Professional Documents
Culture Documents
545 to 551
Pergamon Press Ltd 1980. Printed in Great Britain
Abstract--The imperfect fungus Geotrichum candidura, when grown in batch and in single and two-stage
continuous laboratory culture, has been shown to produce reductions of up to 92% in the BOD loading
of Irish malt whiskey distillery waste. Maximum COD and TOC reductions were 81% and 76% respect-
ively. Yields of fungal biomass of over 30 g 1-1, containing more than 45% "true" protein, have been
obtained. The economic possibilities of employing such a process for waste treatment and perhaps
single-cell protein production seem encouraging.
had taken place and were stored deep-frozen until required determinations were carried out 'after filtration of repli-
for analysis. Prior to such analysis they were filtered cated volumes of the culture through dry, pre-weighed
through Whatman No. 1 filter paper, but it was felt that an Oxoid 0.45/tin membrane filters. These were then washed
adequate determination of suspended solids, reported in and dried to constant weight at 40°C prior to a final re-
the region of 6g1-1 by Campbell (1970), could not be weighing. The culture filtrates were collected and stored
adequately carried out on a small sample. Total carbo- deep-frozen; analyses of pH, BOD, COD and TOC were
hydrate was determined in the filtrate by the anthrone later carried out.
method of Fairbairn (1953); total protein was measured To ascertain if supplementation of the spent wash would
using the biuret method (Herbert et aL, 1971). Free amino be required on a routine basis to obtain maximal reduction
acids were measured by the method of Yemm & Cocking of its organic loading and yield of G. candidura, samples
(1955L and free glycerol by the periodate/acetylacetone collected on four separate occasions were supplemented
colorimetric technique for triglyceride determination with NH,NO3, (4gl - t ) KH2PO,, (4g1-1) urea, 14gl - l )
(Sigma Chemical Company) omitting the saponification or glucose (10 g l- 1), and inoculated and incubated for 40 h
step. Organic acids were detected qualitatively by the as described previously.
paper chromatographic method of Isherwood and Hanes Similar 40 h batch culture studies were carried out on
quoted in Ranson (1955), and total titrable acidity twelve samples of unsupplemented spent wash collected on
determined by titration to pH 7.0 with 0.01 M NaOH separate occasions over the period 1976-1977.
using bromothymol blue as indicator. Total dissolved
phosphorus was determined using the ascorbic acid Continuous culture experiments
method of Herbert et al. (1971), but increasing the concen- (a) Single stage. These experiments were carried out in a
tration of reductant to 0.1 M, and total organic carbon glass, water-jacketed, column fermenter of 350 ml capacity
(TOC) using a Tocsin Aqueous Carbon Analyser (Phase Sep- (Fig. 1), latterly at a temperature of 22°C when it has been
arations Ltd., Queensferry, Flintshire). Biochemical Oxy- established that this was optimal. Aeration and mixing
gen Demand (BOD) determinations were made according were achieved by the passage of pre-humidified air through
to APHA Standard Methods 13th edition (1971)--the dilu- a sintered plate in the bottom of the fermenter, supplied at
tion water containing seed previously "adapted" to spent the rate of 6 culture vols min- 1 to ensure adequate mixing,
wash--and Chemical Oxygen Demand (COD) determina- since this is often difficult to achieve in column fermenters.
tions by the method of Canelli et al. (1976). Copper was Because of the relatively small volume of the fermenter the
measured either by atomic absorption spectrophotometry sample size for biomass determinations and chemical
or colorimetrically after reaction with bathocuproin disul- analyses was kept to a minimum--if possible 25 ml; sam-
phonate (Boehringer Mannheim G.m.b.H). The two copper piing was carried out when a steady state had been
determination methods showed good agreement. attained at each dilution rate studied--normally this
required the passage of three to five culture volumes of
Selection of organism medium. The spent wash, which was supplied to the fer-
In similar situations, e.g. the work of Smith & Bull menter via an LKB Varioperpex peristaltic pump, had
(1976) on coconut processing wastes, the decision was been sterilized after collection to permit its storage; other-
taken to use a microorganism which is already an wise no precautions were taken to ensure axenic operation.
