The Offshore Voyaging Reference Site

An Easy Way to Avoid Engine Failures


There are few things that will ruin a cruise more than an engine failure. But worse still, a sudden engine failure can result in the loss of the boat, or even a life.

There are tons of books and articles on how to properly maintain an engine, but the one tip I never see mentioned is to get the engine oil and marine gear fluid analyzed regularly, something we have been doing for nearly 20 years.

Login to continue reading (scroll down)

Inline Feedbacks
View all comments
Marc Dacey

It’s a good suggestion, just like installing an EGT is a good suggestion. But very few will follow it (well, maybe more readers of you here will). I’ve seen racers kill diesels in five seasons here on Lake Ontario because of their habit of redlining a cold block to the start line in 10 minutes, and then doing it again after the race. The basic needs and proper use of a diesel auxiliary seems to be misunderstood (or ignored). Really, club racers should have some sort of updated, low-compression *gasoline* auxiliary, like the Atomic 4. You can run those on and off all day. Imagine what you’d save in oil analysis!

Drew Frye

Makes sense. I’ve run such programs and supervised the labs. While you’re at it, add periodic coolant testing for glycol cooled engines. Even trace contamination with seawater (20-50 ppm is the condemning limit for chloride) can raise considerable mischief. This is why extended coolant change intervals, common for cars and trucks, don’t work for marine engines.

Testing helps anticipate and prevent serious mechanical failures. However, the most common sort of failure has nothing to do with wear. It is also the sort that happens just as a boat is negotiating a jetty or marina fairway. A filter clogs (you were just bouncing around), a line gets wound into the prop (dropped the chute and for got a sheet, or perhaps a dock line is hanging), or perhaps it simply stalls due to poor warm-up. I once had both engines stall between jetties because I had just changed the fuel filters and forgot to open the fuel valve when finished–I anchored really fast, avoiding harm).

In this case the answer is to allow the engine to warm up at the dock or at anchor before heading out, and to motor the last 1/2 mile when approaching a tricky entrance or jetty. No “jump in and go” or stylish dropping of sails right at the entrance. In this way, stupid, annoying failures happen in a safe place.

(It would be too embarrassing and lengthy to recount all of my dumb failures. Fortunately, none in this catagory were damaging and they have been effectively eliminated.)

Drew Frye

For outboards that tip-up you need to check the down-lock. They are notoriously unreliable until checked (once checked they are safe). Unfortunately, all seems fine in forward, but the moment you shift into reverse they just float up.

Eric Klem

Hi John,

You are certainly more disciplined than me to be doing oil analysis at each change. We have gotten analysis as part of the pre-purchase survey before but you have to trust the owner on how many hours it has been since the last change. I have also owned 3 diesel pickup trucks and have always used analysis to help me figure out what an appropriate oil change interval is for them (interestingly, it was way higher than I expected for all of them so I just change the oil once a year unless I have some very abnormal driving). For true voyagers, I would definitely think that this would be a good idea but I see less of a need for others who have better access to repair facilities. I do recommend doing more than simply looking at the oil level regularly, you can catch some major problems ahead of time by smelling and giving the oil a good look but it is a far cry from proper analysis.

Thanks for the tip on Cat having a service for this. I have always used Blackstone Labs and been happy.


Dick Stevenson

Hi all,
I agree with Drew about the occasional analysis of the coolant.
I test my oil once per year at the end of the season.
I started “sucking” my oil out into a vacuum container a few years ago which made collecting un-contaminated samples more difficult. My solution is to take it from my oil filter which is remote and easy to remove with all the oil still sitting in it: thereby easier to pour out a sample.
Don’t forget the genset. This is an area where an analysis of mine caught excessive wear in my fairly primitive beast and caused me to change oil weights.
I have used Polaris labs for no great reason, but they sound like they do the job.
My best, Dick Stevenson, s/v Alchemy


Hi Dick and others,

I love how I always learn something from this site without having to go through pain for the lesson!

I thought it was a great idea taking the sample from the filter. A couple of follow up questions present themselves:

1. I change my oil filter less frequently than I change my oil (as per the manual). I assume that this is the same for most diesels, but my knowledge is limited only to my Yanmar. Do you change the filters each time you change the oil, or hold off taking the sample until the filter change?

2. Have you ever had a problem with grunge from the filter contaminating the sample?

3. Do you have any tricks for obtaining uncontaminated gearbox samples? I’ve just been using a clear plastic tube attached to a syringe commandeered from the med kit, but it is a tedious and messy process.



