[bwilson4web]

I'm running a series of tests of an oil additive to try and measure the effect of engine operation. This is based upon measuring the fuel consumption of the engine at idle during warm-up and idle during cool-down. So far, I've only been looking at warm-up fuel burn.

The protocol is to start the car and shift into "N" a few seconds after the engine is running on its own. I'm using our 1.5L, NHW11 which shares the same engine as the NHW20 and AutoEnginuity to record the data:

The key data are: coolant temperature, engine rpm, ignition timing and mass air flow (MAF). There are four useful records:

#2 - 0W-20, Mobil 1, 8k service miles, 2 quarts were drained including filter
#3 - 5W-30, Mobil 1, 0 service miles, 2.5 quarts were added
#4 - 5W-30, Mobil 1, ~150 service miles
#5 - 5W-30, Mobil 1, ~150 service miles, added 8 oz additive and 8 oz of 0W-20 to rinse container

One unexpected result is two tests starting under 12C changed the ignition timing from 10 to 5 degrees about 660 seconds into the test. In contrast, two tests were the starting temperatures above 20C kept the ignition at 10 degrees. This suggests at colder temperatures the spark is retarded which would increase the exhaust gas temperature because less work is extracted from expansion stroke. But we are interested in the oil additive.

The next charts compare the new oil versus the same oil with the oil additive:

There was an offset in the spark reduction to 10 degrees offset due to the 11C difference in starting temperature. Furthermore, as the engine ran, the lower heat loss due to warmer ambient temperature means then engine warmed up a little faster.

Increasing the Y-scale, show nearly identical MAF profiles except at the end where the virgin oil suggests a slightly lower fuel consumption compared to the same oil treated:

However, the oil additive instructions report that the additive takes a while to coat the metal surfaces and it suspends gunk and varnish:

"If used when you change your oil, simply add with any premium oil. If used between oil changes, add to engine at least 1,000 miles before your next oil change. This will ensure sufficient time for boundary layer formation and bonding. Add directly to crankcase when the engine is warm, then run engine approximately five minutes."

"This treatment will remove sludge and varnish from your engine. When used in engines with over 100,000 miles, change the oil and replace the oil filter after 1,000 miles to eliminate these contaminants from your engine."

To investigate this further, the time line was offset so the first ICE rpm drop starts at the same time reference:

The plateau of the ICE coolant occurs when the thermostat opens up allowing the rest of the coolant to flow through the engine block. Once all coolant is at the same temperature, warm-up continues.

Changing the Y-axis scale, we see the ignition advance change occurs concurrent with the engine rpm reduction:


Again, increasing the Y-scale, the MAF lines are all but on top of each other:

There is no evidence that treating the oil had an effect during a cold-start, warm-up. This raises a question of whether or not we could detect any difference in oil friction effects.

To test the test, we compared the 0W-20, 8k service miles to the virgin, 5W-30 oil. Changing the time scale to align inflection points, we find a consistent gap between the used and new oil during warm-up:

This data set provided ~2,400 samples showing the lower fuel consuming 0W-20 with 8k service miles versus the stock, 5W-30 with 0 service miles.

Mapping the relative ratios we find the 0W-20 has about a 2% reduction in fuel consumption versus new 5W-30:

Good Prius friend, David Kelly, found a reference that 0W-20 has about 2% lower internal friction than 5W-30.

The oil additive instructions require 1,000 service miles before a second oil change to remove the varnish and gunk it releases from older engines. At that point, it should have coated the moving parts with a low friction layer. But 1,000 miles exceeds the tank capacity of our NHW11 so I'll have to run gas reference tests before and after each tank. Otherwise, there is a risk of measuring a change in winter-summer gas energy content and not the oil additive effect. Also, I didn't add the oil to a 'hot' engine so I may use the second oil change to add the remaining 8 oz.

Bob Wilson

[GeorgiaHybrid]

Bob,

A couple of questions if you don't care...

Is this additive used to reduce internal friction or just as an engine flush to remove varnish? If it reduces internal friction, are you trying to see if you can get a 5W-30 oil to perform as well (friction wise) as a 0W-20 but offer better protection at higher temps?

If a person changes oil on time and uses a good quality synthetic, how much varnish can build up in an engine? I say that as all of mine get changed every 5,000 miles and I use Mobil 1 5W-20 in all of them (yeah, I know I can go to a 10,000 miles OCI running synthetic but I'm too much old school). I also use a scope to look into the cam covers to see how things are going. So far, even the engines with well over 100,000 miles on the clock, they just have a coat of oil on the surface with no deposits.

[bwilson4web]

GeorgiaHybrid said:

Is this additive used to reduce internal friction or just as an engine flush to remove varnish?

I'm testing boron CLS (aka., borax) which historically was also used as a soap enhancement (aka., 20-mule team Borax.) The varish and gunk suspension is probably just a by-product. The real goal is to see if the claimed, boric acid adheasion to metal surfaces actually reduces friction by a measurable amount even after the treated oil is changed. The vendor and Argonne Labs patent suggests it is good for 100,000 miles ... an incredible number.

GeorgiaHybrid said:

If it reduces internal friction, are you trying to see if you can get a 5W-30 oil to perform as well (friction wise) as a 0W-20 but offer better protection at higher temps?

