The 8th RPI for November-December 2016: Per Avseth

Independent geophysical consultant, Oslo, Norway
Adjunct professor in reservoir geophysics at NTNU, Trondheim, Norway

About the influencer

Per Avseth is currently an independent geophysical consultant based in Oslo and an adjunct professor (since 2008) in reservoir geophysics at NTNU in Trondheim. Per is an expert in rock physics and quantitative seismic interpretation. He has been working in the oil industry for more than 16 years, and he started his career as a seismic interpreter at Norsk Hydro Exploration in Oslo in 1994, after finishing his master of engineering degree in applied petroleum geoscience at NTH (now NTNU) in 1993. In the fall of 1994, he pursued graduate studies at Stanford University where he received his PhD in geophysics in 2000. In 2001, after a short post-doc position at the same institution, he started at Norsk Hydro Research Center in Bergen, where he worked as a research geophysicist for 5 years. After that, he was consulting for 6 years (2006-2012), first in in his own consulting company (Rock Physics Technology), and later with Odin Petroleum, both in Bergen, Norway. In 2012, he started working with Spring Energy, a small oil exploration company in Oslo. Spring got bought by Tullow Oil in 2013, where he worked as a geophysical adviser until September this year. Tullow Oil has recently decided to exit Norway, and Per plans to resume as a consultant doing technical services within rock physics and quantitative seismic interpretation for oil companies. However, he plans to dedicate the coming months to update the book he co-authored together with Prof. Tapan Mukerji and Prof. Gary Mavko at Stanford University, titled “Quantitative Seismic Interpretation – Applying Rock Physics Tools to Reduce Interpretation Risk”. Currently his research focus is on the links between burial history, rock physics properties and seismic signatures of reservoirs, seals and source rocks on the Norwegian shelf. 

Your secret recipe and pathway for success in becoming a well-known name in the rock physics community

A combination of many factors are to blame, I think. First of all, I was at the right place at the right time. I started my master’s degree at the Stanford Rock Physics and Borehole Geophysics (SRB) Laboratory in 1994, at a time when AVO, seismic inversion and quantitative interpretation were still bleeding edge technologies in the oil industry. At the same time, the Palaeocene interval in the North Sea was starting to become one of the most prolific targets for oil exploration on the Norwegian shelf. I had just completed a master’s thesis on sedimentology and sequence stratigraphy at the Norwegian Institute of Technology (now NTNU), and also spent a few months working as a conventional seismic interpreter at Norsk Hydro before I came to Stanford University. When I started to learn about rock physics and the intriguing work done by Marion, Yin, Jizba, Han and others who had just finished their PhDs at Stanford a few years before I arrived, I realized that what they had done could be very relevant for a better understanding of seismic signatures of the Palaeocene turbidite systems in the North Sea. When Prof. Amos Nur asked me to stay for a PhD, it was very easy for me to answer yes, as I was very inspired by the new possibilities of linking geology and geophysics using my new knowledge in rock physics (Off course sunny California was also a reason by itself why I wanted to stay). I was also fortunate to work with a lot of bright people at Stanford who inspired me and tought me new things that was very valuable for my own work, including Tapan Mukerji, Jack Dvorkin, Ran Bachrach, Lev Vernik, Manika Prasad, and off course my advisor Prof. Gary Mavko. I also worked closely with the sedimentology group at Stanford when I was there, as my PhD thesis was on seismic reservoir characterization of North Sea turbidite systems. The fellow students at SRB at the time were also a great source of inspiration, as we complimented each other and collaborated to solve problems that were very relevant for the oil industry. So to summarize, I think the key to my success was to be passionate about earth sciences in general, to be open-minded to new ideas, working with and learning from other people who complimented me, and catching the opportunities that were created in the crossroads of different disciplines and emerging technologies. This whole process culminated into a very cross-disciplinary thesis, and eventually into a book that I wrote together with Tapan and Gary, “Quantitative Seismic Interpretation – Applying Rock Physics Tools to Reduce Interpretation Risk”. This book turned out to be a spring-board to fame within the Rock Physics community, and I feel very privileged and fortunate to be part of this community.

