From a safety and security perspective, oil and gas installations are among the most critical and challenging environments. The speed in which an incident can escalate to adisaster means that operators need to get information quickly to be able to react and minimize damages. Advances in video technology and video content analytics have great potential to assist oil and gas installations in detecting fires, leaks, and security threats, as well as help make operations more efficient.
From a safety and security perspective, oil and gas installations are among the most critical and challenging environments. The speed in which an incident can escalate to adisaster means that operators need to get information quickly to be able to react and minimize damages. Advances in video technology and video content analytics have great potential to assist oil and gas installations in detecting fires, leaks, and security threats, as well as help make operations more efficient.
As oil exploration moves to new regions across the world, so does the need for better monitoring of remote facilities, increased protection against terror threats, and pipeline and leak detection monitoring solutions. Exploration and drilling security systems will be the largest segment in terms of spending and adoption followed by refineries and storage facilities. According to research by Frost and Sullivan, the global oil and gas infrastructure security market was estimated at US$19.6 billion in 2013 and will grow to $24.7 billion by 2021. Including network and cyber security, the combined oil and gas security market will reach $30 billion by 2018.
Any damage to an oil and gas facility has grave consequences. Incidents have implications on employee safety, environmental damage, cleaning costs, damage penalties, equipment replacement, and the list goes on and on. This becomes even more challenging when considering the nature of these installations. There are hundreds of potential breach points for leaks and damages: numerous valves and pumps, meters upon meters of pipes. Even within the facility, environments are not the same. A refinery for example will include various types of oil, transported at different temperatures and pressures, from ambient temperatures and pressure up to 130 bars at 100°C.
In addition, sites are often located in remote and harsh environments. All these make the oil and gas industry one of the most demanding clients for safety and security products. One of the limitations of the oil and gas sector is the stringent regulations regarding equipment. For example, equipment for Class 1 Zone 1 areas — the areas closest to the drill site — needs to be explosion proof, making sure it will not cause an explosion in high risk areas. There are similar limitations on mobile devices since cellular radiation can also cause an explosion.
“If you are looking at a typical oil and gas plant, there will be between 300 to 500 cameras to monitor the perimeter and inner workings of the plant. In addition, there will be thermal cameras to monitor flames and detect leaks and specialty cameras for class and zone areas. In an ocean rig there are 16 to 20 explosion proof cameras and 40 to 50 standard cameras,” said William Moore, Senior Account Manager for Pelco by Schneider Electric.
With so many cameras involved, it is impractical to monitor them all the time; therefore, the cameras are monitored based on an alert rule-engine. In case of an event, whether security or safety related, the proper camera will be triggered and brought to the attention of the operator.
The second line of defense: Video analytics
Typically, video analytics are most appropriate for outdoor and wide-area use, when it is impractical to place sensors (i.e. outdoors or in a warehouse with high ceilings). Common analytics in use are smoke and fire detection, leak detection, gas leak detection, flare-size monitoring, and slip and fall (used for employee safety). In addition, more security-oriented analytics such as perimeter protection and theft detection are also used.
Despite technology advances, video surveillance and video analytics alone cannot replace traditional fire detection systems. Usually both are implemented as standalone systems. Analytics are however beneficial for verification and supplying secondary visual inputs to an alert triggered by other sensors. For example, a combination of a thermal camera and a visible spectrum camera can give a better understanding of a situation. “We use the video analytics only as a second layer of detection and we do not trust it to be the first line of alarm. The reason is due to the day camera sensor's capabilities. Our first line of detection is the thermal vision sensor,” explained Tomer Dadon, CEO of Ex-Sight. “The flame detector camera is normally packed inside the pan tilt video surveillance system. This system can connect directly to a video surveillance control center or a SCADA system.” Flame detection analytics enables the user to continuously assess and monitor target sites in the range of a few kilometers. While scanning the target area, the system identifies when a designated preset reaches a threatening temperature threshold. Once a flame is identified, the camera uses multiple alarm mechanisms, including transmitting the fire's coordinates. The specialty of the system is its ability to distinguish between flame and smoke sources which are part of the industrial environment and those which pose security risks.
“We see a strong trend for coupling video analytics with thermal video surveillance cameras,” said Laurent Assouly, Marketing Manager for Evitech. This combination enables a volume protection instead of line detection. Thermal cameras enable long range detection, such as 600 meters or one kilometer, while color cameras usually cannot see beyond 130 meters at night. Detection over color images would possibly raise more false alarms due to moving lights, flashing lights, etc., and would not reveal many details at the end of the fields of view at nighttime.