Rob Gill

Hi Bryce, Dick, John,
Absolutely second your comment on this site, your questions and credit to Dick for his idea. I was also wondering about using oil from the filter – for clarity, isn’t this were the most contaminants are designed to be trapped and therefore might give an overly pessimistic view of the engine condition?

Dick Stevenson

Hi Bryce, Rob and all,
With respect to getting the oil from the filter, I figured the oil was well stirred up and distributed as the filter oil is quite hot. I suspect if you let the oil sit for long periods you might get skewed readings, but even then, I do not know. The analysis is looking at contaminants: metal, coolant, sea water and such, that are in suspension. I am also taking out a sample, just the first few ounces. I am not draining the filter. That said, it is a good question.
If there was any grunge from the filter, it would be on the bottom and not poured into the sample jar. I turn my filter over into a funnel to drain and, over years, have never noticed any grunge coming out and the oil looks the same as all the other oil. I suspect the “grunge” is captured by the filter element and held onto.
I change my filter every time I change my oil. I also change my oil at about 75-80% of the time the manufacturer recommends. Maybe if I were running a fleet of trucks and buying oil by the ton, I would look at ways to extend the time, but for me, the money, in the scheme of things, is not a relevant criteria (1-3 changes per season). For those with access to oil at US prices, this is even more the case.
As a note, my old engine was a Westerbeke with 100 run hours as the oil change mark. Once, in Central America, I had trouble finding oil and called Westerbeke about using a multi-viscosity rather than the straight 30. After a few minutes of inquiry about my motoring habits, they said to just keep going significantly past the 100 hour mark. Their rational was that I (and most cruisers) was much easier on my engine oil as I motor occasionally/regularly for long periods.
As for gearbox oil, I will mention a “tool” that has multiple uses on a boat. It is old fuel containers. They can be cut with a hot knife (I use a propane torch with an old crab knife rather than a lighter weight rope/canvas hot knife—empty and air the container thoroughly first) into any shape you wish to slip under (or hold) most anything. They are very strong and easy to clean (Think winch servicing). One of mine was made to slip under the tranny and is easy to clean so the tranny oil is uncontaminated. It also makes it easy to look for metal particles as the dump is out of the bottom of the tranny where particulates congregate. This inspection is actually all I do.
My best, Dick Stevenson, s/v Alchemy

Marc Dacey

Great minds, etc. I have used a large foil pan of the sort employed at Christmas under my engines for a similar purpose. It can be easily bent to shape and if you have a leak, it’s easily spotted and located.

Dick Stevenson

Hi Marc,
Agreed. For repeat tasks the old plastic fuel containers are great. For others, I keep a selection of throw away aluminum pans from any supermarket. They are light, easily bent and modified, and come in a variety of sizes. They are particularly helpful for containing such ugly tasks as plugged toilet fittings as, with a bit of patience, they can be duck taped to areas where you want a leak caught/contained saving much spread of yuck and clean up. They can also be molded into troughs and gutters for transfer of liquid from difficult to deal with petcocks and such to more convenient locales.
My best, Dick Stevenson, s/v Alchemy

Eric Klem

Regarding whether taking oil from the filter for analysis is okay or not, I would think you would need to test that to be sure but it would not be my preferred solution (people who work on filtration may be able to definitively say). You would be getting oil from both the unfiltered and filtered side. The actual unfiltered oil on its own is probably not that bad but if it takes some of the contaminants caught in the filter, that could skew your results. Oil flows from the outside of the filter to the inside so if you could carefully only take oil from the inside, you wouldn’t have any worries. Whether this is possible is probably installation dependent.

The first thing that I do when I get a new vehicle is install one of the Fumoto valves to make oil changes easy. These also have the advantage of making sampling easy if you have good access to the bottom of the oil pan. On our boat, we have a hose hooked to the drain plug hole which we can use to suck the oil out but could also be used to get a sample by simply disconnecting it quickly from its mount and holding it lower than the oil level.


richard stanard

would like to see comments for your related article on how to stop killing your engine with kindness but says comments are closed…seems to also block access to the registered comments as can’t see how to do that…i must be missing something? have an excellent day…i am…cant miss here in the islands although the wthr lately a little unsettled…hard to beat messing around with boats regardless of loca

richard s., s/v lakota

richard s (s/v lakota)

here is a better description of what is happening…probably operator (me) error…this is a tablet i am using: in the orig article about testing fluids in the commments section at the end of each commment there is a reply button…if i hit that i am asked to login…when i do that the screen jumps to login confirmation with a fine print note reading comments on this article closed, which is when i gave you my orig note earlier today…if i go into the article re killing with kindness then i have access to those commments…let’s just chalk this up to operator (me) error…i never have felt comfortable around any of these computer systems…i don’t even know what they compute…later, r

