Pressure pipe systems often require bends to negotiate changes in direction. The laws of physics dictate that the pressure and mass of the fluid exert an outward force on the bend. This is known as ‘thrust’ and the energy thrusting the bend outward is known as a ‘resulting force’. In water pipes, most of the resulting force is due to pressure, rather than mass and velocity, and it attempts to push the bend outward. If the bend is not to move, then the resisting forces must be greater. In many situations, the forces are so great that it is necessary to configure thrust block anchors, normally from concrete, to counteract the thrust forces and stabilize the pipe. If a pipe is not ‘restrained’, the result of outward movement at the bend is pipe rupture at the coupling. Thrust blocks themselves are expensive to design and build and, due to their own weight, must be built in ground that will support that weight. There are many cases where the thrust block itself moves, causing the same effect: pipe rupture.
How PU Foam Stabilizes Pipe Bends
Closed cell, rigid PU foam is very hard, and can be likened to ‘concrete light’, in terms of its mechanical properties. It is also very light and adhesive, and is often used purely for that purpose. It’s stiffness and compressive strength mean that a normal grown man of 80kg or 175 lbs can stand on a piece of PU foam the size of a cupcake, without it collapsing. The adhesion strength between PU and pipe materials such as GRP is such that a square section measuring 1 square meter or 10.7 sq.ft can support 37 metric tonnes or 81,000 lbs, the equivalent of around 20 average family cars in the US.
This incredible strength and adhesion means that a water pipe bend that is laid in PU foam is glued both to the adjacent pipe sections and into the surrounding soil. Apart from stabilizing the pipe itself, the load spreading capacity of the foam means that the thrust forces are spread over a much greater area, before transferring the thrust forces to the surrounding soil, whilst adding no appreciable weight to the system. Finally, the compressive strength mobilization of the PU means that it will tolerate an extremely high weight of backfill and resist dynamic loading, for example from vehicle and foot traffic.
The VelociFoam thrust block system is very much quicker and more cost effective. PU foam for an anchoring application can be produced at more than 22kg or 50lbs per minute, which is the equivalent of 0.5m3 or 18 cu.ft., or more than a ton of typical concrete. This speed of installation means less disruption to the pipe installation area, and lower terrain / area utilization for building materials, as the PU foam components are part of the mobile installation system.
Validated Results – Perfect Performance
The VelociFoam PU foam thrust block system was used for the first time in the Adamselv project for Statkraft.
We chose to partner with the hydropower faculty at NTNU, one of the world’s leading energy research institutes. Professor Leif Lia is recognised as one of the most acknowledged experts in the field of hydropower. Under his tutelage, the faculty conducted detailed analysis over a period of 3 years, both of the properties of the VelociFoam PU foam itself, and its mechanical behaviour in the field. This independent academic validation is invaluable.
Two 15° bends were installed, with one being wired with high-precision HEP displacement sensors:
The plant was activated in October 2016 and has operated through two full seasons. It is worth noting that the conditions of the Adamselv project are very challenging; the penstock undergoes many loading / unloading cycles, being buried in ground that is often waterlogged and that experiences extreme temperature swings; Adamselv is located in the Arctic wastes of the far north of Norway, where temperatures range from -40°C to +25°C.
The displacement results of the system over two seasons were impeccable. There was no more than 0.2mm of displacement, which was approximately 20% of the 1mm in the estimated results from the work carried out by the MultiConsult engineers in early 2016. The following graph shows the results:
An independent report into the results from the project were published and presented by Professor Leif Lia of NTNU, and can be downloaded here: report.
The calculation model from MultiConsult is available here: bend calculation model.
VelociFoam’s Analysis of Results
Using the validated results, it is clear that the PU foam – bend system performed better than the initial calculations; this was to be expected. In order to comply with the current regulations from the Norwegian regulator, NVE, the engineers must design the system to be ‘static’. At this early stage in the development of the technology, we chose to ‘allow’ an uncontested and universally-accepted static model as the basis for the calculations, in order to be as conservative as possible. The aim of the project was not to produce as accurate a model as possible, but to demonstrate that the system was safe and predictable. That the system performed 500% than calculated was due to the limited number of inputs that were allowed for calculation of the forces resisting the resultant force at the apex of the bend. As stated by Professor Lia and Tor Oxhovd, as well as ourselves, the ‘forces in the bends are distributed away’ through the foam mass. In a horizontal bend (i.e. a bend where the resultant forces are acting in the x-axis, parallel to the ground), the forces in the bend will be resisted by a greater number of systems, with greater size, than MultiConsult used to calculate the behaviour of the system. For example, classic bend force calculations use an ‘allowable length’, because this system used unrestrained pipe joints, meaning that under the classic system, the pipe bend would slip out of the joint to the next pipe section(s). The PU foam adheres very strongly to the surface of the pipe sections, across the joint, meaning that the bend is in effect much longer than the ‘allowable length’ used for the calculation: it is as long as the foamed section of pipe. In a horizontal bend, other resistance is expected from the interlocking interface between the PU foam and the supporting soil underneath the bend. And so on.
The limitations of the static model are known to all parties.
Results from a Dynamic Model
Using our own dynamic model, which takes into account a much greater set of values, including the plastic nature of the PU foam mass, we have seen that standard PU foam can replace concrete thrust block anchors in all but the very extreme of situations.
As an example that represents a fairly extreme scenario in hydropower, we can use the following data:
- Pipe: DN 900 (35”)
- Pressure: 30 BAR (300m head) & 2.5 m/s flow
- Bend Angle: 30° Horizontal Bend
- Resultant Force ‘z’: 1,003 kN
- PU Foam dimensions: 300mm foam thrust block depth
- PU Foam density: 45KG / m3
… with the following ‘bend layout’:
… which gives us the following results:
This means that the maximum movement in the pipe joint is 3.26mm, which is well within tolerance.
Note that these calculations use a standard PU foam formulation, whose most important value is the density. For higher pressures and greater resultant forces, or softer soil types, it is practical and easy to increase the density of the foam around the pipe bend. For example, increasing the density of the foam to 90KG / m3 reduces the maximum movement in the pipe joint to less than 1mm. We can safely and practically increase the density of the foam to 200 KG / m3. The relatively small amount of foam in the bend system means that the increased cost of foam at the bend itself has a negligible economic effect on the project.