Apr-2010

Hazardous — or harmless? Machinery has not changed, but demands have

Based on the example of the ”reciprocating compressor” machine, this article will present current state of the art technology as well as explain how the use of modern, specialised monitoring systems have a positive effect for operators.

Eike Drewes and Jost A Anderhub,
PROGNOST Systems GmbH

Article Summary

It mostly happens unexpectedly and always at the wrong moment: When the reciprocating compressor gave out in a refinery without warning, due to major damage to the motion work, the spare machine was, of course, in the midst of being overhauled. Due to the failure, the hydrogen compression in the desulphurisation process was reduced by 50% for many days. In addition to the high repair costs for the damage, the refinery registered a substantial production loss.

Emergency shutdowns of critical production machines are always an annoyance, causing countless hours of overtime for maintenance staff as well as often eating away at maintenance budgets due to high repair costs. This is before one even considers the lost production if there is no sufficient spare capacity available

Where? What? Why?
One reason for the prolonged downtime was the extensive troubleshooting required. The emergency shutdown was triggered by a vibration sensor on the machine housing. Where, i.e. in which of the 4 cylinders, was the root cause of the damage? What was damaged? Can we trust the alarm at all or was it a false alarm (again)? With the typical single frame vibration sensor monitoring it was not possible to answer these questions. As such, no real informed decision could be made. When the broken piston rod was finally found 5 hours after the shutdown, the most important question was asked: why did the damage occur? How can such a damage and be iated downtime be avoided in the future?
Could the damage have been prevented or at least minimised so that the machine shutdown could have been avoided? To find out if there were measures that could be implemented, the operating company immediately began evaluating machine protection and online monitoring systems, which provide early damage detection and effective protection and thereby minimise the spread of any damage and continuously monitor and protect the compressor. Ideally, it should also be possible to record diagnostic information in order to predict and prevent future damage of this sort through appropriate measures.

At the end of the 1st decade of the 21st century, some machine safety systems already meet the advanced expectations and users can report about actual experiences with such equipment. Although machine safety systems have often been introduced in recent years to turbo machinery, such retrofits are rare on typical oscillating machinery such as reciprocating compressors and plunger pumps. For these types of machine, the relevant international standards only offer operators limited information as an aid to decision making. As it is mainly reciprocating compressors that are a major cost factor in maintenance budgets, due to high repair costs, particular attention is paid to such machines here.

Until now, simple machine safety switches have often been used to monitor vibrating and rotating equipment in equal measure, experience with reciprocating compressors has shown that broken piston rods and corresponding secondary damage cannot be sufficiently identified using this equipment. “Run” or ”Stop“ no longer provides a basis for the safe and efficient operation of machinery. This especially applies to machines that are critical for production. The number of which has increased greatly in recent years due to de-bottlenecking programs the inclusion of stand-by machines in regular production processes.

Statutory requirements
All companies that operate process critical machines must set themselves the task of evaluating the risks when operating the equipment. For the operator, a variety of laws and regulations require the assessment and management of risks associated with the operation of machinery, such as the international standard IEC 61511 Functional Safety - Safety Instrumented Systems for the Process Industry Sector, as well as insurance guidelines and internal company process standards.

Functional safety and the SIL classification
When expanding the provision of safety facilities, availability (and/or Functional Safety) is an important criterion. Functional Safety, i.e. the requirement that the monitoring technology is fully operational when required, can be certified by TUV Germany, for example. Based on IEC 61508 (Functional Safety) the system design is checked in detail, and/or its development process is actively accompanied by TUV experts. The result is a certificate which is valid worldwide and documents the safety requirements met by the monitoring system (Safety Integrity Level, in short: SIL). SI-levels range from 0 to 4, with 0 being the lowest level of availability.

Operators see this SIL classification as a key criterion in the evaluation process for safety devices. In connection with this, the SIL level of protective equipment must also at least reach the required level identified by the risk analysis (e.g. HAZOP). As such, for the user, the scope of the certificate is worth a closer look. Have all monitoring parameters that are necessary for effective machine safety been classified? A SI-level awarded for overspeed protection in turbomachinery is useless in ensuring safety in reciprocating machinery. In this case, for example, the continuous monitoring and analysis of vibrations on the crosshead slide and the dynamic piston rod position are important factors. In high pressure compressors in the LDPE industry, for example, which compress ethylene to over 3000 bar, the position measurement of the plunger is a crucial safety parameter for which operators also allocate a higher SI-Level.

For operators, the expansion of machine safety facilities is only acceptable if a guarantee is given that an increased number of unjustified machine shutdowns will not be experienced. In order to avoid costly false alarms, plausibility checks of the measured signals can be carried out for example. One or more checks of the analysis results BEFORE the alarm is issued safeguard against instances where loose cables, faulty (signal) lines or carelessness cause unjustified alarms or emergency shutdowns. These quality characteristics are crucial for the long term, successful application of modern safety equipment.