[skyship007]

Which engines are you testing ??

My local lab does radiometric particle counts, BUT they still use final oil analysis, cos the particle counters only count metal contamination and most of the debris caught by an oil filter is Silicon or small clumps of Carbon. It's those 2 contaminants that cause most wear and often determine when the oil must be changed.

Also you might be testing synthetic media oil filters. They can be obtained from the US (Mobil, Amsoil or Purolator) for some newer Volvo engines and will produce better initial figures (U curve rather than hockey stick), but alas Volvo sold mostly petrol and not diesel cars in the US, so they don't make one for my 1.9TD at present.

The modern way of testing filters is not radiometric particle counts, as that only works for metals. They use that data mostly to figure out which oil or fluid works best, NOT which is the best oil filter. That's still done with multiple cold starts, redline spikes and sudden shut downs using standard used oil analysis.

My own real world UOA results did produce 16 ppm of Si in 10K km, which if I had done a 20K km run would have put the final figure near the 30 ppm limit (Varies a lot with air quality as I got 8 one year), although the insolubles (General dirt, which is mostly Carbon) only reached one third of their limit. The Iron content was no where near the oil condemnation limit, apart from one run of Catrol Edge 5w40 TD (Terrible Acea C3 cat test). It averages about 30 when the limit is 200 ppm before it becomes a long term negative factor.
The Si in the second run with a dirty filter has always been lower, as has been the Fe figure.

PS: Your post has got my comment and your reply mixed up, if I copy part of something I use Italics to make it more obvious.