August 2010 Vol. 237 No. 8


Improving Gas Controllers Competencies Through Computerized Simulation

Promigas is the only transmission company responsible for transporting natural gas to the majority of towns on the Colombian Atlantic Coast in South America. The natural gas is received from several production companies connected to Promigas’ pipeline transmission system, such as Chevron, Pacific Rubiales, Geoproduction and Solana.

Promigas renders service to more than 4 million users connected directly to its system, or indirectly through the local distribution companies (LDCs). The system, which includes 4 compressor stations, is managed from the control center in the city of Barranquilla. In 2009, it experienced a maximum peak of 0.58 Bcf/d.

The Natural Gas Control Activity
In our operation, gas controllers are responsible for monitoring and controlling the pipeline conditions, carrying out the system’s simulations, responding promptly to emergencies, taking care of the nomination process with clients and producers, and following the rationing codes during emergencies.

Given the complexity of activities involved, the firmness requirements of contractual commitments and our obligation of protecting the safety of communities, the critical importance of the gas controller position should be recognized with a certification process that validates and qualifies the ability of the individual assigned to this post.

With this in mind, we attended the AGA conference in 2007 and participated in the roundtable for Control Room Management (CRM), where we decided to implement Gas Controllers’ Certification according to the published standards to ensure our gas controllers’ competencies and skills. In order to achieve our objectives, we developed a certification process based on a systematic procedure which takes into account both normal and abnormal conditions in the pipeline.

Selection Of Standards
When we started to implement the standards to qualify operators (OQ), we first studied the API 1161 Guide for its applicability. Later in the research process, we concluded that the ASME B31Q standard was more suitable to our natural gas operation, since API 1161 was at that time mainly used for liquids while the ASME B31Q is a continuation of the ASME B31.8, on which our design, construction, operation and maintenance of natural gas systems is based.

Our main objective was to comply with the ASME B31Q Standard by means of a skill assurance program for gas controllers. A program that performs the “Operation of Natural Gas Transmission and Distribution Systems” as a main covered task was instituted. This skill assurance program is designed for operators who exercise this task daily. The program will allow them to recognize and effectively respond to abnormal operating conditions (AOCs), conditions that they could face while performing their job.

We adopted Appendix 6A of the ASME B31Q – 2006 Standard as the basis for identifying AOCs that could be faced by our operators. These abnormal conditions are listed below:

• Natural gas unexpected leaks;
• Fires or explosions;
• Unexpected pressure deviations (for example: pressure increases, decreases or absence or lack of pressure);
• Unexpected deviations of the rate of flow (for example: high flow, low flow, no flow);
• Gas pipeline damages (for example: damages during an excavation activity, atmospheric discharges, tornados, floods, earthquakes, among others);
• Safety equipment triggering out of schedule tests (safety equipment such as: safety valves, emergency systems, shut-down systems due to high pressure, other pressure related shut-downs, high temperature related shut-downs, among others);
• Unexpected changes in the operation status of the pipeline system without having been remotely or directly activated (for example: compressors start-ups, compressors shut-downs, valves opening or closing);
• Interruptions or failures in communication and/or in the control or power systems; and
• Inadequate odorizing or gas smells (odor) reports.

In this skill assurance program for gas controllers, we defined the criteria for training, assessment, documentation and registration. Our training process is based on the last revision of the gas pipeline operations module from the Contren Learning Series, which covers the following specific topics:

• Introduction to the gas pipeline industry,
• Gas pipeline pneumatic equipment,
• Communications for gas pipelines,
• Routine operations in control centers, and
• Abnormal operating conditions,

The Contren Learning Series gas pipeline operations module is part of the operator training and obligations program by the American Petroleum Institute (API) and the National Center for Construction, Education and Research (NCCER), Figure 1. The module was designed to comply with the regulatory requirements for operator qualification. After implementing and documenting the framework for the gas controllers qualification program, (even though not required by Colombian regulation), we continued with the NCCER qualification process in a joint effort with TVG Global Services from Houston.

During the first year of implementing the program, we were able to measure our gas controllers’ qualification by using the curriculum designed by the U.S. natural gas industry through API and NCCER, with positive results. The outcome was an average of 80% in the written test that included an evaluation of controllers’ knowledge of the following gas pipeline topics: the industry, the hydraulics, major equipment, communications, control room routine operations, gas measurement, gas quality and AOCs.

Enhanced Training Process
A year after administering the written testing (table top approach), we realized the need for a tool to assess and train the gas controllers in a virtual control room environment, very similar to the real control room condition. At about the same time, we also strengthened the control room technologies by completing the implementation of a Wonderware InTouch Scada Application, added an in-house developed nomination system and implemented a Real Time Pipeline Model by Gregg Engineering.

Soon thereafter, we had an opportunity to witness a Real Time Simulation-based Virtual Pipeline Trainer developed by Gregg Engineering and installed by Cheniere Energy, an LNG and pipeline operator based in Houston, in an actual virtual control/training room environment. After viewing Cheniere’s implementation and success with their operators, we concluded that this was the training methodology to follow for a pipeline transmission company with some adjustments.

A Virtual Control Room provides Integrated SCADA HMI working in conjunction with a virtual simulated Promigas pipeline System. Our controllers can be trained using our own SCADA HMI on a virtual pipeline model of our own pipeline system, which ensures exactly the same behavior as our natural gas transportation system.

Simulation-based Trainer

To understand how the Trainer evolved and works, let us start with a typical gas pipeline SCADA system implementation. In the actual SCADA, the flow of data is coming from the field RTUs and the set points are flowing from the SCADA system to field RTUs. The time movement is the real world clock.

