By some estimates as many as half of the perforation clusters in geometric shale hydraulic fracturing plans are not effectively stimulated, resulting in under-producing wells.
Obtaining the data needed to better place the frack clusters has required additional logging runs and rig time, until now.
A new process for planning frack jobs relies on data the drillers already gather. As the shale plays in the US have matured, drillers carried out increasingly fewer logging runs, resulting in less data about the lateral.
“Completion engineers were faced with completing wells with a lack of info across the lateral. They didn’t have the information to guide them with stage placement,” says Bill Katon, vice president of sales and operations at Drill2Frac. “The results were less than optimal.”
Geometric frack patterns arbitrarily place perforation clusters at set distances from one another, while a geomechanical design places perf clusters in rock with similar characteristics to improve production.
Drill2Frac takes the geometric frack design provided by the client and optimises that design by moving clusters into similar rock.“It gives data to the entire operations team, which they didn’t have before, that they can use to make actionable decisions on,” says Kevin Wutherich, chief technology officer at Drill2Frac.
Comparing production from wells completed with a geometric frack pattern with those completed based on full logs with frack clusters placed in a way that was appropriate for the rock shows significantly higher output, Katon says.
“We found a way to give (operators) a guide, a tool, using just drilling data that they already had.”
Bill Katon, vice president of sales and operations at Drill2Frac
Engineers had been making frack design decisions with little to no data, which led to the concept of geometric cluster design. Data gathered through running additional logs proved uneconomical and delayed well completion.
Drill2Frac was formed in 2014 to solve that problem. Company co-founder Sridhar Srinivasan says the company took a “radical” approach to turning drilling data into a format that made it possible to optimise frack designs.
The company says its method increases the effectiveness of fracking heterogeneous rock by analysing the rock mechanical data commonly obtained during drilling.
“We found a way to give them a guide, a tool, using just drilling data that they already had,” Katon says.
As such, an operator need not invest in additional logging runs, tools or rig time. Moreover, “instead of using rules of thumb, they can now design all completions based on the well itself”, says Wutherich.
The Drill2Frac process uses surface and downhole information such as rate of penetration, weight on bit, torque, standpipe pressure, differential pressure, and revolutions per minute, along with daily drilling reports, end of well reports and any other available data, such as gamma ray, logging while drilling logs, and deviation surveys.
Anomalies in the logs, such as a motor about to fail, are weeded out in the data editing and validation process.
“We are characterising the rock,” says Srinivasan. Drill2Frac uses the cleaned-up data to calculate the rock’s mechanical specific energy (MSE), a measurement that indicates the hardness of rock, which engineers have used for decades to optimise drilling.
Changes in drilling efficiency across the reservoir can cause changes in MSE. It is important to identify and correct for those changes in drilling efficiency so that the MSE delivered reflects the behaviour of the rock and not the drilling process, according to the company.
“We want to make sure the numbers that go into the MSE calculations are characteristic of the rock,” Srinivasan says.
Proven technique
Rock strength is a valuable input for frack design, Katon says. However, Drill2Frac does not compute rock strength or stress profile. Rather, it uses the MSE as a proxy for those values.
The company uses the resulting MSE “to place clusters in similar rock so they’ll break down at the same time, move them if they have room to move, or change stage length to try to get as many perf clusters as we can in similar rock”, Katon says.
At the time Drill2Frac was founded, clusters were spaced more widely apart, usually between 30 and 75 feet. Now they are more often 15 to 25 feet apart. This leaves little room for shifting clusters to avoid fracking into heterogeneous rock, which means only a portion of the rock may be stimulated, Katon says.
The company aims to work with the operator’s philosophy on completions. Approaches to fracking shale plays have changed in the last couple of years, Katon says. “We’re adapting to help these guys.”
Drill2Frac is currently developing and testing adaptations of their process to improve cluster efficiency in high-density fractured completions.
