What to Adjust When Lots Fail: Processing Parameters for Herbal Products

A lot fails and the pressure arrives all at once. One person wants to turn the heat up, another wants to blend the issue away, and someone else is already asking whether the ship date still holds. In practice, this is the real test of quality culture: not how quickly a team reacts, but how clearly it thinks.

Key Takeaway: When a botanical lot fails, work in order: confirm the result, classify the failure, then change only the process levers that genuinely match that failure type. Moisture and microbial issues usually call for drying, handling, and storage changes. Assay and fingerprint drift usually point to extraction design and heat exposure. Physical test failures are often mechanical before they are formula-related. Stability drift asks for better process control and packaging. Identity and contaminant failures belong upstream in sourcing, not in salvage.

Step 1: Confirm the Failed Lot Before Changing Anything

First, slow the room down. Confirm the failure before touching the process.

A short verification window protects both the team and the material. Good lot evaluation includes retesting unusual results before rejecting a lot, especially when sampling error, instrument drift, or a handling slip could be involved.

Repeat the analysis with the same method using a fresh, representative composite sample, ideally with a second analyst. In parallel, review calibration status, reference standards, system suitability, and the batch record. The goal is confidence in the result, not just a new number.

Open a deviation record while you verify. It keeps the investigation traceable and prevents a common mistake: changing multiple variables before anyone has established what actually failed.

Example: a chamomile lot shows moisture above spec on the first read. A repeat composite sample comes back inside range. The issue turns out to be top-layer sampling from bags left open in humid air overnight. The right correction is tighter sampling and sealed staging, not hotter drying.

• Retest with a representative sample
• Verify calibration and standards
• Review critical process notes
• Pause inventory movement if needed
• Open the deviation before making adjustments

Step 2: Classify the Failure So the Right Levers Become Obvious

Not every failure gives you the same options. The type of failure determines the type of response.

Name the category clearly: identity, assay, microbial, physical, stability, or contaminants. That single step prevents a lot of waste and a lot of ethically questionable “fixes.” A low marker is different from a wrong species, and a moisture miss calls for a different response than oxidation drift.

• Identity failures involve the wrong species, wrong plant part, or a non-matching profile.
• Assay failures involve markers outside the agreed range.
• Microbial failures involve counts above limits.
• Physical failures involve flow, hardness, friability, disintegration, fill weight, viscosity, appearance, or moisture behavior.
• Stability failures involve drift over time: marker loss, off-odors, color change, moisture creep, or oxidation.
• Contaminant failures involve unwanted residues, foreign matter, or other unacceptable inputs.

Example: a holy basil extract has the expected aroma and identity markers, but the eugenol result falls below target while microbes and moisture are fine. That is an assay failure, so the focus belongs on extraction and blending decisions.

Step 3: Correct Moisture and Microbial Failures Through Drying and Handling

When the issue is water, your primary tools are drying, handling, and storage.

Water activity is especially useful here because most microbial growth is inhibited below about 0.60 aw. Moisture control is a stability practice as much as it is a cleanliness practice.

For many botanicals, the practical work starts before the dryer. Harvest timing, cut size, load depth, airflow, and staging conditions strongly shape whether a lot stays steady or moves toward mould. Traditional know-how matters here: cooling the field heat, drying evenly, and sealing promptly after cooling are time-tested moves that protect plant integrity.

Set drying endpoints that suit the material rather than drying “until it looks done.” Dense roots and barks often need a different endpoint than delicate aerial parts, and house standards built from experience are valuable.

Then watch reabsorption. Clean material can pick up moisture again during transfer, milling, staging, or storage. Barrier liners, sealed containers, pallets off concrete, and a dry, steady warehouse environment often solve more than another round of heat.

Worked example: a lemon balm lot keeps failing for mould on intake. The team moves harvest to dawn, reduces field heat, uses forced-air drying, cools fully before packing, and stages the lot in sealed liners instead of open bins. The next lots pass comfortably.

• Review harvest timing
• Check cut size and dryer load depth
• Track both moisture and aw where possible
• Improve sanitation at slicers, mills, and conveyors
• Seal cooled material promptly
• Control warehouse humidity

Step 4: Improve Assay and Fingerprint Results Through Extraction Design

When markers or fingerprints drift, the answer is usually in extraction design.

Extraction outcomes are strongly influenced by solvent choice, solvent-to-herb ratio, time, temperature, and related process variables. When one of those shifts, the profile tends to shift with it.

This shows up clearly in hydroalcoholic work, where ethanol concentration changes can noticeably alter what you pull from the plant. A small percentage change can move an assay even when incoming raw material is consistent.

Heat deserves equal respect. During concentration or drying, heat-sensitive constituents can be lost when temperatures run too high or too long. Many teams focus tightly on extraction and then unintentionally undo good work downstream.

Blending can help when used with discipline. In-process checks on intermediate extracts and pre-final blends catch variation early, when small, intentional adjustments still keep the finished preparation coherent and true to its intended character.

Example: a hawthorn extract shows low vitexin. You increase hydroalcoholic strength, adjust solvent ratio, add a short second pass, and keep concentration temperatures gentler. The pilot returns to range without flattening the broader fingerprint.

