Oil Selection

by Richard Atwell
(c) Copyright 2007-2011

badge

Which oil should you use? It's a tough question to give specific answers to and highly debatable but let's see if we can come to some sensible conclusions to maximize engine longevity.


Advice:

You'll hear arguments against putting Brand X in your engine, or hear that synthetics retain heat and cause leaks or other comments but keep in mind that few people have actually experienced these problems and most are simply passing on what they've heard or read. There's sometimes a little truth in their statements but you have to understand the context they were put it and this is often lost when information is passed.

Some brands perform better than others but there is a logistical problem standing in the way of figuring out which truly are better: how many thousands of miles, how many engines and how many gallons will be required to obtain the necessary accurate data to distinguish them? It's an expensive and time consuming prospect.

Is a more expensive oil better? Perhaps you are paying more because the sales volumes are lower or the company advertises a lot and passes that cost onto the consumer. Perhaps it really is better. What about the short term testing others have done? Do the motorcycle oil tests you read apply to an automobile without the wet clutch? Maybe those engines are a tougher environment and the tests are valid.

When catastrophic failure occurs due to Brand X, quite often that tip is passed along with no mention of the oil-change interval or climate. Was the oil changed after 20k miles? Was it too thick to begin with? And what about the application: oil that is good for the drag racer may be no good for your bus.

A lot of old school knowledge about oil naturally percolates down to the younger generations. Is Kendall oil really any better? Is the recommended oil brand even formulated the same as the oil from 30 years ago to make that statement? See if you get an answer to justify the claim but quite often you won't get any facts other that "it's always worked for me". Conversely, if that person is still the original owner and driving around with the original engine their advice is probably worth listening to.

IMHO, not a lot can be gleaned about the performance of various oils in the VW engine by listening to the experience of other people. For every group of owners that have had a good experience, there is a group who has had a bad one but more often than not people tend to stick with the brand that works for them and/or follow the herd. Luckily, there are only a handful of brands to choose from when you have to decide what you need and what your budget will dictate.

Don't get me wrong: oil formulations vary among brands and perform differently. However, a bigger issue is figuring out which grade to use in your bus. It would seem like this topic has been beaten to death and a consensus built but options remain very divided.


Which grade when it's hot?

Back in the days when Muir was still fixing VWs and oil came in cans, multi-grade oils were still somewhat experimental. Eventually the oil companies managed to formulate them into the flexible product we use today. By the mid-70's, VW had published an engine oil viscosity specification chart that superseded all prior advice on oil selection and they were able to recommend multi-grade oil. Because VW engine technology is stuck in the 70's so are a lot of opinions about oil.

The idea behind multi-grade oil is simple. Take 10w30 as an example: behave like a 10W oil at 32F (in order to protect the engine during startup) and behave like a SAE30 oil at 212F for protection during normal running temperature. This is something that a single-grade oil cannot do.

The chart is a guide to answer the following questions:

VW makes recommendations about grade (weight, viscosity, etc) but there is often a debate whether or not this information out of date. This is what the owner's manual has to say (I've drawn a red line at the top of the 20W50 range where the multi-grades apparently are no longer effective):

The 1978 chart:

SE chart

According to the chart, you shouldn't be driving around in 75F temperatures (24C) or hotter weather unless you are using SAE 40 in the crankcase. VW owner's have taken this to heart and still put SAE30 or SAE40 into their engines fearing they may overheat if they use 20W50 (as the chart suggests). It's a valid concern but what's the reason for the recommendations in the chart?

In the 70's the 20W50 formula came with an SE rating from the American Petroleum Institute (API). When used, it couldn't take the heat and would break down too quickly in the aircooled engine (probably a lot of other engines too). This is what the rating means (source Valvoline):

SE: This category was recommended for certain 1971 vehicles as well as most 1972 vehicles. This classification offered more protection than the SD group of lubricants and was suitable for severe-duty applications. This classification is used in place of SD oils, but it is now obsolete.

