methane recovery from water hyacinth through anaerobic activated sludge process

Methane Recovery from Water Hyacinth through Anaerobic Activated Sludge Process

N. 8araswat and P. Khanna Centre for Environmental Science and Engineering, Indian Institute of Technology, Powai, Bombay, 400076, India

The concepts of phase separation, anaerobic activated sludge process, and alkali pretreatment have been in- corporated in this investigation with the objective of de- veloping rational and cost-effective designs of diphasic anaerobic activated sludge systems, with and without alkali treatment, for methane recovery from water hyacinth (WH). Evaluation of process kinetics and optimization analyses of laboratory data reveal that a diphasic system with alkali treatment could be designed with an alkali Pretreatment step (3.6% Na2C03 + 2.5% Ca(OH), (w/w) of WH, 24 h duration) followed by an open acid phase (2.1 days HRT) and closed methane reactor with sludge recycle (5.7 days HRT, 7.7 days MCRT) for gas yield of 50 L/kg WH/d at 35-37°C. Likewise, a diphasic system without alkali treatment could be designed with an open acid phase (2 days HRT) followed by closed methane reactor with sludge recycle (3.2 days HRT, 6 days MCRT) for gas yield of 32.5 L/kg WH/d at 35-37°C. Detailed economic analyses bring forth greater cost-efficacy of the diphasic system without alkali treatment and reveal that the advantage accrued in terms of higher gas yield is overshadowed by the cost of chemicals in the diphasic sytem with alkali treatment.

INTRODUCTION

Water hyacinth (WH) has a high productivity coupled with an excellent pollutant removal and methane gen- eration potential that makes it amenable to utilization in integrated pollution control and energy conversion systems in rural areas. This energy potential is not exploited fully in the conventional single-stage anaerobic digestion systems due to inherent limitations of the traditional design procedures and the lignocellulosic nature of the substrate.

Modifications of the single-stage process involving physical segregation of biochemically diverse acid and methane bacteria and microbial or chemical pretreatment of feed have shown to achieve better substrate conversion. Several researcher~'-~ have demonstrated the advantages in terms of improved gas yields, conversion efficiency, and stability as well as the reduction in hydraulic retention time (HRT) achieved through the segregation of phases in anaerobic digestion. HRT in the methane phase could be reduced further in the

Biotechnology and Bioengineering, Vol. XXVIII, Pp. 240-246 (1986) 0 1986 John Wiley & Sons, Inc.

anaerobic activated sludge system, with concomitant maintenance of a high mean cell residence time (MCRT), by the recirculation of sludge' facilitating higher substrate conversion efficiencies and cost-effective biogas system designs.

Three types of pretreatment are reported in literature, viz., enzymatic hydrolysis, acid hydrolysis, and alkali pretreatment, for improved digestibility of lignin-rich biomass. These methods could be used either as a full treatment step for complete hydrolysis to sugars (as in the case with enzymatic and acid hydrolysis6'*) or as an alkali pretreatment step for breaking certain links in the hemicellulose-lignin polymeric system so as to provide increased diffusivity to hydrolytic enzymes.' While the enzymatic hydrolysis methods suffer from the need for stringent pretreatment steps, which in itself could be a difficult and expensive proposition, the acid hydrolysis methods require large capital in- ves tmen t~ .~ Alkali pretreatment methods have ac- cordingly been adopted by several researchers"-'* for achieving increased volatile acid and gas production in anaerobic digesters. An economic analysis of these methods has, however, not been reported so far in the literature.

This investigation addresses itself to the development of rational and cost-effective designs of diphasic anaerobic activated sludge systems for methane recovery from water hyacinth incorporating phase separation, sludge recycle, and alkali treatment (AT). Process kinetics has been evaluated experimentally for a diphasic system, with AT (Fig. 1) comprising an AT step followed by an open acid phase and a closed methane phase with sludge recycle; and for a diphasic system without AT (Fig. 2) comprising an open acid phase followed by a closed methane phase with sludge recycle. Design parameters have been ascertained through optimization analyses of laboratory data. Finally, an economic analysis has been carried out to evaluate whether the benefits that accrue in terms of increased gas production offset the expenditure incurred in the additional step and chemicals used in the diphasic system with AT.

