Seismic Reflections

Slide 1

• This unit goes into some more detail on what causes a seismic reflection and the characteristics of the seismic response
• In other words, what do the “peaks” and “troughs” on a seismic section mean
• No need to explain this figure – will be covered in the lecture

Slide 2

• The ideal seismic response would give us information about the stratigraphy in the subsurface at the same scale as an outcrop
• Here the beds are about a foot thick – the ideal seismic line would show us this level of detail
• Unfortunately, we do not live in an ideal world
• Seismic Reflections do not allow us to “resolve” (be able to distinguish) strata at this scale

Slide 3

• There is a hierarchy of layering within sedimentary rocks – the strata
• This hierarchy (or scale) is shown on the left
• The smallest scale of layering is the lamina
• Two or more related lamina form a slightly thicker stratal unit – a lamina set
• Several lamina sets form beds
• Several beds stack to form bed sets
• ETC.

• Seismic data can not image (resolve) beds or bed sets, at least not normally
• However, they are able to image parasequences, parasequence sets, and larger-scale stratal units
• So seismic data is limited in imaging finer-scale stratal units
• However, the advantage of seismic data is the areal coverage it provides
• For many sedimentary basins, we have 2D or 3D seismic data covering the entire extent (area) of the basin

Slide 4

• The amount of resolution of the seismic data – the thickness of stratal units that can be distinguished – varies by the shape of the seismic pulse that was used to acquire the data and the velocity of the rocks
• Since velocities tend to increase with depth:
• We can resolve thinner stratal units at shallow depths (e.g. 10 meters) than we can at intermediate depths (e.g. 25 meters) and
• We can resolve finer stratal units at intermediate depths (e.g. 25 meters) than we can at great depths (e.g. 40 meters)
• As shown on this slide, the seismic tends to “integrate” or average the layering at a scale of 10s of meters
• The seismic response can tell us that the upper part of this outcrop is predominantly sand, while
• The lower part of the outcrop is predominantly shale
• There are finer-scale layers in the outcrop (beds and bedsets), but we would not be able to distinguish these with seismic data

Slide 5

• Here we have some of the most common terms related to seismic data
• The white “sine wave” is a simple wavelet – the shape of the acoustic wave that travels down through the earth and is reflected back up to receivers on the surface
• The wavelet consists of movement that is part compression (positive values as recorder by sensors on the surface, i.e., receivers) and part rarefaction (negative values)
• Amplitude (A) is a measure of how big the wavelet is – the magnitude of the excursion to the right of zero (compression = positive ) or to the left of zero (rarefaction = negative)
• Lambda (λ) is the wavelength of the wavelet – its length in feet or meters
• The Period (P) is the time for the wavelet to travel one wavelength
• Pulse Duration (Dp) is the time that it takes for the wavelet to pass a particular reference point
• The next slide has a few simple equations that relate some of these parameters

Slide 6

• Equation 1 tells us that the Period is equal to 1/Frequency
• Equation 2 tells us that the Wavelenght is equal to the Velocity times the Period or, using equation 1, the Wavelength equals the Velocity divided by the Frequency
• Equation 3 tells us that the distance (or the depth) is equal to the velocity times the time divided by 2
• Why the division by 2?
• It is because the acoustic wave travels the distance twice – once down and once up

Slide 7

• In review, the essence of Seismic Reflections is that
• We generate energy at the surface (e.g., we set off a charge of dynamite)
• The energy travels down through the earth
• At a boundary between one rock unit and another, there is a change in either the velocity of the rocks or the densities of the rocks, or both
• We represent the acoustic properties of a rock layer by a parameter called impedance
• Impedance = velocity times density ( I = V * ρ)
• Where there is a change in impedance (e.g., top of the yellow layer), a fraction of the energy “bounces” or is reflected
• Most of the energy continues down (is transmitted)
• At the next change in impedance (top of the brown layer) some of the energy “bounces” or is reflected

• Let’s say that the acoustic energy corresponds to a compression (positive numbers) followed by a rarefaction (negative numbers)
• In this case:
• At a boundary where the impedance increases (lower layer has a higher impedance than the upper layer) the reflected energy will be a compression followed by a rarefaction – on the seismic section a black peak followed
by a white trough
• If there is a decrease in impedance at a boundary, the reflected energy will be a rarefaction followed by a compression – on the seismic section a white trough followed by a black peak

• On this slide there is an increase in impedance at both boundaries – hence both events on the seismic trace are a black peak followed by a white trough
• On slide 1 there is an example on the right where there are 2 boundaries with an increase in impedance (between layers 1 and 2 and also between layers 3 and 4) and one boundary where there is a decrease in impedance (between layers 2 and 3)

