Expert Viewpoint with Oliver Mullins
Heavyoilinfo.com spoke to Oliver Mullins, Schlumberger Scientific Advisor and Wireline reservoir domain champion, about asphaltenes, compositional gradients, flow assurance and other fluids related issues pertaining to heavy oil.
Oliver Mullins at a book signing event at the IPTC, Malaysia, December 2008.
Mullins latest book “The Physics of Reservoir Fluids, Discovery through Downhole Fluid Analysis” was published in December 2008. The publication includes overviews that can be understood by non-experts and sections aimed at oilfield practitioners. Most of the material in this feature article is explained in more detail in the new book.
· Diagnosing asphaltene deposition
Oliver is moderating our Asphaltene Forum. Ask him a question: here.
Compositional variations
Heavy oil exhibits large compositional variations between different reservoirs and compartments of an oilfield, and there are often compositional gradients within a compartment. Compositional variations can lead to significant variations in fluid physics, such as viscosity, which have a major impact on hydrocarbon production. So what causes these compositional gradients?
It is generally assumed that biodegradation, which is now known to be predominantly anaerobic, is the root cause. Reservoir temperature seems to be the key controller, as it must be less than 80 degC (176 degF) for biological activity to occur. In general, shallower oil is heavier than the oil found in deeper, hotter reservoirs. Microbes living in oilfield waters preferentially consume alkanes at oil/water interfaces. This consumption leads to higher concentrations of asphaltenes close to the oil/water contact, yielding fluid gradients, and a small increase in asphaltene concentration leads to a big rise in viscosity.
Biodegredation is not the only activity influencing compositional gradient. John Stainforth, Shell, proposes that reservoir composition reflects the charge history of a reservoir: in a normal sequence of burial, heavy fractions come out first, with increasing temperature and longer burial times, lighter fractions are produced. Lighter factions may migrate to the top, displacing previously charged, heavier factions to produce a “layer cake” that is heaviest at the bottom.
It has been assumed that, over the millions of years of geological timescales, reservoir fluids will tend towards a state of equilibrium; but this may not be the case. William England, BP, suggests that the diffusion process is extremely slow so mixing can be inhibited, especially if fluids are not flowing through each other. Migratory paths, such as high permeability streaks and faults, can bypass the bulk of fluids in a reservoir, thereby inhibiting mixing. By contrast, convection can be very effective at mixing fluids; however, efficient convection requires high permeabilities and particular reservoir geometries such as tilted sheet sands.
Equilibrium applies in a geological sense and implies that a reservoir has reached an unchanging state where minor changes in ambient conditions can yield subtle shifts in the equilibrated state. In contrast, metastable states are often dislodged by subtle changes in ambient conditions. A state of equilibrium may exist in the fluid columns of some reservoirs, but this is often not the case. Dynamic processes acting on reservoir fluids often preclude equilibrium. These processes include biodegradation, current reservoir charging, water washing, and (pressure dependent) leaky seals. We cannot assume equilibrium, especially in heavy oil columns where the high viscosity impedes mixing. To accurately predict production, models must take into account the dominant physics incorporating potential dynamics. Proper interpolation between, and extrapolation from, wells cannot be performed without a good understanding of the physics and mechanisms involved, but there is disagreement, even among the experts, about these mechanisms. For heavy oil, some champion biodegradation, while others champion reservoir charge history as dominating. If leading experts cannot predict reservoir fluids, then how can an operating Asset Manager?
Asphaltene plugging
The avoidance and management of asphaltene plugging is one of several challenges frequently encountered in heavy oil reservoirs. A standard approach used by operators is to acquire a sample, test it in a laboratory, then use the results to design an appropriate production system. Mullins considers this approach to be flawed; it is comparable to assuming that lab test results from a single sidewall core are representative of the properties of a whole reservoir rock formation, which no-one would accept. We should ascribe to the oil column the same possibilities of compositional variation that we do for rock, but to do this, we must develop the workflows required to properly address it.
Downhole Fluid Analysis
Mullins asserts that “there is no such thing as the oil in a reservoir”. He recommends that, in order to characterize vertical and/or lateral variation of fluid, particularly within a heavy oil reservoir, Downhole fluid analysis (DFA) should be performed in several wells. The vertical coverage (downhole sample interval) must be adequate to characterize the variation. Before bringing samples to surface with a tool such as an MDT Modular Formation Dynamics Tester, it is necessary to determine whether a gradient exists. If it can be shown that no gradient exists, the acquisition of multiple samples is likely to unnecessary. Acquiring high-quality MDT samples, shipping them to a laboratory and obtaining an extensive lab report can take a long time, often more than six months. Knowledge gained from DFA in real-time enables the design of current sampling jobs to be immediately adjusted to match the complexity of the reservoir and optimize the time taken for the whole process.
