Diamondoids Are a Basin Modeler’s Best Friend

Most of us are familiar with Marilyn Monroe’s rendition of the song “Diamonds are a Girl’s Best Friend” from the 1953 hit movie “Gentlemen Prefer Blondes”. However, if Marilyn had been a petroleum system basin modeler, she might have sung a slightly different version, i.e., “Diamondoids are a Basin Modeler’s Best Friend”. And why might she have sung this? Because she would have known how much more accurate her basin models would have been by performing various types of diamondoid analyses on oils from the basin she was modeling.

What knowledge might Marilyn have gained from her diamondoid studies? Well, she would have known: (1) if deeper source rocks in the basin are contributing to the accumulations, (2) the approximate vitrinite equivalent of those deeper sources, (3) the approximate relative amounts of deep charges to the reservoirs versus shallower charges, and (4) she would have gained a better understanding of migration pathways from the various generational pods into the reservoirs.

In addition, by studying large diamondoid distributions (which Marilyn would have especially liked due to their size and beauty), she would have been able to identify the deep source. And for all these reasons, she probably would have been singing “Diamondoids are a Basin Modeler’s Best Friend” while putting together her very accurate basin models. 

So, let’s look at each of these diamondoid-related advantages to basin modeling in more detail.

Suppose Marilyn had been asked to model one of the basins in Brazil, let’s say Santos. And since Marilyn is part of history, let’s go back to the time when the pre-salt accumulations in the basin had not yet been discovered. Based on published biomarker studies (many by Mello et al), she would have been familiar with the multiple source rocks in the basin, including the lacustrine and its multiple facies and the overlying marine units and their facies. She would also have been aware from the literature and statistical geochemical oil family studies she had purchased from BPS, that most of the oils from the post-salt reservoirs in the basin had been correlated to marine source rocks using biomarkers. 

The obvious thing to do at that point would have been for her to use only the marine source rocks in her model and assume that if there is lacustrine source below the salt, then lacustrine oil is not making it up into the post-salt reservoirs. But nobody likes diamondoids like Marilyn Monroe the basin modeler, and so she obtains diamondoid analyses of post-salt oils from the Santos Basin from BPS/BTI. The first thing she looks at are diamondoid and biomarker concentrations in each of the samples (a measurement known as QDA), and then she plots the data on a diamondoid-biomarker plot like the one shown in Figure 1.

And what does Marilyn find?  She finds that some of the post-salt oils in the study have elevated diamondoid concentrations along with abundant marine biomarkers (Fig. 2, actual data). This tells her that besides the relatively shallow oil-window marine charge indicated by the biomarkers, there is a deep charge getting into some of the post-salt reservoirs. The QDA excites Marilyn for several reasons. First, her models will be more accurate because she will know to include an active deep source, and secondly, there is a possibility of the presence of active lacustrine source rocks generating below the salt.

The next thing Marilyn wants to know for her model is whether or not the deep source is lacustrine or marine. For this she compares the distributions of large diamondoids (measurements known as QEDA; Fig. 3) to see if the diamondoids in the high-diamondoid concentration samples correlate with the marine or lacustrine rocks. Having run end-members oils, she is then able to correlate the diamondoids to the lacustrine source as shown in Fig. 4 (again actual data). These QEDA correlations are done exactly like biomarker correlations, only instead of using ratios of various biomarker, ratios of large diamondoids and large diamondoid isomers are used. This analysis reveals the deep source because, unlike biomarkers, diamondoids are more abundant in the more mature, deeper charge.

From her diamondoid analyses, Marilyn knows that there is an active deep source in the Santos Basin, and she knows that the deep source is the lacustrine below the salt. She also knows from the biomarker studies the maturity of the shallower marine source. The next thing she needs to do is get some idea of how mature the lacustrine source is. She cannot see it on the seismic, and the heat flow over time below the salt is uncertain. However, she loves diamondoids and knows that they will love her back by giving her an estimate of the minimum equivalent vitrinite reflectance of the deep source.

How will she do this?  Well, she will look at the samples with the elevated diamondoid concentrations and determine the minimum amount of cracking these oils have undergone. The idea is simple really. If the diamondoid concentration is 3ppm in uncracked lacustrine oils from Brazilian basins and the diamondoid concentration is 6ppm in the analyzed sample, she calculates that half of the oil has been cracked to gas, causing the diamondoid concentrations to double.

However, she also knows that oil cracking creates some diamondoids, so to compensate for this, she uses the graph in Figure 5 to convert the amount of cracking determined from diamondoid concentrations to actual cracking determined from laboratory cracking experiments (liquid weight differences before and after heating). And from the same graph, she can read off the equivalent vitrinite reflectance (Easy Ro) necessary to do that amount of cracking. It is this Ro that she uses for the deep source in her model.

