996-997 wet-sump engine reliability: enter your stats here!
Discussion
I think that's a good thing actually... Mine's just failed (on a Turbo) and the thought of lots of little bits of sharp metal circulating around the cooling system is not a good one! At least the plastic impeller is unlikely to rip through the rubber hoses and punch holes in the 3 radiator cores!
kVA said:
I think that's a good thing actually... Mine's just failed (on a Turbo) and the thought of lots of little bits of sharp metal circulating around the cooling system is not a good one! At least the plastic impeller is unlikely to rip through the rubber hoses and punch holes in the 3 radiator cores!
The latest cheap aftermarket pump I bought for my Peugeot 406 TDI from ECP had a cast bronze impeller on it. Bronze would give you a nice warning of failure with the noise - a small amount of bronze filings will not do any harm - we're not talking about live jigsaws here - just specks of inert metal 
A complete impeller disappearing around the cooling system in shards of plastic would concern me far more because a) no warning = very inconvenient, b) no impeller = no cooling = overheating and/or stranded, c) Depending how the plastic shatters you may fail to fish out all the bits..
I guess a water pump is really a maintenance item rather than a 'replace when it fails' item, a bit like the timing chain on a TT.
Re the cooling issues mantioned earlier and the lower temperature thermostats idea, an observation: even with new radiators, my 996 (early type) runs hotter, ie perhaps 10c above its normal guage level, to the point that the fans cut in, on warm days, say >18c unless roadspeed is over 45mph or so. Reason: abismal air outlet design resulting in low flow. Ie design cock-up, again. Later 996s and all 997s have slots in the front wheel arch liner to boost flow.
Bought my 996 three and a half years ago. It had 40K on the clock. It now has 65K on the clock. All I've had to do was change the coolant reservoir which cracked and a front rad (I had radiator grills installed at the same time.
The RMS is starting to weep slightly so I'll get that changed when the clutch needs to be done. However, the clutcvh seems to be bullet proof. It isn't showing any signs of needing to be changed, even after 65K miles.
Amazing car.
Pots
The RMS is starting to weep slightly so I'll get that changed when the clutch needs to be done. However, the clutcvh seems to be bullet proof. It isn't showing any signs of needing to be changed, even after 65K miles.
Amazing car.
Pots
Another vote for reason and not the usual bash the 996 brigade, very sorry for all those with issues but you are buying a high end car and as Baz said not seen a car that can't have engine rebuilt. Would you buy a 60k Ferrari and not expect any big engine bills over the next 3 years?
A small percentage of Porsche 996 owners use this board and only the ones with problems are going to bother posting anyway.
For the record my 996 was faultless
Apologies Iphone and beer combo to blame for poor quality posts
A small percentage of Porsche 996 owners use this board and only the ones with problems are going to bother posting anyway.
For the record my 996 was faultless
Apologies Iphone and beer combo to blame for poor quality posts
Edited by lotuselan7 on Saturday 11th December 05:04
Edited by lotuselan7 on Saturday 11th December 05:06
lotuselan7 said:
Another vote for reason and not the usual bash the 996 brigade, very sorry for all those with issues but you are buying a high end car and as Baz said not seen a car that can't have engine rebuilt. Would you but a 60k ferrari and not expect any big engine bills over the next3 years?
A small percentage of Porsche 996 owner’s use this board and only the ones with problems are going to bother posting anyway.
For the record my 996 was faultless
Could you fix your quoting in you post 1-2 posts above?A small percentage of Porsche 996 owner’s use this board and only the ones with problems are going to bother posting anyway.
For the record my 996 was faultless
I'm not really a 996 basher, but I know the issues.
The 996, 997, 986, 987 all have the same flawed engine which would be very lucky to see out 100k miles. The faults have been know for years, IMS bearing seizure, overheating, cylinder wall cracking, RMS seal failures and knocking big-end bearings.
To have this in the pinnacle manufacturer of Germany is actually pretty sad TBH. They had years to fix it and failed. I'm sorry but I like Porsche cars - especially the Boxster, but to have a random unheralded requirement of a new engine every so often stops me buying one (unless I can afford a brand new onw - which I can't), especially when my well thrashed BMW has 165k miles on the clock and still goes as well as new.
