Prevention is better than cure
Introducing high fidelity simulators for LNG plant verification and validation.
Private sector employment in the US increased by more than 1 million jobs between 2007 and 2012 (approximately 1%). Oil and natural gas industry jobs increased 40% during that same period according to the US Energy Information Agency (EIA).1 During that time, LNG production increased by 25% and the EIA forecasts that demand and production will continue to increase until at least 2035. With this kind of projected growth, LNG plant owners/operators must find ways to shorten design and build times for new plants and plant expansions without sacrificing safety. Dynamic, real-time simulation can play a significant role in meeting these challenges.
Traditionally, dynamic simulators have been used for control room operator training. These simulators replicate the control room environment and distributed control systems (DCS) user interface. The benefits of simulation training are well known and well documented across the energy industry. But beyond operator training, today’s high-fidelity simulators can also serve plant owners with their growth strategies.
The use of dynamic, real-time simulation for verification and validation (V&V) is a process that has been proven as effective in the nuclear energy industry and it transfers well to LNG facilities. Using an approach similar to that of the nuclear industry, LNG companies can evaluate the feasibility of the design and engineering for a new plant or expansion before it is built. In addition, verification of plant modifications helps ensure the expected operating results. This ultimately saves money throughout the design and construction process.
When used for V&V, simulators reduce design cost and project risk for:
- Validation of plant system process designs.
- Design and validation of digital instrumentation and control (I&C) strategy.
- Design and validation of human-machine interface.
- Use as a platform for human factor engineering (control room design).
- Validation of plant operation procedures.
Understanding simulation V&V
The US Army began using software V&V in the 1970s as part of its Anti-Ballistic Missile System programme. V&V standards were subsequently developed by the Institute of Electronics Engineers in 1986 and 1993,2 and by the American Nuclear Society in 1987.3 The American Society of Mechanical Engineers V&V standard is in review.4 V&V is now widely used by government organisations such as NASA and the Department of Defense.
As coding and simulation technology have developed over the years, so has V&V. Today, high definition (HD) simulators, also called high fidelity simulators, create a V&V platform to help identify design errors before they are incorporated into plant construction and speed commissioning.
For clarification, the AIAA Guide for Verification and Validation of Computational Fluid Dynamics Simulations defines V&V as:
- Verification – the process of determining that a model implementation accurately represents the developer’s conceptual description of the model and the solution to the model.
- Validation – the process of determining the degree to which a model is an accurate representation of the real world from the perspective of the intended uses of the model.
For years the oil and gas industry has relied on steady state simulation to design individual systems around overall plant production objectives. While suitable for individual system design, there is a trade-off between engineering analysis, real-time performance and continuous operation. These systems lack the feedback from integrated plant operations. Even with the latest advances in computer technology, the computational fidelity of steady state models cannot faithfully duplicate real-time performance across the entire range of plant operation. Real-time models must return accurate results though the full range of system operations. However, steady state codes generally are not numerically robust under all operating conditions to be able to do this.
Dynamic, high fidelity simulation gives owners/operators the ability to holistically view a plant’s operations in real-time. The modeling shows the integration and interaction between the plant’s piping, equipment and control system, and it reproduces the integrated plan dynamic performance for a wide range of start-up, steady state and shutdown operations.
Having the model work as an integrated system allows engineers and operators to see primary, secondary and tertiary effects. These model developments have been spurred by users’ need to defend the validity of simulators’ responses to local, state and federal regulators, and to the control system vendors that use the simulators to validate their products.
When built in conjunction with design data, this type of simulator helps validate the interface design between systems. The following questions can be answered:
- Has the controls vendor faithfully implemented the strategy and coded the system correctly?
- With the control strategy provide the designed results?
- Will the operating procedures work in practical application?
- Will the plant behave as expected, especially the secondary and tertiary effects that may arise from a plant upset?
- Are there better ways of displaying information to make the plant operator controlling this particular process more effective?
