Biotechnology Letters Vol 5 NO 2 113-118 (1983)

ANAEROBIC DIGESTION OF A PETROCHEMICAL EFFLUENT

T.J. Britz*, L.C. Meyer and P.J. Botes

Dept. Microbiology, University of the Orange

Free State, Bloemfontein, 9300, South Africa

SUMMARY

Downflow fixed bed reactors operated at 35"C, were sucessfully used for the

anaerobic digestion of a petrochemical effluent. COD reductions of 93-95:Z

were found at an optimum retention time of 2.3 days and a loading rate of

4.7 kg COD/m3/d. The amount of biogas produced was 0.88 m3/m3/d (STP),

with a methane content of 90-96%.

LNTRODUCTION

In today's industrial society, it has become increasingly important to pre-

vent the pollution of our limited water resources by providing adequate

treatment of effluents from industrial sources. Anaerobic waste treatment

is one of the major biological waste treatment processes in use, and has

been employed for many years in municipal sewage treatment units. The

search for greater efficiency and better economy as well as the interest in

methane as a renewable energy source, has led to the study of new types of

anaerobic reactors. In recent years considerable attention has been focus-

sed on a new range of reactors for the treatment of low strength industrial

waste water. One of the most successful of this range is the downflow

fixed film reactor (Kennedy and Van den Berg, 1982).

In South Africa a unique low strength petrochemical effluent is produced

by the Sasol oil-from-coal process. This low strength effluent contains

approximately 8-13 g/e volatile fatty acids. Large amounts of these

fatty acids are produced per annum and have to be disposed of by other

T,C zns. T'i_~ paper reports on the use of downflow fixed bed reactors for

the anaerobic treatment of low strength petrochemical effluent.

113

A = Substrate B = Peristaltic pump C = Thermometer D = Gas collector E = Liquid level control F = Temperature sensor G = Heating tape with

temperature control H = Reactor effluent I = Reactor column with

bacterial carrier

Figure 1 Diagram of the downflow fixed bed reactor

Details of the downflow fixed bed reactors are given in Figure 1. The

reactors consisted of 1 m glass columns, each with a working volume of 3.5

liter. The operational temperature was 35°C. Inside each reactor an inert

cylindrical, polyethylene bacterial carrier was placed. The surface area

of the support material was 3,500 cm2. The substrate was continuously

pumped in at the top of the reactor at the required rate while the reactor

effluent was removed from the bottom. The reactor liquid level volume was

electronically controlled (Britz et al., 1982). Gas exited at the top of

the reactor and volumes were determined by a means of a brine displacement

system.

At the start-up, the reactors were filled with sludge from a local munici-

pal plant and slowly fed with a diluted fatty acid solution, which corre-

sponded to the fatty acid composition of the factory effluent. With the

114

start of gas production, the dilution was gradually reduced until the fatty

acid solution corresponded to 50% of the effluent fatty acid concentration.

Approximately 8 weeks were required, at a hydraulic retention time of 4.0

days, to obtain a 90% COD removal.

Substrate

Effluent, produced during the Sasol oil-from-coal process, was used as sub-

strate. During the period that the reactors were tested, the effluent had

a total fatty acid concentration of approximately 9 000 mgfk?, comprising

(m/v> : acetic acid - 70%; propionic acid - 15%; isobutyric acid - 2%;

n-butyric acid - 8%; isovaleric acid - 1% and n-valeric acid - 3%. The

effluent alsocontained trace amounts of phenols, alcohols, ketones and emul-

sified oils. The pH of the substrate was adjusted to 6.0 with sodium hy-

droxide. Ammonium sulphate and potassium hydrogen phosphate was added at a

concentration of 0.05% (m/v). The COD value of the waste water was 11 000

w/E. A mineral analysis showed the following: less than 4.0 mg/& - K,

Mg, Cu, Mn, MO, N03-, SO =, 4

NH 4, Na, Ca, Cl and more than 20.0 mg/E - Fe..

Parameters

The following parametersonthe substrate and reactor effluent were moni-

tored: chemical oxygen demand (COD); soluble COD (after centrifugation

at 12 000 x g for 15 min); volatile fatty acids; total solids (TS); vola-

tile solids (VS); non-volatile solids (NVS); ammonia-nitrogen; Kjeldalhl-

nitrogen; feed rate; gas production and the gas composition. The analyyses

were done according to Standard Methods (1976). The fatty acids were de-

termined gas chromatographically, on a Hewlett Packard chromatograph modsel

58308, equipped with a flame ionization detector and a column (1.8 m x

1.5 mm ID) packed with Porapak Q, 80-100 mesh. The column temperature was

195°C and detector temperature 250°C with inlet temperature of 210°C. Ni-

trogen was used as Carrie-r gas. The gas composition was determined gas

chromatographically on a Perkin Elmer chromatograph, model 820, equipped

with a thermal conductivity detector and a column (4.0 x 0.3 mm ID) packed

with Porapak N, 80-100 mesh. Oven temperature was set at 55'C. Hydrogen

was used as a carrier gas and peak areas were compared with standards. The

amounts of gas produced were corrected to standard ,temperature and pressure.

