Choosing the frame
A plate HTST and a tubular HTST both regenerate, heat, hold and cool. The plate does it in a 3 mm channel between gasketed 316L plates; the tubular in a 1 to 2 inch bore. Everything that separates them follows from that: heat transfer per square foot, pressure rating, what fits through the channel, how fast the wall fouls, and what it costs to buy and to open.
Most fluid milk lines are plate and should be. Most cream, mix, pulp and long-run lines are tubular and should be. The wrong choice shows up as a plant that cleans twice a shift or a frame two thirds empty.
A plate channel wants turbulent flow and gets it easily on milk at 2 cP. As viscosity climbs, the velocity needed to stay turbulent climbs and the pressure drop with the square of it. Around 20 cP the pack is either laminar, with starved outer channels, or running 40 to 60 psi to stay turbulent. Wide-gap plates push the crossover to about 100 cP. A 1.5 inch tube-in-tube is happy to 2,000 cP and has no outer channels to starve.
The practical breaks: skim to whole milk, plate. Cream to 40 percent, wide-gap plate or tube. Standard ice cream mix and egg nog, wide-gap plate; premium or high-stabilizer mix, tube. Pulp, cocoa, oat or almond bases with fiber, tube, because a 3 mm gap plugs on the first fiber ball no matter what the numbers say.
Fouling is about wall temperature, and at the same approach both constructions put the same wall in front of the milk. The difference is what a deposit does. In a 3 mm channel a 0.3 mm protein layer on each plate takes 20 percent of the gap; pressure drop climbs sharply and the plant cleans. In a 1.5 inch tube the same layer takes under 2 percent of the bore. The tubular unit fouls at the same rate and runs two to three times as long before anyone notices.
That is the case for tubes on high-solids mix, fortified milk, and any plant that wants a 16 to 20 hour run. The fouling margin is in the bore, not in the plate count.
A plate regenerator at 90 to 95 percent in a section a few feet long has no tubular equivalent. Tubes regenerate 80 to 90 percent in two to three times the length at a higher price per square foot. On a two-shift milk line regeneration is the largest operating cost in the pasteurizer, and plates recover it most cheaply.
Plates also open: a fouled section, a failed gasket or a pinholed plate is fixed on the frame in an hour. A tubular unit is CIP-only or brushable straight tube. Plates lose on pressure: a gasketed frame is honest to 100 to 150 psig, and a homogenizer upstream or a product needing 40 psig of back pressure at 175 F pushes past that.
Buying plates because plates were quoted, and tubes because tubes were on the last line. A plant running 8,000 lb/h of stabilized mix on plates cleans every four hours and blames the CIP. A plant running 60,000 lb/h of skim on tubes has a 40 foot regenerator recovering 85 percent and pays for hot water it did not need. Mixed lines, milk in the morning and mix after lunch, are where a wide-gap plate earns its keep: it handles the mix and regenerates the milk at 88 to 90 percent.
Send the product list with viscosity or fat, the flow, the run length wanted and the shifts. The construction usually picks itself.
Call and talk it through with an engineer: 1-805-484-2992
Each row sets the plate pasteurizer against the tubular one on the same duty. Read down the column that matches your product, then send your viscosity and flow with the quote request so we can pick the right route.
| Condition | Detail |
| Product fit | Plate: milk, skim, juice under 20 cP. Tubular: cream, mix, pulp, fiber to 2,000 cP |
| Flow range | Plate 5,000 to 150,000 lb/h one frame; tubular 2,000 to 60,000 lb/h per skid |
| Temperatures | 40 F in, 161 to 175 F setpoint, 38 to 40 F out |
| Hold | 15 to 25 seconds in an external holding tube |
| Utility | Plate: hot water at 5 degree approach. Tubular: hot water or steam via U-tube set |
| Regeneration | Plate 90 to 95 percent; tubular 80 to 90 percent in two to three times the length |
| Pressure rating | Plate 100 to 150 psig, gasket-limited; tubular 150 to 300 psig |
| Run between CIP | Plate 6 to 10 hours on milk, 3 to 5 on mix; tubular 12 to 20 hours |
| Finish / class | Both 3-A, 316L, 32 Ra; EPDM plate gaskets; tubular seals only at tri-clamps |
| Access | Plate opens on the frame; tubular is CIP-only or brushable straight tube |
Yes; the PMO regulates the hold, the pressure relationship, the diversion valve and the records, and none of that depends on the shape of the surface. A tubular line carries the same sealed timing pump, holding tube, safety thermal limit recorder and leak-protect valve as a plate line. A double tubesheet regenerator adds a vented gap so a failed joint shows outside the product path, which inspectors like because a tube leak can not be found by opening a frame. Expect to be asked how you will inspect welds you cannot see; the answer is a hydrostatic test record by shell at a set interval, and a brushable straight-tube design where it matters.
