Pseudogeophysics

Seventy-five years ago, the first paper of the first issue of the journal Geophysics was published: Black magic in geophysical prospecting by Ludwig Blau of the Humble Oil and Refining Company. If you are an exploration geoscientist, you must read it. Then go and get someone else to read it too.

The paper is remarkable for a few reasons, apart from simply being first:

• it is scientific, but warm and humorous, and utterly compelling
• it is subtle but devastating in its condemnation of some of the day's technology
• the critical approach Blau obtusely describes is still relevant to us today

How to crack a nut

There are two parts to the paper: a brief spotter's guide to black magic, followed by eighteen examples from the author's own experience.

Blau's guide to the characterisitcs of a nutty inventor is timeless. It presages John Baez's wonderful Crackpot Index. Here are the highlights:

• The inventor has been working alone for many years, usually about 20
• The inventor has no formal training, regarding this as a hindrance
• The inventor has many Nobel prize-winning scientist friends
• None of these friends understand the contraption in question (owing to their hindrances)

Thus it was proved...

Yet more enlightening is Blau's categorization of geophysical technology. He identifies five modes of detection, from the merely implausible to the downright bizarre:

• Particle radiation, akin to α or β radiation
• Non-gravitaitonal forcefields, attracting 'bait' oil
• Radiant vibrations, detectable by skilled divination
• Electromagnetic waves, readily received by a radio
• Sexual emanations. No, really.

But it's the vivid descriptions of the contraptions and their inventors that light the paper up. Blau's brief but scathing reviews are so drily delivered, one imagines he must have been a man of few, but golden, words.

Here is Blau is describing the conclusion, and coup de grâce, of a hotel meeting with a pair of gentlemen peddling a stick which, when primed with a capsule of oil (or any other sought-after substance), points decisively to the nearest reservoir:

When the “bait” was changed to whiskey, the device in the hands of the inventor stubbornly pointed to a leather bag lying on the bed; the inventor asked his friend how this could possibly be explained since they had finished the last bottle that morning and he had not bought more. Upon opening the bag a pint bottle was revealed and the friend admitted having bought it that afternoon without telling the inventor about it. Thus it was proved that the device was not manipulated or influenced by the operator.

I would give my eye teeth to have been a fly on the wall during that scene.

References

Blau, L (1936). Black magic in geophysical prospecting. Geophysics 1 (1). DOI:10.1190/1.1437076

I am very grateful to the Society of Exploration Geophysicists, Tulsa, OK, for permission to reproduce the first page and quote freely from this paper. The quoted passages are copyright of the SEG. The image of Houston, dating from 1927, is in the public domain and was obtained from Wikimedia Commons. The drawings are original.

Geophysical stamps 4: Seismology

This is the last in a series of posts about some stamps I bought on eBay in May. I don't collect stamps, but these were irresistible to me. They are 1980-vintage East German stamps, not with cartoons or schematic illustrations but precisely draughted drawings of geophysical methods. I have already written about the the gravimeter, the sonic tool, and the geophone stamps; today it's time to finish off and cover the 50 pfennig stamp, depicting global seismology.

← The 50 pfennig stamp in the series of four shows not an instrument, but the method of deep-earth seismology. Earthquakes' seismic traces, left-most, are the basic pieces of data. Seismologists analyse the paths of these signals through the earth's crust (Erdkruste), mantle (Mantel) and core (Erdkern), right-most. The result is a model of the earth's deep interior, centre. Erdkrustenforschung translates as earth crust surveying. The actual size of the stamp is 43 × 26 mm.

To petroleum geophysicists and interpreters, global seismology may seem like a poor sibling of reflection seismology. But the science began with earthquake monitoring, which is centuries old. Earthquakes are the natural source of seismic energy for global seismological measurements; Zoeppritz wrote his equations about earthquake waves. (I don't know, but I can imagine seismologists feeling a guilty pang of anticipation when hearing of even quite deadly earthquakes.)

