The mill does not create quality

The olives reach the mill in perforated crates. From a distance they all look alike: green, purple, black, covered by a few leaves left over from the harvest. But sinking a hand into two different containers is enough to see that they are not delivering the same material. One batch is cool, dry and firm, and the drupes resist the pressure of the fingers. Another is warmer: some fruits are bruised, others show a hole, a yielding flesh or the beginning of a degradation that is not always obvious to the nose yet.

Both can weigh the same. They do not have the same future.

Once the olive leaves the tree, time begins to work against it. Harvesting can damage its surface, transport can crush it, and the heat that builds up on the nets or inside a poorly ventilated container speeds up the changes. If the fruit has been attacked by pests or picked up off the ground, the available margin shrinks further. The miller does not receive a neutral material waiting to be given value: the miller receives a quality that already exists and is already vulnerable. The task is not to create it out of nothing. It is to avoid losing any more of it.

Quality arrives before the lorry

The oil begins in the grove. Variety, soil, water availability, pruning, the run of the weather and the moment of harvest all contribute to the composition of the fruit. Even inside a single holding, though, the olives do not ripen together: some cultivars develop early, others late, and in certain varieties ripening is staggered, with fruits on the same tree sitting at different stages.

The outside colour is not enough to describe what is happening in the flesh: an olive that has just turned colour at veraison can keep a high firmness, while another that looks much the same may already have altered the balance between water, fats, phenolic compounds and aroma precursors. Picking very early does not automatically produce a better oil: it can favour certain aromatic profiles, but it can lower the yield and make extraction harder. Waiting increases the amount of oil in the fruit and exposes it more to over-ripening and to late attacks.

The decision belongs first of all to the grower, but a good miller does not stand outside it. The miller knows how the cultivars that reach the plant behave, knows which pastes are harder to separate, and knows what follows from a harvest concentrated into a few days or spread over too many weeks. The relationship between grove and mill should not begin when the lorry pulls into the yard, but earlier, while harvesting and processing can still be organised as a single process.

Taking delivery is already production

The miller’s first operation is not switching on the crusher. It is looking. A batch has to be identified, weighed and linked to whoever brought it in, and cleanliness, integrity, the presence of leaves, soil, foreign bodies and signs of infestation all have to be assessed. Not everything belongs in the same cycle: sound olives and compromised olives should not be treated as though the difference could disappear inside a machine, and different batches can call for separate processing, different timings or, in the worst cases, a refusal.

Saying no is part of the craft.

It is a hard choice in the weeks when the mill runs almost without a break and every delivery is worth hours of production. But accepting any material at all means passing the problem down the line: a degraded batch slows the operations, demands heavier cleaning and produces an oil with defects that no later adjustment will cancel.

The apparent neutrality of the hopper is deceptive. Once the olives have been tipped into it, some distinctions can no longer be reconstructed. That is why the miller has to know not only how to work a batch: the miller has to know when to keep it apart.

The olive does not contain a bottle in miniature

In ordinary speech olives are said to be pressed: the fruit goes in whole, pressure crushes it and the oil comes out like juice from a citrus. But the oil is not a free liquid gathered in a single cavity. It is spread in small droplets through the tissues of the fruit, and to make it separable the plant structure has to be broken first.

After cleaning and de-leafing the olives are crushed, and what comes out is a paste of oil, vegetation water, fragments of flesh, skin and stone. The droplets released are too small and too scattered to be separated efficiently, so the paste is then malaxed. In modern continuous plants it goes on into a decanter, a horizontal centrifuge that uses the different densities of the phases to separate oil, water and solids. A vertical separator further reduces water and impurities in the oil. Hydraulic presses still exist and belong to the practice of some plants, but they do not describe the work of every mill today. In many cases the miller is not pressing: the miller is breaking, re-gathering and separating, and that is where the real decisions lie.

Crushing does not only release the oil: it brings into contact substances that, in the intact fruit, had stayed apart. Breaking the tissues starts enzymatic reactions, and some of them will give rise to volatile compounds associated with green and fruity sensations, while others can contribute to the loss of phenolic components.

