Prius 100,000 mile maintenance

  Part 1: Overview;   tires/brakes
  Part 2: Underhood
  Part 3: Headlights: the Big Schnoz
  Part 4: Inverter pump
==>   Part 5: Coolant testing
  Part 5b: Engine coolant
  Part 6: Transaxle / driveline, references

Coolant testing

Now I'm curious about the state of the coolant that came out of the inverter system. It looks quite clean, with no cloudiness or visible crud floating in it, but there is a definite color difference between it and brand-new fluid. It's more bright orange-red now -- well, *assuming* the same pink SLLC was put in at the factory, that is.

A little while prior to this, small samples have also been drawn from various points around the engine-side cooling system. One came from the drain cock under the storage thermos, one of the lowest points on the system where one would expect suspended crud to accumulate. It looked surprisingly clean, and was stored in the freezer in the same Dunk's cup for a couple of months prior to the inverter-pump job. That was just to play with it and see if it could get cold enough to slush, which it didn't, but of course coolants can get plenty cold enough to cause frostbite and still remain as liquids, which is sort of the point.

On the back of the engine block is another of the three engine-system drain taps, at the low point of the water jackets around the cylinders. Another potential crud accumulation point, given the narrow passages around the block, the high temperatures present in the area, and the mix of metals that the coolant can come in contact with.

But astoundingly, a sample taken from the block looks as fresh and pink as the day it went in at the factory, and is simply dumped back into the reserve tank for now until more in-depth testing can be done.

So how does that old frozen engine-loop sample compare against the inverter coolant and the new stuff? The color looks somewhere in between, I'd say. But after being in the freezer that long, albeit with no obvious change in color or volume there, who knows what further changes it may have undergone. [Or if my food all has a delicate frosting of glycol by now...]

Early searching around turned up a few mentions that coolants with degraded or "spent" anti-corrosion additives would tend to have more ions floating around in them and promote electrolytic effects between dissimilar metals and thus hasten corrosion along. This implied that old coolant might be more electrically conductive than new, and somewhere or other I had heard of it being possible to do a conductivity test with a meter. The old stuff seems to be quite resistive, at almost 10 megohms across the distance between the probes.

But the new coolant, if anything, is *more* conductive than the old. So this probably isn't a valid test at all, at least not here.

There are, however, very relevant electrolytic effects that can eat right through thin metal parts such as radiator cores. Currents can be generated by the coolant fluid itself in contact with the different metals found around the engine block and radiator and heater-core, all acting like a battery; or the car's own electrical system coupled with bad grounds to engines or other parts can wind up pushing some of the return current through the coolant and literally blasting ions off internal metal surfaces. Adding secondary grounds is one mitigation strategy but doesn't always work as expected. The presence of these effects can be tested for in different ways but this isn't really the right place to get deep into it; but do see the references section for some discussion on some of the crazy things that can happen and what people have [often erroneously] tried to do about it.

After this inconclusive goofing around, more research is clearly needed. What the heck is "non-silicate, non-borate, non-amine, hybrid organic acid technology" anyway?? Why is there so much mythology and disinformation floating around about coolants? Well, the fact is that coolant chemistries *and* engine/radiator metallurgy have changed rather radically in the last decade, and the fact that a fill of coolant now has an expected 100,000 mile or more corrosion-resisting service life is somewhat amazing in its own right. It's definitely cut into the classic radiator flush-n-fill service upsell that so many garage guys are used to suggesting to customers, so there are significant forces at work to keep the coolant-change mystique alive as long as possible, but it seems that at this stage of the game it's about as present-day as carburetors and mechanical spark advance.

There are a few fundamental but often subtle facts that come into play, and there's much more to be found under "coolants" in the reference section.

    Water is an excellent thermal transfer medium, but has a limited useful temperature range. Ethylene glycol is about the best anti-freeze and anti-boil agent to help widen the usable liquid-phase temp range, and when mixed with between 40 - 60 % water provides excellent thermal protection and retains the good heat-carrying characteristics of the water. [The less toxic propylene glycol runs a close second in freeze/boil protection, but is more expensive and thus far less commonly used.] However, such a mix alone provides little protection against corrosion, so a package of supplemental coolant additives [SCA] needs to be added as well to be suitable for use in systems with metal piping and passages.

The water used really should be distilled or deionized, as it is much less reactive to metals than tap water which contains random dissolved minerals and ions. This is one reason Toyota's SLLC comes pre-mixed, to be more sure that the right type of water is already in there and doesn't fall prey to just any old thing being added in by a third party.

