🥛 Keeping Milk Safe

Milk spoils faster than almost anything else a smallholder produces, and understanding exactly why tells you where to intervene.

Start with what milk is from a bacterium's point of view. It comes out of the cow at roughly 37 degrees Celsius - body temperature. It is liquid, so organisms move freely through it. It is packed with sugar, protein, fat and minerals, which is precisely the diet a bacterium needs. And it is close to neutral in acidity when fresh, so nothing about it is hostile.

If you set out to design a growing medium for bacteria in a laboratory, you would design something very like fresh milk. That is not a metaphor. That is essentially what laboratory culture media are.

So bacteria that get into milk multiply, and they multiply fast. Where do they come from? From the udder itself, from the milker's hands, from the container, from the strainer cloth, and from the air. Everything in the milking module was about keeping that starting number low, and the reason it matters so much is that bacteria multiply by doubling. A batch that starts with ten times as many organisms does not end up ten times worse in a fixed way - it arrives at any given bad number far sooner.

Now add the African reality. In much of sub-Saharan Africa, ambient temperatures commonly reach 35 to 42 degrees Celsius. Milk left standing does not cool toward a mild room temperature. It sits at or near the temperature bacteria most enjoy, which happens to be close to body temperature. Heat does not merely fail to protect the milk. It actively accelerates the spoilage organisms.

That is the bad news, and it is genuinely bad. But milk is not defenceless in the first hour or two, and this next fact is one of the most useful in this whole course.

Milk contains its own natural antibacterial system, called the lactoperoxidase system. It is present in the milk itself, put there by the cow, and it provides real bacteriostatic protection immediately after milking. Bacteriostatic means it holds bacteria in check rather than killing them outright.

The reviewed literature describes this natural system as capable of preventing a significant rise in bacterial load for roughly the first 3 to 4 hours after milking at ambient temperature.

That is a genuine, sourced grace period, and it is the single most practically useful number in this module. You have roughly three to four hours in which milk is defending itself. What you do inside that window largely determines whether the milk arrives saleable.

But be precise about what the grace period is and is not.

It is not a licence to leave milk out indefinitely. It is a window, and it closes. Once the natural system is exhausted, bacterial growth is no longer being held back and the population climbs against no resistance, in ambient heat that is helping it along.

It is not a substitute for hygiene. The system holds back a bacterial load; it does not remove one. Milk drawn dirtily starts with so many organisms that the natural defence is overwhelmed sooner and the milk is in trouble well inside the window.

It is not a substitute for cooling. It buys you the time in which to cool. That is its whole function in a well-run operation.

And it is not a food-safety guarantee. It does nothing whatever about brucellosis or bovine tuberculosis, which are covered later in this module.

What it does give you is a planning number. Take out a watch and time your own milk's journey from the teat to whatever counts as cooling or delivery in your operation. If that journey takes ninety minutes, you are comfortably inside the natural protection window and your milk should be arriving in good condition - and if it is not, your problem is hygiene at milking, not the cold chain. If that journey takes five hours, you are outside the window every single day, and no amount of care at the cow will rescue it.

That one measurement - how many minutes from teat to cooling - is the diagnosis that tells you which half of this course to work on. Most farmers have never made it. It takes one morning and a watch.

The same natural lactoperoxidase system, incidentally, is the one that a Codex-recognised preservation method deliberately boosts when mechanical cooling is genuinely unavailable. That method is Lesson 3, and it is a real technology, not a folk remedy - but it belongs at a collection centre with trained staff, and there are conditions attached that matter.

You have three to four hours of natural protection. This lesson is about what you do with it.

The target, stated plainly: cool raw milk to below 4 degrees Celsius as soon as possible after milking.

That figure is consistent with the other two temperature numbers you met on the testing bench. Milk received at a chilling station should be below 7 to 8 degrees Celsius. And milk held at 0 to 4 degrees Celsius shows no detectable increase in bacterial count within 24 hours.

Put those three together and the whole logic of the cold chain appears. Below 4 degrees, the bacteria stop multiplying to any measurable extent, and the milk holds its condition for a day. Between 4 and 8 degrees you are in the receiving band - acceptable at the point of arrival, on the way to proper chilling. Above that you are on a slope, and the higher you go the steeper it gets, until you reach the 35 to 42 degree ambient of Lesson 1 where the bacteria are working at full speed.

Now the crucial point about what cooling actually does, because people misunderstand this and make expensive mistakes as a result.

Cooling does not kill bacteria. It does not undo anything. It does not clean milk. It suspends bacterial multiplication, and that is all it does.

