Joseph McMullen and Livia Wiley, Schneider-Electric, USA, explain the importance of simulation technology when designing LNG plants.
Transporting natural gas over long distances is only economically possible in the liquid phase. It is this reason, along with the increased demand of natural gas since the beginning of this century, that has driven the expansion of LNG plants.
Furthermore, recent events have demonstrated how dynamic the LNG market is and will remain. Events such as the Fukashima Daiichi nuclear disaster in 2011, the discovery of deep offshore gas fields, and the boom of shale gas in the last year have led to a staggering market forecast. These factors have pushed companies to liquefy offshore natural gas and led to the US reconverting import terminals to export facilities. By 2020, it is expected that the LNG market will be approximately 10% the size of the crude oil market. This is compared to approximately 1% in 2010. With this expected growth in market share, plants must address the major challenge of energy recovery to minimise cost under increased demand.
The challenge of design
As with any oil and gas installation, the design phase of an LNG plant is done in two successive stages – the pre-project phase and the project phase. For both stages, numerous simulation software tools – both steady state and dynamic – play a critical role.
The pre-project phase is where the most suitable scheme for the gas processing, specifications, local constraints and economic considerations are all chosen. For cost estimate purposes, and to ensure that this equipment can be manufactured, only the sizing of the main equipment is done at this level. In this phase, mainly steady state simulations and in-house tools are used for pre-sizing the main pieces of equipment.
The project phase is where the plant is defined in details based on the scheme chosen during the pre-project. This is also where the simulation scheme is finalised and the details are used to perform the dynamic simulations that check the robustness of that scheme, plant availability and the startup procedure. With safety being the major concern, a detailed analysis must also be performed to properly size the flare and flare network, which is of the upmost importance in an LNG facility.
It is no secret that mistakes in design can lead to lengthy shutdowns or drastic reductions in capacity for plants. The scrub column, an essential piece of equipment that’s main purpose is to remove freezable components to prevent problems during liquefication, is a perfect example to illustrate the difficulty of designing an LNG plant well and the problems that can ensue if done incorrectly.
The scrub column is located after the first cooling, with a propane loop, but upstream of the main cooling. This distillation column has a feed of lean gas and an injection of solvent in the condenser. The scrub column can have different configurations for better energy integration. The variables of operating pressure, though, often pose a critical compromise. Why is this important? High pressure liquefication is favourable in terms of overall plant efficiency, but pressure that is too high can lead to operating problems if the properties of the vapour and liquid phases are too close.
With LNG plants on the rise recently, a trend of scrub column problems is evident in the initial startup of some plants. Although some of these issues were fixed by simply lowering the pressure of the column, others required more challenging modifications – leading to a significant delay. Used to determine which solvent and column pressure are optimal, simulations on the scrub column can help improve energy integration while providing a safe pressure range. Thus, simulations of the scrub column determine a critical, but delicate, line to walk in terms of operating pressure that is essential to reaching satisfactory plant behaviour.
The necessary tools
The process engineering community holds strong opinions about which simulation tools are suitable for designing an LNG plant. It is true that not all tools are created equal, but what makes them suitable for proper plant design? Most would argue that the quality, versatility, and robustness of a simulation tool is what set suitable offerings apart. It must accurately represent the thermodynamics of the actual process with rigorous specifications for, and sizing of, equipment – including separation units, pressure changers, heat exchangers, etc. – while adhering to operating constraints.
So, what simulation tools are needed for the best possible design of an LNG plant?
Thermodynamic modelsFirst, accurate thermodynamic models are of paramount importance to minimise excessive design margins that hamper the operational effectiveness and efficiency of LNG plants. These margins are often due to the low temperature approach in the LNG heat exchanger purity specifications in the various distillation columns. If for some reason a performance improvement is needed for a particular model, a retune of experimental data for some model parameters, such as the binary interaction parameters (BIPs), may be needed. Thermodynamic models frequently used for modelling LNG conditions must include accurate parameters for the following:
- Liquid-vapour equilibrium and vapour density.
- Enthalpy and entropy.
- Liquid density.
- Solid deposit.
Simulation tools that have thermodynamic models to address the aforementioned conditions should qualify the technology as a suitable offering for effective LNG plant design. Dynamic simulation models are also a necessity for effective design stages.
Dynamic simulation models
Dynamic simulators model the transient nature of a plant over time. Because it is rigorous, it can be used to verify process designs and conduct engineering studies. With energy use a major concern, it is important that LNG plants are as heat integrated as possible. This concern stems from the sheer number of recycle loops. When using suitable dynamic simulation tools, the convergence of the distillation columns and recycle loops are relatively easy to reach – even with a poor initialisation. This convergence process is essential to understanding the health of the plant, as severe convergence troubles are often a strong sign that the plant will be unstable and difficult to operate.
Simulation and training
The applications of a process model does not end with dynamic simulation. Once a model is up and running in a dynamic state and emulating the plant, the simulation can be used to execute control system design, controls checkout, startup and commissioning, and operator training. A dynamic simulator without these capabilities limits the lifecycle of the model.
With the ever-shrinking talent pool and the accelerated rate of retirement, plants are tasked with training a decreasing number of engineers and operators to take on added responsibility. Operator training simulators (OTSs) can be used for greenfield and brownfield LNG plants to improve operator capabilities, agility, and reliability. Operators need to know how to operate the distributed control system (DCS) and the plant because missteps can be costly in terms of safety, availability and profitability.
When designing an LNG plant, the simulation and training steps should be kept in mind. Building controls and safety schemes into the design model during the project phase will improve the overall efficiency of the design, reduce overall project time, and set operators up for success. This inherently reduces the overall capital costs and operating costs.
LNG plants are remarkably complex and in high demand of late, with no signs of slowing down. However, without suitable simulation tools, the high energy integration of LNG plants, from several mass and energy recycle loops, and drastic purification and liquefication processes make cost-effective design a significant challenge. To determine the optimal design of an LNG plant, engineers need simulation technology that offers critical process insights to make informed decisions.
Suitable technology for this level of decision making must include three key components. First, it must contain the required thermodynamic models for robust and reliable convergence methods – this includes solid detections. These models help engineers easily compare the different process schemes at the design stage and choose the best one for the plant to run as safe and efficiently as possible. Dynamic simulations are the second key component. Performed early in the project, dynamic simulations check the stability and flexibility of the plant to determine the best possible startup procedure. When searching for the right dynamic simulation tool for a plant, it is important to find one that is both robust and versatile to ensure engineers have a thorough assessment of the plant’s stability and flexibility to select the most appropriate startup procedure. And finally, it is important to keep in mind the lifecycle of the model. Beyond design, how will you use the simulation to improve the efficiency, safety, and profitability of the plant, today and into the future? All of these considerations should be made during the design phase of the project to meet precise, but possibly changing, specifications.
To succeed in this expanding and challenging market, LNG plant engineers and operators must ask themselves, ‘are we using the right simulation tools for plant design?’
Read the article online at: https://www.lngindustry.com/liquid-natural-gas/05102017/simulatingforsuccess/