Video Analytics for Operations Monitoring
Video analytics can also be used for process monitoring and ensuring the facility is functioning properly. “We monitor pump jacks to detect any problems with their operation. By monitoring the cadence of the pump we can detect if it stops moving and alert the operators immediately,” explained Michael Von-Hauff, CEO of Osprey Informatics. This makes analytics exceptionally useful for remote sites that are not manned permanently. Without this capability, a malfunction can sometimes be detected only after hours, or even days if the pump is located in a remote location, when a human inspector visits the site. Such a solution, therefore, saves lost production time and helps turn video into actionable intelligence. Video monitoring can also be used to audit employee behavior during alarms and other safety events — it allows the relevant officers to check if all employees are following the relevant procedures.
.jpg)
Connectivity and systems integration
A growing market trend is connectivity and systems integration. “There are thousands of land rigs, gated and sensed-in, the clients are asking for the possibility to integrate cameras and access control,” explained Moore. “For example, we implemented such a solution in western Texas. A FOB key card swipe at the gate pops up the picture of the truck to the administrator.” The system is also integrated with LPR cameras, cell phones, and iPads and is used for site management and monitoring site visits. “Another thing the customer is looking for is a common platform such as ONVIF,” added Thomas Soderlund, Business Development Manager for EMEA at Bosch Security Systems. “With the old analog systems it was easy to pick any camera and plug it in and it would work due to the 1 volt peak-to-peak signals. With IP devices it is not always that easy since a new firmware or change of model mean additional programming of the main system is needed to be able to provide functionality.”
SCADA Integration
There is a benefit in connecting different systems to the SCADA system. The benefit of an integrated system is to have one common interface for alarm handling and monitoring, compared to moving between multiple systems in order to then build up the picture of the event. In addition this integration can help in halting delivery and limiting a leak's consequences. However, most platforms need to use SDKs to integrate with the SCADA systems, which is often a costly and complex solution.
Avoiding false alarms
There are several options to reduce false alarms. One is to install multiple different sensors for verification purposes. A second system for avoiding false alarms is to calibrate the system against an existing database of detections and false alarms. Tests can be performed using high pressure water cleaning tools, which deliver varying temperatures and pressures. These simulate various sorts of leaks in the different pipes (themselves at different temperatures). These tests can be used as a benchmark for the analytics and then need to be carried out at regular intervals to ensure the continuity of service of the solution. Artificial intelligence analytics check alerts and reduce false alarms by differentiating between real objects and other objects that might cause false alarms such as changes in lighting (e.g., a cloud passing over the sun or a light being switched off or on, long evening shadows, and bright lights pointed at the camera).
Not just video: Other senses take part
Dedicated sensors and video analytics are not the only features used for leak detection — there is also an audio-based detection method. “In a southern Louisiana installation, the cameras have the audio option turned on and in the case of an alert the operator can listen to the pipe to hear if there is a high pitch whistle coming out, indicating gas is leaking from the pipe,” described Moore. “This is already an automatic feature in the non-class and zone cameras. For class and zone areas it is more complicated due to technical reasons. The external microphone has to meet the safety criteria and not cause explosions.” iOmniscient combines with its video analytics smell sensors that can detect gas leaks and audio analysis to identify sounds. “When combined with the analytics from video these systems can provide an enhanced understanding of what is happening in an environment,” explained Dr. Rustom Kanga, CEO of iOmniscient.
Automated response
Information integration from various sensors into one platform is usually described as PSIM (physical security information management). Kanga described a solution where the information is not only pulled together but is also responsible for automated response. “The system, without human intervention, will find the nearest appropriate first responder and provide him with detailed information,” said Kanga. Relevant information can be pushed to the user's cell phone, turning the user's smartphone into a mobile control room. The user can use the phone to perform all the operations that he could have performed in a control room. However, the quality of the entire system depends on the quality of the core analytics. Automated responses reduce reaction time and as such are beneficial for both safety and security but also for operational efficiency — making sure malfunctions are treated fast.
Future trends
As oil and gas exploration continuous to expand further to more difficult environments so will the demand for video surveillance systems. The future calls for durable and reliable systems, able to withstand harsh field conditions and connect with other systems that will be able to give operators more than just security and also optimize operations. The future will no doubt involve more and more integrated systems, making full use of technology for both safety and security.