Going by the classifications alone, oil formulas didn't drastically improve until 16 years later:

SF: Recommended with 1988 and older passenger vehicles. This oil has superior anti-wear properties and enhanced oxidation stability over SE lubricants.

Oxidation is what happens to the oil at elevated temperatures: the cylinder walls get super hot and the oil there tends to vaporize (the engine burns oil).

So we can understand why people don't trust multi-grade oils but here's some evidence that shows VW's oil viscosity chart from the 70's is out dated...VW continued to update the chart.

The SF-era chart:

SF chart

The SG-era chart:

SG chart

Compare both charts to the original in the owner's manual and you can see as the oil formula improved so did its resistance to high temperature (at least according to VW). To summarize the improvement in 20w50:

Now what if VW had continued making air-cooled engines into the mid-late 90's? The oil chart may have looked like this (from a 95 Audi owner's manual):

The SH-era chart:

H chart

But wait...the beetle was still in production in Mexico until 2004. What oil did it use? Here's an excerpt from the German version of the technical manual:

94 guidelines

Translated it reads, "Only use SAE15w40 SF oil approved by VW of Mexico. Much like the beetle manuals of old, there is no climate chart (these guidelines are for Mexico after all).

See where am I going with this? I believe 15w50 is the most suitable oil for the bus engine given all of this information. At least, it's a starting point for new engines that makes sense.

We're up to API SM now (although I'm sure the gains are small) so I think it's safe to conclude that modern multi-grade oils can be substituted for SAE 40 in 100F weather, specifically 20w50.

Does Round 1 go to the multi-grades? It's hard to say so far, so we need to keep investigating...

Note: VW no longer publishes charts in this format. While they tend to recommend 5w40 for all their new vehicles, we can't really use that info for comparison because they are intended for watercooled engines that run much lower oil temperatures.


Viscosity Ranges:

Let's compare the specifics of SAE40 vs. 20W50 to see if we can determine which one might be a better choice. Look at the technical data sheet (TDS) for Valvoline conventional oils:

SAE 40:
Vis @ 100C (cSt) 14.7
Vis @ 40C (cSt) 136.2

20W50:
Vis @ 100C (cSt) 18.5
Vis @ 40C (cSt) 165.9

The viscosity of each oil is tested at two temperatures (104F and 212F). The cSt is the unit of measure for viscosity and the lower the number the thinner the oil is at that temperature when comparing two oils at the same temperatures.

It might seem contrary to VW's oil chart judging by the weight alone but the numbers tell us that SAE40 is actually thinner at 104F than 20w50 (for Valvoline). After reading the TDS for modern oils and seeing the updated charts, I think most of us can agree that chart in the owner's manual is out of date and that the oils have improved.

I put together this graph to show the relationship between different viscosities (Valvoline again). You can see that at the low temperature test there is a huge spread in viscosity but as the oil warms up they get very close to each other. Keep in mind that at the right end of the graphs there is still significance between grades but the linear scale of the Y-axis makes it hard to see.

oil viscosity

Notice that the lines do not cross? Even though there are only two data points, the viscosity of oil has a linear behavior so the oil that you select is going to be important from that standpoint. It also demonstrates that the various grades exist because no one formula can do it all.

Round 2: in hot weather will 20w50 retains its viscosity better than SAE 40 between oil changes?

The old school argument (based on experience from the 70's) is that the long hydrocarbon chains that make up oil get chopped up by contact with the engine parts (the oil breaks down). A multi-grade oil like 10w40 eventually becomes a 30 at running temperature, then 20, etc. While this is probably true over a long period of time it partly holds true of single-grade oils so it's a race to see which can hold its viscosity longer.

Viscosity breakdown is one of the reasons why we change oil at regular service intervals. To truly find out which is best, select a brand and send 4 samples to an oil lab like Blackstone:

An extended drain interval oil change might produce the clearest results. The total cost for the tests will be about $80-90 and it will tell you what oil your 40 year old VW likes but we're not done just yet.


Which grade when it's cold?

