Extended shelf life, heating side
ESL heating lands between HTST and UHT on purpose. Milk, creamers, flavored milk and oat, almond or soy drinks go to 250 to 260 F for 2 to 4 seconds, and the result keeps 30 to 90 days refrigerated with a flavor much closer to fresh than a shelf-stable product. The heating side has three parts: a regenerative preheat, a preheat hold, and a final heater, and the final heater decides the run length.
This page covers the heating train alone. The cooling train, with its own sterility and pressure questions, is on ESL Cooling.
The first section is regenerative: raw product at 40 F against product returning from the hold, lifting the raw side to 195 to 205 F. Then the product sits in a sloped coil for 30 to 90 seconds. Beta-lactoglobulin denatures fastest between 165 and 205 F; if it does that in the bulk during the hold it complexes with casein and stays in suspension, and if it does it on a 250 F wall it sticks. Lines with a preheat hold run three to four times longer, and the coil is cheap.
Plant-based drinks change the recipe. Oat carries starch that gels from about 140 F; almond and soy carry protein that behaves like whey. The hold still helps, but the heater bore is larger and the velocity higher.
From 200 F to 255 F is a 55 degree lift on pressurized hot water, and it is where the exchanger is designed. Corrugated tube-in-tube or a multi-tube bundle keeps the product at 5 to 7 ft/s so the deposit is sheared thin as it forms. Hot water enters at 262 to 268 F, counterflow, so the wall at the product outlet is only 5 to 8 degrees above the product. A heater sized on a 20 degree approach is cheaper and hits its pressure drop limit in two hours.
HeatX sizes ESL final heaters with 20 to 30 percent more area than the clean duty, so that as deposit builds the outlet still reaches setpoint with the valve open wider, not with the water temperature raised. Raising the water to chase setpoint is the most common way to shorten a run, because it raises the wall exactly where the deposit is forming.
At 250 F a gasketed plate heater is possible on thin milk where the plant already stops each shift. The gaskets are the limit: EPDM at 250 F for hours a day ages in months, and the narrow channel reaches its pressure drop limit on a thin deposit. The tubular heater is the standard answer on lines that want 8 to 12 hours between cleanings, on anything over 12 percent solids, and on anything with fiber or stabilizer.
Treating an ESL heater as a hotter HTST heater. HTST sections are sized close, on a small approach, for a run of a shift. An ESL heater sized that way, without the preheat hold and without area margin, meets setpoint for an hour and then needs its hot water raised every hour until the CIP. The line sold as a 12 hour line runs 3, and the real fault was a hold coil never installed and 25 percent of tube length value-engineered out.
Send the product and solids, the flow, the setpoint and hold, the run length needed and the steam pressure for the hot water set.
Call and talk it through with an engineer: 1-805-484-2992
The heating train is described here as the sizing basis: preheat, hold, final heater and the pressure that keeps the product liquid. Compare your setpoint and run length with it, then send your figures for quoting.
| Condition | Detail |
| Product | Milk, creamers, flavored milk, oat, almond, soy and pea drinks, 8 to 25 percent solids |
| Flow range | 3,000 to 40,000 lb/h, one skid per line |
| Temperatures in / out | 40 F raw; 195 to 205 F after preheat; 250 to 260 F at hold inlet |
| Hold | 30 to 90 seconds preheat hold at 195 to 205 F; 2 to 4 seconds final |
| Utility | Pressurized hot water at 260 to 275 F from a steam-to-water U-tube set |
| Approach | 5 to 10 degrees hot water to product at the heater outlet |
| Construction | Corrugated or multi-tube final heater; plate or tubular preheat regenerator |
| Finish / class | 3-A, 316L, 20 to 25 Ra, welded product path |
| Product pressure | 40 to 60 psig at the heater to stay above saturation; 150 psig design |
| CIP | Caustic then acid; designed to run 8 to 12 hours between |
The pressure drop across the final heater and the hot water valve position needed to hold setpoint, both of which rise early and smoothly as deposit builds. Record a clean baseline after each CIP and set the end of run as a fixed rise over that baseline rather than waiting for the outlet to sag. Plotted run by run, the same trend shows whether a recipe change, raw milk quality or drifting CIP is shortening runs, which is the argument a plant otherwise has with its supplier.
It depends on what is on the wall, and the only way to know is to look. Protein-rich deposit from the cooler end of the heater responds to caustic first. In the hottest zone near 255 F the deposit is mostly calcium phosphate, and an acid step first can open that mineral layer so caustic reaches the protein under it. Pull a tube or open a head after the first production runs, see which deposit dominates, and set the order from that; confirm it by whether the pressure drop returns to its clean baseline after each wash. Recheck the order after any recipe change.
Push product out with hot water first, because cooled starch gel sticks harder, then run caustic at full temperature and velocity, and add an enzyme step if pressure drop does not recover. Oat runs usually need a longer caustic cycle than milk. Schedule plant-based runs just before a full clean rather than in the middle of a dairy campaign, so the heater never goes from starch back to milk on an intermediate rinse.
A full clean, a protein-specific verification test, and a heater that drains completely. ATP swabs do not detect allergen protein reliably, so plants use protein-specific tests at agreed points. The heater's contribution is design: a welded product path, no dead legs and full drainability, so residue has nowhere to stay. Put the allergen sequence into the production plan instead of leaving it to the shift.
Yes; if the margin above saturation is thin, product at the hottest wall boils locally even while the bulk stays liquid, and those bubbles speed deposit formation. Water saturates at about 35 psia at 260 F, so roughly 20 psig, and a 40 to 60 psig design at the heater leaves a margin that a pressure drop rising over the run, or a back pressure valve set at its low end, can erode. Check the pressure at the heater outlet, not at the valve, after any change to the back pressure valve or the downstream piping, because extra pressure drop after the heater lowers the pressure the heater actually sees.
First confirm the CIP flow reaches every tube at 5 ft/s or more, then check temperatures and concentrations at the heater, not at the CIP skid. A multi-tube bundle cleaned at production velocity leaves the slower tubes coated, and a long CIP supply line can arrive 15 degrees cooler than the skid reads. If the pressure drop still does not return to baseline after those two fixes, inspect the tubes, identify the residue, and treat it as a CIP design problem rather than a one-off. A heater that never quite comes back clean loses run length every cycle.
Three stages from 255 F to 38 F, treated product always at higher pressure, glycol for the last 15 degrees.
+ Learn More
Steam into the milk or milk against a hot wall: how the two routes to 285 F differ in flavor, fouling, energy.
+ Learn More
From the sterile hold to the aseptic tank and filler: coolers sterilized with the line that stay sterile.
+ Learn More