Troubleshooters.Com(R) & Steve Litt's Overheating Guide Present

Litt's Overheating Hypothesis

Copyright (C) 2000 by Steve Litt
Warning! Read safety precautions on Steve Litt's Overheating Guide


There is probably considerable misinformation on this page. Much of this page is my own theory, and quite a bit of the information I gathered over the phone and Internet from various sources. I just learned about Block Testers today. I take *absolutely no responsibility* for the results, including injury or damage, obtained from using the information on this page. Do not use this information unless you are ready to accept the full responsibility for the outcome. Always use proper safety precautions.

So why have I written information I believe to be partially wrong? To establish a dialog so various people can come together to, over time, solve an incredibly tough class of automotive problem, and to give those already having that problem at least a fighting chance.


Since I wrote it, I've upgraded this information from a hypothesis to a theory. In fact, many Troubleshooters.Com readers have written me to say that this information is absolutely true, and at least one provided what I consider to be some pretty strong evidence.

Therefore, this information has been superceded by the May 2000 issue of Troubleshooting Professional Magazine, which contains much more information and treats this information as a theory and probable fact, rather than as a hypothesis. This page has been left on the Troubleshooters.Com website to show the history of thinking that went into the head gasket and gas related overheating theory.


I hypothesize that exhaust gasses entering the cooling system through a broken head gasket can cause cyclical overheating, unexplained overheating, and coolant loss in the absense of overheat or known leaks. You can buy a "block tester" for around $45.00 that can confirm or rule out the existance of exhaust gasses in your coolant. Troubleshooters.Com intends to act as an information center for cyclical overheating, unexplained overheating and unexplained coolant loss.


Starting in 1999, several Troubleshooters.Com readers described symptoms of cars that would repeatedly overheat and then go back down to normal temperature (cyclic overheating). One such description did this only once, several did it repeatedly, typically on approximately 20 minute intervals. I estimate about five such symptom descriptions, enough to convince me that this is *not* "defective user", but instead is a legitimate problem. I didn't take the first couple describers seriously, and for that I apologize.

The first break happened when one reader  had cyclic overheating problems (the temp would go up while driving, then go back down, then up, then down...) and also had white smoke coming out the exhaust. He installed block sealant fluid, and the cyclical overheat temporarily subsided. This is very indicative that the broken head gasket might be the cause of the overheating, as opposed to the usual case of overheating causing a broken head gasket.

Since then, I've paid close attention to all overheating complaints mailed to Troubleshooters.Com, and have noticed that in quite a few there was a reason to be suspicious of head gasket causation.

My Hypothesis

A hypothesis is an educated guess that must be proved or disproved. It is not a fact, nor even an opinion, professional or otherwise. My hypothesis is that many cases of cyclical overheating (temp gauge going up and down repeatedly while driving at fairly steady speed and incline) are caused by combustion gasses escaping the cylinders into the coolant through a flaw in the head gasket.

I believe it possible that these gasses heat and pressurize the coolant and possibly fill it with small gas bubbles, all of which reduces the ability of the coolant to carry heat to the radiator. The engine heats up. Once the pressure becomes too great, the combustion gasses are "burped" out the radiator cap and into the reservoir, at which time the coolant is once again able to carry heat so the engine cools. But the exhaust gasses are still transferring and building pressure in the coolant, so soon enough the process repeats itself.

Nor is cyclical overheat the only outcome of exhaust in the coolant. On some cars the pressure pushes out coolant rather than getting burped out, resulting in a coolant loss. Note that this can happen even in the absense of overheating. Exhaust gasses in the coolant are a documented cause of unexplained coolant loss at steady medium to high speed driving.

If the coolant loss is massive, the loss of coolant results in overheating. Many readers have written me saying they've had the thermostat, coolant and radiator replaced, and a pressure test indicates no leak, yet it's overheating. I advise those readers to check for combustion gasses in the coolant, as explained later.

