Preventing, detecting and mitigating pipeline commodity releases: Part 2

Pipeline operators are under severe financial and social pressure to avoid incidents that cause commodity releases. In Part 1 issues of pipeline design and management were examined, including external leak detection.

There is no optimal commodity release detection system for all pipelines in every environment. Each pipeline is unique and requires an individual evaluation. In Part 1 of this article the design and ongoing management of pipeline systems was discussed, along with external methods for the detection of commodity release. In this second part we examine internal (computational) methods, and the evaluation of such systems.

Detection

Internal-based pipeline detection

Internal-based pipeline detection looks at conditions inside the pipeline wall to discover commodity releases. More commonly known as computational pipeline monitoring (CPM), this methodology has been around for about 30 years and uses software that takes a variety of measurements available on the pipeline to establish what is happening within the pipeline.

The 2012 American Petroleum Institute (API) Recommended Practice (RP) publication 1130 defines CPM systems as systems that are internally based, utilising field sensor outputs that monitor internal pipeline parameters such as pressure, temperature, viscosity, density, flow rate, product sonic velocity and product interface locations. Which parameters are considered and how they are interpreted depends on the CPM method being applied.

The following is a brief description of the five CPM methods in use on pipelines today:

• Line balance CPM techniques measure the imbalance between the receipt and delivery volumes. The capabilities of its simplest form (meter in/meter out comparison) can be enhanced by correcting the meter readings to standard conditions and by compensating for changes in the line pack (amount of commodity actually inside the pipeline) due to temperature and pressure for each product in the pipeline.
• Real-time transient model (RTTM) CPM models all the fluid dynamic characteristics, including line pack, slack, shut-in and transients, under all pipeline flow conditions. This is a very detailed configuration with very fast calculations and the ability to model hydrocarbons in any phase. The RTTM software compares the measured data for a segment of pipeline with its corresponding modelled conditions.
• Statistical analysis CPM statistically evaluates pressure and flow inputs that define the perimeter of the pipeline in real time for the presence of patterns associated with a commodity release. A probability value is then assigned to whether the event is a commodity release or not. An alarm is generated if the statistical changes persist for a certain time period.
• Pressure/flow monitoring CPM examines the relationship between various sensors’ outputs and applies an algorithm to determine if they indicate an anomaly. Essentially this CPM is what an operator does by nature, looking for unexplained large drops in pressure or flow, but there are applications that look for these anomalies to ensure these large changes are not missed.
• The acoustic/negative pressure wave technique takes advantage of the two negative pressure, or rarefaction, waves produced when the commodity release occurs and the integrity of the pipeline is compromised. This methodology requires installing high response rate pressure transmitters at selected locations on the pipeline.

These five CPM methods can be classified according to two different alarming principles underpinning their detection algorithm.

Conservation of mass methods work on the principle that whatever enters the pipeline must be equal to whatever exits the pipeline, adjusted for change in inventory of the pipeline. The line balance CPM, real-time transient model CPM and statistical analysis CPM techniques can base detection on this method.

Signature recognition methods consider the relationship of system pressures or flows, or recognise anomalies in sensor outputs on the pipeline. The real-time transient model CPM, statistical analysis CPM, pressure/flow monitoring CPM and the acoustic/negative pressure wave CPM techniques can base detection on this method.

General considerations for evaluating CPM systems

No one single commodity release detection system is optimal for the entire range of pipelines operating in widely diverse environments. A comprehensive analysis is necessary to identify which CPM technologies and methods are best suited for the particular pipeline. A simple A-to-B pipeline route might have simpler operations than a pipeline with many active route connections and elevation changes, multiple receipt and delivery points, and reversible flow. The more complex the pipeline, the more flexible the CPM needs to be to handle all possible operational scenarios.

The following list of key factors to consider when evaluating a new CPM (or re-evaluating a legacy system) for its detection capability should be weighted according to their importance to any particular operation:

• Rate of false alarms and misses.
• Sensitivity to pipeline flow conditions such as transients, shut-ins, starts and stops.
• The impact of instrument accuracy and configuration accuracy.
• Personnel training and qualification requirements.
• Required response time.
• Accuracy and precision in estimating location and volume of release.
• Ability to detect pre-existing releases.
• Robustness/high availability.
• Initial cost/tuning costs/maintenance costs.

The most important goal in selecting a commodity release detection system is the ability to identify a commodity release quickly enough to mitigate the safety and environmental risk while also meeting the operator’s overall business objective. This includes the potential value of product lost, the cost of clean-up and potential regulatory fines, potential detriment to surrounding environments, and the cost to reputation and potential impact on future projects.

Specific considerations for evaluating CPM systems

In addition to the overall general considerations that need to be taken into account when evaluating commodity release detection systems, some more specific aspects are applicable to particular pipelines. High consequence areas will have a significant impact on the evaluation, as well as the size of the release that is likely to need to be mitigated.

High consequence, or high risk, areas (HCAs) are defined as areas where a pipeline commodity release will have a significant impact on people, property, the environment or all three. HCAs typically demand higher levels of commodity release detection capability and sensitivity to mitigate the higher risk of significant consequences from a release.

Pipeline companies that have pipelines in such HCAs must conduct a more thorough risk analysis and employ additional commodity release detection measures to enhance public safety and protect property and the environment. Some of these measures can be summarised as follows:

• Automated data collection for over-short analysis.
• Integrated alarm and status information between connected pipelines.
• Use of, or more frequent, operational shut-in tests.
• Additional instrumentation or the relocation of instrumentation.
• Application of, or tighter parameters on, pressure/flow deviation monitoring.
• Higher degree of data integration between operations support applications.
• Higher fidelity commodity release detection applications.
• A multi-tiered commodity release detection approach, where systems work independently of each other.

