"We have noticed that the viscosity of the oil in our circulating oil system has been decreasing. What are the possible causes of this?"

There could be several reasons for a downward trend or even a sudden drop in viscosity without warning. However, the most common cause is adding in a lower viscosity fluid than what’s required. For instance, if you are using an ISO 220 viscosity gear oil but a lower viscosity gear oil (such as ISO 68) of the same brand is added in, this can lead to a sudden drop in the viscosity reading when the oil is sampled.

A decrease in viscosity may also occur when non-lubricants like solvents and diesel fuel accidently get into the lubricant. If this happens, it is a good idea to change the oil.

Another way your lubricant could be losing its viscosity is through the loss or shear down of the viscosity-index (VI) improver. For example, if you are using a multi-grade SAE gear or engine oil such as a 10W-30, this oil contains an additive known as a viscosity-index improver. During use, the VI improvers can sheer down and break apart, causing the viscosity of the oil to decrease. Remember, exposure to high heat is the biggest factor in causing the sheer of the viscosity-index improver.

Multi-grade oils aren’t the only ones that can thin due to high heat. Oils operating at extreme high temperatures can begin to crack thermally. The high temperatures can sheer/crack the oil molecules into smaller molecules, which causes a decrease in viscosity. If you are having trouble with a mineral oil losing viscosity at high temperatures, look at switching to a synthetic oil for the application.  

Keep in mind that viscosity can go up, down or remain unchanged. The list of root causes that can alter a viscosity reading is quite extensive. This is why viscosity has become such an information-rich measure of used oil condition. After all, when viscosity has not changed, you can rightly conclude that the many known viscosity-altering factors are probably not occurring. What's not so good is when viscosity moves suddenly with no obvious explanation or warning. You then must determine what this means and why it happened.

It's safe to say that viscosity will not change without a forcing event or condition that incites the change. While the oil analysis community is aware of the usual suspect conditions or events, some remain undiscovered or at least are not fully understood.