Deepwater Production

Effective riser tension: Fact or fiction?

Be sure your analytical methods comply with elementary physics

Jack H. Bayless, Consultant

The author, with 44 years of industry experience, analyses complex oilfield practices using elementary physics. When a conflict with these basics is evident, he points that out so that improved practices can be developed. This second article – following “Important basic facts regarding riser tension and mud weight,” World Oil, August 2001 – also deals with a complex subject – riser analysis – in an effort to dispel commonly held beliefs that do not comply with elementary physics.

Effective tension. The physical model: For more than 30 years, analysts have used, as a basis for fundamental forces on risers, the concept illustrated in the accompanying figure. The drawing shows: A) a straight riser loaded with heavy mud; B) the same riser with a slight displacement due to a temporary eddy current or temporary vessel excursion; and C) the same riser with a large displacement due to postulated internally generated forces from the hydrostatic pressure of the mud. Casual examination indicates that the side of the riser to the right is longer than the other side of the risers in the other two scenarios. Thus, it seems plausible that this longer side would have more area than the other side.

   The forces: The larger area of the right side would seem to allow pressure to exert more force on the riser to the right. It is postulated that this greater force would cause progressive instability of the riser, requiring large tensions to oppose these forces and stabilize the riser; i.e., the riser would become self-propelled, eventually causing catastrophic failure. As one would imagine, this pipe movement with bending and buckling riser pipe would require large amounts of energy when one considers the massive nature of riser pipe.

   The mathematical analysis: Riser analysts have equations to calculate these forces and have shown mathematically that the “effective tension” must always be positive to prevent riser failure. The amount of tension required is equivalent to the net weight of the drilling mud. There is essentially nothing wrong with the mathematical equations involving triple integrals, and strength of the riser pipe, including provisions for current and wave action, if they comply with Basic Physics – the Conservation of Energy.

Fundamental physics for riser analysis. The First Law – Conservation of Energy: The problem is that the basic assumption is incorrect and the calculated results do not comply with basic-physics criteria. Also, movement and bending of pipe with no energy input is a class of a perpetual motion machine that is impossible from the basic-physics viewpoint.

   Energy balance: If an energy balance is performed on the system, i.e., using conservation of energy concepts, it is obvious that large amounts of energy are required to bend and buckle massive sections of riser pipe. The only source of this energy would be the potential energy in the mud column in the riser itself. But it is not possible for the mud column to drop more than 35 ft due to slip-joint stroke-out from the mid-point, thus delivering a negligible amount of energy to the system.

So, one must conclude that the “effective tension” basis for riser analysis is incorrect. The only way that enough energy would be delivered is that the real tension in the riser pipe itself is insufficient to hold up the wet weight plus a small component of net mud weight. This would put the bottom of the riser pipe in compression. Since the riser pipe is always tensioned with “real tension;” it is impossible to bend and buckle the pipe due to internal pressure.

   Pressure on curved surfaces: It is impossible for a riser pipe to become self-propelled due to forces generated by mud pressure on opposing curved surfaces. A further examination of basic equations involving the forces generated by pressure on curved surfaces –  Mechanics of materials, by Laurson and Cox (1950) – indicates that no net force is generated by opposing curved sections, even though the areas are different. These same equations are used to calculate hoop stress and longitudinal stresses in pressure vessels that have curved sides or ends.

Flex joint analysis. Since no net forces are generated by pressure on opposing curved surfaces, another concept of riser analysis is incorrect. This is the postulated side force on the long side of a flex joint. This postulated force is used by riser analysts to negate side load on the BOP stack at the flex joint due to riser tension applied at an angle. Since no internal force is generated, standard vector analysis should be used to calculate bending moments on the BOP connector due to riser tension. It is obvious that using minimum tension reduces bending on a BOP connector.

Fig. 1. Riser bending due to internal pressure.

Why worry about excess tension? There are several reasons: 1) unneeded tension capacity is not cost-effective; 2) higher tension has higher operating costs; and 3) excess tension causes potential leaks in subsea blowout preventers.

So what tension should be used? This short article is not meant to be a detailed analysis of the forces on a riser. However, some guidelines listed here should be considered. Part of the present computer analysis can be used if based on the correct laws of physics, i.e.:

1. Simple lab models of forces caused by the weight of mud in a drilling riser indicate that about 50% of the net mud weight should be used as a starting point for required tension. Then, additional amounts of tension can be pulled to account for the wet weight of the riser. Then, a safety and efficiency factor of the tensioners would be appropriate as a multiplier (125%).
2. If a current is present, note that forces from net weight of mud in the riser lean into the current at the top and can compensate somewhat for the current. However, operating parameters can be adjusted based on the flex-joint angle, which should always be minimized. And for higher mud weights, no over-pull is necessary for lift off of the LMRP for an emergency disconnect. Note that the LMRP picks up 20% of the net mud weight.

Conclusions: 1) Existing analytical methods for riser tension have caused huge amounts of needless tension capacity and operating expense; 2) excessive tensions have caused excess bending moments on BOP stacks, thus resulting in potential leaks at high pressure; 3) riser equations should be modified and the correct basis used; 4) all mathematical models should be validated by simple lab experiments; and 5) “Effective tension,” as presently calculated, is fiction.