Ducato X250 fuel filter leak

[snowie]

There was a thread on this problem in July last year.

I have had 2 previous occurrences, and had a third last Friday. This is over a 9 year period, with a total mileage of approx 57,000 miles.

This is very low mileage for a commercial X250 van.

I am wondering if the filter housing is just a faulty design that has a short life. I imagine that Fiat have moved on, and are unlikely to be interested in producing a new assembly for a van of this age.

I have seen it recommended that the complete assembly, including filter, should be replaced.

This on the basis that the “lid” does not then need to be tightened. However, this assumes that the correct torque is used by Fiat Spares.

 

I called the AA out and the mechanic stripped the assembly, checked it for damage, and reassembled it. I didn’t see a torque wrench being used. I am not confident that this assembly that has been taken apart and reassembled several times will prove reliable on a long European trip.

 

So, do I take my van to the local Fiat Professional garage or trust my local garage?

Finally, I “heard” somewhere that a metal filter housing assembly was available for the X250, is that so?

 

Regards Snowie

[monique.hubrechtsgm]

snowie leave it to fiat.They are aware of the problem whit their UFI filter housings, being plastic. They use a metal version now who can be retrofitted. But whit other connections. Have a look at coastal motorhomes.

[Keithl]

Snowie,

 

I believe a genuine UFI replacement filter assembly is only something like £70 to £80 so buy a spare and carry with you.

At least then if you do 'spring a leak' you can simply change the whole assembly and carry on your way.

 

Keith.

[Derek Uzzell]

There are quite a few threads about leakage from the plastic fuel-filter housing fitted to Fiat Ducato X250 models with Euro 4 motors.

 

In his posting of 14 August 2015 12:01 PM on this 2015 forum thread

 

https://forums.outandaboutlive.co.uk/forums/Motorhomes/Motorhome-Matters/Fiat-Ducato-Fuel-Filter-Replacement/39103/

 

Nick Fisher (euroserv) mentioned that attempting to replace the problematical ‘Euro 4’ plastic-housing filter-unit with the disposable metal canister-type fuel-filter used on Euro 5 Ducatos would not be a realistic proposition.

 

The reasons Nick gave were that not only were the position and orientation of the two filter types’ fuel-hose connectors completely different, but the Euro 4 plastic housing carried an electrical connection whereas the Euro 5 metal filter did not. Photos 1 and 2 below show the Euro 4 plastic-housing filter-unit (with the orange electrical connection very apparent) and Photo 3 shows the Euro 5 disposable metal fuel-filter canister without any electrical connection.

 

Euro 5 Ducatos began to be marketed in 2011, but for any Euro 4 Ducato X250 with the plastic-housing filter-unit (as on Snowie’s motorhome) my understanding was that it is not possible to swap this for the later Euro 5-onwards metal canister type. (My understanding seems to be at odds with Monique’s advice about retro-fitting, so it might be worth contacting Coastal Motothomes about this.)

 

As Keith has said, a ‘Euro 4’ UFI plastic-housing complete with filter can be obtained for £70-£80 (example advert here)

 

https://coastalmotorhomes.co.uk/fiat-ducato-peugeot-boxer-citroen-relay-fuel-filter-housing-2-3-3-0-complete.html

[Laika.brian]

 

Snowie.

 

I carried out fuel filter element replacement on our 2007 X250 before our trip to UK last summer, so far no issues, have to admit to been very worried on our travels.

 

I believe to do this work correctly the complete unit needs to be removed, a vice is required, the special tool and a torque spanner also, Dimond Dave has a very good video .

 

I purchased a inline filter , to this I attached two short hoses, which should connect to the inlet and outlet fuel pipes and be secured with jubilee clips at the filter. This is my emergency kit.

[snowie]

Thanks folks; having checked locally with Fiat Professional parts department:.........I shall be ordering one, or possibly two OE Fiat assemblies.............from Coastalmotorhomes, no contest.£83 versus £250!

I don’t think a metal one is available as a direct swap so I’ll not persue this.

Regards

Snowie

[Derek Uzzell]

Apparently the electrical connection on the top of the plastic-housing of the earlier (Euro 4) filter relates to a fuel-temperature sensor.

