Cleaning Validation Procedure for Tablets: Step-by-Step, With Every Calculation
A working procedure for validating the cleaning of tablet manufacturing equipment, from grouping products and setting residue limits to sampling, recovery studies, acceptance criteria and ongoing verification.
01Purpose and scope
The purpose of this procedure is to give documented evidence that the approved cleaning procedure removes active residue, cleaning agent and microbial contamination from tablet equipment to levels that are safe for the next product made on it.
Its scope covers shared product-contact equipment across the tablet process: granulation and drying, milling and blending, compression, and film coating, plus any transfer containers and tooling. A typical protocol evaluates the cleaning of the granulation, compression and packing equipment using validated chemical and microbiological methods.10
It builds directly on the general cleaning validation post earlier in this series, which explained the principles. This post turns them into a procedure you can follow and adapt, with the calculations shown in full.
Cleaning Validation: A Practical Approach — Destin A. LeBlanc
A practitioner's guide to building a cleaning validation program, covering residue limit calculations, sampling and recovery studies, and how to write protocols and reports.
It is the best companion to this procedure, since it works through the same sequence of decisions and explains the reasoning behind each limit and sampling choice.
02Regulatory basis
| Document | Body | What it contributes |
|---|---|---|
| Guide to Inspections of Validation of Cleaning Processes (1993) | U.S. FDA | Expects a written, scientifically justified cleaning procedure with residue limits and validated methods1 |
| EudraLex Vol. 4, Annex 15 (Qualification and Validation) | European Commission | Sets cleaning validation expectations, with limits based on toxicological evaluation2 |
| EMA guideline on health-based exposure limits and PIC/S PI 046-1 | EMA / PIC/S | Defines the permitted daily exposure (PDE) approach for shared facilities2,3 |
| Industry guidance (APIC, IPA) | Industry bodies | Practical worked examples of grouping, worst case selection, MACO and swab limits6,7 |
The overall sequence is consistent across sources: define the procedure, choose the worst case, calculate limits, clean, sample, test, confirm with three successful runs, and document.11
03Responsibilities
| Function | Responsibility |
|---|---|
| Production | Executes the cleaning procedure as written and records every step, time and parameter |
| Quality Control | Validates analytical methods, performs sampling analysis, reports results against limits |
| Microbiology | Performs bioburden or other microbial testing where the protocol requires it |
| Quality Assurance | Approves the protocol and report, handles deviations, owns the matrix and revalidation triggers |
| Toxicology / EHS | Supplies or reviews the PDE for each active |
| Engineering | Provides equipment drawings and surface areas, and confirms qualified status |
04The validation procedure, step by step
Work through the ten steps in order. Click each to expand it.
List every product-contact item for each tablet process route and group them into equipment trains, such as granulator → dryer → mill → blender → press → coater. Record the product-contact surface area of each item in cm², from engineering drawings.
- Include tooling, hoppers, feed frames, chutes and transfer containers
- Confirm the equipment is qualified before cleaning validation starts
Build a product matrix with dose, PDE, batch size, solubility and cleanability for every tablet product. Group similar products, then rate them to identify the worst-case previous product. The exact method is in section 07.
- Oral solids can form one group, with potent products treated separately6
- Reassess the worst case whenever a new product is introduced
Calculate the maximum allowable carryover (MACO) by the health-based method, and where your policy requires, by the dose-based and 10 ppm methods. Use the most stringent. Then convert it to a surface limit, a swab limit and a rinse limit (Equations 1 to 9).
- Use the smallest batch size and the largest daily dose of the next product for the worst case
- Also set limits for cleaning agent and microbial levels
Choose swab sampling for risk-assessed, hard-to-clean locations and rinse sampling for large or inaccessible surfaces, and use visual inspection on everything accessible. Most protocols combine them.4
- Define each location, the swab area (commonly 25 cm²) and the number of swabs
- Include a justification for every location in the protocol
Select a specific method such as HPLC, or a non-specific one such as TOC where justified. Validate it for specificity, linearity, accuracy, precision, and a quantitation limit comfortably below the swab limit from Equation 6.
- The method must detect residue at the limit with margin
- See the analytical method validation posts earlier in this series
Spike known amounts of the worst-case active onto coupons of each surface material (stainless steel, coated steel, silicone, polymer), sample with the defined technique, and measure what comes back (Equation 7). Apply the recovery to later results (Equation 8).
- Test every surface type in the train, not just stainless steel
- Use replicate spikes, ideally at several levels around the limit
The protocol fixes the products, equipment, cleaning SOP, limits, sampling locations, methods, hold times and acceptance criteria before any run. Quality approves it before execution.