• Review solvent composition first
• Check ratio, time, temperature, and pH together
• Examine concentration and drying temperatures
• Use intermediate checks before final blending
• Document every approved process change clearly

Step 5: Handle Physical Test Failures with Mechanical Adjustments First

A batch can be chemically sound and still fail physically. When that happens, mechanical fixes usually deserve first attention.

Plant-based powders are naturally variable, so flow, compressibility, and fill behavior often need tuning. Particle size strongly influences flow, segregation, compressibility, and dissolution, so it belongs near the top of the checklist.

For tablets, compression force changes hardness and friability, and excessive compaction can slow disintegration. For capsules, machine settings affect fill weight alongside powder flow properties, so tamping and dosing speed should be reviewed before assuming the blend is the problem.

Day to day, this means checking grind distribution, room humidity, blending uniformity, granulation behavior, static, and equipment setup. Small corrections often restore performance faster than changing the composition.

Example: an ashwagandha capsule run jams and shows poor fill consistency. Investigation reveals a very broad grind and a dry room encouraging static. The team re-mills to a tighter distribution, improves flow, and slows the dosing setup. The run stabilizes.

• Tighten particle size distribution
• Review room humidity and static
• Check blend homogeneity
• Retune compression or capsule settings
• Try mechanical corrections before altering the formula

Step 6: Solve Stability Drift with Better Process Protection and Packaging

If a lot passes at release but drifts later, the product is telling you something about protection.

For shelf-life decisions, real-time stability studies are primary, while accelerated conditions play a supporting role. With botanicals, moderate, lived storage conditions often reveal more than extreme stress.

Review residual moisture, oxygen exposure, light exposure, headspace, pH where relevant, and the barrier quality of the pack. Water-containing formulas deserve especially careful preservation planning because they are technically demanding.

Oxidation-prone products need extra care. Nitrogen flushing, reduced headspace, light-protective packaging, and antioxidants can help slow peroxide rise during storage. For oils, aromatic preparations, and other sensitive materials, these steps protect quality rather than “upgrading” it.

Example: a calendula-infused balm begins showing peroxide drift at month six. The team reduces headspace, switches to amber packaging, nitrogen-flushes the container, and adds antioxidant support. The next stability cycle is far steadier.

• Review release moisture and aw
• Reduce oxygen exposure and excess headspace
• Choose light-protective packaging
• Revisit antioxidant support where appropriate
• Base dating on rolling real-time data

Step 7: For Identity or Contaminant Failures, Fix Sourcing — Not the Batch

Some failures are not invitations to adjust the process. They are signals to reject the material and improve sourcing.

If the plant is misidentified, mixed with another species, or outside your agreed identity window, authenticity has already been compromised. The right response is stronger supplier qualification, clearer botanical and plant-part specifications, and layered identity checks that match the sourcing pathway.

The same principle applies to contaminant failures. When the issue begins upstream, trying to engineer around it later undermines trust and usually creates new risks.

Build better relationships and better standards: seasonal expectations, post-harvest handling instructions, traceable documentation, and verification steps before a lot reaches full production. This is how respect for the plant becomes visible in operations.

Example: a lot sold as saw palmetto shows an off-profile and mixed identity. The team rejects the material, upgrades incoming requirements, and adds a pre-shipment verification step. Later lots arrive cleaner and far more consistent.

• Write clear botanical and plant-part specifications
• Strengthen supplier qualification
• Use layered identity checks where appropriate
• Clarify post-harvest expectations
• Reject lots that compromise authenticity

Step 8: Turn Failures Into Better Systems

A failed lot is expensive. It should also be instructive.

Strong teams do more than close a deviation. They build learning into the system through complete batch records, clear CAPAs, useful trend charts, and periodic review of what repeats.

Over time, this shows which ranges are realistic, which steps are fragile, and which materials demand gentler handling. It also supports honest communication. If a meaningful process change shifts the fingerprint or overall character of a preparation, the language around it should evolve too.

One habit helps: hold regular reviews of recent deviations, compare before-and-after outcomes, and update the relevant SOPs while the details are still fresh.

• Record process details completely
• Trend failures across batches, not in isolation
• Update CAPAs and SOPs from real events
• Review claims when process changes alter product character
• Let the system become wiser over time

Conclusion: Adjust What Fits the Failure, Protect What Matters

When a botanical lot fails, the most useful response is rarely the fastest one. Confirm the result, name the failure clearly, then adjust only the levers that truly fit.

Moisture and microbial problems usually ask for better drying, handling, and storage. Assay and fingerprint drift usually ask for tighter extraction control and gentler downstream processing. Physical failures often improve through particle size control and equipment tuning. Stability drift points toward stronger protection in process and packaging. Identity and contaminant failures call for stronger sourcing and the willingness to say no.

Traditional practice has long treated plant quality as something you protect through careful handling, respectful processing, and clear standards. Modern testing adds useful visibility, but the deeper pattern stays the same: resist shortcuts, correct the right root cause, and keep learning visible in the system.

As a final note, good decision-making here relies on solid sampling, realistic specifications, and a team culture that documents changes as carefully as it makes them. When those foundations are in place, failed lots become less common, and the work becomes steadier for everyone involved.