So the count you start with is roughly the count you still have. Cold milk that was drawn dirtily is cold dirty milk. If you chill a batch that already carries a heavy bacterial load, you have preserved that load beautifully, and it will fail a resazurin test at the collection point just the same.

This is why the milking module comes before this one, and why a farmer whose milk fails should establish which of the two problems he actually has rather than spending money on a cooler that will not fix a cloth-and-bucket problem.

Now an honest gap. You will hear rules of thumb like "cool within two hours or lose a quarter of the value." A precise, Africa-specific figure of the form "cool within X hours or lose Y percent of value" was not available for this course, and this module will not invent one. What is sourced and safe to work from is this: the natural protection covers roughly the first three to four hours, and every hour beyond that without cooling or an activated preservation system is working against an accelerating bacterial population in 35 to 42 degree ambient heat. The correct target, wherever it can be achieved, is below 4 degrees - not merely "cool to the touch."

That phrase deserves emphasis. Cool to the touch is not a temperature. Milk that feels cool against your hand is very likely somewhere in the twenties, because your hand is at 37 degrees and almost anything below that feels cool. A thermometer costs a fraction of what a rejected can costs, and it is the only way anyone knows the temperature of anything.

Now the practical question of where the cold chain actually fails, and the answer is not where most people expect.

The cold chain fails most visibly not at the point of cooling itself, but at the transitions. Three in particular:

First, the walk or ride from the farm to the collection point. This is very often the longest uncooled period in the whole journey, and it is the one nobody counts, because it does not feel like storage. It feels like travel. But milk in a can on a bicycle in the morning sun is milk in storage at ambient temperature.

Second, transfer between containers. Every time milk moves from one vessel to another it warms, it is exposed to air, and it meets a new set of surfaces that may or may not be clean.

Third, the wait at the collection point before the chiller or the tanker arrives. Milk that has been carefully handled all morning can sit for an hour in the sun at the collection point while everyone waits, and lose everything that care achieved.

So the instruction for this lesson is not "buy a cooler." It is: identify and protect the weakest link in your own specific route to market. Do not assume the risk lies at the farm, and do not assume it lies at the plant. Time each stage, write it down, find the longest uncooled stretch, and attack that one.

Sometimes the fix costs nothing at all. Delivering to the collection point half an hour earlier, so the milk is not waiting for the tanker in the heat. Carrying the can in the shade rather than strapped to the sunny side of a bicycle. Wrapping the can in a wet cloth for the journey. Agreeing with your neighbours to combine trips so one person leaves promptly rather than three people leaving late. These are scheduling and habit changes, and they attack the exact stage that is doing the damage.

Sometimes cooling is genuinely impossible. There is no grid electricity, no generator, no solar chiller, and no collection point with a tank within reach. For that situation Codex Alimentarius has published guidelines for a real, tested method: activating milk's own natural lactoperoxidase system to extend the grace period.

This is not a folk remedy. It is an internationally recognised food-safety method with a published Codex guideline behind it, and it has been field-tested in Ethiopia under real smallholder and collection-centre conditions. It is also the one method in this course that comes with conditions serious enough that they must be taught alongside it.

First, what it is. Lesson 1 explained that milk carries a natural antibacterial system that protects it for roughly three to four hours. That system needs two things to work at full strength, and fresh milk contains only a limited amount of them. The activation method adds those two things in controlled amounts, so the system works harder and for longer.

The two compounds, and their doses:

• Sodium thiocyanate, at 14 milligrams per litre of milk.
• Then sodium percarbonate, at 30 milligrams per litre of milk.

They are added in that sequence, with mixing between the additions. Both are food-grade compounds. This is a controlled dosing procedure, not something to be improvised or eyeballed.

The timing condition is strict. Activation must be carried out within 2 to 3 hours of milking. That is, inside the natural three to four hour protection window, not after it has already lapsed. This method extends a protection that is still working. It cannot resurrect milk that has already spoiled, and adding it to milk that is already going off achieves nothing.

Now the results, which are the reason anyone bothers. Shelf life achieved, by storage temperature:

• Roughly 7 to 8 hours at 30 degrees Celsius
• Roughly 11 to 12 hours at 25 degrees Celsius
• Roughly 16 to 17 hours at 20 degrees Celsius
• Roughly 24 to 26 hours at 15 degrees Celsius

Read that column of numbers carefully, because it teaches something beyond the method itself. The cooler the milk is kept, the more the activation buys you. At 30 degrees you gain hours. At 15 degrees you gain most of a day. Activation and keeping milk as cool as you can are not alternatives - they multiply each other.

