Blackstone TBNs

Status
Not open for further replies.
Joined
Oct 31, 2002
Messages
25
Location
Ft. Wayne, IN
There has been a lot of discussion on the TBN and the different methods of determining it, so we wanted to give you our thoughts on the matter. Blackstone started offering the TBN as a regular test about two years ago. At first, we used Dexsil kits for this test. The Dexsil kits are pretty good and fairly repeatable, though they are expensive and take some time to run. Another drawback is the endpoint. This was determined with a color change, and this method is good only if the same person does the test every time, the same way, and if he/she is not color blind
grin.gif
.

About eight months ago, our sample volumes got to the point that it became obvious that we needed to find a better way to run the TBN test. About the only solution to this problem was an auto-titrator. This is basically a computer-controlled syringe that has the ability to add a solution at a very highly defined rate and volume. The titrator also has a meter attached to it that will read either milli-volts or pH.

Once we decided on the auto-titrator (Metter-Toledo) we then had to decide which test method to use to run the TBN. If you have been reading these postings, you know that there are many different ASTM methods for running this test. There are two main problems with these ASTM methods. One, you have to have a good idea of what the value should be before you run the test, and two, the test itself is very slow to complete --ASTM D-4739 can take several hours for one result. (As a side note, all ASTM tests are available at the library, so you can learn about any test method for free.) The slowness of the test makes it uneconomical for a lab to run this method on every sample. So, we chose a method that was fairly simple (ASTM D-974) and modified it. Originally, this test used a color change as an endpoint. We changed it to have the endpoint be a pH of 4.0 on our meter.

This leads us to our next problem: What should the TBN be? There are no standards for the TBN, and there are many different methods for testing. So what do you use? We chose to use virgin oils as our standard, and we adjusted the calculation of the result to read what the virgin oils should be. Our standard oils are Shell Rotella T 15W/40, Shell 10W/30 and Non-Detergent 30W. These were chosen because they are readily available and give us a good range. The virgin Non-Detergent TBN was about 0.5, the Shell 10W/30 was 6.9, and the Rotella T was 11.5 (both stated on their web site). We run these every day, at the beginning and end of every batch of TBN samples, and we get very repeatable results. We have not run any major tests determining the differences between the different ASTM methods or determined the percentage of error between the tests, though we know that when we run the virgin oils, we repeatedly get the results we expect.

As an independent laboratory, the validity of our data is essential to our business and it is also essential to our customers’ confidence. If our TBN results are actually low, then we need to find the reason why; however, if the results just seem lower then they should be for a certain oil type, this is something that we cannot address. This brings us back to the lack of standards for this test. It has been our experience that the repeatability of the test is as important (if not more so) than the actual result, so we are sticking with our new method.

An interesting note: we are not sure exactly how much of a problem a low TBN is – or whether it is even a problem at all. Both ASTM methods D-4739 and D-974 state that “No general relationship between bearing corrosion and acid or base numbers is known.” Does the TBN give good information about the oil? We think yes, though we also think that a low TBN may not necessarily harm the engine or signal the need to get rid of the oil. I hope this helps explain the TBN test from our point of view and answers some of your questions. Please feel free of contact us, or stop by if you are in the Fort Wayne area.

Sincerely,
Ryan Stark and Kristin Huff
Blackstone Labs
 
I want to personally thank you for your efforts to come on here and explain your testing methods and their contribution.
It shows a commitment to quality in your work when you are concerned enough to hear out your customer and or potential customers concerns, and address them.

Thanks
smile.gif


JP

[ July 29, 2003, 06:14 PM: Message edited by: day1si ]
 
quote:

Originally posted by Kristin Huff:
Our standard oils are Shell Rotella T 15W/40, Shell 10W/30 and Non-Detergent 30W. These were chosen because they are readily available and give us a good range. The virgin Non-Detergent TBN was about 0.5, the Shell 10W/30 was 6.9, and the Rotella T was 11.5 (both stated on their web site).

Kristin, if you were to use the Dexsil kits on these, would you get close to 0.5, 6.9 & 11.5? Or more like 3, 10, 14? Im curious as to the accuracy of the Dexsil kits. Thanks.
 
