Uncovering the corrosion mystery

By Marianne Bom

Corrosion is a major cost for the oil industry, and especially one form of corrosion is an expensive troublemaker in the production wells in the North Sea: The under-deposit corrosion. So far it has been impossible to predict and prevent. Two teams of researchers at DTU now collaborate with Maersk Oil on uncovering the mystery.

You can see the scale formed as a small bulge, but just from the look of it you cannot judge what is going on beneath the surface of a steel tube suffering from under-deposit corrosion. Sometimes the steel is relatively unaffected, sometimes it has almost corroded away. If you mill out the scale as the oil industry does, to keep the oil producing tubes open, there is a risk of exposing a hole. Any leak from the production tube due to penetrating corrosion is costly and dangerous as a barrier is lost and has to be repaired.

“We spend a lot of money and time trying to reduce corrosion of steel tubings used for oil production. Under-deposit corrosion is a big issue to us. It is something of a mystery, since we cannot predict when it is going to happen. We have wells producing 20 to 30 years without changing the tubes, but others corrode after two or three years of use. We would really like to understand why this is happening,” says Hans-Henrik Kogsbøll, Production Technology Discipline Lead at Maersk Oil.

During the years, Maersk Oil has tried different strategies to prevent corrosion in the oil fields of the North Sea. One has been to change the steel alloy types used for the tubes.

“We have used both steel types with less and more chromium and both failed,” says Hans-Henrik Kogsbøll, frustrated that no clear cause and effect pattern has been drawn.

That is why Maersk Oil, DTU Mechanical Engineering, and DTU Chemical Engineering have joined forces in a common DHRTC project aimed at getting a better basic understanding of under-deposit corrosion, and how it happens in the well. It is the intention that the new understanding will lead to knowledge on how it can be controlled in the future.

“Normally, when you want to fight corrosion in a simpler set up, you add baking powder (NaHCO3) to get the optimal pH-value to prevent corrosion. Maybe, in the future, we can do something similar in the production wells, although it is off course complicated to inject a scale inhibitor into a one-way system like a well bore,” says Philip Fosbøl, Associate Professor at DTU Chemical Engineering.

All sorts of ideas occur when researchers and people from the oil industry brainstorm on this issue: To use a special coating of the tubes, to change the temperature and/or the pressure in the wells, or to add a scale inhibitor in the injection wells. Future brainstorming may be on more solid ground due to knowledge gained from the joint research project at DHRTC.

The research

The research is done in two tracks to become combined at the end.

The team at DTU Mechanical Engineering have selected 20-25 scaled tubings from 14 Maersk Oil wells in the Halfdan, Gorm and Dan Fields. The selected tubings are statistically representative of the wells in the three oil fields. They are collected from different well depths, and represent different materials and well history.

“We analyse the internal part of the tubing for the nature and composition of scale layers at extreme detail, and we identify the corrosion mechanisms and scale formation that have taken place progressively over the years,” says Professor Rajan Ambat, DTU Mechanical Engineering. “Initially, the inner part of the tube will be covered with the scale on the top. We analyse the layered structure of the scale to determine the types. Then we remove the scale by chemical means and look into what is underneath on the tube. In this way, we can see the under-deposit corrosion, how it was formed, what is the material, and how is it different from well to well.”

The scale layers can be correlated to well history in a similar manner to the way we calculate the history of a tree (dendrochronology) from the growth rings when we cut wood. We have called this scalochronology.

Scale analysis is a very time-consuming work. To analyse one tubing requires one full month. Having worked on this through 2017, the researchers now have a clear picture of the distribution of the scales between the fields, and why the scale happened across the fields due to different well condition, the materials used, and the depths of the wells.

Recreating history

The team at DTU Chemical Engineering is studying the exact same wells in a different way. They have received chemistry data from Maersk Oil on the produced fluid from the wells over the years, in some cases going back to the mid 90’ies. Now their task is to recreate history. Based on their knowledge on chemistry, thermodynamics, and corrosion, they strive to predict what kind of corrosion has taken place in the tubing by using and improving the ‘SCALECERE’-model developed by Associate Professor Kaj Thomsen, DTU Chemical Engineering. Finally, they compare their predictions with the actual scale identified at DTU Mechanical Engineering to see if there is a match.

“Probably we will have to do the iteration back and forth at least five or six times to say: ‘now we are well aware of which type of scaling is being present when a well is producing a specific fluid’. Then we have understood what goes on,” says Philip Fosbøl.

The research teams expect to present the first conclusions by the end of 2018.

Under-deposit corrosion

Under-deposit corrosion represents one of the most damaging forms of corrosion to a tubing system, compared to normal acidic corrosion. It is typically very aggressive, causing deep penetration of the metal surface with reduced general corrosion in the surrounding areas. This is most likely initiated by a combination of various mechanisms: electrochemistry, scaling kinetics, crevice corrosion, reaction kinetics, thermodynamics and possibly other phenomena like microbial induced corrosion.

Cross section of scale

The picture shows a cross section of scale on a production tubing from well, showing different layers of scale growth due to corrosion and precipitation from fluid.

The layers tell the chronological history of the changes in the well similar to rings in the wood layers, the scale-chronology. This scale is approximately 1.5 mm thick.

On top of the picture you see the light gray tube and the scale, that was first formed. The bottom part is the newest scale.