The question we have been asked a million times since we first posted about the breakdown has been – “just what went wrong?”
The symptoms defied a clear and obvious explanation, and even now we are still not absolutely certain.
After all – Zephyr is a 1961 bus with a 1973 engine block that hasn’t had much professional maintenance since the mid-80’s, and the bus had sat mostly idle and neglected for the 15 years before we bought her.
We knew when we bought the bus (for just $8,000!) that we were buying into a mechanical unknown, and we immediately set aside the necessary funds and planned for what we considered inevitable major engine work.
If anything – we were actually surprised that we had made it two years without any serious engine issues. Even several gurus of the bus world had commented at the “strong runner” we had lucked into with no external signs of the engine tired or worn out.
But… On June 14th 2013, we found ourselves stranded on the side of the road, with coolant boiling out all over remote Montana highway 212.
Unlike most overheating scenarios, our engine rapidly overheated going downhill after having successfully climbed a mild mountain pass with the engine temperature having barely budged above normal on the way up.
When we dissected the engine, two of the lower cylinders showed signs of extreme heat damage.
But what was the initial failure that lead to the cylinders melting down? The thermostats and coolant pump checked out as fine, as did the radiator and fan, and all of the other relevant engine internals.
The one thing that was glaringly and obviously wrong when we took the engine apart – the thing that caused our mechanics to exclaim “well, there’s your problem” when they saw it, was our air filter – a Baldwin PA2721.
See for yourself – it doesn’t take a trained eye to realize that something is wrong here:
This filter had been installed less than a year earlier, and had just 4,062 miles on it. And it had been installed with a Filter Monitor attached to indicate when the filter was becoming obstructed and was ready to be changed. According to the Filter Monitor, during our pre-departure systems check the morning of the breakdown the filter still had plenty of life left.
Something had caused the air filter to fail catastrophically.
And though there is no way to know for certain that the air filter failing (and being partially ingested by the engine) was the root cause of our engine overheating and subsequent failure, very bad things can happen when an engine breathes unfiltered air and bits of torn up paper.
Did we have a defective filter? Was this what had killed our engine? Or was the filter an innocent bystander in the engine failure?
When we began to research what might have gone wrong – we discovered that Baldwin had an investigative process where they would look into failures, and they encouragingly offered that “if a Baldwin product is proven to have caused damage, Baldwin will pay the cost required to repair the equipment to its condition at the time the failure occurred.”
Though getting some of our repair work covered under warranty was appealing, we were even more interested in getting to the bottom of what had failed – and what could be done to prevent a future failure.
We promptly contacted Baldwin, and sent them photos of the filter.
The initial response from Baldwin Service Engineering on July 3rd was:
“The PA2721 is designed for outside to inside flow direction. The pictures illustrate the media was displaced due to an inside to outside flow direction. There is not an external wrapper on the media pack of the PA2721 to support the media in an inside to outside flow direction.”
Wait – what?! Had our filter been installed backwards?!!? Had we used an incorrect filter???
Though the GM 4106 was initially designed to use a big messy “oil bath” style filter, our bus had long ago been modified to use a large canister-style filter instead. According to the Detroit Diesel specifications (and confirmed by the mechanics at Interstate PowerSystems, and many of our bus guru mechanically inclined friends), the new filter style and size was a good match for our 8V71 engine.
The Baldwin PA2721 is the third air filter we have used in our bus since purchasing it. Our bus initially came with a Farr EcoLite C-62891-1 U — which had been in use for an unknown number of years and was extremely filthy (but structurally sound) when we replaced it. The second filter was a Wix 46891 – we replaced this after 1 year / 10k miles of trouble-free use.
The Baldwin PA2721 is marketed as a direct replacement for both of these filters, with no caveats given at all about flow direction.
I happened to have pictures of all three filters to compare.
The Farr EcoLite and Wix filter canisters are explicitly labeled as being bi-directional. The Baldwin canister lacks any flow direction labeling whatsoever.
I also noticed that visually comparing photos of the three filters, while constructed similarly – the failed PA2721 seems to have noticeably fewer of the structural support beads along the outside of the filter. In an “inside to outside” flow direction, these would be the primary support to hold the filter together against the airflow.
Also interestingly – if you look at the picture of the PA2721 on the Baldwin E-Catalog, there is a support mesh present on the outside of the pleats that is not present on the actual PA2721 design as we had been shipped.
Had the design been changed?
When I confronted Baldwin with this information, on July 30th Baldwin’s Service Engineering Manager changed his initial assessment:
“In the interim, I have done some research on this style of product with regard to by-directional flow. This product may be used in either inside-out flow direction or outside-in flow direction, just as the filters it is designed to replace. The medias used in air filters, regardless of brand, have a high capacity side and a low capacity side. Typically, the high capacity side of the media is on the side of the incoming air. However, it does not prevent the filter from being flown in the opposite direction. In fact some filters are made specifically to have the low capacity side of the media on the incoming air side of the filter to reach the filter’s accumulative efficiency sooner.”
Baldwin asked us to send them the filter so that they could conduct a full investigation, and knowing how many people were following our engine rebuild story – we stopped speculating publicly on the causes of our engine failure, or mentioning the air filter situation further. We wanted to give them ample time to respond.
I had my own theories:
- Perhaps the PA2721 suffers from a general design flaw that limits its capabilities when used in an “inside to outside” installation, and the filter is not able to meet its design specifications when used this way. After all, even Baldwin Service Engineering was not initially aware that the filter is specced to be used in either directional flow.