Hmmm. Several concerns:

A modified method isn't, technically , a method at all, even though results appear to be repeatable. A litigant would never--rationally or wisely--admit lab results not founded on an accepted method. This is not intended to be nit picking, it just doesn't feel right.

What method does Shell use? Is this a true apples-apples comparison?

Is pH a valid metric for motor oil? Used and virgin? Why 4? What's the basis for that?

If the Blackstone process is capable of returning 11.5 results, why are we seeing UOAs with these 1s, 2s, and 3s?

If the former method returned values so different (relatively) from the new way of doing it now, what does TBN even mean anymore? What does the value actually tell us? It's like a switch from one currency to another--no way to know the true worth or value of a unit unless it is related to something concrete. How is that going to happen for us with TBN?
 
Also, I would like to ask how linear are the numbers you get from the new test. Do they follow a linear model (like TBN4 is 2x stronger than TBN2) or do they follow other kind of model? How do we interpret these number compare to the old method or compare to other lab's method?

Thanks for clarifying the method and I really appreaciate your effort.
 
Ryan and Kristen - I appreciate your candor in explaining how Blackstone measures TBN. However, your response raises several questions about the modified procedure. First, why would you use pH to determine the end point of the titration? I'm not familiar with ASTM D974, but pH can only be used with an aqueous (water based) solvent. pH is meaningless when titrating acid or base in a polar organic solvent. Wouldn't the inflection point on the mV scale(i.e. the transition from + to - or visa versa) be more appropriate to determine the equivalence point? And why would pH 4.0 be used as the end point? Is that some sort of empirically determined value related to the actual solvent system used in the test verses water? Also, I don't understand your statement about using virgin oils as standards for TBN. Total base number is an absolute (mg KOH equivalents/gm of oil) rather than a relative determination.

The important factors for any acid/base titration are the correct calibration of the instrument and the use of proper reference standard solutions that are generally obtained from a certified source. The standard ASTM TBN tests (D974 and D4739) involve dilution of a known mass or volume of oil with a polar solvent to allow dissociation of the acidic and basic components, dosing the sample with a known amount of acid that is sufficient to completely neutralize the maximum amount of basic components possibly present in the oil, and then back titrating the acidic solution to equivalence using a standard solution of KOH. By measuring the amount of KOH required to just neutralize the acid and knowing how much acid was originally added to the sample, you calculate how much residual base was present. So it is essential for the concentration of the reference acid and KOH solutions to be precise. The acids (HCL or perchlorate) are fairly stable, but KOH absorbs CO2 from the atmosphere and rapidly goes off. Unless you know how to do it properly, you can't just weigh out dry KOH pellets and make the reference solution yourself. The reference solution needs to be obtained in sealed ampules from a certified source. My concern is that if old or improperly prepared KOH standards are used, you would be titrating with a lower concentration of base than you think you are using. The results would be skewed toward lower TBN numbers, which is exactly what we are seeing.

PandaBear - The relationship between TBN values should be a linear function. That is part of my concern about using pH to determine the end point. pH is an exponential value calculated by the instrument from the potentiometric output. The instrument in really just a volt meter reading the potential difference between a reference electrode and a probe with an ion specific membrane. So if the pH scale is used to determine "neutrality", a small error will amplified exponentially.
 
BITOG Members: Thanks for your responses. Shell uses D-2896 for testing their virgin oils. Another reason we liked using Shell’s values to check our results is because they actually post their “should be” results on their website. A lot of other companies don’t. We ran their oil with the Dexsil kits and got 7.0 for the 10W/30 (which is close to what Shell says it should be). With the Dexsil kits, the 15W/40 read at 13.0, which is higher than Shell says it should be.

pH is mainly used in aquatic solutions, and it can’t be directly measured on the oil itself. It can however, be measured by a special non-aqueous probe in a solution of oil plus the titration solvent. In fact, we were surprised when the Mettler-Toledo guy suggested we measure the pH of the solution. We thought we would be using a milli-volt reading and an inflection point on a curve, like those stated in ASTM D-4739.

It is actually incorrect to say we modified the ASTM method. When we purchased the machine, we voiced our concerns about being able to run TBNs quickly and accurately with their machine. Our rep assured me that several other labs use this method and that they run hundreds of TBNs a day and get good results.

