Underpressure in the north

05.07.2017
Why is there underpressure in certain formations in both the Barents Sea and Adventdalen on Svalbard?

The drilling site, Adventdalen - Svalbard

The drilling site, Adventdalen - Svalbard

 

Early in the 1980s, large oil and gas discoveries were made in the newly opened areas north of the 62nd parallel.

The Smørbukk discoveries on Åsgard in the Norwegian Sea only had small deviations from hydrostatic pore pressure in the reservoir, while wells drilled on the same type of prospect nearby encountered high overpressure. The overpressured wells were dry.

For many years after, it was therefore a commonly held belief that the overpressured areas on the Halten Terrace were not prospective.

Being able to predict pore pressure so as to plan and carry out safe drilling operations has always been important in the oil industry; however, in the 1980s, some geologists started using their understanding of pore pressure for prospect evaluation. In high-pressure areas, it is probable that hydrocarbons could accumulate in deeper traps even though the seal of the shallower structures was fractured.

The people that cracked the high-pressure code in the Norwegian Sea were eventually given the opportunity to drill deeper prospects, which e.g. led to Saga Petroleum’s Kristin discovery in 1997.

In the Barents Sea, there was hydrostatic pressure in the discoveries on Snøhvit, and virtually all other exploration wells, and hardly anyone imagined that the pressure conditions would offer up surprises when the areas further to the north in the Barents Sea were being explored in 1988.

However, unexpected problems occurred in the first well in the Fingerdjupet sub-basin, 7321/7-1, with large mud loss to the reservoir formations.

There was a powerful underpressure in the reservoir in all three wells drilled in this basin, 8 to 20 bar below hydrostatic pressure.

To my knowledge, naturally occurring underpressure of this magnitude had never previously been reported in offshore basins. What could these underpressures tell us about the geological development?

This also had consequences for exploration: All three wells contained traces of gas, but were dry. Did this have something to do with the underpressure? How is the discovery probability affected by such low pore pressures?

These events were what caused me to realise how important it is for a petroleum geologist to analyse and evaluate the pore pressure data.

 

Depth to base of Cretaceous and water depth in the southwestern Barents Sea. The area without colour is the Loppa High, where the Jurassic and younger beds are eroded. The profile northwest of the Loppa High runs through the three wells in the Fingerdjupet basin. These boreholes are located in the deep part of the Bjørnøya trough where occurrences of gas hydrate are interpreted based on seismic data. Black dots: Exploration wells.

Depth to base of Cretaceous and water depth in the southwestern Barents Sea. The area without colour is the Loppa High, where the Jurassic and younger beds are eroded. The profile northwest of the Loppa High runs through the three wells in the Fingerdjupet basin. These boreholes are located in the deep part of the Bjørnøya trough where occurrences of gas hydrate are interpreted based on seismic data. Black dots: Exploration wells.


Subsidence, hydrostatic pressure and overpressure

If you have a well at your holiday cabin, you can look down at the water surface. It is at the same level as the top of the groundwater.

The downward pressure in the water column increases according to the formula P= ςgz, where ς is the density of the water, g is the gravitational acceleration (9.81 m/s2) and z is the depth difference from the water surface to where the pressure is measured.

When the pressure going down in the subsurface increases according to this formula and is in equilibrium with the water table (or sea surface), we say that it is hydrostatic.

The water table in a well on land will be deeper during a dry summer than a wet autumn, and the groundwater table varies with topography.

The sea surface does not vary much, but the hydrostatic pressure in offshore geological formations is affected by the tide and will adjust according to sea level changes.

Overpressure builds up in a water volume when it is confined inside the pores of a sealed reservoir under subsidence as it is buried under more and more sediment load.

The available pore volume will decline when buried due to compaction of the reservoir. The compaction is caused both by the sediment grains being physically squeezed together and chemical processes that lead to dissolution and precipitation of minerals in the pore cavities.

The water volume in the subsiding pore volume will increase with thermal expansion, and the total volume of liquid could also increase with the addition of hydrocarbons from mature source rocks.

The principles are simple: As a result of the burial, the pore volume will decline while the volume of liquid will have a tendency to increase. If water can escape and there is communication to the seabed (open system), the pressure continues to be hydrostatic. If the excess water cannot bleed off quickly enough, overpressure will build up. The compressibility of water is so low that it only takes small volume changes to create major changes in pressure. If the reservoir is sealed and the pore volume shrinks, the pressure will continue to rise until it reaches the resistance to fracture (fracturing pressure) of the sealed layers.

