Presented on. Mehul Supawala Marine Energy Sources Product Champion, WesternGeco

Transcription

1 Presented on Marine seismic acquisition and its potential impact on marine life has been a widely discussed topic and of interest to many. As scientific knowledge improves and operational criteria evolve, seismic operations must take all feasible measures to reduce acoustic emissions, especially unwanted energy outside of the effective bandwidth which is no use to seismic imaging. Jointly developed by and Teledyne Bolt, the esource bandwidth-controlled seismic source technology reduces the environmental footprint of marine acquisition through a sustainable and flexible seismic source solution. 1

2 The esource bandwidth-controlled seismic source technology is the only source on the market that has been developed primarily to reduce environmental footprint. The design is based on the guiding principle of only emit the energy that is required for seismic imaging. It has been engineered to suppress high-frequency energy outside the effective marine seismic bandwidth, while preserving the lower-frequency components necessary for seismic imaging. The esource technology is available on the open market through Teledyne Bolt and accessible by operating companies. This joint development combines the marine acoustics expertise of with the multidecade experience Teledyne Bolt has in developing and manufacturing reliable marine sources. 2

3 There are two credible research studies on marine mammals, which are widely used in our industry. These studies are titled Aquatic Mammals Marine Mammal Noise Exposure Criteria: Initial Scientific Recommendations by Brandon L. Southall and Draft Guidance for Assessing the Effects of Anthropogenic Sound on Marine Mammals by the National Oceanic and Atmospheric Administration (NOAA). Government and regulatory bodies have used these studies to draft operational guidelines for marine seismic acquisition. The two key metrics that are commonly used are peak pressure (PP) and sound exposure level (SEL). PP is the maximum received pressure and SEL is the total cumulative energy received over a period of 24 hours. 3

4 Studies have shown that marine mammals are sensitive to sound in different frequency ranges. Southall has classified cetaceans in three different categories based on auditory bandwidth low-frequency (LF), mid-frequency (MF), and high-frequency (HF) cetaceans. It is important to note that the majority of cetacean species have auditory bandwidth outside the typical wavelengths that are used to gather seismic information. 4

5 Each source point in a seismic survey typically comprises 3 to 5 subarrays (left) and each subarray contains 8 to 10 seismic source units (right). Based on this configuration, anywhere between 20 to 40 source units release energy every 8 to 12 seconds. In practice, the configuration, timing, number, and size of the sources depend on a variety of factors, including geophysical objectives and operational considerations. 5

6 The left graph plots the energy in all directions for a 4,400-in 3 source (indicated in red) and the high-frequency mammal sensitivity given by NOAA 2013 criteria (indicated in blue). The sensitivity shows that this cetacean group is more sensitive toward higher frequencies, meaning they can take less energy at those frequencies. The right graph is a product of the total source energy and mammal sensitivity giving the weighted power spectrum. We consider that the useful frequency for imaging rapidly declines beyond 100 Hz. It is therefore apparent that standard sources release a lot more energy into the environment that is not used for seismic imaging. 6

7 Each element contains two physical chambers and a piston in the middle. Compressed air fills both chambers and, at the point of trigger, a volume imbalance is created that pushes the piston forward, rapidly releasing the compressed air through the ports. This is what creates the acoustic pulse. 7

8 Here we plot the output of a 290-in 3 standard source, with all energy in the 3- to 25,000-Hz range. The 500-ms analysis window encompasses the total energy released by the source for the entire acoustic pulse (including the bubble). Note the peak pressure of 4.8 bar.m. The second plot shows 80 Hz, plotting only 3- to 80-Hz bandwidth. This frequency range is more typical for seismic imaging at the reservoir level. The shape of the wavelet is similar to the full-bandwidth version, especially for the bubble train. The final plot shows an equivalent 80-Hz low-cut filter. It confirms our observation that the bubble train is well described by the lower frequencies and that the high-frequency energy is confined to the initial pulse, particularly for the rising edge of the pulse. This shows that if we want to control the high-frequency energy output from a source, then we need to control the initial pulse, which means controlling the rate at which the compressed air is released from the source. 8

9 What if we could control the source to limit unwanted high-frequency energy that is released into the environment? And, what if we were able to tune the bandwidth based on geophysical objectives and the type of mammal species in the area? This is the concept behind esource technology, which can be configured to output configuration A on survey 1 and then output configuration C on survey 2. 9

10 Here is a comparison between a standard source (left) and esource technology (right). The esource technology regulates the rate at which air is released from the source through precise control of the piston speed and a novel port shape. 10

11 0 ms Standard source esource technology This time-lapse shows a comparison of a standard 1900LL source (left) versus the esource technology (right). Both of the sources are triggered at the same time for a total duration of 20 ms. Careful inspection shows that the esource technology releases the air at a much slower rate compared with the standard source. This slower rate of releasing the air results in a reduced peak and less high-frequency energy. The difference is evident when comparing the graphs of pressure versus time. Note that the size of the bubble (the green ball) toward the end of the pulse is the same for both sources. This is the point at which most of the low-frequency energy is released. Several configurations of the esource technology are available termed configuration A, B, and C. The choice of each configuration provides different control over the energy release at high frequencies. 5 ms 10 ms 15 ms 20 ms 11