- Acceptance criteria are never changed after the data is seen
- Define what happens on a failure, including the investigation route
Manufacture the worst-case product, clean strictly per the SOP, perform visual inspection, take swab and rinse samples, and test. Repeat until three consecutive runs meet all criteria.11
- Cover the validated dirty and clean hold times
- Record everything contemporaneously
Correct each result for recovery, compare with the limit, and confirm the total carryover across the train is below MACO. The report states the conclusion, any deviations, and the validated state of the cleaning procedure.
- Any failure triggers an investigation and a root cause before more runs
- The report defines the worst-case products and equipment now covered
Set up periodic verification, trending and defined revalidation triggers: a new product, a change in equipment or cleaning agent, a change to the cleaning procedure, or a failure. New products re-evaluate the worst case and recalculate affected limits under change control.5
- Link to the change control and APQR posts earlier in this series
05Tablet equipment and sampling locations
Each stage of tablet manufacture has its own hard-to-clean areas. Switch tabs to see typical locations, noting that the final list must come from your own risk assessment.
Granulation and drying. Wet granulation leaves sticky, caked residue, so bowls, impellers, choppers, spray nozzles and fluid bed product containers, filters and distributor plates are typical worst-case spots. Bioburden is also a consideration after aqueous processing.6
Milling and blending. Blenders have discharge valves, corners and internal welds that retain powder, and mills have screens and hammers that are hard to access. These are targeted swab locations.9
Compression. Common locations are punch faces (upper and lower), the die table and turret surfaces, the hopper and feeder assembly interior, and scraper blades and product chutes, where powder builds up.8,9
Coating. Pan surfaces, spray assemblies and solution lines are the typical hard-to-clean parts. Swab plus rinse sampling is often justified, since spray lines can only be rinsed.9
For rinse sampling, collect the final rinse from a defined outlet after rinsing with purified water. Protocols commonly collect a fixed volume, for example 500 mL, from the rinse outlet of the fluid bed or compression machine.10
Pharmaceutical Dosage Forms: Tablets — Larry L. Augsburger & Stephen W. Hoag (Eds.)
A multi-volume reference on tablet formulation and manufacturing, including granulation, compression and coating equipment and how each works.
Knowing how the equipment is built is what lets you identify hard-to-clean geometry and justify sampling locations in the protocol, which is why it supports this section directly.
06The cleaning procedure itself
Validation tests a written cleaning SOP, so the SOP has to be specific enough to repeat. A tablet equipment SOP typically follows this sequence.
- Line clearance. Remove all product, labels and in-process material, and tag the equipment "to be cleaned" with the previous product recorded.
- Dry cleaning. Remove loose powder by vacuum with a HEPA-filtered system. Avoid blowing with compressed air, which spreads dust.
- Disassembly. Strip removable parts: punches, dies, feed frame, hopper, scrapers, chutes, screens, filter bags.
- Pre-rinse. Rinse with water to remove gross soil, and capture the rinse as waste.
- Detergent wash. Wash with the specified cleaning agent at defined concentration, temperature and contact time, with the defined mechanical action, using validated brushes and cloths.
- Rinse. Rinse repeatedly with the specified water grade until the defined endpoint, and record the volumes.
- Final rinse. Final rinse with purified water, with a sample taken here for rinse testing.
- Drying. Dry parts by a defined method and time, since wet equipment supports microbial growth.
- Visual inspection. Inspect all surfaces under adequate light, including hard-to-see areas, and sign off.
- Reassembly and labeling. Reassemble, label as "clean" with date, time and cleaning record reference, and start the clean hold time.
The four variables of any cleaning step, time, action, concentration and temperature, should each be defined and controlled, because changing one changes the result. Many sites show them as the four sectors of the Sinner circle.
07Worst-case product selection
The worst-case previous product is the one hardest to clean and with the lowest residue limit. Rate each product on the factors below, add the ratings, and take the highest total. Worst-case selection can be based on risk assessment and rating, and one published example identified its worst case by the highest total risk priority number across the products.6
| Product | Solubility in cleaning solvent (5 = poor) | Toxicity / PDE (5 = lowest PDE) | Cleanability (5 = hardest) | Lowest batch size ÷ largest daily dose (5 = lowest) | Total |
|---|---|---|---|---|---|
| Tablet A | 4 | 5 | 3 | 4 | 16 |
| Tablet B | 2 | 3 | 2 | 3 | 10 |
| Tablet C | 5 | 2 | 4 | 2 | 13 |
In this illustration, Tablet A would be the worst case and would carry the validation for its group. The ratio of batch size to largest daily dose matters because the lowest value gives the lowest MACO per dose for a given next product.6 Scoring is a screening tool and should always be backed by a written rationale.