The M9.2 Sumatra-Andaman earthquake of 2004 lasted for an incredible 500 seconds—compared to a few seconds or tens of seconds for most earthquakes felt by humans. Giant events like this are rare (once a decade), and especially valuable because of the broad band of frequencies and very high amplitudes they generate. This allows high-fidelity detection by precision digital instruments like the Streckeisen STS-1 seismometer, positioned all over the world in networks like the United States' Global Seismographic Network, and the meta-network coordinated by the Federation of Digital Seismograph Networks, or FDSN. And these wavefields need global receiver arrays. 

The basic structure of the earth was famously elucidated decades ago by these patterns of wave reflection and refraction through the earth's fundamentally concentric spheres of the solid inner core, liquid outer core, viscous mantle, and solid crust. For example, the apparent inability of the outer core to support S-waves is the primary evidence for its interpretation as a liquid. Today, global seismologists are more concerned with the finer points of this structure, and eking more resolution out of the intrinsically cryptic filter that is the earth. Sound familiar? What we do in exploration seismology is just a high-resolution, narrow-band, controlled-source extension of these methods. 

News of the week

The summer is, unbelievably, drawing in. While you've been gadding about in the sunshine, we've been scouring the geoscience and technology news — just for you! Here are some things that caught our eye in August. 

Spectral stuff from ffA

If you're a regular reader, you know we'll always cover a story about seismic frequency and resolution. A week ago, purveyors of high-tech seismic attributes ffA introduced HD Frequency Decomposition, or HDFD. As you might guess, HD stands for high definition, though it's not clear what that means in the context of seismic attributes. As is fairly normal with resolution-enhancing technology, the claims are eyebrow-raisingly bold:

...[we can] step beyond the resolution limitations of conventional frequency decomposition techniques.
Steve Purves, Technical Director, ffA

You can read about the technique in the nicely illustrated datasheet, but don't expect to find out how it works. I was disappointed to see that it doesn't even mention the type of decomposition; we assume it's some sort of scale-based approach (that is based on a wavelet, not Fourier, transform). Quiz them yourself at SEG next month in Booth C-1644.

By the way, if you're going to SEG and have a smart phone, think about keeping up with the buzz by following some chatterboxes on Twitter—@SEGAnnualMtg, @ParadigmLtd, @ESG_Solutions, @kwinkunks, and @EvanBianco. You can create and account and just follow the conversation, but it's more fun to join in!

More microseismic power for Canada

A good friend of Agile's recently went to work at Spectraseis, a seismic processing firm specializing in microseismic and reservoir monitoring in general. They just opened a Calgary office, so good luck to them in what is a pretty tight market (there must be 50 seismic processing shops in Calgary). The company also announced $3.6M of investment from Credit Suisse, so they are clearly on a roll.

Next gen knowledge management

Repsol, the Spanish oil company and supercomputing powerhouse, is rolling out some new knowledge sharing technology from Autonomy, a fast-growing UK company. There are two pieces: IDOL, an enterprise search engine, and Virage, a rich media management system. 

Autonomy is an awesome company. How do we know this? They have Bayes' Theorem up in neon lights. Yeah, Bayes' Theorem. 

Get geo-referenced maps faster

Elsevier is one of the giants of 'old school' publishing, but also one of the more innovative ones (have you seen their graphical abstracts?). They have just introduced an awesome-looking search site for exploration geoscientists, called Geofacets. It's a set of power tools for finding maps and figures, already georeferenced and ready for a GIS. It even includes IHS's global basins map to search more geologically, and includes what you need to know about rights and permissions. If your company has ScienceDirect access, which they probably do, then you should have immediate access.

Previous news posts from Agile*

HD Frequency Decomposition is a trademark of ffA, IDOL and Virage are trademarks of Autonomy, Geofacets is a trademark of Elesevier. Bayes' Theorem image is CC-BY-SA licensed by its creator mattbuck. This regular news feature is for information only. We aren't connected with any of these organizations, and don't necessarily endorse their products or services.

