Noise and Vibration in Gas Turbine and Cogeneration Power Plants

You’ve got a new gas turbine simple cycle or cogeneration power plant, or just finished an overhaul or modification, and it generates way too much noise or vibration. What happened and how can it be fixed?

This post provides steps to be followed to identify the source of the problem and ways to fix it.

  1. Find out when the noise/vibration problem occurs

We’re all familiar with having a niggling problem with you car, but the mechanic can’t solve it. You can’t blame them if the problem can’t be recreated. Not to have this issue repeated if/when you retain a noise and vibration expert, you should document any and all operation conditions when the problem occurs and/or when it is at its worst. Does it happen at a specific MW? Is it worse when the outside temperature is low?

  1. How bad is it?

As important as finding out a solution, it is also important to find out how bad it is. Is the noise or vibration causing an environmental problem, causing complaints from the neighbours, or is the shaking so bad that you are concerned that the gas turbine or exhaust stack is going to shake apart?

Typically, high vibration levels suggest that damage of the plant components are probable. High noise levels likely will not cause damage over the short term, however damage is possible with continued use without resolving the issue. A noise/vibration engineer is usually retained to quantify the noise/vibration levels and provide an assessment to the severity.

  1. What could be the source of the noise/vibration?

The source of the noise/vibration could either be from the gas turbine or from the exhaust circuit.

The most common gas turbine source is an instability in the combustor. This can be relatively common in lean running low-NOX systems.

There are more possible sources of noise/vibration in the exhaust circuit. With the exception of fluid elastic instability, all of the sources require a structural or acoustic resonance condition to cause a significant problem.

Sources of noise/vibration include:

Simple Cycle or Cogen plants
  • Acoustic Resonance – An acoustic resonance, generally a standing wave, can exist if the shape and dimensions of a part of the exhaust circuit are just right. With this condition, a tone generated from the turbine or in the exhaust circuit will be amplified to become a problem.
  • Vibration Resonance – Similar to acoustic resonance, if there is a vibration resonance of the structure it could amplify a tone generated by another source. Note that without acoustic or vibration resonances, the source of the noise/vibration typically are not significant enough to be a problem.
  • Turbulence – Poor design of the exhaust circuit can cause high vibration levels because of turbulence.
Generally only found in Cogen plants
  • Vortex Shedding – An oscillating flow that is created by the exhaust flow passing through the tube bundles in the exhaust circuit.
  • Turbulent buffeting – Similar to vortex shedding, turbulent buffeting is generated by exhaust flow through tube bundles.
  • Fluid elastic instability – This phenomenon generally only occurs at high flow velocities through tube bundle banks. If the flow velocities are high enough, the system is unstable. With continued use, failure will occur.
  1. How is the source of the noise/vibration determined?

An investigation akin to a crime investigation is usually required to identify the source of the noise/vibration problem. Sometimes finding the culprit can be simple, while other times very challenging.

The process typically takes the form of conducting noise and vibration measurements during well defined control test runs of the plant. The first step is to quantify the noise/vibration levels, defining where and under what operating condition it occurs, and defining its characteristics; its dominant frequency, is it a pure tone?, etc.

The next steps in the investigation will be dependent on the noise/vibration characteristics.

If it’s a low frequency tone

From experience, a low frequency tone, ~60 Hz or lower, is typically a combustor instability problem of the gas turbine. Before blaming the gas turbine manufacturer, it is prudent to get evidence that supports this conclusion. Multichannel measurements including an in-flow acoustic measurement can determine the direction the noise tone is travelling. The travel direction can be 1) the same as the exhaust flow direction, 2) the opposite direction, or 3) stationary (a standing wave). The first would prove that the gas turbine is the source, the other two indicate that the exhaust circuit is the problem.

If it’s a mid/high frequency tone   

Typically mid/high frequency noise/vibration is from exhaust circuit aerodynamic phenomena, such as vortex shedding or turbulent buffeting. However, for this to be a problem it also requires an acoustic or structural vibration resonance. To determine acoustic resonances would entail in-flow acoustic measurements at strategic locations in the exhaust circuit. Some calculations of possible acoustic room modes aid in identifying the in-flow test sites. To identify possible vibration resonances includes impulse response testing. Calculations and/or computer modelling is sometimes used to aid in the identification of structural elements that could be involved in the vibration resonance issue.

If it’s a broadband noise/vibration

Broadband noise/vibration is typically caused by turbulence. The process to identify that this is the likely source is usually by the method of elimination. Testing and measurements are conducted to eliminate all off the other possible sources. If these are all eliminated, it is concluded that the problem is turbulence. Note that this issue could be caused by the gas turbine, the exhaust circuit, or a combination of oth.

  1. What can be done about it?

The solution to the problem is dependent on the problem itself. If the source is the gas turbine, it is recommended to contact the manufacturer to resolve the problem. They may be able to change the operation parameters of the turbine to mitigate the problem.

Some potential solutions to noise/vibration problems in the exhaust circuit are provided below. Note that an expert should be retained to evaluate the problem and come up with the solution(s) specific to your problem.

Acoustical resonances are dealt with by introducing baffles. This effectively changes the acoustical resonance modes away from the problem frequencies.

Similar to acoustics resonances, vibration resonances are addressed by changing the resonant frequencies away from the problem frequency. This typically is performed with the addition of stiffening elements at strategic locations.

Turbulence is addressed by improving the exhaust flow circuit, smoothing out the flow and/or introducing flow straightening elements. CFD (Computational Fluid Dynamics) modelling would aid in the design of the solution and evaluate its effectiveness.

If you would like further guidance through this process, please contact us at [email protected]