As a rule of thumb, if there is a 0.5 to 1 per cent surplus of water in a closed groundwater system, one can expect the pressure to increase by 100 bar. This calculation takes into account compressibility of the rock. The pressure increase in the water will in fact be counteracted by some compression of the network of mineral grains in the reservoir.

 

Can erosion cause underpressure?

In a subsiding basin that is being buried by more and more sediment, it is thus natural for overpressure to build up. In central parts of the North Sea and in the Norwegian Sea, it is common to encounter overpressure with burial of 3000-3500 m or more.

But the geological beds underneath the seabed in the Barents Sea do not subside. Most of the Barents Sea shelf has a history of net erosion through the ice ages, and vast areas had their maximum burial 30 – 40 million years ago.

When subsidence leads to high pressure, is it then reasonable to expect that a removal of sediments will lead to low pressure? This would e.g. happen if the compacted pore network started to expand when it was unloaded. But one can see on both Svalbard and in the Barents Sea that a sedimentary rock that has been compacted will remain compacted even if the load is removed. Many of the processes that lead to compaction cannot be reversed.

In 2009, I performed a comparison of pressure data from the Barents Sea in the Glacipet research project to see whether there could still be correlations between unloading and low pore pressure.

One of the points of departure was that the Fingerdjupet sub-basin has been exposed to 2000 – 3000 metres of erosion since Eocene. This area has deeper erosion than the other areas drilled in the Barents Sea at that time.

Underpressure and tendencies for underpressure in a smaller scale had also been measured in certain other exploration drilling wells. The data I had was assessed in relation to the physical processes that occur during unloading.

  1. If water can move into or escape from the reservoir to the seabed, the pressure will be hydrostatic.
  2. Stress reduction and cooling will take place underneath a surface that is eroding, since the temperature increases with depth, and the unloaded sediment comes closer to the surface. The water volume in the pores will shrink due to the cooling. The reservoir will also contract due to cooling. The net effect of this on a sealed reservoir is not known, but was assumed to be minor, since certain wells in the erosion area have overpressure.
  3. The pore volume can expand somewhat because the stress caused by the weight of the overburden declines. Such an expansion effect is probably most relevant in unconsolidated sediment. In consolidated rocks, changes in stress can lead to cracking and make it easier for gas and liquids to move out of the reservoir.
  4. In a closed volume that originally contained both water and hydrocarbons, the volume of liquid in the pores will decline if gas can escape from the system. This will contribute to underpressure. There are many observations of residual hydrocarbons in the wells in the Barents Sea, which indicates that there have been extensive leaks.

One interesting result of the comparison for Glacipet was that observed underpressure of 1 – 2 bar occurred in reservoir volumes that were either small or had very low permeability and that there also was or had been a gas leak.

The explanation for these small underpressures could be that gas seeps out faster than the water flows back into the pores. A deficit of gas and water in the pores leads to pressure reduction.

There were many gas leaks in the Barents Sea during the ice ages, but in reservoir systems with high permeability and good communication to the seabed, the water pressure remains hydrostatic if the inflow of water is in equilibrium with seepage of gas. The figure shows that the major Jurassic aquifer in the Hammerfest basin has hydrostatic pressure. The pressure is seemingly somewhat higher than hydrostatic, but this could be explained with the pore water having a high salt content and therefore higher density than seawater.

 

Plot of pore pressure against depth in relinquished exploration wells in the Barents Sea. Blue lines: Hydrostatic pressure with water with normal and high salt content. Red rings: Underpressure wells in the Fingerdjupet sub-basin (Fing) and Longyearbyen CO<sub>2</sub> lab. Blue arrows: Pressure points from various local systems with weak underpressure.

 

Plot of pore pressure against depth in relinquished exploration wells in the Barents Sea. Blue lines: Hydrostatic pressure with water with normal and high salt content. Red rings: Underpressure wells in the Fingerdjupet sub-basin (Fing) and Longyearbyen CO2 lab. Blue arrows: Pressure points from various local systems with weak underpressure.

 

The figure shows that around the Loppa High and the Finnmark coastline, both the Jurassic and Upper Triassic stratigraphic sequences are in communication with the seabed over large areas.

These considerations could explain observations of minor underpressures, but did not fully explain the significant underpressures in the Fingerdjupet sub-basin.

A few years later, I received a phone call from Snorre Olaussen, currently professor of Arctic geology at UNIS, a call which opened for approaching this mystery from a new angle.

