COMPUTERIZED ESTIMATION OF BACTERIAL HEAT

PRODUCTION IN A LABORATORY FERMENTER

B. L. Walsh1, J. C. Chow2, E. L. Watson3, and P. M. Townsley4

department of Microbiology, University of British Columbia, Vancouver, B.C. V6T 1W5, department ofAgricultural Engineering, University of California, Davis, Calif., 'Department of Bio-Resource Engineering,

University ofB.C., Vancouver, B.C., and ^Department of Food Science, University of B.C., Vancouver, B.C.

Received 3 August 1978

Walsh, B. L., J. C. Chow, E. L. Watson, and P. M. Townsley. 1980.production in a laboratory fermenter. Can. Agric. Eng. 22: 93-95.

Computerized estimation of bacterial heat

A Dewar flask serves as a laboratory fermenter capable of supporting bacterial growth over the temperature rangefrom ambient to greater than 70°C. These thermophilic temperatures are generated by the growing bacteria, and datafrom the temperature-time curves can be fitted into a computerized program developed for the apparatus to estimate therate of bacterial heat generation and the cumulative heat generated by the microbial activity.

INTRODUCTION

Available methods for estimating theheat production on a dynamic basis inlaboratory fermentations tend to be tooexpensiveor too elaborate (Skinner 1969;Wadso 1974). Very few reports areavailable using simple techniques(Cooneyetal. 1968).

A heat balance equation is used todescribe microbial heat production in alaboratory fermenter. Such a description,utilizing small-scale, inexpensivelaboratory equipment, is necessary tofacilitate investigations of microorganisms responsible for the heat generated during composting (Poincelot 1972)and the thermophilic processing of liquidwastes (Popel and Ohnmacht 1972;Coulthard and Hendren 1973) as performed in large pilot plant fermenters.

Aerobic bacteria growing in thenutrient broth contained within the

fermenter require both a constant airsupply and agitation. The temperature ofthe fermenting broth varies with timeaccording to the activity of the bacteria.By recording and analyzing thistemperature-time curve, the dynamic heatproduction by these bacteria can beestimated and plotted as heat productionrate versus time and as cumulative heat

production versus time.

MATERIALS AND METHODS

Growth ApparatusA 4-L Dewar flask (Fisher Sci.) served

as the fermenter. Heat loss was reduced

by covering the top of the fermenter witha 2-cm thick plexiglas lid. An operatingvolume of 2 L was agitated vigorouslywith a 5000 rpm, 1/100 HP electric motor(Little Giant Pump Co.), driving a3-bladed propeller of 4.5-cm diameterattached to a stainless steel shaft 35 cm

long. Air was supplied by a Bell andGosset Duraire pump (ITT, Monroe, La.)

connected by tygon tubing to a flow meter(R. Gilmont Co.) and humidified bybubbling through water at room temperature. The humidified air was dispersedinto the broth through a porous stonesparger located below the stirrer. Aschematic diagram of the apparatus ispresented in Fig. 1.

Organisms and SubstrateExperiments were conducted using

Trypticase Soy Broth (BBL (30 g/L)) asthe growth substrate. A mixture oforganisms contained in thermophilicallyfermented dairy cattle manure was usedas inoculum into the broth at ambient

temperature (23°C).

Heat Measurement

Temperatures of both the stirred brothand the room air were simultaneouslyrecorded using copper-constantanthemocouples and a Riken Denshi

RECORDER

10-inch (25.4-cm) strip chart recorderagainst the reference junction located inan ice bath. The temperature changesrecorded in the broth were a function of

the microbial activity. The shape of thetemperature-time curve thus obtaineddepends on the following transient stateequation.

<? = ,Q + UA (T-TJ) + flevap ,..at) (1)

where

q

*?motor

T

7oc

dT

+ <?s, <7motor

heat production rate of thebacteria

heat rate of evaporationwater

sensible heat rate of the bubbl

ing airheat rate of the stirrer

temperature of the brothambient temperatureslope of the temperature-time

STIRRER

THERMOCOUPLE

PLEXIGLASS

PUMP

HUMIDIFIER

STONE AERATOR

DEWAR FLASK

NUTRIENT BROTH

AND BACTERIA

of

Figure 1. Schematic diagram of fermentation apparatus for determining heat production bybacteria growing in liquid culture over mesophilic and thermophilic temperatureranges.

CANADIAN AGRICULTURAL ENGINEERING, VOL. 22, NO. 1, JUNE 1980 93

dQ curve

U = overall thermal conductance

A = surface area

M = mass of the systemC = specific heat of the system.

