quality control checklist for coffee bean exports

A practical, field-tested water activity protocol for export-ready green coffee. Sampling plans, meter setup, pass/fail thresholds, troubleshooting high aw, and packaging/shipping steps that prevent mold and OTA claims.

As an export team that’s shipped Indonesian coffee into humid ports and dry warehouses alike, we’ve learned something the hard way: water activity is the fastest, most reliable early-warning signal for mold and OTA risk in green coffee. Nail your aw protocol and you prevent most post-shipment headaches.

The 3 pillars of reliable water-activity control

Clear sampling and acceptance plan. You can’t control what you don’t measure consistently. Decide how many bags per lot you’ll test and what triggers a hold.
Meter discipline. Calibrate, equilibrate temperature, and handle samples the same way every time.
Packaging and transit risk management. Even “passing” lots can fail in-transit if you load hot coffee into a cold container or skip desiccants.

We’ll walk you through the checklist we use before any container leaves our warehouses in Java, Sumatra and Bali.

Week 1–2: Lot validation, sampling plan and targets

In the two weeks leading up to stuffing, we lock specs and test representative samples. That’s because aw can drift during conditioning and bagging.

Targets we use. For specialty shipments, water activity ≤ 0.60 aw. We prefer 0.50–0.58 for added stability. For commercial-grade and robusta, we still push for ≤ 0.60. Moisture content follows SCA guidance at 10.0–12.0%. We hold any lot above 12.5% MC or 0.62 aw.
Why aw and moisture both. Moisture content is the total water in the bean. Water activity measures how “available” that water is for microbial growth and chemical reactions. Two coffees at 11.5% MC can behave very differently if one is 0.52 aw and the other is 0.66 aw. The latter is a mold/OTA risk in warm transit.
Sampling plan by lot size. For micro-lots ≤ 30 bags: test 3–5 bags. For 100–300 bags: test 8–10 bags. For 300–600 bags: test 13–15 bags. Pull probes from top/middle/bottom of each selected bag and composite per bag before testing. If any reading ≥ 0.62, pull two additional bags from adjacent pallets. If two or more measurements in the lot land ≥ 0.62, place the lot on hold and investigate.

Practical takeaway: plan your sampling early, and write down pass/fail triggers. Don’t improvise on stuffing day.

Week 3–6: Meter setup and a repeatable aw testing protocol

Here’s the protocol our QC rooms follow to the letter.

Instrument choice. Lab-grade dew-point meters like METER Group AQUALAB 4TE or Novasina LabTouch-aw are rock solid and fast. Portable sets such as a Rotronic HP23-AW kit or Novasina LabSwift-aw can work for field checks, but verify against a lab meter weekly.
Calibration and verification. Check daily with two salt standards bracketing coffee’s range, typically 0.500 and 0.760 aw. Record drift and clean the chamber with 70% ethanol if residues appear. If you see >0.003 deviation on either standard, recalibrate before proceeding.
Sample prep. Do not grind green coffee for aw tests. Grinding changes surface dynamics and can inflate readings. Use whole beans or lightly crack a few larger beans so they sit flat. Fill the cup about two-thirds, avoiding sensor contact.
Temperature equilibration. Water activity is temperature dependent. Measure at 25 °C when possible. If your room is 22–32 °C, let the sample equilibrate in the chamber until the instrument reports a stable temperature and reading. In our rooms at 28 °C, whole beans usually stabilize in 10–20 minutes. Rushing this step is the number one cause of noisy data. If the sample came from a cooler or a hot roaster room, give it 20–30 minutes.
Replicates and recording. Take two readings per bag. If they differ by >0.010 aw, take a third and average the closest two. Log lot ID, bag number, aw, temperature, moisture content, date/time, operator, and instrument serial.

Not obvious but critical: keep fingers off the inside of the cup. Skin oils can bias readings upward by a few thousandths, which is the difference between pass and hold in tight specs.

What is a safe water activity for green coffee beans?

For shipments with weeks of ocean transit, ≤ 0.60 aw is a practical ceiling. We like 0.50–0.58 aw. Above ~0.65 aw and warm conditions, microbial growth risk rises quickly.

How is water activity different from moisture content in coffee?

Moisture content is how much water is present. Water activity reflects how “free” that water is. Mold needs available water, not just total water. That’s why 11.5% MC coffee at 0.66 aw is riskier than 12.0% MC at 0.54 aw.

Which water activity meter is best for testing green coffee?

If you run a lab, a dew-point system like AQUALAB 4TE or Novasina LabTouch-aw gives fast, repeatable, temperature-corrected results. For field and warehouse checks, Rotronic HP23-AW or Novasina LabSwift-aw are dependable when verified against a lab unit weekly.

Week 7–12: From pass/fail to shipment-ready

This is where your measurements meet packaging and ocean reality.

