New pH Test Offers Benefits over TAN/TBN

Total Base Number (TBN) and Total Acid Number (TAN) were originally developed for quality control monitoring of new oils. However, their common use in the monitoring of used lubricants can occasionally lead to wrong conclusions. For instance, it is known that Total Acid Number provides an indication of acid concentration, but not acid strength. As such, it cannot always be relied upon to provide a dependable indication of corrosion potential of an oil.

There is also difficulty in setting condemning limits with TAN and TBN because of characteristic differences between trending behavior of oils. To overcome these drawbacks a new pH Method is proposed that, according to test data, seems to provide an absolute measurement of the depletion of over-base additives and the true corrosive potential of used oil, independent of oil type. It also appears that the procedure is free of interferences.

Original Method Development. Because acid corrosion is caused by hydrogen ions (H+), the measurement of their concentration gives a good indication of how corrosive an oil is becoming. The test requires the oil to be diluted with a suitable solvent and measured with a glass electrode pH meter. New oil has a pH between 7-8 which decreases steadily over time.

At a certain point the pH begins to decrease more rapidly and it's at this point when the oil needs to be changed. Correlation between pH and TAN/TBN has found that the condemning value for 20% oil in solvent is a pH of 3.5. For 4% solutions the limit is 4.5. This is the concentration used for measuring the initial pH (IpH) relating to the first point on a TBN titration curve.

Mettler DL25 laboratory titrator for determining pH Method or doing TAN/TBN titrations.

The proposed new pH Method is not covered by any standard procedure although the information for obtaining initial pH is described by ASTM D664-89. The pH Method differs from IpH in that is uses a more concentrated oil in solvent mixture (20% vs. 4%).

This reduces data fluctuations associated with weighing errors. And, it was found to produce a stronger signal from the oil compared to that from the solvent. Because of the largely non-aqueous solvent used, during the development stage (initial five minutes) a certain drift normally occurs. Generally after this settling period repeatable readings are obtained from one sample to another, within a minute or so.

Evaluation of the pH Method. The pH Method was first brought to the attention of the author in an article by Dr. K. J. Masters, published by the Institution of Diesel and Gas Turbine Engineers (London, UK ), see references. Having been asked to review and comment on the procedure, it was found to have great commercial and technical potential.

In contrast, the process of determining TAN was relatively time consuming, costly, and in many cases irrelevant to actual machine condition. Therefore, it was decided to experiment with the pH method and, in a very short time span, the advantages became apparent.

To validate the procedure, samples received from various large industrial clients were tested for TAN and pH simultaneously. In many cases the TAN showed a slight change as compared to the previous results or published specification on the oils. Normally, these movements would have received only scant observation. However, the pH results generated considerably more attention in many cases.

Numerous samples that had trended normally with the TAN, were suddenly found to be extremely corrosive, with pH readings well below 3.0. Subsequent machine internal inspections revealed high levels of corrosive pitting on metal surfaces. This provided factual confirmation that TAN testing did not provide a satisfactory assessment of actual oil condition.

Australian Modified pH Test Method. Below is the initial procedure used for the pH Method:

1. Into a clean, dry analysis cup weigh approximately 10g of sample.
2. Manually add 50mL of TAN solvent per ASTM D 664-89 (Isopropyl Alcohol, Toluene, & Water)
3. Mix by agitation & place cup under suitable glass pH electrode linked to a Denver Titrator (or equivalent).
4. After about 3 minutes read stabilized result as pH on titration controller screen.
End of test.

Natural curiosity lead to further investigation using a Mettler DL25 Titrator, which incorporated the ASTM D664-89 method within its software suite. It was assumed that as this instrument was more internationally recognized. Correlation of results revealed that the Denver Titrator pH results were confirmed by the Mettler Initial pH (IpH). The results were expressed in mV and converted to pH. The Mettler was then set to readout pH directly and once again the results paralleled those of the Denver Titrator.

Since these validation tests were originally conducted, many samples have been examined using the pH Method from diesel engines, gearboxes, air compressors, refrigeration compressors, steam turbines, etc. Follow-up inspections of the machine and oil confirmed the conditions initially reported by the pH Method. Today, the pH Method is progressively being used as a standard test within the Lubetech Oil Analysis Division laboratory in Brisbane, Australia.

References:

1) K. J. Masters, PhD, “Lubricating Oil Analysis - what is it all about?” Transactions of The Institution of Diesel and Gas Turbine Engineers, Publication 489, December 1995.

2) K. J. Masters, H Leverton Ltd, Windsor, UK.

3) P G Ball, “Machine Wear Analysis - A Rational Approach to Methods Integration for Maximum Benefits,”. STLE Lubrication Engineering, Vol 54, No 3. March 1998. Presented at the World Tribology Congress in London, UK. 1997.

Reader Response to This Article:

There is a topic in oil analysis which has bugged me for some time, and I find it addressed in one of the articles in the September/October edition. The article New pH Test Offers Benefits over TAN/TBN, discusses a new test for determining the Initial pH value of a used oil sample by using a set-up similar to the standard TAN equipment.

I believe three very different properties of an oil have been used interchangeably for some time: corrosiveness of a used oil, used oil "pH", and Total Acid Number. In this article, the author points out significant differences between a pH measurement of the oil, and the Total Acid Number. He does go on to draw positive correlation between the "pH" measurements, and the observed corrosion of equipment. Let me address some of these issues at the Chemistry level.