Your roll of the dice was lucky, but we know the faults and consequences of this engine design and your own personal luck sadly won't change that for me.
IIRC if you want total reliability these days Lamborghini is hard to beat. That's a high end car...
Globs said:
lotuselan7 said:
Another vote for reason and not the usual bash the 996 brigade, very sorry for all those with issues but you are buying a high end car and as Baz said not seen a car that can't have engine rebuilt. Would you but a 60k ferrari and not expect any big engine bills over the next3 years?
A small percentage of Porsche 996 owner’s use this board and only the ones with problems are going to bother posting anyway.
For the record my 996 was faultless
..which would be very lucky to see out 100k miles. A small percentage of Porsche 996 owner’s use this board and only the ones with problems are going to bother posting anyway.
For the record my 996 was faultless
It must be hard to understand how most engines can be really good yet a small number can fail - but this fact doesn't make all the engines "flawed" because the ones still running are extremely good and can cover over 150K before needing anything (I know I bought one that has just cracked a liner - but otherwise perfect - I am rebuilding it and I am going to continue driving it to see how many miles I can get out of the original parts that remain).
I will shortly be posting information to complete explanations why a very small number fail while others are perfect and they will implicate - in some specific areas - both driver care (or lack of it) and mechanical defects not repaired timely enough - being contributory factors to the failures (and including some tips and parts to improve reliability). Our buyers guide section 4 (www.hartech.org) despite being a few years out of date still contains a lot about these engines and failures and solutions including photos (currently being brought more up to date to include 997's etc).
In a few areas they could have been made a little better (and consequentially be a little more reliable statistically) - but that is easy with hindsight (anyone can be clever with that) and it seems to me that it is the attitude towards assisting (or not) those with problems - that is more flawed than the actual design or numbers failing.
Those board room decisions include not supplying oversized pistons for rebores, not making crankshafts that can be re-ground, not discounting parts when known weak areas fail, not subsidising engine replacement (like BMW did in similar circumstances etc). These create the biggest problem and are nothing to do with the basic engineering that will have been dictated as a policy to the engineers and not the other way around.
However Porsche did put in a huge investment in this model range and generated huge profit from it - but seemed less interested in reputation and customer long term satisfaction than they might have done if the main boards rewards were more linked to long term growth targets than short term gain - but then they had another major agenda on the cards needing cash (taking over VW).
Finally - for those put off buying one because of the extremely small risks - don't please forget that we offer a brilliant way to minimise the risks. Buy a car from us and it is fully prepared and guaranteed, covered with a free period of Lifetime Maintenance Cover, then from £44/month (+Vat) free servicing (including parts), free MOT and free labour for all normal repairs (including engines, gearboxes, brakes, clutched, water pumps etc) FOR LIFE - with no age, mileage or claim limit. IF YOU CANNOT AFFORD THAT YOU SHOULDN'T BE CONTEMPLATING BUYING ONE IN THE FIRST PLACE.
For cars bought elsewhere you need a comprehensive service first and a slightly higher monthly payment - but still no annual up front charges and you can stop whenever you like and it is now transferable.
Baz
I will shortly be posting information to complete explanations why a very small number fail while others are perfect and they will implicate - in some specific areas - both driver care (or lack of it) and mechanical defects not repaired timely enough - being contributory factors to the failures (and including some tips and parts to improve reliability). Our buyers guide section 4 (www.hartech.org) despite being a few years out of date still contains a lot about these engines and failures and solutions including photos (currently being brought more up to date to include 997's etc).
In a few areas they could have been made a little better (and consequentially be a little more reliable statistically) - but that is easy with hindsight (anyone can be clever with that) and it seems to me that it is the attitude towards assisting (or not) those with problems - that is more flawed than the actual design or numbers failing.
Those board room decisions include not supplying oversized pistons for rebores, not making crankshafts that can be re-ground, not discounting parts when known weak areas fail, not subsidising engine replacement (like BMW did in similar circumstances etc). These create the biggest problem and are nothing to do with the basic engineering that will have been dictated as a policy to the engineers and not the other way around.
However Porsche did put in a huge investment in this model range and generated huge profit from it - but seemed less interested in reputation and customer long term satisfaction than they might have done if the main boards rewards were more linked to long term growth targets than short term gain - but then they had another major agenda on the cards needing cash (taking over VW).