If still using another copy of the same SCADA system as Trainer SCADA, but replacing the field RTUs with a virtual pipeline (a transient simulation model), we now have a simulation-based trainer. The pipeline data flow from virtual pipeline to trainer SCADA and set points are flowing from the trainer SCADA system to the virtual pipeline. In Trainer mode, the time movement can be real clock scale or compressed time.
With our past experience with the Gregg Engineering real time model implementation, we realized that it would be very easy to link our SCADA HMI to the virtual pipeline model. This way the gas controllers could be trained and assessed by using the same SCADA HMI that they use on a daily basis. A trainee Interface that looks and behaves exactly the same as their daily tool, rather than a generic HMI to monitor the pipeline hydraulic reactions, was created. This achieved one of our main goals, but we still needed to be able to train our operators on how to recognize and manage AOC’ events.

Figure 2: HMI Trainee Interface showing simulated emergency (AOC)

In addition to pipeline hydraulic responses, gas controllers also have to manage other control room activities (tasks). When an event or an AOC event occurs, controllers will have to perform the following:

1. Notifications – Trainee actions that have no effect on the hydraulic simulation but are still an integral part of control room management. For example, did the gas controller notify the correct company and emergency personnel for the AOCs?
2. Delayed Operations – Pipeline control actions that cannot be initiated through the SCADA system either because they do not have a remote actuator or because of a loss of SCADA communication. An example would be the closure of a remote valve that a field person has to be dispatched to perform.
3. Gas Scheduling – Trainee actions after an event, the trainee is required to ration (re-schedule) the available supply of gas to the customer and then notify customers to curtail their deliveries.

Gregg Engineering enhanced the Trainer with these three additional requirements and provided a simplified Trainee Interface in case the full SCADA MMI Interface was not used. Figure 3 illustrates the enhanced data flow diagram for the new Trainer.

Figure 3: Information Flow in the Trainer

After the Trainer software was enhanced, we configured the virtual gas pipeline and integrated it with our Wonderware InTouch system. We then were ready to assess the skill of all our gas controllers during AOC´s events. The Enhanced Trainer allowed us to train and qualify each of our controllers to recognize and respond to the AOCs while carrying out their normal activities in a simulated control room environment. This was extremely important because most of the controllers never experienced how AOCs would look in the SCADA HMI screens with which they interact every day.

In the first exercise, the group had a positive response and was able to identify operational conditions that could create future risks. These risks were mainly associated with the continuity of the operation, the promptness in emergency attention and the service quality rendered to our clients.

Training scenarios used during the first exercise included:
1. Certainty/uncertainty of pressure measurement locations along the gas pipeline;
2. Validation of by-pass valve status; and
3. Observing the pressure and flow pattern trends during major events and pipeline internal cleaning procedures.

Final Results
The initial assessment exercise with Virtual Pipeline Trainer resulted in many positive inputs for our operation. Since the implementation, we have been able to identify many opportunities for improvement and recognized a number of leverage points for optimizing and improving the efficiency of our processes. The Trainer allowed us to gain valuable lessons at a low cost with no risk to the communities we served. We have greatly enhanced our reliability and safety of the natural gas transportation and distribution service rendered to our clients.

Figure 4: Results from the First Assessment Session Using the Trainer

An important application of this type of tool, Virtual Pipeline Simulation (for computer based evaluation), is the new rule published by the U.S. government in December 2009. This application is a more realistic than table top test and offers the operating company a broader scope of guarantees with regard to the skills of a gas controller when facing emergency events, therefore complying better with the spirit of the law. Figure 4 shows the mean result of the key characteristics evaluated with this implementation.

At the end we would say that the operator trainer took us close to the compliance of the Control Room Management Final Rule, non mandatory in our country, but adapted by Promigas as the best practice for pipeline operations. Regarding table top methods, we would say there is still a gap between both methods and worth it to spend more effort in developing a computerized based solution that ensures your controllers would have full resources to face a real emergency upon getting trained, using a systematic procedure in which a daily basis tool and experiences are combined.

Alejandro Villalba is Operations Manager for Promigas in Barranquilla, Colombia. He received his Industrial Engineering education from Universidad del Norte in Barranquilla and has been with Promigas for 18 years. Villalba has led in-house projects with local and international contractors in developing management information systems for natural gas operations, nominations and other field systems. He has also been instrumental in other technological implementations related to gas control and simulation and was a contributor in preparing the gas quality rule in force for hydrocarbon dew point limits in gas pipelines in Colombia.

Carlos J. Castaño has been with Promigas since 2002 and holds degrees in Mechanical Engineering and Projects Management from Universidad del Norte. Castaño has among his responsibilities the control of unaccounted-for gas, measurement devices operation and controller’s training program. He is certified as Operations and Maintenance Instructor by the National Center for Construction, Education and Research – NCCER –. As a Promigas employee he has led projects of OQ Rule, Online Simulation and Controllers’ Training Software.

Michael L. Istre is the Director of Software Development at Gregg Engineering in Sugar Land, TX. He has been with Gregg Engineering for 16 years and is responsible for the WinFlow and WinTran product lines. At Gregg Engineering, Istre’s responsibilities include all aspects of the simulator’s design from the interfaces to the hydraulic run engines as well as their integration with external software. He has been developing Windows based programs for Gregg Engineering since the early 1990s and was instrumental in bringing the first Windows based pipe simulator to the market. Istre is a graduate of Texas A&M University.


{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}