Biological Wastes 31 (1990) 199-210

Enhancement of Dry Anaerobic Batch Digestion of the Organic Fraction of Municipal Solid Waste by an Aerobic

Pretreatment Step

E. ten Brummeler & I. W. Koster

Department of Water Pollution Control, Wageningen Agricultural University, PO Box 8129, 6700 EV Wageningen, The Netherlands

(Received 4 June 1989; accepted 14 June 1989)

A BSTRA CT

The start-up of the dr), anaerobic batch digestion by the BIOCEL-eoncept of the organic fraction of municipal solid waste (MSW) is unbalanced when a methanogenic inoculum (digested sewage sludge) is added to a total solids' concentration of 35%. The unbalanced conditions are the result of the rapid degradation of easily-degradable compounds" which are present in the organic'

fraction. Enhancement of the first start-up of the dr)' batch digestion was tried by applying an aerobic partial-composting step. By this aerobic' treatment the easily degradable compounds are removed. After the composting step the anaerobic digestion will be limited by the conversion of the ligno-cellulose part o[" the organic fraction. It appeared that at least 19"5% of the volatile solids (VS) should be converted during the aerobic" composting period before acid .[ormation in the digestion was in balance with the methane Jormation. This amount of'aerobically degraded VS means a 40% loss of potential biogas. The loss of a part of the biogas is a major drawback to the partial composting as a method for enhancing the start-up of the dry anaerobic digestion. A shorter composting period which is combined with another start-up method might be a fi, asible method to decrease the energy input of the dr)' digestion process.

I N T R O D U C T I O N

Dry anaerobic digestion of solid organic waste, especially the organic fraction of municipal solid waste (MSW) is of growing importance in the

199 Biological Wastes 0269-7483/90/$03"50 (C 1990 Elsevier Science Publishers Ltd, England. Printed in Great Britain

200 E. ten Brumrneler, 1. W. Koster

field of solid-waste management (De Baere & Verstraete, 1984; Ten Brummeler et al., 1986; Spendlin & Stegmann, 1988; Ten Brummeler et al., 1988). Several reactor types are being developed, based on batch digestion and continuous digestion. The continuous digestion of the organic fraction of M SW needs a high grade of technology (Ten Brummeler et al., 1986). The high investment and maintenance costs will be a major drawback when the implementation of these systems is considered.

Two continuous-reactor types are being developed so far, that is the VALORGA process and the DRANCO process. (Membrez & Nicolet, 1985; De Boosere et al., 1986). Both systems have been tested on pilot-plant scale and semi-technical scale (5-50 m3). However, no experimental data are given which refer to the start-up method of these continuous dry-digesters. Both processes report digestion times of 2-3 weeks at mesophilic temperatures.

Dry batch digestion of the organic fraction of M SW is in essence a simpler method than continuous digestion. The lower cost per ton of waste treated in a batch reactor is a consequence of the simple technology applied. BIOCEL is a recently introduced concept for dry, anaerobic batch-digestion of solid organic wastes (Ten Brummeler et al., 1986; Koster et al., 1988). Several methods for start-up of BIOCEL-reactors have been described previously (Ten Brummeler et al., 1988; Ten Brummeler & Koster, in press). The first phase of the dry anaerobic digestion in a BIOCEL-reactor is characterized by a period of imbalance of the acidogenesis and methanogenesis (Ten Brummeler & Koster, 1989), which results in inhibition of methane formation due to low pH values and high organic-acid concentrations. The main reason for the imbalance seems to be the relatively high conversion rates of certain compounds to hydrogen and organic acids: in particular the degradation of soluble sugars to lactic acid and acetic acid by lactic-acid bacteria plays an important role in the acidification of the process. In the 'fresh' organic fraction of MSW a considerable amount of lactic acid is found, which indicates that the lactic-acid bacteria are already present in the substrate. This is very similar to the situation in silage fermentation (McDonald, 1982).

When the volatile solids degradation is described as a first order reaction rate, the following equations can be given for the balanced, dry anaerobic- digestion process:

d[VS] dt - r v s = - k . [ V S ] (1)

rcn . = -- rvs (2)

where [VS] is the concentration of the remaining volatile solids, rvs is the

Start-up 01 dr), digestion 201

removal rate of the VS, k is the first order reaction-rate constant, VCH 4 is the formation rate of methane from the VS.

