Let's talk engine coolant temperature

Car is a 1997 BMW M3.

Cooling system is brand new with the following modifications:

- BMW S54 radiator
- EMP Stewart high flow water pump
- 80 C thermostat
- New radiator switch (low speed 80 C - high speed 88 C)
- Brand new AC condensor fan
- Clutch fan removed and replaced by a high flow 16" SPAL - #2120 (2360 CFM)
- All factory shrouds in place.
- Spal fan is triggered via the factory radiator switch
- Aftermarket STACK water temperature gauge with sensor located in cylinder head on cyl 2 (factory gauge is much closer to Cyl 1)

Driving in the city, in south Florida the temperature hovers at 195 F and it is pretty steady. Tops at 200 F and goes back down to 190 F when the fan activates.

When steady state cruising on the highway, at about 75-80 mph and 3500 rpms, it goes from 210-220 F all day. It will not go past 220 F.

I have the Stack gauge warning to come on at 215 F, as that as high as I am comfortable running the engine, but as mentioned above it goes all the way to 220 F if I let it go and don't manually turn on the Spal fan.

Am I correct in that 220 F in a straight 6 is asking for trouble? Or is it my ptsd from blowing a headgasket on this engine?

Also, could it be that I am taking the temperature measurement right at the head, while most people with aftermarket gauges do it at the thermostat housing, before the water goes in the head? I guess my temperature measurement is closer to what goes back to the radiator, vs what leaves the radiator.

Any input is appreciated. My next step here otherwise is to trigger the fan right from the gauge at a lower temperature.

Will add pictures in a bit

Gauge temp sensor:

Spal fan:

Diagram showing radiator temp switch location:

While in traffic it will stay here all day (195 F)

While cruising on the highway it bounces between 210-220 F. This is as high as it will go, I only let it get here for the picture and to see how high it would go:

I'd be curious to know the temp coming out the radiator exit. Seems like the rad isn't big enough, or isn't flowing enough water, to lose the heat. Definitely not an airflow problem.

In reply to 02Pilot :

It might take me a bit, but I can probably rig a sensor on the outlet and see. 

Its an new BMW S54 radiator from a late Z3 with that engine, much bigger than the factory S52 one. I would think it would be better than the factory one.

Running hotter on the highway than in the city is odd, are you sure that the fan isn't wired backwards and pushing air out the front of the radiator? Once you confirm that, test again with the fan wired to run at all times and see where the temps are.

220F isn't too hot, as far as I know.  We (Ford) regularly ran that hot, very intentionally.  As long as you have a good rad cap, it's not going to boil any time soon.

But it's normally hot when running slow, not when running on the highway, so +1 for that being really odd.

In reply to GameboyRMH :

Fan is spinning correctly. If I over ride the switch and manually trigger it, it will go down to close to 205 F I think. I can drive it and test tomorrow. 

So your buffered factory gauge is still dead center that tells me you are still in the BMW safe zone.  My X5 according to the OBDII runs about 215 all day.  My Lemons car normally runs about 210 except when we forget to fill the radiator then it runs 250 for hours at Road America.

Changed your water pump lately?  One of mine fell apart, still kind of worked but in two pieces.

Running hot at highway speeds suggests an airflow problem, not a cooling system problem. 

I would guess you are building a high-pressure zone in the engine bay that is blocking flow through the radiator. 

Do you have any missing ducting or a missing air dam under the car? How about the sides of the radiator? Are they blocked or do they allow airflow through into the engine compartment? You need to make sure the high-pressure area at the front of the car is in front of the radiator and the low-pressure area is behind it. 

Edit to say, all air flowing into the engine compartment must go through the radiator. Block all other routes. 

In reply to Toyman! :

Yeah, you're right: airflow, not water flow. My brain is struggling today.

