Analytical predictions of dynamic top tensions of disconnected drilling risers, in which the contents are retained, are often performed assuming that the mass of the riser contents ("mud") is simply lumped onto the bottom of the riser. This approach ignores the compressibility of the trapped fluid column.
The equivalent axial spring constant of a fluid column that results from compressibility has been derived. The derived spring constant is then employed in both time domain and frequency domain axial riser dynamics computer programs to study the effect on dynamic top tensions for both the lumped mass approach and one that treats the fluid column as discrete spring-mass degrees of freedom.
The resulting dynamic top tension predictions are developed. Finally, it is concluded that predictions developed using the lumped mass approach may be in error and non-conservative, particularly for risers greater than about 2000 feet in length.
During emergency storm conditions, it may be necessary to disconnect a drilling riser at the sea floor and suspend it from the heaving vessel. Operational procedures may dictate that the contents of the riser (usually dri11ing "mud"), be preserved and therefore retained in the riser during storm hang off. This procedure may even be employed by some in order to increase the riser wet weight.
Analyses of the "hang off problem" are usually performed to make certain that the dynamic top tensions resulting from riser hang off remain within bounds usually dictated by cable slack or riser buckling on the lower limit, and overstress of the riser or related equipment on the upper.
When the contents are trapped in the riser during hang off, the contents become a part of the heaving system and thus will contribute to the top tensions resulting from vessel heave motions. Riser analysts have long recognized this and have typically simulated the tension contribution of the contents by assuming that the mass is attached to the bottom of the riser. This approach effectively makes the riser more massive and therefore increases the top tensions required to force the riser to adhere to vessel heave motions. However, the mud column actually has flexibility, as well as mass, and cannot be accurately treated as a rigid lumped mass for many cases.
The goal of this work is to demonstrate predictions of dynamic top tensions by simulating the mud mass in two ways which we will call "rigid" and "free" and to show that the two methods, in general, do not agree.
The "rigid" method will employ the mud mass by summing it with the mass of the last element of discretized spring-mass models of sample drilling risers. The "free" method will simulate the mud mass as a system of discretized springs and masses which interface with the lower end of the riser.