When the process shows imbalance the degradation of the VS to hydrogen and acids exceeds the formation of methane. The VS of the organic fraction of MSW is composed of several organic compounds (Ten Brummeler & Koster, in press) which are biodegradable under anaerobic conditions. The first start-up ofa BIOCEL-reactor must be carried out with a limited amount of inoculum (methanogenic sludge), which has a total solids concentration of 4-10% TS (Ten Brummeler et al., 1988). The methanogenic biomass concentration is rather low in these inocula. At 35% total solids, the regular TS-concentration for start-up of a BIOCEL-reactor, the initial acid- and hydrogen-formation rate exceeds the consumption rate of these compounds by the methanogens of the inoculum (Ten Brummeler et al., 1988). If an optimal amount of the methanogenic inoculum could be added the TS- concentration in a BIOCEL reactor would decrease to values much lower than 30%.

When just a limited amount of methanogenic inoculum can be added the concentrations of the easly-degradable compounds have to be reduced to a value which will result in a more balanced digestion with this limited amount of inoculum. No data are available concerning the individual removal rates during dry anaerobic digestion of the biodegradable compounds which are present in the organic fraction of MSW. In Table 1 the relative removal rates are given which are calculated from experiments on sludge digestion. According to Gujer and Zehnder (1983) and Eastman and Ferguson (1981) the reaction rate of sludge digestion, which is in fact the hydrolysis rate of the solid sludge-particles, follows first-order kinetics (Gujer & Zehnder, 1983). For the organic fraction of MSW, which is also mainly composed of solid

T A B L E ! Relative Removal Rates of Several Biodegradable Com- pounds of the Volatile Solids of the Organic Fraction of

MSW

Compound V S ( % )

Soluble sugars 7 Starch l Cellulose 55 Proteins l(~l 5 Lipids 11

r,.~/rc~t4 x 100%"

612 344

11 568

57

"Relative removal rates calculated from Gujer & Zehnder (1983) and Noike et al. (1985), assuming Methanosarcina sp.

the predominant methanogen.

202 E. ten Brummeler, L W. Koster

particles, first-order kinetics are likely to be valid. Assuming that the biomass of the acid-forming bacteria is non-limiting, the relative removal rates of several compounds during dry anaerobic digestion can be calculated from the first-order constants which have been calculated from sludge digestion. As can be seen from Table 1, where the composition of the volatile solids is given, the soluble sugars, starch, and proteins show higher relative removal rates than the methane formation rate. This is a result of the difference in growth rate of the methanogens and the acidogenic bacteria (Gujer & Zehnder, 1983).

It is likely that the compounds with a higher removal rate under anaerobic conditions will also show higher removal rates than the less-easily- degradable compounds, like cellulose, under aerobic conditions. During aerobic composting the production of heat increases the temperature to 49-70°C (Strom, 1985). Apart from a degradation of the easily degradable compounds, the advantages of an aerobic pretreatment for the dry anaerobic digestion might be this increase in the waste temperature. Using this high temperature waste, the external energy input for heating the reactor during the dry digestion could be lower. A major drawback could be the loss of a potential amount of methane during the composting.

In this paper the usefulness of partial composting of the organic fraction of MSW as a pretreatment method for stimulation of the start-up of the dry anaerobic digestion is investigated. The minimum degree of VS reduction during the partial composting which is needed for a rapid start-up will make clear whether partial composting is feasible as a pretreatment method.

METHODS

Chemical composition of the organic fraction

The substrate was an organic fraction obtained from the Recycling Zoetermeer separation plant for MSW. The organic fraction was passed through a sieve (mesh 20 mm) in the separation plant. The total solids (TS) were 48"7% and the volatile solids (VS) 30"3%; both on total wet-weight base. The composition of the volatile solids is given in Table 1. The potential methane yield, as determined in an experiment at 4% total solids, was 80 liter (STP)/kg organic fraction.

Partial composting

The composting process was carried out in a 1.28 m high plexiglass column 0.44 m in diameter. The column was insulated with a polyamide jacket. The

Start-up of dry digestion 203

column was filled with 60 kg of organic fraction, without compression of the contents. The aeration of the contents was by a pump set at a flow of 750 liters/h (20°C).

During the experimental period the temperature of the composting mass was in the range 40-68°C. At intervals, samples of c. 5 kg were taken for the subsequent dry digestion experiments. The total content of the column was mixed before the sampling was carried out. The composting material was maintained at 50-55% TS by additions of tap water to the column, when a sample was taken.