In reply to Toyman! :

As I recall Miatas were somewhat notorious for this if you left the bottom shroud off. I might be mistaken though.  I kind of remember mine acting that way when I test drove it with it off, but its been over 10 years)

Those shrouds...  were they a retrofit from the smaller engine radiator onto that one or?   That said, I could see a low sped recirc issue with the gap SPAL to shroud, but that doesnt explain the higher speed issue.  Do they maybe need flapper-door to allow highway velovity air through other parts?  (doesnt seem likely, but thought process)

high flow water pump...  could it be recirculating coolant too quickly for the heat to transfer to the rad at highway RPM?   The thermostat, is it the right amount of restrictive?  (if you run no thermostat on many engines you run into similar)

In terms of airflow, since you're running an electric fan with its own ducting you don't need the factory fan shroud anymore, you could remove it or cut holes in it to get some airflow improvement.

First, I gotta say it.  (sorry to sound like a jerkface)  Ditch that fan.  Two reasons:  1) it is not shrouded.  It has too much space between the fan and the shroud.  2)  I have never entered one of these conversations where the problem WASN'T the aftermarket electric fan.... and I've been in a lot of these conversations.  Buying a fan based on CFM and size is like buying a spark plug because the thread size seems to fit the hole.  WAY too many factors.  Some fans are designed for parade queens where they have to generate all the flow.  Those fans tend to constrict flow on the highway.  Conversely, some fans don't flow enough at idle but they're great at free flow on the highway.  I'm not saying you picked a random fan, but it's really the only thing you've changed from a known-good recipe.  You've replaced a belt driven fan that consumes up to 10 hp and free-flows 6000 cfm (guessing) with a 1/3 hp electric fan with tiny blades that are 1/4" away from the radiator.  I'm not saying you need to go back to a belt fan, but you can see from my exaggerated comparison why the fan might not be enough.

Secondly, 220 isn't too hot.  Not by a long shot.  You blew a head gasket because of boiling, not because the head reached 240.  You boiled coolant, and parts of the head hit 800 degrees while others stayed at 240.  Think of it like a cast iron skillet.  You can put it in a 1000 degree kiln and it wouldn't warp because it heats evenly.  If you instead focus an acetylene torch in one corner of the skillet, it will either warp or crack.  Same goes for coolant and heads.  When you get boiling, one part of the head gets super heated.

The other thing to keep in mind is that as long as you're not getting boiling, the engine doesn't care what the coolant temp is.  Combustion happens at 2800 degrees.  It doesn't care if the coolant is 200 or 220.  The limiting factor is that aqueous coolant has a relatively low boiling point.  That's why everyone stresses about 235 degrees, not because 240 is where death happens, it's because 240 degrees is where boiling happens... and the BOILING causes the death.

I would say as long as it refuses to go over 220 in any condition, and you're not getting boiling, I think you have a good 15-20 degree buffer (depending on the coolant mix and cap pressure) before you're in trouble.

I'm in agreement with 220F is fine especially if the stock gauge also says its fine. But also with that fan setup is far from optimal. I had a Z3M with the same engine for a year or two. I ditched the engine driven fan and cut out as much of that shroud as I could tofree up exit airflow, kept the stock A/C fan on the front side, and switched to the low temp fan switch. I debated an aftermarket electric fan but ultimately stuck with the stock A/C fan because it was actually pretty powerful and, more importantly, ducted well even if the pusher configuration isn't quite as efficient. That setup had zero issues with any normal driving or mid-summer autocross. I monitored temps with an OBD2 reader for a bit initially, I forget the exact numbers but idling it would cycle between 190 and somewhere around 200 and I don't think I saw anything over like 215, Oh....and that car had a supercharger on it so it was making a bit more power and A LOT more heat than stock. I did run a weak coolant mix with some water wetter type stuff just for some extra insurance.

alfadriver said:

220F isn't too hot, as far as I know.  We (Ford) regularly ran that hot, very intentionally.  As long as you have a good rad cap, it's not going to boil any time soon.

But it's normally hot when running slow, not when running on the highway, so +1 for that being really odd.

Oh jeez, some newer German cars don't even have the thermostat open yet at 220F.

Highly turbocharged German cars. 2 liter highly turbocharged 5000lb sedans German cars.

