PIPELINE WALKING

 

BACKGROUND

High-temperature and/or high-pressure subsea pipelines exhibit the phenomenon of thermal walking, also called crawling or ratcheting. It is axial shifting of short pipelines due to the passage of a thermal transient, usually during start-up of flow, or due to another asymmetric load in combination with pipeline heating or pressure change. A long pipeline that is buckled along its length may behave as a chain of short pipelines within which each segment walks independently. Walking on flat bathymetry usually proceeds from the end where the hot fluid is introduced towards where it flows. Other factors that influence the direction of walking are the slope of the sea bottom and the application of external axial loads such as bottom tension at the touchdown point of a catenary riser.

 

Why Does Walking Occur?

The earliest recognized cause of walking is the passage of a thermal transient along a pipeline, as for instance when hot fluid from a well first traverses a flowline that is initially at ambient temperature. As the hot front progresses, the pipe rises in temperature behind the front and expands towards the free end of the pipeline, the direction of least resistance. As the expanding pipe slips on its supporting soil, the soil resistance builds up compressive stress in the pipeline increasing with distance from the free end so that the maximum compression is at the front of the transient. In order to maintain equilibrium the cold pipe just beyond the heated portion must resist with a corresponding compressive load, achieved by slippage of the neighboring cold pipe in the downstream direction until the frictional build-up in the cold portion matches that in the hot portion. As long as the slippage zone does not reach the far end of the pipeline the pipe stays in place. However if the pipeline is short enough the slippage zone will reach the end of the pipeline when the thermal front is midway down the route. After this point the further heating produced by the continuing hot fluid will cause expansion of the cold half. The hot part will have expanded due to temperature increase reduced by the compression due to soil resistance. The cold half will have already compressed but with the static location being at the far end so the equilibrium point will have shifted into the cold zone by that amount of pre-compression.  After that time the cold part heats up and thermally expands but no further elastic compression occurs. Thus the middle of the pipe has moved by the amount of pre-compression in the previously cold half of the pipeline.  During a later cooling part of a cycle the cooling is essentially uniform over the length of the pipeline so the corresponding contraction is symmetrical and no shifting occurs in the cooling part of the cycle. Thus, over an entire cycle there has been a movement from the hot to the cold direction equal the amount of the pre-compression induced in the cold half just before the front passes through that location. The mechanism of walking is thus greatly dependent on the soil resistance because it determines the accumulation of strain within the pipeline and on the shape of the thermal front. 

A second cause of global movement of a pipeline is slope in combination with temperature or pressure change. This does not require a transient; it occurs even with uniform rise or fall in temperature. The driving mechanism is the active component of gravity in the direction of the pipeline. The slope that causes movement can be remarkably small.  Its influence is not so much due to the magnitude of the force as the imbalance it brings about in the larger force system it modifies. A uniform temperature change or pressure change causes the pipeline to expand or contract and thereby allows slippage to occur between the pipe and the soil supporting it. If the pipe is on level ground, frictional force acts in opposite directions on either side of an equilibrium point in the middle of the route and thus the accumulated strain is balanced.  The superposition of the downhill component of pipeline weight on top of the frictional force shifts the equilibrium point uphill. Thus a longer distance is available to expand on the down-hill side of that equilibrium point than above it. Similarly during the contraction part of the cycle the equilibrium point shifts downhill so that it cannot contract as much as the uphill portion. Thus during both halves of the cycle the pipeline effectively moves downhill.        

Movement into a steel catenary riser represents a third common cause of walking. In this case the end force once again shifts the equilibrium point about which expansion and contraction occur so that the thermal or pressure induced expansion acts on different lengths of pipeline and thus causes a different amount of end movement.

 

How is it Currently Managed?

In shallow water, walking has been restrained by rock dumping, but in deep water such methods have been impractical. Instead, anchors, usually suction piles, provide the restraint to longitudinal movement. These anchors may also serve as an aid to pipeline installation. In most cases, the piles have been located at one end of the pipeline segment, the higher or predicted hot end, however there have also been conceptual designs with the pile near the middle of the pipeline.

Piles incur significant cost both in terms of fabrication onshore and installation offshore. In addition, they constrain the subsea architecture and congest the environs around subsea trees and manifolds.  

 

AN ALTERNATIVE

The proposed alternative is the interposition of a subsea constant-thrust (CT) device between a point on a pipeline and a foundation capable of resisting horizontal load. Its purpose is to eliminate thermal walking while reducing horizontal loading to the foundation.

