How to Check the Quality of your Seismic Data Conditioning in Hampson-Russell Software HRS9 Houston, Texas 2011
QC Data Conditioning This document guides you through the quality control check process used in conditioning their gathers. Mute process Trace mute Noise attenuation Band pass Filter Trim Statics Parabolic Radon Transform 2
QC Trace Mute 1. After we apply the mute to the data, we make a stack with the data that has been removed. To do this, we will subtract the outputted volume from the original data set and then stack the results. This shows if we have removed signal. We should not see noise after the mute is applied. Data Removed Before mute After mute 3
QC Trace Mute We need to determine if the mute process has removed too much or too little data. To check this, subtract the original volume from the muted to see if the correct amount of noise was removed. We will create a trace math script to do this. Before mute After mute 4
QC Trace Mute 2. To create the script we will specify a script name; select the original volume from the User Variables tab. 3. From the Constant and Operators tab, select subtraction. 4. From User Variables chose the muted volume and click OK. 5
QC Trace Mute 5. Click OK to run the process. We will get something like this 6
QC Trace Mute In order to prove that we are not removing signal from the data, we will stack the residual, after the trace math script. 6. Select Process > Seismic Processing > Stack > CDP Stack. 7. Chose the script residual volume as your input, name the output and fill in the remaining parameters as needed. 8. Click OK to run the process. 7
QC Trace Mute The following slides show three situations. Example 1 is a mute that is too conservative. Example 2 is a mute that is just right and Example 3 is a mute that is too strong. Example 1 Example 2 Example 3 8
QC Trace Mute This is the results of the applied mutes for the three situations. Notice how much noise is left in the data on the far offset after each mute. Example 1 Example 2 Example 3 9
QC Trace Mute Here you can see the data removed from each mute. Example 1 Example 2 Example 3 10
QC Trace Mute Below is the stack of the remaining data after the mute was applied. Now we can see the improvements in the data caused by removing the noise in the three situations. Example 1 Example 2 Example 3 Coherent events Coherent events Coherent events noise noise 11
QC Trace Mute The purpose of the trace mute is to remove noise without removing signal. Now we stack the data to appreciate how removing the noise has improved the data. Far Offsets 12
Bandpass Filter We can apply a band pass filter to remove noise. First, view the amplitude spectrum. 1. Select Process > Seismic Processing > Utilities > Amplitude Spectrum. 2. Select the trace rate, time window and offset range. Click OK. 13
Bandpass Filter 3. Determine what frequency you want saved and enter your filter values depending on your amplitude spectrum. 4. Click OK. 14
QC your Band Pass Filter Parameters The following slides will show different scenarios from applying a strong or a moderate Band Pass Filter. Key points to remember: Analyze the amplitude spectrum of your data to define the objective of the study. You may have reset the time window to your target zone or to the entire data set. We will show an example of how to QC the output data in order to visually determine if any signal was removed in the process. 15
Bandpass Filter 1 st case. Low and high cut values of, 8 and 90 are applied respectively, and low and high pass values of 10 and 25 Hz. We are affecting the frequencies between 8 and 10 Hz and 25 and 90, and we let pass everything between 10 and 25 Hz. Un-filtered filtered As you can see in this example to the left, noise has been removed. We have removed much noise inside the red circle and boxes. Next we will review if we are also removing signal from the data. 16
Bandpass Filter Next we will subtract the original data from the filtered data. 1. Select Process > Utilities > Trace Math, to open a dialog and select the input volumes. 2. Next, define the trace range, time window and offset range. 3. To create our script, click <Create New> and Edit. 17
Bandpass Filter 4. Under the User Variables tab, select the original volume. 5. Next select the Constant & Operators tab and select Subtraction. 6. Click OK to run the script. 18
Bandpass Filter This is the result of the operation, as we can see we have removed lots of signal (coherent events) from our data. 19
Bandpass Filter 2 nd case. In the first case the band pass filter was too strong and removed signal. Now we will apply low and high cut values of 2 and 60 respectively, and low and high pass values of 4 and 40 Hz. Affecting the frequencies between 2 and 4 and 40 and 60 Hz, we let pass everything between 4 and 40 Hz. In this exercise we can notice some differences: less noise. Next lets review if we removed any valuable signal. 20
Bandpass Filter For the second case, you can see coherent noise was not removed from the dataset. The more conservative band pass of 2-4-40-60 removed the desired noise and not the signal. 21
Bandpass Filter We recommend selecting a few inlines and crosslines and testing different parameters on the data set, depending on its characteristics. After you are comfortable with the results of your testing, apply those parameters to the entire data. This is a good practice for finding the best result and saving time in the data conditioning process. 22
QC Trim Statics The Trim Statics process can fix migration move-out problems on pre-stack data. It attempts to determine an optimal shift to apply to each trace in a gather. To start the Trim Static process, select Processes > Seismic Processing > Trim Statics. 1. Select the input data, name the output data and specify the time and trace range for the process to be applied. 2. Define the time window where you will be applying the process. The maximum allowed shift is the value that will be applied to each trace. 23
QC Trim Statics The next three slides show examples of the Trim Static process applied to a data set. In each case we are changing the Maximum allowed shift value. Example 1: In this case we assigned a maximum shift value of 5 for the shift. Notice the difference between the before and after data sets. The horizons is better aligned with the trough of each trace in the gather. Before After 24
QC Trim Statics Example 2: In this case we assigned a maximum shift value of 10 for the shift. In this example,the horizon is well aligned to the trough but we are moving some of the overlying events as well. 25
QC Trim Statics Example 3: In this case we assigned a maximum shift value of 25 for the shift. Applying a shift value of 25 we are braking the data. As we can se inside the red circle we have displaced considerably some traces. 26
QC Trim Statics The trim statics process can correct minor migration move-out problems on pre-stack data. But also we have to make we do not apply an excessive shift that will damage the data. 27
Parabolic Radon Transform The Radon Transform is a tool used to perform multiple elimination and noise suppression. 1. To apply the filter select Process > Seismic Processing > Filter > Parabolic Radon Transform. Click OK. 28
Parabolic Radon Transform The residual volume is generated automatically after we run the process. Before Radon (Input data) After Radon (Output data) Residual (noise removed) 29
Parabolic Radon Transform The Radon Filter is a very effective method if we want to remove multiples. This tool can help to improve the data set in order to get a better input for the AVO analysis. 30
Conclusions Throughout this document we have shown the importance of having the best data quality as an input for the AVO and Inversion Analysis. The best practices and application of the tools to achieve this goal will have events horizontally aligned, amplitudes preserved, the signal to noise ratio enhanced and multiples attenuated, etc., to obtain the best AVO response possible. 31
Support Offices 32