Guide To Collecting Representative Sample
The Importance of Collecting Representative Thermal Fluid Samples
If thermal fluid samples are not collected in a representative method you will see artificially high flash point values. This means you will perceive that there is a lower risk from flash points than is actually correct.
This presents a potential risk to your business (its operating insurance), the legislative implications of DSEAR/ ATEX and health and safety and, arguably more important of all, you and your staff.
Thermal fluid is collected at Operating Temperature.
In their document entitled ‘Monitoring Heat-Transfer Fluids: The Sampling ‘Bomb’’, (ref: MP 623 International), dated December 1980, BP Oil states,
‘A truly representative sample of the complete charge can be taken only when the fluid is hot and circulating.’
The document goes on to describe use of ‘the bomb’ – a closed sampling device designed to capture the volatile light ends that would otherwise be “boiled off” if the sample was taken when open to atmosphere.
Systems design varies widely from site to site: pipe diameter, fluid velocity and pipe layout must be taken into consideration. In addition, aged oil can be significantly more viscous, therefore;
- If the sample is taken with the system running and at normal operating temperature it is far more likely that turbulent flow is occurring thus ensuring that a homogenous mix of fractions - within the bulk fluid - is sampled (please see example below)
- Any insoluble contaminants will, for the same reason, be more likely to be suspended within the bulk fluid
- Both high and low molecular weight ranges of fractions will be detected
For typical ISO 32 heat transfer mineral oil
At fluid velocity V(m/s) x pipe diameter d (mm) = 10<
Reynolds number = 2,000 for fluid @ 100°C (212°F) but
Reynolds number = almost 20,000 for fluid @ 325°C (617°F)
At ambient temperatures turbulent flow cannot be guaranteed
- There is a substantial difference in viscosity between a sample at operating temperature and an ambient sample. This will also affect the way the fuel-like light fractions mix
It can only be representative when at temperature - if there was an event it would be from a system operating at normal temperature.
- If there was an event and a subsequent investigation, it would be vital to know that your flash point values and any corresponding advice given were correct - the only way to ensure this is by having hot, closed & circulating samples
A closed sample device such as a ‘bomb’ must be used to ensure that the fluid does not pass through atmosphere. Light ends or volatiles consist of a homologous mix of hydrocarbons with different boiling/flash points.
Where an ‘open’ sample is collected, the most volatile (lowest flash point) species will automatically escape and flash off to atmosphere, instead of being allowed to cool and condense back into the sample where it can be decanted under lab conditions.
In this case, as the lowest flash point material has been vented off, incorrect (too high) flash point values will be seen