow through seed and grain:

.

102

7/28/2019 11560.pdf

2/10

AP= AUBThe s uperficial air velocity (units of cfm /ft?,or ft/min) is equal to the volumetric flow rate (Q )divided by the cross-sectional area of the porous material, tangential to the direction of flow. The two terms

A and B in Equation 1 are empirically determined. Hu kill(1953 ) modified Equation 1 to acco unt fornonlinearities in the results of Shedd:CU

h l ( l + D U )P=In Equation 2, C nd D re empirically determ ined. Although the pressure drop is normalized bymaterial height, Equations 1 and 2 do not account for the effects of com press ible materials such asvegetable wastes. The simples t correlation (Chen , 1980;Higgins et al., 1982; Haug, 1993) for quantifyingair flow through a col umn of compressible m aterials is:

AP =KU"h' (3)In Equation 3, the value of K and the exponents n and j are empirically determ ined. The values of theseexpo nents approach one for laminar flow. The value of j becomes one for non -compressible materials. and. iggins (1982) suggested that j will be one for a depth less than 2.4 meters. Sinc e flow through porousmedia is turbulent (Sc hede gger, 1957) values of the expo nents are generally not equal to one. Morerecently Das and Keener (19 97) demonstrates the effect of moisture cont ent and comp action on pressuredrop. Reynolds (1900) formulated an alternate approach to describe pressure drop as a function ofsuperficial velocity:AP =aU + b U 2 (4 )Ergun (1952) is often credited with the application of this equation, as he provided an exten sivedemonstration of its utility in describing the behavior of oven coke. As in previous equations. Equation 4has empirical constants which must be determined for the specific material. The first term in Equation (1represents the viscous energy losses, and is essentially a simplification of Darcy's Law, which regardsinertial effects as negligible. Viscous energy losses are caused by the contact of a fluid with the surfaceover which it is traveling. The second term represents the kinetic energy losses, caused by rapid lateralmovement of the fluid as it f lows through pore spaces. In completely turbulent flow at high air velocities.kinetic energy losses constitute the principal resistance to flow. Thu s the coefficients in Firgun's e quatio nindicate the relative imp ortan ce of viscous and kinetic energy losses.An important parameter in composting with bulking agen ts is the ratio of bulking age nt tocompostable material. or the bulking agent density (4). One would an ticipate that at a given air flowrateincreasinp the bulking agen t density would in general decrease the pressure drop. Increasing the value of I$would change the empiric al constants in any of the above correlations (Equations 1 - 4). An optimalbulking agent w ould be one which provides low pressure drop (and associated costs) w hile max imiz ing thequantity and quality of the composted m aterial.The goal of this study was to compare quantitatively several bulking agen ts for the compo sting oflettuce and onion waste. The two m easures of bulking agent effectiveness are the press ure drop in a columnof com post, and the compostability of the mixtures.COMPOSTING VEGETABLE WASTES

-- Th e U.S.1997 fresh-cut vegetable production was estimated at 83 0 million pounds (Ray nes,1997). At one processor near Atlanta GA. 83 % of total production consists of lettuce and 8% of onions(Raynes. 1997). The industry is very sensitive to vegetable quality. and therefore w aste ge neratio n almostequals wholesale production. The U.S. annual projected fresh-cut vegetable waste b y-prod ucts are 790million pounds (Ra ynes . 1997), essentially all of which w ill go directly to landfills. With increasing public._

103

7/28/2019 11560.pdf

3/10

pressure to reduce solid waste disposal reduction or recycle of these v egeta ble wastes is of growin gimportance. Com postin g may provide one alternative to landfilling there highly b iodeg radab le materials.In addition to causin g a significant volume reduction, com postin g can yield a useful soil amen dm ent.However, unlike traditional compost materials such as animal and municipal waste, vegetable waste hasseveral characteristics that may present difficulties to the process (H aug, 1993). Specifically. lettuce andonions are high in moisture , have a low C:N ratio, and can generate odor while com postin g due to their highsulfur content.

2"1"2"1 *)2"

0.5-3"

MATERIALS

93 4.2090 4.3094 3.5091 5 .0094 3.2578 10.00

The organic compostab le mixture used in this study consisted of lettuce, onio ns, poultry litter, andpine sawdust. Lettuce and onions were both obtained as fresh w aste from a local vegetable processingplant, the poultry litter from a local farm, and the sawdust from a sa wm ill (Curl Pack, Suwane e, G A ) . Forsize uniformity, the vegetables were manually chopped and screened (2'' x 2"). The volume of material wasmeasured to the nearest 0.003 m3 and routinely stored in 45 gallon covered containers in a cold room(4'C).manufacturing industries. Three polypropylene packing materials (show n in Figu re 1)were studied:Telpack (Telpack Tower Packing, Boston,MA), ripac, and Sad dles (Jaeger Products, Inc., El Dorado,KS). For each packing material two nominal sizes ( I " and 2" ) were studied. Pine bark (Crav en Pottery.. Commerce, G A ) was also used studied. Table 1 shows pertinent param eters on these seve n bulkingmaterials.

