Freeze-dried durian and a low-moisture product evidence file
Siam Diamond

Freeze-dried durian, low-moisture control and shelf-life evidence

Crisp Freeze-Dried Durian Is Not Proof of Safety Throughout Shelf Life

Crispness and a single moisture result describe a sample at one moment; they do not validate microbial controls, the process, packaging or the full shelf life. This article proposes a seven-layer Low-Moisture Product Evidence File without inventing product limits or brand certifications.

1. Crispness describes eating quality, not the complete safety case

Crispness can change as a product picks up moisture and with formulation, storage and pouch integrity. It is a useful quality attribute, but a crisp piece does not show that contamination never occurred or that microbial controls worked as designed.

A safety decision begins with the hazards in the actual raw materials and process—not appearance, texture or a photograph of one piece.

2. Moisture content and water activity are not interchangeable

Moisture content describes the total amount of water, while water activity reflects water available to participate in reactions or support microorganisms. Their relationship depends on the composition and structure of the food, so one result should not automatically stand in for the other.

This article does not assign a target value to a Siam Diamond product. The formula, process, test method, applicable rules and destination market would need to be established first.

Freeze-dried durian, a sealed pouch, a measuring instrument and a low-moisture product evidence file
Editorial illustration; it is not a test result, shelf-life validation or evidence that Siam Diamond makes or sells freeze-dried durian.

3. Preventing growth is not the same as eliminating a pathogen

Codex CXC 75-2015 explains that a pathogen such as Salmonella may be unable to multiply under low-moisture conditions yet remain viable for an extended period. Microorganisms can also be more heat tolerant in food matrices at reduced water activity.

Terms such as dry, freeze-dried or crisp are therefore not a kill step and must not replace an assessment of the controls that address the real hazard.

4. Process validation differs from end-product testing

Codex recommends that, where feasible, a microbial reduction treatment be validated for the food and operational-scale process. Monitoring and verification then demonstrate that the process continues to operate within its established limits.

The same Code notes the limited information that end-product testing provides about hygiene-control effectiveness and says environmental monitoring should be considered as a verification measure.

5. The pouch and time belong in the evidence system

Even when product leaves a dryer in its intended condition, moisture ingress, microleaks, incomplete seals, oxygen, light, temperature and repeated opening can change quality or risk during storage.

Shelf-life substantiation should therefore connect material structure, seal settings, storage conditions, sampling and trends across relevant tests to the actual formula and lot—not rely on a projected date or one endpoint reading.

6. The seven-layer Low-Moisture Product Evidence File

An auditable file should keep each evidence layer distinct while linking all of them through the same formula code, process version, packaging material and lot.

  • Product definition: formula, materials, intended use, process and the rationale for applying a low-moisture-food framework
  • Hazard analysis: raw-material and supplier risk, foreign material, allergens and relevant microbial hazards
  • Process evidence: validated reduction or other controls, monitoring conditions, verification and deviation handling
  • Post-process hygiene: zoning, people and material flows, cleaning method, water control and risk-based environmental monitoring
  • Packaging system: material properties, seal integrity, lot coding, artwork and version control
  • Stability study: storage plan, time points and risk-selected sensory, chemical, physical and microbiological attributes
  • Lot release and lifecycle: actual lot results, deviations, retains, complaints, recall readiness and change control

7. Thai regulatory context and the boundary of this conclusion

Thai FDA's Food Division explains that GMP 420 addresses food production processes, equipment or utensils and storage practices within its stated scope. The same page lists an April 2026 updated guide for final-product analysis used to verify GMP-system effectiveness.

Our evidence-design inference is that establishment or system verification should not be substituted for process validation, shelf-life substantiation or a decision on a particular lot. The real product and applicable requirements must be classified before business use.

Origin context and source record

  1. สำนักงานคณะกรรมการอาหารและยา — กองอาหารGMP 420 และคู่มือรายการตรวจวิเคราะห์ผลิตภัณฑ์สุดท้าย — คู่มือปรับปรุงเมษายน 2569
  2. ราชกิจจานุเบกษาประกาศกระทรวงสาธารณสุข (ฉบับที่ 420) พ.ศ. 2563 เรื่อง วิธีการผลิต เครื่องมือเครื่องใช้ในการผลิต และการเก็บรักษาอาหาร
  3. Codex Alimentarius (FAO/WHO)CXC 75-2015: Code of Hygienic Practice for Low-Moisture Foods — revised 2016, amended 2018
  4. Codex Alimentarius (FAO/WHO)Official Codes of Practice list — CXC 1-1969 and CXC 75-2015

References explain industry context as of publication; they do not constitute product endorsement or legal advice.

Frequently asked questions

01Is freeze-dried durian safe if it remains crisp?

Crispness alone cannot answer that. Hazards, process controls, hygiene, packaging, storage and lot evidence must be considered together.

02Can one moisture or water-activity result prove shelf life?

No. One point describes one sample at one time. Shelf-life substantiation needs a study plan, methods, criteria and linkage to the real formula, package and storage conditions.

03Does a GMP certificate prove every lot complies?

No. Establishment-system evidence, product scope, process validation and actual lot-release evidence are separate layers that must be checked independently.