2.5 Turbo - Oil Fuel Dilution Issues?

V
2010 CX-9 GT
This only applies to GDI (Gas Direct Injection) engines of which Mazda's are not GDI.


Straight from Mazda's website: "Fuel system: Advanced Direct Injection"

It's a direct injection engine. You don't get 310 pound-feet of torque at 2000 rpm any other way.
 
V
Mazda CX-9 Grand Touring 2016
The viscosity is out of spec. The viscosity at 100C for 30 wt should be 9.3 to 12.49 centistokes. Your test showed 8.4 cst. Your Pennz started at 10.3 according to their data sheet. That loss could be from shearing, or it could be from gasoline dilution. Do let us know when the lab responds to your question. The silicon at 21 ppm could be either silane anti-foam additive in the oil or dust that passed through the air filter. The wear metals are nice & low. Shell (Pennz) uses moly in the oil for anti-wear. Castrol uses a titanium compound to do the same job. You'll see the difference if you get an analysis with Castrol. The zinc phosphate compound in the oil is anti-oxidant and anti-wear but leaves a high friction layer on the metal. The addition of either the Molybdenum dialkyldithiocarbamate or (cheaper) similar oil-soluble titanium compound provides a low friction hard antiwear layer.
 

sm1ke

Work In Progress..
Moderator
Contributor
L
Canada
V
'18 CX-9 Signature
The viscosity is out of spec. The viscosity at 100C for 30 wt should be 9.3 to 12.49 centistokes. Your test showed 8.4 cst. Your Pennz started at 10.3 according to their data sheet. That loss could be from shearing, or it could be from gasoline dilution. Do let us know when the lab responds to your question. The silicon at 21 ppm could be either silane anti-foam additive in the oil or dust that passed through the air filter. The wear metals are nice & low. Shell (Pennz) uses moly in the oil for anti-wear. Castrol uses a titanium compound to do the same job. You'll see the difference if you get an analysis with Castrol. The zinc phosphate compound in the oil is anti-oxidant and anti-wear but leaves a high friction layer on the metal. The addition of either the Molybdenum dialkyldithiocarbamate or (cheaper) similar oil-soluble titanium compound provides a low friction hard antiwear layer.
Thanks for the explanation, PTguy. You must have some background in oil analysis, or you must also be a member on BITOG.

The rep finally got back to me. These are his words: "Fuel will only be run if the viscosity is low and flagged. If your viscosity is normal, then the fuel won*t be run. I hope this answers your question."

Sounds a little lazy, and he didn't mention how low the viscosity would need to be to be flagged. Does lower cSt = more viscous? If so, I guess the viscosity wasn't low enough for them to run the sample for oil dilution. That may be their general practice but it doesn't really sit well with me, so I'm going to try Oil Analyzers (also in Edmonton) next time.

I usually do an oil change in March/April and another in October/November. I think I'll send the October/November sample out for a UOA since it will reflect the more aggressive driving I do in the summer months.
 
Last edited:
V
CX-9 GT 2019
Heh, I did a trip to another dealer and that experience was like from another world. I never thought you can get that much of variation between two same brand dealers 20 miles from each other. Seems like the first dealership does not have anyone proficient enough to do anything other than simple part or fluid change.

This time they took the car for testing once they heard oil level was raising and oil had a gas smell. No stupid questions.

They did tests and confirmed that one injector was leaking. Replaced all injectors.

Easy.

Lets see how the next analysis would look like.
 
V
Mazda CX-9 Grand Touring 2016
Does lower cSt = more viscous?
Other way round.The kinematic viscosity test, measured in centiStokes, is by pouring a specified volume of oil, at the specified temperature, through a specific orifice and timing it. The thinner the oil, the faster it flows.
https://www.machinerylubrication.com/Read/411/oil-viscosity

By the way the labeled viscosity grades, called weights, are established in the SAE J300 tables. It gives several specs each viscosity grade must meet including different types of viscosity measures. Any viscosity grade with a W is winter--it doesn't have a hot spec. For example SAE 20 and SAE 20W aren't the same thing. An engine oil that is labeled, say, 5W-30 meets a group of cold specs to be 5W and a group of hot specs to be 30. And...any viscosity grade of 70 or above indicates automotive gear oil. The higher gear oil number doesn't mean it's thicker than engine oil. SAE 40 engine oil is about as viscous as SAE 90 gear oil.
https://www.tribology-abc.com/abc/viscosity.htm#SAE

One of the viscosity tests is the HSHT high shear high temperature test at 150C. It is measured in a different sort of viscometer with a read out in milliPascals. Higher is more viscous. HSHT viscosity relates to both engine wear and oil drag within the engine. Lower wear will result with higher HSHT, but the increased drag results in lower gas mileage. That's a balancing act the oil chemists do when they create one of the energy conserving oils. In any case, the products they sell us all perform very well. (One Pascal equals one Newton per square meter. I don't understand that, either.)
 
Last edited:
V
Mazda CX-9 Grand Touring 2016
crimeariver, good going. You knew you had a problem and got it fixed. At the dealership service desk you're pretty much stuck with the service writer and what they (think) they know. Some are sharp, some are willing to explore a problem, and some aren't. I wonder if you'd have better luck at the first one if you'd asked to discuss the problem with the service manager rather than just a service writer. Anyway, the 2nd place has earned your future business.
 
V
CX-9 GT 2019
One more point on labs. The Blackstone is very popular lab, but they don't do oil analysis even though they provide the "Fuel %" value in the report.
They simply check flashpoint, but their staff seem to think its a reliable test. I have send my previous sample with 3.8% to them in parallel and their recommendation was totally misleading:
 

Attachments

Last edited:

Latest posts

Top