I just got a particularly interesting email from a reader who found excessive hydrocarbons in his coolant, and was able to eliminate his overheating symptom by loosening the radiator cap. I'd imagine that once the exhaust gasses were free to go, they just cruised right out of the radiator. As long as a cooling system is capable of maintaining temperature below the boiling point under normal circumstances, this is quite believeable.

Most Mechanics Do Not Believe This

Note: In the time since this was first written, I've received email from many mechanics saying this information is definitely true. To be fair, it must be said that there are also many mechanics that DO know and believe this information.

I've spoken to several mechanic and found that few believe this hypothesis. For one thing, many of these cyclical overheats occured without evidence of coolant in the oil (yellow gunk around the oil cap) or coolant in the cylinders (profuse white exhaust, especially on startup). As one mechanic put it, "a broken head gasket is a two way street". The idea is that once the engine is shut off, the coolant pressure will exceed the cylinder pressure and force coolent into the cylinders and oil.

I'm not so sure. Envision a head gasket defect so small that it acts like a valve, opening only above, let's say, 50psi. The cooling system bleeds at much lower pressures, so the transfer would be only from cylinder to coolant, but not the reverse, even when the engine is shut down.

Most of the readers submitting these symptoms have been to one or more mechanics, and typically had spent several hundred dollars replacing components. Most overheats are caused by clogged radiators, low coolant levels, bad water pumps, inaccurate thermostats and the like, so a mechanic can successfully diagnose and fix over 90% of overheats without ever considering a broken head gasket. And most broken head gaskets leave obvious clues such as yellow gunk on the oil cap or billowing white exhaust. I believe this is why most mechanics reject this hyptothesis.

How to Test for Combustion Gasses in Your Coolant

You can use a "Block Tester" to chemically determine whether there is exhaust gas in your coolant. These "Block Testers" are sold by NAPA stores for approximately $45.00. The quote I got from my local NAPA dealer was $45.99. He didn't have it in stock, but said he could have it the next day. The relevant NAPA catalog is called "The PSA 2000 catalog" or the "Balkamp Catalog". The catalog calls the Block Tester a "combustion leak tester kit", so that's probably what you should ask for. From what I understand, it comes with a ball, tubes, test fluid, aspirator bulb and engine adapter (cone shaped device you place in your radiator filler cap). If there's exhaust in your coolant the test fluid changes color.

I don't know how to use this test kit or how good its directions are, but my local mechanic expressed the opinion that an ordinary person could use it to perform the test. Needless to say, for safety reasons it must be performed BEFORE the engine overhats (in other words at the beginning of the day).

I don't know how accurate this test is. As more information comes in from Troubleshooters.Com readers we'll get a better idea.

How This Information Can Be Used

Because this information is, at this point, hypothetical theory, I can't recommend using a Block Tester as a General Maintenance test (See for details on General Maintenance in Troubleshooting). After all, $45.00 is not cheap. If overheat is accompanied by an obvious cause such as a leaky or shaky water pump, a totally clogged radiator, a non-functioning thermostat or fan, you'll probably want to fix the obvious cause, especially if it's a relatively new car and did not seriously overheat. However, I would personally use the Block Tester in any case of unexplained overheating, rather than playing the $200/guess "maybe it's this" game with a mechanic. If the car is of limited value, I would spend the $40.00 immediately before investing in a radiator or other expensive component, because if I need to replace the head gasket I'll junk the car instead.

If, in the future, I find that the Block Tester test is accurate and it is verified that combustion gas leakage frequently causes overheats, I will probably begin recommending Block Testers as a General Maintenance test for overheats and unexplained coolant loss.

Prevent Overheating At All Cost

As I read overheat emails from Troubleshooters.Com visitors, I'm struck by the heartaches caused by even a single overheat incident. Overheats can break head gaskets, and broken head gaskets can cause additional overheating, broken starters, broken solinoids and broken flywheels. Extreme overheats, or moderate overheats on cars with aluminum heads, can destroy the heads. A single overheat can cost you well over $2000.00.

