🍞 Shelf Life, Staling and Preservation

A loaf that has gone hard by the second day is usually not dry. It is stale, and staling and drying are two different faults with two different cures. Confuse them and you will spend money on better bags that fix nothing.

Staling is chiefly starch retrogradation. In the oven, starch granules take up water and swell, with gelatinisation beginning around 55 to 65 degrees Celsius, and the crumb structure finally sets when the loaf core reaches 91 to 93 degrees Celsius. From the moment the loaf leaves the oven, that gelatinised starch begins to recrystallise, mainly the amylopectin fraction. The crumb firms up, turns short and crumbly, and loses its spring. At the same time moisture migrates out of the crumb and into the crust, so a crust that came out crisp goes leathery while the inside goes hard. A loaf sealed in a perfectly airtight bag, losing not one gram of water, still stales. That single fact is the whole lesson: wrapping cannot stop staling, it can only stop drying.

Drying is the simpler fault. Water leaves the loaf into the air, the loaf gets lighter, and the crust hardens. Unlike staling, you can weigh it.

Worked example. Take two loaves from the same batch, both 750 g at wrapping. Wrap one, leave one open on the rack.

1. Wrapped loaf on day two weighs 741 g. Weight loss = (750 - 741) divided by 750, times 100 = 1.2%.
2. Open loaf on day two weighs 705 g. Weight loss = (750 - 705) divided by 750, times 100 = 6.0%.
3. Squeeze both. If the wrapped loaf is clearly firmer than yesterday while having lost only 1.2% of its weight, that firmness is retrogradation, not water loss.
4. The 4.8 percentage-point gap between the two loaves is what your bag is actually buying you.

That four-line test costs one loaf and a kitchen scale, and it stops the argument in your bakery about whether the bags are working.

On temperature, be careful. Retrogradation is fastest somewhere above freezing and below gelatinisation temperature, and freezing largely arrests it, so a freezer is a real storage tool. The widely repeated trade rule that a refrigerator stales bread faster than a warm shelf is not verified in the source material behind this course, so do not attach a rate or a number to it. Teach and use the mechanism: freeze bread if you must hold it, and do not assume a fridge is helping you.

What you do control is formulation and handling. Fat retards drying out in doughs and batters. Sugar is hygroscopic and improves shelf life. Full-fat milk powder extends shelf life, while skim milk powder loses that fat-derived benefit. A dough temperature in the 24 to 28 degrees Celsius band is listed as the optimum for keeping quality, and an improperly high dough temperature is a named cause of poor keeping. Wrapping above 35 degrees Celsius is another named cause, and typical cooling to that point takes two to three hours. Overbaking and a cool oven are both on the same fault list.

A named example. A bakery in Tamale, northern Ghana, selling a 750 g loaf, raised fat from 2% to 4% on flour to slow drying. On a 50 kg flour batch that is one extra kilogram of fat. At an indicative USD 2.20 per kg, verify locally, that is USD 2.20 per batch. With a total formula percentage of 176, a 50 kg flour batch gives 88 kg of dough, and at 800 g scaled that is 110 loaves. Extra cost per loaf = 2.20 divided by 110 = USD 0.02. Two cents a loaf to buy a softer crumb on day two is a decision you can make on the spot, and you now know how to price it.

Staling is not a moral failure or a bad flour. It is starch doing what starch does. Your job is to slow it, price it, and sell the bread before it wins.

Two products can hold the same amount of water and spoil at completely different speeds. What decides spoilage is not how much water is present but how much of it is available to micro-organisms. That availability is called water activity, written a-w, and it runs from 0 to 1.0.

Here are the measured values that matter to a small bakery. Fresh pan bread sits at a-w 0.93. Batter cakes sit at 0.81. Soda crackers sit at 0.30. Now set those against what can grow. Most bacteria are inhibited below a-w 0.91, though some have been found as low as 0.75. Yeasts and moulds can grow down to 0.60, and in bakery products mould onset is put at 0.6 and upwards, with product below that generally mould-free. Nothing at all grows below 0.60. For shelf-stable foods that rely on water activity for safety, a threshold of 0.85 or less is cited.