 

https://www.fiatforum.com/ducato/422290-ducato-fuel-temp-sensor.html

[Airstream]

https://www.ebay.co.uk/itm/OEM-Fuel-Filter-Housing-for-Fiat-Ducato-244-250-290-2-0-2-3-2-8-3-0-1368127080/332821745182?fits=Car+Make:Fiat%7CModel:Ducato&epid=23024258689&hash=item4d7db8ce1e:g:fwoAAOSwuGZbrNJk&redirect=mobile

Only £66 here

Ray

[Derek Uzzell]

I notice the following 19 December 2018 comment from a buyer of the filter-unit advertised on the ebay link you provided

 

“What I wanted

 

Check that the cap is torqued to 30Nm, mine wasn't, using Laser 5843 Filter Tool and a reliable torque wrench. Don't be fooled by yellow tamper marks and assume that it is. It will leak otherwise. Removing the old housing isn't easy as the fuel pipes are difficult to disconnect. You need to pinch the grey flat sides of the fuel pipe retainers very hard and pull on the fuel pipe. Some instructions would have been helpful.

 

As people will tend to buy this filter-unit believing that it has been correctly assembled, and probably won’t have a suitable torque-wrench and/or the appropriate tool to hold the unit without potentially damaging it, this comment is more than a little concerning.

 

(For what it’s worth, the filter-unit is a few pence cheaper (£64.99) from here)

 

https://tinyurl.com/ydhozk32

[owendavies54]

I have recently experienced the dreaded Ducato fuel filter leakage problem and two intelligent garages and an excellent Britannia Rescue fitter later I had still not solved the problem. Having spent most of my working life trying to keep a multitude of fluids and gasses inside pipes I thought I would buy the Laser tool and spend a bit of time trying to understand what’s going wrong.

Firstly, the successful erection of any joint is dependent on seal compression and although a fastener torque figure is widely used to achieve it, it is only ever a pretty poor substitute to compression measurement. It suffers from so many potential errors: torque spanner calibration error, thread matching, thread finish, cleanliness and lubrication to name but five.

So, to achieve a tight joint with an O seal what compression do you actually need? Typically between 10 and 40% of the uncompressed thickness so for this seal which is approximately 3.5mm thick the compression should be between. 0.35mm and 1.4mm.

So how do you know how much you have compressed the seal? Measure the thread form and it looks to me as if one revolution of the clamping ring will move it down, compressing the seal, by 3mm.

If we therefore take a mid-point compression of 25% (half way between 10% and 40%) which is equivalent to a compression of 0.25x3.5=0.875mm. we need to work out how far to rotate the ring to achieve this.

First having carried out all the advice that Diamond Dave provides in his excellent video (Google Search for: How do I change a Ducato Fuel Filter?), wind down the clamping ring gently and by hand until you cannot detect any vertical movement between the green inner body and the ring. Exact accuracy isn’t essential and in practice it’s not difficult to detect. Now with a piece of chalk or tape mark the position of the clamping ring and the green body so that as you rotate the ring you can see how far you have gone. If you now take a steel or other flexible tape measure and measure around the circumference of the green body you will find that it is roughly 28cm. If one revolution of 28cm compresses the seal by 3 mm then to achieve a compression of 0.875mm we must turn it 28x0.875/3 cm= 8.2cm. Using the tape measure put a mark on the green body 8.2cm clockwise from the first mark, take the tool and spanner and rotate the ring until you reach the mark. The spanner I used was a ‘bendy bar’ torque wrench because I was interested in comparing the compression with the specified torque figure of 30Nm but you don’t need to use a torque wrench. I noted two points: firstly, that the point was pretty close to 30Nm and secondly that it was at the point where a small rotation resulted in a fairly large torque change. Bendy Bar torque wrenches are pretty inaccurate but since I knew that at 30Nm it had already failed twice it was interesting to know that it was probably within the 10-40% band. I rotated it further to 10.5cm (a torque of about 37Nm) which gave a compression of 10.5/28x3mm=1.125mm or 32%. All I can say is that it isn’t leaking….yet, and if it does, I shall go up to a 40% compression. (Equally I shall carry a short length of fuel hose and two jubilee clips to bye-pass the filter in a hurry.)

There is one other point to make, because the fuel pressure comes onto the seal from the gap between the two parts of the filter body, the seal needs to be sitting in the housing in contact with the outer face so that there is no lateral movement when the pressure comes on. To ensure this happens if you look at the seal you will see it has little bumps on its inner surface. Just make sure that these are facing inwards.

