Oil in ancient caves poses new challenges

In 2013 and 2014 Lundin Petroleum discovered significant amounts of oil and gas in the prospects Gohta and Alta on Loppa high, north of the Snøhvit field.

The reservoir type encountered represent something new on the Norwegian shelf: carbonate and gypsum formations with evidence of pervasive dissolution and cave formation (commonly known as "karst") followed by infilling and collapse during subsequent burial.

Similar reservoirs are known from the Middle East, China and the US.

"Such "paleokarst reservoirs" are quite intriguing from a scientific point of view, but far more difficult to characterize than the reservoirs we are familiar with from elsewhere on the Norwegian shelf. The processes acting during their formation produce extreme contrasts in porosity and permeability over short distance," says Tveranger. A detailed understanding Drill holes and seismic surveys provide only limited information in this context.

"Thus forecasts of spatial distribution of formation properties, a prerequisite for successful well placement and production planning, have to employ other means. A detailed understanding of the processes these reservoirs are subjected to during their formation is the key to their successful development," Tveranger says.

"Predictable reservoir behaviour ensures that optimal recovery with a minimum number of wells. This cuts both installation cost and risk for operators. The need for fewer wells and prolonged operational life for individual fields is also less environmentally disruptive," he adds.

The size of the Alta discovery is promising. Preliminary resource estimates based on wells drilled by the Island Innovator rig in the autumn of 2015 suggest reserves of between 14-50 million standard cubic metres of oil and 5-17 billion cubic metres of gas, according to offshore.no.

Lundin will continue the evaluation of the reservoir with two new assessment wells planned for 2016.

Could the neighbouring prospect on Loppa High, Lakselv, Børselv and Neiden contain similar reservoirs? According to Tveranger, an exploration well is in progress on the latter.

Unique combination

Unusual reservoirs require unusual solutions. This also extends to team composition.

Uni Research CIPR, the Department of Earth Science at the University of Bergen (UoB) and NORSAR have joined some of their competences in an effort to improve characterization of paleokarst reservoirs.

This includes Norway's premier research environment on karst processes, extensive experience in reservoir modelling and simulation and a suite of new tools for seismic forward modelling.

"Simply explained, we start out by establishing what different karst systems look like. This is linked to the initial properties of the rock as well as local tectonics and climate. Knowledge of the processes acting during subsequent collapse, infill and burial can be used to forecast property distributions in the subsurface. This is followed by seismic forward modelling to see what a given system will look like in seismic data from real reservoirs," says Tveranger.

A long-lived research effort

Uni Research CIPR is one of small number of Norwegian research institutions who have worked on this type of reservoirs for a long time through field studies in Texas, Wyoming and Billefjorden on Svalbard.

"Students and researchers at CIPR have built reservoir models and conducted fluid flow simulation studies based on these field studies since 2008," Tveranger points out.

Lundins discoveries in the southern Barents Sea have been described as a potential game changer for the petroleum production in the region.

Although the company in November 2015 expressed their intention to develop the two discoveries, falling oil prices emphasize the need for reliable forecasts to lower investment risks.

Composing a team

Norway's foremost karst expert, Professor Stein-Erik Lauritzen from the University of Bergen is part of the research team.

"Cooperation with UoB is a key element. The link between modern karst and paleokarst reservoirs is well known, but the disciplines working in these two fields (karst research and petroleum reservoir characterization) traditionally don't mix much," says Tveranger.

"However, our close collaboration with UoB allows us to set up untraditional teams in a targeted manner to handle specific problems, or simply explore the vast scientific ground between normally compartmentalized disciplines," Tveranger points out.

Seismic forward modelling

Seismic interpretation of paleokarst reservoirs is challenging. The processes forming them give rise to geometrically complex structures and substantial heterogeneity.

It is difficult to identify and interpret these structures if one doesn't know what to look for.

Here seismic forward modelling of modelled structures from known paleokarst reservoirs can aid interpretation.

"Our collaborators at NORSAR have developed a tool for seismic forward modelling of the reservoir models we build. This will make it possible to identify what different structures look like in the seismic data and thus give the interpreter an idea of what he or she should look for."