An Ethiopian field study, run under real conditions along an actual dairy value chain, confirmed this same range: raw milk shelf life extended from about 7 to 26 hours depending on storage temperature between 15 and 30 degrees. That study also measured what was happening biologically - reductions in total bacterial count of about 0.76 to 0.94 log colony-forming units per millilitre in activated milk compared with non-activated, along with reductions in total coliform count and in E. coli.

That is genuinely African, genuinely field-tested, quantified evidence, and it is among the strongest material in this whole course.

Now the conditions, and they are not optional.

Codex is explicit that this method is appropriate only where technical, economic or practical reasons genuinely prevent cooling. It is a fallback, not a first choice. If you can cool the milk, cool the milk.

It should be applied at a collection centre, by trained staff, under dairy-plant oversight. It should not be improvised by an individual farmer at home. The reason is dosing. Fourteen milligrams per litre is a small quantity that needs proper weighing equipment and a known milk volume. Guessing it means either underdosing, which does nothing, or overdosing, which puts more of these compounds into food than the guideline permits. This is exactly the kind of procedure that goes wrong when it is done from memory with a spoon.

Pasteurisation afterwards removes the residual activation compounds. That tells you what this method is for: it is a bridge to processing, holding milk in acceptable condition until it reaches a plant that will heat-treat it. It is not a way to sell raw milk for longer without further treatment.

And the final condition, which follows from everything later in this module: activation does nothing whatever about brucellosis or bovine tuberculosis. It controls spoilage organisms. Activated milk from an infected cow is still infected milk.

So the practical position for a smallholder is this. You do not do this at home. What you do is find out whether your cooperative or collection centre knows about the method, whether they have trained staff and the compounds, and whether it would help on your particular route - and if you are in a cooperative that has no cold chain at all, this is a concrete, Codex-backed proposal worth putting to your committee.

Most smallholders cannot chill milk on their own farm, and this lesson is honest about what the options actually are and how thin the evidence is on several of them.

Start with the arithmetic that shapes everything. A household with one or two cows produces a small volume of milk. A mechanical chiller costs a large amount of money and costs money to run whether it is full or nearly empty. One or two cows' worth of milk cannot justify a chiller of its own, and no amount of enthusiasm changes that.

So there are three directions to look: passive cooling, off-grid mechanical cooling, and aggregation.

Passive cooling means evaporative cooling. The principle is real and it is old. When water evaporates from a surface, it takes heat with it, and the surface and whatever it surrounds get cooler. A chamber with porous walls kept wet - a charcoal-and-water cooler is the common design - will hold its contents below ambient temperature without electricity, without fuel, and without moving parts. It works best where the air is dry, because dry air takes up evaporating water readily, and it works least well in humid conditions.

This has been genuinely researched for milk storage in arid pastoral areas of Kenya, including for camel milk. That study exists. However, its specific achieved temperatures, its construction dimensions and its shelf-life-extension figures could not be obtained for this course. So this module teaches the design principle, which is sound, and gives no performance number at all. If you want the figures, that Kenyan evaporative cooler study is the document to obtain, and no number should be quoted from it second-hand until somebody has read it.

What that means practically is: an evaporative cooler is worth trying, it is cheap to build, and you must measure its performance yourself with a thermometer rather than trusting a claim. Put a thermometer in it, put another in the shade outside it, and read both for a week. That is a real experiment, it costs you nothing but attention, and it will tell you more about your own conditions than any published figure would.

Second, solar-powered mechanical cooling. This has been piloted and studied with smallholder dairy farmers in Kenya. Again, the specific achieved farm-gate temperatures, the costs, and the effects on farmer income from those Kenyan studies could not be obtained for this course, so no number from them is given here.

What is safe to teach is the structural point. Solar cooling removes the grid-electricity dependency that stops many rural collection points from installing a conventional chiller. That is a genuine advantage in places where the grid does not reach or does not stay on. The cost is a significant upfront capital investment, and this course cannot tell you what it is, because all prices in this course are local and unverified. If somebody offers you a solar milk cooling system, the questions are: what temperature does it actually achieve, at what volume, under our conditions, and can I see one working at a farm like mine? Ask for a demonstration and a thermometer reading, not a brochure.

Third, and most important for most learners: aggregation.

The standard structural solution to "one farmer cannot afford a chiller" is that one farmer does not buy one. Milk from many small producers is pooled at a hub, and the hub owns one chiller - grid-powered, generator-powered, or solar. This is the model behind most successful African dairy cooperative cold chains, and it is why cooperatives exist in dairying far more than in most other farm enterprises.