SAE40 seems "middle of the road" so why not use it all the time? Because the temperature range is limited and it's too viscous below 70F (or thereabout). Look at the oil chart from the 1974 owner's manual:

74 oil chart

Before the multi-grades appeared you had to select an oil based on the seasons. If you left SAE40 in the crankcase as winter came, the engine would barely crank over with the starter and you'd get a lot of bearing wear during startup. If you switched to a thinner grade in winter and left it in, the engine would get too hot in summer. In practice most people switched between SAE30 and SAE40.

In the desert heat (or sometimes climates like Floria), SAE50 could go in the crankcase. In mild climates, SAE30 may stay in the crankcase all year long. Maybe the person recommending SAE30 to you lives in such a climate. Do you live there also?

Some of the beetle manuals suggested using SAE30 all the way from freezing to 100F but we know the same Type 1 engine installed in the bus tends to run hot and needs a thicker oil.

Folks using single-grade oils knowing that engine wear is greatest during startup think that risk offsets the danger of running an oil that is too thin 95% of the time. I don't understand this logic because it's been said that starting your engine is the equivalent of driving 500 miles in terms of wear.

Take a look at this table which repeats the kinematic viscosity data in my graph (from Valvoline):

SAE Grade 5w205w3010w3010w4020w50SAE30SAE40SAE50
Viscosity @ 104F 45.961.9869.7495.41165.985.9132.2222.1
Viscosity @ 212F 8.3610.4610.6714.018.510.9214.720.36

The testing data shows how these oils behave at different temperatures. If they weren't distinguishable, it wouldn't make much sense to identify them separately and sell so many grades of oil.

In low temperatures, the multi-grade oils are far thinner than their single-grade cousins and the differences are even more dramatic at ambient temperatures like 50-60F. Not convinced? Buy two bottles and try to pour them until they are empty. The single-grade will take longer to finish pouring.

This ability of multi-grades to flow more easily at low temperatures, permits more oil flow to the bearings during startup especially in very low temperatures which is better for engine longevity. As the weather gets very cold you may need to switch from thicker oils like 20w50 but in general they work all year long as advertised.

Can you buy 5w50 oils to cover all conditions? Yes, but do they last and do they provide adequate protection? There's a general recommendation that the spread between the numbers should not exceed 30 points (e.g. 10w40 or 20w50). This guideline is somewhat relaxed for synthetic oils but a 45 point spread is avoided by most oil manufacturers.

In my opinion, the benefits of using the multi-grades outweigh the "durability" of the single-grades especially if you change your oil often. I haven't done a single vs. multi oil analysis bake-off but I have sent multi-grade oil for sampling a few times and the reports have come back indicating no significant breakdown in viscosity.

Round 3: multi-grades seems to offer the best of cold and hot temperature engine protection.

Viscosity varies from brand to brand and the classification groups allow for some variation. For example, an oil can be classified as 30 if the viscosity in between 9.3 cSt and a maximum of 12.5 cSt. The groupings are side by side and this tells you that a thin 20w50 can have a similar viscosity to a thick 10w40 but in general the grades tend to shoot for the middle. FYI.

A few years ago 10w40 seemed to fall out of favor and disappear from the shelves but it has re-appeared and my guess is that it has been reformulated. I used this oil for many years in Canada's west coast with no problems. 10w40, 15w40 and 20w50 are the most suitable multi-grades for our engines. Use the chart to figure out when to use 10W30.


Engine Wear:

All these oil recommendations from VW are fine except they forget to mention one important detail: they are the recommendations for a NEW engine. As the engine ages, the bearing clearances increase and you need to use a thicker oil to maintain proper oil pressure. What's good for your engine isn't necessarily good for someone else's engine simply because of the difference in mileage and actual wear.

One way to choose a grade of oil for your climate is to check the oil pressure using the Bentley specs as a benchmark. The engine requirements specify an optimal oil pressure for longevity. VW says that you should see 42 psi with SAE30 oil at 175F when the engine rpm is 2500 (Type 4 engine data).

