Seismic Processing and Interpretation

Seismic Reflection Processing Seismic Reflection Processing Seismic Reflection Processing • Filtering Data: • The tape can be played back with a suitable filter if it is desired to cut off unlikely frequencies due to noise. • A filter is a device or process that modifies the input process such t...

Seismic Reflection Processing Seismic Reflection Processing Seismic Reflection Processing • Filtering Data: • The tape can be played back with a suitable filter if it is desired to cut off unlikely frequencies due to noise. • A filter is a device or process that modifies the input process such that the output has different characteristics. • A digital filter changes the characteristics of a digital signal. • The purpose of signal analysis is to apply operations (filters) to our signals that make the information in them easier to see and understand, while reducing the effects of undesirable components such as noise and echoes. Seismic Reflection Processing • Filtering Data: • Objective: to remove noise and to enhance the vertical resolution of the data. • Seismic data is usually contaminated by noise. • Source of noise include wind, traffic and cultural noise, electrical signals from power lines, ocean waves and the motion of the ship and streamer. • Filtering will improve the visual display of the seismic sections by taking out unwanted noise. Seismic Reflection Processing • Much of the noise in seismic data occurs at a frequency that is distinct from the seismic signal – e.g., wind tends to have a higher frequency, marine or traffic noise has a lower frequency. • This noise can be removed by applying frequency domain filters. • Common filters are: • High pass filter (or low cut filter) - All frequencies above a given frequency are retained, lower frequencies are removed. Seismic Reflection Processing • Low pass filter (or high cut filter) - All frequencies below a given frequency are retained, higher frequencies are removed. • Band pass filter – only data within a given frequency range are retained • Notch filter – removes data within a given frequency range Seismic Reflection Processing • Some noise will have frequencies that are close to those of the reflection signals and can not be removed by frequency domain filtering. • Besides filtering, another method is also used to eliminate or minimize the noise: Multiple geophones & stacking. • Some noise is random, which means that stacking seismic traces will tend to reduce its amplitude, and increase the signal-to-noise ratio (SNR). Seismic Reflection Processing • The summation of signals from a number of similar input channels for the purpose of increasing the signal-to noise ratio is known as stacking. • 2 main types: • i) Vertical stacking: one shot-point and a geophone, both of which are kept at fixed locations. • The outputs of the geophone for a number of shots are added, either in the field or in the digital processing of data. Seismic Reflection Processing • ii) Common-mid(depth)-point (CMP) • The numbers represent the traces on a record by different source-station pairs. • Fold of a CMP is determined by m=N/2n, where N is no. of geophone, n is no of array spacings the source is moved between shots. - improvement S/N-ratio from stacking m traces is sqrt (m). - eg: N=96, n=8: => 96/16 = 6 fold. - improvement of S/N ratio is sqrt(6) ~ 2.5. - Fold = the number of times the same point on a reflector is sampled. For hc exploration: 50, 100, or even 1000. Seismic Reflection Processing • ii) Common-mid(depth)-point (CMP) • The numbers represent the traces on a record by different source-station pairs. • Fold of a CMP is determined by m=N/2n, where N is no. of geophone, n is no of array spacings the source is moved between shots. - improvement S/N-ratio from stacking m traces is sqrt (m). - eg: N=96, n=8: => 96/16 = 6 fold. - improvement of S/N ratio is sqrt(6) ~ 2.5. - Fold = the number of times the same point on a reflector is sampled. For hc exploration: 50, 100, or even 1000. Seismic Reflection Processing Seismic Processing • Common depth/mid point (CMP) stacking, which involves the arrangement of component data for a single depth point side by side. The