Beyond the experts

I presented a poster at the 1IWRP, and it was certainly a change in tone from the technical rigor of most other talks. Since I had a good discussion at the break with a number of people, I thought I would make a video out of it. If you've got six minutes, you can check it out:

In the video I make reference to a few other topics we've touched on earlier on the blog:

• AgileWiki,
• doing science on a phone, 
• why we should embrace openness,
• our cheatsheets and,
• our mobile apps

I hope to be getting into making more videos soon, so let me know if you like the format, and if you have any suggestions. 

The last chat chart

The 1IWRP technical program was closed with a one-hour brainstorming session; an attempt to capture the main issues and ideas moving forward. This was great stuff, and I was invited to jot down the bombardment of shout-outs from the crowd.   

Admittedly, no list is fully comprehensive, and this flip chart is almost laughable in its ruggedness. However, I think it represents the diversity in this crowd and the relevant issues that people will be working on in the future. The main points were:

• Creating a common model and data sharing

• The future of digital rock physics

• Dealing with upscaling and scale dependant measurements

• The use of rock physics for improving sub-salt AVO analyses

• Strengthening the connection between rock physics and geomechanical applications

• Improvements for the next workshop at the University of South Hampton, in 2013.

I have scribed this into a more legible form, and put some expanded commentary on AgileWiki if you want to read more about these points. 

Do you disagree with anything on this list? Have we missed something?

More 1IWRP highlights

As I reported on Wednesday, I've been at 1IWRP, a workshop on rock physics in the petroleum industry. Topics ranged from lab core studies to 3D digital scanners, and from seismic attenuation and dispersion to shales and anisotropy. Rock physics truly crosses a lot of subject areas.

Here are a few of the many great talks that really stood out for me:

Mark Chapman from the University of Edinburgh, submitted a new formulation for frequency dependant AVO analysis. He suggested that if a proper rock physics model of the rock is described, frequency can be decomposed from seismic gathers for improved reservoir characterization. Some folks in the crowd warned that the utility of this work might be limited to select cases with a full band impedance change, but his method appears to be a step beyond the traditional AVO workflow.

Arthur Cheng from Halliburton talked about modeling techniques to estimate anisotropic parameters from borehole measurements. He descibed the state of the art in acoustic logging tools, and used a ray-tracing VSP forward model to show a significant smear of reflection points through an anisotropic earth layer. He touched on the importance of close interaction between service companies and end users, especially those working in complex environments. In particular: service companies have a good understanding of data precision and accuracy, but it's usually not adequately transfered to the interpreter.

Colin Sayers from Schlumberger presented several talks, but I really enjoyed what he had to say about sonic and seismic anisotropy and how it is relevant to characterizing shale gas reservoirs. Fracture propagation depends on the 3D stress state in the rock: hard to capture with a 1D earth model. He showed an example of how hydraulic fracture behaviour could be more accurately predicted by incorporating anisotropic stress dependant elastic properties. I hope this insight permeates throughout the engineering community. 

Rob Lander from Geocosm showed some fresh-out-of-the-oven simulations of coupled diagenesis and rock physics models for predicting reservoir properties away from wells. His company's workflow has a basis in petrography, integrating cathodluminescence microscopy and diagenetic modeling. Really inspiring and integrated stuff. I submit to you that this presentation would be equally enjoyed at a meeting of AAPG, SPE, SPWLA, SEG, or SCA — that's not something that you can say about every talk. 

← Every break heralded a new discussion. The delegates were very actively engaged. 

Today, I am going on a field trip to the Niobrara Shale Quarry. After four days indoors, I'm looking forward to getting outside and hammering some rocks! 

Digital rocks and accountability

There were three main sessions at the first day of the First International Workshop on Rock Physics, 1IWRP. Experimental methods, Digital rock physics, and Methods in rock physics, a softer, more philosophical session on perspectives in the lab and in the field. There have been several sessions of discussion too, occurring after every five presentations or so, which has been a refreshing addition to the program. I am looking for talks that will change the way we do things and two talks really stood out for me. 