If you have an oil temperature gauge, an oil pressure gauge and a tachometer you are closer to determining if your oil is providing adequate lubrication or to help you select another grade. If the reading drops below 28psi you could benefit from switching to a thicker oil.

If you have none of these tools and you find the oil warning light comes on during highway travel, your oil is too thin and you need to select a thicker grade.

Easy right? What complicates testing is the fact that it's hard to hold the engine to 175F for long and it's not clear where VW intended you to take the reading although we can guess. Also, the viscosity of multi-grade oils vary so much you are limited to performing this test with SAE30 to gauge the state of the bearing clearances, check the viscosity charts above and then repeat the test with your multi-grade oil. Looking at my graphs SAE30 is closely bounded by 10w30 below and 10w40 above around 175F so you could use these oils with reasonably accuracy if they were already in the crankcase.

Since an oil pressure gauge is only $30-40 I recommend you use it to determine if you oil grade selection fits your engine's current needs.


Temperature vs. Pressure:

All this data, gives you an idea of the level of protection you will get but it will not predict actual wear levels. What actually happens in the engine depends partly on the oil temperature and pressure.

If the oil is too hot and thins out too much, parts experience greater friction and wear increases. This means we need to use a viscous enough oil to keep this from happening (one that will not thin too much).

While temperature is important to prevent breakdown of the oil, the pressure is equally important because it governs the rate at which the fluid is flowing through the engine. If the pressure is too high (oil too viscous), the oil flow is reduced and the oil cannot carry away the heat fast enough and we get a similar result.

Another risk of excessive pressure is that the oil filter will go into bypass mode and contaminants will get past it. In extreme cases, you may blow out an oil gallery plug on a high mileage engine.

Finding the balance between these conditions can only be discovered through measurement:

The oil is the life blood of the engine and the better you can determine which oil is best, the longer it will last.

By now you are probably thinking, "Is all this effort worth it?" That's another debate (hee hee ;-).


Oil consumption:

I hardly burn any oil in my aged engine but on long highway trips I'll burn a 1/4 quart over 1000 miles because the heat from the cylinder wall will vaporize the oil more quickly.

The owner's manual states that it's normal to burn up to 2.4 qt. per 1000 miles. I think that number is way too high for even a new engine and my explanation for this is that the oils they tested back in those days had a much lower flash point.


Synthetics:

Synthetics like Mobil 1 advertise protection to higher oil temperatures (e.g. 300F) that conventional oils cannot attain. Conventional oil starts to break down before reaching 250F which is well within the reach of the air-cooled engine's operating temperature on a hot day. This is a good reason to use synthetics although the price per quart is $6-7 on average.

Discounts thru Costco/Sam's club used to let you buy a 6-pack for $22 but since Katrina the difference in price between the warehouse clubs and the auto parts shelves has diminished. Synthetics from other companies like Redline, Amsoil, Royal Purple and others cost even more per quart.

Do Synthetics cause leaks? They have been known to do so on older engines with old seals. Sometimes they cause the seals to shrink and thus a leak occurs. Other times because they have tendency to clean away the residue left by conventional oils, they can dislodge a contaminant that was an aid to sealing. Additives are put in the oil to prevent this from happening at least in the good quality synthetics. Most high mileage oils have a similar seal swelling agent to prevent leaks which are natural as an engine ages.

I did not experience any of these problems when I switched from conventional oil at 60k miles. I did replace the push rod tube seals around the same time but they were already leaking on me. I have not had any leaks in the last 4 years.

If you are building a new engine, break in the engine with conventional oil and then switch to synthetic: the seals won't care.

Some synthetics are sold as blends of synthetic and regular oils. They seem to have little advantage for the extra cost. Save your money or spring for the real thing.

Recently, the gap between the high-end and the "affordable" synthetics has shrunken since Katrina. I think some people are deciding that $7/qt for Mobil 1 isn't a good deal anymore and they are moving to "boutique" oils like Redline or switching back to conventional oil to save some money. You can almost always save money by buying oil by the gallon or 5L jug but they are not always available or available in the grade you need.