approved food source and therefore invariably one of the (b) Two stage. Two fermenters similar to that described
fourteen yeasts listed by Smith (1975). In the present above were arranged in series, so that the effluent from the
instance, when the primary aim was to reduce the pollution first became the input to the second (there was no attempt
load of a waste which is itself the by-product of a yeast at biomass separation between the stages). Dilution rate in
fermentation this approach seemed futile. Enrichment cul- the first fermenter was maintained at 0.125 h - 1; by altering
tures were accordingly set up using inocula from a variety the relative culture volumes in the two fermenters (and
of natural habitats including aerobic and anaerobic soils, their air flows accordingly) dilution rates could be achieved
and activated sludge plants treating domestic and food in the second ranging from 0.085 h - 1 to 0.15 h- 1.
processing wastes. In all instances the predominant or-
ganism in such cultures was identified as the imperfect
fungus Geotrichum candidum Link ex Pers.; the strain iso- RESULTS
lated from a municipal activated sludge plant in Coleraine
was chosen, and maintained as a stock culture on spent Analytical details of some of the various spent wash
wash solidified with agar (1.3~ w/v). samples collected during the course of this work are
presented in Table 1 ; those relating to a single sample
Batch culture experiments
of a local grain distillery spent wash lcollected
Batch culture experiments were normally carried out
October 1977) are included for comparison. In addi-
using 25 ml samples of spent wash in 250 ml Erlenmeyer
flasks, inoculated with 0.2 ml of a suspension of stale G. tion qualitative paper c h r o m a t o g r a p h y indicated that
candidum arthrospores, which had been precultured in the only two organic acids were present, lactic and succl-
same medium. The spent wash was collected as required n i c - - t h e latter in much smaller quantities. Specific
(normally monthly) immediately after distillation, and the chemical a n d enzymic analyses for these have proved
supernatant decanted after having been allowed to settle
for a few hours so that its suspended solids content was ineffective to date, presumably because of the general
negligible. It was then sterilized by autoclaving for 20 min complexity of the liquor a n d also because of its high
at 115°C, preliminary experiments having shown that b a c k g r o u n d colour which makes colorimetric
autoclaving of the growth medium had no effect on the measurement extremely difficult. Accordingly, total
subsequent yield of organism. Preliminary work had also
titrable acidity has been expressed as lactate.
indicated that the growth rate of the organism was
retarded if the Erlenmeyer flask contents exceeded 10~o of Figure 2 illustrates a batch culture experiment in
the flask volume. which the unsupplemented spent wash used as growth
The flasks were incubated at 22°C in a New Brunswick m e d i u m h a d been diluted fivefold with distilled water.
reciprocal shaker (120 rev rain- 1). Early batch experiments The growth curve shows two growth phases---the first
were carried out using spent wash diluted rivet'old with
distilled water--in later experiments, and in all continuous e x p o n e n t i a l - - a n d a diauxic lag. Cultures grown on
culture work, the waste was used undiluted. Dry weight other spent wash samples did not always reveal the
r
Total
titrable Total
Free acidity dissolved
Date of Carbohydrate Protein amino acids Glycerol (as lactate) BOD COD TOC phosphorus Copper
collection pH gl-I gl-I g l -I gl t gl-t mgl-i mg 1-1 mg I - l P.P.M. P.P.M
?,
1976
May 3.80 13.5 19.8 2.9 5.0 5.8 28,500 47,000 18,820 1270 <0.2
June 3.25 I 1.0 21.5 3.9 6.4 14.0 35,500 42,250 25,620 740 0.4
Sept. 3.20 6.9 15.1 3.4 4.4 11.6 27,500 47,000 35,000 960 2.0
Nov. 3.30 1!.6 17.3 2.8 5. I 9.2 26,000 48,000 N.D. 1280 1.2
Nov. 3.30 9.5 19.0 3.2 5.1 12.3 30,000 52,000 24,320 1140 N.D.
