Can Flash Point Results Be Correlated Between Test Methods?

Why ASTM standards prohibit correlated flash point results

Correlated flash point test results are invalid and prohibited. Flash point is an empirical property dependent on the specific apparatus defined in the test method. Data indicates that using alternative apparatus or mathematical correlations creates significant, sample-specific biases, invalidating the test results.

Key takeaways

• Flash point is an empirical property, meaning the measured value depends on the test method and apparatus used
• ASTM D1655 and ASTM D7566 explicitly state that correlated flash point results are not permitted
• A flash point result obtained using one test method cannot be mathematically converted into another method
• Laboratories should always report the result from the method actually performed using the specified apparatus

What is flash point correlation?

Some flash point instrument manufacturers advertise the ability to generate “correlated” results that appear equivalent to other ASTM or ISO flash point methods. While this may seem convenient, international standards make it clear that these correlated values are not valid results for the alternative test method.

This article explains why flash point correlation is not permitted, what ASTM standards require and why using the correct apparatus is essential for obtaining compliant flash point results.

Learn more about What is Flash Point Testing? >

Recently ASTM Subcommittee D02.J formally acknowledged that Flash Point correlation is not permitted. This is clearly stated in ASTM D1655 Table 1 Note K and ASTM D7566 Table 1 Note N which states “Correlated flash point test results are not allowed.” Furthermore, clause 11.1.3 states “Use only the apparatus specified by each test method. Some apparatus have automated correlation of test results to other flash point methods. These correlated results are not valid and are not permitted for use in this specification.”

For valid flash point results this rules out:

1. Using a different apparatus other than that specifically described in the test method
2. Using any form of correlation equation between methods or apparatus

Using different apparatus or correlating a result from one method to another violates the integrity of the test and the test method designation, and renders the test result itself useless.

Why flash point is an empirical property

The reason why this is important is because flash point is an empirical property, that is to say that the conditions of the test and the apparatus affect the result itself. In flash point this is easy to explain if you consider that the flammable vapour and air mixture itself affects combustion directly.

In practical terms, factors such as the cup design, ignition source, heating rate and test conditions all influence when a flash occurs. Because each published test method defines these conditions differently, the measured flash point is unique to that method.

Flash point is one of the most significant tests carried out, beyond simple quality control, it is a critical test because it is one of a number of properties that must be considered in assessing the overall flammability hazard of a material. Amongst others, the flash point by Pensky-Martens Closed Cup test methods, ASTM D93 and ISO 2719, are utilised in the Globally Harmonised System of Classification and Labelling of Chemicals (GHS) to determine flammability. Flash point is used in transport, shipping and safety regulations to define flammable and combustible materials and classify the, Pensky-Martens is used in critical aviation fuel specifications such as ASTM D1655 and Defence Standard 91-091, as well as in other material specifications for example in road transport fuels (ASTM D975, EN 590) and Marine Fuels (ISO 8217).

With flammability being such an important safety property, attention to detail and compliance are crucial in obtaining the correct flash point. The only way to obtain a Pensky-Martens Flash Point test result is to use Pensky-Martens apparatus because flash point is an empirical property.

There are many different flash point methods, such as ASTM D56 (Tag), ASTM D92 (Cleveland), ASTM D3828 (Setaflash Flash/No Flash), ASTM D6450 (CCCFP), ASTM D7094 (MCCCFP), ASTM D7236 (Setaflash Ramp). Because flash point is an empirical property it means that the result for a sample from one test method may not be the same or correlate with another test method. This is due to the specific properties of the apparatus and conditions of the test (open cup/closed cup, equilibrium, dynamic etc.) and how they affect the combustion of the volatile components of the sample.

Explore the Different Types of Flash Point Test >

Why a universal correlation cannot exist

Using another type of apparatus leads to sample specific bias’s as is easy to demonstrate through the flash points of various pure chemicals used in verification. Pure chemicals are amongst the easiest to test as they don’t usually contain contamination.

The materials in Table 1 have been tested in globally organised inter-laboratory studies and are robust values.

