September 2016, Vol. 243, No. 9
Features
GIS – Improving Pipeline Risk Assessment
While geographic information systems (GIS) have been around in commercial form since the early 1980s, subsequent technologies for data capture, storage and processing have improved rapidly in both resolution and scope. In terms of data capture, technologies that explicitly incorporate spatial referencing are often referred to as “GIS technologies.”
Perhaps the most widely known are global positioning systems (GPS), initially a military technology that has become a fundamental part of our everyday lives, enabling geo-tagging of photographs, navigation, and registration of the location of specific pipeline assets, even girth welds.
Increasingly, the benefit of GIS-based technology is being recognized by the oil and gas industries and incorporated into pipeline management programs. Satellite imagery and more recently, technologies such as drones or UAVs equipped with light detection and ranging (LiDAR), or multispectral cameras linked to positioning systems, can provide detailed terrain information, such as height, etc. Accuracies (absolute height error) of +/- 5-10 cm can be easily achieved with current commercially available equipment.
The key output of these technologies is data, which can serve as input into any number of analytical processes. However, the real power of GIS lies in the ability to integrate various kinds of data using an explicit spatial location. When appropriately analyzed, summarized, and combined with other data, a wealth of information can be made available to pipeline operators and risk managers. Of course, considerable knowledge and skill is required to ensure that this kind of integration is done correctly; in other words, with attention to accuracy, error and resolution, and analytical methods.
Risk Analysis
GIS software and spatial databases are increasingly deployed within a range of risk analyses. These include assessment of threats such as geohazards (landslides, earthquake hazards and flooding), but can also include analyses of other spatial phenomena such as wind speed, population density, elevation, and so forth. This also makes them ideally suited to analysis of the consequences of pipeline failure, such as leaks or ruptures.
Pipeline risk management involves the estimation of risk and the implementation of measures to control and ensure the risk levels are within acceptable or tolerable spectrums. The risk models used to facilitate such analyses in the field of pipeline integrity range from simplified qualitative approaches to complex quantitative approaches. As regulatory requirements increase, there is a commensurate increase in the desire for more quantitative risk analyses; this results in additional demands for high-resolution input data to satisfy all input parameters and generate usable outcomes.
Qualitative models usually consist of a series of subjective and objective questions that are combined with a set of rules or algorithms. Quantitative approaches require more detailed data and significant attention to the interaction of these data. While quantitative approaches are attractive in principle, they bring additional demands for high-resolution input data to satisfy all input parameters in order to generate usable outcomes. Providing high-quality data is the key to ensuring realistic outcomes.
GIS-Based Threat
Pipelines are subject to a wide range of threats that are often complex. The primary issue is that these threats vary over space as well as over time. Long linear assets like cross-country pipelines in particular are subject to a plethora of differing conditions. Some threats can be characterized as time-dependent and represented as such, while others are much more difficult to represent in terms of their likelihood and/or impact over time. In order to effectively use a GIS, we aim to quantify as much of the process or phenomena as possible.
Third-party damage is widely known to be a significant cause of pipeline failure. Depth of cover (DOC) is affected by a range of activities and natural phenomena, including excavation activities, erosion, cultivation, construction, flooding, ground subsidence or other environmental factors. DOC can be determined by combining height information as captured in a digital terrain model with ILI tools equipped with an inertial measurement unit. The proliferation of drones and UAV’s illustrate the huge demand for cost-effective data capture, and these are often equipped with digital cameras and other sensors.
Figure 1: Change detection using satellite imagery.
Source: Huisman and Gharibi (2015)
These tools and sensors can provide analysis to determine changes along the pipeline ROW within a given time interval. Imagery from 2010 and 2013 (Figure 1) has been compared to assess where significant change has occurred. Different satellite bands can be used to investigate specific processes and phenomena, from identifying evidence of ground disturbances along the pipeline to determination of vegetation stress, to detection of hydrocarbon seepage.
Slope instability is a geohazard threat that can result in bending strain, pipeline movement, and even a full rupture. Factors that contribute to slope failure and known locations of past movements can be integrated into a model to determine the most susceptible locations (Figure 2).
Figure 2: Areas of slope instability along a pipeline.
Predicting Consequences
Estimating the likely effect of a pipeline failure using GIS requires accurate data and models. As an example, imagine a gas release scenario for an onshore pipeline or offshore platform. Parameters such as wind, temperature, release volume and surface friction are needed to estimate the dispersion and likelihood of ignition. Figure 3 shows the use of elliptical hazard zones to take into account wind direction and intensity.
Figure 3: Using elliptical zones to model gas dispersion.
The GIS environment provides a dynamic platform for analyses to predict injuries and other potential consequences. Databases containing housing and population data can be readily integrated to provide accurate and up-to-date information.
High-consequence area analyses use a range of parameters including pipeline diameter, valve shutdown latency and pressure to calculate release volume. Soil/geologic information feed percolation models to estimate absorption and, in conjunction with topographic profiles, can be used to estimate flow pathways and residual spill volumes. These parameters are critical components in consequence of failure (CoF) calculations, including cleanup costs and damage assessment.
Figure 4: Flowlines in HCA determination.
Future Perspective
The ongoing improvements in technology are delivering more accurate and higher-resolution data to engineers, risk assessors and decision makers. This in turn creates a demand for improved methods of managing this data, including storage, data documentation, and search mechanisms to enable access to the data that is required.
Data handling, spatial queries and detailed analysis can all be completed within a GIS environment, allowing regular, detailed assessments to be performed, repeated and visualized. The advantage of the single system is speed; fully quantified risk assessments can be executed in minutes rather than hours or days. The result is an enhanced ability to analyze and visualize risk results and take appropriate actions.
In the near future we can expect the increasing availability of data streams from geosensors or sensor networks around the world to help improve the models used to estimate and predict specific consequences in close to real time.
References
O Huisman and A Gharibi (2015), Change Detection Within Pipeline ROWs using high resolution Satellite imagery. ASME International Pipeline Geotechnical Conference (IPG2015), July 16-17, 2015 Bogota, Colombia.
Authors: Otto Huisman is responsible for the worldwide management of integrity management software projects and extension of the GIS services portfolio at ROSEN Group. He has 20 years’ experience with GIS technology, and has published numerous technical and academic papers. He holds a doctorate in Geoinformatics from the University of Auckland in New Zealand.
Comments