 

Underpressure in Adventdalen

The purpose of the Longyearbyen CO2-lab project, headed by UNIS, was to examine the possibility of injecting and storing CO2 on Svalbard.

Several wells down to a depth of 1 000 metres were drilled, logged, tested and monitored.

The goal for CO2 injection is Jurassic and Upper Triassic sandstone.

Snorre Olaussen, who followed up the drilling, witnessed extreme pressure conditions. Well Dh-4 went through approx. 100 metres of permafrost and into sandstone from the Helvetiafjellet Formation that had groundwater with some overpressure.

The deeper sandstone layers in the Jurassic and Triassic are sealed from the Helvetiafjellet Formation by approx. 400 metres of shale.

Injection testing in 2010 showed that the pressure in the low-permeable Triassic sandstones was extremely low, up to 50 bar below hydrostatic pressure.

Alvar Braathen and his colleagues have described the results from the project in the Norwegian Journal of Geology. They also write about pressure conditions, and state that they have not arrived at a final explanation for the extreme underpressure.

When I talked to Snorre, he was very interested in the Fingerdjupet sub-basin. Could these areas of underpressure have something in common?

 

Schematic profiles through the drillings at Longyearbyen (top) and the Fingerdjup sub-basin with the three wells 7321/7-1, 8-1 and 9-1. The blue dotted line shows the bottom of the permafrost. The underpressure is sealed by permafrost where the layers come up to the surface towards the Sassen Valley toward the ENE. If thawing or other processes at the bottom of the permafrost in this area create a deficit of water in the pores, it will affect the pressure conditions in the well. The profile in the Fingerdjupet sub-basin (cf. map of depth to chalk), shows schematic Jurassic and Upper Triassic reservoir rocks (yellow), depth to the bottom of the gas hydrate area (red line) and maximum calculated depth to gas hydrate during the last ice age (blue line). The black lines are faults.

Schematic profiles through the drillings at Longyearbyen (top) and the Fingerdjup sub-basin with the three wells 7321/7-1, 8-1 and 9-1. The blue dotted line shows the bottom of the permafrost. The underpressure is sealed by permafrost where the layers come up to the surface towards the Sassen Valley toward the ENE. If thawing or other processes at the bottom of the permafrost in this area create a deficit of water in the pores, it will affect the pressure conditions in the well. The profile in the Fingerdjupet sub-basin (cf. map of depth to chalk), shows schematic Jurassic and Upper Triassic reservoir rocks (yellow), depth to the bottom of the gas hydrate area (red line) and maximum calculated depth to gas hydrate during the last ice age (blue line). The black lines are faults.

 

 

Permafrost and gas hydrate – important geological factors in the Arctic

When water freezes to ice, the volume increases by about eight per cent. Ice that freezes in the pores of a sedimentary rock consequently takes up more space than the water and will therefore create overpressure and frost heaves.

Pingos are a type of gigantic frost heave, large mounds with ice in the core. They are formed where the overpressure bleeds off and ice builds up. If the pore water that freezes contains salt, the ice will be fresher than the initial pore water, which leads to concentration of salt in the remaining pore water.

 

There are a number of pingo structures in the bottom of the valley in Reindalen on Svalbard, some of them have crater-like structures with melt water on the top. The pingos are typically a few hundred metres wide and approx. 30 metres high. Aerial view, source Norwegian Polar Institute, http://toposvalbard.npolar.no/

There are a number of pingo structures in the bottom of the valley in Reindalen on Svalbard, some of them have crater-like structures with melt water on the top. The pingos are typically a few hundred metres wide and approx. 30 metres high.
Aerial view, source Norwegian Polar Institute, http://toposvalbard.npolar.no/

 

If permafrost thaws, the resulting water volume will be less than the ice volume, and  in theory underpressure could occur. Transitions between water and ice in freezing and thawing create much larger volume changes in the pore water than the other processes discussed in this article.

The pressure conditions in the drillings at Longyearbyen are so extreme that it is reasonable to suggest that processes related to permafrost contribute to creating the underpressure, while the sealing properties of the Jurassic shale contribute to preserving it.

The schematic profile above shows a proposal for how underpressure could occur.

If one were to verify this hypothesis, it would be interesting to compare the conditions in Adventdalen with data on pressure conditions under permafrost in Siberia and other areas in the Arctic region.

The Fingerdjupet sub-basin is located at such significant water depths that it could not have had permafrost during the last ice ages, but gas hydrate has been widely distributed.

The figure shows similarities between the two geological profiles with underpressure.

