A new way to accurately model and calculate true moveable fluid phase saturation in unconventional, organic-rich shale reservoirs.
New and greatly improved technologies in drilling, hydraulic fracturing and imaging have been the game changers that helped lead the oil industry into the shale revolution. The advances in these technologies have brought costs down and production up, yet efficient development and finding the ‘sweet spots’ in these vast fields still remain elusive. A new way to calculate and model saturations in organic-rich shales described in this article could have profound effects on infill volumes, future field development and global oil supply projections.
© Dlouisephotography/Dreamstime.com
Developing any conventional oil field is a daunting process. Multiple development options often exist and selecting the best one can make huge differences in ultimate recovery and overall profitability. Conventional fields have known extents concentrated in discrete accumulations or pools. Saturations of oil, gas and water are well understood, as is the relative movement of these elements through the reservoir when being produced. Now, imagine developing the resources from a widespread, very low permeability and porosity oil-charged horizon composed primarily of shale. Over the vast areas that some of these reservoirs occupy, finding the ‘sweet spots’ and developing them efficiently at current oil prices is essential if a company and this play is to remain profitable.
To date, saturations in organic-rich shale reservoirs have been calculated using methods ported from conventional reservoir and shale dry gas reservoirs. These methods did not account for oil molecules sorbed in/on organic matter in an immobile, solid state. A new and independent method for calculating saturation patterns in organic-rich shale reservoirs subtracts the immobile sorbed oil component, making it possible to calculate the true fluid phase saturations in these reservoirs. Having the ability to calculate accurate recovery factors has an impact on reservoir drainage efficiency and well spacing. Companies can now optimise landing zones based on these new saturation patterns.
Pore Systems in Organic-Rich Unconventional Reservoirs
(a) shows the total pore system, although only the large circle is actually visible; the 3–5 nm resolution limit of SEM imaging means the smaller and parasitic pores are actually invisible; (b) shows the quantities of Dean-Stark water and organic extract (reported as oil), assuming that the Dean-Stark process is able to access the water molecules in the tiniest pores; (c) presents the same size pore space in an organic-rich shale, showing both the pore and the matrix system. Note the sorbed molecules of oil and gas are both in and attached to the matrix, outside the dotted blue line. Only the molecules inside that dashed line are in the fl uid state. (d) shows the quantities of organic extract and solid organic matter removed by overly aggressive solvent extraction during cleaning. The white space will be reported as pore volume even though it includes sorbed molecules that were part of the solid state.
“While developing our method for calculating water saturations, we looked at both mineral matrix and organic matrix pore systems,” says Andrew Pepper, Managing Director of This is Petroleum Systems LLC. “The big difference between the two is the presence of organic sorption in the latter system. We took pore sizes about 4 nm, or about the resolution of an SEM [scanning electron microscope] image. Current core analysis methods employ Dean-Stark clean and extraction to recover all the water and extractable oil from the samples. The pore system cartoons (shown here as a, b, c, and d) clearly illustrate the differences where the organic matrix pore system has a considerable amount of oil and gas molecules locked or sorbed into/onto the matrix and thus not moveable. The cartoons also clearly illustrate how the Dean-Stark method for recovering the water and oil from cores can give misleading saturations in this type of rock. The problems illustrated in cartoons b, and particularly d, persist in current standard mudrock core analysis.”
Quantifying Sorption in Shale Reservoirs
“Sorption is a bucket term that includes adsorption (of molecules onto pore surfaces) and absorption (of molecules dissolved in the solid organic matter),” says Andrew. “Sorption is an important process in both the conventional and unconventional parts of the petroleum system, since it controls and delays the transfer of molecules generated from kerogen into the true pore space of the rock. Generated petroleum molecules must be expelled from organic matter before they can form a phase able to accumulate in the mineral matrix or organic matrix pore system as an unconventional reservoir fluid; or migrate through the mineral matrix or organic matrix pore system to a conventional reservoir.”
Calculated fluid saturations from the Sb interval in the Permian Basin Wolfcamp well. Sorbed and reservoir fluid quantities were derived from core and sidewall core samples by pyrolysis measurements. The yellow line represents live reservoir fluid saturations. © Andrew Pepper.
Being able to accurately subtract the sorbed oil fraction from the total fluids and predict moveable oil became a priority for Andrew and his team. They needed a way to predict a sorption coefficient from the shale geochemical data. Sorption models were developed in the 1980s and use an estimate that carbon can sorb 10% by weight. These estimates were based on the pyrolysis volatile hydrocarbon yield, usually called S1 in Rock-Eval parlance. When it came to organic-rich shales, Andrew recognised the need to further quantify sorption coefficients and has since built a database that has been validated by data from the Delaware Basin and elsewhere.