Non-organic bulking materials consisted of random packing materials commonly used in chemical

Table 1 - Physical properties of the bulking materials

TripacTripacTelpackTe packSaddleSaddlePine Bark

.Nominal Size I Void space I BulkDensity 1I (%) I (Iblft-')I " 90 6.20

The experimental setup to study the pressure drop consisted of three m ajor comp onen ts: a column.blowers and associated piping, and two manometers (Figure 2). The column (155 cm by 61 cm diameter)was made of reinforced fiberglass (U.S. lastics. Lima. OH). The co ncave co lumn ,floor averaged 23 cmabove the base, yielding an effective depth of 132 cm. Over 409b of this floor was open and co vered withan expanded metal and a fine mesh screen to permit airflow and to suppor t the comp ost material. The topo f the column w as open to the atmosphere. Tw o blowers (Dayton Electric Mfg. Co., Niles, IL ) attached inparallel via 5 cm diameter PVC piping provided air f l o w to the column. A single ball valve controlled thecombined air flow. Tw o incline manometers (Dw yer Instruments Inc.. Michigan City, IN ) were used in thisstudy. A manometer of 0.25 kPa maximum water pressure, calibrated with air flow c overin g theexperimental range, w as used to estimate th e air flow rate. A 0.12 kPa water pressure manometer allowedmeasurement of the column pressure. Both manometers had a 0.0012 kPa precision.METHODS

-- The compost mixture consisted of organic materials in a mass ratio of 19:23: 10:1(lettuce: onions:poultry litter: saw dust) wet basis. In order to achieve this mixture, 0.091 m3 of lettuce, 0.091 m3 of onions.0.04 m3of poultry litter and 0.005 m 3 of sawdust were mixed. A selected volume of bulking material waythen mixed into the organic material. beginning with the lowest density to be tested. The resulting mixturewas carefully placed in the column to avoid packing. Starting with the highest flow rate, the air f l o w valve

104

7/28/2019 11560.pdf

4/10

was set to each of five nominal flow rates (0.0075,0.0107,0.183,0.0274,nd 0.0366 m3/s/m2)andcorresponding pressure drop s across the material were recorded.gently and replaced in to the column fo r a subsequen t set of measurement. After three replicatemeasurements were made for a given bulking age nt density, additional bulking age nt was added to providethe next density.

After a se t of pressurdair flow measurements, the material was taken out of the colum n, mixed

Telpack Tripack SaddleFigure 1: Synthetic Bulking A gent Shapes

132

blowers

\manometers 5.1cm diameter pipe--Figure 2: Experimental Setup for Pressure Drop Tests

105

7/28/2019 11560.pdf

5/10

barameter Lettuce Onions Poultry Litter SawdustMoisture (wet basis) 96.17 91.10 22.70 I 1.33

RESULTS AN D DISCUSSION

42.504.130.201099.667

Table 2 show s the chemica l and physical parameters for the four organic substrates . The lettuceand onions were above 90%moisture content and below a C:N ratio of 15.

43.30 31.90 40.403.76 3.71 0.050.69 0.68 0.3612 9 792

42 1.667 383.380 118.OOO

Table 2: Chemical and physical parameters for substrates

Particle density (kg/m3)Pore space

679.333 963.667 1023.550 66O.OOO0.853 0.521 0.623 0.824

Carbon (dry basis)Nitrogen (dry basis)Sulfur (drybasis)C:N ratioBulk density (kg/m3)

Figures 3,4 and 5 show the pressure drop in the compost column as a function of the superficialair velocity for three selected bulking agents. As expected , increasing the air flow increased the measuredpressure drop. Mo reover, for each of the bulking agents an increase in the bulking agent density resulted ina significantly reduced pressure drop. Significant differences are also observed between different bulk ingmaterials. For example, the m easured pressure drop for a compost with 0.00062 m3 kg of 1"Telpack wasapproximately 0.018 P i d m a t an air, flow of 0.0274m l s . In contrast, the measured pressure drop at anidentical flow and bulking age nt density was 0.025Walm for 2"Saddles and 0.040 P i d m for pine bark.The comp ost without bulking agent yielded a pressure drop of about 0.070 inches kFWm at this flow rate.These differences were va lidated in a statistical analysis which showed a sig nificant difference (at p e0.025) between the trial with no bulking age nt and the trials with other bulking agents, and between thetrials with pine bark and those with synthetic bulking agents. How ever, there was no significant differencewhen compa ring the 1" bulking agen ts to the 2"bulking agents. or when comparing any of the bulkingagents of the same size.Figure 6 shows another way to present the pressure drop measurements for I " Telpack. For eachair flow. increasing the bulking agent density reduced the pressure drop. The difference s in pressure dropbetween the five bulking agent densities was statistically significant (at p