Some of these companies are dropping clusters while others are turning to diverters to get uniform stimulation in heterogeneous rock.
Diverters temporarily plug up the perforations that break down first, thereby allowing the remainder of the rock to be stimulated.
Peregrine Petroleum is one of the operators that has used the Drill2Frac technology.
Benjamin Ackley, completions engineer, says use of the technology on three wells resulted in an estimated 10% to 30% increase in perf efficiency as compared to the planned geometric design.
“We did see good diversion on all the wells when we were pumping and using Drill2Frac in conjunction with diversion,” Ackley says.
“The costs are fairly low with this technology. It makes sense to use it since it should increase your perf efficiency.”
Once Drill2Frac has worked out the MSE values along the entire lateral, the next step is to create a completion design for that reservoir that fits with the operator’s completion philosophy.
In general, the goal is to arrange the frac clusters so they perforate rocks of similar hardness. “We force the perfs into like rock,” Srinivasan says.
While Drill2Frac has provided guidance on wells with hardness facies of HD13 and HD14, most tend to range from HD1 through HD8.
Katon says for one operator’s default geometric frack design, 36 of 80 clusters would contribute to the well’s production, for a dismal success rate of 45%.
He says Drill2Frac’s design for that same well meant that 73 of the 80 clusters would contribute, for a 91% success rate.
“45% to 65% is typical of what we see for a geometric design, whereas in the optimised design we believe we are able to get closer to 90%-plus for perforating efficiency, or cluster efficiency,” he says.
“There’s no reason there shouldn’t be industry adoption. It works and it’s so inexpensive.”
Kevin Wutherich, chief technology officer at Drill2Frac
Wutherich first learned of Drill2Frac’s technology when he worked for an operator and had to evaluate whether the company would try the method on a well.
“The way I looked at it is, engineered [geometric] completions has more than 100 papers, articles about people who have seen improved production from it,” Wutherich says.
“There’s definitely a potential upside on it. The nice thing about this is it’s only manipulating data you already have, so there is such a low cost to it,” he says.
“Even on a moderate well with a 1% improvement on production, you’ll pay out the cost of the service in four months.”
Wutherich says one of the key steps in the Drill2Frac process is correlating the frack depths with driller logs.
“In my experience, people who have tried engineered completions and were not successful were getting data through drilling usually through logging while drilling tools,” Wutherich says.
Even with an engineered completion, they would not see a difference in production.
“Nine times out of 10 they didn’t correlate their depths back to their drilling depths. Maybe they mean to perf at 10,000 feet, but they’re doing it at 10,050 feet. “If you don’t do that correlation, really all your work is fruitless.”
This oversight was likely because people were unaware data correlation was necessary.
“They were assuming their log depth was the same as the depth scale the others were using,” Wutherich says. “If you’re using wireline to get data, you’re on same depth scale.”
According to Katon, everything must be related back to the driller’s depth scale.
“In horizontal drilling, the world revolves around driller’s depths, and if you’re not correlating to driller’s depth, you’ll have some sort of potential error somewhere in the process.”
Such oversights have left operators unsatisfied with engineered completions and less likely to use them on future projects, Wutherich says.
“You can have the best data in the world, but if you shoot on the wrong depth it makes no difference,” he says.
The Drill2Frac process has been used on more than 300 wells in the US and Canada for more than 50 operators.
“We are fairly confident about where this is going in terms of technology,” Srinivasan says.
The first well using Drill2Frac’s technology was completed in September 2014.
The company drew on lessons from other disciplines to find a new way to attack the problem of completions not hitting their targets.
They had to figure out how to quantify the clusters that contributed to production, and come up diverter strategies.
“We have validated it with several clients through production, through comparison with other logs, using fibre optic data,” Wutherich says.
“We have a lot of confidence that it works, especially at the price point that we’re at. There’s no reason there shouldn’t be industry adoption. It works and it’s so inexpensive.”
石油圈