Prevention starts with a temperature gauge. Do not buy a car without a working temperature gauge. By the time the temperature "idiot light" comes on, your engine is very close to destructively overheated. If you own a car with only an idiot light for temperature, consider having a gauge installed.

Once you have a working gauge, glance at it at least once every 10 minutes, or any time the car's operation seems strange. Be sure to shut the car down long before the needle goes "in the red" because engines continue to gain temperature after shutdown. It's much better to spend $150.00 for a tow truck than $2500 for an engine rebuild or replacement, or even $1000 for a simple head gasket replacement.

Read Steve Litt's Overheating Guide. It discusses many other tactics and strategies for overheating diagnosis, as well as a discussion on safety.

New Information: 4/8/2000

It's been 3 or 4 days since I put up the hypothesis web page (this page), and I've gotten quite a bit of feedback. Much of the feedback expresses an opinion that yes, it's possible for combustion gasses to get into the coolant without coolant in the oil or steam out the exhaust. Several of those responses were from acknowledged automotive experts on the web.

I've heard from several mechanics who sniff the radiator filler pipe with their emissions probe. If you sniff the radiator filler cap with smog equipment, do it after the engine has warmed up (so warm it up with the cap off), and BE SURE not to let the sniffer contact the coolant, as the coolant will destroy the sniffer according to at least one respondent.

That same respondent said the most sensitive test is the HC (hydrocarbon) test, although a false positive can occur if you've recently flushed your radiator with a petroleum based product. His opinion is that the HC test will reveal the earliest stages of combustion gas transfer, much earlier than the Block Testers will reveal.

Another respondent mentioned an interesting alternative dovetailing hypothesis. He said the concensus on the Mopar lists is that air bubbles in the coolant give rise to hotspots that cause gasket failure due to uneven cooling. Makes sense! Be very, very concerned if you see bubbles in your coolant. So once again we have a chicken and egg question -- did the head gasket create coolant gas, or did the coolant gas cause head gasket failure. Once again, a test to detect combustion gasses in the coolant answers the question.

My Lucky Break

I may have caught a lucky break in tracking this problem. My 1967 Dodge Coronet (318) doesn't have white exhaust or a yellow oil cap, nor does it overheat, but it does continually exhibit unexplained coolant loss. Could a head gasket failure be transferring combustion gas, which pushes coolant out the cap, into the reservoir, and overflows the reservoir? As soon as I verify the best test for combustion gas in the radiator, I'll be doing the test in front of reliable witnesses willing to publicize the results. If I can provide a single documented example of combustion gas in the coolant, in the absense of evidence of coolant in the oil or cylinders, I will have completely justified the combustion gas test in overheating problems. If you operate an automotive repair facility near Apopka, FL, and would like to run this test and agree to use your name on the publicly recorded test results (published on Troubleshooters.Com and possibly elsewhere), please email me.

My New Outlook

I have not gotten enough information to recommend a combustion gas in radiator test a general maintenance (step 5 of the Universal Troubleshooting Process) item yet. But given the information I now have, on my own personal vehicles, I would be very skeptical of any mechanic refusing to consider testing for combustion gasses in the coolant in the case of an overheat complaint, especially a cyclical overheat, unexplained overheat, or unexplained coolant loss.

This is an Ongoing Investigation

This is an ongoing investigation. Troubleshooters.Com will act as a clearing house for information on this subject. If you have any information, email me. If you have cyclical overheating, unexplained overheating, and/or unexplained coolant loss and you live near Apopka Florida, please email me. Maybe I can set you up with an open-minded shop willing to test for combustion gasses, act as a witness to the result.

If my hypothesis is correct, it MUST become common knowledge, because people regularly spend hundreds of dollars replacing parts willy-nilly in a vain attempt to fix certain overheat problems. We must find and publicize the truth.

 [ Home page | About Steve Litt | Email Steve Litt | Training ]

Copyright (C)1999 by Steve Litt. -- Legal