Apparatus

The dry digestion experiments were carried out in different types of reactors which are described elsewhere (Ten Brummeler & Koster, 1989). The main reactors in which biogas production and biogas composition were monitored had a volume of 6 liters. The pH and organic acids were measured in bottles of 0-5 liter each. Every sample for pH and acids determination corresponds to one reactor of 0"5 liter. During each experiment one 3 liter reactor and 5 small reactors were used. These reactors were filled under the same conditions as the main 6 liter reactors. The main batch reactors were connected to a water lock and a 10 liter gas-tight bag. Between the reactor and the water lock a biogas sampling device with a septum was placed.

Seed sludge

The seed sludge (methanogenic inoculum) used was obtained from the sewage sludge digester in Veenendaal, the Netherlands. The sludge had 4.2% TS. The maximum methanogenic activity, determined according to De Zeeuw (1984), was 0"03 kg CH4-COD/kg-sludge/day at 30°C.

Procedure

The batch reactors were filled with a mixture of organic fraction, methanogenic seed, and if necessary, tap water. The initial seed/substrate TS ratio was 0"04, which is the maximum amount of seed sludge that can be added when an initial TS of 35% is used. The initial compaction was 250 g TS/liter obtained by manual compression. In all experiments the initial TS concentration was 35%. The reactors were flushed for 2 min with nitrogen gas before they were closed. The experiments were carried out in a temperature-controlled room at 30°C.

204 E. ten Brummeler, L W. Koster

Analyses

Biogas composition (CO2, CH 4, Hz, N2, 02) was determined with a gas chromatograph (Packard 407), equipped with a TCD detector and two parallel columns: a column of 1"5 m × 1/8", teflon packed chromosorb 108, 60-80 mesh and a 1"2 m × 1/8" mol sieve 5A, 60-80 mesh. The column-split ratio was 1:1. Samples of 100pl were taken with a glass syringe. The measurement of the biogas volume was carried out at intervals by pumping the biogas out of the gas bags through a wet gas meter. In all graphs the gas volume is given for standard temperature and pressure (STP = 0~'C, 1 atm).

The pH of the samples was determined with a Knik mV-meter and a combined glass electrode (Schott) directly in the 0.5 reactors.

Prior to organic acids analysis the contents of a 0-5 liter reactor were extracted with 1500ml tap water. After shaking for 30 min on a shaking table the liquid was filtered. The filtrate was used for the volatile fatty acids (VFA) determination as described earlier (Koster, 1986). Lactic acid was determined on a HPLC, equipped with a UV detector (Kratos Spectraflow 773) and an organic acids column (Chrompack, 300 × 6-5 mm). The injection volume was 10 pl, the eluent 0.01 N H z S O 4 with a flow of 0.8 ml/min. The wavelength during the detection was set at 210nm, the absorption range 0.050. The concentration of organic acids is given as the concentration in the amount of water present in the original sample and expressed in COD- equivalents. The conversion factors from grams acids to grams COD are 1.067, 1.514, 1.818 and 1"067 for acetic, propionic, butyric and lactic acid respectively.

Total solids and volatile solids

Total solids were determined by drying the sample at 105°C overnight. Volatile solids were determined from the weight of the dried solids after ashing at 600°C for 20 min. TS and VS are given as percentage of the original wet sample (APHA, 1975).

RESULTS AND DISCUSSION

In Fig. 1 the VS-reduction at several composting periods is shown. When c. 20% of VS had been degraded during the partial composting it appeared that, after sampling, the temperature did not regain the thermophilic values observed from 0-20% VS reduction. This indicated that the easily degradable compounds had been removed. Therefore the composting was not continued much further. The maximum temperature was 40-42°C during the period from 20-23"5% VS reduction.

Start-up of dry digestion 205

Fig. 1.

% VS reduction

28-

2/,-

20-

16

12

B

J 0

o " L 0 ' ' 120 ' 160 2oo 2L0 composting period (h)

Volatile solids reduction during the partial composting of the organic fraction of municipal solid waste.

In Fig. 2 the observed biogas composition and biogas production of the non-treated organic fraction are given. The hydrogen content reached 35 vol.%, while methane could not be detected. In Fig. 3 the organic acids and the pH are given for the non-treated organic fraction. From Figs 2 and 3 it can be concluded that without pretreatment or addition of neutralizing chemicals this mixture of seed sludge and organic fraction of MSW will not lead to a steady methane production. The high amount of organic acids, in combination with the low pH, were detrimental to the methanogens.

The effect of the several partial-composting periods is illustrated in Fig. 4,

vot % 100.