It's all how the engine was engineered.  220F would make a small block Chevy very upset because its cooling system is "meh, put a water jacket around stuff and pump some coolant through and it will be okay, probably" and at that point the center cylinders would be suffering film boiling around the exhaust ports and combustion chambers and it would probably knock itself to death under load.  Newer engines do a lot better than that.

As for the OP, this is what my mostly stock RX-7 does - fine around town, really hot on the highway - and making sure all of the grillework was in place helped, because that also functioned as an air director to flow air towards the whole radiator - but the major issue is that the radiator is full of calcification, so heat transfer is rather poor.  When making a constant 30-35hp the radiator can no longer effectively transfer heat even with a 75mph headwind.  In city traffic there is less power, and more importantly that power is staccato, so there is a heat sink catch-and-release effect instead of a constant demand.

In reply to Pete. (l33t FS) :

The 6.0 in my Silverado ran steady on 220. The electric fan in my XJ came on at 217. 

In reply to Toyman! :

The electric fan, if present on the Silverado, came on at 225f for low speed, 235 for high, if I recall correctly.

You'd see this every now and then on speed-talk. Guys used to small block Chevys would freak out when their LS would hit 215 and keep climbing.  Nah, guys, that's totally fine.

Also regarding OP - throw that fan far far away and get a BMW one.  I have used SPAL fans before and they were universally crap.  Just looking at that one, I can tell that it isn't very good because it has a tiny motor, and the blades are just straight fins.  Those will beat the air as much as they actually move it.  Good fans have curved, contoured blades, and the motor should require a 30-40 amp fuse.

Mounting it directly to the radiator is also not helping you, because the fan is only drawing air through the parts of the radiator that are swept by the blades. Ideally the fan blades should be one fan hub diameter away from the radiator, and decently shrouded, so they pull air through the whole core and not just 40% or so of it.

Being unfamiliar with the M3, if it had a belt driven fan, BMW may even have shortchanged the natural airflow situation with the expectation that the fan would be there always contributing some positive flow, even at (American) highway speeds.

Has anybody told you to throw that fan in the bush?  Every time somebody tells me all about their electric fan and saving gas and making horsepower, I think of my Dakota, or my Silverado towing the race car, where you can hear the engine clutch fan cut in and feel the truck labour.  Now, that's not really a good thing, but it does tell me how much horsepower is required to pull that much heat out of the rad.

Or, find a Volvo XC90, hook up your scan tool, and pull it up to 100% duty cycle.  It will make a great deal of noise, and draw many, many amps.

I also raced a Neon that never once ran under 225, and had zero trouble.  Of course, its not German...

In reply to Streetwiseguy :

I converted my competiton RX-7 to electric fan because I could cool it down in grid.

It's from a Chrysler 300M.  Small radiator in a cramped nose with an engine right up against it making it hard for air to escape.  Chrysler did not screw around.  There are two two-speed fan motors that each require a 40 amp circuit.  I amp-clamped the cabling going to the fans and when they are on high speed, they will draw an honest 75 amps.  That's really pushing the limits there, Mother Mopar 

They will also cool a 270hp rotary in real time on a chassis dyno, or during a 60 second low speed high power rallycross run.

I love electric fans but I hate crappy electric fans.  If you want to move air you need POWER.

My favorite for a swap fan is from a '13 GT500.  They are cheap to procure and move a LOT of air.

Toyman! said:

In reply to Pete. (l33t FS) :

The 6.0 in my Silverado ran steady on 220. The electric fan in my XJ came on at 217. 

 

There is a large number of guys that were over on a Chevy Express forum who all drank the kool aid and switched their half-ton vans to the 1-ton radiator, because (gasp) they ran 220 degrees.  There is a widely-held notion in the Chevy van world that GM whiffed on the cooling system and should issue a recall.  My 06 Express rides between 215 and 220, which makes sense since the stat is a 195 or 200 depending on which plant they were built.  That means that right beside the chamber where the sensor is seems to make sense.  My Express runs 215-220 in February on the highway, and it runs the same 215-220 in August towing 6000 lbs up a mountain at 35mph with the A/C cooling that huge interior.  That's a sign to me that it's doing its job.