Read those two lists together and the conclusion is uncomfortable but true. Fresh bread at 0.93 sits above the mould threshold and inside the bacterial range. Bread is not microbiologically stable. Its safety comes from being eaten quickly, and its keeping quality comes from hygiene, packaging and preservatives, never from its water activity. Cake at 0.81 is meaningfully safer. Biscuit at 0.30 is in a different world altogether. That is why biscuits, not bread, are the natural product for a bakery with no cold chain and no daily distribution, and it is one of the most commercially important facts in this whole course.

Mould itself is a post-bake problem. Baking destroys vegetative mould, so every spore on a finished loaf arrived after the oven. The listed causes of mouldy bread are unsanitary equipment, contaminated packaging and wrappers, mouldy racks and tools, and dust exposure. Not one of them is a formulation fault. Controls follow directly: cool on clean racks in clean air, never wrap above 35 degrees Celsius because a warm loaf condenses moisture inside the bag and makes an incubator, keep your wrapper stock covered and off the floor, clean the slicer daily because it touches every loaf you sell, and remove mouldy product from the building rather than leaving it near production.

Worked example. A baker in Livingstone, Zambia, supplies four rural shops about 60 km out, delivering twice a week. She sends 200 loaves a week and gets 36 back unsold and starting to mould.

1. Return rate = 36 divided by 200, times 100 = 18%.
2. At a full cost of USD 0.45 per loaf, indicative, verify locally, the weekly loss = 36 x 0.45 = USD 16.20.
3. Annual loss = 16.20 x 52 = USD 842.
4. She converts half that volume, 100 loaves a week, into a hard sweet biscuit at a-w around 0.30, which does not mould on a shop shelf in three days.
5. If biscuit returns fall to 2%, her weekly biscuit loss is 100 x 0.02 x 0.45 = USD 0.90, against USD 8.10 before. Saving on that half of the route = USD 7.20 a week, USD 374 a year.

She did not buy anything. She changed which product travels 60 km.

One firm warning. Water activity values for buns, scones, pastry and cream or custard fillings were not retrieved for this course, and filled products are the highest-risk line a small bakery can run. Do not guess an a-w for a cream-filled item and do not assume it behaves like the cake it is baked from. If you intend to sell filled products, get water activity measured, or keep them on a same-day, cold-held, short-distance basis only.

The practical habit to build: for every product you sell, write down where it sits on the water activity ladder, and let that decide how far it travels and how long it may sit.

Rope is the spoilage disease that ruins a small bakery in a hot month, and most bakers meet it before they can name it. It is caused by spore-forming Bacillus, chiefly Bacillus subtilis and Bacillus mesentericus, with B. licheniformis, B. megaterium and B. cereus as secondary causes.

The uncomfortable part is where the spores come from and why the oven does not save you. They arrive in the flour, and they withstand up to 130 degrees Celsius, while your loaf core only reaches 90 to 97 degrees Celsius. Water boils at 100 degrees, so the core of a loaf can never exceed 100 degrees at ordinary pressure. The spores therefore go into the oven alive and come out alive. They then germinate in the cooling loaf and enzymatically digest the crumb from the inside.

How it presents. First comes a sweetish, fruity, faintly rotten smell in the crumb, detectable at 12 to 24 hours, or after several days of storage. Then the crumb discolours and goes sticky. Tear the loaf slowly apart and thread-like strands appear between the two halves, like fine spider webbing. That thread test is the confirmation. Once you have seen it once you never mistake it.

The conditions that favour rope are exactly the conditions in a warm-climate bakery. Crumb moisture of 35 to 42%. Crumb water activity of 0.90 to 0.95, which is precisely where ordinary fresh bread sits. Optimal germination at around pH 6.0, with germination still possible down to about pH 5.4 under heavy contamination and warm storage. And warmth, with the disease markedly more frequent in hot seasons. This is not an exotic risk for African bakeries. It is a first-order one.

