Ferment Schedule Modeler — Bulk + Shape + Proof Timing for Yeasted Doughs
Most baking schedules read 'bulk 2 hours, proof 1 hour' as if those numbers are physical constants. They are not. Bulk-fermentation duration is a computed function of five inputs — temperature, hydration, salt, leavener type, leavener dosage — with cold retard operating as a ~5x time-scaling operator. This wizard models the math and returns a concrete timeline (mix → bulk → shape → final proof → bake) plus under-proof vs over-proof visual cues at decision moments. Pair with the bread-hydration tool upstream (dough formula) and the sourdough-starter tool (if using levain).
- 1Dough
- 2Leavener
- 3Conditions
- 4Schedule
- 5Cues
Why ferment time is a function, not a fixed interval
Quick answer: yeast and LAB activity scale with temperature via Q10 ~2 (every 5°C drop roughly doubles ferment time), with dough composition modifying the baseline. A recipe that specifies "bulk 2 hours" assumes one specific combination of temperature + hydration + salt + leavener; any drift from that combination invalidates the clock reading. The wizard computes the correct duration at YOUR conditions rather than asking you to trust a fixed number authored for a different kitchen.
The canonical error pattern: a baker reads a summer recipe in winter, follows the same 2-hour bulk, and produces under-proofed bread — not because the recipe was wrong, but because the recipe silently assumed 24°C kitchen when the winter kitchen is 18°C. At 18°C the same dough needs ~3 hours, not 2. The wizard applies Q10 ~2 to whatever ambient temperature you enter.
Temperature is the single largest lever
Quick answer: Q10 ~2 means biological reaction rates roughly double per 10°C rise (or halve per 10°C drop). Between typical kitchen range 18-28°C, a 5°C swing roughly halves or doubles bulk time. Cold retard at 4°C multiplies ambient bulk by ~5x, which is what makes overnight fridge bulks work on ordinary schedules.
| Ambient temperature | Temp factor vs 24°C | Lean dough bulk (1% IDY, 68% hydration) |
|---|---|---|
| 16°C (cool winter kitchen) | ~1.7x slower | ~8.5hr |
| 18°C | ~1.5x slower | ~7.5hr |
| 20°C | ~1.3x slower | ~6.5hr |
| 22°C | ~1.15x slower | ~5.8hr |
| 24°C (industry baseline) | 1.0x | ~5hr |
| 26°C (warm summer kitchen) | ~0.87x faster | ~4.3hr |
| 28°C | ~0.76x faster | ~3.8hr |
| 4°C (refrigerator) | ~5x slower (Q10 + yeast population drop) | ~25hr |
Hydration and salt modify the rate (smaller levers)
Quick answer: higher hydration accelerates ferment via higher water activity; salt retards yeast activity. These are smaller levers than temperature (typically ±15%) but compound with temperature on long schedules. The wizard applies both modifiers to the baseline duration.
| Variable | Value | Factor on bulk time |
|---|---|---|
| Hydration 55% (low) | lean loaf, tight crumb | ~1.13x longer |
| Hydration 68% (baseline) | classic lean bread | 1.0x |
| Hydration 75% | higher-hydration sourdough | ~0.93x faster |
| Hydration 85% | focaccia, ciabatta | ~0.83x faster |
| Salt 1% (low) | slightly flat flavor | ~0.88x faster (salt retards yeast) |
| Salt 2% (baseline) | industry standard | 1.0x |
| Salt 2.5% | slight flavor emphasis | ~1.06x longer |
| Salt 3% (high) | pretzels, some crackers | ~1.12x longer |
Leavener type + dosage — the second-largest lever
Quick answer: leavener choice alone can shift bulk time by 3-4x (instant yeast vs sourdough levain) even at identical temperature. Dosage scales inversely — halving instant yeast roughly 1.8x bulk time (not 2x because yeast viability has a floor). Preferments (poolish, biga) accelerate bulk by 10-15% because they seed the dough with already-active yeast + lactic acid bacteria.
