Apr-2013

Lessons learned from Mpc revamp of two naphtha crackers

A project overview includes changes made at the regulatory and advanced control level, and results obtained. Techniques and tools used in the project are highlighted

Vanessa Conz, Suzana Fuchs and Priscila Macedo, Moura Braskem
Mark Darby CMiD Solutions Doug Nicholson, IPCOS (UK)

Article Summary

Braskem is the eighth largest petrochemical company in the world and the largest polymer producer in the Americas. The Braskem Triunfo site includes two naphtha crackers with a combined olefins capacity of 1250 KTA ethylene and 685 KTA propylene. AspenTech’s proprietary DMCplus controllers were built for both olefins plants over the past 10 years, but were not fully operational.

In 2010, a project was initiated to review the performance of the advanced control systems and improve their operability, service factor and economic impact. A further objective was to increase the site process control engineers’ knowledge. The 2010 project covered two Braskem olefins units located at the same site in the south of Brazil. The site actually has three olefins plants and one aromatics unit. Two plants, referred to as Olefins 1 and Olefins 2, process naphtha. The third olefins plant uses ethanol as feed (green ethylene). The multivariable predictive control (MPC) project discussed within this context is based on the two naphtha-fed olefins plants.

Leveraging advanced controls
Braskem was formed in 2002 through the merger of several petrochemical and chemical companies. The company has a focused growth strategy and has acquired companies in Brazil and, more recently, in the US, Germany and other countries. Braskem has industrial units in the four largest petrochemical centres in Brazil. It utilises advanced control as a means to reduce variable costs, increase energy savings and improve productivity.

Olefins 1 started production in 1982. It is the largest ethylene producer in the complex and has the capacity to process 312 t/h of naphtha. The plant has 12 naphtha furnaces and two recycle ethane furnaces. Olefins 1 normally operates with high feed rates typically limited by raw gas compressor constraints. It is responsible for the pressure control of the site ethylene supply grid and is therefore affected by consumer demand variation. The base instrumentation is the oldest on site, with too many control loops in manual mode. These characteristics make Olefins 1 challenging for operators with a high number of interventions from the control panel.

Olefins 2 started production in 1999 and has the capacity to process 180 tph naphtha. The plant has seven naphtha furnaces, one being a swing furnace that can be used on either olefins plant, and one ethane furnace. It is more difficult for this plant to handle feed composition variability, as this unit processes more LGP and C4s along with naphtha.

Braskem owns several sites in Brazil. A particular challenge the company faces is job rotation of its people, such as production and control engineers. New plant projects require experienced senior engineers. To succeed with MPC, it is well known that site engineers require the process and tools knowledge to maintain the installed base of applications. The use of projects to accelerate engineer training and learning is strategic to the company.

Both olefins plants already have MPC controllers installed. Olefins 1 has had controllers since 2004 and Olefins 2 since 2001. Table 1 lists the controllers that have been installed. After plant shutdowns and minor revamps, some of the controllers were not operational and others were not working well.

Revamp project
The online factor in Olefins 1 was poor and the benefits in the existing MPCs were not being captured. The objective of the project was to improve the economic and qualitative benefits, as well as to increase the online factor. In Olefins 1, for example, the variability of C2 splitter impurities was high and the C3 splitters had a large propylene loss (>20% mass). In Olefins 2, the goal was to remodel the parts of the controllers that were not working properly from the last shutdown in 2005. In addition, a parallel project installed new actuators on the secondary air dampers of the furnaces in order to control excess oxygen more efficiently. The project was an opportunity to review the furnace models and to add the damper positions as additional MVs to the furnace controllers.

The site had just one engineer responsible for MPC for a number of years. In 2010, the company hired a trainee to become a control engineer, and a production engineer was assigned to the team to help the control engineers maintain the controller online factor. The project was a good opportunity for engineers from the site to learn about MPC project methodology and to develop expertise on advanced process control in order to handle future projects by themselves.

The scope of the project was challenging. There were many columns and furnaces involved, and few engineers to execute the work. To speed things up, and to empower a team environment, the control group decided to hire advanced control consultants. The project was justified on economic return. Although some of the equipment in the scope had MPC controllers, certain controllers were selected to be remodelled due to their poor performance. The contract was made based on consultant hours in order to have flexibility in the project scope (for example, in identifying benefits opportunities during the project). This approach requires discipline to calculate benefits and determine priorities.

Lessons learned and relearned
A multi-skilled project team was formed with control engineers, process engineers, production engineers, instrument technicians and operators. MPC consultants participated in all steps. DCS modifications were executed by the site control engineers. In addition, the site supply chain group provided important information about company optimisation objectives and economic scenarios. The skills diversity of the group was essential to the success of the project and reinforced the conclusion that teamwork is very important for MPC projects. The scope of the project initially involved regulatory control stabilisation to avoid flooding of the distillation columns, to allow operation closer to products specifications and, particularly for Olefins 2, better energy performance on the cracking furnaces. These were the obvious opportunities in MPC. However, working with the consultants gave a better understanding of the entire process and led to other opportunities. In order to prioritise these new opportunities against the initial scope, it was important to measure the benefits on an ongoing basis. This also helped project promotion and acceptance.
To ensure the success of the project, it was crucial to monitor the results and performance in order to help the operators and engineers use the new capabilities in the controllers.