A good stability study for a cleaner answers one question: will this product still look, smell, pour and clean the way you promised after months in a warehouse, a truck and a customer's cabinet? To design one, you pick the commercial package, set storage conditions that match where the product will really live, test the same properties at fixed intervals, and write pass/fail limits before the first sample goes in the oven. This guide walks through each step the way we do it with brands that come to us for formulation and private label work.
TL;DR
- Test the product in its final commercial package (or a smaller one of the same materials), not in a lab beaker.
- Pair real-time storage with accelerated and temperature-stress conditions, and judge each against limits you set in advance from your first good batch.
- Track appearance, pH, viscosity, fragrance, microbial load and cleaning performance, not just how the liquid looks.
- Most household cleaners that are not pesticides have no federal shelf-life testing rule. Pesticide claims change that quickly.
- Re-test when you change a raw material, supplier, process or package.
Table of Contents
- What a stability study actually establishes
- Start with the failure you are trying to prevent
- Choose the package before the conditions
- Set storage conditions that match real life
- Decide what to measure
- Sampling intervals, replicates and pass/fail limits
- Know which U.S. rules actually apply
- What to do when a sample fails
- Our perspective: stability is a manufacturing decision
- Build your study with a manufacturer who can adjust the formula
- FAQ
- Sources
What a stability study actually establishes
A stability study is a series of tests that track how a product changes over time. The goal is to learn whether those changes make it unsafe, ineffective or commercially unacceptable. The output is a shelf life you can defend, and the data behind it.
That last part matters more than most new brands expect. Stability data is what lets you justify a labeled shelf life to retailers, regulators and your own QA team. It is also your protection when a buyer calls about a bottle that has separated or swollen.
Keep one distinction straight from the start. This is not cleaning validation, the pharmaceutical process of proving equipment is free of residues. If you are a brand, startup or commercial cleaning company making a product to sell, a stability study is about your product aging in its package.
Start with the failure you are trying to prevent
Design backwards from the complaints. A product that separates, thins, loses strength or swells its bottle on a store shelf may come back as a return. Every failure mode maps to a test:
| What the customer sees | What to measure |
|---|---|
| Cloudy, layered or separated liquid | Appearance, phase separation, centrifuge or freeze-thaw checks |
| Runny or thickened product | Viscosity |
| Weak cleaning or odd dilution behavior | pH, active level, functional cleaning test |
| Faded or shifted color | Visual color against a retained standard |
| Off smell, slime, bloom | Fragrance integrity, microbial bioburden |
| Bulging, leaking, rusted caps or seams | Package inspection, weight loss, corrosion |
Write this list for your own formula before you pick conditions. A thickened glass cleaner and an enzyme-based laundry product fail in different ways, and the second is far more sensitive to heat and time. We cover that case separately in our enzyme stability controls article.
Choose the package before the conditions
The package is part of the formula as far as stability goes. EPA guidance for accelerated storage studies says to use the product in its commercial package or smaller packages of the same construction and materials. The same guidance asks you to observe both product and container at the beginning and end, and to report packaging corrosion such as perforations, darkening, leaking or rust at the seam.
That is pesticide guidance, but the logic holds for any cleaner. Acidic, alkaline and solvent-bearing products attack closures, liners, trigger sprayers and metal seams differently. If you want a feel for how formula chemistry changes what a product does to a surface, this explainer on acid versus non-acid wheel cleaners shows the same acid/alkaline trade-off from the car-care side.
A few package decisions to lock in before you start:
- Test the exact bottle, cap, liner and sprayer you plan to ship, from the same supplier.
- Test orientations. Cosmetics industry stability guidance notes it may be advisable to test packages upright, inverted or on their side. Lying on its side is how many cleaners travel, and it is where leaks and liner reactions show up.
- Weigh the filled units at each pull so you can see moisture or solvent loss.
- Keep retained controls: sealed samples held in a stable, cool place for comparison.
The same guidance also says study design should consider variation in process conditions. In practice that means more than one production lot.
Set storage conditions that match real life
Most cleaner stability programs combine four things.
- Real-time storage at room temperature. This is the only condition that proves the shelf life you print.
- Accelerated storage at elevated temperature, to see problems sooner.
- Temperature cycling or freeze-thaw, for products that will ship or sit in unheated space in winter.
- Light exposure, if the bottle is clear and the product has dyes, fragrance or light-sensitive actives.