When it comes to coolant temps, you have to remember that the actual temperature is only important relative to the boiling point of your coolant mix and pressure.  If you run straight coolant, and your temps go up to 275, it won't matter a hill of beans because it can't boil.  Back in the day when I got kinda suckered into Evans NPG, my Caddy 500 towing 10k lbs would frequently get to 300 without a radiator cap even installed.  I had to install an oil temp gauge to make sure I wasn't cooking oil.

The name of the game is heat, not temperature.  As long as you have liquid coolant that is:

1. Able to absorb X units of energy from the engine
2. Able to shed at least the same X units of energy in the radiator
3. Remains in contact with the water jacket around the chambers (doesn't boil)

Then the actual termperature doesn't make much difference.  Iron melts at 2800F.  Aluminum melts at 1200F.  If 240 degrees would hurt those metals (or 500 for that matter) we couldn't use them on the stove to cook a grilled cheese, let alone survive nearly 3000 degrees of combustion temps.

What matters is how quickly the coolant can absorb units of heat, and how well the radiator can shed the same amount of heat to the air.

Edit to add:  This whole mindset of temperature being king is because aqueous coolants have been the norm for 85 years, and we're so hyperfocused on the temperature side of the equation because for the last four generations, the temperature was the way we knew there was too much heat, when it's the actual heat that does the damage.  We keep focusing on "oh boy, it's getting to 225 degrees," when what we should be looking at is the trend.  If the temp gauge keeps going up, we know we're putting more heat into the coolant than the radiator can remove.  Since your temp gauge is staying steady, that means yours is succeeding.  If you ever have a really hot day and it starts creeping up, then you know your cooling system is just barely adequate and doesn't leave you much room.  And also remember... engines like to be hot, drivers like them to be cool.  OEMs spend large dollars inventing ways to get engines to get UP to temperature so they can operate in the heat range where they're happy for their given oil viscosity, bearing clearances, and ring gap/emissions/blowby and so the cats stay happy.  Then we go in and put in lower temp stats, bigger radiators, and extra fans to make sure they run cool.

Suggestion for you... if it's easy to do.  Pull the electric off and re-install the belt fan for a week or so.  If your temps still get to 220, you know that the system is doing it's job.  If it suddenly starts reading 185, then you know your electric fan isn't up to the task.

I once saw 230F on my Scangauge in the Expedition while towing on a 90 degree day. It sat there for at least an hour and the factory gauge never moved from the middle, the truck never missed a beat, and the AC kept blasting cold air. 

I would also like to hijack a tiny bit in an effort to help Slippery with his problem.

High volume water pumps.  My assertion is that they neither help nor hurt.  Same goes for restrictions like high flow stats vs normal, or coolant restrictor plates.

So if your coolant can absorb (using completely random numbers here for simplicity) 100 kJ of heat energy per second, and the radiator can shed 100 kJ of energy per second, changing the speed of the coolant shouldn't change anything, because the number of seconds it spends in the engine and the radiator are reduced by the same amount.

Let's say the coolant spends 4 seconds in each the radiator and the engine.  It absorbs 400 kJ in the engine, and shed the same 400 kJ in it's 4 seconds in the radiator.  If you add a high volume pump, it now spends 3 seconds, changing those numbers to 300/300 kJ.

You should (in my brain) be equally reducing the amount of time absorbing as you do shedding.  The net effect should be a reduction in the delta T between engine and radiator, not an increased transfer of heat.  If you slow it down really slow, the engine might get up to 200 before it leaves the stat, but then it has plenty of time to get down to 120 in the radiator.  If you speed it up, maybe it gets to 220 in the engine, but only 180 in the radiator.

My logical brain sometimes hyperfocuses on the A + B and forgets that there are sometimes other factors like C and X.  Just seems to me like a hi-flow pump is pointless.

Gee guys, I leave for a few hours and you guys fill the thread with a bunch of questions and great advise, lol. 

I have answers for many of them and agree with quite a few of the thoughts but will answer all tomorrow morning. 

Thanks!