In terms of the size of the commodity release, API 1149 provides a methodology to determine the theoretical ability of a given commodity release detection application to detect a commodity release of a given size, based on the specifications of a given pipeline. Although an American standard, API 1149 is used around the world either directly or as a baseline for local regulations.

While commodity release detection systems do not necessarily need to achieve the lowest theoretical capability as determined by API 1149, pipeline companies can use the standard to weigh the cost of commodity release detection systems against the risk of undetected commodity releases. Further, API 1149 calculations can assist pipeline operators in determining the benefit of specific pipeline infrastructure enhancements to their commodity release detection capability. For example, it can be calculated what increase in commodity release detection sensitivity can be achieved by adding, replacing or upgrading instrumentation on the pipeline.

Although commodity release detection technology has advanced a long way in terms of detection time and detectable commodity release size, damaging pipeline ruptures and large volume release events have still occurred and have unfortunately been missed. In addition to individual companies taking initiatives to improve their commodity release detection capabilities using the strategies discussed above, the Association of Oil Pipelines (AOPL) has created a Leak Detection Rupture Monitoring project as part of its ‘Pipeline Leadership Initiative’ to develop additional strategies to continue improvements in detection of commodity releases.

A key area of improvement the initiative has identified is executing on the ‘3Rs’: recognition, response and reporting. The AOPL has developed performance standards for the industry to follow in this area, with the target goal of 30 minutes for 3R execution (see Table 1).

Table 1: The 3R’s of detecting a pipeline rupture.

Challenges with detecting commodity releases

The uniqueness of each pipeline creates many challenges that might look easy to overcome when selecting a commodity release detection system, but become critical factors for its successful implementation. When evaluating the needs and effectiveness of systems for detecting commodity releases, the following factors should be evaluated to determine their impact:

• Batched systems with multiple products, multiple phase products or reversible flow systems.
• Transient and steady state flow conditions, turbulent and laminar flow transitions.
• Step change product temperature gradients, elevation-induced hydraulic variations (such as over a mountain or under a shipping channel).
• Varying pipeline diameters, telescoping systems, restrictions, block valves, tees, relief systems, control valves and other unique physical characteristics.
• Multiple pump configurations, whether series, parallel, varying and multiple speed, electric and engine drives.
• Branch connections and multiple inlets, outlets and partial flow alignments.
• Slack line and product separation during static conditions.
• Physical properties and hydraulic characteristics of high volatile liquids (HVL) versus crude versus condensate versus refined products, all operated within a single SCADA console.
• Communication outages, variable signal scan and refresh rates, errant signal and data filtering versus non-HCA system variances.
• Human factors such as operator sensory overload and fatigue.
• Varying individual operating procedures.
• Employee turnover and limited training time for new operators.
• External and internal resource availability.

The degree to which any of these challenges will be mitigated is directly related to the CPM chosen for the pipeline. Others, such as human factors of operator overload and fatigue, will rely on the implementation of control room management, HMI and training best practices.

Mitigation

Minimising the impact of a commodity release is the third aspect of pipeline integrity. A release is normally classified as either major or minor. Major releases are emergency situations that result from a rupture to the pipeline that would have a negative impact on both the environment surrounding the incident site and the general public. These kinds of incidents require resources from pipeline operators, emergency response personnel and third-party party agencies. A minor release is still regarded as an emergency from a process point of view but does not require a high level of alertness and mobilisation of resources.

Whether the commodity release is classified as a major or minor release, the following mitigation process phases would typically be followed:

• Locate: The time frame it takes until the physical location of commodity release has been confirmed could be a very short period if the release is found by a third party (eg, a farmer in a field). However, location by emergency response teams could range from just minutes to a couple of hours in a worst-case scenario.
• Recover: Most of the critical decisions about the containment, routing or general management of the incident site are made within the first eight hours after the commodity release.
• Cleanup: This phase lasts until the incident site is fully cleaned up. This could take days, or even months, depending on the extent of the commodity release.

Conclusion

All pipeline companies in the world have as their primary goal and concern that the transportation of commodities be safe and reliable, while realising that commodity releases will continue to happen no matter how strong the prevention measures. Taking a holistic approach to commodity releases and not looking at prevention, detection and mitigation as independent and separate aspects of pipeline integrity benefits the pipeline company, the public in general and the environment.

Additionally, it is important to realise that there is no ‘one size fits all’ commodity release detection system for all pipelines in every environment. Each pipeline is unique and requires an individual evaluation. Pipeline operators need to weigh business objectives against their threshold for risk. The intersection of those points is where companies will find the appropriate commodity release detection solution. Different CPM methodologies and external commodity release techniques provide potentially complementary commodity release detection capabilities, so different methods, or a combination of methods (a tiered approach), might be the right fit overall.

The following steps are suggested for taking a holistic view of pipeline integrity:

• Step 1: Evaluate the activities associated with prevention. Would any of these activities benefit from what is being done in the areas of detection or mitigation?
• Step 2: Evaluate detection activities. Is the level of sensitivity per requirements? Or would it be beneficial to upgrade or install a complementary commodity release detection system?
• Step 3: Evaluate the emergency response plan to see if there are any inputs from prevention and detection activities that potentially would be beneficial for emergency response personnel to know prior to their arrival at the incident site.

*Lars Larsson is a Senior Product Manager at Schneider Electric – Global Solutions. He holds bachelor’s degrees in process automation from Telemark Technical College in Norway and control engineering from the University of Sheffield (UK). He is a certified Eur-Ing and has an MBA from the University of Durham (UK) to complement his 22 years of oil and gas pipeline industry experience. He has published multiple articles in global journals focused on pipelines.

Image credit: ©James/Dollar Photo Club