![]() |
![]() |
Stone Indentations in District Heating Pipes Caused by Lateral Displacement of the Pipeline - Experimental Studies by Gunnar Bergström and Stefan Nilsson SP Swedish National Testing and Research Institute,
Email: Nilsson |
ABSTRACT
The strength of district heating pipes with respect to localised loads from stones pressing against the casing pipe wall is of growing concern as the district heating industry is striving towards the reuse of excavated material as backfill. Unspecified backfill material can contain large stones which, when located adjacent to the pipe wall, may induce severe indentations in the casing pipe which ultimately lead to failure by crack propagation through the polyethylene material. Experimental tests were run aimed at studying the interactions between a laterally displaced buried district heating pipe, a stone situated adjacent to the pipe wall and the surrounding backfill. In particular, the indentation caused by the force concentration from the stone and how the risk for indentations depends on the stiffness of the backfill was investigated. It was seen that a moderately stiff backfill is clearly the least beneficial in terms of limiting the indentation depth. In a very loose backfill, the reaction forces on the stone will not be large enough to produce significant indentations. On the other hand, if the backfill is densely compacted and the wedge between the stone and the pipe is filled with compacted sand, there will be no force concentration effect and thus no indentation.
Keywords: district heating pipe, stone indentation, backfill material
Introduction
A fundamental functional requirement on a district heating distribution system is that it should be able to convey the energy carrier to the clients without interrupts and to a reasonable cost. For the pipe, this reduces to secure the medium pipe integrity and minimise the energy losses for a sufficiently long time. The preinsulated pipe construction dominating the market today consists of a steel pipe insulated with a layer of rigid polyurethane foam (PUR) protected on the outside with a high density polyethylene (HDPE) casing pipe. The outer casing pipe must remain water tight to prevent seepage of ground water into the pipe with possible steel pipe corrosion as a consequence. Loads with a potential of damaging the casing pipe when buried in the ground has traditionally not been a design aspect. Instead, the proper function of the pipe has been ensured through strict regulations regarding the laying conditions and the composition of the backfill material. The loads on district heating pipes are quite different from what is usual for buried pipelines in general. This is mainly due to the temperature of the water, which, under certain service conditions, might be as high as 140 °C and varies significantly over the year. The temperature variations of the water will induce thermal movements and forces in the pipeline grid, causing lateral displacements at bends and tees. The use of cushions and other devices aimed at absorbing the thermally induced displacements has been advised at bends and tees, and the maximum grain size of the backfill material has been limited to prevent localised loads.
There is a growing interest in the possibility of reusing the material excavated when the trench is opened as backfill. Molin et al (1997) showed that this would lead to substantial cost savings, mainly due to less need for transportation and material haulage to and from the excavation site. It will also be environmentally beneficial due to less consumption of natural resources of fine sand.
When reusing excavated material, one must take into account the possible presence of large stones. Full-scale field trials with coarse materials carried out by Molin et al (1997) showed that large stones can cause distinct indentations in the pipe. Laboratory tests performed by Bryder et al (1996) showed that the risk for immediate penetration of the casing pipe is small, due to the strength and ductility of the HDPE material. There is, however, a risk that the large strains in the casing pipe wall induced by a deep indentation may initiate a crack which eventually propagates through the pipe wall. It is well known that polyethylene, despite its ductile nature at high stress levels, is highly susceptible to the phenomenon of slow crack growth (SCG), which may lead to a brittle failure, see e.g. Kausch (1987) and Williams (1987).
The interaction between the soil and a laterally displaced pipe was originally studied by Audibert and Nyman (1977) and later by Trautmann and O’Rourke (1985). They concluded that the relationship between the increase in earth pressure on the pipe wall and the lateral displacement can be expressed as
![]() |
(1) |
where xC is the lateral displacement of the pipe and asnom is the nominal earth pressure. snom,u denotes the ultimate soil pressure and xc,u is the corresponding displacement. snom is the average earth pressure acting on the pipe wall in the displacement direction. The numerical values of the coefficients a and b was suggested by Audibert and Nyman (1977) to 0.145 and 0.855 respectively.
However, the combined interaction between the soil, the pipe and a stone adjacent to the pipe wall has not previously been studied in detail. In this paper, a series of experiments is presented, with the aim of clarifying how the depth of the stone indentation, the contact force between the stone and the pipe wall and the overall deformation of the pipe depends on the stiffness of the backfill material.
Experimental
The conditions around a pipeline with an adjoining stone being laterally displaced was studied in a series of experiments. The tests were performed by subjecting a district heating pipe in contact with a stone to a transverse displacement in a sand box, measuring displacements, forces and indentations, Figure 1. In order to simplify the arrangement, the displacement was applied in the vertical direction. In reality, a pipeline is normally subjected to major displacements only in the horizontal plane. The direction of displacement should however be of minor significance.