The only hard, published U.S. benchmark we can point you to is from EPA, and it applies to pesticide products. For certain pesticides, EPA accepts a 14-day test at 54°C ± 2°C. If the product is unsuitable for that test or fails it, a one-year room-temperature study is required. Treat that as a reference point for how regulators think about acceleration, not as a universal rule for an all-purpose cleaner.
For a non-pesticide cleaner, you choose the conditions and the duration, then defend them. Heat that your product will never see, or a freeze that your distribution chain cannot cause, makes a study harder to interpret. A product sold across Texas and Minnesota needs a different stress profile from one sold in a single climate-controlled facility. Our own approach to pulls and lots is laid out in the shelf life protocol for cleaners.
Decide what to measure
Common measures for a cleaner are appearance, pH, viscosity, fragrance integrity, microbial bioburden and functional cleaning performance. Testing labs describe the field the same way, as evaluating chemical, physical and biological stability over time under varied environmental conditions. Commercial labs also bundle ambient and accelerated shelf-life studies with performance-claims and package-compatibility testing.
Here is how we think about each one:
- Appearance and color. Compare against a retained sample under the same light. Color drift is often the first thing a customer notices, and color stability deserves its own acceptance limit.
- pH. A drifting pH usually signals a reaction in the bottle, a package interaction or a preservative problem.
- Viscosity. Thickener systems can thin or build over time. Measure with the same spindle, speed and temperature each pull.
- Fragrance. Fragrances oxidize, discolor and fade. Smell is a legitimate test if you use trained panelists and a reference.
- Microbial load. Water-based cleaners need a preservation system that holds up in the real package. See how we match cleaners to preservatives.
- Cleaning performance. This is the one most teams skip, and we will come back to it.
Why performance belongs in the study
A product can look perfect and clean badly. EPA's Safer Choice guidance treats performance as its own evidence requirement: cleaners must clean the relevant soils and surfaces at the most diluted routine-cleaning use direction, using an applicable standardized or pre-approved equivalent test. The agency's Design for the Environment criteria ask for a good-faith demonstration of performance and test results or equivalent evidence. A stability study by itself does not establish efficacy unless you run the performance test on aged samples. So pull aged units and clean something with them.
Standard methods help here. For hand dishwashing detergents, ASTM D4009 uses foam persistence during standardized washing: the result is the number of plates washed to the half-foam endpoint, compared with a reference product. Run the same test on fresh product and on aged product, and the difference is your performance drift.
Sampling intervals, replicates and pass/fail limits
This is where most small studies go wrong. They test once at the start, once at the end, and decide afterward what "acceptable" means.
Do it in this order:
- Make a first good batch and characterize it. Record pH, viscosity, appearance, density and performance. Practical guidance for sellers is to set pass/fail limits from your first good batch, because no regulation reviewed there sets measurements or limits for household cleaners.
- Write the limits down before the study. For example, a viscosity window and a pH window around the first-batch values, no visible separation, no package damage.
- Pull at fixed intervals. Space them more tightly early, when accelerated samples change fastest, and wider later. A 3-lot, 4-pull layout is the minimum we recommend as a starting framework.
- Use replicates. ASTM D4009 recommends controlling water hardness, temperature, concentration, method and analysis, and running at least four replicate tests evaluated against statistically determined confidence intervals. It puts an experienced operator's result at about ±5% precision at roughly the 95% confidence level. That tells you how much of a change is real and how much is test noise.
- Keep the method constant. Same operator training, same instrument, same temperature of the sample at measurement.
A simple worked layout for a water-based all-purpose cleaner might read like this:
| Arm | Condition | Pulls | Tests |
|---|---|---|---|
| Real-time | Room temperature, upright and on side | Start, then spaced through the intended shelf life | Full panel plus performance |
| Accelerated | Elevated temperature chosen to match your risk | Tight early pulls | Appearance, pH, viscosity, weight, package |
| Stress | Freeze-thaw cycles | After each cycle block | Appearance, separation, viscosity |
| Control | Retained, sealed, cool | Compared at each pull | Visual and pH |
The temperatures and durations in the middle are yours to choose and document. We are deliberately not printing a number the regulations do not set.
Know which U.S. rules actually apply
This is the part that surprises founders. For household cleaners that are not pesticides, no federal rule reviewed requires shelf-life testing or a date. Hazardous substances under the FHSA have required precautionary labeling, which is a separate matter.
