Welcome everyone back to LNG Academy Episode 4. Before we jump in, I’d like to acknowledge a few things.
First, the mission of LNG Academy is to spread LNG Education around the world. We aim to be as factual and non-biased as possible and focus our discussion on technical issues, with the hope that some of this knowledge will be beneficial to this great industry that we are proud to be part of and contribute to.
The opinions expressed in this podcast are solely based on those of the hosts. With that I’d like to welcome back my co-host, Mehdy Touil, LNG operations specialist.
Lastly, I’d like to thank the sponsor of this podcast, Capstone Industrial Training Solutions. Capstone are leaders in the LNG and industrial training space. For more information on Capstone please visit capstoneits.com
Now let’s jump into our topic today, a fun one, hot air recirculation.
We know that ambient air temperature can affect the performance of LNG facilities quite greatly. The trend towards air-cooled systems has gained traction in modern LNG plants versus seawater cooling out of environmental concerns whenever possible, as we’ve seen with Woodfibre LNG for example. Weather conditions, particularly in the speed, and direction of the wind, are significant factors that must be considered while analyzing the occurrence of Hot Air Recirculation. Mehdy, How would you define Hot Air Recirculation?
Thank you, Alex.
Most LNG trains are designed with a pipe rack located in the facility’s center, flanked by compressors, cooling heat exchangers, and other pieces of equipment for processing and liquefying natural gas. The air-cooled heat exchangers use large axial-flow fans to move air across finned tubes, which results in the elimination of heat and/or the condensation of the process gas.
Hot air recirculation is a well-documented phenomenon that affects air-cooled LNG plants. The hot air from air-cooled exchangers flows back into the cooler’s intake or the intake of other equipment such as gas turbines whose thermal efficiency is strongly influenced by the ambient temperature.
The issue is more pronounced in multi-train configurations. You have hundreds of air coolers operating simultaneously, either through cross-contamination or re-circulation. To ensure efficient plant performance, the temperature of the air supplied to both the air coolers and gas turbines must be similar to the surrounding air temperature. However, due to the crowded layout of LNG facilities and the substantial amount of heat released on site, there is an omnipresent risk of warm air recirculating and affecting the performance of air coolers and gas turbines. The plant performance could be considerably impacted as seen in the case of Gorgon LNG which was operating at 13% below its nameplate capacity. We’re talking about hundreds of millions of dollars of losses.
Oh my god, those are staggering numbers Mehdy. This is a massive issue. Is it a design or operations issue or both?
The airflow patterns surrounding an LNG facility can be very complex, and when the emissions from various unit’s merge together, it can be challenging to identify the causes of certain problems. Computational Fluid Dynamics (CFD) simulations have become a widely accepted tool for investigating the problem of hot air recirculation. These studies are routinely used to evaluate the potential for this issue during the design and also operation phase of such facilities but also to develop and optimize plant layouts that minimize recirculation.
To accomplish this, a comprehensive 3D model of the site and adjacent terrain is created, which is then used to simulate the airflow across the site in different wind conditions. Employing advanced techniques to analyze data from the model enables the identification of individual hot air plumes originating from specific units of the facility and the computation of their effects on air intakes throughout the site. The simulations enable monitoring of the hot air plumes emanating from all the equipment in the vicinity of the site, to determine whether they are being drawn into the intakes of crucial equipment.
Through simulations, process engineers, designers, and operators can gain insight into the possibility of Hot air recirculation and evaluate the efficiency of measures and modifications that can be implemented to minimize the risk of this happening. It is then possible to attribute the origins of air recirculation across all units and employ this information to create a comprehensive overview of the site’s response to variations in atmospheric conditions and design changes.
Ok so they try to engineer this as much as possible, but it is a complex issue. We have seen in the case of Gorgon LNG, Petronas MLNG or Yamal LNG that the CFD simulations were not enough to avoid the problem during operations.
Any simulation has limitations. The emphasis by LNG project developers is on meeting process needs and CAPEX, while hot air recirculation is not always a priority during the design phase. The rejection of heat to the atmosphere by air coolers can have an impact on the ambient temperature and localized wind conditions in the surrounding areas, which in turn can lead to reduced LNG production. This is particularly true when it comes to the turbine drivers of refrigeration compressors, as well as propane coolers and condensers.