"pH" is a chemical term that arrived at by applying the formula pH = -log[H3O+]. That is, minus the base 10 log of the concentration of the hydronium ion (expressed in moles per liter) in solution. This is used in inorganic chemistry to characterize the acidity of aqueous solutions of acidic compounds. We are able to measure the pH of an organic carboxylic acid ONLY because of the TAN solution that is used.

This solution takes a hydrophobic organic compound (oil), and mixes it with another organic, toluene, which is an effective solvent. Isopropyl alcohol is miscible with the toluene solvated oil, because it too is organic.

However, because of a special condition called hydrogen bonding, alcohol, as some of us may know, is miscible with water. It is in this manner that the water is introduced to the oil. Without the presence of these specific solvents (toluene and isopropyl alcohol) the water would not be able to solvate the carboxylic acid, which is the oxidized oil.

We can indeed, in this manner, arrive at a "pH" value to assign to a given used oil sample. However, I believe that many in the industry wrongly believe that this pH number is an indication of the corrosive potential of the oil. It is only an indication of the strength of the proton-donating compounds present in the sample when properly prepared with TAN solvent. In-situ, oil obviously does not have these conditions available, and behaves quite differently.

The Total Acid Number is named as such very deliberately. Because it is a measure of the TOTAL Acidic components, both weak and strong acids. If one looks at a graphical representation (pH vs. titrant volume) of a titration performed to ASTM 664, two areas of slope sign change are seen. It is at the second slope sign change that the weak carboxylic acids are titrated.

By performing this test in the proscribed manner, one can arrive at a value that effectively describes how much carboxylic acid is present in a sample, that wasn't there before. This assumes that no source of other acidic contamination has occurred.

The probable reason that Mr. Ball has noticed a better correlation between his "pH" measurement and corrosion, is that the Initial pH value as measured at the start of ASTM 664, will be highly influenced by the strong acid constituent (including NOx and SOx in internal combustion engines) and almost not at all by the weak acids.

That is because strong acids readily dissociate in the presence of water, but weak acids will only weakly dissociate. However, if there is significant oxidation of the oil, the Total Acid Number value will still be high, even with a low starting pH, and this will be a significant indicator of oil oxidation.

In summary, we first must understand exactly what the ASTM 664 test was designed to measure: that is, in the absence of contaminants, the oxidation progression of an oil sample can be monitored by comparing the initial Total Acid Number of new oil to the value measured in the used sample, as an indication of the formation of weak carboxylic acid (i.e. oil oxidation products).

And an absolute measurement of the strong acids in a used oil sample can be accomplished by determining the Initial pH in ASTM 664. However, although initial pH may show fairly good correlation in practice to the corrosion potential of an oil sample, it is not a totally accurate indication of expected corrosion since presence of moisture, chemical makeup of the target surface, and the relative strength and chemical composition of the strong acid constituents will all influence the corrosion rate.

I look forward to your comments and value your feedback.

Regards,

Rich Wurzbach
Peach Bottom Atomic Power Station

References
Ball, P.G., New pH Test Offers Benefits over TAN/TBN, Practicing Oil
Analysis, Vol.1, No. 2, Noria Corp., Sept/Oct. 1998.
Petrucci, R.H. and Wismer, R.K., General Chemistry with Qualitative
Analysis, MacMillan Publishing Co., 1993.
McMurry, J., Organic Chemistry, Brooks/Cole, 1984

Response by Author:

To POA,
Thanks for sending the comments from Mr Wurzbach at Peach Bottom Atomic Power Station. I have had my laboratory staff review it and they spoke very highly of his remarks. They have asked me to place the following response for adding to your Website.

Measurement of pH in a non aqueous solvent has little significance in terms of possible hydrogen ion activity because of the unknown liquid junction potential, which can be rather large depending on the solvent. Measurements made in this way are usually referred to as "Apparent pH".

In our tests we are actually looking at "apparent pH", and not aqueous pH. Yes, there is no doubt that in the realms of chemical accuracy that the figures are not real pH values; yet the correlation of figures to unit condition are too consistent to be considered as 'one off' cases.

Remembering also that our laboratory does oil analysis, and therefore "apparent pH" is only one test of a number that determine machine condition, we don't for one moment suggest that a machine be shut down, oil drained, and internal conditions inspected, on a single report for "apparent pH" or for that matter TAN values. What we have tried to look at is a method that can be used in the field, as well as in the laboratory, and is simple and reproducible in numbers.

We have found the numbers to be very reproducible, and when trended give our laboratory another 'tool' for determining early oil condition problems. Sofar all of our data suggests that our "apparent pH" values do proceed TAN values, and thus allow us to monitor machine conditions more closely once the figures start moving significantly. We do take into account structures of machinery, as we do understand that different metallic elements react differently when exposed to corrosive effects.

We base our opinions of the oil being corrosive on actual visual inspections, as well as "apparent pH" values. It is appreciated that the chemistry of the pH Method is not 'by the book', and thus in the future to reduce confusion, we will refer to this Method as 'apparent pH value" There is simply too much evidence to ignore this as a coincidence.

Jim, we think your magazine is excellent, in both quality and content. Hope that you get your Forum up and running soon. Peter