Power of Stacking: Example, Stack of Five Shots (A), and Single Shot (B) (A) (B) Seismic Processing • Seismic processing: • The aims are to enhance the interpretable (useful) seismic information relative to the noise in the signal and place the seismic reflectors in their correct x, y, z space. Reflected Ray Paths Geometry • Position of reflector can be determined by measuring to, travel time at shot point. • When x=0, then h = ½Vto or to = 2h/V • Then eq. (1) can be written as: t2 = x2 / V2 + 4h2 / V2 t2 = x2 /V2 + to2 (2) • To find t, can be solved from eq. (1). Reflected Ray Paths Geometry • So, if plot t2 vs x2, we can determine V1 and h1 from the slope and intercept. • The different (additional correction) in the travel time of a reflection for 2 geophones position is known as moveout, Δt. Reflected Ray Paths Geometry • If t1 & t2 are travel time and x1 & x2 are offset, so: Δt = t2 – t1 ~ (x22 – x12 )/ 2V2to (3) If one geophone at shotpoint (or at x1=0), Δt known as normal moveout (NMO), Δtn then, Δtn ~ x2 / 2V2to (4) The importance of NMO: • Having determined the layer velocity, we can use the predicted quadratic shape to identify reflectors • Then correct (shift traces) and stack to enhance signal to noise Single horizontal reflector QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. Single horizontal reflector Single horizontal reflector QuickTime™ and a TIFF (Uncompressed) decompressor are needed to see this picture. Multiple Horizontal Reflector B) If there are 2 or more reflector layers, we want t(x) for each reflected event: t1 ~ 2h1/V1 t 2 ~ 2[(h1/V1) + (h2 /V 2 )] And so for t3 ….until tn. • If the average velocity for each layer as V1, V2 ---Vn is known, then the thickness for layer h1, h2, ---hn can be calculated. The average velocity Vav can be given as: Vav = Σhj/Σ(hj/Vj) Seismic Reflection Processing • Reflectivity and Convolution: • The seismic trace that will be recorded in the field depends on both the reflectivity function and the seismic pulse that is sent out by the source. • It can be shown that the seismic trace is the convolution of the reflectivity function and the input (source) pulse. Presentation of data: Multiple Horizontal Reflector • Deconvolution (Inverse filtering): • is a method that further removes noise of the seismic traces. Multiple Horizontal Reflector • Deconvolution (Inverse filtering): • is a method that further removes noise of the seismic traces. Multiple Horizontal Reflector • Deconvolution (Inverse filtering): • is a method that further removes noise of the seismic traces. Multiple Horizontal Reflector • Deconvolution (Inverse filtering): • is a method that further removes noise of the seismic traces. Multiple Horizontal Reflector • Deconvolution (Inverse filtering): • is a method that further removes noise of the seismic traces. Multiple Horizontal Reflector • Deconvolution (Inverse filtering): • is a method that further removes noise of the seismic traces. Multiple Horizontal Reflector • Deconvolution (Inverse filtering): • Is a method that further removes noise of the seismic traces. Deconvolution is a data processing step that is used to shorten the length of the input pulse, in an attempt to recover the sharp reflectivity function and therefore increase the vertical resolution of the seismic data. Multiple Horizontal Reflector Dipping Layers Dipping Layers • Travel time curve is a hyperbola, but asymmetric where the axial plane is at x = -2hsinθ • t has different value for geophones at both side of shotpoint • when x = 0, h = ½Vt0 but h is not a vertical depth To find dip angle θ, we need to solve t value from Eq. (5). Dipping Layers t = 2h/V [1+ (x2+4hx sinθ) /4h2]½ » to find θ, which is from the different of travel time for 2 geophones at both side with the same distance. x for down-dip = +s and for up-dip = - s, Then travel time t1 and t 2 can be calculated: Dipping Layers t1 ~ to [1+ (s2+4hs sinθ) /8h2] t2 ~ to [1+ (s2-4hs sinθ) /8h2] » Δtd = t2 - t1 ~ to (s sinθ /h) ~ 2s sinθ /V Value Δtd → dip moveout If we replace Δx for the distance 2s, dip can be estimated: sinθ ~ V(Δtd / Δx) Dipping Layers t1 ~ to [1+ (s2+4hs sinθ) /8h2] t2 ~ to [1+ (s2-4hs sinθ) /8h2] » Δtd = t2 - t1 ~ to (s sinθ /h) ~ 2s sinθ /V Value Δtd → dip moveout If we replace Δx for the distance 2s, dip can be estimated: sinθ ~ V(Δtd / Δx) Dipping Layers t1 ~ to [1+ (s2+4hs sinθ) /8h2] t2 ~ to [1+ (s2-4hs sinθ) /8h2] » Δtd = t2 - t1 ~ to (s sinθ /h) ~ 2s sinθ /V Value Δtd → dip moveout If we replace Δx for the distance 2s, dip can be estimated: sinθ ~ V(Δtd / Δx) Dipping Layers t1 ~ to [1+ (s2+4hs sinθ) /8h2] t2 ~ to [1+ (s2-4hs sinθ) /8h2] » Δtd = t2 - t1 ~ to (s sinθ /h) ~ 2s sinθ /V Value Δtd → dip moveout If we replace Δx for the distance 2s, dip can be estimated: sinθ ~ V(Δtd / Δx) Dipping Layers t1 ~ to [1+ (s2+4hs sinθ) /8h2] t2 ~ to [1+ (s2-4hs sinθ) /8h2] » Δtd = t2 - t1 ~ to (s sinθ /h) ~ 2s sinθ /V Value Δtd → dip moveout If we replace Δx for the distance 2s, dip can be estimated: sinθ ~ V(Δtd / Δx) NMO Correction • Before and after NMO Dipping Layers t1 ~ to [1+ (s2+4hs sinθ) /8h2] t2 ~ to [1+ (s2-4hs sinθ) /8h2] » Δtd = t2 - t1 ~ to (s sinθ /h) ~ 2s sinθ /V Value Δtd → dip moveout If we replace Δx for the distance 2s, dip can be estimated: sinθ ~ V(Δtd / Δx) Dipping Layers t1 ~ to [1+ (s2+4hs sinθ) /8h2] t2 ~ to [1+ (s2-4hs sinθ) /8h2] » Δtd = t2 - t1 ~ to (s sinθ /h) ~ 2s sinθ /V Value Δtd → dip moveout If we replace Δx for the distance 2s, dip can be estimated: sinθ ~ V(Δtd / Δx) www.ocean.slb.com/docs/seabed/Public_Webreport_2010_1 Velocity analysis to determine dynamic corrections: a) CMP gather, b) semblance analysis, c) CMP after normal moveout correction, d) stack with 20 neighbouring CMP’s. Migration • Migration: The process of trying to move reflections back to their point of origin. When beds dip steeply, the wave returns from the reflector from a point not immediately beneath the surface location midway between the shotpoint and each individual geophone but from a point up-dip from this position. The data must be migrated to correct this effect. • In consequence, migration is designed to restore seismic reflectors to their proper x—y position Dipping Layers t1 ~ to [1+ (s2+4hs sinθ) /8h2] t2 ~ to [1+ (s2-4hs sinθ) /8h2] » Δtd = t2 - t1 ~ to (s sinθ /h) ~ 2s sinθ /V Value Δtd → dip moveout If we replace Δx for the distance 2s, dip can be estimated: sinθ ~ V(Δtd / Δx) Dipping Layers t1 ~ to [1+ (s2+4hs sinθ) /8h2] t2 ~ to [1+ (s2-4hs sinθ) /8h2] » Δtd = t2 - t1 ~ to (s sinθ /h) ~ 2s sinθ /V Value Δtd → dip moveout If we replace Δx for the distance 2s, dip can be estimated: sinθ ~ V(Δtd / Δx) Dipping Layers t1 ~ to [1+ (s2+4hs sinθ) /8h2] t2 ~ to [1+ (s2-4hs sinθ) /8h2] » Δtd = t2 - t1 ~ to (s sinθ /h) ~ 2s sinθ /V Value Δtd → dip moveout If we replace Δx for the distance 2s, dip can be estimated: sinθ ~ V(Δtd / Δx) Dipping Layers t1 ~ to [1+ (s2+4hs sinθ) /8h2] t2 ~ to [1+ (s2-4hs sinθ) /8h2] » Δtd = t2 - t1 ~ to (s sinθ /h) ~ 2s sinθ /V Value Δtd → dip moveout If we replace Δx for the distance 2s, dip can be estimated: sinθ ~ V(Δtd / Δx) Seismic Surveying Pre-migration Migrated stack Seismic migratiotion • Original seismic section Migrated seismic section Note improvement of data/image quality Seismic interpretation • Check line scale and orientation. • Work from the top of the section, where clarity is usually best, towards the bottom. • Distinguish the major reflectors and geometries of seismic sequences. Scale and orientation Top down approach • Start at the top of the section, where definition is usually best • Work down the section toward the zone where the signal to noise ratio is reduced and the reflector definition is less clear first second third Reflector character and geometry Continuous reflector truncating short ones Next continuous reflector Reflectors onlapping continuous one Seismic Surveying Seismic Surveying What Are DHIs? DHI = Direct Hydrocarbon Indicator • Seismic DHI’s are anomalous seismic responses related to the presence of hydrocarbons • Acoustic impedance of a porous rock decreases as hydrocarbon replaces brine in pore spaces of the rock, causing a seismic anomaly (DHI) • There are a number of DHI signatures; we will look at a few common ones: – Amplitude anomaly – Fluid contact reflection – Fit to structural contours Typical Impedance Depth Trends In general: 3 • Oil sands are lower impedance than