Hydraulic fracture stimulation in combo with horizontal drilling might well be called the heart of the ongoing production success of the numerous shale plays scattered across the United States – and elsewhere.
There’s still work to do and much to learn.
Plenty of aspects of these unconventional reservoirs continue to be poorly characterized.
One of the thorny problems that must be dealt with using hydraulic stimulation is the complex interaction with natural discontinuities in the rock, as shown by mineback and coring studies.
AAPG member Doug Bearinger, geology adviser for shale gas at Nexen, has been working with a team at the company to get a better handle on what happens in these reservoirs and why.
Nexen is a major player in the Devonian Horn River gas shale in Canada’s northeast British Columbia, where it has been drilling wells for more than five years.
“The extent of the stimulated reservoir volume can be reasonably determined from the application of microseismic monitoring, proppant tracers, offset pressure changers and fluid hits of offset wells,” Bearinger said. “But the extent and geometry of the producing fracture network is poorly understood.
“Low recoveries of slickwater fracturing fluid and chemical fluid tracers suggest that much of the stimulated fracture network doesn’t clean-up,” he said. “Plus, production analysis techniques and reservoir simulations indicate that the effective producing fracture network is significantly smaller than the stimulated fracture network.”
To decipher many of the complexities resulting from stimulation, the team soon recognized that they needed to study the water coming out of the reservoir.
There are messages in this water.
“We just have to figure out how to decode them because it’s all in the chemistry of the flowback, which contains unique messages,” Bearinger emphasized.
Measure for Measure
A key issue for the evaluation is the rate of sampling.
“It takes high frequency sampling to see some of the things we’re looking for,” he said. “A typical sample schedule may not catch it; it may not figure out there’s more than one distinct trend with time.”
He noted that James Pyecroft, a veteran reservoir engineer on the Nexen team, worked with a vendor who did chemical tracing and suggested that in addition to the tracers, they should measure the characteristics of the water.
They began measuring ions.
It was determined that the chemistry from the water that comes back from a new break in the rock from hydraulic fracturing looks different than the chemistry of that out of the natural fractures that have been stimulated. This is attributed to the difference in the process that adds ions to the water that was introduced.
“A number of processes are in play, and I think we have a pretty good indication of what is going at this point,” Bearinger said.
“The simple one is water mixing,” he noted. “If water is already there, you can mix with it – and that is a straight linear mixing exercise.
“The natural fractures are gas-filled, but there is some water in these,” he said. “We’ll see a higher degree of mixing in natural fractures than induced.
“A fresh break in the rock won’t have water on those surfaces, just only a little in the matrix pores, because the matrix porosity is so low in these rock types,” he noted. “By the time you have gas-filled porosity, there’s not much water left.”
“Fluid from a new break makes its way into natural fractures and will encounter more connate water because there will be water on the surface of natural fractures.
“That’s kind of the model,” he said.
A Stimulating Subject
An added observation is that even though water doesn’t flow out of the matrix, ions can move out in an osmosis-like manner – with some moving faster than others.
In the model the research team formulated, sodium chloride or potassium are readily observed in the fresh break because they move most easily during diffusion – and they are the most soluble.
With more water available in natural fractures to afford some mixing, other ions begin to show up in bigger quantities.
Bearinger emphasized that reservoirs are hydrocarbon – bearing because water has largely been pushed away by hydrocarbons. The remaining water is not free flowing, so one of the difficulties in looking at mixing is that you can’t get an actual sample of formation water.
“We would like to know the composition of the water sitting in the pore space,” he said. “One of the challenges in figuring out this story is that we actually don’t have all of the pieces to figure it out.”
Ultimately, this effort is slated to improve characterization of the reservoir post-fracturing, giving the operator a better idea of how much of the production comes from the hydraulic fractures and how much from stimulated natural fractures.
“This starts to tell us something about the architecture of the producing fracture network,” Bearinger said. “We’re stimulating rocks two-and-a-half kilometers below the surface of the earth, and it’s challenging to know what it actually looks like.
“You can core and run image logs in wells and get an idea of what the natural fracture system looks like before you stimulate it.
“We can use microseismic and other techniques to see how far the fracing fluid reaches,” he said, “but we don’t know how much of that we actually produce from.
“There are a lot of suggestions that we produce from a lot less fracturing than we create.”