Slide 8

• This is a simplification of the previous display (slide)
• At a certain location we have various layers with different impedances
• We can calculate the impedance of each layer by multiplying the velocity by the density
• On the far left, we show the impedance as a log curve
• The amount of energy that is reflected is a function of the magnitude of the impedance change across a boundary, a small change in impedance results in a small amount of reflected energy; a large change in impedance results in a larger amount of reflected energy
• We can calculate a parameter called the Reflection Coefficient (RC) using a formula that is given in Exercise 6a, which we will do in a few minutes
• An increase in impedance results in a positive RC
• A decrease in impedance results in a negative RC
• We display the RCs as a log of spikes where
• Positive RCs are plotted to the right of zero
• Negative RCs are plotted to the left of zero, and
• The length of the spike is proportional to the value of the RC (small spike = small change in impedance; large spike = large change in impedance)
• The shallowest spike on the slide indicates a positive RC (to the right of zero) of a moderate change in impedance (a bigger change in impedance at the boundary between layers 1 and 2 then between layers 2 and 3; but not as big a change as between layers 4 and 5
• If we know or an can assume the shape of the acoustic pulse (waveform)...
• Then we can use a mathematical process called convolution to model the seismic response for each of the boundaries individually
• The actual seismic trace is the sum total of all the individual responses!
• As we will discuss further, there can be constructive or destructive interference between the individual responses, something that complicates the life of a seismic interpreter!

Slide 9

• If the frequency content (Bandwidth) is very large, then the pulse approaches a spike and we can resolve fine-scale stratigraphy
• This ideal pulse goes back to slide 2
• Unfortunately, the frequency of the pulses we are able to generate are limited, typically from about 10 to 50 Hz (BW = 50 – 10 = 40)
• Thus our ability to resolve thin beds on seismic data is controlled by the limited bandwidth of our pulse
• A high-resolution survey would have pulse frequencies from about 5 to 60 Hz, or a bandwidth of 55 – much better than 40

Slide 10

• Let’s consider the pulse for a few minutes
• There are two end-member types of pulses
• The first end-member is a minimum phase pulse
• This is the type of pulse that you would get from an explosion or an earthquake
• There is no particle motion before the explosion occurs
• Immediately after the explosion, particle motion will buid to a compressional maximum, then decrease, build to a rarefactional maximum (most negative value) and then go back to zero
• Minimum phase pulses are:
• Causal (real – no motion before wave arrives)
• Front loaded
• The peak arrival time is frequency dependent
• The RC is at the first displacement; maximum displacement (peak or trough) is delayed by ¼ λ

Slide 11

• The second end-member type of pulse is called zero phase
• The shape of the pulse relative to the RC is shown on the slide, a zero phase pulse:
• Is not Causal (not real, since there is motion before the wave arrives)
• Is symmetric about the RC
• The peak arrival time is not frequency dependent
• It has the maximum peak-to-side lobe ratio
• The RC is at the maximum displacement (peak or trough)

Slide 12

• Next, we will explain seismic polarity – i.e., the sign convention
• SEG stands for the Society of Exploration Geophysics
• They have set an industry standard for the definition of polarity for both minimum phase and zero phase pulses
• As this slide shows, for a minimum phase pulse:
• For a positive RC (increase in impedance), the number recorded on the tape should be negative, and
• The first motion should be displayed as a trough
• If a minimum phase dataset is said to be SEG reverse polarity, that would mean for a positivve RC the first motion would be displayed as a peak

Slide 13

• As this slide shows, for a zero phase pulse:
• For a positive RC (increase in impedance), the number recorded on the tape should be positive, and
• The first motion centered on the RC should be displayed as a peak
• If a zero phase dataset is said to be SEG reverse polarity, that would mean for a positive RC the motion centered on the RC would be displayed as a trough

Slide 14

• Let’s review what causes a seismic reflection
• A seismic reflection is generated at any interface between rock layers with different acoustic properties
• These acoustic properties are the velocity and the density of the rock
• Geophysicists use the term impedance (I), which equals velocity * density

• If the change in impedance across a boundary is small, the amount of reflected energy is small
• If the change in impedance across a boundary is large, the amount of reflected energy is large

Slide 15

• OK we are ready for 2 exercises (Exercise 6a and 6b)
• In 6a we will give you the equation for calculating a reflection coefficient and ask you to use this equation to calculate two RCs
• In 6b you will calculate the frequency and wavelength for two portions of a seismic line