Finally - for those put off buying one because of the extremely small risks - don't please forget that we offer a brilliant way to minimise the risks. Buy a car from us and it is fully prepared and guaranteed, covered with a free period of Lifetime Maintenance Cover, then from £44/month (+Vat) free servicing (including parts), free MOT and free labour for all normal repairs (including engines, gearboxes, brakes, clutched, water pumps etc) FOR LIFE - with no age, mileage or claim limit. IF YOU CANNOT AFFORD THAT YOU SHOULDN'T BE CONTEMPLATING BUYING ONE IN THE FIRST PLACE.
For cars bought elsewhere you need a comprehensive service first and a slightly higher monthly payment - but still no annual up front charges and you can stop whenever you like and it is now transferable.
Baz
We have just managed to make section 5 of our buyers guide available on the Internet by going to our web site www.hartech.org and buyers guide section 5 or by this link
http://www.hartech.org/docs/buyers%20guide%20web%2... .
This explains in detail the relevant differences between the 3.4 Cayman S, 3.6 and 3.8 engines (the models that exhibit cylinder scoring problems) and earlier engines, the potential causes and various new solutions and road test results INCLUDING A RELATIVELY INEXPENSIVE SOLUTION.
For those wanting faster answers – whilst increased piston to cylinder wall loading results from increased bottom end torque – at the same time changes to the coolant flow and distribution increase the temperature of localised hot spots and in some combinations of driver style, other older typical wear/age related problems, and general deterioration - push them beyond safe levels in an engine type already running hotter (but previously better balanced) temperatures than older typical designs. The result in some combinations of circumstances is too thin an oil film being unable to keep the piston and cylinder wall apart safely and resulting scoring of the piston and cylinder.
Fitting a lower temperature thermostat (while checking out other contributory factors) will lower the odds of you experiencing this failure while new exclusive pistons and modifications to the inside of the engine during a rebuild - will reduce repair costs further and protect against a reoccurrence.
All this research, design, rebuilding, testing and sharing our results with everyone takes up enormous amounts of time and energy – hence often too little left to promote our excellent standard products and services in a more market orientated – reader friendly style – sorry - but I hope the information will help those unsure what to do to protect their interests, enable them to fully enjoy this fantastic model range and if they are unfortunate enough to experience the consequences (which is extremely unlikely) – remember to call Hartech first before potentially incurring unnecessary expense elsewhere that cannot then undertake the same quality or value for money repairs and preventative measures.
My very best wishes and good luck for the New Year – until then
Baz
http://www.hartech.org/docs/buyers%20guide%20web%2... .
This explains in detail the relevant differences between the 3.4 Cayman S, 3.6 and 3.8 engines (the models that exhibit cylinder scoring problems) and earlier engines, the potential causes and various new solutions and road test results INCLUDING A RELATIVELY INEXPENSIVE SOLUTION.
For those wanting faster answers – whilst increased piston to cylinder wall loading results from increased bottom end torque – at the same time changes to the coolant flow and distribution increase the temperature of localised hot spots and in some combinations of driver style, other older typical wear/age related problems, and general deterioration - push them beyond safe levels in an engine type already running hotter (but previously better balanced) temperatures than older typical designs. The result in some combinations of circumstances is too thin an oil film being unable to keep the piston and cylinder wall apart safely and resulting scoring of the piston and cylinder.
Fitting a lower temperature thermostat (while checking out other contributory factors) will lower the odds of you experiencing this failure while new exclusive pistons and modifications to the inside of the engine during a rebuild - will reduce repair costs further and protect against a reoccurrence.
All this research, design, rebuilding, testing and sharing our results with everyone takes up enormous amounts of time and energy – hence often too little left to promote our excellent standard products and services in a more market orientated – reader friendly style – sorry - but I hope the information will help those unsure what to do to protect their interests, enable them to fully enjoy this fantastic model range and if they are unfortunate enough to experience the consequences (which is extremely unlikely) – remember to call Hartech first before potentially incurring unnecessary expense elsewhere that cannot then undertake the same quality or value for money repairs and preventative measures.
My very best wishes and good luck for the New Year – until then
Baz
Baz, a very interesting article. I know its early days for the 2.9 engine, but in terms of the possible faults you have described would you say this engine is likely to be more reliable. Ie. modified IMS, smaller bores so less chance of cracking, less powerful so less chance of seizing etc.