Finally, in case they change the design, check my measurements of seal and thread size on your filter.

Owen Davies CEng. FIMechE

 

[Airstream]

When I replaced the fuel filter on my 2.3 x250 I was away from the required info just working from memory re Diamond Daves video, in error torqued the filter cap to 30 ft lbs

However it has not leaked so maybe the above info ie torque higher than the official figure works

Time will tell

Ray

[owendavies54]

Thanks Ray, that would suggest that (at about 40Nm) you are somewhere around 35% compression. Between us and the many other articles I have read I am beginning to think that the manufacturer's 30Nm figure, although not silly, is perhaps on the lower boundary of success. I am not an expert on polycarbonates but it may be that there is a lubricant that the manufacturer uses that helps the threads glide more easily and achieves a higher compression for the same torque. Alternatively it is a slightly unusual asymmetric thread form with an almost flat mating/loaded surface and a rounded reverse and it may be that the material is prone to galling (rather like stainless steel) and that the surface finish deteriorates with use so that it glides more easily on first use.

 

Essentially my message is: forget the torque wrench, measure rotation and achieve a compression of somewhere above 30% or about 1.2mm.

[Deneb]

I always lubricate seals between rotationally tightened components with a film of engine oil or light grease before assembling, to prevent the seal being damaged or distorted during tightening. I assume Diamond Dave's video covers that? I haven't watched it.

 

I do however have a new, unused and boxed filter service tool for the X250 filters (amongst others). If anyone would like it, make me an offer. I will have to consult Mrs D on postage costs, being as she is the expert in that area, always selling stuff on eBay.

[Derek Uzzell]

Deneb - 2019-01-26 12:13 PM

 

I always lubricate seals between rotationally tightened components with a film of engine oil or light grease before assembling, to prevent the seal being damaged or distorted during tightening. I assume Diamond Dave's video covers that? I haven't watched it..

 

Yes, Dave Newell’s video does mention lubricating the filter-unit’s threads, and this was discussed (argued over ;-) ) in this earlier forum thread.

 

https://forums.outandaboutlive.co.uk/forums/Motorhomes/Motorhome-Matters/fuel-leak-Ducato-2-3-M-Jet/49789/

 

It could be hypothesised that the reason the UFI filter-unit carries a 30Nm marking is because there are dimensional variations unit-to-unit and/or the exact dimensions of the O-seal cannot be guaranteed.

 

If every filter-unit and O-seal were absolutely identical, there would be no need to torque the clamping-ring to a particular setting. Instead there could be a marking on the filter-unit’s body and another on the clamping-ring, and all that would be needed to obtain a perfect seal would be to screw the clamping-ring on to the filter-unit’s body until the two marks lined up. But, if there are manufacturing/material variations, advising a ‘fixed’ torque (30Nm in this instance) could allow for them.

 

Increasing the advised torque-setting by some 30% (which Airstream has done accidentally) and then suggesting that such ‘overtightening’ might be beneficial because his filter-unit is not currently leaking seems to me like a very big leap in the dark.

[Deneb]

Derek Uzzell - 2019-01-26 2:23 PM

 

It could be hypothesised that the reason the UFI filter-unit carries a 30Nm marking is because there are dimensional variations unit-to-unit and/or the exact dimensions of the O-seal cannot be guaranteed.

 

 

I suspect that tolerances are as, if not more likely to result from dimensional variations in the plastic body and lid due to wear of the production moulds over their life cycles.

 

My last and current cars both have oil filter elements contained within plastic bodies which screw onto admittedly metal parts of the engines, also sealed with O-rings. Both have specific tightening torque figures quoted by the vehicle manufacturers, although it was quite apparent from my post servicing checks on the occasions when I had to have the cars dealer serviced, that their workshops paid absolutely no attention to them.

[owendavies54]

Derek, I think you are being slightly pessimistic here if you will excuse me saying so. Look at my original article and start with the O-Seal design. Obviously I cannot be certain of the material characteristics hence my 'typical' statement that O-seals are designed within the on-duty compression of between 10 and 40%. Don't use torque, use compression, that's the key ingredient to a satisfactory joint assuming that the seal has been correctly designed for a sensible volumetric fill on compression. I respectfully suggest to you that torque will not cater for manufacturing tolerances because it will not guarantee that you achieve the correct compression, that's one of its weaknesses, it only works if all the variable factors are the same as the conditions in which it was originally specified.