Mullins suggests an operating practice whereby the MDT is taken to the middle of an oil column and a sample taken. DFA measurements should then be made above and below that level to look for variation. In addition to determining the need for additional samples, the identification of sudden discontinuities in asphaltene concentrations, among other fluid parameters, can indicate compartmentalization barriers. Since well testing is not performed ubiquitously, DFA methods to identify compartmentalization are very useful, particularly in deep water where, typically, even one well test is very costly. Mullins wants to overcome the antiquated view that a reservoir is one giant tank of homogeneous hydrocarbons in equilibrium, which we now know is often not the case. The incorrect presumption of one homogeneous oil type can lead to expensive errors in production such as improper facilities. Understanding these potential implications should accelerate uptake of DFA and establish new workflows both for exploration and development. Indeed the rapid growth in the application of DFA lends credence to the underlying principles discussed herein.
There is a focus in the industry on reservoir simulation using dynamic reservoir models to predict production rates. Within this context, the most accurate description of the reservoir is often not constructed, nor tested with measurements. Indeed, Mullins notes that, for good reason, it is rare to find experimental human endeavors where error is determined strictly on a computer. New work flows are mandated. The most accurate description of the reservoir is the static geologic model – filled with the best understanding of the fluid model. This accurate model should be used to predict MDT-DFA log data, and then tested by measurements to determine its validity. If the model is in error, the MDT is in the well, so measurements can be made to delineate the failure of the model. This process will provide an operator with a much better understanding of the reservoir.
Diagnosing asphaltene deposition
The causes of, and parameters that control, asphaltene deposition have traditionally been tested in a laboratory. Mullins suggests that problems with plugging should be treated in a way comparable to treating a sick body: “first find the cause”. Possible origins of asphaltene plugging may include depressure, commingling, water injection or mobilization of tar mats. In order to solve the problem, the origin must be understood. A good understanding of compartmentalization is also required for proper diagnosis and identification of solutions. To continue the medical analogy, Mullins suggests that we should compare the current state of a reservoir with an asphaltene problem to its previous state. Rather than only looking at problem areas and implementing a fix, good field areas should also be analyzed to determine why the problem is not apparent in those parts of the field. After thorough diagnosis and implementation of a solution, analysis should be performed to determine the effectiveness of the fix.
Downhole fluid logs
According to Mullins, to properly characterize reservoir fluids, the industry needs a quasi-continuous fluid log. In the near-term, Nuclear Magnetic Resonance (NMR) measurements can provide a guide to locating DFA station intervals. Mullins’ new book “The Physics of Reservoir Fluids, Discovery through Downhole Fluid Analysis,” is designed to guide the use of this new technology, including methods for comparative analysis not only of single well DFA, but also multiwell DFA data across a field. DFA covers a lot more than sample acquisition – indeed it is more about reservoir evaluation.
Integration of several measurements will help to improve our knowledge of downhole fluids. The recently launched Schlumberger InSitu Family downhole reservoir fluid properties measurements provide quantitative information about fluids at true reservoir conditions. The foundation of the IFA compositional measurements remains near-infrared spectroscopy, a technology routinely performed in refineries but until now rare in upstream applications. Acquired with the InSitu Fluid Analyzer service, the measurements are used for real-time DFA. The service provides accurate measurements of fluid composition, including gas/oil ratio (GOR). We would expect changes in fluid density to correlate to changes in GOR and asphaltene concentrations. Petrel provides a platform to integrate the various measurements.
One of the first DFA projects for reservoir reconnaissance was in the deepwater Tahiti field in the Gulf of Mexico, operated by Chevron. A large asphaltene gradient was found in the black-oil column. Schlumberger applied a “colloidal suspension” model that used Archimedes buoyancy principles and other basic physics to predict the asphaltene gradient. The resulting colloidal characterization of crude oil agreed with advanced laboratory asphaltene science. This asphaltene model was then used to predict connectivity helping to prioritize development drilling for targeting the largest compartments first. Indeed predicted asphaltene concentrations in the first development well matched predictions indicating connectivity.
Identifying compartments
Thin sealing barriers that separate reservoir compartments represent a challenge as they cannot be imaged from the surface and may also be invisible to wireline petrophysical logs. The standard industry solution for establishing communication within a compartment has been based on the assumption that, if pressures match between two wells, they must be connected. This is inadequate. While pressure can equalize over geological time, even with pinhole porosity, fluid flow over the life of a producing well requires far higher permeability.
Another common industry misconception is that compartments are large unless proven otherwise. In reality, as with all geophysical scaling of magnitude versus frequency, small compartments are much more common than large ones. Consequently, compartment size must be established not presumed, DFA offers a powerful avenue to address this concern.
