Why we work the way we do

We get from time-to-time questions such as the one below:

I would like to understand some points in your proposal. You say you examine the core to define reservoir properties for several reasons, including things such as clay typing, mineralogy, porosity, permeability, etc. As we have core in this well, we have had the core lab run the following tests;

-       Porosity and permeability, at various confining pressures

-       XRD work for whole rock and clay mineralogy

-       Reservoir pyrolysis; TOC, Kerogen Quality/Type, Thermal Maturity Testing

-       Triaxial compressive tests; Compressive Strength, Young’s Modulus, and Poisson’s Ratio

-       Over 100 core photos

It is an impressive assortment of data. But still, they all have problems and are part of a much more complex puzzle. It would also take a lot of time to go through the data rather than just pushing the button on a computer, as a manner of speaking. Reservoirs are an incredible mix of properties that determine how much oil and gas it contains and how fast and how much it can be produce. The properties range from height above free water to clay composition, textures, such as rounding, sorting, to mineralogy to clay typing and distribution and much more. Our latest software attempts to combine up to 47 series of 0.1m depth incremented core data with that of wireline logs.

If you follow formations over several kilometers, its lithologies can change profoundly. The same vertically, just by depositional facies, but also by formation or member. For example, in our regional evaluation of the Mannville in East and Central Alberta, we developed for most formations (McLaren down to Dina) unique porosity/permeability relations. If you track the Cummings from Saskatchewan to the Fifth Meridian, we find many reservoirs with different porosity-permeability relations (an expression of laterally variable lithology).

In our regional Montney evaluations we found reservoirs that measured a fraction of a mD to ones with Darcies in permeability. Also, parameters such as Rw (formation water resistivity) changed from one block of townships to the next.

Then there are the shales and clays. Montney shales are often a mixture of clay and silts and dolomites and heterogeneous even on thin section scale (multi-modal). The compositional and textural proportions vary greatly. Many shales and sandstones have unique features, even on a microscopic level. For example, a shale may comprise silt laminations (one to a few millimeters in thickness). The laminations have porosities that range from poor to excellent. Their grain size variation affects permeability, while the clays dispersed inside its pore systems is often a product of deposition (detrital clays) and diagenetic processes.

Vsh is a very crude expression of clay distribution comprising of interlaminations (shale/rock ratio) and dispersed clays in the pore system and different clay mineralogy, as well as clay clasts in a rock’s framework. SEM analysis is a technique to observe how clays are distributed inside a sand, but will this distribution persist when descending a point bar sequence? How does it affect reservoir in terms of permeability or fluid distribution? Especially when the latter is a function of pore type and system, as well as its position above a local or regional free water zone. An XRD sample may provide a semi-quantitative distribution of clay minerals of a small, crushed sample that may easily change along the well bore within a meter. 

Some wireline logs may help to better understand these distributions, such as the spectral gamma ray. When trying to calculate the reserves of a reservoir, quantifying such reservoir properties and extrapolating them along the well bore and laterally in the area of study, our work provides better insights. We use detailed measurements of these properties to better estimate oil in place and potential recovery as expressed by reservoir quality. Remember how, in the past, several companies went bankrupt or fell considerably in share market value, or in net asset value because of a 30% reserve write down. Yet, much reserve work has errors easily in excess 30%. Our work is aimed at reducing those risks.

While having considerable expertise in rock description (such as performing measurements under an incident light or cuttings microscope). When comparing them with thin section analyses using a polarizing microscope, there are considerable discrepancies. Even when measuring the same properties (e.g. vclay or grain size or mineralogical content). This is especially true in shaly and fine-grained reservoirs. The following was my half a page answer to the client, which I hope will also show you why we work the way we do:

We integrate our core and normalized wireline log data using programs such as IHS Petra, to make our best net pay and OOIP numbers for each well and area. In my experience, point counted thin section data (the Eucalyptus way), is best to quantify reservoir properties. We often recommend taking 1 thin section sample per meter of core. This creates, combined with our detailed core descriptions, a continuous record of enumerated reservoir properties (at 0.1m depth increment – the same as modern logs). It took us many years to learn how to make consistent core measurements and then relate them to wireline logs. Today we are closer to that goal than ever before. 

If with our work, we can help to extend the productive life of a high decline well by 1 or 2 years, our efforts are more than paid for. But most geologists focus on depositional setting rather than on reservoir properties. Especially when one combines these reservoir characterizations with production performance, profound conclusions may result.

Petrophysical analysis of wireline logs by itself, provides non-unique mathematical solutions of reservoir properties such as porosity or permeability depending heavily on the algorithms applied and parameters assumed. A quick example from our new software project (Curve 3.0) may help to illustrate (see headline figure). Which of those Vclay curves (especially over shaley intervals) is the right one? You only know when you measure Vclay correctly and direct in core and then translate it in a comparable wireline log-like curve. Same for porosity, water saturation, permeability, etc.

The data you have in your well are extensive and should be helpful. But most are indirect measurements using algorithms derived from other formations and from other regions of the world. For example, a porosity log is typically calculated from other measured properties such as rock density. The calculation is often a non-unique solution and dependent on parameters the analyst assumes to be reasonable.

XRD and SEM measurements are focused on a small detail of the rock. Their results depend on sampling frequency and selection method. Even traditional core analysis data has such problems. We try to generate a high-quality continuous data record over the cored interval. We measure the rock properties directly in core and as precise as we can. Then we convert it into wireline log-like patterns for comparison with the actual wireline logs to increase confidence in our answers.

I hope that this helps you understand our passion for reservoir geology.