1977
Jan. 3.20 7.0 17.5 2.8 4.4 12.4 27,500 35,500 21,440 1310 1.9
Mar. 3.10 I I.I 23.0 3.3 5.6 17.6 40,500 51,000 30,020 1380 3.5
May 2.90 20.0 24.1 2.1 5.4 18.4 38,500 58,500 28,660 1300 N.D.
June 3.20 8.0 17.3 2.1 6.1 13.5 24,500 35,500 20,660 1180 4.8
Aug. 3.85 21.2 26.5 2.2 6.9 4.9 33,000 52,000 18,820 1460 4.8
Sept. 3.35 6.7 23.0 3.4 6.2 10.7 28,500 36,000 21,440 1570 0.5
Oct. 3.20 10.4 31.0 4.3 7.5 18.4 43,000 47,500 32,780 1210 0.2
Oct.* 3.20 2.9 4.9 0.6 2.6 2.3 10.500 15,000 4920 1960 N.D.
Hedium
4°r
C0o(ing
30.
~ 2o. 80
60
10-
40
(
00( • -3-0
10 20 3O
Time (hi
Table 2. Biomass production by Geotrichum candidum cul- Table 5. Percentage decrease in BOD, COD and TOC
tured in batch on spent wash with various supplementa- achieved as a result of the batch culture of Geotrichum
tions. Levels of supplementation were: NH,NO3 4 g i- t; candidura on samples of unsupplemented spent wash
KH2PO, 4 g l - t ; urea 4 g l - t ; glucose 10g1-1
BOD % COD % TOC
Date of Biomass Date reduction reduction reduction
sample Supplementation production (g 1- t)
1976
Nov. 1975 NIL 17.6 May 64.9 31.7 22.5
NH4NO3 14.8 June 73.2 38.6 58.2
KH2PO4 18.0 Sept. 74.1 61.5 51.0
UREA 18.4 Nov. 63.5 48.4 N.D.
GLUCOSE 22.4 Nov. 68.3 53.3 55.6
Dec. 1975 NIL 17.6 1977
NH,NOa 17.2 Jan. 74.5 50.6 52.5
KH2PO, 17.2 Mar. 91.4 68.0 61.9
UREA 18.4 May 87.0 77.8 75.6
GLUCOSE 21.6 June N.D. 68.9 58.2
March 1976 NIL 20.4 Aug. 72.1 57.4 61.1
NH4NOa 25.6 Sept. 75.7 76.5 N.D.
KH2PO4 23.2 Oct. 70.4 68.6 69.4
UREA 22.0
GLUCOSE 25.6
August 1977 NIL 27.2 such errors are magnified by the very large dilutions
NH,,NOa 26.0
KHzPO4 29.6 of the waste which were necessary for most determin-
UREA 31.2 ations. In this instance the increases in diluent BOD
GLUCOSE 32.8 and TOC compared with the effluent from unsupple-
mented waste are in good agreement.
Accordingly the data reported in Tables 4 and 5 on
final culture filtrates, it was decided that such addi- the reduction of the organic loading of a series of
tions would not be made. The l o w effluent COD spent wash samples, through the batch culture of G.
value of 18,000 after supplementation with urea candidum, and on the biomass yields thus produced
(Table 3) must be attributed to experimental error-- were obtained without supplementation. Maximal
rate of TOC removal obtained in these experiments
Table 3. The effect of supplementation on the quality of was 0.54 g l - ' h - t and of biomass production
the final culture filtrate after batch growth of Geotrichum 0.68 g 1- t h - 1 after an arbitrary 40 h incubation.