Hermetic bags. Hermetic liners like GrainPro stabilize water activity by blocking moisture exchange. They don’t lower aw by themselves. So don’t use them to “fix” a high-aw lot. Use them to keep a good lot good.
Pallet conditioning. After bagging, give pallets 24–48 hours in a low-RH room with gentle airflow. We’ve seen aw fall by 0.01–0.03 simply by allowing temperature and humidity to equalize across the stack.
Container stuffing. Avoid condensation. If coffee is warmer than the container air by more than ~8–10 °C, moisture can condense on container walls and migrate. Target bean temperature within 5 °C of ambient container temperature. Use 1–2 kg of desiccant per cubic meter as a starting point, and keep bags off walls with dunnage. Ventilate the container before loading on humid mornings.
Documentation. Attach aw and moisture reports to the pre-shipment QC pack. Many buyers now specify aw in the contract alongside SCA moisture limits.

Practical takeaway: the best aw number on paper won’t survive a dew-point trap. Manage temperature differences and humidity inside the container.

How many samples should I test per lot before shipment?

Use the plan above. As a shortcut for busy teams: at least 3 bags for micro-lots, 8–10 for 100–300 bag lots, and 13–15 for bigger lots. Any reading ≥ 0.62 triggers expanded sampling.

Does water activity predict mold and OTA risk during ocean transit?

Aw is the best early indicator we have. Lots ≤ 0.60 aw, packed hermetically and stuffed without condensation, almost never produce OTA issues for us. EU limits remain strict on roasted and soluble coffee, so importers increasingly screen green lots to avoid downstream rejections.

Troubleshooting: what to do if your coffee’s aw is too high

Recondition with air. Spread bags or supersacks in a low-RH (50–60%), 25–30 °C room with fans for 24–72 hours. Retest daily. This is often enough to bring a 0.62 down to 0.57–0.59.
Gentle re-dry. If MC is also high (>12.5%), use a controlled dryer at ≤ 40 °C with good airflow. Over-drying flattens the cup and can create case-hardening. We monitor MC every 2–3 hours.
Reclean and remove outliers. Visually wet or underdried pods can drive up aw. A recleaning pass to remove quakers/defects sometimes drops aw measurably.
Hold or blend only with buyer approval. Blending a 0.64 aw lot with a 0.52 aw lot to “average” the number is risky and can hide pockets of high-aw beans. We only consider blending when traceability allows and the buyer agrees.

If you’re wrestling with a borderline lot or spec-setting for a new buyer, we’re happy to share what’s been working across Sumatra wet-hulled and Bali washed profiles. Need a second opinion? Contact us on whatsapp.

Do hermetic bags lower water activity or just stabilize it?

They stabilize. If your lot is at 0.62 aw, sealing it hermetically preserves 0.62. It won’t pull moisture out. Condition first, then seal.

Common mistakes that create claims (and how to avoid them)

Testing too soon after moving coffee. Beans just out of the sun or a cool store read erratically. Let temperature equilibrate before testing.
Relying on moisture content alone. A “perfect” 11.5% MC lot can still mold if aw is high. Always measure both.
Dirty or uncalibrated meters. A 0.01 aw drift is the difference between stable and risky. Calibrate daily with two salts and log results.
Loading hot coffee into a cool container. That fog on the wall becomes moisture in the bag. Match temperatures and use desiccants.
Skipping hermetic liners for long transit. Jute alone breathes. You’ll track ambient humidity changes across the ocean.

Quick reference: pass/fail and records

Pass. ≤ 0.60 aw and 10.0–12.0% MC. Prefer 0.50–0.58 aw for long routes or monsoon season.
Caution. 0.60–0.65 aw. Retest, condition, and hold until trending down.
Fail/hold. ≥ 0.65 aw or ≥ 12.5% MC.
Record template. Lot ID, farm/region, process, bag count, sampling plan, aw (two replicates, temp), MC method/result, instrument ID, standards verification, date/time, operator, corrective actions, final decision.

Where this applies (and a few real-world notes)

Wet-hulled Sumatra lots can show wider aw variability between bags. We sample more intensively on these, including Sumatra Mandheling Green Coffee Beans and Sumatra Lintong Green Coffee Beans (Lintong Grade 1). Fully washed Grade 1 lots like Arabica Java Ijen Grade 1 Green Coffee Beans or Arabica Bali Kintamani Grade 1 Green Coffee Beans tend to be more uniform at the same MC, but we still verify bag-to-bag to catch outliers. Naturals such as Bali Natural Green Coffee Beans benefit greatly from hermetic liners during export months with high ambient humidity.

If you want to see how we specify aw in contracts or compare stability across Indonesian origins, browse our lots and spec sheets. View our products.

Bottom line. Water activity is your simplest lever to de-risk export quality. Build a disciplined sampling plan, respect temperature equilibration, and ship in a way that keeps a good aw number intact. Do those three and you’ll avoid most mold and OTA surprises, even on long ocean routes.