There are granite millstones, hammer crushers, disc and blade crushers and systems that remove the stone beforehand, and each configuration applies different forces and produces a different paste. Particle size counts too: more intense crushing releases the oil efficiently, but it can generate more heat, fragment the stone or alter the activity of the enzymes that will shape the aromatic profile and the phenolic content. There is no system that is better in absolute terms: the choice depends on the variety, on ripeness, on the type of plant and on the profile to be obtained.

So the stone mill is not more respectful of the material simply because it is old, and the metal crusher is not industrial violence because it works faster. Quality does not lie in the stagecraft of the tool. It lies in the fit between what that tool does and the olive entrusted to it.

The malaxer is not a waiting room

Seen from outside, the malaxer looks like a slow, unspectacular stage: the blades turn inside a closed trough, the paste is moved rather than squeezed, and the whole thing looks like an interval before the real extraction. It is instead one of the most delicate points in the entire process. During malaxation the oil droplets meet and coalesce, which makes separation in the decanter more efficient, and alongside that the reactions started by the breaking of the fruit carry on.

Temperature, duration, speed of movement and oxygen all influence what happens. A malaxation that is too short leaves part of the droplets scattered in the water and solid phases, and lowers the yield. Extending it is not an automatic solution, though: more time means more exposure to enzymatic and oxidative change, and the balance between aromatic and phenolic compounds can shift. The same holds for temperature: warming the paste reduces its viscosity and helps the droplets separate, while excessive or badly managed heat can alter the profile of the oil and encourage the loss of sensitive components.

The atmosphere inside the crusher and the malaxer is another variable to be governed. Oil is vulnerable to oxidation, and from that truth a rule is easily drawn: the less oxygen enters the process, the better the product will be. But in the early stages oxygen takes part in the formation of some volatile compounds that contribute to green and fruity sensations. Removing it altogether can protect part of the phenolic fraction and at the same time alter the aromatic profile, while letting too much in encourages oxidative reactions and reduces some protective components.

Plants that limit contact with air or introduce inert gases do not therefore guarantee a better oil: they give the miller one more lever. In the early stage oxygen builds aroma, in storage it becomes something to protect the oil from, and the craft consists partly in knowing when its role changes.

The miller therefore has to recognise the paste being worked. A mass obtained from green, water-rich olives does not behave like one produced from riper fruit, and a dry year delivers a different material from the previous season’s. Malaxation is not dead time: it is the point where yield and quality begin to negotiate.

Twenty-seven degrees are not a cliff edge

The temperature of 27 °C turns up often in the story of oil as though it were a natural frontier: below that threshold the oil would keep its aromas and polyphenols intact, above it would start to lose them. It does not work like that. The figure of 27 °C has, first of all, a regulatory function. The term cold extraction may be used for virgin and extra virgin oils obtained at a temperature below 27 °C by percolation or centrifugation of the paste. The term first cold pressing is instead reserved for virgin and extra virgin oils obtained below the same threshold with a first mechanical pressing carried out through a traditional hydraulic press system.

They are two different wordings for two different systems, and an oil obtained in a continuous plant should not be told as first pressing because the phrase sounds more artisanal. But above all 27 °C is not the point at which, by adding a tenth of a degree, the material suddenly changes its behaviour. The effects of temperature depend on the interaction with duration, cultivar, ripeness, type of crushing, oxygen and the characteristics of the paste: under some conditions a controlled increase can improve the yield and even help the extraction of phenolic compounds, under others it can reduce desirable aromatic components.

The number on the label establishes the right to use a term. It does not replace the technical assessment of the process. A competent miller does not work to stay theatrically as far as possible from 27 °C: the work is to know and control the real temperature of the paste, keeping it from being decided by chance, by the season or by the heat accumulated in the machines. Quality does not come from the lowest number, but from the right number for that material and that result, inside the limits laid down.

Yield is a temptation

For anyone bringing olives to the mill there is one number able to eclipse almost all the others: the litres obtained. A higher yield seems to prove that the year was good, the harvest well judged and the mill efficient. A lower yield is felt as a loss: part of the oil has stayed in the pomace and will not reach the demijohn.

The miller works inside that expectation. The miller knows the producer has faced pruning, crop protection, harvesting, labour and transport, and knows how hard it is to explain that extracting the maximum quantity does not always coincide with the best result. Time, temperature, crushing, added water and the setting of the decanter can all raise separation efficiency, but every intervention has consequences: extending malaxation or pushing the system towards maximum yield can alter the balance between aromatic compounds, phenolic compounds and other components of the oil.