 

You might legitimately argue about the risk of over-torqueing the ring but again assuming that it has been designed with a standard design margin of 1.5 that wouldn't be of concern until you went over 40Nm and almost certainly the critical design criteria for the ring will be the strength required to transmit the thread root stress into the ring rather than circumferential torque.

 

So why do people specify torque ? The answer here is that it is easily understood and in most cases it's good enough and it's cheap BUT in applications where compression really matters we wouldn't dream of using it alone (or even at all) we use devices such as direct tension indicators and compression sensors and various other devices that directly measure what we need to know. So why don't they mark the rim ? I can't answer that unless the can't always guarantee to fit the same o-seal, hence back to my original thread, follow my advice and always measure the o-seal thickness. In passing I would also say that in a polycarbonate component such as this the manufacturing variation is, I suspect, modest.

 

Finally I ask you what you have to lose? I have watched three good people fail to seal my filter at 30Nm. Ray and I have, by applying a slightly higher torque achieved a degree of success: how long for I don't know but it has so far saved me about £60 which is one or two beers. I also know that across the correspondence there are dozens of people who have suffered from this problem and there are some who have achieved success by the use of PTFE thread tape (which might possibly ease the thread gliding) or by tightening it without a wrench as tightly as they dared. Give it a go, what's to lose if you have a leaking filter? The more people who do it and report success the more confident we can all be.

 

 

[monique.hubrechtsgm]

UFI is a well known brand. On my new diesel is a Mahle fuel filter. No screwed head in metal. Fiat have it also. Go for a retrofit of one of these. And prevent loosing your diesel on the high way The problem is do to a proper leaktest in situ at what pressure. I prefer some Loctite on the threads instead of grease, and silicone on the rubber gasket. High torque can decompressed it. Like on your gas connections. The fuel pressure of of the submersible transfer fuel pump in your tank who keeps running at a leak can be fatal.

[owendavies54]

Thank you Monique. I certainly wouldn't reject the idea of Silicone on the o-seal. Loctite is presumably to overcome the risk of loosening in service. I have marked mine but haven't YET seen any signs of this happening. Your English is far superior to any attempt I might make in your language so I hesitate to ask you to explain, but I am not quite sure what you mean by 'High torque can decompressed it.' Please recognise that I am not suggesting that we compress the seal beyond what one would consider to be an entirely normal and acceptable degree of compression. I just think that the manufacturer's figure which may be fine for a new component, is on the lower boundary of success - sometimes it works well, sometimes only for a short time and sometimes not at all. I have personally had all three cases and I think a slight adjustment is entirely appropriate.

 

 

[Derek Uzzell]

monique.hubrechts@gm - 2019-01-26 4:04 PM

 

...On my new diesel is a Mahle fuel filter. No screwed head in metal. Fiat have it also. Go for a retrofit of one of these. And prevent loosing your diesel on the high way...

 

Monique

 

Where have you found the information that, if a 2006-2011 Fiat Ducato X250 vehicle has the plastic-bodied UFI fuel-filter unit, it will be practicable to replace that unit with a metal fuel-filter (presumably the metal canister filter fitted to later Ducato X250sand X290 models)?

 

I’ve looked at several non-UK motorhome forums (German and French). When those forums have included an enquiry about whether the plastic-bodied filter can be replaced with the later metal version (and the question is quite common). The answer given has always been No.

 

The metal canister filter began to be fitted to Ducatos in mid-2011 and, if it were straightforward to replace the plastic filter-unit with the metal filter, one might expect that option would be well known by now. There is a lot of advice on how to replace the filter cartridge of the plastic-bodied filter-unit, but I’m not aware of any advice saying “Why bother? Just fit the metal filter instead - and this is how to do it.”

 

When leakage from the plastic-bodied filter-unit is discussed, it’s intriguing to read the explanations as to why this happens. For example, that the basic design of the unit’s two primary plastic parts is flawed so that the O-seal is inadequately located and/or cannot be properly compressed. And when leakage occurs after the cartridge has been changed, it’s suggested that this may be due to the proper tools having not been used, or the unit’s clamping-ring being under-tightened or over-tightened.