Pesticide claims are different. EPA identifies storage-stability and corrosion-characteristics data requirements under Guidelines 830.6317 and 830.6320, with an accelerated alternative accepted in certain circumstances. If a product fails the 14-day accelerated test or is unsuitable for it, EPA requires the full one-year study. Fail that, and an expiration date plus an advisory label statement limiting exposure to higher temperatures are required.
The practical trap: one marketing word like "disinfects" or "sanitizes" can move a product from a general cleaner into EPA territory. Read our breakdown of how FHSA labels and a single claim can trigger EPA rules before you finalize label copy, not after.
Safer Choice and DfE programs, covered above, add their own performance evidence expectations if you are pursuing that certification.
What to do when a sample fails
Failure in a study is information, and it is cheaper than failure in a customer's hands. Work through it in this order:
- Confirm the result. Repeat the measurement with replicates to rule out test noise.
- Find the variable. Is it the formula, the package, the fill temperature or the process? Compare lots and orientations.
- Fix one thing and re-test. Change the preservative, thickener, buffer or package, then run the study again on that change.
- Shorten the claim if you must. A shorter, defensible shelf life beats an optimistic one.
The cost of ignoring this is documented. The 2015 Mean Green recall identified affected products by date codes 4225 through 4230, which shows why lot and date coding is what lets a brand trace a problem to specific production.
After launch, treat the study as living. Health Canada's biocide guidance says shelf life should be re-evaluated when the formulation, raw-material supplier, manufacturing process or packaging material changes. That guidance is Canadian and written for biocides, but the habit is sound anywhere: a new bottle supplier is a reason to re-test.
Our perspective: stability is a manufacturing decision
Most stability problems we see are not chemistry problems. They are handoff problems. A formula is developed in one place, filled in another, packaged by a third party, and nobody owns what happens after it leaves the building.
We built SaraWest differently. Our R&D chemists work in the same facility that produces your batches, so the person who tweaks the thickener is a short walk from the person running the filler. If a pilot of 1,000 units shows a viscosity drift at the second pull, you are talking to the chemist, not a broker relaying a message. That access is why we push brands to run stability on pilot lots in the real package before they commit to a truckload.
Our advice is plain. Do not wait for the full order to find out your cap liner reacts with your formula. Spend the pilot on answers.
— Faisal Mansur
Build your study with a manufacturer who can adjust the formula
If you are launching a cleaner, a stability plan is easier when formulation, filling and QC sit under one roof. We offer contract chemical manufacturing with in-house chemists who can adapt one of our 1,200-plus formulas or create a new one to your specification. Prefer to start from an existing base? Our private label cleaning products program covers pilot batches through full truckloads. Tell us what you want made, and a real person will read it.
FAQ
How long should a cleaner stability study run?
Run it at least as long as the shelf life you intend to claim, with real-time storage as the proof. Accelerated conditions can flag problems earlier, but they do not replace real-time data. EPA's pesticide approach shows the pattern: a short accelerated test, with a one-year room-temperature study required when the short test is unsuitable or fails.
Do I need a stability study to sell a household cleaner in the U.S.?
For non-pesticide household cleaners, no federal rule reviewed requires shelf-life testing or a date. Retailers, insurers and your own returns data will still push you toward one. If your label makes antimicrobial claims, EPA pesticide rules may apply, and storage-stability data is part of that.
What should I measure besides appearance?
Measure pH, viscosity, fragrance, microbial load and cleaning performance, plus package condition. Performance on aged samples is the one that proves the product still works, and it is the one most often skipped.
How many batches should I test?
Use more than one. Process variation between lots is a real source of failure, and design guidance asks you to account for it. We suggest three lots as a starting framework, tested at several pulls.
When do I need to repeat the study?
Repeat it when you change the formula, a raw-material supplier, the process or the packaging material. Even a small change in a cap liner or a thickener source can shift results.
Sources
- US EPA: Accelerated Storage Stability and Corrosion Characteristics
- Stability and Shelf-Life Studies for Consumer Packaged Goods
- Shelf life testing for a cleaning product you sell
- Guidelines on Stability Testing of Cosmetics, Cosmetics Europe
- ASTM D4009: Foam Stability of Hand Dishwashing Detergents
- Stability Testing of Cleaning Products
- UL Solutions: Chemical Specialty Product Testing for Household Cleaning Products
- EPA Safer Choice Product Performance Guidance
- EPA Safer Choice and DfE Standard for Cleaning Products
- CPSC: CR Brands Recalls Mean Green Cleaner and Degreaser Products
- Health Canada: Guidance on quality requirements for biocides, stability and shelf life