Rob Gill

Hi John – what a wonderfully simple idea.
In addition to engine condition, I have been wrestling with how to determine the optimum service interval for our Volvo 55HP to reduce cost, wasted effort and the mess (on our 14.5 metre, 12.5 tonne yacht). The Volvo manual recommends 100 hrs oil, coolant and filter changes. Our mechanic says between 150 – 200 hrs should be fine, but hasn’t suggested how we tell other than Eric’s look, smell and feel test. I am considering sending a sample with a 100 hour service interval. Then the next time extending the interval to 150 hours and sample. If this was OK, I will send a sample with a 200 hour service interval. From this I could hope to determine an optimum interval, yes? Not being mechanically minded, I am confident you guys will tell me if I have lost the plot or don’t understand the problem, and I certainly won’t be offended.
But how do I then interpret the engine oil condition section of the report to determine whether the oil needs changing? I found an excellent CAT document on their USA site titled “understanding your results” but it didn’t show the range of acceptable readings. Without this the raw numbers are unhelpful to me, since I got no further than “Ordinary” level chemistry.
Do you think oil (and coolant) condition in relation to service interval is something I could specifically request comment on by the CAT lab staff as part of their service (to suit this secondary purpose)?
Anyway, first kits ordered from CAT NZ thanks.

Eric Klem

Hi Rob,

I have never used the Cat service but when I have gotten oil analysis before and provided a mileage that the oil went, Blackstone told me how much further I could go on that oil. After a few consistent data points, I became comfortable setting a change interval.


Rob Gill

Thanks everyone for your comments on engine oil and coolant change cycles. I will continue with my plan and report back if I remember. I suspect with coolant I will do a complete change at one time Dick, but look to extend the duration as with the engine oil based on the condition of the fluid after 100/150/200 hours of running.

Rob Gill

Hi John et al,
In the comments above I committed to commenting back on my oil change interval results, for anyone interested.
According to the operator manual, our Volvo D255 actually has an oil change interval of 500hrs, the same as Morgan’s Cloud (actually ours is a Perkins built engine according to the little brass plate on the back) or 12 months, which ever comes sooner, not 100 hours as I stated above. So I actually ended up changing the oil after 470 hours and our 6 month cruise of the Pacific Islands. Of these hours, about 150 – 180 hours were spent at anchor, charging batteries and heating water – not ideal engine usage. I duly sent the oil sample off to the CAT lab in NZ.
The results came back with green (no action) across all categories. Interestingly given our engine “abuse,” and compared with John’s MC report above (with less than half our engine hour interval), our soot contamination level was 8 ppm vs MC’s 13 ppm. Almost every other contaminant or engine wear measure was identical to MC’s report. The V40 number was still excellent, and well in spec. I rang and spoke to the CAT lab manager and he told me our oil was in excellent condition and could have performed another 200 hours plus, if we kept interval testing!
This gave me some faith that our engine abuse and charging regime is not detrimental to the engine, providing we stick to the practice of running the engine hard after any extended anchor time – a tip we learned right here on AAC. Thank you everybody!

Dick Stevenson

Hi Rob,
You mention the timing interval of coolant changes among others. With respect to coolant, my understanding is that there are many who want yearly changes of the “blue” stuff and every 2 years for the “orange” long life stuff. For most uses of cruising engines, this seems like an overly conservative interval: this is fairly robust stuff and its quality and properties do not break down quickly. My “interval” is to “suck” out coolant from the expansion tank, the reservoir and the manifold, all easy to get at, and then fill them up with new stuff yearly. That gets a good percentage of the coolant. The first motor will get them circulated and mixed up ensuring good protection without the sometimes messy draining of the whole system and the potential for air locks interfering with distribution.
My best, Dick Stevenson, s/v Alchemy

Dick Stevenson

Hi John,
For sure, one flushes with fresh water when changing coolants of different types (regular “blue” to long life “orange for ex), but I have not known a flush to be recommended or necessary when swapping coolants of the same type. Also, for sure, flushing with fresh water and a complete coolant swap would be certainly be the best practice, but the engineers I have talked with (only a couple) have been ok with what I do. But I will re-think this practice and do some checking around and see what I find.
My best, Dick

Drew Frye

Actually, the most critical factor for marine coolant systems is chloride (seawater) contamination, as I mentioned above.This has NOTHING to do whether the coolant is long-life. For example, Yanmar specifies annual changes on marine engines, but specifies 5 years on EXACTLY the same engine, with the same coolant, when used in a land-based generator set. You can look this up and verify. The potential for seawater seepage through the heat exchanger is the problem. The additives are not being used up.