A different development process
Using the simulator for V&V requires a different approach than a conventional training simulator. Traditionally, all plant design data is collected and used to build the simulator models. V&V simulators require an iterative process of building models where interim deliveries help the engineers validate and adjust their design concept and the plant model grows over time along with the design process. This approach provides value early in the design/ build process of the plant.
Newer code developments have the computational capability to allow simulators to preserve the simulation fidelity and accuracy while running in real time. These tools can run synchronised real-time simulation models for full-scope balance-of-plant systems. Engineering grade simulators solve systems of differential equations and exchange results as feedback between the networks of the multi-physics interactions. The codes are synchronised so the exchange of calculation results and the code execution happen according to the same timing as in the operating plant.
The simulator must match the performance of the real plant within certain tolerances, and that performance must be repeatable. In real-time simulation the model must run transients from real plant data and they must represent the realistic performance at any given time.
Linking the dynamic real-time simulation models to the plant’s DCS provides the best way to validate performance. Traditional loop-back testing platforms used by many DCS vendors do not provide a suitable level of integrated plant performance data to yield accurate, integrated plant response to an operator’s action or a plant malfunction. Dynamic, high fidelity simulators are capable of predicting a plant’s behavior in both normal and abnormal operating conditions. In contrast, older training simulators were designed by building code around the transmitters to produce a desired response that the operator could see. They relied heavily on plant performance data. When the plant is still under construction, no data exists and the simulator must reliably predict plant performance based upon solid first principles modeling.
Human factors engineering
Design based on how humans naturally interact with technology is called human factors engineering (HFE). It is a more developed process of what is commonly called ergonomics.
The simulator allows HFE to accurately test the effectiveness of the user interface for the plant control system. With the advent of DCS, more information about the plant is available to the operator than ever before. How that information is displayed, how various colours relate to plant status, what information is always available to the operator or shift supervisor are all critical to safe and efficient plant performance. As operators respond to abnormal and emergency events in real-time on the simulator, HFE can test the effectiveness of the following:
- Location of information
- Density of information
- Information overload
- Problem diagnosis and resolution
- Alarm management
- Electronic procedures
- Regulator comfort.
Employing human factors engineering (HFE) can reduce operator fatigue and the chances of human error, thus reducing the risk of plant upsets.
For today, simulation-based V&V presents an opportunity for LNG companies to build or expand their facilities to meet growing demand and maximize profitability through more efficient designing, building and commissioning.
The goal of the simulation-based V&V is to find problems when they are least expensive to fix. Having a commissioning team wait while a permissive or set-point problem prohibits plant start-up is not only frustrating, but also extremely expensive. Thus, the financial benefit of simulation is clear.
The complexity of modern LNG facilities dictates the need for simulation for operator training. Purchasing the simulator early in the plant design process and using it to validate plant performance can shave days off plant start-up, paying for the simulator that will also be used as the life cycle training device.
- ‘Oil and gas industry employment rising fast’, available at http://www.energyglobal.com/news/exploration/articles/Oil_and__gas_industry employment_rising_quickly542.aspx#.Us6FXWRdX88
- ‘IEEE Standard for Software Verification and Validation Plans’, IEEE 1012-1986 and ‘IEEE Guide for Software Verification and Validation Plans’, IEEE 1059-1993.
- ‘Guidelines for the Verification and Validation of Scientific and Engineering Computer Programs for the Nuclear Industry’, ANSI/ANS-10.4-1987.
- ASME Code and Standard V&V 30, ‘Verification and Validation in Computational Simulation of Nuclear System Thermal Fluids Behavior’, Footnotes 2-4 from Stoots, C. M. , et. al. , ‘Verification and Validation Strategy for LWRS Tools’, INL/EXT-12-27066, September 2012.
- ‘AIAA Guide for the Verification of and Validation of Computational Fluid Dynamics Simulations’, AIAA Standards G-077-1998e, from Stoots, C. M., et. al. , ‘Verification and Validation Strategy for LWRS Tools’, INL/EXT-12-27066, September 2012.