Heat exchangers and chilling the juice

A plate exchanger achieves its performance by turbulence as much as by area. The corrugations in the plates force the flow to churn rather than to slide along the wall in layers, and turbulent flow transfers heat across a boundary far better than laminar flow does. The same corrugations make the machine sensitive to anything solid in the product: a plate pack has narrow channels and will block on pulp, so it belongs downstream of the coarse solids, not on juice straight off a press.

A tubular exchanger — product through a tube, service fluid around it, or a tube-in-tube arrangement — has larger passages and tolerates solids and viscosity. It has less surface area for its size, costs more for the same duty, and cannot be opened up and inspected plate by plate. Where a juice carries pulp, or where a product is being held at temperature for a defined time, tubular is the usual choice.

At small scale the same physics is available in a coil. A stainless coil in a tank fed with chilled water, or a small plate chiller in a transfer line, will pull a few hundred litres down in temperature perfectly well. The compromise is that a coil in the product is an object that must be cleaned and sanitised like anything else touching the cider, and it is not a shape that is easy to clean.

Juice off a press in a warm September is at ambient or above it — a continuous screw press in particular puts mechanical work into the mash and the juice comes out noticeably warm. Every process that a maker would rather control runs faster at that temperature: enzymatic browning, the wild population beginning to multiply, and the loss of the aromatic fraction.

Chilling at the press outlet therefore does several things at once. It slows the browning, which matters where a pale juice is wanted. It postpones the onset of fermentation long enough for settling to work. It makes a longer working day possible, because juice can be held rather than having to be dealt with immediately. And it lowers the temperature at which the ferment will start, which is the cheapest route to the cool, slow fermentation that preserves aroma.

The economics of doing this well come from regeneration. In a multi-section plate pack, incoming warm product is cooled by outgoing cold product across a set of plates before either meets the chiller or the heater, recovering a large fraction of the energy. It is the reason a flash pasteuriser costs far less to run than the temperatures involved would suggest.

Flash pasteurisation is a heat exchanger with a holding tube. Product is heated through the regeneration and heating sections to a target temperature, held in a tube of a length that gives a defined residence time at that temperature, and then cooled back down through the regeneration and cooling sections. The combination of temperature and time is the process; the equipment exists to deliver it repeatably and to prove that it did.

The proving is the part that distinguishes a pasteuriser from a heater. A temperature sensor at the end of the holding tube drives a flow diversion valve, so product that has not reached the target is sent back to the balance tank rather than forward to the filler, and the whole cycle is recorded. What the target and the hold should be for a given product is a validated food safety decision that belongs with a process authority and the local regulator — it is not a figure to be taken from a general reference and applied.

The consequence downstream is that flash-pasteurised product is sterile when it leaves the machine and is then only as protected as the equipment it flows into. It has to be filled aseptically or it will be recontaminated, which is why a flash pasteuriser implies a filler and a line to match. Tunnel pasteurisation — heating the sealed pack rather than the liquid — sidesteps that entirely at the cost of a longer, hotter treatment and a machine with a large footprint.

Heat exchangers foul. Protein, pectic material and mineral deposit build up on the hot surfaces, insulating them and narrowing the channels, so the machine gradually loses duty and gains pressure drop. The signs are a rising pressure differential and an outlet temperature that no longer reaches target at the same flow. The remedy is a cleaning regime — usually an alkaline circulation for the organic deposit and an acid circulation for the mineral one — run to a schedule rather than when the machine complains.

Gaskets are the plate exchanger’s consumable and its weak point. They perish, take a set, and creep out of their grooves, and a failed gasket either leaks to atmosphere, which is obvious, or leaks internally between the product and service circuits, which is not. Plate packs are pressure-tested and inspected on the manufacturer’s schedule for exactly this reason.

The internal leak is the serious one. A pinhole in a plate or a failed interplate seal puts the service fluid into the product — chilled water, or worse, a glycol coolant. Where the design allows, the product side is held at a higher pressure than the service side so that any leak goes the harmless way, and coolant selection takes account of the possibility. A batch that has had coolant in it is not saved by any subsequent treatment.