Intercomparison of sandboxes used for testing joints for district heating pipes (NT TR 495)

Abstract

  An intercomparison has been carried out between 3 laboratories concerning a type test for joints according to EN489 District heating pipes – Pre-insulated bonded pipe systems for directly buried hot water networks – Joint assembly for steel service pipes, polyurethane thermal insulation and outer casing of polyethylene. The intercomparison has been carried out on smooth pipes as well as on smooth pipes with a joint. At all 3 laboratories the tests were carried out in EN489 sand. In addition, tests have been carried out in filter sand and gravel. At all laboratories stress transducers were mounted during testing. That has made it possible to register the stress required to move the test piece. In order to compare the results the stress on the smooth pipe has been recalculated to stress per m pipe. The stress on the joint has been calculated, as the registered stress on the smooth pipe with a joint has been reduced by the stress arising from the smooth pipe. The limited number of tests shows a clear trend. If the trend has to be documented it will be necessary to carry out tests where several comparable sandboxes, sand types and pipe dimensions and pipes with joints form part of the test. In the light of the tests that have been carried out the following conclusions can be drawn. The steel collar incorrectly stresses the pipe and joint as the steel collar absorbs the stress that is created when pulling the pipe. Marks from wear and tear often appear on the under side of joints that have been tested in a sandbox, but they do not appear on the upper side. In practice that does not happen. Stress is detected by the sand above the pipe in the entire length of the pipe. Therefore, it should be stated in the standard EN489 that the test piece must not touch the sandbox during testing. As sand is compressed up against the ends when the joints in the sandbox are moved, the sandbox has to be long enough to avoid edge effects. Alternatively, the length of the sandbox has to be determined so the edge effects are identical and small. We suggest half a pipe length corresponding to 3.0 m. In prEN13941: Design and installation of pre-insulated bonded pipe systems for district heating a laying depth of up to 3.0 m is recommended. In EN489 a laying depth of 1 m is used. EN489 sand and filter sand have a friction coefficient of app. 0.3. In practice, sand types with a friction coefficient of 0.6 – perhaps even larger – are used. Measurements show that stress on the joint is app. 30% greater in filter sand #2 than in EN489 sand. In order to compensate for the limited friction it is recommended to increase the laying depth to 3.0 m and to use filter sand #2. As filter sand to a great extent behaves as EN489 sand and as it is a standardised merchandise with limited environmental inconvenience from dust, we suggest to replace EN489 sand with filter sand #2.