candidum in spent wash (collected August 1977) Nor was supplementation practised in the experi-
ments carried out on the continuous culture of the
BOD COD TOC
Supplementation (mg 1-1) (mg !- t) (mg 1- t) fungus on spent wash. In a preliminary experiment in
which a fixed dilution rate of 0.125 h -~ was main-
NIL 6000 23,500 7320 tained and temperature varied between 14.5°C and
NH,NOa 6000 26,000 N.D. 40°C the optimum temperature in terms of TOC
KH2PO4 6000 28,500 7320 removal and biomass production, was found to be
UREA 6500 18,000 8360
UREA + KH2PO4 7000 26,000 8880 - 22°C. This was the temperature chosen for the sub-
sequent experiment in which dilution rate was varied.
Table 4. Biomass production, pH, and residual BOD, COD and TOC levels after
batch culture of Geotrichura candidum on samples of unsupplemented spent wash
1976
May 10.6 I0,000 28,000 12,700 6.9
June 21.6 9500 27,000 10,720 6.8
Sept. 16.0 7500 15,000 8580 6.9
Nov. 13.4 9500 25,000 11,380 6.8
Nov. 16.4 9500 25,000 10,810 6.8
1977
Jan. 14.6 7000 20,500 10,180 6.9
Mar. 18.6 3500 20,000 11,440 6.2
May 26.2 5000 14,000 7000 6.9
June . 15.0 10,000 14,500 8640 6.2
Aug. 26.5 6000 23,000 7320 6,8
Sept. 19.5 4500 10,000 N.D. 6.8
Oct. 24.5 8000 19,500 10,040 6.7
550 J. P. QUINt<and R. MARCHANT
Table 6. Analyses from a two stage continuous culture of Geotrichum candtdum on distillery spent wash
o~, Biomass
Dilution increase
rate Biomass in second Filtrate "i, TOC Filtrate % BeD
(h - i) pH (g 1- 1) stage TOC removal BeD removal
It must be emphasized that during this experiment, since this was a non-aseptically operated process it
despite the absence of aseptic precautions, contami- might be preferable if a dilution rate in the second
nation did not occur at any dilution rate. stage were chosen such that the pH of the culture did
Wash-out took place at a dilution rate between not rise so close to neutrality, for at D = 0.085 h -
0.175h -1 and 0.200h -1, while substrate utilization first signs of bacterial contamination were observed.
and biomass production were greatest at the lowest Dilution rates in the two stages of 0.125 h -1 and
dilution rate (0.025 h - 1). A dilution rate of 0.125 h - 1 0.10 h-1 would seem to adequately optimize produc-
however, proved optimal in terms of biomass produc- tivity and substrate utilization respectively.
tivity ( 2 . 2 4 g l - t h -1) and rate of TOC removal Biomass removed from the second fermenter when
( 1 . 5 3 g l - l h - 1 ) ; yield was 1.46g (g C utilized)-1. operated at a dilution rate of 0.125 h -~ was found to
However, at D = 0.125 h -~ BeD, COD and TOC have a "true" protein content (as measured by the
reductions of only 50.0~/~ 30.09/~ and 40.7% respect- Biuret reaction) of 45.5%.
ively were obtained, and it was felt that by the addi-
tion of a second fermenter, coupled in series with the
DISCUSSION
frst, the high productivities and yield associated with
the higher dilution rates could be retained, whilst also The composition of distillery spent wash has
achieving a high concentration of biomass and a low received little detailed attention in the literature,
orgamc loading in the final effluent--features associ- partly because of the variability produced by vari-
ated with the lower dilution rates. ations in malt quality or in the conditions of the
The results produced by the second stage in such a manufacturing process, and partly because it has
two-stage system, in which the dilution rate in the never been regarded as a particular asset by distillers.