A wet- and dry-bulb psychrometer wasused to determine qevap and qsen at theoptimum aeration rate of 0.1 L air/mininto the broth at 60°C. It was found that

less than 2 1 12% error was introduced byneglecting qsen and qevap. Thus, Eq. 1reduces to:

MCdT

dB+ UA(T-TJ)-q„ (2)

To use Eq. 2, the thermal capacity(MC), the heat from the motor (<7raot0r)>and the coefficient ofconvective heat loss

(UA), had to be determined. The firstparameter, MC, was calculated by theequation:

(MC)system = (MC)broth + (A/C)glass

= (2kg)(4.19kJ/kg-C)

(0.838 kJ/kg •C (.60)

= 8.91 kJ/C

(3)

♦e?*)

This calculation assumes that 60% of the

glass was in contact with the liquid.The second parameter, ^Woo was

determined by recording the temperature-time heating curve of aerated,stirred, sterile broth in the fermenter. Itwas found that the broth was heated by themechanical stirrer from 23CC to 40°C in

24 h. Eq. 2 was used to calculate qmotOT.Because the broth was sterile, q = 0.(MC)system was assumed to be 8.91 kJ/C.The slope of the temperature-time curve,dT/dQ, was read at several temperatures.An estimated value of UA was inserted

into Eq. 2 for each value of dT/dQ and T.This estimation was repeated until thecalculated qmolor was constant for allvalues of dT/dQ and T.

The third parameter, UA, was determined by heating the sterile brothelectrically to 65°C. The heating elementwas removed and the temperature-timecooling curve recorded while the brothcooled to 40°C (about 8 h). Stirring andsparging were continued during cooling.Soap was added to the broth to representthe foam that occurred due to bacterial

growth. Values of UA were calculated,using Eq. 2, and the slope of thetemperature-time curves at several temperatures. The results were averaged.

An IBM 370 computer was used tosmooth the data by fitting thetemperature-time points to a cubic splinefunction. A series of slopes, dT/dQ, werecalculated from the resulting equation.

94

RESULTS

Two runs were performed to determineqmotor- The motor ran at constant speedwith an 80-W electrical input. Theaverage <7mot0r = 2.75 ± 0.25 W. Theaverage coefficient of convective heatloss (UA) value of the system wascalculated from four runs. The averageUA = 0.30 ± 0.05 W/C.

The dynamic heat production of abacterial growth experiment was calculated using Eq. 2, and the temperature-time curve, and the measured values of<7motor and UA. Temperature and timewere read from the strip chart record andentered on computer cards. The computerprogram used these data, the gmotor> UAand MC previously calculated, anddetermined the heat production rate andthe cumulative heat production due to thegrowth and metabolism of the bacteria.

The computer plots of temperature,cumulative heat production and ofheating rate, each against time, arepresented in Fig. 2.

DISCUSSION

A temperature exceeding 60°C wasproduced in the laboratory fermenterwithin 24 h (Fig. 2). Temperatures haveexceeded 70°C in other experiments(unpublished observations). Similar risesin temperature within 48 h have beenreported for solids composting of moistened straw (Norman et al. 1941; Carlyleand Norman 1941), and of refuse (Niese1963) using Dewar flasks as the fermentation vessels. These and similar studies

(Rothbaum 1961) on the heat output ofthermophilic bacteria growing on woolhave used adiabatic apparatus consistingof a culture vessel contained within an

incubator. The latter requires preciseelectronic monitoring and control tomaintain its temperature precisely withthat of the culture vessel.

There is little information concerningthe production of high temperatures bybacteria growing in liquid culture. Moststudies on the heat output utilize anexpensive microcalorimeter (Skinner

500

400-

300

;200

olCO

- 12

^O64

/"c

—-«v

O

/)

i6 /

/0

/ // °

4 8 / // °0/ /

' /40

/ a 0.

/ .' O A -

32 / A / A

/ • / 'A -

/ a O A'

24 X i 1 A

O A a'

.J /

"\6 / /' A

- (\ O A

0

QJ

E

:'V t<^_. ... 1 1 1 1

A'

0 '

8 12 16

Time ( Hr b )

10

. 6

4 o

2 „

Figure 2. Temperature cumulative heat production and rate of heat production by bacteriagrowing in 2 L of Trypticase Soy broth in an insulated fermentation vessel: • •,temperature (I°C); o o. cumulative heat produced (Kcal); A .... A, rateof heat production, (Kcal/h).

CANADIAN AGRICULTURAL ENGINEERING, VOL. 22, NO. 1, JUNE 1980

1969; Wadso 1974) or are performed invessels whose heat loss is so great thathigh (thermophilic) temperatures are notreached (Cooney et al. 1968).

The cumulative heat production produced by the bacteria growing over thewide temperature range discussed aboveis also shown in Fig. 2. Approximately420 kJ of heat were produced within thefirst 16 h of incubation. The cumulative

heat production exceeded 544 kJ by 24 h.This figure represents a typical fermentation pattern corresponding to the so-called exponential growth equation (Aibaetal. 1965).

