2009 HCH- Possible Internal Coolant Leak?
I have a 2009 Honda Civic Hybrid that I am working on for a family friend. Car has 49,000 miles. The car is well-maintained and not driven hard. He called me initially after the car would start, run rough, and die out within a minute. By the time I got there, the car started and ran fine. We laughed it off and went our separate ways.
Two days later (last night), I get another call. Engine is running weird, temperature gauge is pegged to the max, and there is no cabin heat. Also, engine smells of overheating. He was within a few hundred yards of home so finished the drive and parked it. Let it cool down and restarted it- he said that it would not idle, but if he gave it gas the engine would smooth out and rev up.
I finally got to look at it today. Checked all fluids and found the coolant tank empty. Began adding water, not realizing how much would be needed. I ended up adding just about a gallon of water to a 1.6 gallon cooling system. No coolant leaks noticed or observed. I asked if he had ever seen a puddle underneath the car, he said no, nothing of the sort.
Car started and ran fine. No DTCs set or stored. Drove around the block with no issues. Did not observe any coolant leaking. Parked it, let it sit for a bit, and it started back up just fine.
Also of note is that the engine oil was changed by the Honda dealer last week. They do a multi-point inspection as part of the service, and made no mention of low coolant (recorded on the inspection form).
I would be set that it is an internal coolant leak, but there are none of the classic indicators- no milky oil on dipstick or oil fill cap, no (abnormal) pressurization of the cooling system, no sweet smell or excessive steam out the exhaust. The only thing I think is happening is that the car starts and runs fine, but once stopped, whatever is leaking opens up due to heat expansion and then when the engine is restarted, that leak allows coolant into the cylinders, which causes the extremely rough running and inability to hold an idle. That's my theory, at least, which unfortunately leads to an engine tear-down on a car that is 4 months out of powertrain warranty.
Experienced automotive folks (and smart folks in general!), what do you think is wrong given the information provided?