water sands and shales 4 • The difference in the impedance tends to decrease with depth • The larger the impedance difference between the HC sand and it’s encasing shale, the greater the anomaly DEPTH x 103 FEET 5 • Gas sands are lower impedance than oil sands 5 IMPEDANCE x 103 10 15 20 Looking for shallow gas OIL SAND 25 SHALE 6 7 8 9 10 Looking for deep oil Data for Gulf Of Mexico Clastics DHIs: Amplitude Anomalies Anomalous amplitudes Change in amplitude along the reflector Low High Amplitude DHIs: Fluid Contacts Hydrocarbons are lighter than water and tend to form flat events at the gas/oil contact and the oil/water contact. Thicker Reservoir Fluid contact event Thinner Reservoir Fluid contact event Courtesy of ExxonMobil DHIs: Fit to Structure Since hydrocarbons are lighter than water, the fluid contacts and associated anomalous seismic events are generally flat in depth and therefore conform to structure, i.e., mimic a contour line Courtesy of ExxonMobil Amplitude Variation with Offset • Amplitude Variation with Offset (AVO) • Variation in seismic reflection amplitude with change in distance between shotpoint and receiver that indicates differences in lithology and fluid content in rocks above and below the reflector. • AVO is a seismic technique that uses pre-stack seismic data, to detect the presence of hydrocarbons in the reservoir. • In reservoir rock, AVO response is dependent on the velocities of P- and S-waves and on density to define the pore space and fluids within the rock matrix. R= v1 1 − v2  2 v1 1 + v 2  2 Amplitude Variation with Offset • AVO analysis is a technique by which geophysicists attempt to determine thickness, porosity, density, velocity, lithology and fluid content of rocks. • A gas-filled sandstone might show increasing amplitude with offset, whereas a coal might show decreasing amplitude with offset. Amplitude Variation with Offset Why Do We Care? Reflection amplitude varies with θ as a function of the physical properties above and below the interface • Rock / lithologic properties • Properties of the fluids in the pores Examining variations in amplitude with angle (or offset) may help us unravel lithology and fluid effects, especially at the top of a reservoir Top of Reservoir Base of Reservoir Impedance Lo Hi Zero Offset Near Offset Full Offset Far Offset Why Do We Care? • Trough is top of reservoir (shale over gas sand) • Peak is base reservoir • How wide is the reservoir? Maps or crossplots of AVO responses can be used to detect pore-fill anomalies i.e. hydrocarbons and map their lateral extent Top of Reservoir Base of Reservoir Impedance Lo Hi Zero Offset Near Offset Full Offset Far Offset Why Do We Care? • Trough is top of reservoir (shale over gas sand) • Peak is base reservoir • How wide is the reservoir? Maps or crossplots of AVO responses can be used to detect pore-fill anomalies i.e. hydrocarbons and map their lateral extent Top of Reservoir Base of Reservoir AVO: Quantified with 2 Parameters We quantify the AVO response in terms of two parameters: • Intercept (A) - where the curve intersects 0º • Slope (B) - a linear fit to the AVO data AVO Curve Angle/Offset AVO Crossplot • Negative Intercept • Negative Slope Angle/Offset For some reservoirs, the AVO response differs when gas, oil and water fill the pore space AVO Gradient (B) Time Amplitude CDP Gather: HC Leg Water Oil Gas AVO Intercept (A) A real data example (b) Picks (a) Gas Sand Zoeppritz Amplitudes (c) (a) Common offset stack over a gas sand, where (b) represents model and picks over trough, and (c) represents model and picks over peak. Estimating A and B from Seismic Data Offset +A +B sin2q Time -A -B (a) Small part of common offset stack. (b) Peak/trough picks vs sin2q, where A=intercept, and B=gradient. Intro to Exercise Goal: To map the extent of the A1 gas-filled reservoir W E A1 Gas Sand Figure 1 Inline 840 Courtesy of ExxonMobil Changes in Amplitude Indicate Fluid Gas Sand Water Sand Traces are ‘clipped’ Figure 1 Inline 840 Courtesy of ExxonMobil Fluids within the A1 Sand Extent of Gas Figure 1 Inline 840 Courtesy of ExxonMobil An unmuted CDP gather displaying an AVO anomaly that appears prominent compared to the events above and below