CCC 0006-3592/86/020240-07$04.00

Water Hyaeinth powder , Lime a n d Sodium Carbonate,

Water I

o ut let r Gas

Sludge recirculation

AR Sludge t o SOB

@ Alkal i t r ea tmen t tank

@Acid Reactor (AR l )

@ B a l a n c i n g Reservoir

@Pump t o MR

@ S e t t l i n g B a s i n

@ S ludge p u m p

@ Sludge d r y i n g b e d

@Methane Reactor (MR-1)

Figure 1. Schematic flowsheet of diphasic system with alkali treatment.

Water Hyacinth Powder

Water

1

@ Acid Reactor (AR2)

@ Balancing Reservoir

-Gas outlet

- Treated Ef f luent

c

SI ud g e Recircu la t i on

AR Sludge to SDB

@ Methane Reactor (MR2)

@ Settling Bas in

@ Sludge Drying Bed (SDB)

Excess MR Sludge to SDB

I

0

Figure 2. Schematic flowsheet of diphasic system without alkali treatment.

SARASWAT AND KHANNA: METHANE RECOVERY FROM WATER HYACINTH 241

MATERIALS AND METHODS

Substrate, Inoculum, and Start-up

A slurry (5%) of ball-milled WH (particle size, 1.25 mm) was used as substrate in all the experiments. The inoculum was obtained from a conventional single- stage anaerobic digester (SR) operating on cattle-dung slurry. The acid phase was started by acclimating the SR inoculum to WH slurry in open reactors (ARI, AR2) and operating these at decreasing HRT till meth- anogenesis was inhibited.

The methane phase was started by acclimating the SR inoculum to the volatile acid-rich supernatant obtained from acid phase effluent after settling and operating in anaerobic reactors (MRI, MR2) at an HRT of 10 days.

Separation of phases was thus achieved through kinetic control and selective inhibition of methane formers in the acid phase. Acidogenesis was inhibited in the methane phase as the feed to methane reactors was devoid of substrate for acid formers.

Experimental

Alkali pretreatment experiments were conducted in a series of flasks at ambient temperature (30°C) and pressure conditions. WH was treated with various concentrations of NaOH, Ca(OH),, and Na2C03- Ca(OH), mixture for 24 h. After treatment, the experimental flasks were inoculated with seed culture from ARl and allowed to undergo acidogenesis for 48 h. Controls were run by using alkali-free media. The efficacy of AT was ascertained by estimating the volatile acids produced in each case.

Acid phase studies were conducted with alkali-treated WH in the diphasic system with AT and nontreated WH in the diphasic system without AT. The experiments were carried out in open semicontinuous reactors (ARI, AR2) in both cases. The effect of HRT on product yield was determined by estimating steady- state concentrations of substrate COD, microorganisms, and volatile acids.

Methane phase studies in both systems were conducted with supernatant from the respective acid phase (ARl, AR2) effluents. The completely mixed contin- uous-flow methane phase reactors (MRI, MR2) were maintained at 35-37°C under anaerobic conditions. The effluent slurry was settled and the sludge recycled to the reactors. The flow and sludge recycle rate were varied to determine the effect of HRT and MCRT on gas yield. Steady-state concentrations of substrate volatile acids, microorganisms, and gas yield were estimated in each case.

Analytical Procedures

All analyses were conducted as per standard meth- o d ~ . ’ ~ Volatile acids were estimated as acetic acid

equivalents by the distillation method, and microorganism concentrations were estimated by analyzing the mixed liquor volatile suspended solids (MLVSS) in the acid and methane phases. The effluent substrate COD in the acid phase was estimated by subtracting the COD of effluent volatile acids from the total effluent COD. The COD of effluent volatile acids was determined as 1.066 times the effluent volatile acid concentration from the equation

CH3COOH + 202 - 2C02 + 2H20 (1)

RESULTS AND DISCUSSION

Alkali Treatment (AT)

Results of AT experiments, presented in Figure 3, reveal that the treatments with NaOH, Ca(OH),, and Na,C03-Ca(OH), mixtures provide 1.6, 1.5, and 2 times higher volatile acid yields, respectively, as compared to the control samples. Further, the maximum yield of volatile acids (4.8 g/L) is obtained on fermentation of WH treated with 3.6% Na,CO3 and 2.5% Ca(OH), (w/w WH) mixture for 24 h. The optimum time (24 h) for treatment was ascertained through another study with treatment durations ranging from 0 to 72 h.

