Let the numbers talk!
Whats the hidden story behind oil drain intervals?
By Roberto Rangel
In the lubricants arena, few topics generate as much interest and debate as engine oil drain intervals. Just how long intervals can be extended, and the resulting impact on lubricant demand, is a strong preoccupation for base oil, additive and engine oil sellers, and for the wholesalers, retailers and installers who deliver lubricants into the hands of consumers.
Often, the discussion begins with a look at average drain intervals. For example, in the article Life in the Fast Oil Change Lane in the May 2014 LubesnGreases, Steve Swedberg described how the fast-lube sales channel has performed in the United States over the years. Pointing to survey data from the fast-lube trade publication National Oil & Lube News, the article noted that typical oil change intervals have been extended from their one-time 3,000 miles to their current length of 4,601 miles, and to 5,073 miles for vehicles equipped with an oil life monitor.
As that article pointed out, these figures are average data. It would be interesting to know the range of results, because obviously some users are following shorter drain intervals and others follow longer ones. For the latter, synthetic and semi-synthetic lubricants may have their greatest impact.
How many cars actually follow the average? As the above implies, there are various distribution criteria involved. A normal or Gaussian distribution would follow the well-known bell curve: half of drain intervals will be longer than 4,601 miles, and half shorter. Theres also the Poisson distribution, which considers the period of time in which a given event will likely occur. Put another way: How many times a year will a car have its oil changed?
To reliably answer these questions, an analysis from a systemic point of view can offer an interesting context. Systemic analysis allows us to find a solution according to the size and properties of a system, with the goal of making improvements and discerning the interaction with other systems. This is the methodology I have been applying for some years to analyze different problems of industry in Mexico, including the challenges of the Mexican lubricant market.
In this case, let us call the U.S. engine oil market the system. Our focus will be on drain intervals (the relevant variable) and well pursue consistent and reliable results which reflect the key point: extended drain oil intervals.
As with any system analysis, the challenges are threefold:
1. What are the relevant variables which have an influence on the system?
2. How are they related?
3. Where are the reliable data sources?
Before explaining the analysis, let me share with you some thoughts I consider every time I face any challenge:
As an individual or a group, we are involved in different activities and focused to achieve goals in an efficient way. During this process, the information, databases and references are gathered from different sources, and if these are sustained in a consistent and reliable manner, they are highly valuable because of their contribution to the final results, and for reducing the uncertainty range of values for the system variables. As well, difficulties like unavailability of historical data need to be handled appropriately. Therefore, if you were to distill into one word the main problem to be handled, that word would be reliability, especially of the information sources.
Coming back to the three challenges mentioned, it is not difficult to identify the first group of variables. The primary ones would be a) registered automobiles; b) total fuel consumption (gasoline); and c) average miles traveled per automobile. The reliable source for this information is the U.S. Federal Highway Administration. We also need to know the total product supplied of lubricants (sourced from the U.S. Energy Information Administration); and a reliable estimate of passenger car motor oil consumption, such as that from the industry consultancy Strategic Resources Inc.
With the above data, we can move to the next group of variables, which I call secondary ones because they are derived through calculations involving the primary ones. For example, for each automobile we can calculate the yearly average fuel consumption, average miles per fuel-gallon consumed, average number of oil changes, and average miles per drain.
This dynamic system is consistent because miles-traveled simultaneously generates the consumption of fuel and engine oil (plus other items such as spare parts and ancillary products). We now can find out from different paths the important indicators for each one. This is what in mathematical terms would be called a simultaneous solution for an equation with several variables -the model. In this model, I use formal tools like graphs and statistical analysis.
If we take the fuel consumption path and make the appropriate calculations, associating remarkable historical facts since the 1970s (shown as red arrows on Graph 1), we can see the relevant variables involved and the key vehicle trends.
What this could tell us about the system depends on the observers perspective; nevertheless I will point out two remarks. First, external and internal factors do impact the system from a vehicle mobility point of view, mainly because fuel price volatility is usually associated with economic and global political crisis and obviously affects lube consumption.
Second, the average miles-per-gallon-of-fuel-consumed has been improving as a result of actions taken at several levels, including sales of vehicles that meet the fuel consumption limits and the rising market share of four-cylinder cars. Fuel quality improvements would be an additional one.
These figures for average miles traveled per automobile are a consistent and reliable reference for analyzing the lubricant oil consumption path as well.
