Variability & Spray Package Corrosion

Written on: November 1, 2020 by W. Stephen Tait

Hello, everyone. Variability is a fact of life that complicates corrosion testing. It occurs both within a given manufacturing batch (within-batch) and between different manufacturing batches (batch-to-batch) with both formulas and packages. Within-batch and batch-to-batch variables that affect spray package corrosion susceptibility and formula corrosivity could include one or more of the following:

1. Formula chemical composition:
• Active ingredient concentration
• Corrosion inhibitor concentration
• Surfactant concentration
• Fragrance concentration
• Formula or contaminant water pH
• Downward active ingredient assay drift with increasing time
• Formula or contaminant water pH drift with increasing time
• Loss of product efficacy with increasing time

2. Package materials and components:
• Coating thickness inside the package
• The number of holes and defects in polymer coatings and laminate films
• Cracks in the aluminum foil used for laminated foil bags
• Tin porosity in tin coatings (holes) on steel (electrodeposited tin plate [ETP] aerosol containers)
• Aerosol valve crimp dimensions

The different types of variability could determine when spray package corrosion occurs and the corresponding corrosion severity. Severe corrosion causes a significant reduction in the package service lifetime and less severe corrosion causes a smaller reduction in the service lifetime. Figure 1 provides three graphs for the corrosion risk as a function of time:

1. Storage stability testing on traditional metal aerosol containers
2. Storage stability testing on bag-on-valve packages
3. Electrochemical corrosion tests on all types of spray packages

These graphs were generated with data from numerous corrosion tests and their associated actual, commercial corrosion.

The data in Figure 1 are cumulative data from all types of package corrosion (e.g., general corrosion, pitting, coating blisters, etc.) collected from decades of test data on a wide variety of spray formulas. Notice that the corrosion risk decreases as the test time increases. Notice also that the time for low risk with electrochemical corrosion tests is significantly lower than the associated storage stability tests (30–90 days versus one year, respectively).

Please note that the data in Figure 1 for electrochemical corrosion tests are only valid for the Aristartec technology. The correlation (risk) for other types of electrochemical test protocols either have not been established, or are very low (e.g., 84% risk for the driven can test). Please also note that the storage test data in Figure 1 are also only valid for a specific protocol.

Figure 1: Corrosion risk as percent occurrence versus test time















Suggested best evaluation practices
Therefore, how does one determine how much the within-batch variability and the batch-to-batch variability affect spray package corrosion susceptibility and formula corrosivity? The answer:
By conducting corrosion tests with the appropriate number of replicate
samples as a function of time.

The number of replicate samples that need to be examined or measured for corrosion decreases as the test time increases, as indicated in Figure 1 (risk decreases with time). In other words, the number of samples to examine/measure is inversely proportional to test time—which in equation form is:

Number of replicate samples α (1/test time)

Examining one or two replicates at short storage times, such as three months, has essentially no statistical confidence (essentially 100% statistical risk). My favorite number is 18 replicate samples for package examinations when the time is < 9 months, then a smaller number of replicates after nine months.

Please treat these replicate sample numbers only as guidelines. The appropriate number of replicates for your formulas is determined by the type of formula chemistry, plus the type of spray package materials and components used with your formulas.

Spray packages from multiple replicate commercial batches should also be evaluated to determine if scaling up from the laboratory to commercial manufacturing causes variability that also affects spray package corrosion.

Admittedly there is often not enough time or resources to conduct the comprehensive corrosion testing program outlined in this month’s Corrosion Corner. However, not doing so increases the risk of surprise corrosion with a corresponding possibility of a very expensive product recall and loss of customer good will.

Please keep in mind that Corrosion Corner only provides an overview of corrosion science, engineering, control and prevention and is not intended to teach how to establish and conduct a comprehensive corrosion testing, control and prevention program.

If you have any questions, call 608-831-2076 or email Thanks for reading and I’ll see you in December. SPRAY