- Or… perhaps our particular PA2721 suffered a manufacturing defect where insufficient beads of hot-melt were applied to keep the filter structure intact when under “inside to outside” pressure. This would explain why our filter had fewer of these support beads than the comparable Farr and Wix filters.
After weeks of analysis, we heard back from Baldwin.
All Baldwin could conclude was… the filter had gotten wet.
We checked with both Choo Choo Express Garage that had installed the Baldwin filter, and Interstate PowerSystems that had removed it. And all the mechanics agreed that the filter was installed in such a way as to be completely removed from any possibility of direct water exposure.
One of these assessments we were sent:
“I don’t know where water intrusion would get up and into the filter, situated where it is at up in the engine compartment. I would think that if it got soaked enough to collapse while running that you would have had major major issues with the engine itself also at the time it occurred.”
Seeing as the filter failed after less than 10 months (and just 4,062 miles) in service – the only explanation left (other than a manufacturing defect) that seems plausible based upon Baldwin’s analysis is that the filter was unable to withstand the humidity and condensation of a typical winter in Florida, where we spent autumn through spring while dealing with a family member passing away.
Baldwin Goes Deeper
I continued to correspond with Baldwin, and on September 23rd they told me that they had done a more in depth analysis, and had concluded:
“We have revisited this investigation and conducted further examination into the returned filter. We had a portion of the housing removed to provide us with a larger view of the ruptured area of the media and the condition of the media in the vicinity of the rupture. From our initial review of the media, it is apparent that the media was subjected to moisture, as indicated by the moisture stains on the media. Please see the attached pictures showing the yellow staining of the media. Removal of a portion of the housing where the media ruptured revealed additional moisture staining of the media. When the media of an air filter is subjected to moisture, its restriction to airflow can increase exponentially, varying on the degree of moisture. Structural integrity testing of the design of this product has shown the design to endure at least 63” of water with no evidence of media rupture. Given the fact that the air filter should be serviced when a restriction of 20”-25” is reached, the structural integrity of this design will endure more than 2X the recommended maximum restriction the filter should endure in service.
Through review of the information we have received, it is evident that the media was subjected to moisture. We are unable to determine how much moisture or the method by which the media became wet. Nonetheless, the evidence indicates that the excessive restriction created by the wet media and the compromised structural integrity of the media contributed to the media rupture.”
I remain very puzzled as to why the filter minder was not indicating any signs of imminent failure – the filter minder the morning of the breakdown was reading in the green zone (between 15″ – 22″ – closer to the 15″ mark).
If the filter had become obstructed due to moisture, the filter minder should have indicated a problem long before the critical 63″ of water rupture point that Baldwin’s letter references.
We made it clear to Baldwin that we were less interested in getting any sort of reimbursement than we were in getting to the root cause of the filter failure, so that we can avoid similar problems in the future – and help others avoid a similar fate as well.
Baldwin invited me to send them any further questions, and this is what I told them I would love to have answered:
- You mentioned the structural integrity testing of the PA2721 filter shows the filter to be able to “endure at least 63in of water with no evidence of media rupture” – has this testing been performed with both inside-out and outside-in flow directions? Is there a difference in the range of results based upon flow direction?
- I notice that the PA2721 filter as pictured on the Baldwin e-Catalog site features a visible mesh on the outside of the filter that would have likely prevented this sort of failure. At what point was the design changed to remove this mesh, and why was it removed? Was the structural integrity testing performed again after this change? How did this change impact the results?
- In your initial email on this issue, you said: “There is not an external wrapper on the media pack of the PA2721 to support the media in an inside to outside flow direction.” What is this “external wrapper” that you are describing here? Is it present or not on the PA2721 as shipped?
- The PA2721 as shipped has a protective mesh on the inside of the filter, but not on the outside? Why is this mesh on only one side of the filter?
- What environmental testing has been done on the Baldwin filters in various humidity and temperature conditions? Is there any reason that a Baldwin filter might not be able to perform adequately after a Florida winter? Are these usage guidelines documented anywhere?
I haven’t heard anything yet, but if I ever hear back from Baldwin again, I will update this post.
Avoiding Future Failures
I have been holding off on posting this final failure analysis update for a while now – hoping to reach a more solid conclusion on just what went wrong with our engine.
But in truth it remains a mystery whether the engine failed because of bad air and debris, or whether the filter failed as a result of the engine overheating for some other reason. There is too much else that can go wrong with an old engine to know for sure.
Perhaps the filter had actually failed previously, maybe even months before the breakdown. If so – months of dirty air could have contributed to the engine failing, and with a tear in the filter that would explain why the Filter Monitor was not showing any excess restriction.
I had thought that Baldwin’s analysis would help answer some of these questions, but instead the initial inconsistent answers I received left me with very little trust remaining in Baldwin’s filters and specifications. If even Baldwin’s own Service Engineering team does not know that the PA2721 filter should be able to be used bi-directionally, how can I trust that the design had ever been properly tested?
I hoped to end this story by sharing how Baldwin had uncovered a design defect around “inside to outside” air flow that they were aiming to fix across the board, or perhaps they discovered there had been a manufacturing defect in our specific filter that was rare.
But either way, I wanted to see that Baldwin was learning from our experience and was aiming to make better products in the future.
Every company and every product line has failures – the key is to learn from them. I’m not convinced that Baldwin has learned anything here.
Without further answers about Balwin’s designs and testing process, it seems as if the only conclusion we can reach and bit of actionable advice we can give is to recommend that Baldwin air filters be avoided, particularly in humid environments – and especially when used for “inside to outside” airflow.
And after this experience, we are sure that we will never be letting a Baldwin filter of any sort near our fresh and fully rebuilt engine ever again.