Using figures from Strategic Resources Inc. for annual passenger car motor oil consumption from 2003 to 2008 (published in LubesnGreases in January 2009), we can see that PCMO represents around 30 percent to 32 percent of total lubricants supplied, as reported by the Energy Information Administration over the same period.
Finally, in order to lay bases to evaluate the distribution of the oil drain interval variable, I plotted two sequential trends. Graph 2 shows the trend for average number of oil changes per vehicle on one axis, and also shows the trend for the average miles between of oil drains.
As you can see, drain intervals have been climbing in the last 12 years, reaching around 5,100 miles per drain in 2012 – which is pretty close to the figure reached by National Oil & Lube News (5,073 miles per drain). This climbing tendency is a reflection of what has been happening in the U.S. market, the increasing use of synthetic and semi-synthetic engine oils, and the improvements to conventional engine oils as well as the wider adoption of oil life monitors.
As this shows, we have achieved similar results to the ones reported in the May LubesnGreases, by means of different procedures – which proves consistency – and as a consequence, reliability shows up.
The magic of this research was that it didnt stop here – it was imperative to mathematically support that there was a distribution of users around the average miles per drain number.
The challenge was to find out the adequate range for the variables number of oil changes per vehicle and miles per drain. The results for 2012 are shown on page 28, in Graphs 3 and 4.
As these suggest, a higher number of oil changes implies higher mobility. In Graph 3 the number of oil changes per vehicle ranges from 1 to 24. The right side area from 2.16 is where user vehicles drove more than 11,000 miles in 2012, according to FHWA. The area on the left side represents the opposite; those vehicles drove less than 11,000 miles.
Similarly with Graph 4, we start with 5,100 miles as the average oil drain interval in a range that runs from 2,500 to 22,000 miles. The right-side area (greater than 5,100 miles) represents the vehicles in 2012 that were sensitive to the usage of high quality oils (synthetic or mineral oil based). On the left are those that for some reason were not.
That brings us to the final stage of the process: the sensitivity analysis. Sensitivity analysis focuses on making comments on the results from the system in the context of considered assumptions: What if …?
For example, we might ask what if users top up their oil every 3,000 miles? For vehicles having high miles between drains, the top-up effect is relevant depending on the engine condition, which determines the oil consumption. Under high top-off circumstances the actual oil drains per year could differ from what was projected.
As well, by doing the same analysis shown in Graphs 3 and 4 for previous years, you will find that if plotted and compared one after the other (as a fingerprint), that the peaks of the wave seem to be moving in the grid. In the case of the oil changes/year number theyll tend towards the origin, and in the opposite direction for miles per drain interval. This situation is acceptable and understandable when the dynamic indicates improvements for the system. For instance, the reduction in the oil-change number is a benefit from environmental and end-user perspectives. This decrease could be the outcome of factors like the promotion of better lubricants technology – or from the reduction of vehicle miles driven as a consequence of economic depression.
To be honest, I would rather be involved in the first scenario. Generally speaking, from a marketing and sales perspective it is a challenge in any economic activity to establish strategies to balance the benefits/cost/volume/revenues issues. The scheme of promoting the use of higher technology implies that there is a market niche where final users have the economic capacity to buy expensive products. By contrast, in an economic depression there would be fewer sales possibilities – but of course, always there are exceptions.
Because the goal of this article is to highlight what lies backstage behind todays extended oil drain intervals, Ill offer some conclusions.
First, in order to structure and sustain a solution, we can apply a proven process, involving the identification of the main characteristics of the system; the probable interaction with other systems; the relevant variables; and the most reliable information sources.
Second, we found that the variable average drain oil interval can be shown as a statistic distribution in terms of time, number of users and adequate range of drain intervals.
Next, the sensitivity analysis stage can give us a better understanding of the results.
Finally, this structure of analysis allows us to project the system performance in the short, medium or long term, depending on the assumptions considered.
Above all I would say, let the numbers talk!
Consultant Roberto Rangel has over 26 years of experience in the automotive and lubricant additives industries, including with Spicer Group, Lubrizol, Castrol and Bardahl. He currently advises ANELA, a lubricant-additive association in Mexico which represents 14 large companies, and also provides training on entrepreneurship at two universities. For information or a table of the mathematical structure of the model described here, e-mail him at robertorangelg53@hotmail.com or phone + 52-55-55608629.