 

Why Does Imposition of a Constant Load Control Walking?

Purposely applying a constant load acts through a similar mechanism as do the causes of walking by shifting the equilibrium point within the pipeline route; however it is sized and directed to balance those other effects. The load is chosen to cancel the movements incurred by the other effects mentioned above so that there is negligible total movement over the cyclic history of the pipeline. A smaller load may be chosen to limit the extent of walking rather than completely cancel it.

 

The Deepwater Embodiment

In deepwater, one method for application of near constant force is using a piston-in-cylinder in which the chamber is evacuated or contains gas at a pressure that is negligible compared to the hydrostatic pressure in the depth of water in which it will reside, and in which the driving force is provided by the external pressure of the ambient ocean water. In practice, one atmosphere fulfills the negligible internal pressure requirement for a deepwater installation. The area of the piston is sized based on the external hydrostatic pressure to produce the correct required longitudinal load. For an evacuated cylinder, the pressure differential will remain constant irrespective of stroke. If the chamber is left with one atmosphere, this gas will compress through the stroke, however the pressure and pressure change will be minor compared to the unchanging external pressure and thus the total will be close to constant.

 

Where to Put It?

The location, deployment, and use of a constant-thrust device will depend largely on the force level to be achieved. Whether used at a pipeline end with a PLET or along the route with an inline sled is a matter of convenience and competing demands of the subsea architecture. When the thrust required is moderate, as with a PLET for very short lines, the constant-thrust device would preferably be interposed between the sliding unit of the PLET and the PLET’s foundation part. For higher longitudinal loads that challenge foundation stability, the constant-thrust device would be interposed between the PLET sliding portion and the fixed point on a pile or similar anchor external to the PLET. In that case it can be part of the pile top itself, or included in the rigging between the pile and PLET, or embedded in the hook/yoke. The constant thrust eliminates cyclically reversing loading to the foundation as well as reducing its magnitude.

 

Analogous to PLET ends, the constant-thrust device may be inserted between the sliding and stationary components of an inline sled or between a point on the pipeline that moves relative to an external anchor that is stationary. If an inline sled is already required for independent reasons such as support of a heavy component or restraint of torsion, it can serve to also provide longitudinal restraint. Since inline sleds usually do not slide relative to their pipelines, a sliding mechanism must be included to decouple the pipeline longitudinally from its foundation to allow insertion of a constant-thrust device between them.

 

How Does It Affect Installation?                

During installation of a standard PLET, the sliding portion is locked in position, usually at the cold and unpressured position, and released only after the PLET is on-bottom and the continuing pipeline is laid. The same is the case when utilizing a constant-thrust device. If utilizing a cylinder type constant-thrust device, when the PLET is on the lay vessel, the gas within the cylinder will equalize with the external atmosphere if left at atmospheric pressure or close to it and thus the piston will not be energized and the sliding portion of the PLET will be in a neutral position. Similarly, if the cylinder is evacuated, the pressure difference and thrust generated will be small and will not overcome the friction in the PLET sliding mechanism. Once the PLET is on-bottom, the hydrostatic pressure of the seawater will load the piston, thus the locking mechanism must be able to release in a controlled manner, as the sliding portion will attempt to move forward and begin stretching the adjacent length of pipe. This can easily be achieved using a two-chamber CT device, and opening a flooding valve to the pressure-side.

                                                                                                                       

An Example

Direct comparison with a real world example is a good measure of the benefits that can be achieved. Consider a 12" flowline running 40km from a subsea development in 1100m water depth uphill to a platform in 400 m water depth. There is a relatively level approach extending 8 km from the platform preceded by a curve. To avoid excessive buckling strains, buoyant engineered buckles are to be introduced at 2- 3 km spacing. The buoyant sections break up the pipeline into segments making walking possible that has been calculated to be 10 m. Applying the alternative method for this configuration there will probably need to be a constant tension unit on each segment, however the load to be handled by each will likely be small enough to be handled by a small clamp-on inline sled. A good location for such sled is near the top of a hill just a short distance down from the nearest buckle. Before the initial buckle forms the preload will be almost as much compression on the up side as tension on the restrained side because the pipeline stiffness would be similar. Once the buckle forms that side would soften until the bulk of the constant thrust is applied as tension to the top of the hill.