first fermenter was maintained at 0.125 h-1 and that The present data indicate that carbohydrate and
in the second varied between 0.085 h-1 and 0.15 h - 1 protein form the bulk of the dissolved organic content
are shown in Table 6 together with the values simul- of spent wash: the wide variations encountered in
taneously attained in the first fermenter, since it their levels is probably an indication of the varying
proved impossible to maintain a steady state in the degrees to which saccharification of the grain polysac-
first fermenter over the entire 300-h duration of the charide can proceed during the malting and mashing
experiment. (The reasons for this are currently under processes. It is anticipated that most of the anthrone-
investigation and are possibly related to shifts in positive material would consist of limit dextrins and
population morphology or a result of the sel~tion of maltotetraose and more complex oligosaccharides,
an improved strain of the organism in the highly while Patel & Ingledew (I973) have shown that wort
selective environment of a high dilution rate con- protein is not used as a carbon source by Saccharo-
tinuous culture regime.) myces carlsberoensis during brewery fermentations.
On the basis of Table 6 it can be seen that with an The pH of tbe liquor is consistently low despite the
overall retention time between the two stages of very considerable buffering effect exerted by its com-
19.75 h it was possible to effect an 87% removal of poaents. The lactate fraction of the total titrable
B e D : this corresponds to a rate of TOC removal of acidity is due to the inevitable contamination of the
1.16 g 1- ~ h - I and to a yield of 1.48 g (g C utilized)- i. fermentation by the rich distillery microfiora. Both
Overall productivity was 1.72 g 1-1 h - 1. However, succinate and glycerol, however, are by-products of
Treatment of whiskey distillery waste 551
yeast metabolism (Oura, 1977) and are formed in Research Studentship awarded to one of us (J.Q.) by the
strictproportion to the amount of ethanol produced. Northern Ireland Department of Education. We also grate-
fully acknowledge the co-operation of the staff of Old
The copper content of the wash--derived from the
Bushmilis Distillery Ltd., and of Mrs. Mary Parr of the
distillery fittings--varies considerably, but does not Northern Ireland Department of Agriculture Freshwater
appear to bc correlated with its p H as might have Biological Investigation Unit who carried out the TOC
been anticipated. analyses.
B O D : P ratios arc low--on average 30:1--~espitc REFERENCES
the extremely high values for B O D and C O D loading
American Public Health Association 0971) Standard
which indicate that malt distilleryspent wash is per- Methods for the Examination of Water and Wastewater.
haps the most concentrated organic waste produced American Public Health Association, Inc., New York.
in the British Isles (cf. Tomlinson, 1976). While tbe Anon (1973) Annual Report of the Banff, Moray and Nairn
disparity between B O D and C O D values shows that River Purification Board, p. 18. Elgin, Morayshire, U.K.
Birch G. G., Parker K. J. & Worgan J. T. (Editors) (1976)
a significantproportion of the organic content is not Food from Waste. Applied Science Publishers, London.
biodegradable, it yet seems one of tbe most appro- Brown D., McKay R. & Weir W. (1976) Some problems
priate substrates with which to conduct a biological associated with the treatment of effluents from malt
treatment process combining removal of organic whisky distilleries. Prog. Wat. Tech. 8, 291-300.
matter with its refixation as high-protein biomass. Campbell W. (1970) Malt whisky distillery waste. Eft.
Water Treat. J. Sept. 1970, 509-513.
The fact that such a process can be carried out by Canelli E., Mitchell D. G. & Pause R. W. (1976) An im-
an organism previously cultivated for human con- proved determination of chemical oxygen demand in
sumption because of its high protein content and water and wastes by a simplified acid dichromate diges-
good harvestability(Ledingham et al., 1946) and that tion. Water Res. 10, 351-355.