However, frequently occurring issues in designs can be easily avoided. When it comes to site-wide planning, it is important to consider factors such as the prevailing wind direction and adequate spacing between different units, in addition to aligning LNG trains and banks of coolers. The positioning of gas turbine intakes and the exhaust stack heights also require careful consideration. To prevent upstream obstacles from interrupting the supply of fresh air to the units, it is also advisable to always position the coolers at a higher elevation than the surrounding equipment. Arranging equipment that generates hot air into common groups can assist in decreasing Hot air recirculation by combining the plumes, which enhances their buoyancy and enables them to rise quickly away from the site. You can also ensure that units with different exhaust velocities are not placed nearby, such as avoiding combining forced and induced draft air coolers in the same bank. Also, as modular construction is gaining traction in the LNG industry, providing adequate distance between modules and the main pipe rack is crucial, as spacing will promote and improve airflow to the coolers.
Interesting, sounds like my whole fluid mechanics courses. So what about the performance of the individual air-cooled exchanger?
The guaranteed performance of every condenser unit is reliant on the presence of enough air at specified temperatures at the intake. The specific number of condenser units necessary for a particular plant layout, task, and spatial limitations determines how the bay will be constructed on the actual site. However, the performance of a multiple-bay design may differ from that of a single bay due to anomalies in airflow distribution and hot air recirculation. It should be noted also that while the parameters of wind speeds, direction, and the relative humidity are location-dependent, historical data might not be always available to accurately establish localized wind patterns. In some cases, the selection of a suitable site for the construction of a liquefaction facility mandates the deployment of weather stations to measure and monitor key meteorological variables over extended periods, at an extra cost that developers unfortunately are not always willing to add to their CAPEX.
Yeah, the production loss is an obvious way to diagnose hot air recirculation issues. Are there other ways you can diagnose this?
Although modifying the design of a virtual plant is considerably more cost-effective than implementing alterations in an actual facility, most LNG operators will unfortunately experience the impact of hot air recirculation on production in the post-design phase when it is too late to implement structural changes. LNG facilities are located in coastal areas that experience variations in wind patterns throughout the day. By having a technique to map the effects of recirculation during these fluctuations, operators can have a useful resource to diagnose issues and optimize production. Some hardware modifications can be implemented as well to reduce the severity of recirculation like the installation of chimneys or windscreens on the impacted condensers. These changes should be considered with caution as they may also reduce inlet airflow. The effect on the cooler performance could therefore be worse than the gain recovered on hot air recirculation. To minimize hot spots, it is also possible to enhance circulation by adding or removing structures from the facility whenever feasible.
Beyond the analysis of weather, CFD data, and cooler performance, establishing a diagnosis for an existing facility is a complex exercise that involves the installation of dozens of temperature sensors at the coolers intakes, in an evenly distributed fashion to “visualize” the phenomenon throughout the day. Once this information is acquired, the next logical step is to support the plant APC application with a predictive model that would allow the operator to take actions in advance to mitigate the hot air recirculation impact and optimize LNG production, like changing refrigerant flow or modifying its composition using preset values.
Wow thanks Mehdy. Those are a lot of practical examples. Thank you so much for that are there any other tools that can help mitigate the impact of hot air recirculation?
CFD simulations are computationally intensive and running them can be time-consuming and expensive. This is where AI-accelerated CFD applications come in and could help solve the problem of Hot air recirculation. By leveraging the power of artificial intelligence and machine learning, these applications can significantly reduce the time and computational resources required to perform simulations.
Machine learning models can also be trained on historical data from operating plants, including process variables as well as live data from online sensors. By analyzing this data, the model can detect patterns and correlations that may not be apparent to human operators, allowing for early warning of potential hot air recirculation events and enabling proactive control measures to prevent production losses.
Either from an early design perspective or from an operational one, the future of process engineering will be AI-driven. I’m not talking about ChatGPT which is far from being a reference for engineering. So the technology expected to become an essential tool for solving operating problems and optimizing the design of LNG facilities. Although I would say that we’re still years away from that scenario.
Good point. So I think the savings you would get with increased production would definitely help fund the advancements of a lot of these technologies. So we can see, like you’re saying, these accelerate but still very much in its infancy. We still have to rely on what we can see from our design and human operations. Thank you Mehdy, this has been a really great topic. I’m sure it’s one of those things that has been a thorn in your side as an LNG operator all these years. Hasn’t it?
Ok well, thank you that’s it for today. So thank you everyone for tuning in to Episode 4 and we’ll catch you on the next episode!