Addressing Fire and Gas Detection
15 January 2015
Fire and Gas Detection (FGD) is integral to safety on offshore oil and gas platforms, but ageing rigs are facing a variety of FGD system maintenance and upgrade challenges. Andy Tonge, Sales Manager of Hima-Sella, explains the problems and proffers a solution.
In the event of an outbreak of a fire or a gas leak, an FGD system performs two crucial tasks. It alerts personnel / operators and activates automatic response systems; suppression for fire and containment or extraction for gas. FGD is also essential to functional safety, and can be used to trigger an Emergency Shutdown (ESD).
On ageing platforms, however, there is a growing need for more advanced FGD systems.
These need to:
• Be more reliable, in terms of not raising false alarms and erroneously triggering a disruptive and costly ESD;
• Be more dependable, in terms of working when they should;
• Be more responsive, in terms of reacting in near real-time (as delaying an ESD by even a fraction of a second can make a huge difference);
• Provide better support rehearsals and drills, by way of training personnel how to deal with a variety of emergency scenarios;
• Provide better quality of information, so that operators can make ‘informed’ decisions regarding the potential hazards before sending in responders (if at all); and
• Be able to be modified without disrupting production processes and/or removing existing levels of protection.
The above requirements have come about, in most cases, by the increasing problems the operators and owners of ageing platforms are experiencing in just maintaining their FGD systems, let alone making modifications.
Let’s take an example which, although hypothetical, will demonstrate many of the issues facing the operators of real platforms.
Consider a medium-sized, immobile offshore oil rig. It was built during the early 1990s and was fitted with circa 1,500 detection devices to serve, say, 100 plant areas and fire zones across five decks; namely heli, drilling, mezzanine, production and cellar.
These hazardous areas would have been defined using the standards of the time, which will have applied during the design and construction phases of the platform. In the early 1990s, the standards were replaced by the ATEX standard, and the areas would have been re-evaluated under ATEX to allow the use of ATEX-certified equipment.
The Kittiwake platform
The devices serving these areas/zones include sensors for detecting smoke, gas (such as Hydrogen-Sulphide [H2S] or methane[CH4]), heat (e.g. ‘rate of rise’) or flame, based on the presence of either infrared (IR) or ultraviolet (UV) light.
All sensors were state-of-the-art at the time of installation and their Mean Time Between Failure (MTBF) stats looked good on paper.
The majority, if not all, of the FGD devices employed on our platform have a number of binary outputs to indicate their status (e.g. powered & healthy and alarm condition) and are hard-wired back to a number of programmable logic controllers (PLCs). Though please note, many of the FGD systems built in the early 1990s were custom, microprocessor-based embedded systems.
Our hypothetical FGD system also includes a number of matrix panels, of which one is a master containing indicators and reset buttons. There are also a number of data-loggers and printers around the platform.
So that’s our off-shore rig, which incorporated the technologies - and met the regulatory requirements - of the day.
During the early years of operation, the FGD system served us well, though sensor inspection, maintenance and calibration was a chore; and in the event of an alarm, documenting it for audit purposes involved wading through numerous time-stamped printouts.
Many of the triggers were false alarms, typically raised by flame detectors that were being fooled by, for example, sunlight reflecting on rippling water. Also, as time passed, we started to have reduced levels of confidence in the gas detectors; to the extent that we’re now questioning their ‘sensitivity’, after such a long period of service.
Changing devices typically required shutting down part, if not all, of the FGD system; and, for safety, platform operations. Spares were consumed and as time progressed it became increasingly difficult to source replacements, as the OEMs had ceased making the sensors. We now find ourselves facing and considering obsolescence factors.
Compounding matters were changes in regulations and standards. Yes, it was good that ATEX was launched as a more unified standard for equipment in explosive atmospheres; and we willingly complied with the requirement that all new devices installed into an ATEX-rated zone be ATEX-certified. But, by implication, our legacy detectors, although industry accepted at the point of installation, are not certified for use within zones now classified under the ATEX scheme.
