Hi all,

I too am having trouble with the cooling system on my 2.5TD. The symptoms are that the temperature gauge is always high, 95-100, and at the same time the bottom hose coming out of the radiator is ALWAYS stone cold. I took the liberty of creating a schematic diagram for the cooling circuit to try to get my mind around its operating principle. The diagram is attached.

The pump pushes water into the side of the engine block. It travels around the block and up into the cylinder head. The water then appears out of a hole in the head into the 'coolant outlet tank' as Citroen calls it. Its the aluminium block with all the sensors on it at the top right hand corner of the engine viewed from the front. The largest pipe connection is connected to top of the large radiator. The oulet hose from the radiator is connected back to the inlet of the pump where the thermostat is located. The water is now back where it started.

Let me draw your attention to the circuit which takes the hot water from the head, through the heater matrix and back to join the water returning to the pump housing. As you can see the connection is in the vicinity of the sensing bulb of the thermostat.

In a 'normal' car the thermostat is usually located in the position where you can see the large hose coming out of the 'coolant outlet tank'. The sensing bulb is located in the stream of hot water coming out of the head. If the engine is cold and the thermostat shut then the water is only circulated through the head and the heater matrix. Once it heats up the thermostat opens and water is allowed to pass into the radiator pushing cold return water back to the pump. The hot and cooled water is mixed inside the pump by the impeller and the resulting blend is pumped around the engine. That's dead simple to understand.

In the 2.5 as you can see in the diagram the thermostat is located in the return on the inlet to the pump. When the engine is cold and the thermostat is closed the water returning from the heater matrix washes over the sensing bulb and eventually it heats up and opens the thermostat. Upon doing this the cold water returning from the radiator flows over and around the sensing bulb, only mixing with the return from the heater matrix some way downstream of the bulb location. So my point is that as soon as the thermostat opens just a little bit the cold water surrounds the bulb and closes it again. The bulb can only ever see (apart from when its closed) the temperature of the water coming back from the radiator. If open at all it is insulated from seeing the true engine temperature by the fact that the cold radiator return water envelopes the sensing bulb of the thermostat. Its specified operating temperature is 85 closed - 100 fully open. If the water travels through the radiator very slowly the outlet temperature will approach the temperature of the air flowing over it ie anywhere below freezing to +30 DegC (winter to summer). In which case it will never, and can never, reach 85 degrees never mind 100 degrees. This philosophy can never work. The thermostat is closed if the water the bulb is immersed in is below 85Deg.

My next point is that the thermostat can only ever see, at best, the return temperature of the heater matrix which will be by definition several degrees lower then the water going to the matrix and consequently the temperature of the water in the head. How does this logic work?

My next and last point is that on a normal car the thermostat (in truth the thermostatic switch) which controls the fans is mounted half way to 3/4 of the way down the radiator on the outlet side. This means that if the air passing over the radiator naturally is not enough to cool the water then the elevated leaving temperature brings on the fan(s). On the 2.5 the thermostat (in truth a temperature sensor) is mounted on the 'cooland outlet tank' and therefore sees water at head temperature ALL the time. This sensor, via the Bitron controller, brings the fans on OK but if there's no water passing through the radiator then its a waste of time. I've seen both fans running and the outlet from the radiator is stone cold.

My conclusion is not a palatable one but it is this:
The cooling circuit of the 2.5 is designed wrongly.

Now I know thats a brave statement to make but I invite comments either in support of this or to tell me how stupid I've been explaining of course how easy it is to understand its correct operation. And don't think I haven't asked myself how all the 2.5's in the world survive if I'm correct. I don't know the answer to that other than on every Citroen forum I know of there are always people complaining about the cooling circuit on the 2.5's.

I have contaced Citroen UK to talk to their technical department. They would like to charge me £1.50 for the privelege of telling them their design is wrong. I declined and have e-mailed their customer services to elicit a reply from their technical deparment. I have contacted them through the web portal page on their site which both prevents me attaching my sketch drawing and accessing their e-mail address. I don't suppose they will reply but I'll keep you posted.

I have put this post in the general Xm section and not in the diesel section deliberately. I am interested in anyones reply not just the diesel owners.

Cheers

noz

Hi Noz

Well, usually your knowledge of all things XM is far superior to mine, so I generally hesitate to comment, but from your description the thermostat works in a similar mode to that of the 2.1 td, although that has only one thermostat, lacated at the r/h rear of the engine.