Baz -- I read that article last night, it is very long but it gives a real insight into what might happen to these engines and most importantly what are the contributory factors. Just fascinating, thank you.
More than ever if you are unlucky to find yourself with such an engine issue and no warranty cover, then locating someone who will 'properly' assess the problem and then fully correct it is the dilemna humble owners find ourselves in. At least now the technical issues are identified and even a non-techie could ask the right questions to identify who had thew ability to sort their engine out.
More than ever if you are unlucky to find yourself with such an engine issue and no warranty cover, then locating someone who will 'properly' assess the problem and then fully correct it is the dilemna humble owners find ourselves in. At least now the technical issues are identified and even a non-techie could ask the right questions to identify who had thew ability to sort their engine out.
CONCLUSION
The Internet has thrown up a useful confirmation of our findings when a contributor called “GT4” showed how to read the coolant temperature by another system by re-setting the existing digital air con control unit on the pre face lift 996’s.
The result confirmed that our own test readings (that showed higher engine coolant temperature readings than the dash board temperature gauge) were right and showed some other interesting facts (that we had also noticed) that add weight to our general conclusion that the engines can run too hot in some circumstances in the cylinder/piston face area.
You will remember (if you read section 5 of our buyers guide) that instead of 100% of the coolant passing each cylinder to cool it – only about 1.5% to 4% passes each individual cylinder and then mixes with the majority of the coolant that goes through the cylinder head before travelling on a relatively long journey back to the radiator and even later reaching the thermostat and back inside the engine again.
This means that if there is a sudden temperature rise inside the engine (as a result of suddenly driving much faster) it is the cylinder walls that will try and transfer this heat first (because the coolant in the cylinder head is far less effective) and because the coolant in the cylinder area is travelling relatively slowly - that temperature rise will be higher. The coolant from the cylinders then mixes with about 80% of the coolant going through the head (the temperature of which will not yet be affected) – so the temperature of the resulting mixture will be only slightly elevated as it travels back to the radiator (where it is then cooled) before even reaching the thermostat. The result must be a system that results in much higher immediate cylinder wall temperatures and much hotter pistons because it reacts much more slowly to dissipate that sudden increase in cylinder temperature than more traditional engine designs.
Now the change in cylinder head gasket design has made that balance different in the 3.4 Cayman, 3.6 and 3.8 engines by about 50% - or put another way – the amount of coolant (as a proportion of the rest going through the cylinder heads) has been reduced in those engines compared to a 3.4 engine – by about half – less through the block and more through the head.
This must mean that the temperature rise it has to handle is higher, the proportion of it mixing with the coolant coming out of the cylinder head is less and the resulting mixture going on to the radiator is cooler – thus creating a further response delay in quickly bringing those temperatures down again (because the thermostat can only react to the whole coolant mix temperature and not the rise just in the cylinder area).
Now if we go back to looking at the difference in the actual coolant temperature readings (between the dash board gauge and either the air con digital signal or our test gauges) – we find that whereas they rise from cold almost together – once the dash board gauge reaches about 80 degrees it becomes very slow to read any sudden rises (that the other instruments pick up immediately) and starts to lag behind the true reading as the temperature rises further.
It means that owners don’t realise that during some dramatic changes in driving – temperatures are rising and falling very quickly and to high maximums – and bearing in mind that this rise came about from a very small amount (that was in the cylinder block) mixing with the cooler majority that is in the head – when we see a sudden rise of 10 degrees C in the mixed temperature (using the digital or test gauges) – the rise in the small amount of cylinder coolant temperature must be many times higher than that and certainly could be in that area of boiling and creating air bubble screening 9all of which our tests confirmed).
Now the problem pistons have a PTFE type coating (which some 3.4’s have and some don’t but has not been a problem in 3.4’s) that has always delaminated on the pistons that have not seized - in engines that have scored bores.
So our conclusion still is that the change in the head gasket design has altered the coolant flow ratios and speeds resulting in higher cylinder wall temperatures while the pistons are transmitting higher cylinder wall loads and the consequence is either that the additional load and heat gradually de-laminates the coating that then jams against the pistons rings (causing scoring) or the increased load/unit piston face area is too much for the oil film to support.
