Core Formulas

How to Calculate Enzyme Yield, Recovery, Overfill, and Drying Load

The five core bioprocess calculations every enzyme engineer runs, worked end to end in real activity units and mass.

Industrial enzyme production is a chain of conversions from grams of substrate to active units in the drum. The metrics you must compute are fermentation yield on substrate (Yp/s), volumetric productivity, overall downstream recovery, potency overfill, and the energy to dry the concentrate. Start by fixing units. Enzyme strength is reported as titer in activity units per volume, for example 12,000 LU/mL, while substrate feed is tracked in kg of glucose equivalents. Getting the unit basis wrong is the single largest source of yield error, so lock the activity assay method and its reference standard before any math begins.

Fermentation yield on substrate Yp/s equals total product formed divided by total substrate consumed. If a 100,000 L batch reaches a titer of 40 g/L of secreted protein, that is 4,000 kg of enzyme. If the run consumed 68,000 kg of glucose equivalents from batch and feed, Yp/s = 4,000 / 68,000 = 0.059 g/g, or 59 g product per kg sugar. Volumetric productivity divides titer by fermentation time: 40 g/L over 120 h equals 0.33 g/L/h. The Fermentation Yield calculator handles the substrate bookkeeping across initial charge, fed-batch carbon, and evaporation losses.

Overall downstream recovery is the product of each step yield. A route of centrifugation (0.95), ultrafiltration (0.92), polish filtration (0.90), and formulation (0.98) gives 0.95 x 0.92 x 0.90 x 0.98 = 0.77, so 77 percent of fermenter activity reaches the bulk. On 4,000 kg of protein carrying 100 million total activity units, that is 77 million units recovered. Use the Downstream Recovery calculator to chain step yields. A single step falling from 0.92 to 0.85 pulls overall recovery from 0.77 to 0.71, a 7.8 percent relative loss you can trace to one unit operation.

Products sell on guaranteed activity, so you overfill to cover assay variability and shelf-life decay. If label claim is 5,000 GAU/g and you must guarantee it through 12 months with a 6 percent activity decay and a 3 percent assay CV at 95 percent confidence, the fill target = 5,000 / (1 - 0.06) x (1 + 1.645 x 0.03) = 5,319 x 1.049 = 5,580 GAU/g, an 11.6 percent overfill. The Potency Overfill calculator converts your measured decay rate and assay CV into that fill target, so you neither give away product nor risk failing label claim at expiry.

Drying dominates energy in most trains. Water removed = feed mass x (1 minus the ratio of inlet solids to outlet solids). To take 1,000 kg of 20 percent solids concentrate to 95 percent solids, final mass = 200 / 0.95 = 210 kg, so water removed = 790 kg. At a specific energy of 3.2 MJ per kg of water for spray drying, that is 2,528 MJ, roughly 702 kWh thermal per batch of that stream. The Drying Energy calculator compares evaporation load across inlet temperatures and solids targets, so you can weigh a higher UF concentration step against dryer duty.

Sizing filters uses flux and area. If required throughput is 100,000 L per batch over an 8 h window, that is 12,500 L/h. At a design flux of 50 LMH, or liters per square meter per hour, area = 12,500 / 50 = 250 m2. Add a 20 percent fouling margin and specify 300 m2. The Filtration Capacity calculator solves for area, time, or flux when you fix the other two, and flags when flux decay across the cycle pushes effective throughput past the batch window, which is where undersized membranes stall a schedule.

Tie the numbers together in sequence. Start with Yp/s to get kg of protein, apply Downstream Recovery for units in bulk, size Filtration Capacity and Drying Energy for the process train, then apply Potency Overfill to hit label claim. Carry activity units, not just mass, through every step, because specific activity shifts as impurities are removed in purification. A model that tracks both mass and units end to end will reconcile within 2 to 3 percent of the plant balance. A larger gap almost always means an assay basis mismatch or a unit conversion error upstream, not a real process loss.

Published 2026-07-02.