“We noted that the higher amount of sorbent (kerogen carbon) and the higher the sorption coefficient, then the more petroleum will need to be generated to overcome sorption capacity before expulsion from the organic matter can occur,” says Andrew. “This factor becomes very important when comparing sorbed petroleum quantities in organic-rich shales like the Wolfcamp in the Permian Basin of Texas to organic-lean reservoirs similar to the Middle Bakken Member in North Dakota.
Sorption coefficients defined using solvent extract (i.e. toluene used in Dean-Stark core cleaning) are also typically higher by a factor of two in organic-rich shales when compared to pyrolysis yields. This is because solvents extract significant amounts of non-hydrocarbon (resin/NSO and asphaltene) compounds in addition to hydrocarbon (saturate and aromatic) compounds from the organic matter in shale. For this reason, we rely on pyrolysis hydrocarbon yields which are similar in composition to the oils produced from shale reservoirs.”
The data show that the sorption decreases with thermal stress, which probably reflects decreasing molecular weight and the effect of higher temperature decreasing sorption. In shale sweet spots, which tend to be at high thermal stress, sorption coefficients can be as low as one third the value at the low thermal stresses encountered at the edge of the source kitchen. Obviously, it is important not to subtract too much sorbed oil from the oil budget, otherwise the true fluid saturation will be underestimated.
Quantifying Fluid Saturations in Unconventional Reservoirs
“After being quantified by subtracting the sorbed oil right from the total pyrolysis volatile weight, the non-sorbed petroleum weight of each sample is transformed to a bulk volume of dead oil in the rock; and then to a live reservoir fluid bulk volume using an appropriate formation volume factor. If porosity data are available, saturations are then calculated,” Andrew adds. The calculated fluid saturations presented in the figure on the left are typical of a productive shale sweet spot where there is a significant amount of non-sorbed oil throughout the rock. Fluid saturations are locally high in streaks of high quality reservoir, though overall low.
Petroleum to Water Mobility
“Modelling by Okui and others using relative permeability curves found an interesting correlation between relative permeability and irreducible water,” says Andrew. “By studying conventional reservoirs with differing absolute permeability, they found increasing irreducible water with decreasing permeability. ‘Conventional’ water-wet reservoir relative permeability curves require low water saturation, typically less than 50%, in order for petroleum to flow preferentially to water. However, with decreasing absolute permeability and increasing irreducible water, they projected that petroleum should be able to flow preferentially to water at higher and higher water saturations.”
“Using the results presented in these graphs (below), and given the shale permeability and saturation, water cuts may be estimated in advance of the completion of a lateral well,” says Andrew. “The results confirm that with lower permeability, more and more of the water occupies tiny pores that are disadvantaged in the flow process: it is the location of the water – not just its volumetric saturation – that determines the relative flow of petroleum and water from a reservoir.”
This graph demonstrates that conventional reservoirs with up to a Darcy of permeability will need low water saturations (less than 50%) to produce oil. However, for unconventional reservoirs in the 10 to 1,000 nD permeability range typical of many mudrocks, petroleum remains preferentially mobile relative to water at extremely high water saturations; even as high as 80–90% Sw. © Andrew Pepper.
This graph compares actual unconventional reservoirs in the 10 to 1,000 nD range. Points M (Marcellus gas-condensate play in West Virginia) and E (Eagle Ford high GOR play in Texas) are from mudrock reservoirs that return frac-water, with no significant formation water production. The model predicts, correctly, that they should produce water-free petroleum at ~60% Sw. Note that with ‘conventional’ reservoir relative permeability (from the previous graphs) this level of saturation would result in a about a 50:50 petroleum:water mix at downhole conditions! Points Sb, St, W, and B are from Wolfcamp targets in the high GOR play in the Delaware Sub-basin of the Permian Basin in West Texas. Note the low oil saturations in this very prospective play. © Andrew Pepper.
A Game Changer for Unconventional Reservoirs
For companies exploiting and exploring organic-rich shale reservoirs, being able to predict water saturations before drilling will certainly be a game changer. For some of the Delaware Basin wells that encountered hard-to-explain water production and for some oil-bearing rocks that failed to produce fluid, the concepts presented here help explain certain trends for shale production. Early pore fluid analyses did not capture the whole picture nor did it recover all the water. Correcting the pore fluid calculations and separating sorbed from fluid phase is key to understanding how fluids move in very low permeability reservoirs.
How does this apply to the unconventional field? We now can conclude that since presumed saturations have been incorrect typically by a factor of two or more, recovery factors are much higher than originally assumed. Using only the potentially mobile fluid phase petroleum translates to lower oil-in-place and higher than expected recovery factors. Therefore, frac jobs are accessing a larger fraction of the net mobile fluid which has a huge impact on reservoir drainage efficiency and well spacing. “Enhanced profitability can come from drilling fewer wells to get the same EUR,” speculates Andrew. “There is evidence from reservoir simulation that current infill (daughter) wells may be delivering production rate without recovering significantly more than the parent wells would do, over time.”