Figure 1. Schematic drawing of test arrangement
Pipes of dimension DN 65/160 (steel pipe diameter DS = 76.1 mm and casing pipe diameter DC = 160 mm) were used for all tests. The pipes were buried in a dry sand with a maximum grain size of 4 mm. The composition of the sand was in accordance with the directives in the European Standard EN 489 (1995) for the testing of district heating pipe joints. The pipe displacement was effected by the pulling of two steel rods fixed to the steel pipe with the aid of hydraulic jacks. The displacement force was under control by measuring the oil pressure in the hydraulic system.
Figure 2. Artificial stone with built-in force transducer
The stone used to produce the indentation was placed, at the centre of the pipe, in contact with the casing pipe facing the displacement direction. To enable the measuring of the contact force, an artificial stone was moulded using a polyester compound, and a force transducer was built into it. To obtain a realistic stone surface texture, a bandy ball was used for shaping the mould. The body of the stone was spherical in shape with a diameter of approximately 65 mm. The “tip” of the stone, in contact with the pipe wall, was formed by a stainless steel insert with a diameter of 25 mm, Figure 2.
To enable the measuring of the indentation depth, three displacement transducers were mounted at different locations between the casing pipe and the steel pipe, Figure 3. One was placed directly underneath the stone, measuring the maximum indentation of the PUR foam. The other two transducers were placed 50 mm and 250 mm respectively away from the stone, thus measuring the radial compression of the PUR foam caused by the earth pressure increase on the upward side of the pipe.
Figure 3. Displacement transducers mounted on test pipe
The experiments were run with three different types of conditions as regards the stiffness of the sand, Table 1. The variations in stiffness were accomplished by compacting the sand to a greater or lesser extent, and it could be further increased by applying a static pressure to the concrete tiles on the sand surface.
Table 1. Summary of sand box tests
![]() | |||
Test designation |
Fill height, m |
Sand density (kg/m3) |
Surface pressure (kPa) |
![]() | |||
Stiff |
0.3 |
1829 (102 %) |
100 |
Stiff |
0.3 |
1683 (94 %) |
100 |
Moderately stiff |
0.3 |
1684 (94 %) |
— |
Moderately stiff |
0.8 |
1667 (93 %) |
— |
Loose |
0.3 |
1625 (91 %) |
— |
![]() |
The density of the sand during the test was determined by weighing the entire sand box. Figures within parentheses indicate the degree of compaction relative to the maximum dry density as measured according to the Swedish Standard SS 02 71 09 (1975).
Results and Discussion
The measured earth pressure vs. displacement in Figure 6A agrees fairly well with the Audibert-Nyman description (1) and a large difference in stiffness for the three backfill alternatives is seen.
The deformation of the pipe, both as regards the magnitude and the deformed shape, depended to a great extent on the stiffness of the backfill. The photographs in Figure 4 show the profiles of the pipes after completed test runs. The pipes tested in moderately stiff backfill became severely flattened at the side facing the displacement direction, Figure 4 C-D, while the stiff backfill alternatives retained their circular shape, Figure 4 A-B. The pipe displaced through the loose backfill alternative did not exhibit any visible deformations. The difference in deformed shape is the consequence of a difference in how the earth pressure reaction is distributed over the pipe circumference. Clearly, a stiff backfill spreads the pressure more evenly. The flattening of the pipe front, which can be seen for the moderately stiff cases, indicates a more pronounced pressure concentration towards the centre line of the pipe.
Figure 4. Photographs of deformed pipes after completed tests. Displacement direction to the left. (A) and (B) Stiff backfill. (C) and (D) Moderately stiff backfill.
Table 2. Pressure ratio for the different backfill stiffnesses
![]() | |
Backfill |
sstone/snom |
![]() | |
Stiff |
0.16 |
Moderately stiff |
0.50 |
Loose |
1.77 |
![]() |
The force pressing the stone into the pipe wall is a direct consequence of the earth pressure increase due to the lateral pipe displacement. The body of the stone accumulates the earth pressure acting on it, and transfers this onto the point of contact. One may therefore assume that the force on the stone is strictly proportional to the increase in earth pressure, and this was actually confirmed by the measurements. Figure 6B shows the average earth pressure on the stone, sstone vs. the averaged earth pressure on the pipe wall, snom. sstone was calculated as the force measured with the built-in transducer (cf. Figure 2) divided by the area of the stone (Æ65 mm) in the displacement direction. The ratio sstone/snom was nearly constant, with different values depending on the backfill stiffness, Table 2. If the pressure distribution over the pipe circumference is constant or convex symmetrically with respect to the centre line of the pipe, the average pressure on the stone should always be greater than or equal to the nominal earth pressure, Figure 5. This was however contradicted by the results in Table 2. This discrepancy is likely caused by the force measurement and the shape of the stone. If the stone is partly resting on sand in the wedge between the pipe wall and the stone body, the force felt by the transducer represents only a part of the total force acting on the stone. As confirmed by Table 2, This effect would be less pronounced in loose backfill, since the rigid measuring tip of the stone concentrates more of the force if the filling in the wedge is less stiff. Calculations by Achmus (1995) show that the earth pressure distribution on a pipe in an actual pipe trench is not necessarily symmetrical, but may have a maximum slightly towards the trench bottom. However, due to the symmetry of the test arrangement, Figure 1, an asymmetrical pressure distribution is not expected in this case.