ZDDP:

Remember all my talk about the brand you choose not being so important? Times are a changing but it's not 100% clear what's happening and what is the best action to take.

Oil contains many performance enhancing additives:

Quite a list, isn't it? While oil refining improves all the time, the biggest leap since synthetics were developed have come from the chemistries of the additive packages that are incorporated into engine oil. These are the steroids that distinguish modern oils from the older classifications.

One of the most effective anti-wear agents is ZDDP. It has been proven effective and has been present in oil for a long time. It is a compound made from Zinc (Zn) and Phosphorous (P) and while there are several variations in chemical structure they all do the same thing: to prevent wear when the base oil can't. Only a tiny amount around 0.15% is required for the compound to be effective. Increase the level of Zn by too much though and spark plugs can foul.

The amount of ZDDP varies from brand to brand. While some manufacturers substitute other compounds for the missing percentage, others just reduce or increase the level outright. This is one distinction between brands that actually matters and the internet has allowed car owners to send off oil samples for analysis and share the results online in forums like this:

Visit the Bob is the Oil Guy Forums

So what's the problem? The most recent API rating (SM) limits ZDDP levels in an effort to reduce pollution. How does this occur? As oil consumption increases, the amount of organic P present in the oil increases. As the oil vaporizes and is burned in the intake, the 3-way catalytic converter becomes coated and its effectiveness is reduced.

The oil chemists and engineers haven't figured out exactly how low the ZDDP levels should be reduced to, so they have arbitrarily reduced them by 1/3 to around 0.10% in order to curb emissions. This is a problem for older cars like ours that have no catalytic converters to protect and realize benefits from ZDDP in high wear areas like the camshaft lobes. The V-8 hot rod community with their high lift cams first raised the alarm bells on this issue.

So, just like the unleaded scare in the 80's, is the sky falling again or is there a real concern? Should you be changing your oil more often? Will the engineers figure out a way around this or do we need to go out and hunt for oils with the highest levels of ZDDP? Do synthetics with their high film strength alleviate the concern over low ZDDP levels?

What does this mean for the eventual API SN specification and when is it due to appear? Your guess is as good as mine. So what can you do? Find an SL rated oil? They're almost all gone now.

Recently Castrol has introduced a reformulated Syntec 20w50 that they say has higher levels of zinc for classic cars with flat tappet cams even though it is SM rated. Apparently they can do this because 20w50 is not used anymore in engines of new cars and it comes with a disclaimer that, "SYNTEC 20W-50 does not meet the catalyst compatibility requirements of vehicles manufactured since 1993".

Note: some of the more expensive oils don't have API ratings because they have additives levels that are too good for the SM rating! If you were buying a $1 bottle of oil without the rating I'd be worried but not with an expensive brand like Redline etc.


What's the latest trend?

Although automotive oils have largely changed to adopt SM levels of ZDDP, 5w40 and 15w40 "diesel" oils have done so to a lesser degree and most of the motorcycle oils are even better off. The diesel oils tend to cost more than regular oil and the motorcycle varieties cost a lot more.

If you aren't concerned about the cost of oil ($8/qt) and you want higher levels of ZDDP and don't want to measure and mix additives, get Redline 20w50 or 10w40 for motorcycle (black bottle w/ silver label). If you like to use 15W50 (Redline only makes a 15w50 in the blue label bottle (autos) not one for motorcycles), buy two of each and mix up some "15w45" or whatever that combination produces.

Zinc levels are 0.25% and Phosphorus levels at 0.21% which are about double the typical SM rated oil and a lot more than the old SL levels. I have no idea what the long term effects are of using oils like this in your vehicle will be. It maybe as small as a change in gas mileage.

redline Zn content redline 20w50

References:

History:

03/16/07 - Created
09/08/11 - Fixed broken photos, added translate button, updated footer