If we can succeed, in a simple manner, in separating the produced water from the crude oil in the well itself, then we can drastically improve the economics of many oilfields. Hydros H-SepTM concept aims to achieve this by exploiting the natural gravitational separation in a horizontal section of the well. The water settles to the bottom of the casing while the oil floats on top, so in principle it is merely a matter of tapping from top and bottom at the right speed. Together with its partners Kværner Oilfield Products (separator) and Weir (downhole pump), Hydro is aiming to install this in a test well on the Brage field early in 2001.
Measurement can grant success
If the tapping is not done just right, an imbalance will slowly accumulate, and the separator will fill up with either water or oil. The key to guiding the oil and water stream out of the separator will therefore be to measure the level of the phase transition between oil and water.
SINTEF Instrumentering has worked for many years to develop sensors for similar problems in oil wells, and was commissioned by Kværner to develop a level sensor for the H-SepTM project.
Interdisciplinary approach in practice
Many physical principles can be exploited in order to find the position of such a phase transition. One of these is to use the big difference in electrical properties between oil and produced water. Both the conductivity and the dielectrical constant of water are much bigger than those of oil, allowing for several useful methods. The most important thing is to find one that is simple and robust and is not affected by the downhole environment. Kværner had already decided to exploit a capacitance measurement principle, and since SINTEF had long experience with capacitance sensors in oil wells, it was this method that was chosen. However, the measurement method is only one of very many details that have to be right before a sensor in a well environment is a success. Kværner has set a standard of a five-year life and a maximum temperature of 175oC. If this is to be achieved, the sensor needs to be extremely robust to withstand high pressure, acidity and high temperatures. The demands made on the materials are formidable. The chosen solution is based on a ceramic packaging together with duplex steel and Inconel. Inside are electronics specially developed to withstand high temperatures over a long period. Our development project so far has experience with more than 2000 hours of operation at 175oC for critical components of this sensor.
Many opportunities to fail
Sensors in well environments do not have a particularly good reputation as regards long-term reliability. This is true regardless of who has supplied the them. Damage to cables with leakage as the result is perhaps the commonest failure. The H-SepTM project is based on a commercially available communication system. Between the level sensor and the communications system run Inconel tubes welded to the sensor capsule. Seal testing under high pressure and fluctuating temperatures are a major part of the quality assurance. Another important test is the combination of high temperature and powerful vibration.
It is not only “catastrophic failure” it is important to test for. Capacitive principles are based on the field distribution around one or more measurement electrodes, which alters when water comes up the “windows”. How this field distribution relates to changes in the fluid phases’ movements along the sensor needs more study.
In Nansen’s footsteps
When Frithjof Nansen was exploring the Arctic Ocean, he encountered “dead water”. His ship “Fram” moved as if in molasses. As the good scientist he was, Nansen did not content himself with mythological references, but investigated the phenomenon. The explanation was underwater waves. Under certain conditions, water of different densities layered itself, and the interfaces between the different layers can create waves as the ship passes. It was the resistance put up by these underwater waves to the “Fram” that explained the monstrous grip on the ship.
The same thing happens with the interface between oil and water deep in the well, though more a matter of fire-demons than water-monsters. And this wave formation affects the measurements. In the same way, different surface tensions of the fluids will create errors, not to speak of the sensor housing’s moisture characteristics. More knowledge of the modification of the surface properties of ceramics and more hydrodynamic calculations are therefore important details in the development process.
The prototype works well
In order to test the separation principle under the most realistic possible conditions, a prototype of the separator has been run together with a prototype of the sensor in a high-pressure loop at Hydro’s Research Centre in Porsgrunn. The results are very encouraging. The meter works well, and the separation functions as planned in a closed loop in which the level measurement is used to regulate the flow of oil and water out of the separator.