Figure 5 Earth pressure acting on pipe and stone
The depth of the indentation caused by the stone for the different backfill alternatives is shown in Figure 6C, as the measured force on the stone vs. the indentation depth. di is the depth of the actual indentation, i.e., the difference between the displacement of the stone into the pipe and the radial compression of the pipe due to the earth pressure as measured far from the stone. The dashed line in the diagram indicates the measured response from a deformation controlled single stone indentation, i.e., with no backfill surrounding the indenter, previously reported by Molin et al. (1997). In the stiff backfill, the indentation was close to zero. This indicates a lack of force concentration caused by a complete sand filling in the wedge between the stone body and the pipe wall. The indentation depth in the loose backfill, on the other hand, followed largely the path of the single stone indentation curve, indicating an almost complete concentration of the force onto the point of contact. The moderately stiff alternative was clearly an intermediate case also as regards the situation around the indentation zone. The final plot, Figure 6D, shows the consequence of a lateral pipe displacement in terms of indentation depth, i.e., di vs. the steel pipe displacement, xS. Due to a faulty position indicator, the “loose” curve becomes vertical at xS ≈ 11.3 mm. In reality, the steel pipe displacement continued to grow, and the maximum indentation depth of 1.3 mm actually occurred at a displacement of roughly 50 mm. The stiff alternative never reached a greater steel pipe displacement than approximately 10 mm due to the large reaction forces developed and the limitations in load capacity of the test equipment.
Figure 6. Results from measurements: (A) Nominal earth pressure snom vs. transverse casing pipe displacement xC. (B) Average pressure on stone sstone vs. nominal earth pressure snom. (C) Contact force P vs. indentation depth di. (D) Indentation depth di vs. transverse steel pipe displacement xS.
Conclusions
The earth pressure induced from a lateral motion of the pipe is highly dependent on the stiffness of the backfill. A loose backfill will allow the pipe and the stone to move together under a small pressure reaction yielding only a minor stone indentation. In a backfill with very high stiffness, the flexibility of the foam insulation will cause the steel pipe to deform the foam essentially without moving the casing pipe, which implies no force concentration and thus no stone indentation. In a moderately stiff backfill, at a compaction to normally used levels, substantial indentations were observed during the experiments. This is a consequence of the successive compaction of the backfill during the transverse motion of the pipe, the resulting earth pressure acting on the stone and the concentration of the force onto the point of contact. Thus one may conclude that the moderately stiff backfill material is the least beneficial in terms of limiting the indentation depth. A very loose backfill will never be able to accumulate a large enough reaction force to produce any significant indentation. On the other hand, if the backfill is very densely compacted and the wedge between the stone and the pipe is filled with compacted sand, there will be no force concentration effect and thus no indentation.
Acknowledgements
The research work presented herein was funded by the Swedish District Heating Association (Svenska Fjärrvärmeföreningen), the Swedish National Energy Administration (Statens Energimyndighet) and SP Swedish National Testing and Research Institute (SP Sveriges Provnings- och Forskningsinstitut).
References
1. Achmus, M. (1995). “Zur Berechnung der Beanspruchungen und Verschiebungen erdverlegter Fernwärmeleitungen“, PhD Thesis, University of Hannover, 1995.
2. Audibert, J. M. E. and Nyman, K. J. (1977). “Soil restraint against horizontal motion of pipes”. Journal of the Geotechnical Engineering Division, 103(GT10), pp. 1119-1143.
3. Bryder, K. L. et al. (1966). “Casing of preinsulated district heating pipes - Functional requirements”. Research Programme of the Danish Ministry of Energy.
4. European Standard EN 489 (1995). “Preinsulated bonded pipe systems for underground hot water networks - Joint assembly for steel service pipes, polyurethane thermal insulation and outer casing of polyethylene”.
5. Kausch, H.-H. (1987). “Polymer fracture”. Springer Verlag, Berlin.
6. Molin, J., Bergström, G. and Nilsson, S. (1997). “Laying of district heating pipes using existing soil material”. Swedish District Heating Association, FVF FoU 1997:17, in Swedish.
7. Swedish Standard SS 02 71 09 (1975). “Geotechnical test methods - Compaction properties - Laboratory compaction”. in Swedish.
8. Trautmann, C. H. and O’Rourke, T. D. (1985). “Lateral force-displacement response of buried pipe”. Journal of Geotechnical Engineering, 111(9), pp. 1077-1093.
9. Williams, J. G. (1987). “Fracture mechanics of polymers”. Ellis Horwood Limited, Chichester.
![]() | |
© 2002 ejge |