The threshold question - how many cows or how many litres a day it takes before a shared chiller becomes economically viable - could not be sourced for this course, and no figure is given. That is a genuine gap, and it matters, because it is exactly the number a cooperative committee needs when deciding whether to invest. What you can do is build the case with your own numbers: how many members, how many litres a day between them, what the chiller costs to buy and to run locally, and what the current losses to rejected and downgraded milk actually amount to. That last figure is the one most committees never calculate and it is often the largest.

And here is the discipline to apply to any cooling investment at all. Go back to the weakest link you identified in Lesson 2. If your longest uncooled stage is the two-hour wait at the collection point for the tanker, a cooler on your own farm does not touch the problem. If your longest stage is the three hours the evening milk sits at home before the morning collection, then on-farm cooling is exactly right. Spend money on the stage that is actually failing, and you will not find that out without the timings.

This lesson is the most serious in the whole course. Read it carefully, and then act on it, because the risk here falls not on your income but on the health of the people who drink your milk, including your own children.

Two diseases matter most in raw milk.

Brucellosis in cattle is caused by the organism Brucella abortus. Bovine tuberculosis is caused by Mycobacterium bovis. Both are real, documented diseases of African cattle herds. Both pass from cattle to people.

How they reach people: through raw milk and raw dairy products, and through direct contact with infected animals, with birth fluids, and with aborted material. That second route matters for you personally as the person handling calvings and abortions, quite apart from what anyone drinks.

What they do to people:

Brucellosis in a person causes a debilitating, often chronic fever illness. Understand what those words mean in practice. It is not a short illness. It comes with fever that rises and falls, drenching sweats, joint and back pain, exhaustion, and it can go on for months or return again and again over years. People lose their ability to work.

And here is the part that makes it so dangerous in rural Africa specifically: those symptoms look exactly like malaria. A person with brucellosis presents with fever and aches, is tested for or simply treated for malaria, does not get better, and the real diagnosis is missed, sometimes for a very long time. Brucellosis is routinely misdiagnosed for this reason. It is treatable when it is correctly identified, and it goes on wrecking a person's life when it is not.

Bovine tuberculosis causes tuberculosis disease in people that is clinically similar to ordinary human-strain TB, and it can affect lymph nodes, bone and other tissues beyond the lungs. It too is underdiagnosed, because it looks like the TB that everyone is already testing for.

So both diseases share the same terrible property: they hide behind a commoner illness that health services are already looking for. That is precisely why a farming family can be affected for years without anybody connecting it to the milk.

Now the honest state of the evidence, and this course is going to be straight with you rather than impressive.

Brucellosis and bovine TB have been studied in African cattle in at least Ethiopia, Uganda, Tanzania and Burkina Faso. Surveillance work exists, the studies are peer-reviewed, and both diseases are found in African dairy cattle populations. That much is confirmed.

But the specific prevalence percentages from most of those studies - the Arsi zone study in Ethiopia, the Mbarara milk basin study in Uganda, the Dar es Salaam and Lugoba study in Tanzania, the Burkina Faso study, and a Tanzanian systematic review that pooled many studies - could not be obtained in a form this course could verify. So this course prints no percentage prevalence figure for brucellosis or bovine TB for any African country. Anyone who quotes you a national rate should be asked which study, which population, and which year.

There is one exception, and it is worth your full attention because it is directly useful.

A Ugandan study compared pastoral and zero-grazing dairy cattle systems - 497 pastoral cattle and 226 zero-grazing cattle, 723 animals in total. It found animal-level brucellosis seroprevalence of 34.0 percent in the pastoral system, against only 3.3 percent in the zero-grazing system. At herd level, 100 percent of pastoral herds were positive against 5.5 percent of zero-grazing herds. Abortion rates in seropositive cows were 23 percent in the pastoral system against zero in the zero-grazing system.

That is roughly a tenfold difference in individual animal seroprevalence between the two housing systems, in one country, in one study. The likely reason is that zero-grazing sharply reduces nose-to-nose contact and shared pasture and water between herds, which is how brucellosis spreads between animals.

Do not over-read it. It is one comparison, in one place, at one time, and it does not prove that zero-grazing prevents brucellosis. But it is real, numeric, African evidence that your housing system and your herd's disease exposure are linked, and it adds a serious animal-health argument to everything the housing module already said.

Finally, what you must do, and where this course stops.

