HTST Hold Time After a Flow Increase
A small dairy runs a 6,000 L/h HTST that pasteurizes 3.5% milk at 72 °C for 15 s. The holding tube is 22 L of 50-mm-ID sanitary tubing with no flow-control device; the recorder shows 72.5 °C throughout the run. To finish an oversized order, the operator raised flow from the normal 5,000 L/h to the unit's 6,000 L/h nameplate and ran for 4 hours. Two days later the finished-product phosphatase result comes back at 350 µg/mL phenol equivalent against a limit of 10; coliforms are negative. Milk density is 1,030 kg/m³ and viscosity about 0.8 mPa·s.
1. Calculate the actual holding time at 6,000 L/h and determine whether the run was legally pasteurized. Address whether the flow is turbulent enough for the mean residence time to be meaningful.
Read the worked answer
Holding time. Volumetric flow = 6,000 L/h ÷ 3,600 = 1.667 L/s.
t = 22 L ÷ 1.667 L/s = 13.2 s
Required is 15 s, so the run is 12% short. The tube was sized for the job at the original rate: 22 L ÷ (5,000/3,600 L/s) = 15.8 s, only a 5% margin. Raising flow to the nameplate maximum consumed that margin.
Is the mean time meaningful? Check the Reynolds number:
A = π(0.050)²/4 = 1.963 × 10⁻³ m²; v = 1.667 × 10⁻³ ÷ 1.963 × 10⁻³ = 0.849 m/s
Re = (1,030 × 0.849 × 0.050) ÷ 0.0008 = 5.5 × 10⁴
Well above the turbulent threshold for sanitary tubing, so the velocity profile is nearly flat and the mean residence time is a fair representation of the fastest substantial fraction of the fluid. If this had been laminar (Re < 2,100) the 13.2 s average would hide a fast core with far less residence time, and the shortfall would be worse than 12%.
Verdict. The product was held at 72 °C for about 13.2 s, not 15 s, so it is not legally pasteurized, regardless of what the temperature recorder shows. The recorder only proves temperature; nothing in the system measured flow.
2. Explain the phosphatase result and set out the product disposition and notification actions you would take.
Read the worked answer
Root cause. Alkaline phosphatase is inactivated slightly more slowly than the target pathogens, which is why it is used as the pasteurization indicator. A result 35× the limit means a substantial fraction of the product received sub-lethal heat treatment. The temperature chart looks perfect because the failure mode was residence time, not temperature - the flow-diversion valve is temperature-actuated and cannot see a flow excursion. The 4-hour run at 6,000 L/h is the entire suspect window.
Actions, in order:
- Hold everything. Place all packaged and bulk product from that 4-hour run, and back to the last satisfactory phosphatase result, on hold - no shipping, no blending into compliant lots.
- Define the window from records. Pull the charts, flow logs and the previous diversion/tracer test to bound the affected volume rather than assuming 4 hours.
- Rework or condemn. Re-pasteurize through a verified 15-s hold (or longer) with documentation, or divert the product to a non-dairy-legal use. Do not simply re-run for the same 13.2 s twice and call it 26.4 s - confirm with the state regulator that the rework protocol is acceptable before shipping.
- Notify. Report the deviation to the state regulatory agency under the applicable milk sanitation program; the plant's own HACCP/PMO records require a documented deviation, corrective action and verification.
- Verify the fix before restart with a conductivity or dye tracer test on the hold tube, not by calculation alone.
3. Propose and size the physical fix so the unit can legally run at 6,000 L/h, and state the verification you would require before returning to production.
Read the worked answer
Required hold-tube volume at the maximum rate:
V = 6,000 L/h ÷ 3,600 × 15 s = 25.0 L
Present volume is 22 L; add at least 3 L, and preferably 10% more (about 27 L) so normal flow variation cannot eat the margin the way it just did.
Check the consequences of the longer tube:
- Velocity stays at 0.849 m/s, so Re stays ~5.5 × 10⁴ - still turbulent, no new distribution problem.
- Added pressure drop over 3–5 L of 50-mm tube is small (a few tenths of a psi at this velocity), well within the booster pump's capability, but confirm the pump curve and that the diversion valve still operates crisply.
- Confirm the extra length still drains and is cleanable in place; no dead legs.
Institutional fix. Geometry alone is not enough - the operator changed flow because nothing prevented it. Add a flow meter with a low-flow-to-hold interlock, or a fixed-speed pump/metering orifice sized for 5,500 L/h, so the hold time cannot be reduced by an operator decision. Update the batch record to capture flow as well as temperature.
Verification before restart: conductivity or salt-tracer test measuring the actual minimum residence time at 6,000 L/h, a documented phosphatase run on the first production, recalibration of the recorder and flow meter, and a written revised operating procedure with the new maximum rate posted at the control panel.