NEG Food Manufacturing Hub
Process Notes |
ISSUE #004 |
|
| |
|
Deep-Dive · Food Safety & Process Engineering
The Maths Behind Safe Food: How to Validate a Thermal Process
Hitting 72°C for 15 seconds is not thermal process validation. It is a legal minimum for one product category. Validation is the documented, scientific proof that your specific process, running on your specific equipment, destroying your specific target organism in your specific product matrix, achieves the required log reduction every time. If you cannot demonstrate that — you have a compliance gap.
This issue covers the science (D, z, F-values), how to select your target organism, how to size and qualify a hold tube, how to integrate lethality across a real time-temperature profile, and the five questions you should be asking your co-man or equipment supplier before you sign off any thermal process.
|
|
| 00 |
Validation vs. Monitoring: Not the Same Thing |
These two words are used interchangeably in production — they should not be. The distinction has direct regulatory consequences.
| PROCESS VALIDATION |
PROCESS MONITORING |
| Done once (then re-validated on change). Establishes that the process can achieve the required lethality. |
Done every run. Confirms the validated parameters are being maintained in real time. |
| Output: Validation protocol, inoculated pack studies, thermocouples, signed documentation. |
Output: CCP records, hold time logs, temperature charts, deviation reports. |
You need both. Control without validation is assumption. Validation without control is history.
|
| 01 |
The Three Numbers You Need to Know |
Thermal process validation is built on three interrelated parameters. Understanding them is the difference between filling in a form and actually knowing whether your process is safe.
|
D-value — Decimal Reduction Time
Time (minutes) at constant temperature to reduce viable organism population by 90% (1 log cycle). Organism-specific, temperature-specific, and critically — product matrix-specific.
Formula: D = t / log(N₀/N)
Example: D₇₂°C for Salmonella in whole liquid egg ≈ 0.008 min — but in high-fat chocolate: significantly higher. Fat protects.
|
|
z-value — Temperature Sensitivity
°C required to change the D-value by one log cycle. Low z-value = highly temperature-sensitive organism. A small temperature increase dramatically reduces D.
Formula: z = (T₂ − T₁) / log(D₁/D₂)
Listeria monocytogenes z ≈ 7°C · Salmonella z ≈ 5–7°C · C. botulinum spores z ≈ 10°C
|
|
F-value — Equivalent Lethality
Total accumulated lethality expressed as equivalent minutes at a reference temperature. F₀ uses 121.1°C with z = 10°C for sterilisation. For pasteurisation, your reference temp matches your process design temp.
Formula: F = D × log(N₀/N)
A 5-log reduction of an organism with D₇₂°C = 0.008 min requires F = 0.008 × 5 = 0.04 min (2.4 seconds) at 72°C — hence the 15s standard includes a safety margin.
|
|
Connecting the Three
DT = DTref × 10((Tref−T)/z) — Use this to convert D-values between temperatures when your process temperature differs from published literature values.
|
|
| 02 |
Selecting Your Target Organism |
You validate against the most heat-resistant relevant pathogen in your product. "Relevant" depends on pH, water activity, product type, and intended consumer. Get this wrong and your entire validation study is invalid.
| PRODUCT / PROCESS |
TARGET ORGANISM |
LOG RED. |
| Milk (HTST) |
Coxiella burnetii |
≥ 5 log |
| Juice (pH < 4.6) |
E. coli O157:H7 |
≥ 5 log |
| RTE sauce (pH > 4.6) |
Listeria monocytogenes |
≥ 6 log |
| Low-acid shelf-stable |
C. botulinum |
12D cook |
| UHT sterile |
C. botulinum |
F₀ ≥ 3 min |
The pH 4.6 boundary is critical. Below it, C. botulinum cannot produce toxin. Above it, you must treat every product as potentially supporting spore outgrowth.
|
| 03 |
The Hold Tube: Sizing It Correctly |
The hold tube ensures every particle of product is held at the target temperature for at least the minimum required time. Most food safety failures in continuous thermal processes originate here — undersized, improperly insulated, or never actually timed.
|
Process Flow
| FEED PUMP |
→ |
HEATER |
→ |
HOLD TUBE |
| ↓ |
| COOLER |
← |
FDV |
|
FDV (Flow Diversion Valve) — diverts product back to balance tank if temperature drops below the critical limit. The hold tube is the lethality zone. Everything before it is preheating. Everything after is preservation.
|
Sizing the Hold Tube
The hold tube must be sized for the fastest particle, not the average flow velocity. In turbulent flow (Re > 4,000), the correction factor is 0.82 — meaning the fastest streamline moves at 1/0.82 of the mean velocity. In laminar flow, it moves at 2× mean — you need double the tube length. Design always for turbulent flow.
|
Hold Tube Length Formula (Turbulent Flow)
L = (Q × thold) / (0.82 × A)
| L = tube length (m) |
Q = volumetric flow rate (m³/s) |
| thold = required hold time (s) |
A = cross-sectional area (m²) |
|
|
Worked Example
Product: Whole milk | Process: HTST 72°C / 15 s | Flow: 10,000 L/hr | Tube diameter: 50 mm
Q = 10,000 / 3,600,000 = 0.00278 m³/s
A = π × (0.025)² = 0.001963 m²
L = (0.00278 × 15) / (0.82 × 0.001963) = 25.8 m minimum
In practice: size to ≥ 28 m, insulate fully, and verify with physical timing at maximum rated flow.