On the 2.1 the system works extremely well, you have to drive very hard and very long for the temperature to go much above normal. I have never towed wth it so dont know what effect that would have. I think the 2.1 design was at least in part to prevent air being trapped in the head, which of course is the primary cause of cracking, and it seems to work, I have never heard of a cracked 2.1 head and I have certainly never had one, which is more than I can say for the the 1.9.

I appreciate that the heat output of the 2.5 is greater, but I think it has a bigger radiator. I have heard other 2.5 overheating complaints so I know it must be a problem. If the water temperature around the bulb, as you say, never reaches the coolant temperature, a lower temperature thermostat would seem to be whats required, especially as you say that the return pipe from the rad is mostly cool, which would indicate adequate cooling capacity, but I assume that they dont make one?

I am not really being very helpful here I know, but I thought it might help if you knew someone was sympathising with you.

Peter.

Hi Noz,

what you write makes sense to me, but I'm not knowledgeable enough on the nuts and bolts of the cooling system. What I can think of however is, if your theory is right, then simply replacing the original thermostat with a lower grade one, eg one opening at 65 degrees, would cure the problem, wouldn't it? Or at the really hot instances at least, ie when fresh liquid in the engine is needed mostly. At these cases, even the returning water from the rad would hit the 60s I guess. Experimenting with various grade thermostats would be necessary to strike a good compromise. But, are there any thermostats out there for the job?

regards,
George

Hi Noz and All

I seem to remember reading somewhere that Citroen designed the XM engine coolant system to obtain an unusually rapid temperature rise to get their engines to the steady state operating temperature as fast as possible. This was apparently to get the maximum service life out of a lightweight engine. This might explain why some engines last well while others suffer head poblems. I think I will be a lot less tolerant of varying/over temperature problems in future. At present the only one that gives me concern is the V6 were the temperature climbs to about 95 in town driving. The 2.5's stay rock solid a fraction below 90.

Regards

XMexc

Interesting theory this one...
My V6 does as well 95 in city -so would go much higher if it was not for the fans cutting in-. Around 87 on motorway. I plan to fully flush it, if summer comes this year...

George

PS sorry Noz, bit OT.

Hi George

I have come to the same conclusion as you. Give my V6 coolant system a good flush out sometime this summer. It has the brown gunge on the sides of the header tank so I guess the whole of the cooling system has similar problems. As none of the 2.5's had this brown gunge problem I had assumed that it was just a fact of life with aging petrol engines until I had a chat to Citroenxm the other day.

Regards

XMexc

Allo

Interesting reading guys- I have a different experience with my 2.5 which may provoke interest...my temp gauge never goes above 90 - and then only in really slow& heavy traffic.

Is this related to Norrie's theory about thermostat placement, do I have a serious problem or is my gauge in need of replacement?

Good luck with Cit UK though - keep us posted.

Hi Chadders, Noz and All

The temperature behaviour of the 3 of my 2.5's that I have driven extensively on the road is exactly as you describe, rock solid a fraction below 90 except very occasionally a tad over 90 in extended crawling traffic. Mileage on these cars ranges from 140k to 50k so I assumed this was the correct behaviour for a 2.5TD in normal maintenance state. I do not tow with the cars but I understand that this raises the engine temperature of a 2.5 by 5 to 10 degrees. I suspect the the thermostat on Noz's car is not opening correctly so the temperature behaves like a car under high load such as towing but that is just a guess from the description.

Regards

XMexc

Hi Noz and All

Thought a little more about the 2.5 and cooling problems. Due to local IT difficulties (Only computer available is the wifes overburdened laptop for some months now) I cannot risk downloading Noz's diagram.
With a 2.5 when refilling the coolant if the full pressurising proceedure is not followed to the letter then an air lock is a certainty with adverse performance. Failure to use a pressurised head will lead to air locks. As you do this check you also monitor that the fans are cutting in correctly.
How do you check for an air lock other than monitor for correct working?
10 year old water pumps may have degraded impellors. Anyone got any experience of this? There is no easy check.
From the description the thermostat system may be designed to operate as and on /off switch. The 5 degree temperature drop through the heater that Noz identifies seems about right to me.

