Environmental Research Foundation  [Printer-friendly version]
July 27, 2010

LOOKING FOR CO2 BURIED AT SLEIPNER

By Peter Montague

The Sleipner project in the North Sea, off the coast of Norway, is
the model that SCS Energy says it hopes to emulate with its PurGen
project proposed for Linden, N.J., which will bury 500 million tons of
CO2 beneath the Atlantic Ocean.

The Sleipner project was developed by BP (the oil company formerly
known as British Petroleum) and the Norwegian state-owned oil company,
Statoil (now called StatoilHydro). Sleipner is a production platform
anchored in the North Sea where natural gas is pumped from below the
sea floor. At Sleipner, the natural gas comes up mixed with naturally-
occurring CO2. Because Norway has a tax on CO2, BP and Statoil made
the decision in 1996 to separate the CO2 from the natural gas on the
pumping platform and then to pump the CO2 back into the ground below
the floor of the North sea. Since 1996 they have pumped about 1
million tons per year into the Utsira formation, which is a 200-meter-
thick layer of sandstone and mudstone.

In 1999, 2001 and 2002, Sleipner's buried CO2 was "imaged" using
seismic techniques. The goal was to draw a 3-dimensional picture of
the buried CO2, to find out how it has moved and where it has gone. In
2007, a group of British scientists examined the data from 1999, 2001
and 2002 and published a lengthy technical paper on their findings
in
2007.

The scientists reported a discrepancy between their model and actual
measurements of the permeability of the Utsira formation. Permeability
is a measure of the ability of a porous material (such as sandstone)
to transmit fluids. Permeability will determine how rapidly CO2 pumped
below the sea floor will spread horizontally. As they said, "A number
of factors could cause the apparent discrepancy between estimated and
measured permeabilities." (pg. 174) They listed 4 possible
explanations for the discrepancy:

1. The first possible explanation could be that the CO2 has flowed
into a dome-shaped cavity and so has not flowed laterally as rapidly
as was expected.

2. The second possibility could be that the actual measurement of
permeability was not representative of large-scale permeability in the
Utsira formation; in other words, when they took actual physical
measurements of permeability in the Utsira formation, they sampled a
spot that was not typical of large-area permeability.

3. The third possibility was that "a significant fraction of the
injected CO2 is not stored in the layers imaged." In other words, much
of the injected CO2 has gone somewhere else.

4. The fourth possibility is that assumptions in the mathematical
model are wrong.

The authors of the report commented on possibility #3 as follows:

"The volume of CO2 imaged would need to be reduced to 19% of that
injected for the whole-reservoir estimate of permeability to be
compatible with the lower bound of measured permeabilities and the
seismic data suggest this is unrealistic."

In other words, to make sense of the measured permeabilities and the
permeabilities derived from their model, they would have to assume
that 81% of the injected CO2 had gone somewhere outside the area that
they imaged.

So their modeling affirms that at least 19% of the injected CO2 has
stayed put, and the scientists believe that more than 19% has stayed
put but the discrepancy between measured permeabilities and calculated
permeabilities remains unexplained.

They conclude (pg. 175):

"Reservoir permeabilities estimated from the shape of the
accumulations [of CO2 in the Utsira formation] are below the range of
measured values and it is not yet clear to what extent this
discrepancy is a result of limitations in the modeling, a difference
between permeability on the scale of the reservoir and the scale of
the measurements, or whether less CO2 is stored in the layer imaged
than estimated from the seismic studies."

To put it bluntly, no one is able to say for sure whether CO2 injected
at Sleipner has leaked or not.

Summarizing this study in a PowerPoint presentation (see slide #12),
Peter M. Haugan, director of the Geophysical Institute, University of
Bergen (Norway), Bergen Marine Research Cluster, has made this point
explicitly:

"Layers have started filling gradually. Leaks occur through thin
mudstones. Model-data consistency requires either CO2 permeabilities
order of magnitude lower than measured on core samples or CO2 layer
thickness from seismic are overestimated. Possible that CO2 saturation
is small and CO2 has escaped."