Gases can be, flammable and therefore lead to fire and explosion, they can be asphyxiants and they can cause other occupational health related problems (asthma/lung sensitisation, skin irritation and soft tissue damage).
Gases are used in many industrial processes, but gases may also be produced as an unwanted by-product in some activities (oil production, refining and chemical manufacturing). In all cases it is very important that these processes are managed professionally to eliminate harmful gas leaks. Where it is not practicable to eliminate the risk of a gas leak then the organisation must take steps to effectively mitigate the risks from these escaped gases. In this article we will focus on what steps that can be taken to minimise these risks, particularly around the use of gas detectors.
Harmful gases and their characteristics
Some of the most common gases produced are carbon monoxide, chlorine, hydrogen fluoride, hydrogen sulphide, ammonia and flammable substances (natural gas). Typically, these gases are found in the production of iron and steel, chemicals and petrochemicals, oil refining, natural gas sourcing, refrigeration and water treatment.
It is important to be able to detect some gases, such as those which are toxic or flammable, more urgently than others. There are a number of factors that affect the need for speed of detection, namely:
- The type of gas
- The temperature, pressure, quantity and toxicity
- Its proximity to employees and public
- The effectiveness of counteractive measures taken against it
- The speed and effectiveness of medical intervention
Certain environments require much more finely tuned methods of detection and protection, as a leak would have much greater ramifications, for example, petrochemical plants carry a high risk of gas leakage, which can lead to explosion or fire that not only damages equipment but puts lives at serious risk. Meanwhile, nuclear power stations use gases, such as CO2, as coolants to transfer heat to generate steam and to prevent reactors overheating. Any loss of CO2 could reduce the efficiency of heat removal and cause overheating. In both of these situations, gas detectors are used not only to keep the process going, but to keep people safe by detecting any gas releases.
The coke produced in making iron and steel gives off carbon monoxide. Process fluids or solvents produced in the chemical and petrochemical industries can be extremely flammable themselves, as can the actual chemicals used in the processes like chlorine or benzene.
Oil refining hazards include hydrogen sulphide, hydrofluoric acid and other flammability issues, while the oil and gas drilling industry is also extremely hazardous. Disasters in recent years expose the need for prior warning systems that pick up even the most minute escape of gas.
“disasters in recent years expose the need for prior warning systems”
Some gases are used for fire suppression and inert blanketing, like nitrogen and CO2. Some oil storage tanks have a blanket of nitrogen sitting on top of the fluids to minimise the risk of a fire starting. Nitrogen and CO2 are asphyxiants and must be handled with great care. Many sensitive electronic systems are protected by N2 and CO2 suppression systems. The gases can extinguish fires by displacing oxygen with non-flammable gases. The gases do not damage the electronic systems the way that water or powder might. I once investigated an incident in a record keeping organisation where the CO2 drench system was activated by mistake. There were people in the storage rooms that were being flooded with gas. In this case everyone was able to escape safely, but this is a considerable risk. As CO2 is much heavier than ambient air, we had to check very carefully for potential “pockets” of CO2 which had not dispersed naturally. Forced ventilation was brought in to ensure all areas were made safe.
Any system which uses, generates or contains a gas under pressure can develop leaks. But the significance a leak has on the business’s ability to maintain normal activities will depend on the type of gas being used, its pressure and temperature, any products or reactions generated by the process, where the leak is located, and the quantity of gas that could be released.
The location and sensitivity of a detector will depend on where the gas is likely to escape from and when. Although leaks are the result of a host of issues, the most frequent causes are human error, corrosion (internal and external), wear or faulty equipment, poor maintenance, or accelerated chemical reactions that increase pressure.
The risk management process for selecting gas detection often involves gas dispersion modelling, using likely locations where the substance could escape as a template. Varying densities, volumes, and temperatures of the gas are tested, along with differing weather conditions, to find out how the gas cloud is likely to form and disperse in the event of a leak. A change in atmospheric conditions can have profound effects as highlighted in a recent gas explosion I investigated.