 

One interesting caveat (stated authoratively) was that only the UFI-made cartridge should be fitted as a replacement, as all other-brand cartridges are slightly longer and this prevents the filter-unit’s ‘lid’ from mating correctly with the unit’s lower section.

 

There are also caveats about there being differences between the the metal canister filters (not just whether they have a heater or not) with it being advised that a metal filter suitable for a Ducato X290 won’t be suitable for a metal-filter-equipped Ducato X250 even though the filters look similar.

[monique.hubrechtsgm]

Decompression on a gasket not O-rings is complex. On your gasbottle you have a rectangular gasket who should be compressed by hand coupling. But Truma and GOK market leaders have a plastic wrench. A metal tool is not allowed be cause of spark risk. But oil tankers and fefineries use Stelcon - bronze metal tools to stop leaks in explosive areas and are spark free. Not sure UFI uses a O-ring. Unlikely.

[monique.hubrechtsgm]

Derek you, not me noticed a canister disposable metal filter for the ducato. Like you said like a oil filter. A commonrail diesel filter has a bracket and location. The canister type does not separates water. and has no electric contacts. And have connections coded : Diesel tank- fuel cooler- tandem pump- fuel rail return. The cold fuel and hot fuel monitoring is outside via the ECU. Whit winter diesel up to - 26 deg c That is what i have now in my van.

[Steve928]

Whilst Fiat moved to the metal canister filter for their Euro5 Ducatos, PSA continued with the plastic UFI filter until the change to the BlueHdi Euro6 powerplant. My current 2015 Boxer, therefore, van still has the plastic filter housing under discussion.

I think that I've now done 4 filter changes (on this and the previous van) using the correct tool and without issue thus far.

I used to carry the tool and a spare filter in the van as some form of insurance but I've now adopted a different method: using 2 x 8mm male quick release fuel connectors, a length of hose, a T piece and a bicycle valve adapter I made a test loop to pressurise the canister after a filter change using my cycle track pump.

I wasn't able to determine at what pressure the in-tank fuel pump operates but experience (and holding my thumb over the end of the inlet pipe with the ignition turned on..) suggests it would be well under 10 p.s.i. I go to 30 p.s.i and if it holds at that consider it good to go. It always has thus far.

If I had a spare filter housing I might do some testing with various pressures and torques but for now that's good enough for me.

[Derek Uzzell]

monique.hubrechts@gm - 2019-01-27 3:55 PM

 

Derek you, not me noticed a canister disposable metal filter for the ducato. Like you said like a oil filter. A commonrail diesel filter has a bracket and location. The canister type does not separates water. and has no electric contacts. And have connections coded : Diesel tank- fuel cooler- tandem pump- fuel rail return. The cold fuel and hot fuel monitoring is outside via the ECU. Whit winter diesel up to - 26 deg c That is what i have now in my van.

 

Monique

 

This is what you said above

 

snowie leave it to fiat.They are aware of the problem whit their UFI filter housings, being plastic. They use a metal version now who can be retrofitted. But whit other connections. Have a look at coastal motorhomes.

 

Perhaps you could explain, please, what "They use a metal version now who can be retrofitted” was intended to mean?

[owendavies54]

Steve, Your pressure test kit sounds a a really sensible step. Convention has it that one tests mechanical seals in service to about 1.2 times their normal working pressure. One tends to avoid going to 1.5 times (or beyond) which would be near the proof test of the whole body. My judgement is that maybe your 30 p.s.i is a touch high and risks driving any cracks that might exist at the highest stressed areas of the component such as the thread roots towards failure. In an ideal world one conducts pressure tests such as these hydrostatically (but for obvious reasons you don't want water in the body nor would diesel be a sensible test medium) so that if there is a failure there isn't the release of the stored energy in compressed air that would cause it to shatter in one's face. If one is forced to use air one would increase the pressure in steps checking for leaks or distortion in say 5 psi steps to reduce the risks (sorry to preach a bit here but if it did shatter at 30psi a shard of polycarbonate in someone's face would not be funny). In passing you have probably inadvertently demonstrated that the clamping ring is a pretty robust feature which can accept a bit more than 30Nm without flinching. Thanks for this information - 'rush along steady' as they say in this part of Devon.

[monique.hubrechtsgm]

Splendid Steve about your leak test of the fuel filter.