Shelf-Life Failure in a 50,000 L/day Pasteurised Milk Plant
A 50,000 L/day toned milk plant uses HTST at 78 °C for 15 s, a plate cooler to 4 °C, and fills 500 mL pouches on a form-fill-seal machine. Over three weeks, 8 of 40 retention samples stored at 6 °C failed the 5-day shelf-life test with off-flavour and curdling, yet every batch released with a negative phosphatase test and a standard plate count below 30,000 cfu/mL on the day of production. The failed samples showed psychrotrophic counts of 10^6 to 10^7 cfu/mL, and coliforms were detected in 6 of the 8 failures. Raw milk from 14 villages averages an MBRT of 5.5 h, and 12% of tankers arrive above 8 °C. The post-fill chlorinated water spray runs at 0.4 ppm free chlorine against a specification of 1-2 ppm, and two filler valves were replaced last month. Retailers are demanding 12 days of shelf life at 6 °C.
1. Diagnose whether this is under-pasteurisation or post-pasteurisation contamination, citing each piece of evidence and explaining what it excludes.
Read the worked answer
Phosphatase-negative plus a release-day SPC below 30,000 largely excludes gross pasteurisation failure: under-processed milk would retain phosphatase and show a high count on the day of production, not on day five. The failure signature is psychrotrophs reaching 10^6-10^7 cfu/mL from a low starting count during five days at 6 °C, with coliforms in 6 of 8 failures. Psychrotrophs that grow in the cold and coliforms that appear after pasteurisation together point to recontamination downstream of the plate cooler - most plausibly the filler, given two valves replaced last month and the post-fill spray at 0.4 ppm against a 1-2 ppm specification. Raw milk quality is an amplifier rather than the cause: MBRT 5.5 h means a high initial load, and 12% of tankers above 8 °C add to it, so any recontamination starts from a large pool. It is not either/or - the raw supply provides the load, the filler introduces it, and the cold chain grows it. Note what the evidence does not yet show: whether the pasteuriser holding time was correct on the eight specific batches.
2. Set out the corrective actions in priority order and state how the HACCP plan should change.
Read the worked answer
Priority one is the filler: swab the two new valves and all filler heads, check O-ring and seal integrity, and validate filler clean-in-place rather than assuming it. Restore the post-fill chlorinated spray to 1-2 ppm free chlorine with a titrated daily log and an alarm at the low limit. Priority two is the raw milk supply: divert or reject tankers arriving above 8 °C, and record the 12% failure rate as a formal supplier non-conformity with a corrective-action deadline. Priority three is verification: hold and test all retained stock from the affected three weeks and re-read the pasteuriser chart records for the eight batches to exclude a short holding time that phosphatase would not have caught. On HACCP, the HTST remains the process CCP and is holding; add a second CCP at the filling and packaging step with critical limits of free chlorine 1-2 ppm and a verified filler sanitation swab result, monitored every shift. A CCP added after the incident is only credible if the monitoring is verifiable and someone signs it.
3. Assess whether the retailers' 12-day shelf life at 6 °C is achievable, and state the conditions and evidence you would require before committing to it.
Read the worked answer
With an MBRT of 5.5 h and a recontamination-prone filler, 12 days at 6 °C is not achievable now. Psychrotrophs at 10^3/mL on day zero with a generation time of roughly 6-8 h at 6 °C reach 10^6 by day four or five and 10^8-10^9 by day eight or nine, which is where proteolysis, bitter off-flavour and curdling appear - exactly the pattern in the failures. Fixing the filler alone will not buy 12 days, because the raw-milk load sets the starting point: a 12-day target requires psychrotrophs at or below 10^3/mL on day five, which in turn needs raw milk under 24-48 h old with MBRT above 6.5 h and a verified 4 °C chain. Recommendation: commit to 9 days at 6 °C now, and move to 12 only after three months of post-fix data showing that day-five psychrotroph counts stay below 10^3/mL across all product lines. Do not run a challenge test on product from a plant with an active recontamination fault and call the result a shelf life.