A well workover team were replacing downhole monitoring systems and as part of the process were producing quantities of natural gas back to the surface. This work was being conducted in a large flat plain where there was a constant breeze blowing. It is likely that the crew was not aware that this breeze was dissipating any natural gas that was escaping and preventing it from ever reaching a lower flammable/explosive limit. However, one summer evening the breeze stopped, and the air was very still, now escaping natural gas was able to pool and eventually find a source of ignition. Of course, you cannot and should not use the weather as a protection factor, but the significance here was the major impact that occurred because of a small change in weather conditions. So, the impact of the weather should be considered as part of the risk management process.
‘Hazops’ – hazard and operability studies – may also be used to examine a company’s equipment and its operation, to determine the possible points at which gas may be released.
“wear and tear are often related to poor inspection and maintenance”
Wear and tear are often related to poor inspection processes and inadequate maintenance. Failed pipe work, poor piping joints, leaking pumps or valve seals, and vents and drains can become the culprits if they are not kept in good condition. The critical issue here is to have a robust preventative maintenance scheme in place. This should ensure that all relevant equipment is inspected, tested and, if necessary, certified on a periodic basis (to be determined through a combination of HAZOP and Manufacturer’s instructions). Any defects or irregularities should be reported in immediately so that further investigation and restoration may take place. This is particularly critical with the monitoring of internal and external corrosion.
As an example, with either buried or surface pipelines the internal condition can be inspected using intelligent pigs, for the external corrosion it can be more challenging. It is not practicable to dig up a 20 km pipeline, so decisions have to be made on digging test pits for routine inspection. Surface pipelines may on the face of it seem easier to inspect externally, but again, they can be routinely coated in insulation. It is not practicable to remove 20 km of insulation to inspect the entire line, again decisions have to be made about the number of inspection points to be made. Once the inspections are complete, it is likely that some forms of corrosion will be detected and the critical element there is to ensure that these points have increased scrutiny to monitor the speed of corrosion and of course the business must decide when the levels reach a point where preventative repairs have to be made.
Some types of gas can also cause issues in enclosed spaces and workers can find themselves in a situation where the atmosphere can’t support breathing, so gas detection will be a part of the risk reduction measures.
Workers must be fully aware of all the gases used, produced, or discharged in the areas where they work. Working in an enclosed/confined environment is a specific hazard and must be very strictly controlled, particularly if gases may be present.
For example, even a task as seemingly harmless as setting concrete gives off carbon dioxide, which can act as an asphyxiant in the right conditions. Organic food or waste can remove oxygen, producing hydrogen and methane as it decomposes, and process plants like oil refineries can release toxic gases such a hydrogen fluoride and flammable hydrocarbons.
Gas detection methods
Most systems can be broken down into two categories – fixed and portable.
Portable electronic personal gas detection monitors allow an individual to work in a potentially hazardous area, by sounding a warning alarm when the gas rises to an unsafe level and action must be taken. These safe working levels are often defined in national legislation, international trade associations and manufacturers safety data information. Be careful to check the valid levels for the specific country or region you are working in as there may be some small variations in levels from country to country.
Portable detection is used to help workers undertaking tasks such as maintenance, where employees might disturb or release gas that wouldn’t normally be detectable in that environment. One example here would be people cleaning out degreasing tanks (trichloroethylene). The gas is often absorbed in the greasy residue on the bottom of the tank, however, when a worker starts to disturb the residue to collect it to remove it, the gas vapours can be given off and as it is heavier than air, it could collect in the breathing zone of the worker.
A personal gas detector could be used to detect small quantities of gas vapour and action can be taken before it becomes hazardous to the worker. Portable devices can also be used for sampling, where workers need to measure the quality of the atmosphere before entering an area. But again, as stated above, the sampling atmosphere may be fine before the person goes to work, but it could soon become hazardous once work begins, so it is important to understand when sampling gas testing is appropriate and where continual sampling is required. This is very much part of the risk management process.
Fixed systems are preferred in process safety and static systems where the layout is more constant. In refineries for example where the risks associated with gas leaks are known and constant. Fixed systems tend to protect and entire area or process and provide a warning to many people, not just the individual wearing a portable monitor. The position/placement and type of fixed monitor used will depend on many factors, for example if the gas is heavier than air then the monitor would need to be below the leak point.
