2013 Mazda3 TSB says fuel in crankcase.. do I change the OEM oil spec?

My 2013 Mazda3i with 4 Cylinders 2.0L FI DOHC still has indications of the Mazda Technical Service Bulletin No: 01-005/15, Last Issued: 06/19/2015, Subject: CHECK ENGINE LIGHT ON… WITH INCREASED ENGINE OIL LEVEL. I bought this CPO Mazda with just 25K miles from the dealer with no mention of any DTC being cleared but the 6+ quarts of “crankcase fluid” [dare I call it “oil”] looked like it had never been changed. So much for a 140 point Mazda CPO inspection. The TSB recommends a PCM reprogramming = $100 = Not Part of the CPO warranty. I did the full oil & filter change after the PCM update with only 4 quarts instead of 4.5 quarts. Nothing but warm weather since and the dipstick has still climbed from 1/2 to FULL. Apparently, PCM reprogramming is insufficient to overcome the poor job of Mazda valve timing & fuel injection engineering. Now I need to consider the outcome. Do I use the 0-20w grade oil Mazda specifies in the interest of fuel economy or do I use a higher viscosity to offset the thinning action of fuel in the crankcase. Mobil lubricants told me to consult Mazda and did not indicate which Mobil product best accommodates fuel contamination. I am fed up with Mazda customer service and Mazda dealers. What would your recommendation be when the crankcase is diluted with fuel? Do you increase the oil viscosity? Warmer weather persists. When the fuel evaporation from the crankcase through the PCV diminishes during colder weather, then some consideration of oil viscosity will need to be made. Perhaps I should leave the oil filler craced open to allow the lighter distallates = fuel = to exit the crankcase.

From TSB:


  1. Verify engine oil level is increased and diluted with gasoline.
  2. Replace the engine oil.
    NOTE: Oil filter is not covered under this warranty repair.
  3. Reboot the IDS to clear memory before reprogramming.
  4. Using IDS 95.05 or later software, reprogram the PCM to the latest calibration (refer to “Calibration Information”
    table) by following the “Module Reprogramming” procedure.



Tester , got it . I misread . Going to delete my post.

Is there over 80,000 miles on the vehicle?


If there’s a concern the oil is losing lubricity because of gasoline in it, I would use the thickest oil recommended for the conditions (10W30, not 5W20, for example, if both are recommended.) I’d also use the an oil with the narrowest range (10W30 instead of 5W40, for example, if both are recommended.)

Wow, and I was impressed by Mazda’s direct fuel injection. I think yours has the Skyavtiv-G system… otherwise ignore. I don’t know the answer, but…

Apparently the reprogramming is to fix a known problem… the FI sprays gas in the cylinders when the engine is shut down. Did you have this done? If shut off often, the problem is much worse. (Makes sense… more shutdowns, more times to squirt gas.) I would make sure they check that the high pressure fuel pump is actually cutting fuel delivery. DI requires 2 fuel pumps, the tank model and a high pressure unit, mounted on the firewall. It’s driven by the engine, and it appears it has a cut-off valve.

You didn’t mention how long it took for it to squirt 1/2 quart of gas into your crankcase. That is excessive, even for this engine… and I’m surprised the engine runs w/ 1/3 gas/oil mix for lubrication. Some of the boards simply suggest changing oil every time the oil level climbs… a silly “patch”.

An apology is appropriate for my delay in this response. I wanted to take time for careful consideration of your reply.

The mileage was 25,506 on 6/1/2019 when the recent oil & filter change was done. Today it is 29,554. Please review my post to the other replies for more details on the analysis.

An apology is appropriate for my delay in this response. I wanted to take time for careful consideration of your reply and assembling the information for the response.

Thank you for the oil viscosity recommendation. I will likely use 5w20 in the next oil change if increases in the crankcase oil level continue. The loss of lubricity with fuel dilution is my concern. My notes for my last oil change show 4 quarts of Mobil 1 5w30 oil [instead of specified 0w20] with some uncertainty on the proper viscosity to use. The manual specifies 0W20 for USA/Canada SKYACTIV-G 2.0, MZR 2.0 or MZR 2.5 and specifies 5W30 for the MZR 2.3 DISI Turbo and 5W20 for the Non-USA/Canada SKYACTIV-G 2.0, MZR 2.0 and MZR 2.5. Your recommendation for the highest recommended [5w30] with the smallest range makes 5w20 seems like the right choice. I am baffled about non-USA change of viscosity unless they are targeting a fuel efficiency standard in the US/Canada.

An apology is appropriate for my delay in this response. I wanted to take time for careful consideration of your reply and assembling the information for the response.

Thank you for the description of the fuel pressure path &logic. This 2013 Mazda model code is M3SIA… 2.0 L DOHC 16valve SKYACTIV-G L4 with Variable valve timing. The Mazda tech service bulletin [TSB] does say to update the programming to correct the fuel in the crankcase problem. The Mazda Certified Pre-Owned warranty covered neither this fuel-in-crankcase TSB nor the TSB [should have been a recall] for headlight flicker when brakes are applied that also specified a programming update for the correction. This TSB required diagnostic trouble codes to be considered. An oil level that was so high that it could not be read on the dipstick was sufficient to convince me to spend $100 on the dealer programming update to attempt to fix both issues. Being early winter when I discovered the overfilled crankcase, I simply changed the oil without replacing the filter. Luckily the seals have no apparent damage and the headlight flicker is now barely noticeable though still present.