08Calculations and equations
Each equation has a worked example using one consistent scenario, so you can follow the chain from MACO to the swab result. Scenario: Product A (previous, worst case) has a PDE of 0.5 mg/day and a lowest standard daily dose of 5 mg. Product B (next) has a minimum batch size of 100 kg and a maximum daily dose of 2000 mg/day. The shared product-contact surface area of the train is 40,000 cm².
Health-based MACO (PDE method)
PDEA = permitted daily exposure of the previous product (mg/day), MBSB = minimum batch size of the next product (mg), MDDB = maximum daily dose of the next product (mg/day).4,5
Worked. MBS = 100 kg = 100,000,000 mg. MACO = 0.5 × 100,000,000 / 2000 = 25,000 mg.
Dose-based MACO (1/1000 of the lowest dose)
STDA = lowest standard daily dose of the previous product (mg/day), SF = safety factor, traditionally 1000.4,5
Worked. MACO = 5 × 100,000,000 / (1000 × 2000) = 250 mg.
10 ppm criterion
Ten parts per million is ten milligrams of residue per kilogram of the next product.4
Worked. MACO = 10 × 100 = 1000 mg.
Selecting the limit
Use the most stringent value your policy allows. Where health-based limits are required, MACOPDE is the controlling basis, and the other two are shown for comparison.2,3
Worked. The minimum of 25,000, 250 and 1000 mg is 250 mg (dose-based).
Surface residue limit
Lsurface in µg/cm², MACO in mg, SSA = total shared surface area of the train (cm²). The factor 1000 converts mg to µg.5
Worked. 250 × 1000 / 40,000 = 6.25 µg/cm².
Swab limit and the concentration in the extract
Aswab = area swabbed (commonly 25 cm²), Vextract = volume of extraction solvent (mL), Climit in µg/mL. The limit per swab is the surface limit multiplied by the swab area.5
Worked. Lswab = 6.25 × 25 = 156.25 µg per swab. In 10 mL, Climit = 15.63 µg/mL.
Swab recovery
Determined for each surface material on spiked coupons. Use the mean of replicates.
Worked. If 150 µg is spiked and 127.5 µg recovered, recovery = 85%. At the limit, the method should therefore read 15.63 × 0.85 = 13.28 µg/mL.
Corrected residue from a swab result
Cmeasured = concentration found in the extract (µg/mL). Compare the result with Lsurface from Equation 5.
Worked. A swab reads 4.2 µg/mL. Residue = (4.2 × 10 × 100 / 85) / 25 = 1.98 µg/cm², which is 31.6% of the 6.25 µg/cm² limit.
Rinse limit
SSArinsed = surface area contacted by the rinse (cm²), Vrinse = total rinse volume (mL). This assumes the residue is evenly rinsed into the whole volume.
Worked. 6.25 × 40,000 / 20,000 = 12.5 µg/mL for a 20 L rinse.
Carryover check in the next product
This confirms the chosen MACO leads to a daily exposure below the PDE (health-based) or the intended dose fraction (dose-based).
Worked. Cnext = 250 / 100 = 2.5 ppm. Exposure = 2.5 × 2 g/day = 5 µg/day, which is 1% of the 500 µg/day PDE, and equals 1/1000 of the 5 mg standard dose.
Equipment train check
The total carryover across the whole train must stay below MACO, even if one location has a higher residue than average.6,7
09Interactive calculators
The defaults reproduce the worked example above. Change the inputs to see your own limits.
A. MACO and residue limits interactive
Implements Equations 1 to 6, 9 and 10. Batch size is in kg, and doses are in mg/day.
B. Recovery and swab result evaluation interactive
Implements Equations 7 and 8. If you enter spiked and recovered amounts, the calculated recovery is used. Otherwise the manual recovery field is used.
These calculators illustrate the arithmetic only. A real protocol uses toxicologist-derived PDEs, qualified surface areas, validated methods, and acceptance criteria approved in advance. Never use a web calculator as the sole basis for releasing equipment.
Validated Cleaning Technologies for Pharmaceutical Manufacturing — Destin A. LeBlanc
Covers the cleaning side of the problem: how detergents, temperature, mechanical action and automated cleaning systems work, and how to design a cleaning process that is capable and repeatable.
It is the right book for section 06, since a cleaning procedure that cannot reach the limit will fail validation regardless of how well the calculations and sampling are done.
