Login Register

Access the GIFSQ Portal

Select your user type to log in or register a new account.

Student Portal

Access your food safety courses, certifications, and exams.

Instructor Portal

Manage courses, view student submissions, and grade quizzes.

Company Portal

Manage corporate setup, view employee logs, and access QA services.

How to Write a HACCP Plan for Soft Cheese | GIFSQ

How to Write a HACCP Plan for Soft Cheese

Soft cheese — brie, camembert, fresh mozzarella, queso fresco, feta — is where Listeria monocytogenes goes to thrive. High moisture, moderate pH, weeks of refrigerated shelf life during which Listeria happily multiplies at 4 °C. This category has produced some of the deadliest listeriosis outbreaks on record. Regulators know it: the EU sets strict L. monocytogenes criteria for RTE foods, the US runs zero-tolerance for the pathogen in ready-to-eat product.

So let’s be blunt about what this plan is. The pasteurizer works — that’s the easy part. Everything after it is the plan: the brine, the ripening room, the packaging hall, the drains, the condensation dripping off the ceiling at 3 a.m. That’s where soft cheese safety is won or lost. Codex 12 steps, with the weight where it belongs.

Step 1: Get a cheesemaker on the team

QA, production, maintenance, sanitation, procurement — and someone who actually understands cheesemaking. Ripening cultures, brine behavior, surface flora: these behave in ways a generalist QA manager won’t predict, and the hazard analysis will miss what the team doesn’t understand. One member with formal HACCP training. Document it.

Step 2: Describe the cheese completely

Each variety: milk type and treatment — state pasteurized explicitly (raw-milk soft cheese is a separate universe of risk and needs its own plan). Cultures and ripening flora. Coagulant. Salting method. Brine parameters. Ripening time and conditions. Final pH and water activity targets. Rind treatments, surface inoculation. Packaging, shelf life, storage ≤4 °C. If you can’t describe the ripening, you can’t analyze its hazards.

Step 3: Name the consumers — including the vulnerable ones

General population, explicitly including pregnant women, the elderly, and immunocompromised people. Soft cheese is singled out in consumer guidance for these groups in multiple countries. Your severity ratings aren’t being dramatic — they’re reflecting reality.

Step 4: Diagram it, walk it

Receiving → raw milk storage → pasteurization → cooling → culture/coagulant addition → coagulation → cutting/stirring → molding/draining → salting/brining → ripening → packaging → cold storage → dispatch. Add brine maintenance, culture prep, rework, CIP. Walk it with the cheesemaker. Brine tanks, ripening rooms, packaging lines — all on the diagram, all verified. Sign, date.

Step 5: Hazard analysis — interrogate everything after pasteurization

  • Biological: raw milk pathogens — handled by pasteurization, if it works. L. monocytogenes post-pasteurization contamination — the defining hazard, arriving via environment, equipment, brine, people, packaging. S. aureus enterotoxin if the starter fails. Salmonella and STEC in raw-milk variants (separate plan, remember).
  • Chemical: antibiotic residues, allergens (multiple milk species, added ingredients), CIP residues, brine contaminants.
  • Physical: metal, plastic, wood from ripening boards.

Go step by step through every post-pasteurization operation and ask: how does Listeria get in here, and could it grow? The ripening room — cool, humid, product sitting for weeks — deserves the hardest look. Condensation overheads? That’s in the analysis too.

Step 6: CCPs

  • Pasteurization — validated time-temperature for cheese milk.
  • Brine control — concentration, temperature, maintenance (filtration, heat treatment, replacement frequency). Brine touches every cheese; neglected brine has caused outbreaks.
  • Ripening and cold storage temperature — per your validated profile, ≤4 °C or the validated ripening conditions.
  • Metal detection — post-packaging.
  • Post-pasteurization environmental hygiene: usually prerequisite-based, verified by an aggressive EMP. Run the decision tree, document the call — don’t just default it.

Step 7: Critical limits — including the brine

  • Pasteurization: e.g., 72 °C for 15 seconds minimum (HTST) or your validated vat equivalent.
  • Brine: salt concentration at or above the validated minimum, temperature ≤4 °C, with defined maintenance — filtration, heat treatment, or replacement frequency. Listeria survives in brine. “We change it when it looks dirty” is not a critical limit. We’ve heard it said. It didn’t survive the audit.
  • Ripening/storage: your validated temperature profile; final pH and aw targets where applicable.
  • Metal detection: validated test-piece sizes.

Step 8: Monitoring

Pasteurization: continuous recording, per-shift checks. Brine: concentration (salometer or refractometer) and temperature at defined frequencies; maintenance logged when done. Ripening rooms: continuous temperature and humidity recording. Cold storage: continuous recording. Metal detection for every unit, with test pieces challenging it at the defined frequencies. Named positions — the brine checks need an owner, not a hope.

Step 9: Corrective actions

Brine out of spec: hold everything brined since the last good check. Correct and re-verify the brine. QA dispositions the held product — evaluate, re-brine within validated parameters, or destroy. Temperature deviation in ripening: hold, assess, disposition. Pasteurization deviation: divert, reprocess, or destroy. Every action names who decides, and requires the cause to be investigated — not just the product dealt with.

