Just got an email that asked “What did you miss at the Society of Exploration Geophysicists Annual Meeting?” so you know I had to ask the title of this post. A lot of the great things I saw at this year’s SEG conference happened off the main floor:
1. The first geophysics hackathon: The two-day hackathon held by Agile Geoscience and its outcome are detailed very well by Matt Hall in these posts. What I got out of it was somewhat different from what other participants did. I left programming (Fortran, Visual C++, MFC) behind a decade ago for full-time seismic interpretation, reservoir characterization and technology development. Getting my first flavor of Python in 2003, spending a lot of time with geographers and digital map developers in 2009-10 and, lately, a strong desire for more open, nimble and intelligent geophysical apps have me interested in coding again.
The hackathon introduced me to the mechanics of Scientific Python or SciPy, “open-source software for mathematics, science, and engineering,” and very powerful and visual scientific-programming interfaces – Enthought Canopy, Continuum’s Anaconda and iPython Notebook. After playing with some online examples of SciPy and NumPy, I was able to understand the syntax behind and reproduce Zoltan Sylvester’s grain settling code as well as generate some geologic surfaces. Granted, I have a long way to go before quickly translating geophysical formulae from mathematical notation to code and There Will Be Debugging, but it’s a start.
The best part was just being around scientist-programmer hybrids to talk open-source tools and rocks in a creative, comfortable and open co-working space in East Downtown Houston called StartHouston. Every single day, I work with at least a dozen brilliant scientists and engineers on pushing the geologic and geophysical frontier, but all the data is proprietary and analysis performed on bulky, commercial software in offices and meeting rooms, i.e. the physical, technological and intellectual opposite of Open. As Gaël Varoquaux says in this timely post on publishing scientific software, “the public availability of code is a cornerstone of the scientific method, as it is a requirement to reproducing scientific results … like [Galileo's] telescope, it also builds upon scientific progress and shapes our scientific vision.” Thanks to the Concentric Circles Of Awesomeness known as dGB Earth Sciences, Enthought, and OpenGeoSolutions for sponsoring this weekend of innovation and great food (tacos, banh mi, po boys and mole enchiladas – don’t ever change, EaDo!) and current and former luminaries of the SEG for observing and offering support.
I got to judge the hackathon entries along with Enthought’s Eric Jones, dGB’s Paul de Groot and SEG’s Chris Krohn, Dennis Cooke and Peter Annan. As a developer, expert end user and someone who works in exploration and field appraisal now, I appreciated Team OpenGeo’s app for its quick and easy forward-modeling ease, Team Sweetness’s entry for rapid velocity uncertainty estimation at the drillbit and Team Teapot Dome’s stab at parsing the contents of LAS files, the data-management bane of subsurface professionals all the way from exploration to production. More on these apps in this Agile post. Great job, Matt and Evan! Hope to be a coding team member at the next hackathon.
2. Women’s Network committee meeting and breakfast: You’ve read me on the SEG Women’s Network before. We exist because we are professional women versed in a male-dominated discipline working in an industry with, as our keynote speaker this year described, “codes, processes and decisions made by men for men.” This year’s breakfast speaker was Sophie Zurquiyah, the Senior Executive Vice President of CGG’s Geology, Geophysics & Reservoir Division, and she gave a very honest and direct talk on how companies can promote gender diversity and individual career management. Beyond the ethical case, there is a business case for caring about the careers of female technical workers: at a time when CGG alone requires 400 functioning geoscience professionals for current and upcoming work (probably projecting across the course of the next decade), the company that grooms competent men and women in successful careers, doesn’t lose them along the way and promotes them to first-line management roles will be that much far ahead.
A tip that Zurquiyah reiterated over the course of her talk is one I’ve internalized only in the last few years and cannot emphasize enough now: “Take charge of your own career. State out loud what you want. You don’t get what you don’t ask for.” Just because you are a good geoscientist and do your job well doesn’t mean management is going to discover you in the lineup and shove you and your career to the oilfield Oscars. Women also have a tendency to self-limit, we second-guess ourselves while more junior men are asking their superiors how to be the next team lead, vice-president and CEO. Again, women operate in a very “narrow mudweight window” (Zurquiyah’s geomechanical term for the glass ceiling), the age-old roadblocks of biases and perceptions do still exist and men promote those they understand, i.e. other men, so manage your own career, make yourself understood and work towards that stated goal. She also reminded us that there are many men who aren’t prejudiced and egotistical, so identify such men as enablers and build a network of crucial mentors and contacts. Like with the hackathon, surround yourself with good, like-minded people.
