The Choice of Suitable Online Analytical Techniques and Data Processing for Monitoring of Bioprocesses - Measurement, Monitoring, Modelling, and Control of Bioprocesses - page 256

variety of bioprocess systems. These findings show the versatility for investigation
of almost all types of cell metabolism, viz. aerobic, fermentative, anaerobic,
anoxic, photoautotrophic and mixotrophic. Changes in the measured heat flow rate
profile were effectively used by several research groups to understand anomalies
encountered in a bioprocess system such as diauxic growth, change in physiology
of organism, dual substrate limitation and substrate and toxic inhibitions [ 19 - 21 ].
Furthermore, the estimation of heat yield coefficients could confirm the existence
of these anomalies, e.g., values of heat yield due to cell growth (Y Q/X ) and oxy-
calorific coefficient (Y Q/O ) could provide quantitative information on diauxic
behaviour and metabolic shifts in an ongoing bioprocess [ 12 ]. The development of
a high-sensitivity biocalorimetry [ 22 ] proved its ability to monitor weakly exo-
thermic biochemical reactions encountered in anaerobic and WWT systems; For
instance, Liu et al. first reported the existence of endothermic microbial growth by
cultivation of the acetotrophic methanogen Methanosarcina barkeri [ 23 ]. Also,
Daverio et al. successfully monitored acidogenic and methanogenic phases of
anaerobic granular sludge originating from an up-flow anaerobic sludge blanket
(UASB) digester [ 24 ]. Calorimetric investigations in micro-algal cultures, viz.
Chlorella vulgaris and Chlorella sorokiniana, were carried out using an improved
'photobiocalorimeter' [ 25 , 26 ]. The heat flow measurements were utilised to
quantify the stored chemical energy (converted from incident light) inside algal
biomass and to estimate photosynthetic efficiency.
Recent studies by the authors proved the robustness of a heat flow signal
compared with process signals acquired in parallel to PAT process analysers such
as dielectric spectroscopy and exhaust gas analysers [ 35 ]. Apart from monitoring,
the measured heat flow rate signal can also serve as an input to control the
bioprocess in order to improve the product yield, e.g. initiating limiting substrate
feed during fed-batch culture and/or the induction phase. This is illustrated in
Fig. 1 , which represents a typical heat flow rate (power-time) profile during
aerobic respiratory growth of Kluyveromyces marxianus. Since there is no fer-
mentative by-product formation during respiratory metabolism, the majority of the
heat generation is from the cell growth process. This phenomenon can be inferred
from Fig. 1 , since the measured heat flow rate signal clearly depicts distinct phases
of cell growth. The logarithmic growth phase of K. marxianus corresponds to an
exponential rise in the heat flow rate, while substrate (glucose) limitation leads to a
dramatic fall in the heat profile at approximately 8 h after inoculation. This shift in
heat profile was used as the signal to start the fed-batch operation in order to
maintain the growth trajectory of K. marxianus in the exponential mode (Fig. 1 ).
A further improvement has been achieved via feedback control employing heat
flow rate measurements and a proposed real-time fed-batch control [ 35 ]. A simple
estimator was developed for biomass and specific growth rate using heat flow rate
measurements, and its reliability was investigated in a fed-batch process in real
time [ 36 ]. The robustness of feedback control to maintain specific growth rate at a
desired set value employing such estimators is shown in Fig. 2 . It can be seen that
the

average tracking error between the controlled and the actual set value

(0.21 h -1 ) of the specific growth rate is 0.03 h -1

over a 5-h period. These results