The quest to squeeze more capacity from existing assets undoubtedly is as old as the chemical industry itself — and efforts certainly aren’t flagging. Indeed, debottlenecking projects are proliferating around the world.
For example, the raft of 2019 project announcements includes the debottlenecking by Solvay, Brussels, Belgium, of its hydrogen peroxide plant in Jemeppe-sur-Sambre, Belgium, with similar projects to follow at its sites in Bernburg, Germany, and Voikkaa, Finland; efforts by Oxea, Monheim am Rhein, Germany, to continue improving and debottlenecking manufacturing processes at its five existing carboxylic acid production units in 2020 in preparation for the goal of bringing a sixth world-scale production plant on stream in 2021; and a project by BP’s joint venture Lotte BP Chemical Company, Ulsan, South Korea, to add 100,000 metric tons (m.t.)/yr via debottlenecking, bringing acetic acid capacity there to 650,000 m.t./yr. In addition, Eastman Chemical, Kingsport, Tenn., has completed debottlenecking of its diethylhydroxylamine unit in St. Gabriel, La., that increased capacity by 15–20%, with other similar projects in planning stages.
In North America alone, efficiency drives by the chemical processing and oil and gas industries has prompted $1.3 billion in debottlenecking projects, according to research by Industrial Info Resources, Sugar Land, Texas.
The Power Of Simulation
“Debottlenecking and retrofitting in general are ideal situations in which to use process simulation. They are definitely strong leads for our software,” says Stéphane Déchelotte, CEO of ProSim, Toulouse, France.
This is especially true, he adds, when a customer already has a plant that is up and running and so can use its own data to validate the simulation of its process. “From there you can fully validate the simulation as a test case on a digital twin and identify exactly where and how the debottlenecking needs to take place.”
Two recent projects cited by Déchelotte used the firm’s ProSimPlus Energy software package to debottleneck and improve plant energy consumption.
One involves French agro-industrial group Avril, Paris, which heavily focuses on renewable chemistry and biofuels. The company’s Diester biodiesel brand, developed over the last 20 years, is produced mainly from rapeseed oil and is incorporated in a proportion of 8% in diesel fuel distributed from French service stations.
“We worked with Avril to reduce bottlenecks in a biofuel plant. The outcome was a dramatic reduction of more than 40% in energy consumption with an ROI [return on investment] of less than a year. This project involved modifying the distillation column and associated heat exchangers,” Déchelotte explains.
The second was for Air Liquide, Paris, on a steam methane reforming plant. The focus here was on the reformer and heat exchangers. The simulation project identified ways to reduce energy consumption by 10–20%, depending upon plant operations.
“You never know at the start of a simulation project what you are going to find, but usually the final savings are substantial,” he adds. “Both projects were performed within six months and the cost of each was under US $50,000. So we’re not talking about a huge investment. The cost of a study is very small relative to the possible savings.”
A simulation study usually uncovers a whole range of potential improvements but a company may not choose to do all of them, especially ones that require a plant shutdown, Déchelotte notes. After a company decides which changes it wants to make, ProSim carries out another simulation using just those parameters.
One niche business that is very important to ProSim is debottlenecking of nitric acid manufacturing plants, especially in North America. Its ProSimPlus HNO3 software package, honed for 30 years, is aimed at the traditional dual-pressure and mono-pressure nitric acid manufacturing processes, together with the nitrous vapor absorption units used in fertilizer, explosives and adipic acid manufacture (Figure 1).
“Usually we are working with quite old plants where there isn’t a lot of money to invest. So the companies are constantly carrying out retrofitting and debottlenecking projects,” Déchelotte explains.
“There are typically three aims to such projects: a reduction in emissions; a reduction in energy use; and an increase in production. With some changes, typically to columns and condenser coolers, we can achieve all three of these targets. It’s a very busy market, mainly in the U.S., and we have seen a huge increase in sales of the related software.”
A number of factors contribute to this success, he believes. General-purpose simulation software struggles with the complexity of the physical and chemical phenomena involved in nitric acid processes — typically gas-phase chemical reactions whose thermodynamic properties aren’t easy to model. The specificity of equipment used in such plants makes the processes tricky to simulate, too. Also, the software should take into account sizing parameters, for example pipe volumes and the spacing of absorption column trays, at early stages of process design because these can affect mass and energy balances if production is to be increased at a later stage. Operating companies that have developed in-house programs for the main unit operations often lack the flexibility to look at the entire production process.
The massive growth in availability of plant data spurred by growing adoption of the industrial internet of things (IIoT) promises better simulations, he adds. “Our software is used offline, but we are always interested in technology that lets us take advantage of all the data available on the plant. We can mine this and use it to validate our models. Being able to use many years’ worth of detailed data to do this gives us very, very accurate simulations so the reliability of our software is improving all the time.”
The Full Lifecycle
It’s also enabling the emergence of unified lifecycle approaches to simulation. This involves extending one process model throughout the entire lifecycle of the plant, from concept to operations. Its aim is to do away with the problems caused by legacy architectures and operating systems as well as the need for a high level of software programming skills and to take advantage of the emergence of the IIOT and artificial intelligence.
Such next generation platforms provide a digital twin of the plant through the process lifecycle that today’s tools can’t.
Many current simulators typically only support a single phase of the lifecycle and often are based on thermodynamics of different simulation vendors and disparate calculation methods, according to Aveva, Cambridge, U.K. This not only leads to lack of trust in the results but also causes substantial rework in having to build a new simulation model in each new tool. As a consequence, the results of each model are hard to compare.
Launched in March 2017, the SimCentral simulation platform is Aveva’s answer to this problem.
One engineering/procurement/construction (EPC) firm that adopted the platform saw dramatic benefits with a cooling water revamp project at a North American refinery. When the refinery operator asked, “What if we completely remove this area or switch this equipment?,” the EPC replied, “Give us two hours.” Prior to adopting SimCentral, such an evaluation would have taken two weeks. The EPC also noted that the learning curve isn’t steep, even for engineers with no prior experience of using process simulators.
Hyundai Engineering, Seoul, Korea, chose the same technology after seeing how quickly and efficiently it could model complex processes.
The company was looking at how best to design, engineer and build vacuum transfer lines. These can be challenging given their high velocity and two-phase flow, leading to poor separation in the vacuum towers that they feed.
“We chose SimCentral simulation platform after seeing how quickly and efficiently complex processes could now be modeled,” says Hyundai.
The platform’s ability quickly to handle and avoid critical velocities helped process engineers to find the right design faster, reducing both the time and cost of the simulation exercises, notes Aveva.
The platform also has received positive feedback in workshops with chemical companies, including for ease of debottlenecking process utility systems by fluid flow network simulation, better collaboration between modeling and control experts by early understanding of process controllability through unified dynamic simulation, and a 50% reduction in simulation effort across the entire plant lifecycle.
Seán Ottewell is Chemical Processing's Editor at Large. You can email him at firstname.lastname@example.org.