Huge engine temperature fluctuations

[Boomslang]

Hello Canalworld - I've been a liveaboard boater for the last 5 years and have been reading this forum for about the same amount of time. I've never needed to post before as most of the problems I've come up against have been experienced by someone else and answered already. This time however I'm stumped!

 

I have an Isuzu 55. When we are running the engine the temperature comes up to 81 degrees and will sit there all day long. I've been on the Thames punching against the tide and after about an hour it steadily heats up and if left will trigger the overheat alarm. This is easily rectified by dumping the hot water from the calorifier. The point is I know that my cooling system set up is good enough.

 

This summer we were out cruising and for the first ten days everything was as usual, engine temp constant at 81 deg. Next day the temp was steady at 81 but at some point the temperature quickly crept up to 95 degrees, stayed there for about a minute and then dropped back down to about 75 degrees before returning to 81 degrees. I thought nothing of it but the same happened the following day. I tried running the engine without the filler cap and this stopped the problem from occurring, however it also stopped the operating temperature being reached.

 

THis is now a recurring problem but is now getting more and more extreme. The temperature often shoots up 110 degrees and then drops back down to 60 degrees. The way this happens looks to me like a blockage that clears, or a coolant pump getting stuck. If I run the engine at idle it doesnt overheat, it will sit at 81 degrees as if there is no issue, it could be that the calorifier is operating as a good enough heat sink to deal with idle. I have never run it for long enough to heat the calorifier up.

 

I thought this might be a sticky thermostat so I changed it - the stat was gunked up but putting the new one in didnt solve the problem. I thought it might be the coolant pipes, so I have changed these, although not the pipes to/from the skin tank - which I guess might be my next thing to do.

 

I think I probably need to check my coolant pump as well but from what I have read this is not something I want to attempt without ruling out everything else first.

 

Is this the right approach? Anyone seen similar behaviour? Any pointers would be greatfully recieved.

 

Jared

 

 

[RLWP]

I have never run it for long enough to heat the calorifier up.

 

Jared

 

 

 

My guess is that you have an airlock in the calorifier

 

Richard

[by'eck]

Remake electrical connections to temp sensor and gauge. Buy cheap IR/laser temp gauge from eBay and check temperature at same point as sensor.

 

If engine temp did reach a true 110^oC I think you would smell hot oil, coolant would be ejected etc.

[Jess--]

although I have a different engine (lister) I had a similar problem that was due to a skin tank half full of air

[bigste]

Classic symptons of an air lock.

[OldGoat]

Two points:-

Your engine is likely to run hotter when going upstream on the Thames because

• You're going faster than on the canals
• You're punching against the flow
• The skin tanks are invariably undersized and inadequate for that task (happened to us on a hire boat many years ago - the operator expressed 'surprise')

As stated above an air lock is most likely and the only reliable way to control that is to fit a header tank in the system and check it regulalrly.

[bizzard]

You say the thermostat was gunked up, I'm wondering if the system is gunked up elswhere too, probably coupled with air locks. It might be worth draining and flushing out and back flushing with a hose pipe the whole system, the engine, cooling tank and colorifier all separately and re-filling and bleeding with fresh coolant. Personally I would not use tap water in the system but distilled, de-ionized or rain water mixed with the antifreeze, ''especially if your tap water is hard and lime scaley'', in this case doing this will keep the system much cleaner and much free'er of rust and corrosion.

[DHutch]

Some engines have issues with the coolant pumps impeller coming loose on the shaft, and hence the pump effectively cuts in and out slightly randomly. may well not be the case here, but if you are still having issues its something to consider. Else as said it sounds a lot like a airlock type issue, assuming its not electrical....

 

In terms of 110deg ejecting coolant, a combination of the cooling loop likely being under mild pressure, and maybe not the most accurate gauge, its not beyond possibility. Equally if there is a stack of air in it the effect will be similar. Water under half a bar of pressure will boiler at 112 if memory serves, and 1bar becomes 121.

 

 

Daniel

[bizzard]

The lime or limescale and calcium content of tap water is quite easy to prove, simply proved by weeing in a bucket over a few days and emptying it when necessary and then examining your wee's calcium content left behind in the bucket, you'll be surprised and wonder how you never got Gall stones.. No matter how you drink the water whether it be just the water or in tea, coffee or with diluted pop it will be the same, in fact I've discovered by experiment that heated tap water ie in tea, coffee ect makes matters worse, as it does in an engine or boiler.

Ahem, after the experiment with tap water has been completed swap to drinking distilled water, rain water, beer ect only for a few days whilst weeing in the bucket and you'll find that the lime- calcium content left behind is much, much reduced, although there will always naturally be a little calcium in your wee.