It is evident that the maximum rate of

heat production by the growing bacteriawas achieved during the period 10-12 hafter the start of the experiment (Fig. 2).This maximum rate of approximately40.6 kJ/h (5.65 X 10-3W/mL) wasreached when the temperature of thefermenter was in the range of 48°-54°C.The heating rate decreased after 12 h, andreached a minimum of 10.5 kJ/h (14.7 x10-4W/mL) about 17-18 h after the startof the experiment. Although the heatingrate of the bacteria decreased during thistime period, the temperature of thefermenter rose to about 63°C. At 17.5 h,additional substrate (TSB) was added tothe fermenter, resulting in an immediateincrease in the heating rate of the bacteriaand an increase in temperature to 65°Cwithin the next 6.5 h. Measuring theheating rate is probably the easiest meansof detecting the immediate effects ofadditions of substrates or inhibitors on the

activity of the bacteria under study(Cooney et al. 1968; Walsh andTownsley, in prep.).

Although cell numbers were notdetermined for this particular experiment,

similar experiments (Walsh andTownsley, in prep.) have indicatedbacterial numbers in the range of0.45-3.4 x 108colony-forming units permL (cfu/mL) at temperatures from 50° to54°C. Assuming a bacterial count of 1.0x 108 cfu/mL, calculations from theabove data would indicate maximum heat

outputs around 56.5 x 10~12 W/cfu in thesame temperature range. This value fallswithin the following ranges of bacterialheat outputs taken from the literature:

2.2-168 x lO"12 W/cfu calculatedfrom data on different species ofbacteria growing in milk (Berridge etal. 1974).6.7-3977 x 10"12W/cell for a mixtureof organisms growing on wheat strawCarlyle and Norman 1941).Alternatively, knowledge of the heat

ing rates can allow calculations todetermine close approximations of thenumbers of viable bacteria at given timesor temperatures.

Although further refinement is neededfor more precise work, the apparatusdescribed does provide a practical meansof studying the organisms involved incomposting or in thermophilic fermentations. The computer evaluation of heatingrates allows study on the effects ofparticular substrates, inhibitors andgrowth conditions (Walsh and Townsley,in prep.) on the achievement of thermophilic temperatures by microorganisms.

ACKNOWLEDGMENT

Acknowledgment is given to the financialassistance provided by Agriculture CanadaGrant No. 06-3023.

AIBA, S., A. E. HUMPHREY, and N. F.MILLIS. 1965. In Biochemical engineer

CANADIAN AGRICULTURAL ENGINEERING, VOL. 22, NO. 1, JUNE 1980

ing, Chap. 4, Kinetics. Academic Press,New York, N.Y. 330 pp.

BERRIDGE, N. J., C. M. COUSINS, and A.J. CLIFFE. 1974. Microcalorimetryapplied to certain species of bacteriagrowing in sterilized separated milk. J.Dairy Res. 41: 203-215.

CARLYLE, R. E. and A. G. NORMAN.1941. Microbial thermogenesis in thedecomposition of plant materials. II.Factors involved. J. Bacteriol. 41: 699-

-724.

COONEY, C. L., D. I. C. WANG, and R. I.MATELES. 1968. Measurement of heat

evolution and correlation with oxygenconsumption during microbial growth.Biotechnol. Bioeng. II. pp. 269-281.

COULTHARD, T. L. and G. A. HENDREN.1973. Pilot plant studies on the thermophilic bacterial treatment of animalwastes. Paper presented at the CSAEAnnual Meeting, 21 Aug., Victoria, B.C.

NIESE, G. 1963. Experiments to determinethe degree of decomposition of refusecompost by its self-heating capacity. Int.Res. Group on Refuse Disposal. Bull. No.17.

NORMAN, A. G., L. A. RICHARDS, andR.E. CARLYLE. 1941. Microbial ther

mogenesis in the decomposition of plantmaterials. I. An adiabatic fermentation

apparatus. J. Bacteriol. 41: 689-697.POINCELOT, R. P. 1972. The biochemistry

and methodology of composting. Conn.Agric. Exp. Sta. Bull. No. 727.

POPEL, F. and C. H. OHNMACHT. 1972.Thermophilic bacterial oxidation of highlyconcentrated substrates. Water Res. 6:

807-815.

ROTHBAUM, H. P. 1961. Heat output ofthermophiles growing on wool. J. Bacteriol. 81: 165-171.

SKINNER, H. A. 1969. Theory, scope andaccuracy of calorimetric measurements. InH. D. Brown, ed. Biochemical mic

rocalorimetry. Academic Press, New York,N.Y. pp. 1-32.

WADSO, I. 1974. A microcalorimeter forbiological analysis. Sci. Tools 21: 18-21.

95