A Denver-based startup company is trying to fine-tune a technique that would give shale producers instant gratification by explaining one small aspect of the behavior of fracturing fluids during a horizontal well completion.
The subject of this research is known as the water hammer effect and, though others have investigated, no one has yet made complete sense of it. Founded in 2014, the oil and gas analytics firm Well Data Labs recently completed a study that aimed to get closer to unraveling the mystery.
Because the signature of the water hammer signal is determined by the makeup of the wellbore and near-wellbore fractures, the engineering team behind the research believes the raw data can be crunched into a report that describes the quality of a well’s connectivity with the producing reservoir.
From there, engineers could draw conclusions on how much proppant or sand was really placed into a particular stage and do so quickly.
“The main advantage here is that the data is already being collected on every job,” said Josh Merritt, a petroleum engineer with the company. He explained the water hammer effect occurs whenever there is a sudden change in the velocity of the fracturing fluids being pumped downhole.
Most often this happens as the pressure pumps slow their rate or shut off, sending a pressure pulse through the wellbore. When that pulse is reflected back to surface, it is recorded by sensors or gauges, although Merritt pointed out the data are rarely used as a meaningful input.
Well Data Labs will be sharing the results of its study which it says identified clear trends based on the water hammer signal in a technical paper at the SPE Oklahoma City Oil and Gas Symposium being held 27-31 April.
Telling Differences
If the water hammer signal is proven to be a reliable diagnostic tool, engineers would potentially have another justification to make on-the-fly adjustments to their hydraulic fracturing designs. Additionally, they may be able to use this data point to help predict a well’s productivity or integrate it with hydraulic fracturing and reservoir models.
Jessica Iriarte, a petroleum engineer with Well Data Labs, said the company’s interest in the water hammer signal was piqued as it began seeing substantial differences in the strength of the signals between its clients’ data. It turned out that most of the disparity could be explained by whether a horizontal well used sliding sleeves or plug-and-perfs—the two most common completion approaches in the shale sector.
“That was very interesting to us and is actually how the whole study started,” said Iriarte, adding that the research has since revealed there are a number of other determining factors beyond downhole mechanisms. “Our aim now is to identify what is affecting that signal, and we see that sometimes it might be the completion, sometimes it might be the fractures, or it could be reservoir properties.”
Among the study’s findings is that wells using sliding sleeves have a far more consistent water hammer signal than wells using plug-and-perf, which showed variations for each fracturing stage. In the latter group of wells, the company determined that when no signal was received at the end of treating a stage, it meant that the proppant was not successfully placed.
For certain wells, the company was able to correlate a higher decay rate and amplitude of the signal to interwell communication, an indicator of a possible frac hit. Thirty-day production rates were also examined and tied to faster decay rates, which was theorized to be a result of formation permeability and the size of the fracture network area.
Merritt said that in addition to production data, its clients’ chemical tracer results were also used as validation. “It was really neat to see some of those things line up as we put the data in,” he said.
Limitations and Future Study
Though this study has yielded some new insights, there remain too many unanswered questions to use the water hammer signal in real-time decision making. One of the remaining hurdles involves figuring out a way to calibrate the analysis of the signal to fit the particulars of each area or formation interval being drilled and completed.
Another possible limiter involves how many perforations are used per cluster stage. The study suggests this diagnostic approach is more applicable to wells designed with single-entry clusters, meaning there was only one perforation per stimulated cluster. One perforation reduces the noise and increases the certainty of the correlations.
Merritt said future studies will likely turn to modeling, which should provide a clearer picture of how the wellbore and near-wellbore area alter the signal. The good news is that a lot of the factors needed for this more advanced work are not hard to obtain. “We know what the casing sizes are, we know where our perforations exist, and tying in a larger data set with the models will add a lot of strength” to the reliability of this approach, he said.
This first study also did not benefit from the inclusion of reservoir characteristics or rock properties, but those are two parameters that the engineering team wants to integrate next time because they relate strongly to how fractures propagate. Estimated calculations of the surface area inside the open fracture network will also be useful since the volume of space inside the rock matrix is understood to directly affect how much of the water hammer wave is absorbed.