Alex Herger, Marcia McNutt and Sophie Zurquiyah have all stressed to us that partner and family support is critical, whether you are the only breadwinner or half of a dual-career couple. Zurquiyah’s acceptance of senior positions required family flexibility, discussion and support; her husband took different types of jobs or didn’t work, depending on her career at the time, and it is ok for the leading career to alternate. The key is that work at home is shared (or outsourced), and you’re not a bad spouse and mother if you can Manage It. My mother managed the hell out of her career, our family and more back in the 1960s-90s, while fighting Arab, Indian and European gender bias almost everday, so I can’t believe we’re having the same conversation in America in 2013. Yet, when I hear that Exxon is extending benefits to married same-sex couples and that large oil companies are starting to hold their own internal women’s network gatherings at which they talk about executive career paths, work-family balance and alternate work schedules, I feel we have made great strides. As for the SEG Women’s Network committee, I am honored to do my small part alongside Eve Sprunt, Nancy House, Anna Shaughnessy, Louise Pellerin, Manika Prasad and the growing group of women and men who understand the importance of removing unnecessary sexism in our industry as well as supporting and promoting professionals because that is what we are – professionals.
This year, I am running for SEG Women’s Network treasurer, because I am good at it (thanks, Krewe du Vieux!) and feel that the more I give to this particular effort, the more it will give back. So, join the network and vote for me!
3. The SEG Integrated Quantitative Earth Workshop: The bulk of this four-hour event was the presentation of the four best talks from this summer’s Integrated Quantitative Earth Forum held in Boston, but Rocky Detomo put the WORK in workshop this time around by asking the audience to think and write about three questions during the course of each talk:
- What have you seen and suggest as an opportunity for integration?
- What data sets can be opened for analysis and easier access?
- What data standards/formats are hurdles to efficient analysis?
The four “winning” papers were Geologic Controls of AVO Systems in the Niger Delta: Impact on Exploration Evaluation by Shell’s Chris Wojcik, Integrated Stratigraphic Concepts as Functional Templates in Reservoir Models by Exxon’s Matthew Casey, Anisotropic Model Uncertainty Quantification in Seismic Tomography by Schlumberger’s Konstantin Osypov and Structural Interpration in Color by Western Geco’s Andreas Laake.
Chris Wojcik’s paper was most interesting to me. We know by now the value of Direct Hydrocarbon Indicators (DHI) in derisking prospects, but it is the non-unique anomaly presence (as well as the anomaly absence or, as Chris calls it, “the curious incident of the non-barking dog”) that begets the need to really integrate across disciplines, especially with geology, basin modeling and petrophysics, and scales within those disciplines, all the way from thin section through core and logs to seismic. The same rock taken through different pore pressure or burial history gives different results. Different rocks at similar depths and pressure-temperature regimes look the same. A response, or lack thereof, is instructive but not indicative (are you listening, geophysicists?), so it’s important to understand the variability of geological scenarios and appropriately characterize the reservoir.
NURBs, or Non-Uniform Rational B-splines, are making their way from the world of 3d modeling and animation into geomodeling, I see. It makes sense given that NURBs are easily parameterized and lend themselves to functional form modeling of geological bodies (channel complex, channel lobe, shoreface, ramp). There is then no reliance on a geocellular grid and makes static and dynamic models easier to update and less memory-intensive to run. [Exxon paper]
Konstantin Osypov gave a nice but very rapid talk on depth and structural uncertainty representation, and how it may be properly carried through the life cycle of processing and interpretation. While he didn’t say this explicitly, the way we describe and monetize a reservoir is somewhat silly in that we carry one geo-model with some attached statistical perturbation, but it is essentially one model and easy to fall in love with (are you listening, geologists?). What is an “optimistic or pessimistic geo-model” in the face of more scientific, risk-neutral geophysical models with built-in risks and uncertainties onto which you can attach reservoir properties? It’s akin to the horizon and fault clouds Matt Hall and I discussed earlier that week, i.e. a fuzzy collection of interpretations as the model object rather than one flexed surface with attached uncertainties. Reservoir volumetrics calculation and uncertainty estimation have always seemed very chicken-and-eggy to me, and there is definitely lots more to learn.
The IQ Earth initiative now has a forum and workshop as well as the new Interpretation Journal, a joint effort of SEG and the American Association of Petroleum Geologists (AAPG). What is missing from the “measures of success” is an open data set curated and offered by the SEG. Statoil, as the IQ Earth initiative’s corporate sponsor, was to offer up a North Sea data set, but that was squashed by their lawyers, natch. While searching the internet for free data sets at the hackathon, I came across the Norwegian University of Science and Technology’s Norne Field Benchmark Project with operational data freely provided for research and public access by Statoil. Granted, it isn’t seismic, well data and static reservoir properties, but if Statoil has released Norne field production data, why not release the rest of it instead of asking SEG to get some?
4. Catching up with friends and making new ones. I finally met fellow 52 Things author Rachel Newrick, albeit very briefly, and purchased her monograph (co-authored with Larry Lines) Fundamentals of Geophysical Interpretation. Yonghe Sun, the editor of the new Interpretation Journal, and I chatted for a long while and he encouraged me to edit a special section on Inversion For Reservoir Characterization if I indeed want to see such a section in the journal.
Next year’s SEG conference is in Denver, but I am probably going to save conference time for the Canadian equivalent (CSEG). Lots of Work Work ahead this week, so hope I can write soon about Newrick’s book, fall (as in “autumn”) rituals (as in “kitsch”) at my house as well as doing and talking science in a proprietary corporate culture. A sneak peek: Competitiveness is impossible when everyone has the same technology, so be careful when you say you want to “standardize.”