Control works on three fronts, and the first is chemistry rather than cleaning.

1. Acidify. Formulate to pH 5.4 or below, or below 4.6 through fermentation. Sourdough and acidifying agents containing acetic acid are effective, as are acetic acid salts such as sodium diacetate and calcium acetate. The sourced dose for an acidity regulator in hot months is 0.2 to 0.3% on total flour. Propionic acid or monocalcium phosphate is given at 0.1 to 0.5%.
2. Cool fast and never stack. Stacking loaves during cooling traps warmth and moisture, which is the rope incubator. Cool in a single layer on open racks with air moving.
3. Break the reseeding loop. Disinfect slicing equipment with a 2% vinegar solution, and eliminate contaminated leftover bread from the production cycle entirely. Returned ropey bread brought back into the bakery reseeds everything you bake next week.

Worked example. A bakery in Kano, northern Nigeria, loses batches every March and April. The owner mixes 75 kg of flour a day.

1. Acidity regulator at 0.2% of total flour = 75 x 0.002 = 0.150 kg = 150 g per day.
2. At the upper end, 0.3% = 75 x 0.003 = 0.225 kg = 225 g per day.
3. At an indicative USD 3.00 per kg, verify locally, that is USD 0.45 to USD 0.68 a day.
4. His daily output from 75 kg flour, at a total formula of 176 and 800 g scaled pieces, is 75 x 1.76 = 132 kg dough, divided by 0.8 = 165 loaves. Added cost per loaf = 0.68 divided by 165 = USD 0.004, less than half a cent.
5. One ruined 165-loaf day at a selling price of USD 0.90 costs him USD 148.50. The treatment costs him USD 0.68.

For the slicer, a 2% vinegar solution means 2 parts vinegar in 100 parts of solution. To make 5 litres: 100 mL of vinegar topped up to 5,000 mL with clean water. Wipe the blades and the crumb tray at the end of every day, not at the end of every week.

Rope is the clearest case in baking where a problem is solved by lowering pH rather than by scrubbing harder, though scrubbing still matters, because contaminated bread coming back through the door starts it all again.

Packaging is the last thing that happens to your bread and the first thing your customer judges. Done badly it actively shortens shelf life, so it deserves a written routine rather than a habit.

Start with cooling, because wrapping is downstream of it. Do not slice bread above 35 degrees Celsius, because it caves in and the slicer tears it. Do not wrap above 35 degrees Celsius, because poor flavour and poor keeping quality follow. Typical cooling to that point takes two to three hours. And do not stack loaves during cooling, because stacking traps moisture and warmth and favours both rope and mould. Cool in a single layer, on clean racks, in moving air.

That three-hour window is a production planning fact, not an inconvenience. If your delivery leaves at 06:00, your last loaf must be out of the oven by 03:00. Bakeries that wrap warm bread are almost always bakeries that scheduled the bake too late, and they pay for it in returns.

Now the wrapper itself. Contaminated packaging and wrappers are a named cause of mouldy bread, alongside unsanitary equipment, mouldy racks and tools, and dust exposure. So bags are food-contact material. Store them sealed, off the floor, away from flour dust, and never under the mixer. Do not reuse bags returned from shops.

Worked example. A bakery in Blantyre, Malawi, sells 110 loaves a batch and has been selling unwrapped over a counter, with 12% going stale unsold.

1. Bag cost, indicative USD 0.03 each, verify locally. Label cost USD 0.01. Packaging per loaf = USD 0.04.
2. Packaging per batch = 110 x 0.04 = USD 4.40.
3. Current losses = 110 x 0.12 = 13.2 loaves at a selling price of USD 0.90 = USD 11.88 per batch.
4. If wrapping halves stale returns to 6%, losses become 110 x 0.06 x 0.90 = USD 5.94.
5. Saving = 11.88 minus 5.94 = USD 5.94 per batch, against a packaging cost of USD 4.40. Net gain USD 1.54 per batch, about USD 40 a month on a six-day week.