Church B. D., Erickson E. E. & Widmer C. M. (1973)
aseptic precautions need not be preserved to maintain Fungal digestion of food processing wastes. Food Tech-
a monoculturc would seem to increase considerably nol. 27, 36-42.
its economic viability. Church B. D., Widmer C. M. & Espinosa R. (1977) Fungal
Even if it is judged purely as a waste treatment conversion of carbohydrate wastes to animal feed pro-
process it would appear to have advantages over the tein-vitamin supplements. Proceedings, 8th National
Symposium on Food Processing Wastes, 355-388, U.S.
evaporation methods currently employed. Effluents of Environmental Protection Agency, Washington, U.S.A.
a strength no greater than "evaporator condensate" Fairbairn N. J. (1953) A modified anthrone reagent. Chem.
could certainly be achieved by a batch or continuous Ind. 72, 86.
two-stage mode of operation, and with the advantage Herbert D., Phipps P. J. & Strange R. E. (1971) Chemical
analysis of microbial cells. In Methods in Microbiology
that they would not require neutralization or nutrient (Edited by Norris J. R. and Ribbons D. W.) Vol. 5B, pp.
supplementation before final treatment by conventio- 209-344. Academic Press, London,
nal biological filtration. Indeed the provision of bio- Jarl K. (1969) Symba yeast process. Food Technol. 23,
mass feedback would almost certainly make a single- 1009-1012.
stage continuous process feasible, and at quite high Ledingham G. A., Oayson D. H. F. & Balls A. K. (1946)
BIOS Final Report. 236, H.M.S.O., London.
dilution rates. Mackel C..I. (1976) A study of the availability of distillery
In terms of removal of organics the results of the by-products to the agricultural industry, 46 pp. Report,
present study compare very favourably with those of School of Agriculture, University of Aberdeen, Scotland.
other workers who have investigated similar pro- Oura E. (1977) Reaction products of yeast fermentations.
Process Biochent April 1977, 19-35.
cesses. Shannon & Stevenson (1975) using eight fungal Patel G. B. & Ingiedew W. M. (1973) Trends in wort
and yeast strains achieved BOD removals from brew- carbohydrate utilization. Appl. Microbiol. 26, 349-353.
ery effluents of 15.5-56.2% after laboratory batch cul- Ranson S. L. (1955) In Moderne Methoden der Pflanzenana-
ture for up to 8 days, while Updegraff et al. (1973) lyse (Edited by Paech K. and Tracey M. V.) Vol. 2, pp.
obtained a TOC reduction of 50-80% in batch and 546--547. Springer Verlag, Berlin.
Shannon L. J. & Stevenson K. E. (1975) Growth of fungi
48% in continuous culture in coffee wastes using a and B.O.D. reduction in selected brewery wastes. J. Food
variety of fungi. Most efficient removals appear to be Sci. 40, 826-829.
those of Church et al. (1973) who achieved 95% BOD Smith M. E. (1975) Ph.D. Thesis University of Kent, Can-
reduction in corn and pea canning wastes using Tri- terbury, U.K.
Smith M. E. & Bull A. T. (1976) Studies on the utilization
choderma viride and Gliocladium deliquescens in a
of coconut water waste for the production of the food
non-aseptic continuous process operated at a dilution yeast Saccharomyces fragilis. J. Appl. Bact. 41, 81-95.
rate of 0.05 h = 1 in 10,000 gallon tanks. Initial BOD Tomlinson E. J. (1976) The production of single-cell pro-
values, however, were relatively low, about tein from strong organic waste waters from the food and
1600 mg 1 - 1 . Interestingly, G. candidum was an impor- drink processing industries. Water Res. 10, 367-376.
Updegraaf D, M., Espinsoa R., Griffin L., King J.,
tant contaminant in the process, occasionally becom- Schneider S. & Rolz C. (1973) The production of animal
ing dominant. feed by the growth of fungi on wastes from the coffee
Intensive studies on optimization of the conditions and rum-distilling industries. Dev. Ind. Microbiol. 14,
for biomass production and on the nutritional value 317-324.
Wilkin G. D. (1977) Products of Scotch whisky synthesis.
of the product are currently being carried out. Process Biochem May 1977. 22-35.
Yemm E. W. & Cocking E. C. 0955) The determination of
Acknowledgements--This work was supported by a amino-acids with ninhydrin. Nature 80, 209-213.