Yes, and on a mixed product list it is often the right answer: plate regenerator and cooler for cheap area, tubular heating section for run length. The regenerator is where plates earn their place, at 90 to 95 percent in a few feet, and the heating section is where the wall is hottest and the fouling margin of a 1.5 inch bore matters. The hydraulics then span both: the timing pump sees the plate pressure drop plus the tube pressure drop, the booster interlock and the 1 psi relationship are proven across the whole raw path, and the CIP must reach 5 ft/s in the tubes while the plates run at their own cleaning flow, which usually means a CIP pump sized for the tubes with a bypass around the plates.
Plates are opened and dye checked or pressure held one side at a time; tubular units are hydrostatically tested by shell, and a double tubesheet shows a joint leak at its vent while running. Plate cracks start at the port necks and the contact points you cannot see from the frame, so a visual pass is not a test; a fluorescent dye check on each plate on a schedule, or a pressure hold with the opposite side open, is. Tubular inspection is less frequent and harder to do in place, which is why on a tubular regenerator we quote the double tubesheet: the gap between the two tubesheets is open to atmosphere and a failed tube-to-tubesheet weld drips there during the run, where an operator sees it. Fit the vent with a sight glass or a conductivity probe and it becomes a continuous joint-leak detector that a single-wall plate frame does not offer; a pinhole in a tube wall still shows only on the scheduled hydrostatic test.
Plates cost gaskets and labor on a schedule; tubes cost more CIP chemical and energy when they are used where plates would regenerate better. A tubular unit has no wear parts beyond tri-clamp seals, but on thin milk its regeneration is 80 to 90 percent against 90 to 95 on plates, and at 40,000 lb/h each point of regeneration is about 48,000 Btu/h bought twice, once as hot water and once as refrigeration. Price the two on ten years of gaskets, chemical, energy and downtime, not on the purchase order; for stabilized mix the tube wins that sum easily, for skim on two shifts the plate does.
Tubular tolerates a change of product better; a plate frame adapts by adding or swapping plates, up to the point where viscosity or particulates outgrow a 3 mm channel. A 1.5 inch tube-in-tube that runs milk today will pass 40 percent cream, stabilized mix or an oat base with fiber tomorrow, though the 175 F, 25 second mix pair still needs a longer holding tube and a new timed setting. A plate frame can take wide-gap plates in one section or more plates on the bar, but a jump from 2 cP milk to 100 cP mix usually means a new frame. Weigh the five year product plan: if cream, mix or plant-based bases are likely, a tubular heating section now costs less than a second pasteurizer later, and the plate regenerator stays for the milk.
Yes; a tubular system holds several times the product of a plate frame on the same duty, and that volume goes to drain or rework at every start, stop and changeover. A 2 inch tube holds roughly 0.15 gallons per foot, so a 40 foot regenerator pass holds about 6 gallons before the piping is counted, and a plate section of the same duty holds a fraction of that. On a line with eight changeovers a day the water-milk interface alone can cost more product than the CIP savings return. Ask for the hold-up volume with each quote and multiply by your changeovers per week before comparing; where short runs dominate, a plate frame with a tubular heating section only keeps most of the hold-up small.
Two legal routes for fluid milk, 161 F for 15 seconds or 145 F for 30 minutes, and the exchanger each needs.
+ Learn More
Raw milk heated by pasteurized milk through a plate, 85 to 95 percent, pressure always pointing the safe way.
+ Learn More
Tube length from velocity and hold time, a valve that diverts in under a second, and a chart that proves both.
+ Learn More