12 This graph compares the frequency amplitude spectra from the three different esource technology configurations compared with the standard source. The esource technology configuration A is the least aggressive and only starts to significantly attenuate energy beyond 100 Hz. Configuration C is the most aggressive and starts attenuating at closer to 40 Hz. [Median amplitude spectra: Single source: 150-in 3 ] [2,000 psi, 6-m source depth, near-field hydrophone (NFH) below source] [Vertical db scale in upa.m/hz] 12

13 The red curve shows the average power spectrum with high-frequency cetacean weighting applied for a standard source array firing every 10 seconds. The area under this curve is proportional to the SEL. It is very clear that this area drops significantly when using the esource technology configuration A, with further reductions using configurations B and C. 13

14 We can also estimate the potential exposure for marine mammals in proximity to the different seismic sources. Using guidelines from the NOAA 2013 report, the injury zone is the area around the source within which the mammal has the potential to get injured. There are several ways to model this, which also depends on the type of cetacean class. In this study we looked at high-frequency cetaceans and assumed both the source and animal remain static (relative to each other) for 1 hour, with the same 4410-in 3 source firing every 10 seconds. A simple ocean propagation model is used. 14

15 The frames encompass a distance of 1 km in each direction away from the source location, as indicated by the small blue dot at the sea surface. The dashed blue lines show a 200-m lateral reference distance from the source. The red surface represents the potential injury zones for a mammal based on the SEL criteria from NOAA. Any mammal within the red area has the potential for injury. In addition, a green surface is also visible. This is the peak pressure (PP), but for this model, it falls within the SEL zone. It is immediately clear that the esource technology configuration A has significantly reduced the red surface of the SEL zone. In terms of lateral distance, the surface now fits almost entirely within the 200-m reference. Configurations B and C shrink the SEL surface even further, such that the PP surface becomes the key criteria. 15

16 A side-by-side field trial was conducted with esource technology (configuration A) and standard sources. A total of 18 source units deployed in 3 subarrays, shooting flip-flop at 25-m interval. The source array volume for both the esource technology and standard source was kept exactly the same 4,335 in 3. 16

17 From an operational point of view, the rigging of esource technology to the subarrays went smoothly and no issues were identified during deployment and recovery from the vessel. For this field trial, only 4.5 hours was required to replace standard units with esource technology. A total of 2,000 shot points were covered in the tests, and there were no failures. 17

18 Pictured here is the combined esource technology (left) and the standard source array (right), configured and ready for use on the Amazon Conqueror as part of field trials. 18

19 Shown above are far-field signature estimates computed using near-field hydrophone (NFH) and positioning information. The far-field signature for both is nearly identical, with the esource technology having a slightly lower peak and gradual rise. The amplitude spectra show very similar responses and reduce high-frequency content beyond 100 Hz, as expected. 19

20 Pictured here are premigration stacks from two shots one using a standard source and one using esource technology. The stacks look identical for all practical purposes. 20

21 Pictured here are the stacks after migration. Again, the images are very identical. All events and features of interest are clearly evident in both images. 21

22 Pictured here are a couple more examples after final processing using bandwidth extension. We anticipated the need to use a matching filter to match the target spectrum, but we discovered that a global scalar is all that is needed to match the target spectrum. 22

23 Using the same dataset, pictured here is a linear frequency scale (left) and a log scale (right) focusing on the low frequencies. Compared with the standard source, the esource technology spectrum shows a slightly higher signal amplitude at low frequency (below 4Hz) and slightly lower amplitude for high frequency. Both signal-to-noise ratio (S/N) estimates for the standard source are consistently higher; hence the S/N separation is similar for both the esource technology and standard source. This leads to the expectation that, although overall output is slightly different (as expected), after subsequent standard processing, we can shape both sources to the same target spectrum and maintain equivalent S/N characteristics. 23

24 After migration, the S/N has been improved as expected. The difference in amplitude between the two datasets has been proportionally reduced. 24

25 Finally, here is the spectra after bandwidth extension, applying a 1.07-amplitude scalar to esource technology data. Signal spectrum overlay now shows a very high level of similarity for both. There is no need to shape the sources to the same target spectrum. The noise spectra are similar, but the standard source suggests slightly higher amplitude for the noise. 25

26 In conclusion, esource technology performs better than the industry-standard conventional seismic source operationally, the source is practical data analysis shows equivalent results across seismic bandwidth as of February 2017, esource technology is currently in action on a 11,500-km 2 multiclient survey in Australia. 26

27 Presented on For questions and answers, please listen to the recorded webinar at slb.com/esource 27