Slide 16

• Here is the first part of Exercise 6a
• This slide has the acoustic properties for rocks above and below an interface – in this case shale on top of sand
• Let the students perform the calculation

Slide 17

• This is the answer for the first part of Exercise 6a

Slide 18

• Here is the second part of Exercise 6a
• This slide has the acoustic properties for rocks above and below an interface – in this case shale on top of carbonates
• Let the students perform the calculation

Slide 19

• This is the answer for the second part of Exercise 6a

Slide 20

• For Exercise 6b, we will use this seismic section
• There are blow-ups of 2 windows
• We want you to calculate the frequency and wavelength of the seimic in each window
• The relevant equations are in the upper right
• We get the apparent (observed) frequency for each window by counting the number of cycles (1 cycle = a black followed by a white) over a certain time interval (e.g., how many black-white couplets occur over 0.1 seconds)
• We have an empirical formula to get the dominant frequency given the apparent frequency
• Once we have the dominant frequency, we can calculate the wavelength (λ) using the third equation

• Give the students a little introduction to the exercise, and then some time to calculate

Slide 21

• We will talk about this in greater detail in Unit 11, but seismic reflections tend to parallel stratal surfaces
• We can use reflection terminations to identify and mark unconformities
• Changes in the characteristics of a reflection (e.g., amplitude, frequency, continuity) indicate changes in depositional facies

Slide 22

• Why do reflections parallel stratal surfaces?
• Recall that reflections are generated where there is a change in acoustic properties
• Either the velocity of the rocks change
• Or the densities of the rocks change
• Or both
• Let’s look at a thick outcrop in West Texas – 1200 ft or 365 meters in relief
• From bottom to top, there are 4 formations
• The Pipeline Shale – guess what the lithology is?
• The Lower Brushy Canyon
• The Middle Brushy Canyon
• The Upper Brushy Canyon
• Each member of the Brushy Canyon consists of shale with silt and sand layers – sand content increases Lower to Middle, and Middle to Upper
• Consider where there would be sharp changes in impedance
• Note some white, ledge-forming layers (e.g., just below the Middle Brushy Canyon label)
• This is a relatively sand-rich layer
• We could walk out this layer for several miles
• If we sampled this layer, say every ¼ mile, we would find that
• the first sample might be 75% sand, the next 73%, then 72%, 70%, 71%, 68%, 65%, 66%, 64%, 62%, 60%, etc.
• The point is that the sand content is changing, and also the acoustic properties, but these changes are very gradational
• There are no sharp physical surfaces laterally across which the acoustic properties change significantly
• Now consider if someone repelled down the cliff and took sediment samples every 2 meters
• The first sample might be a shale, next a shale, then a silt, a sand, a shale, a shale, a sand, a shale, a silt, a shale, a sand, a silt, etc.
• The point is that there would be more abrupt changes in acoustic properties vertically
• Some significant changes would occur at the larger-scale stratal packages, i.e. at boundaries between parasequences, and between parasequence sets, and between sequences
• Thus it is reasonable that the reflections we see on seismic sections are generated at parasequence boundaries, and at parasequence set boundaries, and at sequence boundaries
• You may be thinking:
• Is there NOT a seismic response as we pass from one environment of deposition (EOD) to another EOD?
• YES there is
• A reflection will follow, for example, a boundary between one parasequence and the next parasequence
• The characteristics (attributes) of the reflection (say a peak) will change as the sedimentary facies above and below the parasequence boundary changes
• For example:
• A shale on top of fluvial rocks might result in a moderate reflection amplitude,
• Changing to a high amplitude reflection where there is shale on top of nearshore sands,
• Changing to moderate amplitude where there is shale on top of offshore silts
• Changing to low amplitude where there is shale on top of offshore shale

Slide 23

• A word of caution...
• There are other seismic reflections out there that may not be stratigraphic in origin
• For example:
• Fluid Contacts
• Fault Planes
• Multiples
• Others

Slide 24

• It is time for another exercise
• Exercise 6c
• The next slide is a brief introduction

Slide 25

• You are going to start to make a synthetic (modeled) seismic trace
• You will use a very simple pulse – a sine wave – which is a minimum phase pulse (on left)
• And you will have 3 reflection coefficients
• +0.20 at 0.108 seconds
• -0.10 at 0.144 seconds
• +0.15 at 0.204 seconds
• Using a chart, you will model the seismic response to each RC individually
• Then you will sum the individual responses to get the synthetic (modeled) seismic trace