Eglisau Power Station

Customer: Nordostschweizerische Kraftwerke, NOK (today: AXPO)
Report: 071012, December 10, 2007
Publications: 2009 IOMAC 3, Portonovo (Ancona), Italia

The Eglisau River Exploitation Station with its seven 4.5 MW turbine/generator units was built on the Rhein River separating Switzerland and Germany here between 1905 and 1910. The owner plans to replace the turbine/generator units and to thereby increase the units‘ rotational speed. Therefore, the question arose whether or not this could lead to vibrational problems with the power station building. In addition to a Finite Element study undertaken by a third-party engineering office rci dynamics performed two experimental modal analyses on the station in October 2007. On the one hand, the global behavior of the building was aimed at. The local behavior of the section with the units No. 6 and 7 located in the river mid part was to be investigated on the other hand.

IMG_7172 foto power house outside Kopie

Maschinenhaus innen IMG_3271

An owner of a power station is not really eager to shutting down his station for a vibration test. A first attempot to identify the global station building dynamic characteristics was therefore performed with everything in operation. The result was, however, negative. As a result of some negotiation, the owner agreed to shut down his station, but only once and not twice as usually planned. It takes several hours to shut down or run up the turbine/generator units with also keeping the requirements concerning the water level on the upstream side of the river (constant water level!). The global and the local experimental modal analysis had hence to be performed in one 24 hours time slot. The publication mentioned at the beginning contains a detailed discussion of the test results for the global analysis without and with the units being in operation.

Instrumentation layout as used for the global analysis; setup 1 is shown (references blue, rovers green).

A 2D-measurement point grid with 12 measurement points on three levels (turbines, generators, roof) was used for the global ambient analysis. Special attention was paid to the two dilatation joints separating the building into three parts. On each level, a reference and a rover were located. This resulted in 11 setups with 12 sensors. The time window length was chosen to 30 minutes per setup.

19 natural global modes of the power house could be identified with frequency, shape and damping coefficient in the frequency range f = 1.86…39.2 Hz .

Global mode 3, f = 3.28 Hz, zeta = 0.91%.

A 3D measurement point grid with 18 measurement points on two levels (turbines, generators) was used for the local ambient analysis. On each level, a reference and a rover were located. A third rover was to used to optimize the total test duration. This resulted in 12 setups with 15 sensors. Again, the time window length was chosen to 30 minutes per setup.

Instrumentation layout as used for the local analysis; setup 1 is shown (references blue, rovers green). The local part is identical with the lower part of lefthand third of the powerstation as shown above.

Local mode 4, f = 5.00 Hz, zeta = 1.0 %.