These results compare well with similar studies reported in the literature,” where treatment of corn stover with 0.663% Na2C03 and 0.5% Ca(OH), (w/v) mixture for 48 h followed by acid fermentation of 5% slurry

52 O O r

20001 I I I I I I 0 0.01 0.02 0.03 0.04 0.05

Alkali Concentration ,(g OH-/g W H )

Figure 3. Volatile acid production on acid fermentation of water hyacinth treated with various alkalis: (0-0) NaOH, ( x --- x ) Ca(OH),, (&A) Na,CO,-Ca(OH),.

242 BIOTECHNOLOGY AND BIOENGINEERING, VOL. 28, FEBRUARY 1986

for an equal duration produced 4 g/L volatile acids. A lower concentration of Na2C03-Ca(OH)2 mixture and shorter treatment duration is reasonable for WH in view of a lower volatile solids lignin content of

as compared to 20% in corn stover.I2 8% 1 1.12. I4

Acid Phase

The optimum hydraulic retention time in the acid phase (ARI) of the diphasic system with AT has been evaluated from the maximum rate of product formation criterion through the following steps:

1 . Steady-state concentrations of substrate and microorganisms at various HRTs have been used to estimate the kinetic coefficients in Table I through es- tablished procedures. Is,'

2. The theoretical effluent substrate concentration (S) has been calculated from the equation'j

3. The product output rate (P) has been calculated from the equation2

(3)

Figure 4 depicts the rate of product output for various HRTs in ARl. The optimum HRT is identified as 2.1 days with corresponding theoretical product output of 6.8 g/L d. A maximum volatile acid yield of 5.6 g/L d corresponding to an HRT of 2 days was achieved in the experimental work.

Data analyses for the diphasic system without AT provides an optimum hydraulic retention time of 2 days in the acid phase (AR2). The variation of volatile acid concentration with HRT, shown in Figure 5, in- dicates that a yield of 3.1 g/L d could be obtained in this system as compared to 5.6 g/L d in the diphasic system with AT.

a(Ue + me) (So - S) O(U, + Ue + me) P =

Methane Phase

The methane phase has been optimized for maximum gas yield at minimal cost based on costs of the methane

Table I. Kinetic coefficients for acid phase with alkali treatment.

Kinetic coefficient Value

Half velocity constant, K , Maximum rate of substrate utilization

per unit mass of microorganisms, k Growth yield coefficient, Y Endogenous decay coefficient, Kd Maximum specific growth rate, pa True product yield constant, CI

Substrate utilization coefficient for

Substrate utilization coefficient for en-

Maintenance coefficient, rn

growth, U,

ergy, Ue

12.8 g COD/L

53.56 d-' 0.0233 mg/mg COD 0.2068 d- ' 1.25 d- ' 0.7872 mg/mg

27.25

15.645 8.8717 d - '

T ,Theoretical

0 ,(days)

Figure 4. system with alkali treatment.

Estimation of optimum 0 for acid phase (ARI) in diphasic

'0 0.5 1.0 1.5 2.0 2 . 5 3.0 3.5 4.0 e,(days)

Figure 5. Volatile acid production in acid phase (AR2) of diphasic system without alkali treatment.

reactor, pumping for sludge recycle, and the sludge drying bed. Optimum HRT and MCRT in the methane phase (MRl) of the diphasic system with AT have been ascertained with recourse to the following steps:

1 . Steady-state concentrations of substrate volatile acids and microorganisms at various HRTs and MCRTs have been used to estimate the kinetic coefficients reported in Table 11.

