Using continuous structural acoustic emission monitoring systems, offshore wind farm operators can now detect cracks within monopile foundations of wind turbines, says Tim Bradshaw
Due to their relatively simple design and ease of installation in shallow-to-medium (0-40m) water depths, monopiles are the most commonly used foundation structures for offshore wind turbines.
The monopile structure itself is a cylindrical steel tube that extends above the waterline and down into the seabed. The pile penetration depth is adjustable to suit the actual environmental and seabed conditions.
The tower is supported by the monopile, either directly or through a transition piece.
A limiting condition of this type of support structure is the overall deflection (ie, lateral movement along the monopile) and vibration, as the structures are subjected to large cyclic, lateral loads and bending moments (due to the current and wave loads) in addition to axial loads (eg, vertical loads due to the structure being supported).
The diameter of a monopile can range to over 7 metres, with wall thickness as much as 160mm.
Current monopile inspection methods
A number of methods can be used to inspect for defects in monopile structures. These include diver and/or ROV (remotely operated vehicle) based inspection.
However, both of these methods are time consuming, costly and generally unreliable when compared to AE monitoring.
A diver or ROV has to visually inspect for cracks or other defects, which is challenging due to potential issues with visibility and marine growth. There are also the inherent safety risks when using a diver.
Vibration monitoring is another method often used to detect problems in monopile structures. This involves looking at the resonance (low frequency movement) of the structure to see if any changes occur. However, a defect may be required to be of a significant severity for the resonance to change, which means it is difficult to detect problems early.
Acoustic Emission (AE) is the term used to describe the high frequency signals (not audible to the ear) that result when material cracks or yields. These signals are usually short transients and can be detected using high frequency sensors in the same way that earthquakes are detected using low frequency sensors.
The advantage of the AE monitoring approach is that the technology is able to identify active defects, providing information on when these defects are active (i.e. under what load conditions) and their location on the structure being monitored. AE is used worldwide to monitor offshore structures, bridges and pressure vessels as part of inspection and risk-based inspection schemes.
AE monitoring of monopiles
Mistras has developed, and successfully deployed, a continuous structural monitoring system to provide fatigue crack detection capability within the monopile foundations of offshore wind turbines.
The AE system is permanently installed on the asset, continuously monitoring the AE signals being generated by the monopile to provide real time detection and location capability for any active gross fatigue crack sources.
Prior to developing the AE system, Mistras identified the following factors as critical:
* Simple to install as possible, with minimal impact on the asset and enabling easy access for maintenance
* Provide 100% coverage of the monopile for any active sources within the installed sensor array (water’s surface to the mudline)
* Collect AE signals, process and store the data on a specific AE data acquisition system
* Filter potential noise sources from day-to-day operation on the asset
* Have the potential to interface with SCADA to collect additional parameters used during AE data analysis
* Monitor movement of the asset for correlation to AE activity
* Be remotely interrogated via a network connection to allow system and data management
* Automatically provide statistical updates to the customer via direct or website interface on the status of all assets being monitored
* Provide vertical position (±0.5m) of any active sources within the monopile.
The AE system developed by Mistras utilises a string of underwater AE sensors immersed in the monopole water column. This allows signal transmission from any active discontinuity on the monopole structure (that is within the sensor array) to be transmitted to the sensors through the water column. The physical setup of the complete system is defined by verification trial measurements of the system performance to ensure sufficient sensitivity.
The AE system comprises the following separate components:
* The sensor string
* The data acquisition system
* The website
* Linear array sensor string.
The sensor string is the array of sensors that are distributed down the centre of the monopile water column. This array is delivered to site as a manufactured component ready for installation.
The use of a string of sensors within the water column in the monopile is utilised for both ease of installation and for coverage.
The installation, inspection and maintenance of the sensor string can be completed without any requirement for access into the water. In addition, the ability to monitor the whole circumference of the monopile with the same sensitivity but a minimal number of sensors. The number of sensors and spacing are defined by Mistras to provide sufficient sensitivity to active discontinuities.
If a suspected active source is identified, a 3D array can be installed, the objective of which is to provide a 3D source location, providing not just vertical but circumferentially source position information.
Data acquisition system
The data acquisition system (DAQ) is installed in the Transition Piece in an available area. The DAQ consists of an AE monitoring system enclosed in an IP65 case. The DAQ system and industrial PC are designed specifically for acquisition of AE data, providing 16 channels of AE data acquisition in order to provide spare channels and the capability to monitor the 3D array if required.
A second DAQ system is provided as backup. A battery back-up unit and UPS is also provided for the DAQ system.
A two-axis inclinometer is also mounted internally in the DAQ system, which provides information on the inclination of the tower. This gives some indication of the movement of the structure to correlate to the AE activity.
Each DAQ system is set up to continuously monitor signals on all the AE sensors and inclinometer simultaneously, undertaking real time location of signals and rejecting any signals that do not locate linearly along the length of the sensor array. This helps to minimise noise signals and therefore minimises the size of data files collected. The system is also capable of identifying if the sensors are working correctly.
The MISTRAS AE system has alarms defined within its software to identify activity on the monopile that is of interest and requires further investigation. In order to minimise false alarms, the system is programmed to identify the signature received from a boat pushing on to the structure. In this event, the system will inhibit the release of alarm messages in the form of emails or website messages.
In addition to the raw data files collected, each DAQ system generates statistics files that summarise the data collected every 30 minutes. These files are small text files designed to be used for data transfer to a website while minimising bandwidth demand on the network. These files are automatically transferred by the system to the Mistras servers for processing and to allow data display on the customer’s secure website.
To allow the customer to have visibility of alarms and trending data generated by the monitoring systems, Mistras provides a secure website. This interface is designed around customer requirements. Access to the website is through a secure log in using HTTP SSL security protocol.
The website clearly shows the status of all AE monitoring systems in a systems status grid. An alarm log is also provided, whereby alarms can be reviewed, acknowledged and managed.
The website also provides access to all historical statistical monitoring data up to data that is at maximum 1 hour old. Cumulative alarms for an asset against time can also be displayed, as well as the ability to view cumulative events/hits. Access to historic reports can also be accessed in a format that can be searched.
Various levels of reporting can be provided by Mistras. The first level includes the automatic reporting by the system via alarm emails and the data displayed on the website.
In addition, Mistras can provide a daily check on each system using remote desktop to view each system and reporting on the time of test, activity since last inspection and any activity of interest. Hardware tests such as automatic sensor tests can also be conducted at this time.
If an AE alarm is triggered, Mistras can acknowledge the alarm on the website and communicate the system of interest. The data from the alarm period can be recovered over the network for further analysis.
A monthly report can also be produced to summarise the performance of all systems over the last month.
Tim Bradshaw, Director, Mistras Group, UK operations.