Regards

XMexc

Hi all,
Noz says on 29.3.2006 that he is having trouble with the cooling system on his 2.5TD. Referring to his long mail, I can answer as follows.

Para 1. My engine’s bottom hose also is always cold at the radiator end when I check it after a normal run, including motorway, but some miles after coming off it

Para 2. I agree, but would add that the reference to thermostat means the main thermostat in this case.

Para.3. Agreed.

Para 4.Agreed , but not relevant to the set up on a Citroën 2.5TD

Para5. There is an assumption made here that the impeller is only directing the water in one direction – into the cylinder block. Surely by the design of the pump housing, the water is swirling round the thermostat bulb, and at times may be hot enough to open this thermostat.

Para5. The thermostat will not only see the water from the main radiator, but this is now plus the water from the cylinder head outlet (before the small thermostat) via the de-aerator (header tank) plus water from the cylinder head outlet after the small thermostat opens at 85degrees.

The diameter of the TWO inlets to the water pump is greater than the diameter of the outlet from the pump to the engine so water in the pump is under pressure and is only released to the engine when it has been delayed in the pump body i.e. after it has heated the main thermostat bulb. The return pipe to the deaerator is very hot (feel it!) because it is a direct supply from two cylinder head outlets, one before the small thermostat and one after. Note that the small pipe from the small thermostat in the cylinder head outlet must be an outlet not an inlet (as the sketch by noz.)
Normal circulation must therefore add hot water to the top of the radiator and this will add more to the hot water in the main radiator. It would seem that the small radiator and its circulation through the intercooler is therefore a self-balancing arrangement at it can flow in either direction, depending on whether the small thermostat is open or not. Likewise the oil cooler circulation is also self-balancing, though with a much larger cylinder head outlet, uncontrolled by the small thermostat.

The small radiator therefore keeps the engine and the intercooler at a steady running temperature until greater demands are made on the cooling system, requiring the greater cooling as provided by the fans, (on the small radiator), and by the large radiator beyond the ability of the fans on the small radiator. Between them, they will quickly reduce the engine running temperature.

As an observation, I have driven a similar car with the small radiator blocked off, which gave no indication of overheating except when towing,(when the selection of a lower gear had the effect of reducing the temperature to normal levels), which to me seems to indicate that in the UK under normal ambient temperatures there is little need for the large radiator at all!

So, is the problem that noz relates, actually one of four things; the gauge is reading high or its sensor is at fault, or the small radiator is blocked, or its hoses are clogged up?

As the saying is –think on!

Hi all,

The next thrilling installment in the 2.5TD cooling woes....

Peter, Thanks for the sympathy. As we all know XM owners need plenty of that.

George, Unless this debate throws up some other solution I'm heading towards a mod which would turn the cooling/thermostat principle into that which is used on every other car in the world ie sensing bulb located in the circuit just where the water leaves the cylinderhead.

XMexc, interesting theory. I don't suppose you can remember where you heard/read that? It doesn't quite answer my challenge though unless I've still got the thought process wrong ( which of course is highly likely)

Chadders, Sounds like your cooling circuit is behaving perfectly. Presumably you are aware of the fans running when either towing or in the summer months?

XMexc, I suspect you may be right about the non- or slow- opening of the thermostat (see below for explanation). Not sure what you mean about "pressurisation procedure" - where is this written? There are 3 routes to expel air from the system - The air vent at the top of the Coolant Outlet Tank, the air vent on the heater circuit by the bulkhead and the air vent on the rubber hose from the intercooler. The first one is permanently open and works continuously to bleed any air back to the coolant reservoir. The other two must be opened manually to vent any trapped air. I have checked the impeller by sticking my fingers in through the inlet. I can't detect any undue wear but I suppose its possible. The thermostats definitely modulate rather than on off because the mechanism is a wax stat which relies on expansion proportional to temperature. The Bitron control over the fans is on/off but with 3 stages - both off, both slow, both fast. Thats step control (albeit a bit coarse) but I suspect the electronics in the Bitron Unit uses PID control which would compensate somewhat for the coarseness of the output.