 

The ice on fire

A lump of gas hydrate looks like ice. It has a crystalline structure where gas molecules are trapped in a cage of water molecules. In nature, the hydrate usually contains methane, with a small percentage of heavier components such as ethane and propane. Other gases, e.g. CO2, can also form hydrates. If one were to thaw one cubic metre of methane-hydrate, about 160 cubic metres of methane would be released at atmospheric pressure. Methane hydrate is only stable at high pressure and low temperature. At atmospheric pressure, the gas will quite quickly fizzle out of the ice structure.

Gas hydrate is not stable under atmospheric pressure. The methane gas that is trapped in the hydrate flows out and could ignite.

<< Gas hydrate is not stable under atmospheric pressure. The methane gas that is trapped in the hydrate flows out and could ignite.
(Photo: www.usgs.gov)

 

Naturally occurring gas hydrates have been extensively studied the last 20 years. They are regarded as a potential gas resource, and gas released from decomposing hydrates may be a factor contributing to greenhouse effects. In the early 1990’ies I was introduced to gas hydrates as a geological factor by Anders Solheim who was then a researcher in the Norwegian Polar Research Institute. He had discovered a field with crater structures in the Bjørnøya trough, using bathymetry and shallow seismic data. The km-size craters looked like giant pock-marks but were formed in solid Triassic sedimentary rocks. Anders published his results together with his coworkers in 1993, suggesting that the craters were formed by blow-out of gas from decomposing gas hydrates. This hypothesis has been supported by later investigations, and much larger crater areas have been discovered. This is described by Karin Andreassen and a number of co-authors in an article in Science from June 2017.

At locations with considerable water depths where hydrate is stable on the seabed, pingo structures will form, analogous to permafrost pingos. If pressure and temperature change, the gas hydrate core of the pingos will decompose and the result is a crater at the sea floor.

In the Bjørnøya trench, with a water depth of 400 metres and a temperature at the seabed that is nearly 0 degrees, methane hydrate is stable in the sediment at a depth of 600-700 metres. At maximum glaciation during the last major ice ages, the glaciers calved out at the edge of the shelf on the entire Norwegian shelf. The water pressure under the ice was great enough for gas hydrate to remain stable at the sediment surface.

Consequently, the crater fields suggest that large volumes of gas hydrate existed in the Bjørnøya trough, possibly limited to areas with significant methane seepage.

 

Gas hydrates and volume change

One of my colleagues, Oddbjørn Nevestveit, made a calculation of what would take place in a shallow gas field if temperature and pressure change so that gas hydrate becomes stable. Methane and water in hydrate form take up less volume than methane and water separately. When the gas converts into hydrate, the pressure will drop dramatically if the gas field is located in a reservoir volume with poor communication to the seabed or major groundwater systems.

 

The curve shows the volume occupied by water and methane when 1 m<sup>3</sup> methane is dissociated, at different burial depths. Vertical axis shows depth below sea level in meters. Methane hydrate is stable in deep, cold water settings, in sediments down to a few hundred meters below the sea floor. Assumptions: Pure methane hydrate, pressure increase 1 bar/10 m, no methane dissociation in water.

The curve shows the volume occupied by water and methane when 1 m3 methane is dissociated, at different burial depths. Vertical axis shows depth below sea level in meters. Methane hydrate is stable in deep, cold water settings, in sediments down to a few hundred meters below the sea floor. Assumptions: Pure methane hydrate, gas expansion factor equal to depth divided by 10, no methane dissociation in water.

 

Such calculations indicate that large losses of fluid volume, and consequently- underpressure could be caused by gas hydrate formation. In the Fingerdjupet sub-basin, gas hydrate has existed in larger volumes, but it decomposed once the ice age ended. If the reservoir volume were completely sealed the entire time, the volume of gas and water would have returned to the starting point when all of the hydrate had thawed. If the pore volume in the sediment were then also preserved, the pressure would return to normal. But such complete sealing and preservation of pore volume is hardly realistic. The crater formation further east shows that large volumes of fluids escaped from the system, and it seems likely that underpressure in the aquifer could be preserved. More mapping and more complex models are required in order to understand more about these processes.

There is little data on the effects that formation and thawing of gas hydrate have on pressure and liquid flow in the subsurface. The processes are of significance for those examining the possibilities for recovery of gas from gas hydrate and for exploration and production of hydrocarbons in areas that have been exposed to glaciation. And – most likely – for those looking for explanations to the underpressures in the north.


Topics: Geology