| Leavener | Typical dosage | Relative bulk time @24°C (baseline 1% IDY) |
|---|---|---|
| Instant dry yeast (IDY) | 0.5-2% of flour | 1.0x @ 1%; 1.8x @ 0.5%; 0.7x @ 2% |
| Active dry yeast (ADY) | 0.5-2% of flour | ~1.05x (5-10min hydration lag) |
| Fresh/compressed yeast | 1.5-6% of flour (3x IDY weight) | ~1.0x at equivalent dose |
| Sourdough levain (100% hydration) | 15-25% of flour | ~3.5x (wild yeast + LAB slower than commercial) |
| Poolish preferment (100% hydration, 12-16hr) | 20-40% of flour pre-fermented | ~0.85x (pre-activated yeast seeds dough) |
| Biga preferment (50-60% hydration, 12-18hr) | 20-40% of flour pre-fermented | ~0.90x |
Cold retard is a time-scaling operator, not an independent variable
Quick answer: placing dough at 4°C (refrigerator) multiplies whatever ambient duration you would have at the same temperature by ~5x. This is not "cold slows ferment" as a vague idea — it is Q10 compounded between 4°C and 24°C (factor of 4x) plus a ~25% yeast-population drop at cold onset (additional 1.25x) = effective 5x multiplier. The wizard applies this math to bulk and final proof separately depending on retard plan.
| Retard plan | Bulk stage | Final proof stage | Total duration (lean, 1% IDY, 24°C ambient) |
|---|---|---|---|
| No retard | ~5hr @ 24°C | ~1.25hr @ 24°C | ~6.25hr |
| Cold bulk (overnight fridge bulk) | ~25hr @ 4°C | ~1.25hr @ 24°C | ~26hr |
| Cold final proof (overnight fridge proof) | ~5hr @ 24°C | ~6hr @ 4°C | ~11hr |
| Both cold (long pizza dough) | ~25hr @ 4°C | ~6hr @ 4°C | ~31hr |
Cold retard is used for two reasons: (a) schedule flexibility — a baker can mix Friday evening and bake Saturday morning, (b) flavor development — LAB continues producing lactic and acetic acid at cold while yeast slows, shifting the acid:CO2 balance toward flavor. Classic sourdough uses cold final proof; classic Neapolitan pizza uses both-cold for 24-72hr.
Shape timing — shape at ~65% of bulk-progress, not at bulk end
Quick answer: the textbook instruction "shape after bulk fermentation" is imprecise. Shape at approximately 65% of predicted bulk time (dough at ~75% peak volume, still has gluten structure). Past that, the dough has passed its peak extensibility window and shape will not hold. The wizard surfaces shape time in the schedule timeline.
Earlier shape (50% bulk) gives a tight, dense loaf — under-developed gluten. Later shape (90%+ bulk) gives a slack, puddling dough that will not hold shape during final proof. The 60-70% window is where gluten is developed enough to hold shape but extensible enough to take a final shape without tearing.
What this model does not capture
The Q10 coefficient is empirical — real-world dough temperature diverges from ambient (large batches retain metabolic heat; small batches cool fast). Dough temperature, measured with an instant-read thermometer in the centre of the mass, is the authoritative input; ambient is an approximation. Experienced bakers aim for 24-26°C final dough temperature right after mix rather than letting ambient dictate.
The wizard does not model: (a) autolyse time impact (30-60min pre-mix rest with just flour + water has complex effects on gluten + ferment — slight ferment acceleration), (b) fold-frequency impact (more folds = more gas distribution = slightly shorter bulk), (c) yeast cold-damage from refrigeration beyond 72hr, (d) specific-flour micronutrient effects (whole-grain accelerates ~15%; rye ~20%), (e) altitude effects on CO2 retention (above ~1500m dough proofs faster + may over-proof visually with less actual ferment progress). Use the wizard as prediction + calibrate against 3-5 actual bakes at your specific conditions.
Sources and further reading
Ken Forkish, Flour Water Salt Yeast (Ten Speed Press, 2012) — pp.41-73 ferment timing tables + schedule math for home bakers. Jim Lahey, My Bread (W. W. Norton, 2009) — no-knead ferment kinetics at low yeast dosage. Jeffrey Hamelman, Bread: A Baker's Book of Techniques and Formulas 3rd ed. (Wiley, 2021) — pp.52-89 formula math + ch.6 preferments (poolish, biga, levain-levain). Nathan Myhrvold and Francisco Migoya, Modernist Bread (The Cooking Lab, 2017) — Vol.3 ch.10 fermentation science + Vol.2 ch.5 temperature control. Harold McGee, On Food and Cooking 2nd ed. (Scribner, 2004), ch.10 microbiology of fermentation. For cold-retard schedules: Maurizio Leo, The Perfect Loaf (Penguin Random House, 2022) ch.8 cold-retard sourdough.
Ferment Schedule Modeler Tool v1 · canonical sources cited inline above · runs entirely client-side, no data transmitted