Challenges you see in moving rock physics to the next level 
(This can be unresolved issues in rock physics in general or in a particular field you are working on)

I think we will see an even broader use of rock physics in the future. For instance, we are starting to see the use of rock physics in the processing stage of seismic data. By linking basin modeling and rock physics, more reliable velocity models can be established as input to migration of seismic data, but also to constrain seismic inversions. With increasing amounts of data, we will see more use of statistical and mathematical tools (data clustering, eigenanalysis/PCA, neural networks, singular value decomposition, etc.) to detect clusters and patterns in the data. It is important not to forget the physics when we screen and explore big data. Statistical rock physics has been very efficient to honor both physics and uncertainties. But very often we select a favorite rock physics model, and then do sensitivity studies of the model parameters. With increasing computer capacity, we can use several models at the same time, and invert not only for the model parameters, but also determine the optimal rock physics model for a given case/interval. Prof. Tor Arne Johansen and Dr. Erling Jensen at University of Bergen recently developed a way to do this, called “inverse rock physics modeling”, which is very promising. 

With the new developments in broadband seismic, there is a potential to better resolve the reservoirs, but also there is a potential to extract more rock physics information from the lower frequencies. We still don’t fully understand the effect of fluids in general, and mesoscopic flow in particular, at very low frequencies (i.e. a few hertz). Another issue that remains unsolved in rock physics, is the upscaling of pressure effects. We often assume that pressure sensitivity is given by microcracks and grain contacts. But how do we capture pressure sensitivity for larger scale fractures, and what about stress sensitivity in interbedded sands/shales during depletion or injection of thin-bedded reservoirs?  Do we assume zero stress change in the shales, or will there also be stress changes in the interbedded shales? And how do we quantify and upscale these? 

Another challenge that keeps coming back, is the effect of anisotropy. Often we lack information about the Thomsen parameters when we deal with marine seismic data. New logging tools are developed in recent years that can provide essential information about anisotropy, but unfortunately the oil industry often avoid collecting this information for budget reasons. Luckily, and thanks to the oil shale/shale gas industry, there has been a lot of progress in shale rock physics based both on experimental and theoretical work in the recent years, and with good geological knowledge we can now model shale anisotropy quite reliably. 

A closer link between geomechanics and seismic rock physics, or static and dynamic properties, is needed to better honor the effect of stress anisotropy that is often present in shallow reservoirs, for instance in the Barents Sea where there have been large uplift episodes. We still don’t have good rock physics models to capture the textural effects of diagenesis in shales. How is micro-crystalline quartz cement in shales affecting the velocities in different directions during the smectite to illite transition? And what if we have organic rich shales? How do we disentangle the various factors affecting organic rich shales during burial, including diagenesis, maturation, pore pressure build-up, microcrack generation, and hydrocarbon expulsion? 

At the end of the day, our biggest challenge is the rock physics “bottle-neck”, meaning that we normally only have 2 or at most 3 (quasi-)independent geophysical observables (f.ex AVO Intercept and Gradient, or acoustic impedance and Vp/Vs), but we have many more unknown geologic parameters. There are two main roads to reduce uncertainties, solve ambiguities, and improve our predictability; either get more independent observables (CSEM, gravity data, Q-attributes), and/or to constrain our inversions/predictions with more geological knowledge. We need to continue to acquire more data and integrate more disciplines. With increasing focus on subtle, stratigraphic traps, we need to better understand the link between sequence stratigraphy and rock physics, at the same time as we take advantage of the broadband seismic technology. Do our rock-physics based lithology predictions match with what is expected from sequence stratigraphic principles?

Advice for early career scientists (rock physicists, geophysicists, etc.)
(This can be in term of inspiration or direction you see young scientists should focus on)

I would advise early career scientists to think outside of the box, and dare to challenge the conventional wisdom. New hydrocarbon discoveries will be made by geoscientists who are creative and innovative and come up with new ideas. Don’t only follow the existing “best practices”. Geophysics is a field that allows for creative thinking, and there are plenty of unsolved problems yet to be solved in our industry. And dare to cross borders between disciplines; the most interesting problems to work on are often hidden in the borderline area between existing disciplines. I think computer coding and math skills will be increasingly important relative to earth sciences for the future rock physicist, especially if you want to explore big data. Rock physics is also very important during production and 4-D time lapse analysis, so if you are an explorationist working with rock physics, know that you can use a lot of your knowledge also in reservoir characterization and monitoring studies. Finally, with increasing awareness and focus on climate issues and clean energy, I think that rock physics can be important outside the conventional oil industry, so if you lose your oil job, don’t panic. Rock physics is already important for CO2 sequestration, but can be important also in geothermal studies, geohazard analysis, water aquifer monitoring, etc. So to all earth science students or young geo-practitioners: Rock physics has a great future regardless of the future of the oil industry! 

We thank Per for his continuous contributions to the rock physics community.