× Auto Repair Pros Member Benefits TechHelp Knowledge Base Forums Resources My iATN Marketplace Chat Pricing About Us Join Industry Sponsors Video Members Only Repair Shops Auto Pro Careers Auto Pro Reviews
Join Now
International Automotive Technicians Network
Fun with Diesel Exhaust Fluid
Posted to Technical Discussion Forum on 4/23/2014 31 Replies

Hi fellow TDFers.

Recently, the specification values provided for Diesel Exhaust Fluid (DEF) in GM eSI were questioned, specifically as related to this thread by Paul Johnson.


The post and ensuing thread discussed possible reasons why a DEF system pump failure was not covered under the vehicle manufacturer warranty, because the recorded value for the vehicle was far below the specifications for DEF and the system was considered to have been contaminated with water.

Of concern was testing methods and accuracy, what equipment was used, whether the technician knew how to use it and why the cost of repair might have been customer pay on such a low mileage vehicle.

Rather than focus on the possibilities of honesty/dishonesty or competency/incompetency which are quite plausible, the real interest was how the numbers were achieved and with what tools and tests. There are on-vehicle systems tests to run when DEF quality issues are identified, but the concern of what hand-held tools are generally used in a GM dealership arose.

Having graced those service bays for many years, I am quite familiar with most of the tools that have been around for some time, although correlating exact tool numbers can be a little difficult without a full description.

Of particular interest, was the fact that GM specifications in eSI, listed the acceptable range for DEF as 1.310-1.3843 @ 20º C (68º F). Since pure water is identified as being 1.3330, there was definitely some confusion.

Reviewing DEF specs from a number of supplier sources, confirmed that pure DEF specifications have a Refractive Index range of 1.3814-1.3843. I had encountered this some time ago, when the measurement issue was discovered, by another instructor and the numbers didn't jibe, but I didn't pursue it any further than making a mental note. It seems that when using refractive index to display specific gravity there is some difference that is just beyond my need to research further.

So, I sent a memo to my GM contacts that quietl founds its way across various brand quality managers desks to an individual charged with investigating. The results of his research, supported a change in the listed specifications for Duramax and Cruze diesel application documents in eSI this morning. Since eSI documents are live, updates can be effected quite quickly, often within 24 hours.

Fast forward to the end of my work day this afternoon. Armed with the Leica refractometer from my program which was exactly the same model as supplied to the GM dealership where I worked for many years, I stopped in at the GM training centre to gather more information to hopefully better understand where the errors or anomalies in the specifications arose from.

Since I also discovered that there were also anomalies in the ratio of de-ionized water to urea in SI, reference to the actual percentages of de-ionized water to urea has now been deleted.

I had the opportunity to measure this ACDelco supplied DEF [Yara DEF] used in the Duramax course and also some YARA DEF. The latter fell off the back of a truck, quite literally and was picked up off the highway by my co-worker. [YARA DEF] [Yara DEF].

The plan was to gather information by measuring both DEF samples using the Leica refractometer, Reichert refractometer and the Kent-Moore supplied Adblue DEF refractometer to compare readings in an effort to make sense of the numbers.

First things first, exploring these refractometers brings to my attention that the Kent-Moore refractometer [Samples for testing] has an AdBlue Urea scale, with a waterline set at 20 º C and a scale that reads from 15% to 40%. Sorry, it is near impossible to take a photo of the scale. There are plenty of sources for images of the DEF refractometer scales.


Hmmm, none of the testing in eSI lists testing using the percentage scale. This refractometer was supplied for training. I didn't see anywhere in diagnostics any range for urea. The only range reference had been the values in question.

Reading about refractometers versus specific gravity testing of various fluids (relative density), appears to identify that there is some variance in values. Searching the Internet, I discovered that there are specific gravity floats available for just about any imaginable fluid. I'll keep this in mind later during testing. I'm simply attempting to identify if and where the error is.

Testing all of the refractometers to ensure calibration accuracy at 20º C in my home kitchen science experiment this afternoon yielded accurate calibration to the water lines, using the calibration de-ionized water sample "D" in the temperature controlled environment. [Temperature] Outside temperature value was in the kitchen.

Since Paul Johnson had wondered what the differences might be between tap water and de-ionized water, I included a couple of extra samples.

So, over to the kitchen table , set up with the following samples in clean glass egg cups, were Eggzibit "A" from our kitchen water cooler (sorry, I couldn't resist the pun) , Exhibit "B" Municipal tap water, Exhibit "C" YARA Air1 DEF and Exhibit "D", the de-ionized calibration water from the Kent Moore DEF refractometer kit.

Prior testing of both the ACDelco and Yara DEF samples had resulted in identical measurements of exactly 32.5% to specifications. So, for this test exhibit "C", I used the Yara DEF. So, it appeared that when tested with the DEF refractometer, the sample met specifications exactly.

Since the Leica and Reichert Battery and Coolant refractometers tested both the same, the Leica was used for the tests. All of the water samples tested exactly the same, to the calibration line, including the Kent Moore DEF refractometer water line.

Testing DEF using the Leica refractometer yielded 1315 which might possibly support why the GM eSI specification for DEF was originally listed as 1.310 -1.3843. Note: On refractometers used for specific gravity measurements, it is very common to leave the decimal place out for ease of viewing. so, 1315 = 1.315, which is not far from the specification previously listed at 1.310.

So, I now wonder if when DEF was first tested using a refractometer, if one of these battery and coolant refractometers was used, resulting in the lower specified number in eSI, than that of the DEF manufacturers. My result with the Leica was also backed up by the same measurements on the identical model Reichart refractometer.

After all of this, my conclusion is that to be sure of DEF quality using a hand-held tester, to use a DEF specific refractometer that indicates % of urea, not a common battery and coolant refractometer. Depending on how much one is willing to spend, DEF hydrometers can be purchased in the traditional float style hydrometer, analog refractometer as used in this post, or digital.

I would like to test DEF using a dedicated float type DEF hydrometer and if available, a refractometer specifically identified for DEF measurement, because my suspicions are that the battery electrolyte specific gravity range does not provide sufficient accuracy when using a refractometer that is not dedicated to testing the specific fluid.

I suspect that since the SG value achieved on the Leica was so close to the original listed DEF specs in SI, that this tool was possibly originally used to identify that value. Just my opinions, based on the numbers achieved multiple times using clean containers and test instruments. Looking at combination refractometers that can be used for DEF, coolant and battery electrolyte testing, each as its own dedicated scale. Food for thought.

At the end of it all, even given any inaccuracies in the measuring tool used, achieving the 1.295 value by the service technician would support an excessive ratio of water to urea.


Martin from British Columbia

Files Referenced:

thumbnail thumbnail thumbnail thumbnail thumbnail

31 Replies Received (View Replies)