So to sum up. What's good for your body, 'minus antifreeze of course', must be good for your engine.

[Bee]

The lime or limescale and calcium content of tap water is quite easy to prove, simply proved by weeing in a bucket over a few days and emptying it when necessary and then examining your wee's calcium content left behind in the bucket, you'll be surprised and wonder how you never got Gall stones.. No matter how you drink the water whether it be just the water or in tea, coffee or with diluted pop it will be the same, in fact I've discovered by experiment that heated tap water ie in tea, coffee ect makes matters worse, as it does in an engine or boiler.

Ahem, after the experiment with tap water has been completed swap to drinking distilled water, rain water, beer ect only for a few days whilst weeing in the bucket and you'll find that the lime- calcium content left behind is much, much reduced, although there will always naturally be a little calcium in your wee.

So to sum up. What's good for your body, 'minus antifreeze of course', must be good for your engine.

Well I was about to try this fascinating experiment but then I thought, 'I'll just have a look in the kettle' Sure enough my kettle seems to be developing gallstones.

[alan_fincher]

In respect of what temperature it will boil at......

If there is a working pressure cap, this will increase the temperature before boiling takes place . Sources show that for every pound of pressure exerted on the coolant in the system, the static boiling point of the coolant is raised by approximately 3° F. 7 lb caps are typical, so this alone would increase the boiling point by something like 21° F, (13° C). So if a gauge is accurately recording 110° C, it should not boil.

This assumes plain water as a coolant, though - if it is an antifreeze mix, then the boiling point will be even higher.

Look at tables like this

If we look at the case for an 8 psi cap, and a 3% mix of antifreeze then boling point is 240° F, (nearly 116° C).

(Stating the obvious: This is why you don't remove a 7 or 8 psi cap from a system measuring at over 100° C - if it wasn't boing brfore, it ma well do so in a most dramatic and dangerous fashion if you do!).

[David Mack]

Sources show that for every pound of pressure exerted ...

You'll have the Units Police after you! I think you mean "for every pound per square inch of pressure exerted..."

[alan_fincher]

You'll have the Units Police after you! I think you mean "for every pound per square inch of pressure exerted..."

 

Yes, I know that, but I said it is what some sources say, (which they do - I cut and pasted it).

[David Mack]

 

Yes, I know that, but I said it is what some sources say, (which they do - I cut and pasted it).

Just like others talk about Amps per hour.

[matty40s]

Have you bled your skin tank,?? this is a common problem.

[Boomslang]

All - many thanks for your replies, all food for thought. I believed I had given enough information to head off most of the points raised but, as is often the way, what I thought I had written, how I thought people would read it and how people actually read it have ended up being completely different things!! I

 

have used the bleedpoint at the top end of my skin tank, near where the pipe comes out to return coolant to the engine, so as far as I can see it can be considered bled. Incidentally water almost instantaneously came out of the bleedpoint so very little air was present. I also isolated the calorifier, hooked it up to the hose and pushed water through it, there were no air bubbles in the water and I subsequently topped up with antifreeze. So I think I have eliminated airlocks.

 

Anyone disagree? Anything else I should do?

 

I understand boiling point elevation so am fairly confident it is not a dicky connection and that my temp readings are good. I wouldn't expect 110degrees to blow the pressure cap - if it did this would be a serious flaw in the way the system is engineered. The warning light comes on at 110degrees and you need a buffer between being warned and it going 'pop'.

 

I think the most likely answer is gunking (coupled possibly with air that I cant get out of the system) and/or the impeller. I'm going to back flush the system first to try and eliminate the obvious but if this doesn't work does anyone know how bug a job checking/changing the impeller is?

 

Jared

[Tony Brooks]

Quote: "The point is I know that my cooling system set up is good enough"

 

Quote: "have used the bleedpoint at the top end of my skin tank, near where the pipe comes out to return coolant to the engine"

 

Well assuming its a normal vertical skin tank can anyone see what the two quotes from the original poster implies?

 

 

 

I would suggest the first quote is far from correct (assuming a vertical skin tank) and it may be an idea if the OP posts a photo or photos showing the whole cooling system so we can have a good look at it.

[nicknorman]

Quote: "The point is I know that my cooling system set up is good enough"

 

Quote: "have used the bleedpoint at the top end of my skin tank, near where the pipe comes out to return coolant to the engine"

 

Well assuming its a normal vertical skin tank can anyone see what the two quotes from the original poster implies?

 

 

 

I would suggest the first quote is far from correct (assuming a vertical skin tank) and it may be an idea if the OP posts a photo or photos showing the whole cooling system so we can have a good look at it.

Well it could just be a typo / bit of imprecision. I think what you are getting at is that the hot water should go in at the top of the tank, and exit at the bottom.

