October 2012, Vol. 239 No. 10

Features

Integrating Leak Detection Into SCADA Systems

For pipeline operators, deciding to implement leak detection is easy. However, determining how to accomplish it is hard. Leak detection is a good tool for quickly raising alarms for theft, identifying product loss due to wear and corrosion to the pipe, and/or preventing environmental events caused by spills. Whether driven by safety and production concerns or regulatory requirements, the method chosen for leak detection has a significant impact on its effectiveness.

Approaches to leak detection can be categorized as either “external” or “internal” methods. Integrating leak detection into existing SCADA systems can provide alarm management of a leak event as well as record keeping and trending before and after the event. Locating the leak with a precise location facilitates quicker response and repairs. Choosing the approach method defines costs for new and existing systems and the practicality of adding leak detection to existing systems.

External Leak Detection Equipment
External leak detection involves the installation of detection equipment on the pipeline. Detection equipment can monitor the dynamics of the flow for changes that would indicate a leak. This type of approach uses instrumentation installed on the pipe to detect changes in sonic patterns or vibration. The ability to pinpoint the location of the leak is a function of the location of the instrumentation.

Typically, the leak can be located to a section or area between instruments. Visual monitoring can employ CCTV with analytic software to monitor changes in the images captured. Installing instrumentation on new systems is costly due to the number of devices required and the communication network required. The costs go up significantly when adding leak detection to existing systems. External methods of leak detection may fail to meet expectations for several reasons:
• The installation process is quite intrusive, because it requires digging up buried pipe.
• Monitoring is very rudimentary and is a function of the pipeline SCADA operation. Operators observe an event and react to it. The speed of response is a function of how accurate the identification of the leak’s location is, and when it alarms. Some detection approaches require more significant leaks to trigger an alarm.
• The equipment is only effective if the device is in the exact right location. For example, if using vibration or sonic detection, the instrumentation must be installed frequently enough that the area of the leak can be pinpointed, and even then, you can only identify an area of concern. Where the pipe is above ground, CCTV coverage is limited even with analytic software.
• Adding a significant number of detection instruments to a system increases the operational complexity and the number of maintenance items. Considering the cost of constant monitoring, equipment repairs and the complexity of installation, external leak detection carries both a high capital expense and ongoing cost. The additional operational expense required to support this system of instruments and communication networks can be significant.
• Providing power sources to drive remotely installed monitoring equipment and instrumentation poses a host of challenges and costs.

The Internal Approach Alternative
The alternative to an “external” approach is an “internal” method. Internal methods use algorithms that incorporate data from several points along the pipeline that are already being monitored for flow. It uses instrumentation that, in many cases, would already be available in the system monitoring the flow, pressure and temperature at each end of the pipe. The external approach requires additional instrumentation along the length of the pipe. The internal approach uses data from the inlet and outlet to determine when and where a leak occurs.

There are two methods for integrating internal leak detection with SCADA systems. With a pipeline of any length, there is measurement at each end reporting the product volume, and the pressure at which it flows. Under the first method, wave propagation calculations are performed using time of flight. A spontaneous leak will cause a change in the pressure wave even when only small volumes begin leaking.

The pressure wave travels through the pipe at the velocity of sound. The time it takes for the pressure wave to travel to each monitoring point (inlet and outlet of the pipe) will be different, allowing the precise leak position to be determined. The time between the event and the alarm is a function of the speed of sound and the speed at which the data is processed. The control room can dispatch a response team to the leak almost immediately.

The second approach uses the gradient intersection method. Under this method, it is assumed that if there is no leak, the pressure at the beginning of the pipe will always be higher than the pressure at the end. Under conditions with no leak, the pressure drop will be very steady. If a leak does occur, the pressure before the leak will be higher than the pressure after the leak. The intersection of the two gradients determines the location.

The advantage of internal leak detection is that even a pipeline that has been in the ground for 20 years can be monitored in this way. Existing SCADA systems are, without a doubt, already measuring pressures and temperatures at each measurement point (inlet and outlet). These data points are used to calculate volume.

With an internal approach, existing data is simply taken from SCADA and put into a module that is running the above-described algorithms in a leak detection software application. The application module then provides information about any leaks back into SCADA where it can be handled like any other alarm. With this information, the back office can tell where on the line even a minor leak has occurred, and exactly where it started. The response is quick and the location is pinpointed.

Once a software application for leak detection is selected based upon a pipeline’s specific requirements, installation is simple. It is not necessary to create separate data capture or local area network (LAN) communication systems, or additional end devices. There may not even be separate hardware or screens that must be maintained in the SCADA system. The leak detection software is installed on a server in the process control network. Interfaces with the SCADA application are established and the system is configured.

During the installation process, the SCADA system, architecture and screens are evaluated, and then a reporting mechanism and screens are built. The system is programmed to send alarm notifications upon identifying a leak, and can even automate shutdowns.

Building on the automation enabled by existing SCADA systems enables the quick identification of the exact location of a pipeline leak. By automating shutdown, internal methods of leak detection mitigate both environmental and production concerns. With no external equipment to be inspected, repaired or replaced, internal methods also represent both a lower initial capital expenditure and lower operational costs.

The Author
Jim Fererro is vice president of GlobaLogix, a Houston-based oilfield automation and control company. www.globlx.com