2. Minimum mean cell residence time (0:) has been estimated as 3.2 days from the equationI5

(4)

3. A safety factor (SF) of 2.4 has been applied to


Table 11. Kinetic coefficients for methane phase (MR1).

Kinetic coefficient

Half velocity constant, K ,

Maximum rate of substrate utilization per unit mass of microorganisms, k

Endogenous decay coefficient, Kd Growth yield coefficient, Y

Maximum specific growth rate, p,,,

Value

111.18 mg volatile acids/L

0.4249 d-' 0.0777 d-' 0.9375 mg/mg

0.4 d - ' volatile acids

8: for estimating 8, design (8%) as 7.7 days from the expressi~n '~

(5 ) 4. Average values of X r (15,000 mg/L) and X (3000

mg/L) have been plugged in the following equation' to calculate 8 for various recycle ratios r ranging from 0.03 to 0.15:

8: = 6 x SF

(6)

5. Values of sludge density W, total head H, hydraulic flow Q , and efficiency r ) have been plugged in equation (7) to calculate break horsepower (BHP) for various values of r:

Xr X

8 = O,(l + r - - r )

WQrH BHP = -

757 (7)

6. The 8 and BHP values obtained for various recycle ratios from equations (6) and (7) have been converted to nondimensional parameters r/r*, 8/8* and BHP/BHP* by dividing all r values by 0.03, all 8 values by 2.816 (which corresponds to r = 0.03), and all BHP values by 0.0168 (which corresponds to r = 0.03). These parameters have been used to construct the nondimensional plot in Figure 6. The intersection of the two curves identifies optimum design r and 8 values as 0.061 and 5.7 days, respectively.

OO do 1 2 3 4 5

r/c * Figure 6. Evaluation of optimum rand 8 in methane phase (MRI) of diphasic system with alkali treatment: (0-0) 8/8* versus r / r * , (A-A) BHP/BHP* versus r/r*.

7. Using the optimum r value identified in step 6, the costs of the methane reactor and sludge drying bed (SDB) have been calculated as a function of 8,. These costs are plotted in Figure 7. The intersection of the two cost curves provides a 8, value of 7.7 days. This value converges with the 8, obtained from safety factor considerations in step 3 and identifies optimum design 8,.

Data analysis for the diphasic system with AT therefore provides optimum 8, B,, and r values of 5.7 days, 7.7 days, and 0.061, respectively, in the methane phase (MR1).

Data analysis for the diphasic system without AT provides optimum 8, 8,, and r values of 3.2 days, 6 days, and 0.116, respectively, in the methane phase (MR2). The variation in gas production with 8, in the two systems is depicted in Figure 8. The maximum gas yield in the diphasic system with AT is recorded as 50 L/kg dry WH/d as against 32.5 L/kg dry WH/d in the diphasic system without AT.

Maximum substrate conversion efficiencies of 20- 25% in the acid phase and over 97% in the methane phase could be achieved in this investigation. The low efficiency in the acid phase is attributed to a low food- microorganism ratio adopted due to fluid dynamic considerations and is corroborated by the high endogenous decay coefficient ( K d = 0.2068 d-*) reported in Table I.

The biogas contained 69% methane. Fertilizer solids (0.70 kg/kg WH) with a nitrogen, phosphorus, and potassium content of 2.03, 0.069, and 1.81%, respectively, were recovered in the acid phase of both systems. The pH in the alkali treatment, acid phase, and methane phase effluents ranged from 7 to 7.5, 6 to 6.5, and 7 to 8, respectively. The alkalinity in the methane phases hovered around 3000-4000 mg/L. Therefore, neither system necessitated pH adjustments at any stage.

Iooo t / , I Optimum ec =7.7 days I

OO 1 '4 I { e,,(days) Ib 1; 1; 16 l ls 2b

Figure 7. Evaluation of optimum 8, for methane phase (MRI) in diphasic system with alkali treatment: (0-0) cost of methane reactor versus O,, (A-A) cost of SDB versus 8, (Rs 100 = 8.14 U.S. dollars).


5 5 1 I 0 0

40t / / d

235 5 t I

10

I 1 , 4 0 2 4 6 8 10 12 14 l f i L

Bc,(days)

Figure 8. time in methane phase: (0-0) MR1, (A-A) MR2.