10Acceptance criteria and hold times
| Parameter | Criterion | Basis |
|---|---|---|
| Visual | No visible residue on any accessible surface | Applied to every run and every surface |
| Active residue (swab) | Corrected residue ≤ Lsurface (Eq. 5, Eq. 8) | Most stringent MACO |
| Active residue (rinse) | Concentration ≤ Crinse (Eq. 9) | Same MACO, whole-train basis |
| Train total | Σ residue × area ≤ MACO (Eq. 11) | Confirms overall carryover |
| Cleaning agent | Below the limit set for the agent, by a specific or conductivity/TOC method | Toxicological or defined limit |
| Microbial | Bioburden within the protocol limit where risk warrants | Especially after aqueous processing6 |
| Reproducibility | Three consecutive successful runs | Standard validation practice11 |
Hold times. Validate the maximum dirty equipment hold time (from end of production to start of cleaning) because dried residue is harder to remove, and the maximum clean equipment hold time (from cleaning to next use) because cleaned equipment can pick up contamination. Each is tested by holding the equipment for the maximum time, then cleaning or sampling as defined.
11Checklist and pitfalls
Readiness checklist
- Using stainless steel recovery for every surface. Silicone gaskets, coated pans and polymer parts retain residue differently, so each needs its own recovery.
- Sampling only easy locations. Choosing flat, accessible surfaces rather than the worst-case geometry produces passing results that do not mean much.
- Mixing units. Batch size in kg, dose in mg and areas in cm² are easy to confuse. Convert before calculating, and state units in every protocol line.
- Leaving the matrix out of date. A product added without reassessing the worst case can invalidate the limits the validation depends on.
12Specimen quality forms
A protocol summary and a sampling and results record, ready to adapt to your own procedure.
Form CVT-01 — Cleaning Validation Protocol Summary (Tablets)
Specimen only — not a controlled document.
| Item | Value | Calculation reference |
|---|---|---|
| MACO (PDE / dose / 10 ppm) and selected | Eq. 1 to 4 | |
| Shared surface area (cm²) | Engineering drawings | |
| Surface limit (µg/cm²) | Eq. 5 | |
| Swab limit (µg) and extract limit (µg/mL) | Eq. 6 | |
| Rinse limit (µg/mL) | Eq. 9 | |
| Dirty / clean hold times | Protocol |
Form CVT-02 — Sampling and Results Record
Specimen only — one row per sampled location, per run.
| Location | Surface material | Method | Measured (µg/mL) | Recovery % | Residue (µg/cm²) | Pass / Fail |
|---|---|---|---|---|---|---|
These specimen forms illustrate typical content only. Your quality system's document control procedure takes precedence over this format.
Handbook of Analytical Validation — Michael E. Swartz & Ira S. Krull
A handbook on validating the analytical methods behind cleaning validation, covering specificity, accuracy, precision, and detection and quantitation limits.
The swab limit in Equation 6 only means something if the method can measure it, so this supports step 5 of the procedure, where the method has to be proven for trace-level residue.
13References
- U.S. Food and Drug Administration. Guide to Inspections of Validation of Cleaning Processes. July 1993. fda.gov
- gmp-compliance.org. "Shared and Dedicated Facilities: EMA Publishes Final Guideline on Setting Health Based Exposure Limits (PDEs)." gmp-compliance.org
- Pharmaceutical Inspection Co-operation Scheme. Guideline on Setting Health Based Exposure Limits … in Shared Facilities, PI 046-1. picscheme.org
- CASRAI. "Cleaning Validation: Worst-Case Grouping, MACO, and Acceptance Limits." casrai.org
- Leucine. "MACO Calculation in Cleaning Validation: A Complete Guide." leucine.ai
- Indian Pharmaceutical Alliance, Sub-Group 4. "Cleaning Methodology and Validation Best Practices Document." ipa-india.org
- APIC. "Guidance on Aspects of Cleaning Validation in Active Pharmaceutical Ingredient Plants." 2021. apic.cefic.org
- PharmaValidation. "Tablet Compression Machine Cleaning Validation Protocol and Acceptance Criteria." pharmavalidation.in
- Web of Pharma. "Cleaning Validation in Pharmaceutical Manufacturing." September 2026. webofpharma.com
- Guidelines for Pharma. "Cleaning Validation Protocol for Tablets" (example protocol). guidelinesforpharma.blogspot.com
- PharmaGMP Guide. "Cleaning Validation in Pharma: 10 Proven Steps, Complete SOP, MACO Calculation Guide." pharmagmpguide.com
No comments:
Post a Comment