Most of us think about measuring moisture by...
When using these processes, we take for granted that the results will be the product's actual moisture.
Not always so.
Recently, I was reading about a procedure the Brazilian Agriculture Ministry (BAM) developed to get a very, very tight tolerance moisture determination for coffee beans.
The BAM procedure specifies placing the coffee beans in an oven for a long time -- 104° C for 24 hours. This test was the standard against which all other coffee moisture determination methods were compared.
However, it was found that significant amounts of other volatiles were being released during the 24-hour process, which resulted in higher-than-expected moisture readings. The oven test standard thus over-stated the actual moisture.
The BAM also used an oven method using ground beans
In addition to two oven methods and the infrared loss-on-drying instrument, a Karl Fischer Titration was used. The Karl Fischer Titration measures only moisture and eliminates the effect of volatiles. Because the coffee beans need to be dissolved in a solvent media, the beans had to be ground -- thus inducing the aforementioned grinding error to this moisture analysis.
(If an evaporator oven had been used in concert with the Karl Fischer Titration, the whole beans would have been heated to a high temperature. The volatiles and moisture would then have been extracted and the residual gas collected in the Karl Fischer unit. Only the moisture would be measured using this method, thus eliminating the grinding error.)
The moisture range in the beans tested was 4.2% to 13.5%. Based upon these analyses, two significant inferences were made:
1.The effect of volatiles evaporating is more significant as the moisture levels decrease. This indicates that volatile material amounts tend to be the same at all moisture levels, thus representing a greater weight loss proportion for lower moisture samples.
2.Grinding causes moisture loss, which becomes a greater proportion of the total for lower moisture samples. In these lower moisture samples, it was found to be more significant than the volatiles.
Most of our solids moisture measurement applications are low volatile situations. When grinding is needed, the processes usually have a wide enough tolerance to not require grinding loss corrections. Even though most of you do not have to contend with this kind moisture test adjustment, I thought you would be interested in the implications of these factors on the true moisture level in a sample.
Let me know if you have questions or comments.