For the above reasons, upgrading all FGD devices, on a sensor type-by-type basis, on our ageing platform starts to make sense, particularly as today’s detection devices are at least one generation removed from those installed in the 1990s. For example, today’s flame detectors typically employ triple IR technology and are ‘solar blind’ (and therefore not tricked by sunlight).
However, our 20-plus-year-old FGD system is not particularly accommodating. Its condition and physical architecture are not cut out to get the most from modern detectors.
Figure 1: The FGD System would act as an autonomous local protection system. Several such local systems could be distributed around a site/platform, all of which would be addressable and controllable from a unit
Technology aside, there has been a fundamental shift in ‘where’ FGD decisions are made. As mentioned, the majority of our original sensors have binary outputs. They make the call, locally, on whether or not to raise an alarm. Most modern detectors are more akin to condition monitors, and provide a digital representation of a key condition; for example, temperature or the concentration of a particular gas. They are, in essence, expecting ‘intelligence’ at the heart of the FGD system to make the call on whether or not to raise an alarm. That intelligence can also:
• Provide an indication that maintenance is due;
• Provide an ‘early warning’, by way of acknowledging that conditions are changing (ahead of raising an alarm); and
• Mitigate against false alarms by taking into account the levels being reported by other detectors in the same area/zone.
The Next Stage
A practical solution for the dilemmas we, as the operator/owner of our hypothetical rig, face is to install a new, more robust and future-proof FGD system alongside the legacy one.
This would allow us to migrate the protection for each zone/area, on a sensor type-by-type basis, to the new system. The new system would also accept, as inputs, the same binary signals (from the legacy sensors) seen by the legacy FGD system.
For example, a new FGD system could be crafted using multiple SIL3-rated logic solvers, which are suitable for use in Ex-Zone 2 and are available with up to several hundred I/O points for hardwired connection to detectors, alarms and call points (see Figure 1).
Several such units could be distributed around a platform, such that each area/zone would have its own autonomous local protection system.
Each could then be networked via a robust comms protocol to a Programmable Electronic System (PES).
The PES should also be SIL3 capable, with continuous performance capability so that hardware and software changes can be made without taking the system off-line, making it ideal for safety-critical production processes that can ill-afford downtime.
Under such a system there could be a scalable network of FGD devices; all controllable from a centralised location.
The Bigger Picture
Figure 2: Above, a Supervisory Control And Data Acquisition (SCADA) system is used to capture and report extensive intelligence from the field devices, real-time
Our new FGD system would not only be able to cope with online detector replacement (“old for new”, whilst simultaneously enhancing the functional capabilities) but it would also be able to cope better with significant modifications.
For example, the operational lives of many ageing rigs are being extended in light of field expansions, made possible thanks to advances in subsea drilling tie-back technologies. The acceptance and processing of oil from subsea well heads requires making fundamental changes to the platform, and the FGD system must keep pace with these changes.
Our new FGD would also help transform our platform into an information hub, and make possible the provision of information (via a one-way data-link) to shore-based operators and stakeholders (see Figure 2).
Also, our new FGD system would be able to provide non-stop protection throughout, and play a significant role in, the Functional Safety Management (FSM) of the platform; throughout its entire lifecycle.
Though pitched in the context of a hypothetical rig, the issues/problems discussed in this article are real; and are expected to escalate over the course of a platform’s lifetime. It is therefore not so much a case of should the legacy FGD system be replaced but rather a case of when should it be replaced?
Hima-Sella is currently providing an upgraded fire and gas detection system for the Kittiwake oil and gas platform in the North Sea. The system features a HIMA HIMax PES, which will cater for an estimated 1,300 I/O with scope for expansion.
The fire and gas detection system’s functionality is being designed using SILworX – HIMA’s integrated configuration, programming and diagnostic environment – and the majority of the signal conditioning will be performed directly on HIMax’s I/O modules. The HIMax’s ‘NON STOP’ feature will allow changes to be made whilst the system is online and without interrupting processes on the platform.
The overall project also includes the supply of new flame, smoke and gas detectors.