Step 10: Verification — this is where soft cheese plans live or die

A rigorous, zone-based EMP for Listeria spp. — zones 1 through 4, defined sites, defined frequencies, and defined response procedures for positives. Sampling without response procedures is just documenting that you knew. Calibration of thermometers, salometers, pH meters. Phosphatase verification of pasteurization. Finished-product testing for L. monocytogenes per your sampling plan and regulatory criteria. Brine micro testing. Independent review of CCP records. Supplier verification for milk and cultures. Trend all of it, and feed the trends back into the plan.

Step 11: Records

The plan, the hazard analysis, pasteurization records, brine logs, ripening room charts, deviation records, calibration logs, EMP results with investigations, product test results, training records. EMP positives and what you did about them are among the most-scrutinized records in dairy auditing. Make them complete.

Step 12: Reassess — and let the EMP drive it

Annual minimum. Triggers: new variety, new culture, brine system changes, ripening room modifications, any Listeria positive in zone 1 or 2, finished-product positives. Treat EMP trends as a standing input to reassessment — the monitoring program talks to you. Listen.

Things that actually work

Design the EMP before the audit is scheduled. Sites, frequencies, responses — running for months, with history. A program started the week before the audit is transparent, and not in a good way.

Treat brine as a food-contact surface. Because it is. Filter it, heat-treat or replace it, test it — on a schedule that reflects the risk, not the convenience.

Zone ruthlessly. Raw-to-post-pasteurization separation: footwear, tools, clothing, traffic. Every crossover is a Listeria delivery route. Design them out.

Validate the ripening profile. Time, temperature, humidity — shown to control pathogen growth in your cheese. Tradition is not validation, however old the recipe.

War stories

The brine that seasoned everything. A plant’s brine system hadn’t been fully changed in months — “topped up” regularly, tested never. Listeria spp. positives started appearing on product-contact surfaces downstream. The investigation led back to the brine, which had become a pathogen reservoir bathing every cheese. Brine maintenance is a CCP for a reason.

The positive nobody acted on. Listeria spp. found on a ripening room drain. Logged. Filed. Nothing else. Six weeks later, a zone 2 positive on a rack. Still nothing — “it wasn’t monocytogenes.” Here’s the thing: Listeria spp. in the environment means conditions support Listeria, and the response procedure exists precisely for that signal. Finding it and doing nothing is worse than not sampling — it proves you knew.

The condensation rain. Ripening room ceiling dripping onto uncovered cheese during draining. The hazard analysis mentioned “environmental contamination” in general terms but nobody had looked up. Look up. Overheads belong in the analysis and in the sanitation schedule.

Raw-milk thinking in a pasteurized plan. A plant making both raw-milk and pasteurized soft cheese, with one hazard analysis covering both. The raw-milk hazards — which are severe and numerous — got diluted into the pasteurized framework, and neither product was properly assessed. Separate products, separate plans.

Common mistakes

Topping up the brine forever. Months of “topped up,” never tested, never fully changed — until Listeria spp. positives traced back to the brine reservoir bathing every cheese. Brine is a food-contact surface: filter it, heat-treat or replace it, test it on a risk-based schedule.

Filing positives without acting. Indicator Listeria found on a ripening room drain — logged, filed, nothing else. Six weeks later a zone 2 positive; still nothing — “it wasn’t the pathogen.” Any Listeria in the environment means conditions support the genus; finding it and doing nothing proves you knew.

Forgetting to look up. A ripening room ceiling dripping onto uncovered cheese while the hazard analysis mentioned “environmental contamination” in general terms and nobody had looked up. Overheads belong in the analysis and in the sanitation schedule.

Covering raw and pasteurized with one analysis. One hazard analysis for both raw-milk and pasteurized soft cheese — the raw-milk hazards diluted into the pasteurized framework, neither product properly assessed. Separate products, separate plans.

Treating tradition as validation. A ripening profile inherited from the recipe’s origin, never shown to control pathogen growth in this cheese. Tradition is not validation, however old the recipe — validate the profile in your cheese.

Crossing the raw-to-post-pasteurization border. Shared footwear, tools, clothing, and traffic between raw and post-pasteurization sides — every crossover a Listeria delivery route. Zone ruthlessly and design crossovers out.

Checklist — before you call this plan done

  • [ ] Team includes genuine cheesemaking expertise; training documented
  • [ ] Product description states pasteurized milk, cultures, pH/aw targets, shelf life
  • [ ] Flow diagram includes brine system, ripening, rework; walked, signed, dated
  • [ ] Hazard analysis centers on post-pasteurization L. monocytogenes, step by step
  • [ ] CCPs: pasteurization, brine (concentration + temperature + maintenance), ripening/storage temp, metal detection
  • [ ] Brine limits defined numerically with maintenance schedule — not “when dirty”
  • [ ] Monitoring with named owners; continuous recording in ripening and storage
  • [ ] Corrective actions: hold, investigate cause, QA disposition authority
  • [ ] Verification: zone-based EMP with response procedures, calibration, phosphatase, product testing, trend review
  • [ ] EMP trends and positives feed reassessment; annual review scheduled