Selection, step one
A dairy exchanger is chosen three times before anyone calculates square feet. The first choice is the construction family, and it comes from what the product does in a narrow channel: whether it stays thin, thickens when cold, carries solids, or falls apart under shear. The second is whether a single gasket may separate raw milk from pasteurized milk. The third is how much heat the frame should hand back from hot pasteurized product to cold incoming raw.
Plants that skip straight to flow and temperature usually find the mistake at startup. A cream heater built on standard 3–5 mm plates loses pressure within a shift as fat builds on the contact points. A yogurt cooler with tight channels sends product to the filler thin, and QA holds the batch. A regenerator laid out without thought for the pressure differential loses its booster pump to the differential interlock every time the pressures swing.
Settle the three decisions below, in this order, and the sizing work in the next step has firm ground under it.
Fat, solids, pieces and shear sensitivity
Plate, wide-gap plate, tube-in-tube or straight tube
Differential alone, or double wall as well
The percentage that pays back on your run hours
The common error is choosing by the heaviest product the line might run once a year, or by the lightest one it runs every day, without saying which. A frame sized for skim that later takes 40% cream will not deliver its rating, and a frame sized for cream runs skim with far more plates than it needed.
Tell us the full product list and the share of hours each one runs. We size for the hard product and check that the easy one still fits the pack without overshooting temperature.
Many plants run white milk, then flavored milk, then half-and-half through the same pasteurizer. A gasketed plate frame usually carries all three while the heaviest product is half-and-half or light cream. Once heavy cream joins the list, a separate wide-gap section or a dedicated cream unit tends to cost less then oversizing the whole pack for one product.
The changeover order matters as well. Running the lightest product first and the heaviest last keeps fat off the plates for most of the day and leaves CIP one predictable soil to remove at the end.
Neither the 3-A plate standard (11 series) nor the tubular standard (12 series) picks the construction for you, but both constrain it: drainability, access to product surfaces, sanitary gaskets and finish. On a Grade A line the PMO adds the regenerator pressure rule, and that rule is what drives the double-wall question in the first place.
Where the symbol is not required, a food-grade build in the same materials and finish is often enough; the construction rule does not change either way.
Sometimes the right match is a plate frame for the milk sections and a tubular cooler downstream for a cultured stream. Splitting the duty keeps each product in the channel it tolerates, and it often costs less than forcing one construction to cover both. We quote it that way when the numbers say so, and show the single-unit option beside it.
Call and talk it through with an engineer: 1-805-484-2992
It can if the cream is light. Half-and-half and light cream run in standard plates at a cost in pressure drop. Heavy cream at 36–40% fat cooled toward 40°F thickens enough that a standard channel distributes it poorly and fouls; that duty belongs in wide-gap plates or a tubular section, which can sit in the same line beside the milk frame.
For the same water-like duty, usually yes, because tubular surface carries a lower film coefficient and needs more length. On viscous or particulate product the comparison turns around: a plate that has to be opened and cleaned mid-shift, or that damages the product, costs more over a year than the tube that would have replaced it.
Normal flow is rarely the problem. Most Grade A regenerators run single-wall plates with the pasteurized side held at least 1 psi above raw by piping layout and a differential controller. Startup, shutdown, diversion and the changeover to CIP are where the margin is tested; where it is hard to hold in each of those, plants add double wall behind it.
Often only partly. The base is heated to 185–195°F, held for several minutes, then cooled to about 108–112°F to incubate, so the hot side never comes back cold. Some of that heat can still preheat the incoming base, and a modest regeneration section usually pays for itself; chasing 90% on a cultured base rarely does.
Wide gap plate vs tube for dairy: why thin milk runs in plates, cream and mix in wide gap, and broken curd or pieces in tube-in-tube or straight tube.
+ Learn More
When a double wall heat exchanger belongs in a PMO regenerator: a second barrier between raw and pasteurized milk, on top of the 1 psi pressure differential.
+ Learn More
Regenerative heat exchanger sizing for milk: what 85, 90 or 94% regeneration means for plate count, heating and cooling load, and payback on run hours.
+ Learn More