Metabolic heat generation is directly proportional to the oxygen uptake rate (OUR) of the culture. The relationship is Q_met = OUR x V x DeltaH_O2, where DeltaH_O2 is approximately 460 kJ per mole of oxygen consumed (this is the oxycaloric equivalent and is remarkably constant across organisms). For a CHO culture with OUR of 2 mmol/L/h in a 1000L bioreactor, metabolic heat is approximately 256 W. For E. coli at 20 mmol/L/h, this jumps to 2,560 W. OUR can be measured directly from exhaust gas analysis or estimated from cell density and specific oxygen consumption rates.
The Log Mean Temperature Difference (LMTD) is the effective driving force for heat transfer between the process fluid and the cooling jacket. It is calculated as LMTD = (DeltaT1 - DeltaT2) / ln(DeltaT1/DeltaT2), where DeltaT1 = T_process - T_coolant_in and DeltaT2 = T_process - T_coolant_out. The heat removal rate is then Q = U x A x LMTD, where U is the overall heat transfer coefficient and A is the jacket surface area. For a 37C process with 15C coolant inlet and 25C outlet, LMTD is approximately 16.4 K. Using chilled water (4C) increases LMTD significantly, improving cooling capacity without requiring more surface area.
This is one of the fundamental challenges of bioreactor scale-up. Heat generation is proportional to volume (V), which scales as D^3, while cooling surface area (jacket) scales as D^2. As the bioreactor diameter doubles, volume increases 8-fold but surface area only increases 4-fold, creating a surface-area-to-volume mismatch. At small scales (2-10L), the jacket alone provides far more cooling than needed. At 1,000-5,000L, the jacket may become marginal, and internal cooling coils, baffled jackets, or external heat exchangers may be needed. At 10,000L+, sophisticated cooling strategies are essential, especially for high-OUR organisms like E. coli.
The relative contribution depends on organism, agitation intensity, and scale. For mammalian cell culture (low OUR, gentle agitation), agitation heat may represent 20-50% of total heat at large scale. For microbial fermentation (high OUR, vigorous agitation), metabolic heat typically dominates at 60-90% of total. Agitation power scales as P = Np x rho x N^3 x Di^5, meaning it is extremely sensitive to impeller speed (cubed) and diameter (fifth power). Switching from a Rushton turbine (Np=5) to a pitched blade (Np=1.3) dramatically reduces agitation heat while maintaining adequate mixing if properly designed.
Consider internal cooling coils when the jacket cooling capacity safety margin falls below 1.5x the total heat generation. This typically occurs at scales above 500-2000L for microbial fermentation, or above 5,000-10,000L for mammalian cell culture. Internal coils provide additional heat transfer surface area with high U values (600-1000 W/m2K due to direct contact with agitated fluid). However, coils reduce working volume, can create cleaning challenges, increase contamination risk, and may interfere with impeller flow patterns. Alternative strategies include using chilled water, dimple or half-pipe jackets for higher U values, or external recirculation heat exchangers.
The overall heat transfer coefficient depends on vessel material, jacket design, fluid viscosity, and agitation intensity. Typical values are: glass vessels with simple jacket: 100-200 W/m2K; stainless steel with standard jacket: 300-500 W/m2K; stainless steel with dimple jacket: 500-700 W/m2K; stainless steel with half-pipe jacket: 400-600 W/m2K; internal coils: 600-1000 W/m2K. These values decrease with increasing fluid viscosity and decreasing agitation speed. Fouling over time can reduce U by 10-30%. For conservative design, use the lower end of the range for your configuration.