Variation of gas production with mean cell residence

Comparison of kinetic coefficients for acid and methane phases in this investigation with values reported for other substrates in the literature bring forth several interesting aspects. Due to the complex nature of water hyacinth, the maximum specific growth rate p, of 1.25 d- ' for acid formers is observed to be substantially lower than the values of 7.2, 3.84, and 1.7 d-' reported lw7 for acid phase digestion of glucose, sewage sludge, and pure cellulose, respectively.

The maximum specific growth rate p, of 0.4 d-' for methane formers compares well with the value of 0.43 d- ' reported4 for methane phase digestion of con- fectionary waste. Further, it is observed that the maximum specific growth rate of acid formers for water hyacinth is only 3 times higher than that for methane formers, whereas for pure substrates such as glucose the order of magnitude of pa is reported' to be almost 10 times larger than pm. This implies difficulty in separation of phases by kinetic control alone, and ac- cordingly the strategy of selective inhibition of methane formers by maintenance of acid phase in an open reactor in the present study appears appropriate.

The half-velocity constant K , of 12.8 g/L in the acid phase is almost half the values of 24 g/L for single- stage digestion of alkali pretreated biomass-waste blend" and 26 g/L for acid phase digestion of sewage sludge.* The K , value of 11 1 mg acetic acid/L for the methane phase is also observed to be lower than the values of 164 mg butyric acid/L and 369 mg acetic

acid/L reported4 for methane phase digestion of con- fectionary waste, probably due to the entirely soluble nature of the methane reactor feed. This, in turn, brings out rapid conversion of substrate in the methane phase.

ECONOMIC ANALYSIS

Economic a n a l y d 8 for plant capacities between 2 and 71 m3 biogas/d reveals that the unit cost of biogas production ranges from Rs 4.6/m3 to Rs 5.2/m3 and Rs 0.6/m3 to Rs 1.09/m3, respectively (Rs 100 = 8.14 U.S. dollars), for diphasic system with and without AT. Also, the investment payback periods range from 8.7 to 15.4 months and 4.5 to 8.5 months, respectively.

Notwithstanding lower capital cost for a diphasic system with AT, in small capacity plants (2-4 m3/d), the unit costs are still observed to be higher due to the high cost of chemicals. A perusal of Figure 9, showing variation in annual costs with plant capacities, reveals that the difference between costs in the two systems is even more pronounced for higher capacities predominantly due to the chemical cost component in diphasic system with AT.

Cost estimations for an existing single-stage WH plant based on the KVIC design" operating at a hydraulic retention time of 30 days with a gas yield of 20 L/kg WH/d at 37°C bring out the unit cost of biogas production in this system as Rs 5.30/m3 with an investment payback period of 25.2 months for a capacity of 2 m3 biogas/d.

The designs based on the present investigation are thus more cost-effective than those available in the literature.

140[ 120

I 0 10 20 30 40 50 60 70 80

P l a n t Capac i ty , (m3 biogasld)

Figure 9. Total annual cost for various plant capacities: (0-0) diphasic system with alkali treatment, (A-A) diphasic system without alkali treatment.


CONCLUSIONS

Methane recovery from water hyacinth has been studied so far only in conventional single-stage anaerobic digestion systems characterized by high capital cost, long hydraulic retention times, and low gas yields. The present study establishes higher gas yields with recourse to a diphasic anaerobic activated sludge system at substantially lower hydraulic retention times, thereby providing more cost-effective designs. Alkali pretreatment, however, proves to be an expensive proposition. The benefits that accrue due to higher gas yields in the diphasic system with alkali treatment are overshadowed by the high cost of chemicals. Further research is proposed to reduce the cost of the AT step in this system with recourse to low-cost alkalis.

This research was initiated by the Indian Institute of Technology, Bombay, in 1980 and later supported through a project by the Department of Non-Conventional Energy Sources, Ministry of En- ergy, Government of India, in 1982. The authors acknowledge the contribution of Ms. N. S. Naik and Mr. A. P. Annachhatre toward the experimental work.