John is right. A fresh water flush is required. No flushing chemicals, though. And if you add water, use DI water, just like for a battery (this is now required by the ASTM and OEM specs).

Marc Dacey

I think it’s clear that most of the people posting here are more fastidious and attentive to change cycles of lubricants and coolants than is probably strictly necessary. I changed the oil on my Atomic 4 every “25 hours whether it was necessary or not”, because the open sump and the angle of said sump was just one jug’s worth of oil and I rarely put 25 hours on in a season. So it was just one more task at haulout or prior to launch, but it was cheap insurance in the long run as is more frequent changes on diesels. One thing I would point out is that oil, filter and coolant changes get one up close to the engine and it’s easier to see rusty clamps, loose connections or potential leaks. So if nothing else, a cursory eyeball of the motor can’t hurt and could help avoid problems down the road. Speaking of which, I realized recently that the ubiquitous “selfie stick” can be used in the engine bay to get a phone’s camera right under the block, sump and front and rear seals to check for leaks or corrosion. So the damn things have a function for the non-narcissistic skipper after all.

Maurice Hollingsworth

First, thank you John, Phyllis and AAC participants for such a valued site. As one who lives amply inland and sails – lake, coastal, and a little offshore – whenever possible, I find AAC a really valuable resource.

I appreciate the recommendation to have a regular oil analysis, something I have done previously but only at survey time. This will improve in the future.

I welcome thoughts on oil change routines relative to the season. For example, the manual specifies oil changes every so many hours. However, I usually finish the sailing season with less hours. I don’t like to have my boat sitting with used oil for extended periods between sailing seasons. My solution is to change oil and filter at the end of a sailing season (which makes great sense to me). I also then change the oil (and filter?) again at the start of the next sailing season. This second oil change recognizes that the oil has sat for many months through varied heat and humidity conditions (e.g. Florida). This approach likely truly defines “fastidious”, but does this second oil change have value?

I have some experience with engines, but I am no mechanic or engineer so I welcome thoughts on this approach. Am I simply into overkill mode, or does this approach have value?

Steven D'Antonio

The oil analysis drum is one I’ve pounded for years, I’m a believer, and I’ve practiced what I’ve preached. When I ran a boat yard I implemented a regular oil analysis program, and trained my staff how to properly draw samples. I’ve written and lectured about this subject for both boat owners and the marine industry on dozens of occasions. I’ve completed tribology training administered by Polaris, one of the premier independent labs (more on the value of independent labs, and training, in a moment). Now, if only it were all that simple.

There is a huge level of skepticism in the marine industry regarding fluid analysis, crankcase oil, transmission and hydraulic fluid and coolant, and in many respects that skepticism is well-placed, because the vast majority of marine professionals have no clue how to properly draw samples or complete a sample form or, most important of all, interpret an analysis report, they simply aren’t trained. Invariably, they scan the report for red or yellow bars, and if none are found they pronounce all is well. This isn’t necessarily malicious or willful, however, it does demonstrate incompetence, which can lead to costly and even life threatening errors.

I’ve taken hundreds of samples in my career, and supervised the taking of hundreds more. In many cases those taking samples commit error after error, they re-use sample tubes or drag them along the bottom of the oil pan, they cross contaminate samples by allowing oil to reach the vacuum pump, and then fail to clean it before taking the next sample, they use non-standard sample jars or, among the most egregious of violations, they draw samples from the oil change pump during an oil change. additionally, most of the forms that accompany samples, which I review, are completed incorrectly. One of the most common errors is transposing ‘unit time’ and ‘lube time’, which can yield dramatically inaccurate results, as seemingly very high lube time will be yield very high contamination levels, which the lab’s computer algorithms will identify as “normal” or predictable, when in fact a problem may exist. I’ve encountered this scenario on several occasions and was able to change the report’s results from GREEN to RED by simply having the lab correct these numbers. Some lab analysts will catch this error, but most sail through the system unnoticed, in spite of the fact that it’s impossible to have more time on the lube than the unit.

