Processing Can’t Outrun Physics: Where the Food-Security Argument Holds, and Where It Quietly Breaks

A region that grows enough food but can’t keep a third of it is the kind of paradox that should keep policy planners awake. The figure most often cited for Sub-Saharan Africa — that as much as 37% of food produced is lost between the field and the fork — traces back to an FAO estimate, and while the exact number is contested (more on that below), the direction is not. Somewhere between a third and a half of what the region grows never gets eaten. That is not a production problem. It is a preservation-and-movement problem.

So when Wael Khoury, Managing Director of Tetra Pak Southern Africa, argued in a recent op-ed that the region keeps reaching for the wrong lever — pouring effort into growing more while ignoring the processing capacity that would stop the losses — he is starting from solid ground. His piece is one of the rare interventions that is genuinely right about the problem. Which is exactly why the proposed solution is worth examining carefully, because being right about the diagnosis is not the same as being right about the cure.

The argument has a quiet move in it. Buried in the middle is the claim that aseptic processing can preserve dairy, juice, and liquid foods without refrigeration — and that, given how expensive and unreliable cold chain is beyond the major cities, this can “revolutionise” food distribution to rural and remote areas. Read quickly, it sounds like processing lets you skip the cold chain. Read slowly, it says something much narrower. The gap between those two readings is the whole subject of this article.

The part that is simply correct

Start by granting the argument its full strength, because most of it is sound.

Processing genuinely extends shelf life. Turning a perishable into a shelf-stable product — drying, milling, canning, or aseptic packaging — buys time, and time is the single most valuable currency in a food system that loses goods to spoilage. A shelf-stable product can sit through a border delay, survive a truck breakdown, and reach a market that fresh produce would never have lived to see. In a region where cross-border freight is slow and fragmented, that durability is a real economic advantage, not a marketing line.

Processing also does things that cold storage alone cannot. It stabilises farmgate prices by smoothing the gap between glut and scarcity — a processor can absorb an oversupply that would otherwise be dumped. It creates downstream employment in plants, quality assurance, and logistics, in a way that primary agriculture, by itself, does not. And it improves a country’s trade position, because processed goods are less vulnerable to the delays that destroy fresh exports.

On the central structural point, Khoury is also correct: South Africa, and the region more broadly, has under-invested in processing relative to both primary production and transport infrastructure. Treating processing as a peripheral commercial concern rather than as food-security infrastructure is a real blind spot. None of that is in dispute here.

The aseptic technology is real too. Ultra-high-temperature (UHT) treatment heats milk to at least 135°C for a few seconds, killing all microorganisms and spores, then fills it into pre-sterilised, light- and oxygen-proof packaging. The result is liquid milk that keeps for six to nine months at ambient temperature without ever seeing a fridge until it is opened. That is not a gimmick. It is one of the genuine engineering achievements of twentieth-century food science, and in a country with an unreliable grid, the appeal is obvious.

So the steelman stands: the problem is real, processing is under-built, and the technology works. Now the complication.

The bypass is a road, not a highway

Here is the analogy worth holding onto. Aseptic processing is not a wider highway that moves all your food faster. It is a bypass road — and a bypass only helps the vehicles that can use it.

The vehicles that can use it are a narrow class: homogenisable liquids and a handful of low-moisture goods. Milk, juice, liquid egg, some soups and sauces, grains and pulses that can be dried. For those, the bypass is excellent. The trouble starts the moment you ask what cannot take that road, because that list is most of the food economy.

You cannot UHT a grape. South Africa’s entire deciduous fruit and citrus export economy — the table grapes, the oranges, the apples that earn the country billions in foreign exchange — is sold as fresh. Freshness is the product. There is no shelf-stable version of a premium export orange that anyone in Rotterdam wants to buy. For that entire category, the cold chain is not an inconvenience to be engineered around; it is the product’s life-support system from packhouse to port to overseas shelf.

You cannot aseptically package a carcass. Fresh and frozen meat depend on an unbroken temperature regime from abattoir to plate, and that chain is precisely where South Africa’s structural weaknesses live — at slaughter capacity, at traceability, at biosecurity. (We have written separately about how abattoir closures have created rural cold-chain dead zones, and why the country imports poultry it could in principle produce; processing does nothing to close those gaps.) Canning and biltong exist, yes — but they are different products serving different markets, not substitutes for the fresh and frozen meat trade.

And you cannot make a vaccine shelf-stable by wishing it so. Biologics, most vaccines, and a large share of temperature-sensitive pharmaceuticals live or die inside a +2°C to +8°C window. There is no aseptic carton that suspends that requirement. For the entire health cold chain, the “bypass” simply does not exist.

So the honest version of the claim is this: processing substitutes for the cold chain of one product class — shelf-stable liquids and dried goods. It does not substitute for the cold chain of fresh produce, fresh and frozen protein, or pharmaceuticals. For those, which collectively dwarf the aseptic-eligible category, the cold-chain problem is exactly as large after you build the processing plants as it was before.

The South African trap: the worse the grid, the more seductive the bypass

This is where the local context turns a reasonable idea into a potential trap, and it is the part of the argument that deserves the most scrutiny.

