Hydraulic fracturing has long relied on microseismic monitoring, tracers, and fiber optics to infer fracture geometry and complexity. Yet every stage of every well already contains a widely overlooked signal: the pressure oscillations that occur at pump shutdown, known as water hammer (WH).
In our recent SPE Journal publication (SPE-225459), we introduced a novel technique that transforms this free but underestimated signal into a powerful diagnostic tool. By treating water hammer as a damped harmonic oscillator and analyzing its pressure response with the Continuous Wavelet Transform (CWT), we extract damping coefficients that correlate directly with fracture complexity in the reservoir.
he workflow applies the CWT using a complex Morlet wavelet to isolate the dominant frequency ridge of the WH signature, then calculates the damping coefficient from the logarithmic envelope decay. When applied across multiple fracture stages in Marcellus wells, the damping coefficient trends correlated strongly with natural fracture density interpreted from image logs. This provides operators with a cost-effective, pressure-only method to evaluate induced fracture complexity stage by stage.
Value for Unconventional Reservoirs
Unconventional reservoirs depend on maximizing stimulated reservoir volume (SRV) through the creation of complex fracture networks. Traditionally, operators rely on indirect diagnostics such as production logging, tracers, or costly microseismic surveys to estimate fracture effectiveness. The WH-CWT method directly leverages existing pressure data to provide real-time, low-cost insights into fracture complexity, even in the absence of other diagnostics.
This is particularly valuable in resource plays where thousands of stages may be completed annually and marginal wells cannot justify the expense of advanced diagnostics. By linking damping behavior to fracture tortuosity and natural fracture interaction, operators can rapidly identify which stages achieved high complexity and which underperformed, supporting design optimization, well-to-well benchmarking, and economic decision-making.
Model Assumptions and Limitations
The model assumes that water hammer can be represented as a partially damped harmonic oscillator with an added exponential decay term to capture fluid leakoff effects. Under this assumption, damping is governed primarily by fracture tortuosity, interaction with natural fractures, and leakoff into the formation. The approach further assumes:
-
Immediate pump shutdown is necessary to avoid overlapping signals from stepped closures.
-
Consistent fracturing design across stages allows damping coefficient variations to be attributed mainly to differences in induced complexity rather than treatment changes.
-
The complex Morlet wavelet adequately captures oscillatory behavior and phase shifts in the WH signal.
While the method has shown strong correlation with fracture density logs and robust performance in Marcellus wells, operators should note that gradual pump shutdowns, noise, or highly heterogeneous reservoirs can complicate interpretation. Even in such cases, overdamped signatures often provide useful qualitative indications of poor fracture-wellbore communication.
Key findings include:
-
Damping coefficients serve as reliable indicators of fracture complexity and natural fracture intensity.
-
WH responses can be classified from high-amplitude oscillations to overdamped signals, reflecting changes in fracture tortuosity and communication.
-
Nonlinear optimization methods, especially basinhopping, successfully matched modeled WH responses with field data, achieving R² values above 0.86 in most cases.
This new approach leverages data already collected during every hydraulic fracturing stage, requiring no additional field instrumentation. The information is freely available and can be analyzed for each job. The only operational step is a hard pump shutdown for approximately five minutes—no special gauges, sensors, or extra procedures are needed. By reframing water hammer as more than background noise, this method unlocks valuable insights for fracture diagnostics, completion design optimization, and stage-by-stage performance evaluation.
For full methodology, mathematical formulation, case studies, and correlation with fracture density logs, see the peer-reviewed article in the SPE Journal: SPE-225459 or watch our the following webinar : Decoding Induced Fracture Complexity: Water Hammer Damping Analysis with Continuous Wavelet Transform