So, now Marilyn knows that there is an active deep source in the Santos Basin along with the shallower marine source. She also knows this deep source is the lacustrine and she has an estimate of the vitrinite reflectance at which it was generated.

The next thing she wants to do is get some idea of the relative charge from the shallower marine source versus the charge from the deep lacustrine source for model calibration purposes. And again, Marilyn turns to her best friend, diamondoids! She knows that uncracked oils in the basin have a diamondoid concentration of about 3ppm. However, in some of the reservoirs, the diamondoid concentrations are near 9 ppm. And it isn’t difficult for her to reason that in an oil with 9ppm diamondoids, 3ppm come from the low maturity oil, but the other 6 ppm must come from the deeper charge. If the diamondoid baselines for the marine and lacustrine are both about 3ppm, this tells her that about 1/3 of the charge is from the shallow source and 2/3 is from the deep source. Obviously, this is a bit of an oversimplification, but the method does give an idea of the relative charges. In addition, she knows that the deep charge, being very mature, will have a high GOR and the gas may or may not be trapped in the reservoir.

Our diamondoid-loving basin modeler now has a model for the Santos Basin which includes a deep sub-salt lacustrine source at a calibrated vitrinite reflectance, and she has calibrated the model to charge the various reservoirs based on estimated relative charges determined from diamondoid concentrations.  And she has done all this before the pre-salt discoveries were even made, all because she knew what kind of information she could get from diamondoids.

So, let’s compare her model with one in where the modeler did not realize that “Diamondoids are a Basin Modelers Best Friend”.  This modeler only has the marine sources in their model because that is what the biomarkers are telling them. As such, their models are very different than Marilyn’s, and a comparison of the two is shown in Figure 6. 

This figure only shows the post-salt, but even in the post salt, there are accumulations which Marilyn generates that do not occur in the one-source model.  Furthermore, there are higher liquid volumes, higher GOR’s, lighter oil, etc. in Marilyn’s model. And obviously, huge subsalt accumulations will occur in Marilyn’s model that will not be present in the other.

Years later, after the discovery of the huge fields in the pre-salt, Marilyn constructs new models. And again, she turns to her best friend, diamondoids.  An analysis of diamondoid concentrations (QDA) in oils from the pre-salt Santos fields shows that the fields nearest to shore contain relatively low diamondoid concentrations, and as such, no evidence of oil cracking. Biomarkers show that these fields were generated from the lacustrine source at oil window levels of maturity. The lower diamondoid concentrations indicate that there is no oil coming from other more deeply buried depocenters.  However, as one moves offshore, although the biomarkers are similar in concentrations and ratios, the diamondoid concentrations increase, indicating co-sourcing from a deeper pod. Knowing the geology of the area, Marilyn can reason that the shallower lacustrine charge is coming from lacustrine source pods located to the west of the fields towards the shoreline. This pod is charging all the fields. However, elevated diamondoid concentrations in the more outboard fields indicate that there must be a deeper, more mature kitchen further offshore, also charging these deeper-water fields. This cracked oil and gas must be migrating from deeper water back to the east into the outboard fields.  

A summary of migration directions based on diamondoid analyses, geology and basin modeling is shown in Fig. 7.  Interestingly, the very high CO2 concentrations in some oil fields correlate with high diamondoid concentrations, indicating that the CO2 is using the same migration pathway as the oil and gas from the deeper depocenters.

The example given here is for the Santos Basin of Brazil. However, stacked source rocks occur in almost every prolific petroleum basin in the world and the same benefits to basin models bestowed by diamondoid analysis of Santos Basin oils, can be obtained in a similar manner for any basin in the world. In this way, new and more accurate models can be created in even the most mature, well-explored basins, and these new models will contain new plays not recognized previously.

So, to sum up, remember the words of Marilyn Monroe the basin modeler:” Diamondoids are a Basin Modeler’s Best Friend”. And one last thing: Marilyn is well aware that a bad diamondoid analysis is worse than no diamondoid analysis at all, because it can lead to incorrect conclusions. Diamondoid concentrations must be measured to 0.1 to 0.2 ppm. For example, one must be able to discern uncracked oils with 3ppm diamondoids from cracked oils with 4 ppm diamondoids with confidence. For this, only the best analyses using the best standards allow for the accuracy needed.  The same is true for large diamondoid distributions using QEDA. For such analyses, largely due to their unique deuterated diamondoid standards, BPS/BTI provide diamondoid concentration accuracies to 0.1ppm and are the only providers of QEDA analyses.

Share this post