The logic of the op-ed runs: cold chain beyond the cities is expensive and unreliable, therefore lean on shelf-stable processing instead. In a country with load shedding, two-week fuel reserves, and a grid that the whole industry plans around, that “therefore” is intensely attractive. The worse the electricity supply gets, the more appealing it sounds to simply skip refrigeration altogether. And that is precisely why operators and planners should be careful, because the appeal is emotional as much as logical, and it papers over two physical facts.

First, the aseptic plant does not escape the energy problem — it relocates it. UHT processing is heat-intensive by definition: you are taking liquid to 135°C and holding it, then running sterile, tightly controlled filling lines. That is a serious, continuous electrical and thermal load, sitting in the same load-shedding environment as everyone else. You have not removed an energy-dependent step from the chain. You have swapped a distributed refrigeration load for a concentrated processing load, and concentrated loads are not obviously easier to keep running through Stage 4.

Second — and this is the fact the bypass framing hides completely — the raw milk still needs a cold chain to reach the plant. Raw milk that has warmed and soured is not suitable for UHT processing at all; the standard tests that reject low-quality raw milk exist precisely because spoiled input cannot be rescued by heat. So before a single shelf-stable carton exists, there is a farm-to-factory cold chain doing exactly the job the op-ed implies we can skip. The bypass does not remove the cold chain. It moves it upstream, from farm-to-shop to farm-to-factory, into a part of the system that the “no refrigeration needed” headline never mentions.

There is even a quieter physics footnote. Shelf-stable does not mean indestructible. UHT milk stored cool keeps for around 34 to 36 weeks; stored at 30–37°C — ambient conditions across much of the region for much of the year — that drops to roughly 16 to 20 weeks before taste, colour, and sedimentation degrade it. The product is more forgiving than fresh milk. It is not immune to heat. The cold chain’s job shrinks; it does not vanish.

Who benefits from the framing

It is worth saying plainly, and without any accusation, where this argument comes from. It comes from the Managing Director of a packaging company — and Tetra Pak is, more or less, the global definition of aseptic packaging. The claim that “processing is national infrastructure, on par with roads and ports” is a category-expansion argument, and the category it expands is the one the author sells into.

That does not make it wrong. Vendors are often right about their own field, and a packaging executive understands aseptic processing better than almost anyone. But it does mean a reader should weight the infrastructure-priority ranking with the source in mind. When someone tells you their product category should be treated as essential national infrastructure, the claim may be true and the incentive is also real, and both things can be held at once. The Curator’s standing rule applies cleanly here: follow the physics first, then notice who benefits from the framing. The physics says processing handles one product class. The framing invites you to hear it handling the whole problem.

The real question isn’t “either”

The mistake embedded in the “lever Southern Africa keeps ignoring” framing is that it sets up a choice that does not exist. Processing and cold chain are not competing line items fighting over the same budget. They are sequential parts of one system, each handling the products the other cannot.

Shelf-stable processing should absolutely be built out — for the liquids and dried goods where it is the right tool, and as a genuine resilience layer for rural distribution of those specific products. The op-ed is right to demand it. But it should be built alongside a functioning cold chain, not instead of one, because the fresh produce that earns the export revenue, the protein that feeds the domestic market, and the vaccines that protect the population all still need the chain that processing cannot replace.

The right question is not “processing or cold chain.” It is: which products go down which road — and have we actually built both roads? South Africa’s real failure has been treating either one as optional. Building processing capacity while the cold chain decays in the provinces that grow the most simply relocates the losses. Fixing the cold chain while ignoring processing leaves the liquid-and-dried category exposed. The food we lose will not be saved by one lever, because no single lever touches all the food.

A third of the harvest is already in the ground. Keeping it is not a matter of finding the one clever bypass. It is a matter of building the unglamorous, parallel infrastructure — cold and shelf-stable, fresh and processed — that an honest accounting of the physics demands. The lever the region keeps ignoring isn’t processing. It’s the word and.

Sources & References

• Why Processing Is the Food Security Lever Southern Africa Keeps Ignoring — The op-ed by Wael Khoury, Managing Director of Tetra Pak Southern Africa, that this article responds to. Business Explainer, May 2026.
• Is Post-Harvest Loss Significant in Sub-Saharan Africa? — World Bank analysis tracing the 37% production-to-consumption figure to FAO (2011), with the important caveat that self-reported farm-level losses are far lower.
• From Farm to Fork: Strategies for Sustainable Reduction of Post-Harvest Losses in Sub-Saharan Africa — Peer-reviewed review citing recent estimates of 30–50% annual post-harvest loss in the region.
• Food Loss and Waste Data Gaps in Fruit and Vegetable Value Chains — Frontiers review documenting the wide discrepancy between modelled FAO loss estimates and measured field data.
• UHT Milk FAQ — Tetra Pak’s own technical explanation of UHT processing (≥135°C) and aseptic packaging, confirming the multi-month ambient shelf life.
• Shelf-Stable Milk: Conscientious Convenience — GEA processing engineering overview confirming the six-to-nine-month ambient shelf life of unopened UHT milk.
• Changes in Stability and Shelf-Life of UHT Milk During Long-Term Storage at Different Temperatures — Peer-reviewed study showing UHT milk shelf life falls from 34–36 weeks (chilled) to 16–20 weeks at 30–37°C.