That margin is thin, which is the honest answer: wrapping pays here, but only just, and only if it really does halve returns. Run the arithmetic with your own numbers before you commit, and measure the return rate for a month afterwards to check the assumption held.

Date marking next. Under the Codex labelling standard, a date of minimum durability must be declared, showing when the product remains fully marketable and retains its stated qualities, using day and month format where durability is under three months and month and year otherwise. Special storage conditions must be declared if the validity of the date depends on them. Codex does exempt fresh bakery items from date marking, but this is a Codex provision, not a right. Many countries do not grant that exemption, or grant it only for unwrapped product sold on the day of baking. Wrapped, dated, shelf-stable bread almost always requires a date mark. Ask your own national authority before you rely on the exemption.

Finally, traceability, which costs a pen. Record which flour lot, meaning miller, batch number and delivery date, went into which production date, and which production date went to which customer or outlet. Two columns in a book. Without that link, a recall is impossible, and a bakery that cannot recall is a bakery betting its licence on nothing going wrong.

Build the routine as a checklist on the wall:

1. Out of oven, single layer, clean rack, air moving.
2. Probe a loaf. Below 35 degrees Celsius before slicing or wrapping.
3. Bags from a sealed box, taken with clean hands.
4. Date mark applied, storage instruction printed.
5. Flour lot and production date written in the traceability book with the outlet.

This lesson has two halves, and the second half is more important than the first. The first half is what preservatives do. The second half is why this course will not tell you how much to use.

Calcium propionate is the industry-standard bread mould inhibitor. It inhibits mould and it inhibits ropy bacteria, which makes it useful against both of the spoilage problems in this module. Crucially it has little effect on yeast and does not interfere with fermentation, which is why it, rather than sorbate or benzoate, became the bread preservative. It contains about 21 g of calcium per 100 g of product. Typical bakery usage for technical effect is 0.1 to 0.3% on flour weight, with higher levels described as not uncommon, and one straight-dough source gives 0.25% at 65% water absorption and 0.40% at 67% hydration.

The practical trap is pH. Propionate works best below pH 5.5. A well-fermented dough falls from about pH 5.3 fresh to about 4.5 after fermentation, which is comfortably in range. A fast, lightly fermented dough may never get there, the preservative then underperforms, and the baker blames the supplier. If your propionate seems not to work, check your fermentation before you change supplier.

Now the second half, and read it carefully.

This course cannot tell you a legal dose, and no honest course could from these sources. The Codex GSFA maximum levels for propionates in bakery wares were not retrieved. All African national preservative limits were not retrieved. The only maxima available are European figures, and even those are marked unverified here, because the source cites a regulation number that appears to contain a typographical error and the figures were not independently checked. They are in any case European numbers with no force anywhere in Africa. Presenting them to you as a dose would be inventing a law.

So the core instruction of this lesson is procedural, not numerical: before you use any preservative in a product you sell, get the permitted list and the maximum level from your own national regulator, in writing.

Here is how to do it, and it takes an afternoon.

1. Identify your national food authority. It is usually a national food and drugs authority, a bureau of standards, or the food safety unit of the health ministry.
2. Write or email one short letter naming your product, for example wrapped white pan bread, 750 g.
3. Ask four questions. Which preservatives am I permitted to use in this product? What is the maximum level for each, and is it expressed on flour or on the finished product? Must the preservative be declared on the label, and in what wording? Is any registration or approval required before I sell?
4. Ask for the answer in writing, or for the clause and document number so you can obtain it yourself.
5. File the reply. Show it to an inspector rather than arguing from memory.