115

RESULTS AND DISCUSSION

After adjusting the pH of the petrochemical effluent, a red-brown

precipitate formed. This was ascribed to the precipitation of iron and

other trace metals. With continuous agitation on a magnetic stirrer this

was used as substrate to feed a preconditioned reactor. The effectivity

of this reactor was found to deteriorate with time and stabilized at a COD

reduction of approximately 40% (Table 1). This low performance was as-

cribed to the precipitation of the minerals which possibly caused a growth

limitation for the microbial population.

A second reactor was fed with effluent that had been filtered through a

Whatman no. 1 filter to remove the precipitate and to which a mineral solu-

tion had been added, to prevent any growth limitations to the microbial

population. This reactor operated satisfactorily and for a period of 120

days a COD reduction of more than 90% was found (Table 1).

These results showthat the petrochemical effluent can be effectively

treated by the anaerobic digestion process using a downflow fixed bed

reactor. The data, shown in Table 1, was obtained at the hydraulic reten-

tion time (HRT) with the highest COD reduction. With the shortening of the

HRT beyond this point a sharp decrease in COD reduction and corresponding

increase in volatile fatty acids, especiall-y propionic acid, was found.

Thus the HRT with the highest COD reduction was considered as the optimum

for the system under the described conditions. It is possible that with

a bigger microbial surface area or different film supports, shorter HRT's

could be obtained as shown by Van den Berg and Kennedy (1981). The results

also showed that the microbial population, especially the methanogens, were

successfully retained on the support material. The reactor effluent COD

was found to consist of mainly soluble COD with the difference in soluble

and total COD being less than 2.0%.

The methane content of the gas produced by this type of reactor was very

high compared with the results of other workers. The collected gas con-

tained approximately 90-96% methane. Since the petrochemical effluent was

initially neutralized, the fatty acids were thus in the salt:form. During

the digestion process the acetate ions are assimilated by the microbial

population and the sodium ions are set free into the surrounding medium

where sodium hydroxide is regenerated. The pH of the reactor effluent was

found to be around 8.5. This could be the result of the regenerated sodium

hydroxide. The carbon dioxide formed in the reactor, and not used by the

116

methanogens, will react to form sodium carbonate. The excess carbon dioxide

is thus removed from the reactor and mostly only methane is set free explai-

ning the high methane to carbon dioxide ratio.

Table l- Reactor Performance (The data given was obtained at the hydraulic retention time with the highest COD reduction)

Parameter Substrate

- Unfiltered Effluent Filtered Effluent

Hydraulic retention time (d) 3.5 2.3

Loading rate (kg COD/m3/d) 3.15 4.7

COD removal (%> 40 93 - 95

Biogas production rate (m3/m3/d) 0.15 0.88 (SE')

Methane content (X) 60 90 - 96

Volatile fatty acid removal (X-) 40 93 - 96

pH reactor effluent 8.32 0.2 8.62 0.2

CONCLUSIONS

1. The results show that effluent from a petrochemical industry is amenda-

ble to treatment by an anaerobic digestion process. The HRT's could

probably be shortened by the increase in microbial surface area.

2. The precipitation of iron and trace metals caused a growth limitation

for the microbial population. This limitation could be removed by

filtering the effluent and adding a mineral solution.

3. The methane content was high (90-96%) due to the self-scrubbing effect

of this type of reactor.

ACKNOWLEDGEMENTS

The financial assistance of the Central Research Fund of the U.O.F.S.,

Council of 5.1ientific and Industrial Research and Sasol is acknowledged..

We are gra. ul to the Fedmech Foundation for generously providing a study

grant for L.C. Meyer. We are also indebted to R. du T. Burger, Dept. Ground

117

Science, for the atomic absorption analysis and to H.J.J. van Vuuren for

reviewing the manuscript.

REFERENCES

1. APHA, AWWA, WPCF. (1976). Standard Methods for the examination of

water and waste water. Washington, D.C., 14th ed.

2. Britz, T-J., De Witt, B., Hugo, A.B., and Meyer, L.C. (1982). Lab.

Practice, In Press.

3. Kennedy, K.J., and Van den Berg, L. (1982). Biotechnol. Lett. 4,

137 - 142.

4. Van den Berg, L., and Kennedy, K.J. (1981). Biotechnol. Lett. 3,

165 - 170.

118