This does not mean that yield and quality are always enemies. A well-designed plant can improve both compared with an inefficient process, and contemporary technologies try to separate more oil while cutting long times, heat and water. It does mean that the last recoverable fraction is not free: obtaining it may require putting the paste through conditions that will change what ends up in the bottle.

The miller should not chase the highest yield in absolute terms. The yield to look for is the one compatible with the quality the producer wants to achieve. The difference between the two is where the craft meets the market.

The decanter separates, it does not judge

After malaxation the paste still holds everything together: oil, water and solids. The decanter uses centrifugal force and the difference in density to separate them. In two-phase systems it generally produces oil and a wet pomace that also contains part of the vegetation water. In three-phase systems it separates oil, pomace and a distinct water phase, normally using more process water.

The machine performs a physical separation: it does not know whether the result corresponds to the best possible balance. Speed, throughput, quantity of water and the characteristics of the paste have to be compatible with the system, and a setting that suits one batch may not work with the next.

Water needs attention too. It is used to wash the olives, in some plants it takes part in the separation stages, and it has to be clean and fit for food use. But a greater quantity does not mean more careful processing: many phenolic compounds have an affinity with the water phase as well, and water used during extraction can encourage their transfer out of the oil. The miller has to use it where it is needed and in the quantity the plant requires, without turning the process into an indiscriminate wash.

At the decanter outlet the oil is visible. That does not mean it is already stable, classified or ready to be forgotten in a container. The machine has separated the phases. Judgement begins afterwards.

Cloudiness is not authenticity

Freshly extracted oil can be cloudy. It still contains micro-droplets of water and solid particles from the fruit, and its opalescent look is associated with freshness and with craft work. A filtered oil, clearer, seems instead to have undergone a treatment able to impoverish it. It is a convenient opposition, and often a wrong one.

Filtering is not refining. Filtration is one of the physical treatments permitted for virgin oils and serves to remove suspended water and solid microparticles. Those particles are not only an aesthetic signature of the original material: they can carry micro-organisms, enzymes and substances that go on reacting during storage. The water present encourages hydrolytic and fermentative processes, and the sediment at the bottom can develop sensory defects and force repeated rackings, which expose the oil to air.

A well-executed filtration can therefore increase the stability of the product, but it has to be carried out limiting contact with oxygen and using suitable materials. Natural settling also separates part of the residue, and it takes time, control and rackings. There is no obligation to choose the same solution every time: a small lot to be used up quickly can be handled differently from an oil that will have to stay stable for months, cross a commercial supply chain and be shipped a long way.

What does not hold is the automatic equation between cloudy and authentic, between clear and industrial. Cloudiness describes what is still suspended in the oil. It does not certify how the fruit was grown, how quickly it was processed or how competent the miller was.

Extra virgin is not an impression

When the first oil comes out of the separator the colour can be intense, the smell vegetal and the taste aggressive, and it is easy to declare it extraordinary. It is harder to know what it is. Extra virgin is not the legal synonym of freshly made, green, pungent or produced in a small mill: it is a category defined through chemical and organoleptic parameters.

Free acidity, the peroxide value, the spectrophotometric indices, the ethyl esters and other values have to fall inside the limits laid down. The oil also has to pass sensory assessment: the median of the defects must be zero and the fruity attribute must be perceptible. One low figure alone is not enough. An oil can have a contained acidity and present a sensory defect, it can smell intense and fail a chemical parameter, it can look clear without belonging to the category declared.

Free acidity, moreover, does not correspond to the sour sensation perceived in the mouth: it is a chemical parameter that has to be measured. Colour can deceive too: an intense green depends on the cultivar, on ripeness and on pigments, and on its own it neither proves that the oil is extra virgin nor makes a more golden oil inferior.

That is why the miller has to taste, sample and analyse. Not every test will necessarily be run in the mill’s own laboratory, but the miller has to be able to read the results, recognise a defect and separate the lots according to their real characteristics. Classification does not happen in front of the flowing oil with a glass held up to the light, but by putting process, analysis and sensory assessment together. Enthusiasm can accompany the first tasting. It cannot replace the control.