That had occurred to me but is beyond the scope of my current test because I have no baseline on 'protection' to use. If I had been testing each oil change, over time, I might be able to use the rates of material per service mile to see if there is a reduction but I didn't do that with the NHW11. Now I have the option of starting that with my wife's car but ...

I am giving serious thought to dropping the engine oil pan and giving it a through cleaning. However, there is no gasket and I would have to scrape off the old and apply new sealant. Except for oil in the gallies and head, this would be a big improvement and perhaps allow me to start a long-term study. But there aren't any serious oil change and analysis databases around for folks who share my interest to keep such data (<ahem> Jason.)

GeorgiaHybrid said:

. . . If a person changes oil on time and uses a good quality synthetic, how much varnish can build up in an engine?

This is a hard, multivariable problem beyond my resources. There may be some fleet studies but I haven't searched the lliterature. For example, varnish is apparently a by-product of hot oil and water, either vapor or liquid. A significant portion will come from piston ring blow-by and that is a function of driving profile ... a lot of cold-starts having a worse effect than few cold-starts for the same service miles. It really is a hard problem and nott one I look forward to.

GeorgiaHybrid said:

I say that as all of mine get changed every 5,000 miles and I use Mobil 1 5W-20 in all of them (yeah, I know I can go to a 10,000 miles OCI running synthetic but I'm too much old school). I also use a scope to look into the cam covers to see how things are going. So far, even the engines with well over 100,000 miles on the clock, they just have a coat of oil on the surface with no deposits.

Sounds like a clean machine!

Bob Wilson

[bwilson4web]

The additive requires 1,000 miles before a second oil and filter change to remove the trapped and suspended varnish and gunk. The challenge is I had not stored two tanks, 20 gallons of gasoline, to keep the fuel constant.

Given the variation in ethanol and refinery output, gasoline does not have a constant, heat energy. Worse, this is the time refineries switch from a winter to summer blend and the gas station tanks are just a mix during the transition. So my next best alternative is to measure the change in fuel energy with each tank and use the accumulated offsets to adjust between the pre-treatment and post-treatment engine fuel consumption.

Two warm-up cycles were completed including a drive and cool-down. The series #6 is the last of the first tank and series #7 is the first of the new tank after 10-15 miles to replace the fuel in the gas lines.

The first chart shows the data adjusted so the end of "N" occurs on the right edge:


By careful timing, the starting temperatures were identical but the shift from "N" to "D" was not well controlled. However, when the thermostat opens at ~85C, there is a notch and plateau while the rest of the engine coolant keeps the engine at a constant, operating temperature:


Close examination shows the thermostat opening is an excellent candidate for a constant temperature state to use for fuel rate measurement:


We can see the fuel burn rate plateaus at these point:


So using my old eyes:

2.75 gm/sec - 1st tank, MAF rate
2.71 gm/sec - 2nd tank, MAF rate

(2.75 - 2.71) / 2.75 = 1.46% higher energy of second tank gas

I'll have to do the same calculation between the 2nd and 3d tanks . . . in about 450-500 miles or another three weeks. Then I'll have to wait to the end of the 3d tank to do the last oil change and consolidate all of the data.

One good side effect is now I have a much easier protocol to evaluate relative gasoline energy content than the earlier, hill climb tests used two years ago. Unlike the hill climb tests that run the engine at peak power levels, these tests are not affected by relative octane ratings.

Having developed this technique for gasoline metrics, in the fall, I will repeat this test with every ZVW30 tank. There is a latent, engine knock that happens intermittently for a brief 10-20 seconds. This happens when the first cold weather arrives and the car has been parked for a cold-soak after a brief, not full warm-up, run the day before. But no one has identified a root cause. By conduction fuel tests with each tank, we may be able to finally understand what is going on.

Bob Wilson

[bwilson4web]

Having throughly looked at the data and followed the vendor's recommended instructions, my observations are:

• oil change particle reduction - there is strong evidence that changing the oil substantially reduces engine fuel consumption, ~2.25 gm/sec -> ~2.07 gm/sec MAF.
• boron CLS - appears to accelerate re-suspension of particles from engine gunk and increased drag, ~2.07 gm/sec -> ~2.25 gm/sec.

I can find no evidence of direct friction reduction of boron CLS without taking extra steps to reduce particle suspension. The boron CLS, boric acid, appears to have a substantial detergent effect and increases particle suspension from any gunk already in the engine. These re-suspended particles increase engine friction.

I am planning to drop the engine pan and clean out all accessible places. This should also substantially reduce the ~25% carry forward of old oil. Then retest to see if engine idle friction is reduced. Note that the vendor of boron CLS has no instructions about dropping the engine oil pan and cleaning out any legacy gunk.

Bob Wilson

[bwilson4web]

Here is the data in graphical format:

bwilson4web said:

Having throughly looked at the data and followed the vendor's recommended instructions, my observations are:

• oil change particle reduction - there is strong evidence that changing the oil substantially reduces engine fuel consumption, ~2.07 gm/sec MAF.
• boron CLS - appears to accelerate suspension of particle suspension and increased drag, ~2.25 gm/sec.

I can find no evidence of direct friction reduction of boron CLS (aka., boric acid) without taking extra steps to reduce particle suspension. The boron CLS appears to increase particle suspension and engine friction.

One additional observation is how the existing gunk in the engine can so quickly reduce the effect of an oil change. The 25% carry forward and engine pan material quickly brings the new oil to the same friction level as the old, particle loaded oil.

Bob Wilson