IBP2154_12 SYSTEMS AND STRATEGIES FOR SUBSEA LEAK DETECTION Torleiv Stien Grimsrud1, Kristin Nergaard Berg2, Torild

Nissen-Lie3

Copyright 2012, Brazilian Petroleum, Gas and Biofuels Institute - IBP

This Technical Paper was prepared for presentation at the Rio Oi & Gas Expo and Conference 2012, held between September, 1720, 2012, in Rio de Janeiro. This Technical Paper was selected for presentation by the Technical Committee of the event according to the information contained in the final paper submitted by the author(s). The organizers are not supposed to translate or correct the submitted papers. The material as it is presented, does not necessarily represent Brazilian Petroleum, Gas and Biofuels Institute opinion, or that of its Members or Representatives. Authors consent to the publication of this Technical Paper in the Rio Oil & Gas Expo and Conference 2012 Proceedings.

Abstract
Advancements in technology and energy scarcity have pushed oil exploration into new territories and resulted in ultradeep extraction and developments close to vulnerable resources. Subsea technologies are becoming increasingly complex and the margins for errors have diminished. Stakeholder pressure has increased, comprehensive regulations addressing accidents are established and operators are held more accountable than ever. Spills can cause significant environmental, financial and reputational damage. All these developments present the oil sector with new challenges with regards to leak detection. Conventional techniques might not be sufficient and a structured approach to the issue is required. This paper describes commonly used leak detection technologies and divides them between topside and subsea techniques. Limitations and flaws are included as it is vital to build complete detection systems consisting of technologies that complete and complement each other. When designing a system it is crucial to get a complete overview of relevant requirements and criteria. This includes authority regulations, company requirements, environmental concerns, facility limitations and field specific factors. Quantitative risk assessments can map areas and elements that are exposed to risks. In order to find appropriate technologies that meet the requirements, it is beneficial to perform an analysis of the best available techniques (BAT) for the various areas of the installations. It is likely that a combination of techniques is necessary to fulfill prerequisites and it is recommended that area sensors are supplemented with point sensors that cover the critical elements of the installation. Leak detection also comprises organizational measures and it is important to integrate leak detection into facility management systems. The final part of the paper describes implementation and operating philosophy. Subsea leak detection is getting progressively more common and it is vital to record and document the field performance for the technologies in order to develop and improve the techniques.

1. Introduction
Within the next decades, there will be considerable changes in the subsea sector. Investment projections predict skyrocketing activities within subsea engineering and new possibilities have ascended from the evolving technology. Together with these developments, new challenges have arisen. Offshore leak detection technology for subsea installations is still problematical, but significant developments are taking place within this field. Recent accidents have also demonstrated that leaks can have considerable consequences for the oil companies. Certain operators licenses to operate in the affected countries have been jeopardized due to leaks. This paper aims to demonstrate the importance of suitable leak detection. It will present a brief overview of the most common offshore leak detection technologies and describe a structured process for designing complete offshore leak detection systems. It will also present how leak detection should be implemented and included in the overall operational plans for the installations. The focus of this paper is on continuous leak detection techniques and is therefore not including infrequent examinations or recurring inspections.

______________________________ 1 MSc, MA Hons, Consultant DNV 2 MSc, Senior Engineer DNV 3 MSc, Head of Section - DNV

Rio Oil & Gas Expo and Conference 2012

2. Background
Recent events have demonstrated the importance of adequate leak detection, but there are particularly three trends that are amplifying the significance of detection systems: Subsea technologies are becoming increasingly complex. Templates and manifolds are more advanced than ever, subsea production has evolved and pipeline grids are covering large distances. This also presents the sector with new challenges with regards to potential spills. Visual observation is often unviable and advanced technology is needed. Energy scarcity has caused increased exploration activities and new fields have been discovered. Several of the new fields are located in areas with hostile weather conditions, technical difficulties, at extreme depths and with considerable distance to available infrastructure. It includes marginal fields and sections that have previously been economically or technically unrealizable. Exploration areas also incorporate the Arctic and other vulnerable regions where the precautionary principle can be applied. Increased interconnectedness and communication technology have caused more pressure from external stakeholders. Transparency has grown together with the power of the public, and safety, accountability and corporate responsibility are essential for companies in order to preserve their licenses to operate. Spills can cause both significant environmental, financial and reputational damage.

The result is a situation where technology is complex, environmental risk is increasing and stakeholder pressure is high.

3. Policies, Regulations and Requirements

The trends have resulted in strict regulations regarding leak detection in the Norwegian offshore sector. According to the discharge permits issued, leaks should be detected within 12 hours on unmanned installations. Manned installations usually face a three hour detection criteria, but facilities in vulnerable areas will have to detect spills within one hour. The regulations state that the remote measurement of acute pollution imply systems which, regardless of visibility, light and weather conditions, can detect and map the position, area, quantity and properties of the pollution. Also, the status of safety functions shall be available in the central control room. The Integrated Pollution Prevention and Control directive of the European Union also set certain requirements for the concerned operators. Industrial installations are required to ensure accident prevention and damage limitation through the use of all appropriate pollution-prevention measures, i.e. the best available techniques (BAT). While the regulations regarding leak detection varies between countries, events in the last few years have demonstrated how important it is to have a functional and reliable system in place in order to identify leaks and act on spills as early as possible. The technologies can also be used to document the event and hence create a more accurate picture of the accident. Considering the increased consequences of oil spills and the high environmental risk, leak detection is a field where it could be beneficial for the operators to go beyond compliance.

