The process of scientific investigation is a much more complex and human endeavor than we usually talk about in technical papers, where we clean up all of the dirt and hide the loose ends.
The true stories are, I believe, much more interesting and instructive fraught with blind alleys and politics, as well as science. In this “story”, I will try to tell it the way I believe it happened.
Fifteen years after the start of my first investigation (1985), I propose to you that in a carbonate depositional system, the detection and quantification of windblown dust, represents an important and overlooked sedimentologic input process independent of and competitive with the carbonate sedimentologic process.
Hence, it can be used, when calibrated, to improve our understanding of the carbonate depositional system. Today we recognize that windblown dust is an extremely areally extensive process and appears to have a unique mineralogic signature – its potassium mineral content – that makes it easy to detect with well logs at amounts far below visible detection in core.
Clastic “Silt”: A Bump in the Road on the Way to Computing Anhydrite/Dolomite Ratios: During the many years in which I have been involved in carbonate reservoir description, I have found that the interpretation of the depositional environment from standard carbonate sedimentologic parameters does not result in a unique outcome.
In fact, these sedimentologic descriptions can even lead to drastically different depositional environment interpretation outcomes.
My investigations into windblown dust were initiated by two motivations: First, I was looking for another source of depositional information that was less subject to the uncertainty of carbonate sedimentologic description, especially in the face of dolomitization destroying depositional fabric.
I was aware that intermittently, the geologists with whom I worked occasionally described some clastic input. These clastics, at least, would not be altered by the dolomitization process.
Second, I was forced by my attempts to solve the mathematical response of the well logs in anhydritic dolomite (The DAK Model for the Permian Basin Clearfork; Clerke et al., 1995), to add another dimension to the problem.
As in particle physics, when the attempts to describe the transformation of the input particles into the output particles using known physics fails, the existence of a new and unknown and unnamed particle is conjectured. Then the search for a new particle begins.
In 1985, I was running well logs to compute Dolomite to Anhydrite ratios, (in analogy to computing sand/shale ratios in clastic sediments) in a Permian Basin anhydrite cemented dolomite reservoir on the Central Basin Platform of Texas. With core data, we knew the core porosity.
Yet I could not get the inversion of the log response to reproduce the known porosity, my porosities were too high? Then, I remembered the descriptions of the intermittent presence of intervals of minor clastic content.
Core descriptions showed that these intervals of minor clastic content were exactly those intervals where the log inversion could not reproduce the known porosity.
Of course, this was all indicative of a mineral whose grain density was lower than either dolomite (2.87) or anhydrite (2.96), as quartz would be (2.65).
I added to the solution matrix, the quartz mineral volume of the formation. Problem, the mathematics and logs available would not support the direct determination of the quartz content in a robust way. An answer could be obtained but it would have very high uncertainties, and I wanted a low uncertainty result.
We had preliminary confirmation that there was a clastic mineral at the root of the problem, and now I expanded the search to include a possible secondary mineral in the clastic assemblage that would give a more unique and low uncertainty result, and found it - potassium minerals -.
Potassium content in clays and feldspars of windblown dust are significant and can be detected with the Potassium Log.
Extensive X ray diffraction (XRD) data were acquired. Of the hundreds of samples on which we performed XRD, we found only ten samples that contained enough clastic minerals to attempt to further try to quantify the clastic assemblage, and only three with more than 20 percent clastic minerals!!!! Were ten plugs out of hundreds enough to characterize the clastic input process occurring over 1500’ of carbonate section?? We would soon see.
Can Lessons from Turbidites Apply to Carbonates?
It's funny how you end up in situations that you were being prepared for in a marvelous and unique way, that is the hand of God for sure. I had been hired into Shell Research after completing my Ph.D. in Physics.
I spent a year in New Orleans working for Shell Offshore in 1983, before reporting to Shell Research. There I had done the first reservoir characterization on the last major deltaic field discovered in 1000' of water - Cognac Mississippi Canyon 194 - and was involved in the very first deepwater turbidite reservoir characterization efforts with the Cougar turbidite discovery.
In turbidite systems, we learned rapidly the value of computing, tracking and mapping the sand-shale ratios as an important tool for reservoir characterization and a seismically detectable attribute.
This concept was the spark that lit the fire for my desire to compute the anhydrite-dolomite ratios in the Permian Basin carbonate reservoirs.
If necessity is the mother of invention, we should have invented more in the Permian Basin then, because we had plenty of motivation. We were infill drilling existing waterfloods and installing CO2 floods in many old reservoirs.
The injected water was fresh, the original water was salt saturated. We would run porosity and resistivity logs and have not a clue as to what they were telling us.
The production engineers were producing fresh water when we thought it should be oil, and producing oil from our “wet zones”. What a mess, I had transferred into! And I wanted no part of generating a petrophysical product that was considered useless.
So I chose to investigate and determine something where I might have some chance of being successful - something that wasn’t a moving target -: the formation mineralogy. The relationship we had discovered for the windblown clastic “silt” - we call it the silt volume function, S(K) - could be modeled using the newly developed (1985) Potassium log, K:
- S(K) = 0.0 if 0.0 < K < 0.005
- S(K) = 25 x (K - 0.005) if 0.005 < K < 0.045
- S(K) = 1.0 if 0.045 < K .
With this relationship, the changing behavior of the rock grain density could be handled, the clastic “silt” volume could be computed along with the anhydrite and the dolomite. The porosity comparisons were now robust and the uncertainty was low.
This sounds like a good place to end the story and it was satisfying in 1986. We were able to map the Silt Volume in a field (Figure 2) and compare it to models of carbonate systems, and with the help of a colleague, Dr. Herb Yuan, we were able to deploy a visual convincer for all of this mathematics by taking a thin veneer along a core slab face, powdering it and doing foot by foot mineralogy, and display this along side the mathematical solution.
Wind Blown Dust Goes Global: Carbonates have occupied only part of my time since this early endeavor but I continued to do projects in reservoirs where clastic “silt” and anhydrite volumes played a major part, and I still track publications that have to do with windblown dust.
I had investigated relative abundance of potassium minerals to quartz within the regional setting of the Permian Basin, but I still kept in my mind the question as to whether this was a region specific solution or something stronger.
A figure in Herve Chamley’s book - Clay Sedimentology -, knocked my socks off: windblown clastic dusts of all ages, geography and sediment provenance contain potassium minerals! It was becoming clear that something about windblown dust was very generic worldwide and provenance independent.
Now this was intersting. Not only that but Prospero at the Rosenstiel Center in Miami was busy demonstrating that windblown dust from the Sahara desert was making it all the way to South Florida!! Now this was very intersting.
Conclusion: Windblown dust is another sedimentation rate of extremely wide areal extent and apparently having a fairly stable mineralogic expression. In a carbonate system, there is no other mineral containing potassium until the very highly evaporitic potash deposits.
It disperses over a very wide area, the aggregation is relatively increased when it is deposited in a low carbonate sedimentation rate portion of the systems tract.
Windblown dust may be most easily detected by using the Spectral Gamma Log Potassium curve. Of course, it can also be detected mineralogically at greater time, expense and manpower. The Spectral Gamma Potassium Log can be obtained in cased or openhole wells with suitable preparation steps taken.
Geologic interpretation of windblown dust volumes can significantly reduce the uncertainty in carbonate reservoir descriptions and is useful in reconstructing interpretations of the carbonate systems tract.
E. A. (Ed) Clerke, Ph.D.
© 2001 Mena Report (www.menareport.com)