7/28/2019 11560.pdf

6/10

W I$ =0.00025 m3/kgA I$=0.00044m3/kg

0.00 0.01 0.02 0.03 0.04 0.05U (m/s)

Figure 3: Measu red pressure drop ac,ross colum n per unit material height (AP) as a function of superficialair velocity (U) for 1" Telpack. Solid curves represent regressions of data to Equation 4 n lext.

0.04 e=0.00062m3/kgC U =0.00094 m3/kg

0-03 - A I$=0.00125 m3/kg emEy, 0.02\Un

0.01

0.000.00 0.01 0.02 0.03 0.04 0.05

U (m/s>Fipure 1:Measured pressure drop across column per unit material helght C L I P )s a (unction of superl'lclalair velocily (U)or 1" Saddles. Solid curves represent regression3 of d a~ ao Equation 4 in I c x t .

107

7/28/2019 11560.pdf

7/10

0.12 1 0 ~=0.000003/kg A0 6=0.00062 m3/kg

Q = 0.00094m3/kg0.10 4r\ 0.08EUa5 0.06\

aQ0.04

0.02

0.000.00 0.01 0.02 0.03 0.04 0.05

U (m/s>Figure 5 : Measured pressure drop across column per unit material height (AP)as a function of superficial

air velocity (U)or pine bark and without bulking agent. Solid curves represent regressions o f datato Equation 3.,

0.12 1 1Im A U=O . O 075m / s0.10 1 C U =0.0107 m/s+ U=O.O183m/s

A U=O.O274m/s U=O.O366m/sa

Q m

0.08

0.04

0.02

0.00 0 0 0 0 0 0(u d (D CD 00 0 0 0 F0 0 0 0 0 0 099 9 90 0 0 09

cuF800 8 0_

Figure 6:Mrlrsured pressure drop across colum n per unit material height ( A P) as a function o f hulking agentdensity (@ ) for I Telpack.

108

7/28/2019 11560.pdf

8/10

0.5 -0.0 f I I I I I I

.... \"\.... *'\........ \,\'.. ...... *....o - - - - ~ ~ +..........A---=-4

000009

0CD008

Figure 7: Linear coefficient a versus bulking agent density (a).

The relationship between $ and u was modeled with the following empirical relationship:

in Equation 5 . C I and Cz are fitted parameters and a. is the linear coefficient in Ergun's equation for thecase of no bulking agent. Thus, the dimensionless ratio la0 - aYuo is a measure of the fractionalimprovement in pressure drop: for a bulking agent which provides no improvem ent in the linear coefficientover the case of no bulking agent this ratio will be equal to zero, while for a bulking agent which essentiallyremoves pressure drop this ratio will approach one. If the bulking agent density approaches zero ( bC 0)then the term CIb/(C?+$) approaches a value of C I for any type of bulking agent. For this case thefractional improvement in pressure drop should approach I ,so that the value of C , determined by f i t t ing thedata in Table 3 should be close to I . When (I=C?, he fractional improvement in pressure drop given b yEquation 5 becomes equal to CI / l. If C 1 = . then the fractional improvement in pressure drop i s 0.50 when=C2. Thus. this second empirical constant, C?. s a measure of the bulking agent density required todecrease the linear form of Ergun's equation by 50%. The greater the value ofC, alculated. the higher thebulking agent density necessary to affect a change in pressure drop. Table 3 shows the resulting best-fitvalues for the em pirical constants ('2,and C2) n Equation 5 . The resulting equations are also shown inFigure 7 for these selected bulking agents. As Table 3 shows. the value for the constant C , for each bulkinsagent was indeed approxim ately equal IO . Of the bulking agents tested. the pine bark had the highestvalue of C z.while 1"Telpack (lowest r' and three data points) and I " Saddles (five data p oints) had thelowest value of C2 . These results dem onstrate and quantify the reduction in pressure drop afforded by thesynthetic bulking agents over pine b a rk.