Techmanagain,
Para 5 , The design of the pump is not completely clear from the cutaway drawing. The water is directed into the eye of the impeller by a casting in the shape of a smooth cone. The pump is based on the centrifugal principle with the water sucked into the eye of the impeller and discharged at the impeller tips. The circular motion of the impeller is unlikely to impart turbulence at the inlet of the eye and the discharge from the impeller is in another separate part of the casting so I must disagree on this point. Without going into too much detail some of your next comments may be answered by the content of my attachment. I have taken the outlet tank off and photographed the inside. The flow directions can be better understod and agree with my first guess. As for your conclusions on possible causes I am now leaning towards a dicky thermostat. I have bought a new one and will fit it soon. If the problem goes away then so might this debate.

Please find attached a pdf document showing a comparison of the old and new thermostats, a photograph of the inside of the Coolant Outlet Tank and a blown up schematic of the same tank. The flow directions are much easier to understand now that I have seen what is inside the tank.

However, my original hypothesis still stands - as soon as the large thermostat opens, cold water from the radiator washes over the sensing bulb "protecting" it from being able to see the return temperature of the heater matrix or header tank. I will try to take a photo of the pump casting with the thermostat fitted to demonstrate how far the sensing bulb protrudes into the path of the hot water from the matrix and header. By feel, it is not very much.

Thanks to all who have taken the time to reply. After all, the only bad debate is no debate at all.

Cheers

noz

ps 3.0V6 thermostat on ebay right now. Item number 4629699884

Hi Noz,

Interesting problem.

Attached are a number of scans from the Citroen manual showing you are spot on with your schematic and also detailing the refill / bleed procedure for the 2.5TD. Note there are 5 bleed screws in all. They also show what xmexclusive indicates about the pressurised refill - although it's actually a header tank providing the pressure. I hope these are a help. I'd also urge others to follow the refill / bleed procedure to the letter - as xmexclusive states - as I believe this is a killer with 2.5TD's if not done properly because you end up with air pockets around the valve seats which causes overheating resulting in a cracked cylinder head.

I haven't had a chance to check the actual operation of my own car (because it's now too late) but I wonder if the following might help from my observations of the topic.

The point you seem most concerned about is the operation / design of the main thermostat. It may well be the case that the bottom hose remains cold most of the time but the water in the radiators themselves will be much hotter than the max of 30 degrees you suggest - a) because the radiators aren't that efficient and b) because they have hot water circulating from the other circuits. When the main thermostat reaches 85 degrees (if working correctly) it will start to open but the quantity of water passing through it will be small. The cold water that's been sitting in the hose from the bottom of the radiator may or may not shut the 'stat again almost immediately - but the converse of this is that the hot water in the circuit passing over the 'stat wax operating device will open it again almost immediately. If the cool water does close the 'stat quickly there must have been flow (from the bottom of the radiator) and this flow will continue until the (small amount of) cold water in the bottom hose has been circulated at which point the water from the radiator itself will pass over the stat - and this is likely to be considerably warmer - therefore causing less frequent closing of the 'stat. In practice this will all happen very quickly.

The other point which seems not to have been mentioned specifically is that even when the main 'stat is closed there can be a considerable amount of cooling done by the 2 rads due to the circulation from the other circuits. Both rads are always available to cool whatever water is allowed to pass through them - whether that includes the main 'stat or not. (I disagree that the small rad can operate in isolation as it always draws water from the large rad.)

The cooling fans don't work in quite the way you mention. The first fan (o/s I think) cuts in at low speed between 98-93 degrees. At 101-98 degrees the first fan operates at high speed together with the second fan (also at high speed). This sounds a bit like an aircraft from memory.

It seems to me that changing the main 'stat is going to solve your problem - although there's also the possibility of a blockage.

I suspect the majority of overheating problems 2.5TD owners experience are also caused by faulty 'stats and / or coolant circuits that are restricted by debris.

One final word of warning to everyone regarding the bleed / refill procedure. If air pockets do get trapped around the valve seats the temperature guage shows normal during standard driving - you simply have localised hot spots inside the head. The elevated temperature only shows up when it's too late - the head is already cracked and you have a compression leak into the cooling circuit.

P.S. - sorry about the Windows Bitmap format of the scans. This was all I could do with the computer available right now. They do open using the Windows Imaging Accessory - but someone else migh be able to convert them and re-post a more useable document.

Good luck

Roy

OK - I can't seem to get the attachments to upload. I'll e-mail the m to Noz and hope he can post them here.

Cheers

Roy

Coolant System diagram

Refill / Bleed Procedure 1 of 3