80

60

L'O/ ÷

0 ~ l ~ , , ~ , - , ~ - , ~ ~ , , 0 20 40 60 80 100 120 11,0

sum biogas (t STP)

-80

-60

-¢0

-20

0

time (days) Fig. 2. Biogas production and biogas composition during dry anaerobic digestion of the fraction MSW without partial composting: (D) sum biogas; (~) % CH4: (©) % CO2:

(+) % H 2.

206 E. ten Brummeler, L W. Koster

organic acids ( g COD/I )

z4-

20

0 2O

0

=÷

6o time (days)

~H(-) ?

6

5

L,

Fig. 3. Organic acids concentration and pH during dry anaerobic digestion of the organic fraction of MSW without pretreatment by means of partial composting: ( + ) acetic acid; (©)

butyric acid; (I-1) lactic acid; (/k) total organic acids.

where the relative methane yield curves for the experiments with different composting periods are shown. The relative methane yield is the sum of the methane produced at any time divided by the potential methane yield of the organic fraction, as determined in a batch assay at low solids. The loss of a potential amount of methane during the composting was calculated by subtracting the methane yield of the composted VS from the methane yield of the untreated organic fraction. Significant methane production rates were observed (Fig. 4) when at least 20% of the VS had been degraded during the

(CH~f/EH4mox) = 100 (%) 100

80- . f x

60 " x~* ~*x x~***x-*x''x~='~

.x j

" 1 I+0- /,.*

20 . ~

0- 0 20 40 60 80 100 120 140' 160'180

time (days) Fig. 4. Relative methane yield curves of the organic fraction of MSW with several VS reduction grades in the composting pretreatment: (V) 4 % VS-reduction; (D) 11.9 VS- reduction; (O) 14.5%; VS reduction; (/X) |9-5% VS-reduction; ( x ) 23-5% VS-reduction.

Start-up of dry digestion 207

or( anic acids (g 00D/[ ) 50- .

d

30 I

20-

~ S . . . ........ Q

time (days) Fig. 5. Effect of the VS reduction by partial composting on the initial formation rate of acids during subsequent dry anaerobic batch digestion of the volatile solids: (a) 23.5% VS-reduction; (b) 19.5% VS-reduction; (c) 14% VS-reduction: (d} 11-9% VS-reduction;

(e) 4% VS-reduction.

composting period. The highest methane production rate was found at a VS- reduction of 23-5%. In Fig. 5 the relation of the initial VS degradation-rate (rv~) and the composting period is given. The initial r w was determined by calculating the sum of the VS which was degraded to acids, hydrogen and methane during day 0-day 12. When the sum of the degraded VS is plotted against time the value of the initial rvs is given by the maximum slope of the curve. (Fig. 5, Fig. 6). The initial VS degradation (~ initial acid formation rate) was lowered by the partial composting. In particular, the formation

Fig. 6.

rvs(g C0O/l.doy) 6

0 0 /+ 8 12 16 ~ 24.

% VS reduction Effect of the VS reduction by partial composting on the removal rate r w during

subsequent dry anaerobic batch digestion of the volatile solids.

208 E. ten Brummeler, L W. Koster

organic acids(gCO0/I)

0 10 20 30

pH(-)

-8 .7 .6 .5

,J.

time (days) Fig. 7. Organic acids concentration and pH during the dry anaerobic digestion of the partial composted (23.5% VS-reduction) organic fraction of MSW; (O) lactic acid; (O) acetic

acid; (+) butyric acid; (A) total organic acids; ( x ) pH.

rate of lactic acid was strongly affected by aerobic treatment. Lactic acid could not be detected at VS reductions of 4% or higher which was accompanied by low hydrogen-production rates. These results suggest that soluble sugars, which are the precursors for lactic acid in an unbalanced dry digestion of MSW, were degraded during the aerobic period. The initial methanogenic activity of the biomass (0"49g CH4-COD/liter/d) was sufficient when 23-5% of the VS was aerobically degraded. The initial acid- formation.rate was still higher than the total methanogenic activity of the inoculum, but the digestion started immediately. The reason for this is not clear. In Fig. 7 the organic acids and pH are shown for the experiment with 23"5% VS reduction. The pH values and the organic acids concentrations are characteristic for a more-or-less balanced digestion process; i.e. pH values around 7 and the total organic acids and hydrogen are readily converted without the excessive build-up found under unbalanced conditions (Fig. 2).