Mark Knackstedt from Digitalcore in Australia, gave a visually stunning presentation on the state of the art in digital rock physics. You can browse Digitalcore's website and and view some of the animations that he showed. A few members of the crowd were skeptical about the nuances of characterizing microcracks and grain contacts near or below the resolution limits, as these tiny elements have a dominating role on a material's effective properties.  

In my opinion, in order to get beyond 3D visualizations, and the computational aspect of pixel counting, digital rock physicists need to integrate with petrophysicists to calibrate with logging tools. One exciting opportunity is deducing a link between laboratory and borehole-based NMR measurements for pore space and fluid characterization. 

In an inspired and slightly offbeat talk, Bill Murphy 3 from e4sciences challenged the community to make the profession better by increasing accountability. Being accountable means acknowledging what you know and what you don't know. He offered Atul Gawande's surgical writings as a model for all imperfect sciences. Instead of occupying a continuum from easy to hard, rock physics problems span a ternary space from simple to complicated to complex. Simple is something that can be described by a recipe or a definite measurement, complicated is like going to the moon, and complex is like raising a child, where there's an element of unpredictability. Part of our profession should be recognizing where our problems fall in this ternary space, and that should drive how we deal with these problems.

He also explained that ours is a science full of paradoxes:

• Taking more measurements means that we need to make more hypotheses, not fewer
• Ubiquitous uncertainty must be met with increased precision and rigor
• Acknowledging errors is essential for professional and scientific accountability

The next time you are working on a problem, why not estimate where it plots in this ternary space? It's likely to contain some combination of all three, and it might evolve as the project progresses. And ask your colleagues where they would place the same problem—it might surprise you. 

First-of-its-kind workshop

Next week I am going to the Colorado School of Mines to attend the First International Workshop on Rock Physics, 1IWRP. The name certainly is a bit curious: surely there must have been conferences in rock physics in the past, right? Yes, but the title points to the notion that this conference is the first time a meeting of petroleum rock physicists has happened independently of a technical society such as the EAGE or SEG. I don't see this as a specialist community revolting against these organizations, but it reflects an increase in attention and enthusiasm for the application of rock physics in the industry. Interestingly, this conference sprang directly from a discussion group on LinkedIn.

I will be presenting a poster entitled, Create a software culture beyond the experts, as part of the sesssion devoted to Transferring rock physics knowledge and technology to asset teams. It's a topic well-suited to this audience, because, even though rock physics is a key component in many areas of petroleum geoscience, it still remains partially obscure, or under-recognized as a take-it-or-leave-it niche. Mine will be a softer, community-building appeal for knowledge sharing beyond our distinguished specialists.

The technical agenda has been posted to the conference website. The conference has received solid corporate sponsorship; more than $40k according to the announcements. With 58 technical talks and 25 poster presentations over four days, plus two field trips, it is shaping up to be an intense week. A few talks that I am particularly interested in are:

• Frequency-dependent amplitude-versus-offset analysis, Mark Chapman
• Velocity Evolution during Controlled CaCO3 Precipitation and Dissolution, Ralf Weger, et al.
• Anisotropic static and dynamic moduli from a pair of shale plugs cut parallel and perpendicular to bedding, Douglas Miller & Richard Plumb
• Anisotropic permeability in fractured reservoirs from frequency-dependent seismic AVAZ data, Aamir Ali & Morten Jakobsen
• Use of sonic and seismic anisotropy to characterize resource shales, Colin Sayers

If anything in the conference catches your eye, drop me a line in the comments and I will do my best to capture notes on it. I will be reporting highlights throughout the week on the blog, so please be sure to check back and follow along. 

The photograph of the Colorado School of Mines campus is the work of Wikipedia user Cperko.

Update on 2011-08-22 14:36 by Evan Bianco

The poster is now available in the downloads section or by clicking on the thumbnail above. You can also watch a 6 minute video on my presentation in the post called, 'Beyond the experts'.