If we could run ASTM D-4739 by the book, we would; however, this method is just not feasible in a production laboratory. We run hundreds of samples a day, and to run any ASTM method by the book generally takes a very long time. Some sacrifices have to be made in order to get the samples run in a timely matter and still keep a reasonable level of accuracy. We are not sure how the other labs run this test, though if they any are running these methods exactly by the book, we would be surprised.

When we said we use Shell products as our standard, this does not mean we calibrate the machine with these oils. That means we check the results with them. The machine is actually calibrated with three buffers (pH 4, 7, and 10), and these are readily available.

The main problem with our results has been with used oils. The consensus has been that our results are low, compared to the old method and other labs’ results. This bothers us and it is something that we are looking into very carefully. As we mentioned before, there are no standards for the TBN that we know of, so it’s hard to say what is correct. Almost all other tests we run have standards available, and these make it possible for us to know if the answers we’re giving you are correct. The lack of standards have been the major hurdle we have to get over, and we are working hard with you and Terry Dyson to get some results that make sense. Please feel free to call us at 260-744-2380 to discuss this in more detail!

Sincerely,

Ryan Stark & Kristin Huff
 
I had a nice conversation with Ryan Stark at Blackstone and I think I now understand some to the problems they are having with the TBN assay. They do a direct titration of the diluted oil sample with an HCL-isopropanol solution until an end point of pH 4 is reached. They are following a specific recommendation from the field guy from Metler-Toledo, the outfit that makes their automatic titrater. Using this method with virgin oil, they get TBN values that are close to the manufacturer's published values. However, with used oil, this method gives lower TBN values than would be obtained by the ASTM D 4739 method. They are presently experimenting with a new oil diluent that consistently gives a higher TBN reading with used oil. I think it is fair to say that Blackstone TBN values are useful for trend analysis, but THEY ARE NOT DIRECTLY COMPARABLE TO RESULTS OBTAINED BY THE ASTM D 4739 METHOD.

I agree with Blackstone that TBN is less important for OTR trucking than for passenger cars. The thing to remember is that acids in oil can only cause corrosion in the presence of moisture. If an engine is run essentially all the time, moisture is never an issue.

Now for passenger car UAO, especially where the purpose is to compare how well different oils hold the ability to neutralize acids (and how much active detergency remains), TBN is an important metric.
 
Drstressor and others,

The new Blackstone TBN technique is very similar, nearly exactly the technique used by many labs(that are respected for their ASTM accuracy) and sold as the ASTM 4739 method , difference is the time alloted to the sample before deriving a result. Most forget to leave off the term "modified".

Reality is that every lab derives a technique that most closely aligns the standard to the need.

Call another lab and pin them down on the exact technique and what modifications they do to the standard.

For those that want ASTM standardized TBN testing and are willing to pay $30 to $50 JUST for that test, you are more than welcome to order them from us !


Or you can use a modified method pay much less and the difference in the result will make NO meaningful difference in the overall analysis result and accurate interpretation of same.

As a matter of fact I have used trended analysis successfully in testing that 1000's of $'s of standardized bench testing failed to reveal. The KEY is interpretation of reasonably accurate data.

Kind of like using a gold plated hammer when a regular iron one will work.


As an analyst I can say with a great degree of satisfaction that my customers using the Blackstone labs tests are getting an accurate and repeatable test protocal that is affordable and enables me to determine the health and condition of both oil and engine or component. If not I would not do business with Blackstone Labs.


When I read this thread I was absolutely amazed in a positive way that Blackstone labs was willing to air their proprietary techniques out on this board with all to take a crack at. Ryan and Kristin have treated me the same way and I am not easy to please when it comes to my customers needs.

Thanks Ryan and Kristin for your honesty.

[ August 02, 2003, 05:02 PM: Message edited by: Terry ]
 
They are presently experimenting with a new oil diluent that consistently gives a higher TBN reading with used oil. I think it is fair to say that Blackstone TBN values are useful for trend analysis, but THEY ARE NOT DIRECTLY COMPARABLE TO RESULTS OBTAINED BY THE ASTM D 4739 METHOD.


Kristin and Ryan: I do hope that you'll make it known to us when you have developed a procedure that will allow closer comparison and correlation of VAO values with UOA values. That you are bothered about the current state of affairs and that you are looking in to it is encouraging. Thank you for your candor.
 
Status
Not open for further replies.
Back
Top