Why 90-Degree Phase Shift Sections Can Sometimes Be Used as an Alternative to Inversion

1. Advantages of 90-Degree Phase Shift Sections


In practical applications, 90-degree phase shift sections can approximate the results of post-stack inversion and have shown good performance across various regions. Their main advantages include:
Simple and efficient operation with fast computation speeds;
No requirement for well data or seismic horizon constraints, making them suitable for offshore areas with sparse or incomplete well logs;
Maximum reliance on raw seismic data, minimizing the influence of wells and modeling assumptions;
Seismic reflectors aligned with lithologic boundaries, making reservoir indicators more intuitive.

Reference: Zhao Haifeng et al., Application of 90° Phase Shift Technique in Neogene Reservoir Prediction in Bohai Oilfield

 
2. Basic Principle and Wedge Model Forward Modeling


(1) Problem Statement


According to the simplified convolutional model of seismic data, each interface (i.e., reflection coefficient) corresponds to a seismic reflection. When an anomalous geological body (e.g., a reservoir) with significant impedance contrast to the background exists, the upper and lower boundaries will show opposite-polarity reflection coefficients (unlike resolution analysis where both coefficients are of the same polarity), as illustrated in Figure 1. The upper boundary of a double-interface system is a positive reflection coefficient, and the lower boundary is negative.
 

Figure 1: Zero-phase wavelet reflection from a single (top) and double (bottom) interface

It is evident that the seismic section alone does not easily reveal the top and base of geological bodies, especially when the thickness or reflection strength varies—posing challenges for interpretation.

A 90-degree phase shift involves rotating the seismic phase by 90 degrees (see Figure 2). This adjustment moves the main lobe of the composite wavelet (from top and base reflections) to the center of the reservoir (see Figure 3), enabling identification of reservoir position based on where energy is concentrated.

Reference: Liu Chuanqi et al., Estimating Residual Phase of Seismic Data Using Phase Shift Characteristics of Ricker Wavelets
  
Figure 2: Zero-phase and 90-degree phase-shifted Ricker wavelets

For dual-interface reflections, the 90-degree phase shift makes the seismic response symmetric about the center of the body. The top and bottom boundaries become reflections of the same polarity, and the largest trough (or peak if using –90° shift) aligns with the reservoir center (Figure 3). This technique is especially useful in clastic depositional settings where sand bodies are embedded in shale (“sand-in-mud” environments), allowing vertical reservoir placement to be visually inferred from the seismic section.
 

Figure 3: Reflections of single and dual interfaces after 90-degree phase shift
 

(2) Wedge Model Forward Modeling


Two forward models were designed using ColchisFM (see Figure 4): one simple wedge model (top) and one thinly interbedded wedge model (bottom), both with the same total thickness and low-impedance sand units. With 2% added noise and a 25 Hz Ricker wavelet, seismic forward modeling was conducted.

From the resulting seismic sections, we observe that as the wedge thickens, both top and base reflections strengthen and eventually become separated due to tuning effects.

In the –90° phase shift sections, red zones (negative amplitudes) align well with the sand body’s geometry. Within the tuning thickness range (marked in yellow), the phase shift technique effectively delineates the total thickness. Measured seismic time thickness is about 30 ms, which, assuming a subsurface velocity of 3000–4000 m/s, corresponds to a true thickness of approximately 45–60 meters.

This demonstrates that where reservoirs exhibit a clear post-stack seismic response, 90-degree phase shift sections can effectively indicate reservoir presence. However, for thinly interbedded reservoirs, resolution limits may lead to inaccuracies in net thickness interpretation.
 

Figure 4: Wedge and thinly interbedded wedge models with corresponding –90° phase-shifted sections
 

3. Considerations When Applying 90-Degree Phase Shift


1. Determining wavelet polarity: Before applying a phase shift, the polarity of the seismic data must be identified to ensure correct compensation. This step can be tricky when well data is missing or unreliable. Correlation with marker horizons, seafloor reflectors, or regional phase consistency may help.
2. Understanding post-stack seismic response of the reservoir: Use well-to-seismic ties and integrated analysis to define the expected seismic character. In some cases, P-impedance may not differentiate the reservoir, but Vp/Vs ratios may show strong contrasts. Since we often use elastic impedance (EI) stack data, reservoirs may still present distinct seismic responses, warranting the use of far-offset data for 90° phase shift analysis.
3. Software implementation: Some software platforms define 90° phase shifts differently (positive vs. negative), potentially causing polarity mismatches. Be cautious and verify before interpretation.
4. Limitations: The 90-degree phase shift is primarily a qualitative and rapid screening tool and cannot replace full inversion. It shares the same frequency bandwidth as seismic data (thus lacks low-frequency information), making it inadequate for thick target interpretation. It also lacks well and model constraints, reducing accuracy for thin-bed delineation. When used to support horizontal drilling, precise positioning of the reservoir top requires caution. In contrast, inversion incorporates low-frequency trends and provides more accurate and quantitative results.
 

4. Technical Differences Between 90-Degree Phase Shift and Trace Integration


Trace integration (also referred to as relative impedance) is another valuable seismic interpretation technique. When data quality is high, it may offer similar or even superior results compared to 90-degree phase shift sections. However, since trace integration is sensitive to initial energy levels, poor consistency in preprocessing can lead to lateral imbalances between traces.

Reference: Zhang Jun et al., Theoretical Interpretation and Field Application of 90° Phase Shift in Thin-Bed Interpretation

This article does not elaborate on trace integration, but users are encouraged to explore and compare both techniques using ColchisFM’s built-in wedge modeling and tuning curve analysis modules.



Comments