Metal corrosion inside radiators and engines is greatly accelerated by an acidic environment, which tends to pull even more ions into solution. Thus, coolant formulations tend to run fairly basic, and have a figure known as "reserve alkalinity" which is the capacity for neutralizing acid formation. But if motor oil or blowby gases manage to get into the coolant through some mechanical fault [head gasket failure?], it's pretty much game over.

Silicate compounds are a traditional corrosion-inhibiting protectant additive, such as found in the canonical "green stuff", but have a tendency to drop out of solution and form gels or grits after too long inside a cooling system. This can foul and clog components, wear out water pump impellers and seals prematurely, and cause all kinds of havoc. So silicates are now less in favor as additives than they used to be.

Organic acids such as carboxylates tend to form thin barrier layers on metals and thus have good anti-corrosion properties and generally longer protection life, but tend to bring down the pH of the mix they're in [but not to truly acidic levels].

Hybrid organic acid formulations, called HOAT, retain a certain amount of the silicate [or in some cases, phosphate] based corrosion inhibitors and try to strike a long-life balance between silicate/phosphate and organic acid levels to achieve the longest protection life for typical metal parts and materials found in a cooling system. But that still doesn't tell me if Toyota SLLC has SOME silicates, or NO silicates.

Anything more in-depth than this quick summary, however, is beyond my own chemical savvy and the references speak to it with far more authority than I can. The perfect liquid-coolant would be one with the heat-carrying capacity of water, the liquid-phase stability of ethylene glycol, the lubricity of light oil, and otherwise be completely inert. News of such a witch's brew has *not* yet come to Hahvahd, so everything we have now is a set of workarounds. The bottom line is that although coolant formulation is often complex, these words and terms shouldn't be so mysterious, and the expected lifetimes and benefits of these different types of coolants are becoming somewhat better understood by now but usually at an "only use this stuff!" level.

That's still not going to stop your corner service guy from stroking his chin and going "36,000 miles? You need new coolant" as a way to drum up a little work, and we need a second opinion. Thus, the question remains of how to more accurately test what's going on with a given chemistry in coolants of a given age.

Here's one answer: more test strips. The Acustrip company has been in the fluid-testing biz for a number of years, not just for antifreeze but water quality and some other areas as well. Their strips have multiple reagent pads and measure several characteristics at once, which is useful because the number of different coolant mixtures out there can produce quite different results.

There's a bag of silica "do not eat" in with these, to help their shelf life.

So again, it's supposed to be as simple as dip-n-compare.

In all of the subsequent testing results, the bottle is always left in the picture as lighting and white-balance varies between shots. It's still pretty clear where the tests are landing, and the company claims high reading accuracy stats even by mostly-colorblind people.

Here's the old inverter coolant. The big picture provides enough readable detail on the bottle that hardly any explanation is needed here. The only caveat is that for HOAT coolants, the "reserve alkalinity" figure is a bit thrown off and may be safely ignored in favor of the pH reading.

Glycol percentage and pH look pretty happy, though. Note that a pH slightly toward the acidic side is still listed as okay, which is a little puzzling. But pH is a logarithmic scale, so I guess things down to 6.0 or so aren't *that* destructively acidic yet.

Okay, so what does *new* coolant look like? Not that different, it turns out.


For yucks, the batch of "previously frozen" engine coolant is fished out and tested, to begin getting an idea of what might be happening on the internal-combustion side of things. But given the weirdness this sample has been through, I'm not sure I can trust it.

The percentage pad has just a hint of pink in it, possibly due to a little of the glycol evaporating away; the pH still looks reasonably happy. But let's try to get a fresher sample, and from a different location.

The engine-radiator reserve tank is probably the wrong place to test as it generally contains coolant that hasn't been circulating through the system. We want to get right into the stream of the nasty stuff if possible. In this car the radiator cap opens into a short tube which goes down into the radiator itself; squeezing the hose down below to push the level a little higher lets some spill back into the overflow tank and brings up some of the fluid from the radiator core itself.

This looks a little different. More orangey, and a little cloudy.

Glycol percentage is still acceptable, but the pH is starting to drop just a little in this sample. And that turbidity is a little more like the appearance I was expecting from some of the low points in the system. So while it's still within spec for now, this will need to be changed on the sooner side. That's a more involved procedure and uses much more coolant -- close to 10 quarts, because of the storage bottle, and possibly a new thermostat thrown in too. And some scantool hackery to run pumps and push coolant around the right way to purge air. That will become "page 5b" of this set but for now it will wait. Even a partial change would give the anti-corrosion content a big boost, as noted in some of the references.

Or maybe even a mix including the old inverter coolant, which may be less penny-wise and pound-foolish than it sounds.

Go to Part 5b: Engine coolant

_H* 091005