The rate of fuel into the crankcase for summer driving is lower than in winter. The TSB says the problem only occurs below 17F but that is wrong. On 6/1/2019 the most recent oil & filter change was done. My notes show 4 quarts of Mobil 1 5w30 oil [instead of specified 0w20 for USA/Canada SKYACTIV-G 2.0, MZR 2.0 and MZR 2.5] took the level to 1/3 above lower extreme on dipstick range after restarting the engine. The Mazda Owners Manual says use 4.4 quarts for oil & filter change so the dipstick reading made sense. Since the oil change, the level rose to the full mark on 6/22/2019 with an average local June hi/low 77/60. Today 08/11/2019 the dipstick reads 2/3 to full level with an average July hi/low 86/68F. However, reading the Mazda dipstick presents a challenge compared to all other cars [>15] I have owned. That most recent reading had oil the entire length of the reading zone plus some on one side of the dipstick and 2/3 full on the other side. Apparently, during higher temperatures, the oil level is well-behaved.

My normal driving needs typically last less than 30 minutes. The engine never really gets an extended length of time at the maximum crankcase “bake out” temperature. Warm up time [until the “cold engine” light goes out] is not nearly as long in summer as in winter. I have toyed with the idea of the window shade in front of the radiator [SemiTractors & Saab 850 2 cycle use it] in the winter but the Mazda front end design discourages it. The cost of gas for a 3 hour idle to achieve a bake out of the crankcase in winter time might be less expensive than the synthetic oil expense for the “patch”. Do you think that “cracking” the oil filling cap would shorten the time [and CO2 output] needed for that method?

I will post the outcome of the oil level readings as cooler tempertures begin after this month thru the winter months. I did swap out the battery for the 24F form from the much smaller 35 form and that did definitely reset the computer [why would the dealer not do that as part of the computer reprogramming]. If the oil level rises enough in the upcoming readings to raise concern, I will post the outcome of both a window shade and a crankase “bake out” method.

Are you reporting the oil level from measurements you do at the same point of use of your car? What I mean is, do you always check the oil before you start the engine in the morning? Or when you get home? Or at lunch time? The oil level will vary at different moments depending on lots of factors. And, is the car always in the same place when you measure? Parking on any sort of a hill or slope will affect the oil level on a dipstick, so it’s important to try and be consistent.

You have an excellent point. The time needed last weekend for a bank visit [7 -10 min] was insufficient for the entire oil volume to return to the crankcase. The bank was closed and its empty lot gave a chance in a busy Saturday to check = level car with oil on all of one side of dipstick and other side read half. Today I check again cold & level after sitting overnight and the dipstick reads full on both sides of the dipstick. A dramatic difference in the reading.

Whoa… So your Mazda is “making oil” as its sometimes called in mechanic circles. This is No Bueno as they also say…

I’m having difficulty understanding the logic behind a software change and the impact it has on a direct injected engine making oil. Also how much fuel are we talking about here and how fast ? I believe others have asked but I dont think I read the answer anywhere as of yet.

At any rate I’m sure software is in the chain of events somewhere but in the case of direct injection I just dont see how software plays a part in the internal leaking of an extremely high pressure mechanical fuel pump that is injecting fuel into your engine.

Since this over fueling is occurring and NOT severely affecting the drive ability of the engine in the form of smoke and poor fuel mileage etc etc… That super strongly suggests that this crankcase filling is not occurring via the combustion chamber and has to be entering the engine from somewhere else, like from under a valve cover or on the head or where ever they installed the mechanically driven high pressure fuel pump.

Wherever the pump is and whatever portion of that pump is exposed to the crank case oil supply…must be the place that it is leaking internal to your engine. Pump shaft seal? It has to be something of this nature.

1 Like

Thank you for your question. The “baseline” for fuel accumulation in the crankcase has not yet been established. I purchased this 2013 Mazda with a Mazda “Certified Pre-Owned Warranty” (CPO) less than a year ago in 2018. Research on the TSB for light flicker upon brake application took me to an “aha!!!” moment when I saw the TSB for fuel in the crank for low temperatures. I then realized that the reason I was having trouble reading the dipstick was because the “oil” level was “off the charts”… so to speak. The 140-point CPO warranty had failed to simply change the oil. The dealer made a point of replacing the tires and wiper blades but not the oil!!. The oil on the dipstick so high, even after a cold night, that a reading could not be taken and the dipstick appeared almost clear. When I drained the sludge-laden oil, I realized the oil was, if not the free first-time in warranty oil, the original OEM oil. The oil filter had “Mazda” markings. Can “oil” levels get so high that the “sludge” in the oil does not show on the dipstick? Maybe, if there is significant fuel content, then only the lighter parts may show on the dipstick, especially after a long night to separate.

Why did Mazda engineering miss this obvious flaw and its cause?

Who knows… but the seals seem to be holding ok.

The CPU software update was done in December 2018, at my expense, and nearly corrected the headlight issue. A prior post suggesting fuel line pressure release into any cylinder seems to make sense. My recollection of the rings I put onto Fiat cylinders during an engine rebuild nearly 50 years ago used rings that were not continuous around the circumference of the piston. Colder weather would allow the walls to remain cooler and allow the fuel to remain liquid. The unstoppable engine of gravity then draws the liquid fuel past the rings and into the crank. After 5 years of this, I would anticipate the result that I saw. Warmer climates would prevent the effect. Warmer temperatures vaporize the fuel, even after engine shut down, and the fuel vapor escapes the cylinder in the same path of its entry. Warmer climate also allows the crank to reach higher temperatures during short trips I typically take to allow crank ventilation to clear the fuel. The software change may simply shutoff fuel injection earlier in the valve timing to prevent “post-shutdown drip”. No difference in power or pick-up [what little there is performs actually better than the carburetor version] has been detected since the software upgrade.

As cooler weather approaches, I will post any dramatic changes in oil level rises.

The last Mazda I owned used a carburetor.

This mistake will not be repeated.

Thanks again.

The reply suggesting a relief of fuel pressure into the cylinder during