109

7/28/2019 11560.pdf

9/10

Bulhng Agent c1 c- 2Telpack 1" 1.129 0.000340Tripack 1" 1.191 0.000394Saddle 1" 1.097 0.000242Telpack 2" 0.922 O.ooOo74Tripac 2" 1.155 0.000447Saddle 2" 1.223 0.00042Pine Bark 1.250 0.001322

r20.9810.9650.9920.7930.9371 OOo1.Ooo

For the pine bark and each of the three 2"synthetic bulking agents, the value of this second-ordercoefficient decreased with increasing bulking agent density. In other words, for these four bulking agents astheir density increas ed, the pressure drop became more linear with superficial air velocity. As expected. theaddition of any of these bulking agents serves to reduce the kinetic energy losses.CONCLUSIONS

The pres sure drop data across a column of compost show s a high correlation to Ergun's equation.For all bulking agents an increase in the bulking agent density (4) resulted in an almost linear decrease inpressure drop. The sy nthetic bulking agents resulted in a significantly lower pressure drop than the pinebark. The 2"synthetic bulking a gents resulted in lower pressure drop han the 1" agents. No one particular -synthetic bulking agent shape stoo d out as better than the o thers across he 1 an d 2"sizes.The synthetic bulking agents provide an advantage over the pine bark by reduc ing the headdifferential across a compost reactor. The reduce d head differential results in decreased fan powerrequireme nts, and thus decreased costs. Synthetic bulking agents cannot provide the nutrient and ab sorbemadvantage s of biode gradable agents; however, the synthetic bulking agents can be recovered aftercomposting w hereas th e pine bark will deteriorate over time. Additional testing should be conducted todetermine the suitability of Ergun's model fo r other compost materials and bulking agen ts. and to determineif parame ter C? varies with compost material.ACKNOWLEDGEMENTS

The authors thank Sarah Lee and Sheila Powell for their technical support. In addition thefinancial support of the Georgia Food Processing Advisory Council (FoodPAC)and the GeorgiaExperiment Stations i s gratefully appreciated .REFERENCESCalderwood.D.L. 973. Resistance to Airflow of Rough. Brown, Milled Rice. Transactions of the ASAE16(3):525-527.Chen. S. 1980. Resistance of Sewage Sludge Compost IO Ai r Flow. Mas ters Thesis. Rutgers University,Department of Agricultural Engineering.Coker, A.K. 1 9 9 1 . Understanding the Basics of Packed-Column Design. Chemical Engineenng Ropess.November 1991 . pp.93-99._ _ Cooper. S.C. and H.R. Sum ner. 1985. Airflow Resistance of Selected Biom ass Materials. Transactions ofthe ASAE. 28(4):1309-1313.

110

7/28/2019 11560.pdf

10/10

Das, K. and H.M . Keener. 1997. Moisture Effect on Compaction and Permeability in Composts. Journalof Environmental Engineering, 123(3):275-281.

Ergun, S. 1952. Fluid Flow Through Packed Columns. Chemical Enginee ring Progress, 48(2):89-94.Jeris, J.S. and R.W. Regan. 1973. Controlling Environmental Parameters for Optimum Composting. Pan

IT: Moisture, Free Air Space and Recy cle. Compost Science. MarcMApril 1973. pp. 8-15.Haug, R.T. 1993. The Practical Handbook of Compost Engineering. Lew is Publishers, Boca Raton. Fla..USA. 717 pp.Higgins, A.J., Chen, S., nd M.E. Singley. 1982. Airflow Resistance in S ewage Sludge Com postingAeration Systems. Transactions of the ASAE. 25(4):1010-1014,1018.Hukill, W.V. 1955. Radial Airflow Resis tance of Grain. Agricultural Eng ineerin g. 36(5):332-335.Keener, H.M.,ansen , R.C., and D.L. Elwell. 1993. Pressure Drop Through Com post: Implications forDesign. ASAE Paper No. 934032.Raynes, Mark. 1997. Personal Comm unication. Fresh Express vegetable processors.Reynolds, 0. 1900. Papers on Mechanical and Physical Subjects. Cam bridge Uni versity Press.Scheidegger, A.E. 1957. The Phy sics of Flow Through Porous Media. The Macm illan Co mpa ny, NewYork.Shedd. C.K. 1951. Some New data on Resistance of Grains to Airflow. Agricultural Engineering.17(6):1144-1149.Shedd, C.K. 1953. Resistance of Grains and Seeds to Air Flow. Agncultural Engineering. 34(9):616-619.Sturos, J.B. 1989. Airflow Resista nce of Chunk woo d. F orest Products Journal. 39(3):19-24.

Suggs. C.W. and A. L anier. 1985. Resistance ofWood Chips and Sawdu st to Airflow. Transactions of theASAE. 28( 1):293-301.Vegetables and Specialties Situation and Outlook. VGS-270. 1996. Economic Research Service, U.S.Department of Agriculture, Washington. DC.

List of Symbols Useda. b. A. B. C. D. j. k, n =empirically determ ined parameters i n pressure drop correlations (Eqs. -4)AP =pressure drop per foot depth of grain (inches H2 0/ ft )I$=bulking agent density (ft-'/Ib)U =superficial air velocity (ft/min)

1 1 1