This justifies the changes we make to the coolant flow when we rebuild the engines and also verifies the benefit that fitting a lower temperature thermostat to engines that are presently Ok will have – by compensating for the slow reaction time of the cooling system (because it then starts off at a lower temperature) and allows the slow reacting cooling system to recover back to a sensible running temperature before it has risen far enough to cause damage.
Baz
The Internet has thrown up a useful confirmation of our findings when a contributor called “GT4” showed how to read the coolant temperature by another system by re-setting the existing digital air con control unit on the pre face lift 996’s.
The result confirmed that our own test readings (that showed higher engine coolant temperature readings than the dash board temperature gauge) were right and showed some other interesting facts (that we had also noticed) that add weight to our general conclusion that the engines can run too hot in some circumstances in the cylinder/piston face area.
You will remember (if you read section 5 of our buyers guide) that instead of 100% of the coolant passing each cylinder to cool it – only about 1.5% to 4% passes each individual cylinder and then mixes with the majority of the coolant that goes through the cylinder head before travelling on a relatively long journey back to the radiator and even later reaching the thermostat and back inside the engine again.
This means that if there is a sudden temperature rise inside the engine (as a result of suddenly driving much faster) it is the cylinder walls that will try and transfer this heat first (because the coolant in the cylinder head is far less effective) and because the coolant in the cylinder area is travelling relatively slowly - that temperature rise will be higher. The coolant from the cylinders then mixes with about 80% of the coolant going through the head (the temperature of which will not yet be affected) – so the temperature of the resulting mixture will be only slightly elevated as it travels back to the radiator (where it is then cooled) before even reaching the thermostat. The result must be a system that results in much higher immediate cylinder wall temperatures and much hotter pistons because it reacts much more slowly to dissipate that sudden increase in cylinder temperature than more traditional engine designs.
Now the change in cylinder head gasket design has made that balance different in the 3.4 Cayman, 3.6 and 3.8 engines by about 50% - or put another way – the amount of coolant (as a proportion of the rest going through the cylinder heads) has been reduced in those engines compared to a 3.4 engine – by about half – less through the block and more through the head.
This must mean that the temperature rise it has to handle is higher, the proportion of it mixing with the coolant coming out of the cylinder head is less and the resulting mixture going on to the radiator is cooler – thus creating a further response delay in quickly bringing those temperatures down again (because the thermostat can only react to the whole coolant mix temperature and not the rise just in the cylinder area).
Now if we go back to looking at the difference in the actual coolant temperature readings (between the dash board gauge and either the air con digital signal or our test gauges) – we find that whereas they rise from cold almost together – once the dash board gauge reaches about 80 degrees it becomes very slow to read any sudden rises (that the other instruments pick up immediately) and starts to lag behind the true reading as the temperature rises further.
It means that owners don’t realise that during some dramatic changes in driving – temperatures are rising and falling very quickly and to high maximums – and bearing in mind that this rise came about from a very small amount (that was in the cylinder block) mixing with the cooler majority that is in the head – when we see a sudden rise of 10 degrees C in the mixed temperature (using the digital or test gauges) – the rise in the small amount of cylinder coolant temperature must be many times higher than that and certainly could be in that area of boiling and creating air bubble screening 9all of which our tests confirmed).
Now the problem pistons have a PTFE type coating (which some 3.4’s have and some don’t but has not been a problem in 3.4’s) that has always delaminated on the pistons that have not seized - in engines that have scored bores.
So our conclusion still is that the change in the head gasket design has altered the coolant flow ratios and speeds resulting in higher cylinder wall temperatures while the pistons are transmitting higher cylinder wall loads and the consequence is either that the additional load and heat gradually de-laminates the coating that then jams against the pistons rings (causing scoring) or the increased load/unit piston face area is too much for the oil film to support.
This justifies the changes we make to the coolant flow when we rebuild the engines and also verifies the benefit that fitting a lower temperature thermostat to engines that are presently Ok will have – by compensating for the slow reaction time of the cooling system (because it then starts off at a lower temperature) and allows the slow reacting cooling system to recover back to a sensible running temperature before it has risen far enough to cause damage.
Baz
Gassing Station | 911/Carrera GT | Top of Page | What's New | My Stuff