|
|
| 04 |
Integrating Lethality Across a Real Process |
Your product does not jump instantly from ambient to hold temperature. The come-up and cool-down phases both contribute lethality. Ignoring them is conservative but wasteful. Integrating them is accurate — and required for any high-acid or borderline process where you need every fraction of a log reduction.
|
Step 1 — Lethality Equation (IFTPS)
L = 10((T − Tref) / z) × RT
| L | Lethality — equivalent minutes of exposure to reference temperature |
| T | Product temperature at holding tube exit (°C) |
| Tref | Reference temperature — 121.1°C for sterilisation |
| z | z-Value — 10.0°C for sterilisation |
| RT | Residence time of product in the holding tube (min) |
|
|
Step 2 — Average Particle Velocity
vavg = (4 × Q) / (π × Di²) × CFTherm × CFSteam
| vavg | Average particle velocity (m/s) |
| Q | Product maximum volumetric flow rate (l/h) |
| Di | Internal diameter (m) |
| CFTherm | Correction factor for thermal expansion — default 1.06 (70°C to 143°C); set to 1.0 if flow meter is at inlet |
| CFSteam | Correction factor for direct steam injection — set to 1.0 for indirect heating |
|
CFSteam for DSI (simplified for water at 140°C)
CFSteam = 1 + (1.8 × ΔT) / 1000
ΔT = product temperature rise due to steam addition (°C)
|
|
|
Step 3 — Reynolds Number & Residence Time
Re = (ρ × vavg × Di) / μ
Re < 4,000 → Laminar: CFProfile = 2.0 | Re > 4,000 → Turbulent: CFProfile = 1.2
RT = L / (v × CFProfile)
| ρ | Product density (kg/m³) |
| μ | Fluid viscosity (Pa·s) |
| L | Total length of holding tube (m) |
| RT | Residence time (s) — input back into the lethality equation above |
|
Numerical Integration: How It Looks in Practice
| Time (s) |
Temp (°C) |
Phase |
L(T) at z=7, Tref=72 |
F contrib (s) |
| 0 |
62 |
Come-up |
0.034 |
0.034 |
| 1 |
67 |
Come-up |
0.198 |
0.198 |
| 2–17 |
72 |
Hold |
1.000 |
15.000 |
| 18 |
68 |
Cool-down |
0.278 |
0.278 |
| Total F-value |
≈ 15.51 s |
This is how you generate your F-value from a real plant trial. If F > D × required log reduction, your process is validated at those conditions.
|
| 05 |
5 Questions to Challenge Your Co-Man or Supplier |
A co-manufacturer or equipment supplier who cannot answer these clearly has either never validated properly or is hoping you won't ask. Ask.
| 1 |
"What was your target organism and on what basis was it selected?"
The answer must reference your specific product pH, water activity, and regulatory category — not a generic organism they use for everything.
|
|
| 2 |
"What D-value did you use, and is it referenced to our product matrix?"
D-values from literature are measured in buffer or growth media. Your high-fat, high-sugar, or high-protein matrix will change them. They should know this.
|
|
| 3 |
"Can I see the hold tube length calculation and the timing verification protocol?"
If the hold tube was sized from a manufacturer's spec sheet without a site-specific timing study at maximum flow, the validation is incomplete. Full stop.
|
|
| 4 |
"How did you handle lethality credit in the come-up zone?"
If they say “we don't count it” — fine, that's conservative and safe. If they count it without a measured temperature profile and numerical integration, that's a problem.
|
|
| 5 |
"When was the validation last re-run, and what change triggers a revalidation?"
Formulation changes, equipment swaps, flow rate changes, and CIP cycle modifications can all invalidate a prior study. If there's no change-control trigger list, the validation is static documentation in a dynamic process.
|
|
|
|
From the Process Notes Resource Library
HACCP Template Pack — EU & FDA Edition
10 audit-ready templates covering everything from hazard analysis to CCP monitoring records — dual-jurisdiction, aligned to EU Regulation (EC) No 852/2004 and FDA 21 CFR Part 117. Word docs + interactive Excel workbooks included.
|
|
Already covered HACCP from first principles in Issue #002 — these templates are the practical output of that framework, ready to fill in and submit.
|
|
|
🎁 Free Download
The HACCP Starter Checklist
Before you write a single page of your HACCP plan, you need to know if your foundation is solid. This one-page checklist covers the 12 prerequisites you should have in place before starting your hazard analysis — from supplier controls and sanitation SOPs to allergen management and training records. Print it, walk the floor, and know within 20 minutes where your gaps are.
|
|
|
Next Issue — #005
Scale-Up Without Breaking the Product
A retailer wants 50,000 units. The six variables that derail most food scale-ups — and the pilot trial protocol that catches problems before they reach production.
|
|
If this was useful — forward it to one food engineer who needs it. That's how this grows. And if you have a process validation headache you can't solve, hit reply. I read every one.
Galvin Eyong
Food System Expert · Process Notes
processnotes.beehiiv.com
|
Process Notes
Practical food engineering, every Tuesday. |
|
|
