Fast data transfer via Modbus TCP

Ferdi van den Brûle

With an annual processing capacity of almost a million tons of ethylene, the ethylene cracker in Moerdijk is one of the largest of its kind in Europe. The primary task of the redundant SIL 3 controllers is to shut down critical sections of the plant (ESD = emergency shut-down). They are also used in F&G (fire and gas) applications as well as BMS (burner management systems).

In addition to demanding safety, Shell wanted the ethylene cracker to offer maximum availability. Unwanted downtime leads to breaks in production and consequently to higher costs. “We need systems with high levels of availability and fault tolerance to ensure that the plant keeps running at all times. However, if a problem does occur in the plant – through overheating or excess pressure, for instance – then the system has to respond immediately in order to prevent explosion or fire,” explains Willem Weeda, the system engineer responsible for safety engineering at the Moerdijk ethylene plant.

The ethylene cracker was recently re-vamped, which involved replacing the old safety technology with 17 safety-related H51q systems. The re-instrumentation was carried out in autumn 2006 during a regular system shut-down performed every four to six years for maintenance purposes. Weeda continues: “The Hima safety systems fulfil our reliability and availability expectations completely.”

The systems handle around 2500 safety-related signals. Four of the H51q controllers are used to monitor one medium-pressure and three high-pressure boilers. Seven other systems monitor the ethylene plant’s 18 cracking furnaces. The remaining six controllers are employed in a butadiene recycler, a hydrotreater, a storage unit and two other units.

The safety systems are coupled to the Yokogawa process control system via Modbus TCP. Modbus uses Ethernet, which has the advantage that it enables the use of network components that are now standard, such as switches and Ethernet network cards. These in turn allow data rates around twenty times higher than were previously possible (10 Mbit/s). In the course of this project, the H51q systems were fitted with new CPUs to permit self-learning in redundant systems. When a CPU is replaced – which can be done while the plant is still running – all data and programs are automatically imported into the new CPU, thus eliminating the need for manual synchronisation. The system as a whole fulfils the requirements of SIL 3, even during replacement. Another benefit of these CPUs is their reload function, which allows programming changes without stopping the system. The scope of supply from Hima included the safety-related controllers, project management for the systems, hardware and software engineering and on-site service. Another important aspect for Shell was the smooth collaboration between Hima and Yokogawa Netherlands, who were the main supplier for the instrumentation. The schedule for re-instrumentation was, after all, very tight. The official order to deliver the safety technology was placed with Hima by Yokogawa in autumn 2005, and by summer 2006 the factory acceptance test (FAT) had already been completed at the main Hima works in Bruehl, Germany.

The controllers in the H41q/H51q family enable unrestricted use up to SIL 3, even with a single channel. Furthermore, these 2oo4D/QMR systems are TÜV-certified for applications up to SIL 3 in accordance with the international standard IEC 61508. Their modular design, scalability, redundancy and broad range of I/Os together mean they are easily adaptable to the application-specific conditions of the process at hand. Communication between the safety controllers is via safe Ethernet, while communication with the process control system takes place via OPC, Profibus and Modbus.

cpp 431

Shell Nederland Chemie B.V.

HIMA Paul Hildebrandt GmbH & Co.KG

SPS/IPC/DRIVES 2007