A named example. A baker in Kigali, Rwanda, buys a preservative blend from a wholesaler who tells her verbally that 0.5% is normal. She writes to the national standards body and gets a clause reference back. Whatever the number turns out to be, she now has two things her competitor does not: a defensible file and the correct unit basis, because a limit expressed on finished product is not the same as one expressed on flour. On a 176 total formula, 0.3% on flour is only 0.3 divided by 1.76, about 0.17% on dough, and a baker who confuses the two can be over the limit while believing they are under it. That conversion, which is simple arithmetic, is where small bakeries get caught.

Meanwhile, everything in Lessons 1 to 4 of this module still works and none of it needs a permit. Cool fast, do not stack, wrap below 35 degrees Celsius, clean the slicer, acidify against rope, keep wrappers clean, and pick products whose water activity suits your route. Preservatives are the last lever, not the first.

Most small bakeries pick a shelf life the way they pick a price: by copying whoever is next door. That works until the day a customer is ill, an inspector asks how the number was arrived at, or a shop returns forty loaves and blames your date. A shelf life you tested is a shelf life you can defend.

The method is simple and needs no laboratory.

1. Bake one ordinary batch, on an ordinary day, with your ordinary flour. Not a special batch.
2. Set aside twenty units. Wrap and store them exactly as your customer will: same bag, same shop shelf, same lack of air conditioning. If your bread sits in a kiosk at 32 degrees Celsius, your trial must sit at 32 degrees Celsius. A trial run in a cool back room proves nothing.
3. Weigh five of them at wrapping and reweigh the same five daily, so you can separate drying from staling using the calculation from Lesson 1.

1. Open one unit every day. Record four things: crumb firmness by squeeze, visible mould, any sweetish or fruity smell that could be early rope, and whether you would personally sell it.
2. Tear each opened unit slowly apart and look for the thread test from Lesson 3.
3. Write the day of first failure for each fault separately.

Now choose your number. Your shelf life is the earliest failure of any kind, minus a safety margin.

Worked example. A bakery in Mombasa, Kenya, runs the trial on wrapped 700 g white bread in March.

1. Day 2: crumb noticeably firmer, weight loss 1.4%, still sellable.
2. Day 3: firmer still, weight loss 2.1%, still sellable, no smell.
3. Day 4: faint sweet fruity smell in one unit. Thread test very slightly positive.
4. Day 5: two units clearly ropey, one showing mould at the crust.
5. Earliest failure = day 4, rope.
6. Safety margin of one day. Declared shelf life = 3 days from baking.

That is an honest number, and it is honest because it was produced in March in Mombasa. Repeat the trial in the coolest month and you may find first failure at day 6. Do not then declare 5 days all year. Either declare the shorter figure permanently, or declare seasonally and manage it, and remember that rope is markedly more frequent in hot seasons.

What a declared date buys you commercially. Suppose that bakery sells 900 loaves a week to shops on sale-or-return, at a selling price of USD 0.85 and a full cost of USD 0.50.

1. Before the trial, shops held bread five days and returned 15%: 135 loaves at USD 0.50 cost = USD 67.50 lost per week.
2. After the trial, the bakery prints a three-day date and delivers three times a week instead of twice, in smaller drops.
3. Returns fall to 6%: 54 loaves at USD 0.50 = USD 27.00 lost per week.
4. Weekly saving = USD 40.50. Extra delivery run, indicative fuel and time, USD 12.00 a week.
5. Net gain = USD 28.50 a week, about USD 1,480 a year, and no customer ever buys a ropey loaf.

Finally, the storage instruction. If the validity of your date depends on how the product is kept, that condition must be declared. Print something the customer can actually follow, such as store in a cool dry place away from direct sunlight, or keep wrapped. Then check that your own shops obey it, because a date is only valid under the conditions you tested.

One caution to close the module. Whether a date mark is legally compulsory on your product, and in what format, is set by your national authority, not by this course. Codex gives day and month for durability under three months and exempts fresh bakery items, but many countries do not grant that exemption. Ask, get the answer in writing, and file it beside your trial record. Those two pages together are the strongest shelf-life position a small bakery can hold.