Finding the Petroleome
Schlumberger is developing, with many academic colleagues, a new technical field to better understand the composition and physics of heavy oil. Mullins notes that light-ends have been properly treated chemically in the oil industry. Light-end characterization of differing alkanes, H2S, CO2 and other gases is rooted in chemical fundamentals. However, the heavy-ends have been lumped as “asphaltenes,” an operational, not chemical, definition. Until recently, there has been no agreement regarding the basic chemical properties of asphaltenes. As Francis Crick, a co-discoverer of the structure of DNA, famously stated “If you want to understand function, study structure.” We must understand the chemical structure of heavy oil if we are to better understand its flow and predict its production. We are now at the threshold of following genetic research and finding the “Petroleome”, enabling a proper predictive science, as is becoming standard in medicine. This capability leverages two recent fundamental advances. First, as detailed in several highly cited papers from the Schlumberger-Doll Research (SDR) center in Boston, asphaltenes are monomeric, not polymeric. Second, ultrahigh-resolution mass spectroscopy, in conjunction with two-dimensional gas chromatography, is close to providing a complete chemical listing of all components in a heavy crude oil.
Wettability
Wettability can be a major issue and is a particular challenge in the Middle East where transition zones hold huge reserves. Some components of heavy oils are interfacially active, thus surfactant, between oil and water and/or oil/rock and/or water/rock. The Max Plank Institute for Marine Microbiology has discovered that the heaviest organic matter dissolved in seawater is a derivative of asphaltene, and that the oceans probably contain between 10 and 100 GigaTonnes, which may be important in the global carbon cycle. This is comparable to the amount of asphaltenes in all the Athabasca tar sands. Asphaltenes oxidize when near ocean surfaces, so they are assumed to originate from natural seeps in deep water. Schlumberger has started a 3-year research project working with Oceanographers to understand these complex molecules and investigate the chemical dynamics of asphaltenes, both in seawater and in oilfields. Knowledge gained may help our characterization of reservoirs, including chemical activity at the oil/water interface. It should also improve our understanding of process involved during water and steam-based enhanced oil recovery (EOR) projects. Knowing the chemistry of these systems will make it plausible to design more effective recovery techniques, leveraging or overcoming impediments to wettability caused by natural fluid chemistries. A small amount of soap can enable the mixing of large volumes of water and oil, so there are very good reasons for building a better understanding of the behavior of natural surfactants in water.
Fluids analysis expertise
Mullins is part of a global team that includes experts in the principles and practical operations of fluids analysis and reservoir characterization. This reservoir-focused team addresses not only the analysis of downhole fluids, but also the acquisition of data during projects, enabling tools to be used in the field to their optimum potential. Indeed, with the current low oil prices, it is more essential than ever to utilize the best available expertise to extract the full value from oilfield services technology.
Oliver is moderating an Asphaltene Forum. Sign in and ask him a question here: Forum link
Biography of Oliver Mullins
Dr. Oliver C. Mullins is a chemist with a PhD from Carnegie-Mellon University. He is the primary originator of Downhole Fluid Analysis (DFA), a significant new service in the oil industry. The corresponding tools exploit near-infrared, visible and fluorescence spectroscopy and are being used to uncover compartmentalization and hydrocarbon fluid complexities in subsurface reservoir formations. His current position is Reservoir Domain Champion for Wireline Headquarters, Schlumberger.
DFA acceptance by the industry is reflected in Dr. Mullins being a Distinguished Lecturer for both SPE and SPWLA. Dr. Mullins also leads an active research group in asphaltene and petroleum science which has resolved several important controversies. He has authored a book on DFA and co-edited 3 books and 9 book chapters on asphaltenes. He has coauthored 130 articles and is co-inventor on 50 allowed US patents. His hobbies include skiing, scuba, biking and blues saxophone.
Further Reading
Betancourt, S. S., Creek, J. L., Cribbs, M. E., Dubost, F. X., Mathews, S. G. & Mullins, O. C. Predicting Downhole Fluid Analysis Logs to Investigate Reservoir Connectivity. International Petroleum Technology Conference, December 2007. Paper 11488-MS.
Cubitt, J. M., England, A. & Larter, S. R. Understanding petroleum reservoirs: towards an integrated reservoir engineering and geochemical approach. Geological Society (2004).
Jones, D. M., Head, I. M., Gray, N. D., Adams, J. J., Rowan, A. K., Aitken, C. M., Bennett, B., Huang, H., Brown, A., Bowler, . F. J., Oldenburg T., Erdmann, M., Larter, S. R. Crude-oil biodegradation via methanogenesis in subsurface petroleum reservoirs. Nature 451, 176-180, 10 January 2008.
Mullins, O.C., Sheu. E.Y., Hammami, A. & Marshall A.G. Asphaltenes, Heavy Oils and Petroleomics. Springer, New York (2007).
Stainforth, J.G. Practical kinetic modeling of petroleum generation and expulsion. Elsevier Ltd. (2008).
