2015 CX-5 2.5 SkyActiv coolant in oil

2015 CX-5 GT
After another 2200 miles, the coolant-indicating numbers are the same, but the oil interval was 1000 miles longer. To me, it indicates that this engine can finish its service life with 2,000-mile oil changes. Blackstone report attached.
Other thoughts:
1) The heater core is now producing heat at about 80% of its original capability. I'm planning to replace the heater core and flush the coolant once more to make it safer/more comfortable for Winter driving.
2) If you have this 2.5 SkyActive engine, I would urge you to send your used oil for sampling at every oil change. Catching it (relatively) early saved major damage to my engine.
3) For those whose UOA already indicates high Iron shedding, I would advise a FilterMag be attached to the oil filter. This will only catch Ferrous metal but will capture tiny amounts that can pass through the oil filter. Ferrous metals cause the most damage inside the engine. (A regular magnet won't work; only rare-Earth magnets retain their strength as temperature rises).
4) One thing I wish i had thought of before using the Head Sealer is re-torqueing the cylinder head bolts. I recently saw a Subi engine that failed due to this. I realize they are completely different engines but I believe this one was modified for higher-compression. At 13:1 (US engines) and 90,000 miles, I believe it is worth considering.
Has anyone checked the torque on head bolts for the 2.5 non-turbo SkyActive?
Do any of you Gurus know the torque spec and tightening sequence?

Once I find this information, I'm planning to start well-below the spec, increase incrementally, and determine which ones ones are loose (if any), and by how much. Of course, I'll share my findings immediately.


  • BlackstoneMazdaMay22anonymous.jpg
    181 KB · Views: 49
2020 CX-5 Tour
Head bolts are torque-to-angle using 4 passes, which means that the final torque value is not known and cannot be checked using a torque wrench. However, given that you found zero leak down, IMO the probability of finding a loose head bolt is basically zero as well, so I suggest just calling them good.

But if you feel compelled to check them, then use 40 ft-lbs, which is just 6 over the initial (non-angle) torque, and should not be anywhere close to approaching the actual final torque. As previously stated, doing this will not confirm that the final torque is correct, but it will tell you that none of them are significantly loose or 'backing out' (which again was almost certainly already confirmed by your leak down test results). Sequence is basically inside-out, but won't matter with this low torque setting, which won't do any actual torquing.

You don't want to ever 'experiment' by trying higher torque settings on head bolts, because if any of them are turned beyond the final torque, that might cause a much larger problem than you currently have now.
2015 CX-5 GT
What I hear you saying is.......
Zero Leak Down = Leave Them Alone!
Thanks, edmaz - I'll take that advice!
18 Mazda CX5 AW
From reading your posts, it sounds as if a miniscule leak that is not visible to the naked eye is considered "Severe" by UOA analysis standards.
And that by the time one can see signs of antifreeze in the oil or vice versa that it is probably too late to save.

Never had this happen on any of my previous vehicles but with the potential for these Mazda engines to have this problem,

Sounds like all the more reason to get the oil analysis done on a periodic basis even if it's every other oil change.

Thanks for posting your experience and observations.
'15 6, '06 5
If you're not going to do anything to correct the coolant leak (and I don't say you should or you shouldn't), then you should probably stop wasting money on the UOAs. If it's purely a hobby, fine, but if it's about detecting trouble or being able to make proactive corrections (which you're choosing to not do), that's basically never cost-effective on personal automobiles.


2016.5 CX 5 Touring AWD, 2015 SCION XB
...or maybe the first lab report report was simply a glitch. I've used Bars leak before, I can't see it fixing a high compression engine.