Quality goes on being used up

In photographs the mill is polished steel, pipework and a green thread running into a vessel. The least photographed part arrives when processing stops. Residual paste, water, fragments of skin and stone stay in the parts of the plant that are hard to reach, and the deposits on the surfaces do not become nobler because they come from an olive: they begin to degrade, and inside a hopper, a pump or a malaxer they pass odours and defects on to the next batch.

Cleaning takes time: stopping the flow, opening the plant, reaching internal surfaces, checking detergents, seals and drains, and keeping the product from being contaminated by the previous residues or by the substances used to remove them. During a short season every stoppage looks like a subtraction from productive capacity, and in fact it protects the capacity that comes after. Cleaning is not the work that begins when production is over. It is a stage of production.

The same holds for the surroundings: oil takes up foreign odours easily, and the processing and storage rooms have to be protected from fumes, fuels and perfumed detergents. The mill does not have to look rustic: it has to be clean enough not to leave in the oil the memory of what it worked before.

When the oil enters the tank the mechanical processing may be finished. The responsibility is not. The oil goes on being exposed to oxygen, light, heat and residues of water and particles, and these factors do not improve it with time: they consume its aroma, its stability and its protective components. Tanks have to be of inert materials and easy to clean, and stainless steel has become the standard because it does not absorb, does not give off odours and allows effective hygienic management.

A conical or sloping bottom keeps the sediment away, closure limits the entry of foreign bodies and contact with air, and inert gases in the headspace above the oil reduce exposure to oxygen. Temperature has to be controlled too: heat speeds up oxidative reactions, while excessive cold can cause the temporary crystallisation of some components, reversible but awkward to manage during packaging and handling.

The miller also has to keep a partly empty tank from standing for long with a large volume of air above the product: every draw-off changes the ratio between oil and headspace, and a vessel well protected when it is full becomes more vulnerable as it is emptied. Storage is not a neutral corridor between extraction and sale: it is another stage of processing, only slower and less visible.

Wine can be designed to develop over years. Oil is not waiting to become riper. It is only waiting to be used before time takes too much away from it.

In Toscano IGP the miller leaves a trace

Olio Toscano IGP (the protected geographical indication for Tuscan extra virgin olive oil) is not defined by the origin of the olives alone. The product specification sets out permitted varieties, the production area, the ways of harvesting and processing, chemical and sensory characteristics and traceability rules. The olives have to be picked directly from the tree, by hand or with mechanical means, washed with water or air and worked through mechanical and physical processes that do not alter the quality characteristics present in the fruit. Extraction has to take place inside the territory laid down.

Growers, millers and packers enter the control system, and every stage is documented through identifiable inputs and outputs. The miller is therefore not an anonymous service between the field and the bottle: the miller is one of the operators through whom the denomination becomes verifiable. The lots have to be kept separate, the processing recorded, and a certifiable batch prevented from being mixed up with other material.

Traceability does not improve the flavour. It does, though, one indispensable thing: it makes it possible to reconstruct who took charge of the product and what happened to it. The requirements at consumption, too, are stricter for some parameters than the general limits of the extra virgin category.

So the denomination does not stop at declaring “oil made in Tuscany”. It establishes which material may carry that name and which responsibilities have to go with it. The miller does not apply the geography at the end, by putting a label on the container. The miller guards it while the oil crosses the plant.

What the miller decides not to do

Many of the miller’s skills show up as actions: adjusting a crusher, setting a temperature, deciding the length of malaxation, correcting the decanter, separating a lot. An equally important part of the craft consists in what the miller decides not to do.

• Not leaving the olives standing because the plant is easier to organise the next day.
• Not mixing a sound batch with a compromised one to avoid processing them separately.
• Not setting the same programme for every cultivar and every stage of ripeness.
• Not extending the process just to recover a few more litres.
• Not calling first pressing an extraction that uses no presses.
• Not treating 27 °C as a formula able to replace all the other decisions.
• Not selling cloudiness as automatic proof of authenticity.
• Not leaving sediment, air and light to work on the oil after the machines have stopped.

The mill cannot give back to the fruit what the field, the olive fly, the heat or the waiting have already taken from it. It can, though, stop the loss from continuing unchecked. That is the least spectacular part of the craft: not adding something that did not exist before, but keeping what already existed well enough for it to be recognised in the bottle.

The mill does not create quality. It decides how much of it will arrive alive at the end.