One of the most critical operations during well construction is the cementing procedure, where drilling fluid is displaced by cement, normally with one or more spacer fluids in between. Due to the curing nature of the cement slurry there will be only one opportunity to cement the well properly. Although one for top hole cases can fill cement in from the top in a remedial operation, this possibility cannot fully compensate for a non-optimal initial cement job. Furthermore, it cannot be applied to other well sections. In those sections, complex squeeze cementing operations may be necessary. Consequences of improper annular cement can be leakage during production phase and extensive costs when the well is to be plugged for abandonment after the production phase. To ensure that the risk of poor cement is minimised it is important to use the best procedures to place the cement properly. Most models in use assume that the annulus is homogeneous. This is not always the case since washout sections appear during drilling. The effects of these on cementing are not sufficiently studied and considered in models and procedures.
Here we present and discuss results from fluid displacement experiments in a laboratory flow loop, illustrating annular displacement of drilling fluid by spacer (or spacer by cement). Model fluids with realistic densities and rheological properties have been used in a test setup with a transparent annular section. The wellbore is represented by a 10 m long test section, where the annulus has a 6,5” outer diameter and an inner string of 5” that can rotate. A washout section is represented by a 2 m long section of the outer pipe with a larger diameter of 11”. These diameters are representative for the lower parts of a well were high wellbore inclinations are common. In these sections the inner pipe cannot be assumed concentric at all times, so both concentric and eccentric positions have been tested. Experiments reported here were conducted at 60 degrees inclination. The test section was instrumented with conductivity probes in an array around the perimeter at 4 separate positions along the pipe, including the inlet and outlet of the washout section. Together with a camera along the test section, this provided information about the motion and shape of the liquid-liquid interface through the test section.
Results show that the displacement front changes significantly when entering the washout zone compared to the regular annular section. Due to the larger flow area the density differences between displaced and displacing fluids become more important in the washout section, while momentum effects dominate in the regular section.