Therefore, where pre-purchase surveys or inspections are concerned, I don’t blame brokers for being skeptical when a report comes back for an engine that otherwise runs well, predicting imminent failure, even though that’s entirely possible.

My experience with the dozen plus labs I’ve used has also left me somewhat wary. While I’m careful to avoid painting this characterization with too broad a brush, and it is anecdotal, I’m cautious about sending samples to an engine manufacturer/dealer lab for in-warranty engines, especially those with a problem. I’ve seen examples of the labs erring on the side of, “change the oil, re-sample in 100 hrs” until the warranty runs out, or the thresholds for wear metals, and fuel dilution are higher than those of independent labs. Therefore, my recommendation for sampling on in-warranty engines with potential problems is to use the engine manufacturer’s/dealer’s lab, as well as an independent lab, and for the latter be certain to retain the chain of custody of the samples. This recommendation holds true for all pre-purchase analyses as well, the sample should either be taken and retained by you, or someone solely in your (and only your) employ, your mechanic or surveyor, and be certain those taking samples, including you, are well versed in proper sampling technique. Make certain the forms are properly completed, and then find a resource who has been trained to interpret the results.

Finally, if you are sampling regularly for your own monitoring purposes, consider adding a sampling valve, thereby eliminating the need for a vacuum pump or sample tubes. This approach yields the least likelihood of sampling contamination. Also, be certain you request a measure of TBN or total base number, in the analysis, this what counteracts acidity in oil, which, barring all other contamination issues, is the one thing that causes oil to “wear out” and require replacement. Those using by-pass filtration, and extended drain intervals, use TBN as the trigger for replacement. My instinct is we throw away a tremendous amount of perfectly good oil. For small oil sumps and engiens that aren’t run much this may not be much of an issue, but for some of my full time cruising power vessel clients, who put 1000 or more hours a year on an engine, unnecessary oil changes can add up. If the engine is in good tune and isn’t generating high levels of metal or ingesting contaminants, TBN becomes the primary factor in the need for oil replacement.

Other than that, this is all very straightforward;-)

With John’s permission I’ll post a link to an instructional oil analysis video I created, as well as an article I wrote on the subject.

Steven D'Antonio

Thanks John, they are…

Oil analysis instructional video:

Oil analysis article (a little heavy on the tech side, written for the marine industry):

Boat owner version:

Dick Stevenson

Hi Steve,
The video was excellent. Much appreciated. And, like John, I intend to spend some time absorbing the articles.
If the info is ready-at-hand, could you say where to get the pump and suggest the size tubing to go with it?
Happy New Year, Dick Stevenson, s/v Alchemy

Marc Dacey

And this is why I pay for this site. Oil analysis is a pretty complex topic even in terms of grasping the (many) parameters, so shedding light on how it works is of great value to those of us planing on a putting a few thousand hours on our motors in the next few years.

Steven D'Antonio


Every lab sells the vacuum pumps and tubing, they are generally inexpensive. This is the pump I have, it’s very rugged Tubing comes in a variety of sizes, however, most pumps use an O ring and compression nut, so they will accept most if not all tube sizes. Make certain the tube you order (they are available in 1rolls, 100′, or pre-cut sections) will fit easily into your smallest dipstick tube. If you are going to sample often, buy the roll, again it’s inexpensive.

Kit Laughlin

Excellent articles and video, Steve: sincere thanks.

John Cobb

Hoping somebody here can help me out with a couple of questions about diesel engine oil.

I’m facing my first oil change on a Yanmar 4JH2E. The Yanmar Operations Manual says to use an oil API classification of CD. Turns out that classification is obsolete. However Shell’s Rotella T5 appears to be a suitable replacement for the old API classification of CD. So I’m considering using that oil. However it’s a synthetic and I’m not sure if the engine has synthetic oil in it now or not. Would there be any problem switching to a synthetic if the engine has always been operated with a straight mineral based oil? It’s got about 2000 hrs on it.

I pretty confident that the previous owner used a 30w oil but the Rotella T5 seems to be only available as multi viscosity. Would switching to a multi viscosity hurt anything?

Third question: The Yanmar ops manual says to use the same oil in the tranny as in the engine. However I’m wondering if that is really acceptable with the newer synthetic and multi viscosity engine oils? I’m thinking I might stay with a 30w oil there. Comments?

John Cobb

Good idea. I need to quit overthinking this I guess. Thanks.