We have gotten a locomotive oil sample back showing 32 ppm zinc. Did a re-sample and it came back the same. The previous quarterly samples all were at 1 ppm. We do not want any zinc, as it will corrode the silver bearings in the engine. Does anyone know where this zinc could possibly be coming from? The engines are topped off with a dedicated pump, and I change the drums myself, so I know it was not a wrong product scenario. Any thoughts?

If you are absolutely certain that there was not a cross contamination issue (locally) when top up took place on the engine then you might want to send out a sample of the new oil to see if there is a contanination issue with it. Does the lubricant supplier package the oil on site? If so then there is the possability that a previous product was not flushed completely before starting a run on the product you are using, thus contaminating the oil with the zinc, which is a common anti-wear additive in many different types of oils. You might want to check that first.

Zinc itself is not corrosive to your silver bearings - its the ZDDP additive that is corrosive. Actually zinc metal could provide cathodic protection to silver coating in galvanic situation. You can run/have lab run RULER or FTIR to determine if Zn present as additive or inorganic. Does pump use different type of oil? Could pump that is dispensing oil be wearing/seal leaking?

We actually took samples from the pump in the drum ( dedicated pump ), and directly from the drum itself. I will have the lab check to see if the Zn present is an additive or inorganic. Thanks for the advice.

We have EMD diesels also and I have been in the same mess you are in. Once it was do to someone using a Galvanized funnel and the other was a result of someone replacing a fitting on my pump that was galvanized.

I always like to look at analytical areas first. What is your calibration Range for Zinc? If you ar calibrating upto 2000 ppm to cover the additives from other oils then there is a good chance your calibration curve does not pass through zero. This could bias your results. Being that you are looking at Zinc as a contaminent in this situation you calibrate in a much lweor tighter range so that your curve does indeed go through zero. Just a thought