NOMENCLATURE A H k

Kd KS

m P Q Qr

r SO S SF U P

u, W X X , Y

effluent volatile acid concentration in acid phase (mg/L) total head (m) maximum rate of substrate utilization per unit mass of microorganisms (d ’) endogenous decay coefficient (d-’) half-velocity constant (mg/L) maintenance coefficient (d-’) rate of volatile acid production (mg/L d) flowrate (L/d) flowrate in sludge recycle line (L/d) recycle ratio influent substrate concentration (mg/L) effluent substrate concentration (mg/L) safety factor substrate utilization coefficient for growth substrate utilization coefficient for energy sludge density (kg/m’) microorganism concentration (mg/L) microorganism concentration in sludge recycle line (mg/L) growth yield coefficient (mg/mg)

Greek letters a 8

true product yield constant (mg/mg) hydraulic retention time, HRT (days)

OC 0: p, pm 7 efficiency (%)

mean cell residence time, MCRT (days) minimum mean cell residence time (days) maximum specific growth rate for acid formers (d-’) maximum specific growth rate for methane formers (d-’)

References 1.

2.

3. 4.

5.

6.

7.

8.

9.

10.

11.

12. 13.

14.

15.

16.

17. 18.

19.

S. Ghosh and F. G. Pohland, J. Water Poll. Control Fed., 46, 748 (1974). S. Ghosh, J . R. Conrad, and D. L . Klass, J. Water Poll. Control Fed., 47, 31 (1975). C. C. Claussen and J . L. Gaddy, AIChE, 74, 56 (1978). M. L. Massey and F. G. Pohland, J. Water Poll. Control Fed., 50, 2204 (1978). A. W. Lawrence and P. L . McCarty, J. Sun. Eng. Piv., A X E , 96, 757 (1970). G. T. Tsao, in 3rd Annual Biomass Energy Systems Conference Proceedings, SERI/TP-33-285, (1979), p. 77. M. Mandel, in 3rd Annual Biomass Energy Systems Conference Proceedings, SERI/TP-33-285, (1979), p. 281. D. I. C. Wang, in 3rd Annual Biomass Energy Systems Con- ference Proceedings, SERIITP-33-285, (1979), p. 61. H. Tarkow and W. C. Feist, Symposium Papers, “Celluloses and their Applications,” 156th Meeting A.C.S., New Jersey, (1969), p. 197. D. L. Klass, S. Ghosh, and J. R. Conrad, Symposium Papers, “Clean Fuels from Biomass, Urban Refuse, Agriculture Wastes,” IGT, Orlando, Fla. (1976). S. Ghosh, M. P. Henry, and D. L. Klass, Biotechnol. Bioeng., Symp., 10, 164 (1980). R. Datta, Biotechnol, Bioeng., 23, 61 (1981). APHA, AWWA, WPCF, Standard Methods for Examination of Water and Wastewater (American Public Health Association, Washington, D.C., 1976). 3. A. Chandler, W. J. Jewell, J. M. Gossett, P. J. Van Soest, and J. B. Robertson, Biotechnol. Bioeng., Symp., 10,93 (1980). Metcalf and Eddy Inc., Wastewater Engineering Treatment, Disposal and Reuse (Tata-McGraw Hill Publishing Company Ltd., New Delhi, 1979). D. L. Klass and S. Ghosh, Symposium Papers, “Clean Fuels from Biomass and Wastes,” IGT, Orlando, Fla., 1977), p. 323. S. Ghosh and D. L . Klass, Proc. Biochem., 13, 15 (1978). P. Khanna, A. P. Annachhatre, N. S. Naik, andN. Saraswat, Ministry of Energy and Syndicate Bank Competition on Low Cost and Efficient Model of Biogas Plants (Environmental Sci- ence and Engineering Group, IIT Bombay, 1984). KVIC Report, Gobar Gas Scheme, “Research and Development Work under Biogas Programme: Retrospect and Prospect,” Khadi and Village Industries Commission, Bombay, India, July 18, 1983.


methane recovery from water hyacinth through anaerobic activated sludge process

Documents