“gas detectors can cover a number of typically predictable sources of leak”
If they are used outside, they may need to be very close to the potential leak point to ensure that the ambient weather conditions do not dissipate the gas before detection. Perhaps the most common types of fixed gas detection devices are smoke alarms and carbon monoxide detectors. Smoke alarms are generally placed on the ceiling in proximity to kitchen devices as the smoke will generally rise quickly due to the heat generated. Carbon monoxide monitors are more likely placed at “head height”, the breathing zone of the people potentially affected.
Another advantage of fixed gas detection is that the gas detector will monitor an area constantly, so that if an operator is not present full time, the gas leak will still be detected, and the appropriate emergency response can be undertaken.
Gas detection outside requires many detectors to provide blanket coverage across a wide area around a plant or a release point. This is because without the restriction of building walls and structures, the gas can escape in any direction depending on wind and weather conditions at the time. In these circumstances, gas detectors can cover a number of typically predictable sources of leak, such as corrosion points, sampling points which might get left open by human error, and road, rail or marine loading and unloading locations.
The sensitivity of the gas detection device must reflect a substance’s lower explosive limit or toxic levels, to give early warning of any danger.
Planning to manage gas leaks
There are three principal considerations to manage:
Process control – to ensure that the presence, pressure, reactions and concentration of any harmful gases are controlled so that the gases do not escape into the environment.
Loss prevention – implementation of robust design, maintenance, inspection, and corrosion prevention to ensure that gases are contained.
Loss preparedness – gas detection and process monitoring to identify the loss of gas if it occurs and suitable emergency arrangements to protect those who may be affected.
If gas does leak, then early detection is key. Automated systems have reached a level of maturity now where if gas is detected then entire systems can be shut off automatically to prevent any further escape. Of course, you do not want to shut down a plant because of a defective monitor/detector, so systems will often have a voting system where three or more independent detectors and allow the system to identify faulty sensors and thus make a decision to “warn” as opposed to shut down. If one detector activates and two do not, the system may not go into automatic shutdown but rather prompt the operator to investigate the tripped alarm. Of course, this alarm must always be treated as genuine until the operator can confirm that it was spurious. If two detectors trip, then the voting system would vote two to one to shut down.
Some systems will still prompt an operator action to shut down a process or part of a process. This is most common in systems when hazards can be created by shutting down (e.g. cooling systems) you don’t want to minimise one risk but create a worse one. In many complex plants, systems (parts of systems) can be shut down centrally (control room) or locally on site.
Basic training for workers in hazardous industries takes two forms – an operational response in the event of an alarm, and an emergency response to limit the spread, or the effect of gas that has already escaped.
Various training programmes exist for the design, installation and operation of detectors. Employees also get general training on how to work in a hazardous environment, and what to do if there is a leak to avoid asphyxiation, injuries from explosion of fire, or skin damage due to toxicity.
“training to prevent disaster involves regular emergency response practise”
Anyone working in confined spaces will also receive specific training in detection and assessment of the gas types present. Testers working in these conditions will obviously be trained in the specific detection equipment they are using.
When workers arrive on a potentially hazardous site, the detailed safety induction training will highlight how to behave in an emergency, the sounds of the different alarms, what they mean and how to respond. Safe routes and mustering points will be identified. In the case of a gas like H2S, personnel will be instructed to don emergency escape respirators and make their way upwind to a place of safety.
Plant with an H2S risk will usually have a hi-vis windsock to help people to identify the safe upwind direction. An escape respirator will normally have a short duration usage as it is only intended to provide enough time to reach a place of safety. It should not be used to affect a recovery or rescue. Specially trained emergency responders will normally be provided with full face mask self-contained breathing apparatus. The personnel involved in continuous gas detection may well be members of this specialised response team.
If fire and explosion is a significant hazard, then additional PPE would be required, for example, anti-flash suits and intrinsically safe equipment.
Part of the training process to prevent and deal with potential disaster involves regular emergency response practise – this includes testing alarms and carrying out practise evacuations.