The aerobic pretreatment decreased the initial VS-removal rate, which is shown by the lower initial acid-formation rate. The easily-degradable parts of the substrate are apparently degraded during the composting period as was presumed above. It can be concluded that the start-up of the dry anaerobic batch digestion of the organic fraction of MSW can be accelerated by an aerobic pretreatment step. However, the amount of which has to be degraded during the partial composting means a loss of 40% of the potential methane yield. The composting time (artificial aeration) needed for this degree of VS reduction was 2 weeks. The long composting period and a significant loss of the potential methane means a major drawback for

Start-up o[dry digestion 209

implementation of the partial-composting step in the BIOCEL-process, as the treatment costs will be increased. A shorter partial-composting period, which results in increased temperature, in combinat ion with another start- up method might be more feasible. The heat which is released during the composting period could then be used for heating the waste before the digestion process. As was described earlier, another first start-up method, i.e. a start-up when digested organic fraction is not available, for dry anaerobic batch digestion is the addition of compost and a methanogenic inoculum (Ten Brummeler et al., 1988). However, a large amount of aerobic-stabilized organic fraction is needed when this method is applied. As both start-up methods show different technological advantages, the choice for a start-up method will depend on the economic implications for the costs of the process.

R E F E R E N C E S

American Public Health Association (1975). Standard Methods for the Examination of Water and Waste Water (14th edn) APHO, Washington DC.

De Baere, L. & Verstraete, W. (1984). Anaerobic fermentations of semi- solid and solid substrates. In Anaerobic Digestion and Carbohydrolysis of Waste, ed. G. L. Ferrero, M. P. Ferranti & H. Naveau., Elsevier Applied Science, London, pp. 195-210.

De Boosere, S., De Baere, L., Smis, J., Six, W. & Verstreate, W. (1986). Dry anaerobic fermentation of concentrated substrates. In Anaerobic Treatment, a Grown-up Technology. Industrial Presentations (Europe), BV, Schiedam, pp. 479-88.

De Zeeuw, W. J. (1984). Acclimatization of anaerobic sludge for UASB-reactor start-up. PhD thesis, Wageningen Agricultural University, Wageningen, The Netherlands.

Eastman, J. A. & Ferguson (1981). Solubilization of particulate organic carbon during the acid phase of anaerobic digestion. Journal WPCF, 53, 352-66.

Gujer, W. & Zehnder, A. J. B. (1983). Conversion processes in anaerobic digestion. Wat. Sci. Tech. 15, 127-67.

Koster, I. W. (1986), Characteristics of the pH-influenced adaptation of methanogenic sludge to ammonium toxicity, J. Chem. Techn. Biotechnol., 36, 445-55.

Koster, I. W., Ten Brummeler, E. & Zeevalkink, J. A. (1988). Anaerobic digestion of the organic fraction of municipal solid waste in the BIOCEL-process. In ISWA 88 Proceedings of the 5th International Solid Waste Conference, ed. L. Andersen & J. Moeller, Academic Press, London, pp. 71 6.

McDonald, P. (1982). The Biochemistry of Silage. Wiley and Sons Ltd., London. Membrez, Y. & Nicolet, R. (1985). M6thanisation en continu d'Ordures m6nag6res

ou autre D6chets fi haute teneur en Matiere s~ches. Gas-Wasser-Abwasser, 6fi, 782-4.

Noike,. T., Endo, G., Chang, J. E., Yaguchi, J. I. & Matsumoto, J. I. (1985) Characteristics of carbohydrate degradation and the rate limiting step in anaerobic digestion. Biotechnol. Bioeng., 27, 1482 9.

210 E. ten Brummeler, I. W. Koster

Spendlin, H. H. & Stegmann, R. (1988). Anaerobe behandlung von Biomiill, Miill und Abfall, 5, 204-15.

Strom, P. F. (1985). Effect of temperature on bacterial species diversity in thermophilic solid-waste composting, Appl. Environ. MicrobioL, 50, 899-905.

Ten Brummeler, E., Koster, I. W. & Zeevalkink, J. A. (1986). Biogas production from the organic fraction of Municipal Solid Waste by anaerobic digestion. In Materials and Energy from Refuse, ed. A. Buekens & M. Tels, KVIV, Antwerpen, Belgium, pp. 6.49-6.55.

Ten Brummeler, E., Koster, I. W. & Zeevalkink, J. A. (1988a). Dry anaerobic digestion of the organic fraction of municipal solid waste. In Advances in Water Pollution Control ed. E. R. Hall & P. N. Hobson, Pergamon Press, Oxford, pp. 335-44.

Ten Brummeler, E. & Koster, I. W. (in press). The effect of several pH control chemicals on the dry batch digestion of the organic fraction of municipal solid waste. Resources, Conservation and Re~Tcling.