The simulator above plays a run back. Here you are the operator. A 20 L E. coli fed-batch runs in real time, one second of yours is four minutes in the vessel, and as the cells grow their oxygen demand, acid and heat climb with them. Keep dissolved oxygen, pH and temperature in spec with the levers a plant operator has. Same cells and vessel physics as above, with an acid, heat and foam balance added.
The Monod model relates specific growth rate to the limiting substrate: μ = μmax·S/(Ks+S). Coupled with dX/dt = μX and dS/dt = −μX/Yx/s it reproduces the classic batch growth curve. This simulator integrates that system in real time with RK4 and animates the vessel, so you can see how μmax, Ks and yield shape the trajectory.
Batch adds nothing: cells grow until substrate runs out. Fed-batch feeds substrate over time (constant or exponential) so biomass keeps accumulating and volume rises. Chemostat flows medium in and culture out at rate D; at steady state μ = D, and above the critical dilution rate the culture washes out. This simulator runs all three on the same kinetics.
Washout is when cells leave the vessel faster than they grow, so biomass falls to zero. At steady state D = μ, and since μ cannot exceed μmax, raising D above Dcrit = μmax·SF/(Ks+SF) washes the culture out. The productivity curve (D·X) peaks below Dcrit — that peak is the best continuous operating point.
As biomass rises, oxygen uptake (OUR = qO2·X) climbs toward the transfer ceiling and DO falls. When DO drops, growth is scaled by an oxygen Monod term DO/(KO2+DO), so the culture self-limits — the same density ceiling real vessels hit, set by kLa. Predict your kLa with the OTR & kLa estimator.
It is a teaching and exploration tool built on unstructured Monod-family kinetics with representative literature parameters. It captures the qualitative behaviour and trade-offs very well, but it is not a validated organism-specific digital twin — real design needs parameters fitted to your strain and medium and, for cases like E. coli acetate overflow or CHO multi-substrate metabolism, organism-specific models. Use it to build intuition and screen scenarios.
Yes. The operator challenge above runs a 20 L E. coli fed-batch in real time, one second of yours to four minutes in the vessel. You hold DO between 20 and 80% with agitation, air flow and oxygen enrichment, pH at 7.00 ± 0.10 with a base pump, and temperature at 37.0 ± 0.5 °C with the jacket coolant, while oxygen demand, acid and metabolic heat all climb with the biomass. Three shifts build from oxygen alone to a night shift where every loop is manual and a foam surge and a feed pump fault arrive unannounced. The debrief scores how long each loop held spec and compares your final biomass with an automatic controller on the same recipe.
A sudden rise in DO, often called a DO spike, usually means the carbon source has run out. When glucose is exhausted the cells stop growing and their oxygen uptake collapses within minutes while oxygen transfer carries on, so DO shoots up. Operators and feed controllers use that spike as the signal that the batch phase is over and the feed should start. In the operator challenge it happens around 6 h, and DO falls back as soon as the recipe starts feeding.
Antifoams act on the gas-liquid interface, which also slows oxygen transfer. Silicone antifoams lower kLa by roughly 30 to 50% above about 30 ppm, and PPG-based agents by around 15 to 30%, so each dose trades foam control for oxygen transfer capacity. That is why dosing on a foam level alarm beats dosing on a timer. In the operator challenge every dose visibly cuts kLa and the antifoam decays over a couple of hours, so late in a high-density run you are balancing foam against oxygen. More in bioreactor foaming troubleshooting.