[Tony Brooks]

Well it could just be a typo / bit of imprecision. I think what you are getting at is that the hot water should go in at the top of the tank, and exit at the bottom.

 

It was a rhetorical question designed to make the OP think a bit but that is the point, if what the OP has written is true then everything else he has said is open to question at the very least. The bit about rivers makes me suspect an undersized skin tank but that should not impact upon the temperature in canal use.

[smileypete]

Agree with other comments re skin tank size - what are the dimensions?

 

Might also be the thermostat is upside down, or doesn't have a big enough bleed hole. Once the engine is warmed up a blast of hi revs in idle might help to shift trapped air, then keep checking and topping up the header tank when back at idle. A forum search via google should turn up some relevant topics, eg:

 

https://www.google.co.uk/search?num=100&q=Isuzu+overheating+site%3Awww.canalworld.net

https://www.google.co.uk/search?num=100&q=jiggle+pin+site%3Awww.canalworld.net

 

ETA I'd also get a cheap IR thermometer, will give some useful info on engine head and skin tank temperatures - if the skin tank return temp is rather high under load then it could be it is in fact undersized, if it's quite low there could be an obstruction or air lock.

Edited February 5, 2016 by smileypete

[Boomslang]

Tony - i was being imprecise, or too hasty perhaps.

 

The reason I said my set up is good enough is that it has been running fine for over 5 years, on canals and rivers, including against the current.

 

Something has changed/broken/got stuck etc in the system. It's not the thermostat or the pipes because I have changed them; they are all installed correctly and are genuine Isuzu parts.

 

I will post again once I get it fixed...

 

Thanks,

 

Jared

[Tony Brooks]

So is the cold return to the engine at the top of the skin tank? If so where is the hot flow into the skin tank connected and what type of skin tank do you have?

 

What other symptoms of overheating do you have apart from the temperature gauge and have you used an infra-red thermometer on the to confirm the gauge reading?

 

If you have not checked the actual engine temperature and there are no other symptoms like water/steam being ejected from the pressure cap I would suggest the most likely cause is a faulty temperature sender. They are not the most reliable of things.

 

The symptoms you give are indicative of air in the system but not if no coolant or steam is being ejected from the engine.

 

I assume that you have a bi-coil type gauge and if so an intermittent negative at the gauge would cause the needle to periodically jump to full scale deflection and then drop as the connection remakes itself.

 

I have looked through the posts and have not noticed any other symptoms given apart from the gauge reading.

 

 

Your comment "the stat was gunked up" might indicate that at some time in the past the engine had overheated badly enough to cause some of the wax pellet in the thermostat to leak out but if this had happened you would have been suffering from symptoms of late thermostat opening that would have been cured by fitting the new stat. There is still a question over which way up you fitted the thermostat. Is the little canister that holds the wax pellet on the engine side or the skin tank side of the main disk?

 

Normally thermostats just get a soft layer of rust coloured deposits on them, if your "gunk" is anything more and is not wax from the pellet then it could indicate a head gasket failure but again I would expect that to result in loss of coolant. Not simply overheating at intervals and then curing itself while still running.

Edited February 8, 2016 by Tony Brooks

[Jess--]

when my skin tank was air-locking I was getting massive temp variations while running, however contrary to what you would expect the temperature only went high at low revs, increasing revs reduced the temperature to just over it's normal working temp.

 

At the time I was single handed on a short ( > 15 min) trip and in an area where I couldn't stop and allow it to cool / refill so I just plodded on.

 

I got shouted at 4 times for going past boats too fast because I was passing boats with the engine running at 1700 rpm, they didn't realise I had dropped into neutral 100 yards earlier and I was really only drifting past them at a lower speed than I would normally be travelling at tickover.

[by'eck]

So is the cold return to the engine at the top of the skin tank? If so where is the hot flow into the skin tank connected and what type of skin tank do you have?

 

What other symptoms of overheating do you have apart from the temperature gauge and have you used an infra-red thermometer on the to confirm the gauge reading?

 

Not sure if this was a trick question to catch out the OP. If not surely the cold outlet from skin tank should be at the lowest point.

 

Agree a cheap IR/laser thermometer would help in sourcing the problem and like you suspect the gauge and/or sensor as suggested in post #3

[Tony Brooks]

 

Not sure if this was a trick question to catch out the OP. If not surely the cold outlet from skin tank should be at the lowest point.

 

Agree a cheap IR/laser thermometer would help in sourcing the problem and like you suspect the gauge and/or sensor as suggested in post #3

 

Not a trick question but one I raised earlier based on what the OP had written and so far does not seem to have understood. Even if he gets the present "problem" sorted if what he wrote is true then this really should be looked at.