American Oil and Gas Reporter - February 2015 - 143

a series of events near the facility in the
early 1960s. Researchers conducted water
injection experiments at the Rangely oil
field in western Colorado later in that
decade, the U.S. Geological Survey reports.
A 12,000-foot well was drilled at the
Rocky Mountain Arsenal in 1961 to dispose waste fluids from facility operations,
USGS says. Injections began in March
1962. Seismic events began on April 24
with a magnitude 1.5 event. By the end
of December 1962, more than 190 earthquakes had been reported, the service
says, although none caused surface damage. More than 1,300 earthquakes were
recorded at Bergen Park, about 15 miles
west of Denver. Injections stopped at the
arsenal disposal well early in 1964.
The Rocky Mountain events were
overshadowed in October 1966 by a
magnitude 4.6 earthquake along the Colorado-New Mexico border, USGS continues. Denver felt another series of
earthquakes the following month, with
smaller events (below magnitude 3.0)
registering through April 1967. On April
10, a magnitude 5.0 event was felt in
Denver and nearby cities, USGS says,
and two weeks later, a magnitude 4.4
earthquake hit Boulder, Co. The strongest
event occurred Aug. 9, 1967, USGS indicates, with a magnitude 5.3 tremor
registered in Denver. Three months later,
another five moderate events hit Denver,
before a magnitude 5.2 event occurred
Nov. 26, causing minor damage throughout Denver.
Reacting to the Rocky Mountain Arsenal earthquakes, USGS undertook a
study at the Rangely Field, looking at
the feasibility of using controlled variations
in fluid pressure to influence the number
of local seismic events. According to a
report of that study, researchers controlled
fluid pressures using water in a secondary
recovery program. By analyzing frictional
properties of reservoir rocks and measured
in situ stress levels near pre-existing faults
in the field, they were able to determine
the critical pore pressure above which
earthquakes were induced by injection
and below which seismicity decreased.
According to the survey, changes in the
number of earthquakes recorded each year
appeared to correlate with changes in the
quantity of fluid injections. Between November 1962 and January 1970, 976 earthquakes were recorded near the oil field,
with 320 greater than magnitude 1.0.
Hydraulic Fracturing
While hydraulic fracturing is not considered a prime candidate for causing
seismic events, McGarr says there have
been moderately large events-some of

magnitude 4.0-related to reservoir stimulation techniques.
"There are several Canadian fields
that have produced several hydraulic fracturing-induced seismic events: one in
Northeast British Columbia and one in
Northwest Alberta," he says. "In Oklahoma, the largest one I am aware of was
a magnitude slightly less than 3.0. In the
United Kingdom, there was a magnitude
2.3 event. The reports are fairly rare.
Going back 10 years, there only are a
handful of events believed to have been
induced by hydraulic fracturing."
Hydraulic fracturing is a very controlled
operation, McGarr notes. The aim is to
produce a very even permeability increase
by generating a small system of cracks
and fissures to allow oil or natural gas to
flow to the borehole; the microearthquakes
that result from these operations are much
too small to be felt at the surface, he
says. Operators don't want larger earthquakes, which would produce a large
fault running through the formation. "A
single fault does not accomplish the objective of producing oil and gas from
throughout the unconventional reservoir.
(Operators) really take a lot of effort to
control these operations," he acknowledges.
A 2014 study reported in the journal
Seismological Research Letters, written
by Instrumental Software Technologies'
Friberg correlates fracturing with a series
of small earthquakes south of Clendening
Lake in Harrison County, Oh. Nearly
400 seismic events were recorded between
Oct. 1 and Dec. 12, 2013. The report
claims the 10 that registered as positive
magnitude earthquakes-the largest hitting
magnitude 2.2-coincided with fracturing
operations at nearby wells affecting a
previously unmapped fault.
The Ohio Department of Natural Resources had records of at least seven horizontal wells that had undergone fracturing
near the lake, the study reports, with
three drilled into the Point Pleasant formation 2,422 meters deep. "Despite the
absence of a fault encountered during
the drilling logs, adequate pressures would
have been introduced to cause positive
magnitude earthquakes along a pre-existing
unknown structure that was in a state of
near failure," the study says. "In the case
of the Harrison earthquake sequence, the
induced seismicity indicated faulting
below the target formation."
According to the Ohio DNR, the seismic events show a vertical east-westtrending fault directly below the three
fractured wells. It adds the fault is in the
Precambrian crystalline basement formation, and not in the Paleozoic formations
surrounding the wells, with evidence

showing a possible hydraulic pathway to
the basement fault.
Induced Events
Wastewater injections have been cited
in seismic events in Oklahoma, Colorado
and Ohio. McGarr says in the case of
Oklahoma, disposed fluids were injected
into the very porous and permeable Arbuckle formation, in which liquids could
migrate substantial distances. He says a
study let by Katie Keranen of Cornell
University revealed fluids migrating as
much as 40 kilometers from the injection
points, where they were believed to have
contributed to seismic events.
According to USGS, a human-induced
magnitude 5.0 earthquake near Prague, Ok.,
48 miles east of Oklahoma City on Nov. 5,
2011, likely triggered a magnitude 5.7 event
with an epicenter 44 miles east-northeast
of Oklahoma City the following day. The
agency says the earthquakes involved ruptures to a part of the Wilzetta fault system,
a complex fault zone 124 miles in length.
The 5.0 foreshock occurred near active disposal wells, less than a mile below the surface. By increasing stresses across the fault
zone, the foreshock triggered the main
shock as well as thousands of aftershocks,
the survey describes. The agency points
out 30 percent of the aftershocks occurred
within the Arbuckle's sedimentary section.
Analysis indicates fluids were injected
into effectively sealed compartments during the disposal wells' 18 years of injection,
USGS says, adding that the increase in
net fluid volume lowered effective stresses
on reservoir-bounding faults. "This case
indicates that decades-long lags between
commencing fluid injection and the onset
of induced earthquakes are possible," the
survey states. "The progressive rupture
of three fault planes in this sequence suggests stress changes from the initial rupture
triggered the successive earthquakes, including one larger than the first."
Since March 2011, the Youngstown,
Oh., area has experienced 12 low-magnitude seismic events along a previously
unknown fault, and the Ohio Geological
Survey says those events are adjacent to
a Class II deep injection well that began
operating in December 2010. The Northstar I well was drilled 200 feet into Precambrian basement rock, the agency reports, adding the operator agreed to plug
the well's basement section.
A magnitude 2.7 event was recorded
on Dec. 24, 2011, and preliminary findings
indicate the earthquake occurred 2,454
feet below the injection well, OGS says.
The well was shut down on Dec. 30. The
next day, a magnitude 4.0 event occurred,
and state officials ordered a moratorium
on the three deep injection wells adjacent
FEBRUARY 2015 143

American Oil and Gas Reporter - February 2015

Table of Contents for the Digital Edition of American Oil and Gas Reporter - February 2015