Although the differences between methods may appear small, they are not consistent from one sample to another. This changing bias is exactly why a universal mathematical correlation cannot exist.

	Flash point of pure chemicals (oC)
Test method designation	D56	D92	D93	D3828	D6450	D7094	D7236	Max Δ
Description	Tag	Cleveland	Pensky-Martens	Setaflash	CCCFP	MCCCFP	Setaflash Ramp
n-hexadecane	–	138.8	133.9	–	–	129.2	134.8	9.6
n-decane	50.9	–	–	49.7	–	–	–	1.2
n-undecane	67.1	–	–	65.9	–	–	–	1.2
2-butanol	–	–	–	20.7	–	24.4	–	3.7
dodecane	–	–	–	–	79	80.3	–	1.3
anisole	–	–	–	–	43.9	44.3	–	0.4

A study by ASTM Subcommittee D02.08b on Flammability and the UK Energy Institute TMS SC B-4 Flammability Panel further substantiated this finding in a study on different aviation fuels, and although in a very limited temperature range it further illustrates the point. In Table 2 below the difference from D56 to other flash point methods depends on the sample and the method. Sample specific bias is different depending on the sample, making a correlation equation completely unworkable and impossible to achieve.

	Average Flash Point (oC)					Average Bias (oC)
Sample	D56	IP 170	D93	D7094	D7236	D56	IP 170	D93	D7094	D7236
4	37.2	36.5	36.8	37.2	36.3	0	-1	0	0	-1
8	41.1	42	42.7	43.2	41.4	0	1	2	2	0
7	45.1	46.2	46.8	47.5	47	0	1	2	2	2
3	55.8	56.6	57.8	57.9	57.1	0	1	2	2	1
6	60	60.8	62.7	62	61.8	0	1	3	2	2
2	60.7	60.3	59.5	60.2	58.6	0	0	-1	-1	-2
5	61.3	62.3	63.2	63.5	62.6	0	1	2	2	1
1	61.8	62.6	63.9	63.7	63.1	0	1	2	2	1

The changing bias shown in Table 2 demonstrates that there is no fixed correction factor between flash point methods. A result obtained using one method cannot simply be recalculated to produce a valid result for another method.

What this means for laboratories

Simply put, if different apparatus from that described in ASTM D93 is used then the flash point is not valid according to ASTM D93. Alternatively, a way of saying the same thing, a flash point measured on some other apparatus, such as MCCCFP, cannot be recalculated as if it was from ASTM D93. No correlation is valid or permitted.

Some vendors are stating that their apparatus, which is used in test methods ASTM D6450 and ASTM D7094 can generate “correlated” results purporting to be ASTM D56, D93, D3828 or other standard test methods. This is wrong and users of this test method may not realise that this is not a valid practice in flash point testing. It violates the integrity of the test and the test method designation. Further, a modified version of a test method, for example with a different ramp rate, is no longer a published test method and has no precision or validity.

In order to further alert users to this subject the following text is being balloted by ASTM Committee D02: This test method is cited in the globally harmonized system (GHS), product specifications and government regulations. Use only the apparatus specified by this test method. Results using alternative apparatus or correlation equations are not valid for this test method.

Laboratories working to ASTM, ISO, IP or regulatory specifications should therefore ensure that flash point testing is always carried out using the apparatus specified in the published test method and that reported results correspond to the method actually performed.

Choosing the correct flash point tester

If your laboratory needs results compliant with ASTM, ISO, IP or regulatory specifications, always use an instrument designed for the specific test method you are performing.

Stanhope-Seta manufactures flash point testers covering methods including:

• ASTM D56 (Tag)
• ASTM D92 (Cleveland)
• ASTM D93; ASTM D8175 (Pensky-Martens)
• ASTM D3828; ASTM D7236; ASTM D8174 (Small Scale)
• ASTM D6450 (CCCFP)
• ASTM D7094; ASTM D8606 (MCCCFP)

Each instrument is designed to perform the published method rather than generating correlated values, helping laboratories maintain compliance with international standards and regulatory requirements.

View the full Stanhope-Seta Flash Point range >

Learn more about Flash Point testing and choosing the correct instrument >