Sunday, June 6, 2010

Down to the Core

I've had the rare priviledge in the course of my project work to be able to truly ground-truth interpretations from seismic data through comparisons with core samples. For onshore projects that may not seem like so much of a priviledge as the norm, but in deepwater environments, the combination of data is much less common, particularly in the early exploration stages of which I'm involved.

I have my interpretation of stratigraphic intervals and the kinds of processes that may be at work based on the regional setting, observed site-specific structures, and by characteristics revealed through various seismic attributes. Then I have core data to add to that interpretation, refining the model of what we think we're seeing. The core data includes various geotechnical properties measured through standard lab tests, such as shear strength, water content, atterberg limits, etc. These lab tests, standard and advanced, are destructive to the sample. For my project we also have a radiocarbon and paleontological dating program planned, to try and understand the depositional rates and timing of events. Selecting the sample locations, considering both time and space (where in the core and from which core), is a critical step towards putting the geologic story back together and involves close inspection of the interpreted seismic data, correlation of geotechnical properties between the physical sample and the geophysical record, and understanding what an age at a particular portion of the core can and can't tell you. Digging out samples for age analysis is obviously destructive to the core also. If you have plans to undergo geotechnical testing, you have to make sure you aren't using material you also want to get an age from. It definitely takes some thinking and planning out.

There are some things that can be done to analyze a core without destroying the sample. A multi-sensor core logger, or MSCL for short, is a tool that allows for geophysical measurements of an unsplit, cylindrical core. Working in the marine environment, we've used MSCL for sub-samples of box cores, as well as gravity cores, piston cores, and jumbo piston cores. Compressional wave (P-wave) transducers allow for velocity measurements of the sediments in the core. Gamma ray attenuation is measured through the core and used to derive sediment density. The labs I've worked with can provide various levels of detail for logging, such as 4 cm samples or higher resolution 2 cm sampling. If the core is later split, or material was sampled from the core ends during acquisition, the densities may be calibrated with the lab measurements taken directly. Velocities and densities also play important roles in how the seismic data behaves, so with the MSCL data, the measured geotechnical properties, and the seismic section you can start to make some correlations. Accurate and precise calibration of the MSCL for your samples and close QA/QC of the data are key to getting the most useful data.

Ultimately, if you want to understand the small-scale features, you have split open the core and take a look at the sediment types and preserved structures. This, too, is a destructive process. Once the core is split, you can't put it back together again.

One of the best things you can do with your split core is to photograph it. As with all geologic-intended photographs you will need something in the photograph to reference for scale. Some people place a measuring tape or ruler along the core edge, others mark directly on the core liner in regularly spaced intervals. In the process of describing your core, you are likely to scrape, smudge, cut, poke, and sub-sample the sediments. You might even taste the sample (Yes, a lot of geologists will taste the dirt). In addition, an open core will dessicate and oxidize much quicker than an un-opened core (even in controlled storage environments), which means color, water content, and other geotechnical characteristics can change. While a core photograph cannot capture the geotechnical properties, it can preserve color and structural relationships to which you can refer to long after your sample has been picked through or discarded. A photograph is an important piece of your compiled data set that may prove useful even beyond its original intended use for your project.

It has been my experience that most labs and researchers rush through the steps of capturing those photos. Some are done with basic point-and-shoot cameras, with little attention to lighting, resolution, or focus. Even photos done meticulously with a high-resolution camera suffer from amateur photographers in control of the picture-taking. My recent project data included high-resolution core photographs taken every 10 cm down the length of the core. A lab technician cropped each image and stitched them together to re-create each core section. The files were 100+MB bitmap images exported out of Adobe Photoshop. The resolution on the images was undeniably good, but most sections were coarsely scraped creating ridges and shadows on the split section, the high water content and lighting caused problematic reflections, and poor light distribution made many of the sections dark and the variable colors difficult to distinguish.

When preparing to photograph your core, there are several things to consider. First, the cores are best viewed in natural light. Flourescent lights tend to add hints of blues to the photo. Incandescent lights tend to add hints of yellow to the photo. Neither are desireable if you want to look at the natural colors. Adjusting the white balance compensates for your light source and allows you to mimic natural sunlight when done correctly. Just a note: Color is not always the best geologic tool in the box, as it can be decieving at times. Be careful with what assumptions you make based solely on color.

Second, you want to look at a fresh surface, which suggests you may need to scrape the sediments exposed in your split core. Scraping the core surface is a delicate procedure. You want to scrape across the core, not down the core. Scraping down the core could contaminate your sample, mixing younger and older sediments as you drag the spatula down the core. Scraping across minimizes that potential, assuming you clean the spatula between each stroke. Many people who scrape a core, particularly if the sediments are soft, mistake the process to be similar to plastering and instead of cleaning the surface to reveal the sediments and structures, they smear the sediments, distorting or disguising the structures. It takes a steady hand and the proper spatula angle to really get it right. Circular polarizing filters can reduce reflectivity of the core surface. Multiple strobes at different angles can minimize or eliminate shadows created by uneven surfaces.

Third, things like focus, scale, and resolution are extremely important. You want to make sure each photo is taken from the same distance with the same focal point. This is true particularly if you plan to stitch the images together. You want to minimize the distortion caused by the camera's perspective view by selecting the right lens, zoom level, and core position. A mounted camera is an absolute necessity, as there is no way anyone can hold the camera still enough or at the appropriate angle for every snapshot. Carefully plan the lighting for the photos and minimize changes in the settings as you are capturing your photos. If you are adjusting things throughout your photo shoot, images you capture at the beginning won't match up well with those you capture at the end. Take a lot of test shots and once you have the settings worked out, then capture your suite of images to preserve your entire data set. When correlating your findings are important for you, you want to make sure you are comparing apples to apples.

So, a few recommendations for those pursuing a core sampling program:
1. Consider MSCL data for your cores, particularly if you have other geophysical data you may be trying to ground-truth.
2. When your core is split, carefully prepare a freshly exposed surface for viewing by meticulous scraping done properly.
3. Train the photographer (or hire a professional) to use the camera and other equipment to its full potential.

You might think these are minor concerns, or that any part-time tech could do this work. But if you were to sit in the core viewing sessions I've had with a room full of geologists and engineers, with nothing but the photos and the seismic sections you would realize how drastically different the interpretations can be based on what you see in the photo. If you never saw the core in person, you might think the reflective streak was a lamination and it's cross-cutting what looks to be other laminations. You may not realize it was caused by irregular scraping and poor lighting conditions. It can make a big difference in your interpretation.

Obviously, you need to take your individual program into account and make decisions about your sampling and testing program that will be suitable for the purposes you need. Your methodology for a single core versus what you may do for hundreds of cores could be drastically different. One thing I believe, though, is that it is worth your effort to do it right and do it well the first time.

If you don't have time to do it right, you must have time to do it again. Only with a core, you usually only get one try.

No comments:

Post a Comment