Featured Article in March Issue of Pipelines International Magazine
An article describing the accuracy of 3D Toolbox was included in the March 2014 issue of the magazine “Pipelines International”. The article describes how the 3D Toolbox performed in comparison to traditional measuring methods as well as to a coordinate measuring machine.
A direct link to “Pipelines International” for a condensed version of the article:
http://pipelinesinternational.com/news/measuring_anomalies_with_the_3d_toolbox/86662/
Pipelines are the safest, most effective method of transporting oil and natural gas. In order to maintain pipeline safety, monitoring systems are installed to ensure a rapid response in the event of a failure, and in-line inspection (ILI) tools are used to locate and identify defects to enable corrective action that prevents failures.
ILI tools generally incorporate sophisticated inspection technologies into a tool that moves inside a pipeline; the inspection technologies typically measure parameters that can be used to calculate the current wall thickness of the pipeline at specific points. Sophisticated software determines the type and severity of anomalies, such as corrosion or dents, and helps the operator of the pipeline determine what – if any – corrective action is required. Because of the measurement uncertainty associated with ILI data, direct inspection of anomalies is frequently required to verify their severity.
Direct assessment often requires excavation of a portion of a pipeline. Given the frequency of anomalies identified in ILI runs, it is not uncommon for a pipeline to require several hundred excavations annually at a cost of tens of millions of dollars. Once the pipeline is excavated and the pipe surface is prepared, the anomaly can be inspected. After spending millions of dollars in monitoring and safety systems, ILI runs, and excavation costs, the severity of most anomalies is determined through manual measurements using a pit gauge.
An alternative method of measuring anomalies on pipelines uses the 3D Toolbox, an analytical software package paired with a 3D imaging system that is based on a technique commonly referred to as ‘structured light illumination’. Once the 3D data is acquired, the data is then analysed for areas of metal loss. Measurement error can be introduced in either the 3D data acquisition or the 3D analysis step.
Accurate 3D measurements form the basis for determining metal loss and metal deformations. However, accurate 3D measurements are not sufficient to determine the extent and impact of damage on a pipeline, as the original condition of the pipe needs to serve as a baseline for the 3D analysis. With the 3D measurement representing the current condition of the pipe, the data are separated into both damaged and undamaged regions of the pipe. Once the undamaged points have been identified, these points can be used to form surfaces, which then become the reference against which damaged areas are measured to determine the extent of metal loss or metal deformation. The overall process is:
- Acquisition of the raw 3D data; - Analysis of the 3D data, including establishing the reference surface (shown in green), determining defect areas and depths, and calculating the river-bottom profile; - Generate RSTRENG calculation and report.
All measurements have errors. The table on the right is a summary of the major sources of error associated with the use of 3D data to determine metal loss and metal deformations.
In order to determine the measurement errors, known values must be established. For this study, a co-ordinate measurement machine (CMM) was used to accurately measure the surface of pipes with corrosion and pipes with dents. The CMM measurements were traceable to National Institute of Standards (NIST) and were accurate to within 0.04 milli-inch (mils), or 1 micrometer (μm). Because the accuracy of the CMM was expected to be almost two orders of magnitude better than for all the other techniques, all CMM measurements were considered to be the true value. Five areas were investigated in this study: three areas of corrosion, and two pipes with dents. In addition, ten defects were selected in each area of corrosion. The deepest point for each of the ten defects was measured by a conventional pit gauge, the 3D Toolbox, a 3D structured light system from Seikowave, and the CMM. The CMM was programmed to take data in a rectangular array with a circumferential pitch of 50 mil (1270 μm) and an axial pitch of 100 mil (2540 μm). Because the CMM was not programmed to search for the deepest points of corrosion, there was some probability that the CMM data might not measure the deepest point for all defects. The errors associated with the data from all five errors were in agreement. For brevity, this article will focus on presenting data from one of the corrosion samples.
The pit gauge data was acquired over the course of a day. The CMM data took more than three days to acquire. All data using the 3D Toolbox was acquired within a few hours.
Figure 3 shows a comparison of the CMM data points and the 3D Toolbox data points. Within a one square inch area, there are 200 CMM points. In this same area, the 3D Toolbox produces a 3D model with 2,500 points. Overall, the agreement between the CMM data and the 3D Toolbox data is within 4 mils (101.6 μm) for 77 per cent of the data; 90 per cent of the data was within 6 mils (152.4 μm). The major deviations occur in areas where the sampling frequency of the CMM was not high enough to capture the change in the shape of the pipe; typically, this occurred in areas where corrosion was present.
While the comparison between the CMM and the 3D Toolbox data showed excellent agreement, this result did not verify whether or not the 3D Toolbox could accurately determine metal loss. To verify this, the 3D Toolbox analysis software was used to determine the depth at ten specific locations. The depth values calculated by the 3D Toolbox analysis software were then compared to the depth measurements made by the CMM. The 3D Toolbox and CMM depth measurements were also compared to pit gauge measurements.
CONCLUSION
The 3D Toolbox is a measurement system that comprises of a 3D imaging system, the 3D digital pit gauge, and a set of software tools for acquiring and analysing 3D data of corrosion, dents, and other pipeline defects. This study focused on establishing the error of the raw data collected by the 3D Toolbox and the depth calculations determined by the 3D Toolbox analysis software. Unlike most 3D imaging systems – where error is reported on a global basis – this study focused on determining error on a point-by-point basis. This is necessary because maximum allowable operating pressure (MAOP) calculations are driven by the deepest points of corrosion, and not by an average metal loss over a large area. Consequently, understanding the error associated with individual measurement points is critical to determining if the MAOP calculations are valid and, ultimately, if the pipeline is safe.
Based on the data, the deviations from the CMM points were determined to have an average value of zero and a standard deviation of 4 mils (101.6μm). The ability of the 3D Toolbox to accurately determine the depth of a corrosion pit was calculated to be within 2.2 mils (55.9 μm) of the CMM, on average.