On August 1, 2016, EPA issued the final research report: Review of Well Operator Files: Hydraulic Fracturing Operations (EPA/601/R14/004, July 2016). This new report examines the records of nine companies conducting hydraulic fracturing at 23,200 oil and gas wells between September 2009 and September 2010. Interestingly some of these oil and gas wells were hydraulically fractured more than once during that time span, so there were actually 28,500 occurrences of hydraulic fracturing over the one year period that EPA examined. EPA calls these occurrences: hydraulic fracturing jobs.
To understand the report, one must know a little bit about how oil and gas wells are constructed. After a well is drilled, a metal pipe called a casing is set into the hole created by the drill. Usually the space between the outside of the casing and the rock is filled with cement. The space between the casing and the rock is called the annular space. The annular space can be cemented for the entire length of the well or it can be cemented just in certain sections. Cement helps to protect groundwater by preventing fluids from moving up the annular space and into rocks containing protected groundwater resources. Cementing is also used to prevent gas and other fluids from moving up the annular space and causing a ruckus at the surface. Sometimes a device known as a packer is also used to seal off sections of rock containing groundwater. Packers are mechanical devises which look a lot like plungers. Before wells are hydraulically fractured, the casing next to the rock formation containing the oil and gas is perforated so the pressurized water and chemicals can enter the rock at great force and produce fractures. Occasionally however, what are called “temporary casing strings” are used for the hydraulic fracturing process instead of the permanent cemented casing. These casings really are temporary and are removed after the process is completed.
Of the hydraulic fracturing jobs EPA examined, 90 % had cemented casing, 5 % had packers, and 5 % had open holes. In 10 % of the jobs temporary casing strings were used. Temporary casing strings were used both with the open holes and in some wells with permanent cemented casings. Generally when a temporary casing string was used in an oil and gas well with a cemented permanent casing, the wells were undergoing recompletions or refractures. Recompletions are where an existing well which was previously hydrofractured at one depth is later refractured at another depth. Refractures are where the same depth in the well is being refractured. Recompletions and refractures are most often done in older wells (EPA gives a median age of 6 years for the wells examined in their report.)
This is all important knowledge, because how wells performed during hydraulic fracturing is dependent on the integrity of the safeguards like concrete, casing and packers put into place to prevent movement of fluid into protected groundwater zones. EPA calls the success of these safeguards – “mechanical integrity.” To determine the mechanical integrity of the 23,200 wells, EPA examined four different tests conducted by companies during the hydraulic fracturing process. The first type of test is a casing pressure test. This kind of test is conducted before the well is hydraulically fractured and is used to determine the mechanical integrity of the well. Basically pressures to be used in the hydraulic fracture process are applied to the well. If the well maintains its integrity then there should be no mechanical failure of the safeguards protecting groundwater during the actual hydraulic fracturing.
The second type of test is surface treating pressure monitoring. This is simply the monitoring of the pressure being applied to the well during hydraulic fracturing. If there is a sudden drop in surface treating pressure during hydraulic fracturing, it is likely one or more of the safeguards have failed.
The third test is annular pressure monitoring, which is simply the monitoring of pressure in the annular space during hydraulic fracturing. Changes in the pressure in the annular space should tell you if the wells safeguards have been breached.
And the fourth test is microsesimic monitoring, where monitors are set out to determine the extent of fracture growth in the subsurface. Microseismic monitoring can tell you if vertical fractures are long enough to intersect groundwater zones above the hydraulically fractured rock area.
As you can imagine, it was kind of hit or miss as to what sort of testing had been conducted at the wells. Two or more tests were conducted at 76 % of the 28,500 hydraulic fracturing jobs EPA examined. Casing pressure tests were conducted at 57 % of jobs, surface treating pressure monitoring at 97 %, annular pressure monitoring at 53 %, and microseismic at 0.5 %. So other than measuring surface treating pressure during the actual hydraulic fracturing process, there is simply not a lot of data being collected. In fact the annular pressure tests were so frequently insufficient; EPA opted not to even use the data for analysis. So we’re down to 3 tests now and one of them, the microseismic test, has very little data.
EPA relied mostly on the other two tests (casing pressure and surface treating pressure) and notes taken by companies kept during the hydrofracturing process to determine mechanical integrity of wells that had been hydraulically fractured. Reported in company notes were failure of the mechanical integrity of the well in 3 % of all the hydraulic fracturing jobs. Failures occurred in new wells (2%) and in recompletions and refractures (6%). EPA found that jobs using temporary casing strings had the largest number of failures. In fact 20 % of the jobs using temporary casing strings had some sort of failure. Failure generally meaning one or more of the safeguards failed. Now many wells of course don’t have just one casing, sometimes there are multiple casings set in the well, so failure doesn’t mean fracturing fluid escaped into a groundwater zone. In fact, EPA noted in their report that once failure was noted by a company, the job was stopped and the cause of the safeguard failure was addressed. But in 0.5 % of the hydraulic fracturing jobs, information in the companies’ files indicated fracturing fluid moved into the annular space at the location of groundwater.
EPA did learn something from the microseismic monitoring they examined. Among the 100 jobs where it was done, EPA used the data to estimate the vertical extent of fractures. It showed vertical fractures grew anywhere from 380 to 1340 feet. EPA could find no case where vertical fractures extended into the groundwater zone in the areas being hydraulically fractured. In fact EPA estimated fractures reached no closer than 5000 feet from any groundwater source.
What did EPA learn from the casing pressure tests and surface treating pressure monitoring? In 48 % of the hydraulic fracturing jobs where casing pressure tests were conducted, the pressures recorded during the actual hydraulic fracturing process exceeded the casing pressure tests by 1000 to 1999 pounds per square inch.
Pretty interesting information. The most interesting fact to me is there were 142 hydraulic fracturing jobs out of the 28,500 jobs examined, where the mechanical safeguards in the well failed and hydraulic fracturing fluid entered groundwater zones. Is a 0.5 % failure rate of wells during hydraulic fracturing an acceptable failure rate? This will have to be determined by state governments who oversee oil and gas operations in their states and the citizens who live in the vicinity of oil and gas fields.
On August 11, 2016, the Environmental Protection Agency’s (EPA) Science Advisory Board issued their Peer Review of the EPA Draft Report on Impacts of Hydraulic Fracturing on Drinking Water. The peer review contains the Science Advisory Boards’ official comments on the EPA Draft Report. EPA announced it will use the Science Advisory Boards’ final peer review comments, other scientific literature completed since the release of the EPA draft, and public comments to complete their final report within the next few months.