Herd testing, certifying a herd as free of these diseases, notifying the authorities of a case, and any policy on what happens to a positive animal are all set by your national veterinary and public-health authorities. They differ by country. This course gives you none of them, and cannot.

What you do is go and ask your district veterinary officer or the equivalent national veterinary authority: is there a brucellosis or bovine TB testing programme my herd can join or must join, what am I required to report and to whom, and what is the position on selling raw milk here. Write the answers down with the date and the name of who gave them.

And until you have those answers, treat all raw milk as potentially infected and heat it. That is the subject of the final lesson.

Prevalence figures differ from place to place and could not be established for this course. The control measure does not differ. Heat kills Brucella and Mycobacterium bovis, and the physics of that is the same everywhere in the world. This lesson gives you the real numbers.

The standard pasteurisation time and temperature combinations:

• Vat, or batch, pasteurisation: 63 degrees Celsius, held for 30 minutes.
• HTST, high-temperature short-time: 72 degrees Celsius, held for 15 seconds.
• Ultra-pasteurisation: 138 degrees Celsius for 2 seconds.
• UHT and aseptic processing: no single fixed time and temperature. These are validated for each product and each piece of equipment by a qualified process authority.

Notice the trade in that list. Lower temperature means longer time. Higher temperature means shorter time. Both combinations achieve the same result, which is why there is more than one right answer. What is not acceptable is a low temperature for a short time, or a temperature you have not measured.

For a smallholder or a small cooperative, the vat method at 63 degrees for 30 minutes is the achievable one. It needs three things: a thermometer, a heat source, and a way to hold the temperature steady for half an hour. That is realistic at village scale in a way that flow equipment running at 72 degrees for 15 seconds is not - you cannot time fifteen seconds by hand in a moving stream of milk, which is precisely why HTST requires a machine.

The practical method for vat pasteurisation is a double boiler arrangement - the milk in one vessel standing in a larger vessel of hot water - because milk over direct heat scorches on the bottom and the flavour is ruined. Heat to 63 degrees, watching the thermometer. Start the thirty minutes when it reaches 63, not when you put it on. Hold it there, stirring, for the full thirty minutes. Then cool it quickly and keep it cold.

The thermometer is not optional. There is no way to judge 63 degrees by hand or by eye. Milk at 55 degrees looks and behaves exactly like milk at 63, and one of those is safe and one is not.

Now the single most useful thing in this lesson for a household with no equipment at all.

Boiling milk on a stove reaches 100 degrees Celsius. That exceeds every one of the pasteurisation combinations above by a wide margin, and it is more than sufficient to destroy Brucella and Mycobacterium bovis. A smallholder with no pasteurisation equipment who brings milk to a full rolling boil is taking an effective, low-cost, immediately actionable safety step.

Full rolling boil means the milk is genuinely bubbling all over, not merely steaming or scalding at the edges. Milk boils over easily and quickly, so watch it, but bring it right up.

That one habit, in a household drinking its own milk, is the difference between a real risk of brucellosis and bovine TB and effectively none from that route. It costs firewood and attention. Nothing else in this course delivers so much protection so cheaply.

Now the limits, and there are three that matter.

First, pasteurisation does not undo dirt. It kills organisms; it does not remove dirt, does not remove antibiotic residues, and does not improve milk that is already sour. Milk that is dirty when heated is heated dirty milk. Everything in the milking and cold chain modules is still required.

Second, pasteurised milk is not sterile and it still needs cooling. Heat treatment sharply reduces the organisms present; it does not eliminate everything, and recontamination after heating is entirely possible from a dirty container. Cool it and keep it covered and cold.

Third, and this is a judgement about other people. Family or on-farm raw consumption is a risk decision each household makes for itself, with full knowledge of what is in the previous lesson. Selling untested, unpasteurised milk to other people's households - especially to children - is where the documented zoonotic risk becomes most serious, because those buyers cannot assess a risk they know nothing about and have no way to check your herd. Do not treat pasteurisation as optional in a commercial setting.

And the boundary of this course, stated once more. Whether pasteurisation is required by law, whether herds must be tested or certified, and what must be notified to whom are all decided by your national veterinary and public-health authorities. This module has given you the biology and the physics. It has not given you a legal requirement, because that is not this course's to give.

The same applies, without exception, to every veterinary drug you ever give a cow. The withholding period during which her milk must not be sold or consumed is set by the specific product's label and by your national authority - never by this course, never by a neighbour, and never by a general rule about an active ingredient. Read the label, ask the vet or agrovet who supplied it, and when you cannot find out, withhold the milk and keep asking. That is the standing rule, and it does not change.