< Articles & News

4D Seismic: Why Different Is Different

July 29, 2015   |   11:38 AM
Big Data Analytics
Teradata Articles

I met an ex-colleague on a flight recently, and as we were catching up, the subject of data management for 4D seismic data came up.  My ex-colleague said that although the Oil & Gas industry have been talking about 4D seismic (where repeat images of the subsurface are combined to see how reservoir fluids have moved) for years, things aren’t that different to warrant  a new approach.

Not that different? Really? That reminded me of a quote I’d read, “Different stuff is different.”  Dave Dellanave, a trainer, coined that phrase (although he used a different “s” word). When Dave says “Different s…. is different”, he is saying that exercises, such as front squat, back squat, KB goblet squat, that appear to target major muscle groups in the same way, are actually different.  What’s different is an individual’s reaction to each exercise, and small changes to how an exercise is performed can be the difference between not being able to do an exercise at all, and pain-free execution.

pipe
Dave lays down this challenge: “the next time that someone, or your own ideology, places an artificial limitation on you based on the idea that everything is the same, stop and ask more questions.  With better questions comes more progress and better results”.

So, back to 4D seismic; yes, it’s still seismic.  But it’s different. Here’s how:

Repeatability

Position differences from survey to survey are one of the biggest sources of uncertainty in 4D interpretation.  With trenched ocean bottom cables, the receiver positions are fixed.  The source positions are still dependent on the ability of the source vessel pilot to hit the correct spot – but that is so much easier in a small source vessel than in a traditional acquisition vessel with 15km of cables hanging out the back.  And you know what this difference in repeatability means?  It means that a 4D effect – a production effect – can be seen even on the raw data, before processing.  That’s a huge difference – the ability to immediately inform a decision on injection regimes and drilling targets.  Minimising bypassed oil by right-time monitoring increases the percentage of oil in place that can be recovered – easily worth a few extra million barrels.

Acquisition Choices

The oil company can now choose when, and how to acquire data.  In the Permanent Reservoir Management (PRM) world, the oil company can choose everything – ocean bottom cables, layout, trenching, source vessel, source air gun configuration, acquisition window in terms of environmental aspects like seasons, and acceptable weather window.  They choose what data comes onshore and when, and how much they interact during the survey acquisition.   Shot data is received daily during acquisition, meaning the seismic data itself can be checked for quality while the vessel is still out.  This in turn means vessel time can be managed effectively – minimising acquisition costs while making best use of limited time windows to gather the most important data at the desired quality

Data between seismic shoots

Consider the additional passive seismic data that may be gathered from the array when no active shooting is taking place.  You’ve no doubt heard the phrase “one man’s noise is another man’s signal”. But in discussions around the use of passive data, I have even heard “one man’s noise is another man’s (seismic) source”.  Is it possible to use platform noise as a virtual source for reflection seismology?  Even if all it can illuminate is the overburden, just consider the impact this will have on the safety of operations.  Over-pressurised zones will be visible before anyone or anything is put at risk.

Try new things

How do we best put these differences to work in the acquisition, processing and interpretation of such data?  One way is to just try things out – but learn quickly and focus on what works.

I wish I could lay out a perfect step-by-step workflow that meets all eventualities.  But truth is, the technology is ever changing – the speed at which we can stream this data back onshore; the additional sensor data we can employ to identify remaining sources of “noise”; the data management options that underpin our ability for storing, manipulating and analysing these vast datasets.  It feels like every day something new becomes possible.

What I do know is that if we allow our historical assumptions about traditional seismic acquisition to put artificial limitations on what is possible with PRM seismic data, we are putting the brakes on our own ability to progress.

We have the chance to improve the way we acquire, process and interpret results – all in a more timely manner – and therefore improve our ability to better inform operational decisions like injection regimes or plans for infill wells in a truly operational timescale.

Different stuff is different.  Let’s take advantage of that.