4. Topside Leak Detection

The following chapters describe some of the most commonly used techniques and certain emerging methods are also included. The various techniques are divided between topside and subsea technologies, and the use of mass balance, which is detection within the installations, is included in the latter. 4.1. Visual Observation Oil at the sea surface creates a distinctive sheen that can be visually observed. Appearance of the sheen (silver, rainbow, brown/black) gives an indication of the thickness of the spill. On manned installations spills could be detected by visual observation and even by odors, either from installation or from a standby vessel. Visibility, light and weather conditions will be factors that can limit visual observation. Due to uncertainties concerning weather conditions and the perceptiveness of the observer, it is difficult to quantify possible leaks sizes or time before a leak is detected though visual observation. It is unclear if gas leaks rising from the water column to the sea surface would be visually detected. Depending on the leak flow rate, boiling water (gas bubbles) would be observed. However, the size of the bubbling area and the distance at which it could be observed is very uncertain. 2

Rio Oil & Gas Expo and Conference 2012 4.2. Infrared Oil absorbs solar radiation and emits some part of it as the thermal energy mainly in the thermal infrared (IR) region (8-14 m). Oil has a different spectral signature than water, and depending on the film thickness, oil appears hot (thick oil), cold (thin oil) or is not detectable on infrared (very thin oil). The detection works independent of light conditions and is commonly used for estimating thickness and quantities of oil spills. Very little information is available about the applicability of IR detection for offshore gas leaks. IR is used in the industry for the detection of gas leaks from process systems, but it is unclear if the technology would offer benefits for the detection of gas rising from the water column to the surface. Infrared detection is available on a wide range of systems, including handheld cameras and fixed installations. It is important to distinguish between un-cooled and cooled systems. While un-cooled systems are far less sensitive than advanced cooled IR systems, they are also significantly cheaper, can be used instantly and do not have coolers requiring regular replacing. 4.3. Oil Spill Detection Radar An oil detection radar system is based on a normal marine radar and a special oil spill software that detects oil on the sea surface. This detection method requires wind in order to generate capillary waves. Too much wind will get the oil mixed into the water and it will be harder to detect the spill. Various actors are operating with different specifications and performance possibilities of radars. There are several considerations affecting the performance of a certain radar, such as the volumes and properties of the relevant oil spill, distance, waves and height of the antenna. The range of a radar is typically 6-12 km from an installation (height of antenna: 50+ m) and 4 km from a vessel. Some systems are expected to cover up to 15 km in ideal conditions. A radar system on a facility will function in wind speeds between approximately 2 m/s and up to 10-15 m/s, and location accuracy is better than 250 m. The radars operate independently of light, but there are a number of phenomena that can create interferences and set off false alarms. This includes calm patches with little wind, fresh water slicks and sea weed. 4.4. Satellites The satellite detection system used in Norway is using synthetic aperture radar (SAR) technology and is able to cover large areas. Satellite surveillance is not interfered by darkness, fog, rain or snow, but is depending on the resolution and the area of the oil slick. They are able to detect oil on the water at wind speeds ranging between approximately 2 and 10-15 m/s. The orbiting intervals and the number of satellites are limiting the picture frequency and it is therefore not providing continuous coverage that can comply with the detection requirements on the Norwegian Continental Shelf. However, satellites can provide operators with valuable information and are beneficial for detecting leaks from pipelines.

5 Subsea Leak Detection Techniques

5.1. Mass Balance DNV (2010) provides a description of mass balance systems. Such detection methods are based on monitoring the pressure drop between pressure sensors or flow meters installed in the subsea production system or export pipelines. The number of sensors required is evaluated in a pre-study by the mass balance supplier. Comparing flow in and out of the system will reveal significant loss of flow. The limitation for a mass balance system is that it will not detect small and sometimes even medium leaks. This will provide cover for leaks that are greater than a few per cent of overall flow. Depending on the flow mix, system in place and on the facilities, the detection threshold can be 5% of the total flow rate (DNV, 2010). Even if individual meters typically achieve a better accuracy than 5% of the total flow, the overall system relies on a number of meters, and on a pipeline model. The pipeline model allows taking into account changes in the fluids characteristics, e.g. pressure and temperature, as well as multiple input and output in a system. Therefore, the overall mass balance system accuracy is lower than the accuracy of each of its individual components. 5.2. Capacitance Capacitive sensors measure change in the dielectric constant of the medium surrounding the sensor. The capacitor is formed by two concentric, insulated capacitor plates in the same plane, the one being a disc and the other being a surrounding annulus. The sensors capacitance is directly proportional to the dielectric constant of the medium between the capacitor plates. The dielectric constants of seawater and hydrocarbons are very different. If the sensor gets in direct contact with hydrocarbons, this will show as a change in the measured capacitance. The capacitance method is a point sensing method. Positioning the leakage relative to the sensor will not be possible. The sensitivity compared to 3

Rio Oil & Gas Expo and Conference 2012 size of leak will be dependent on distance to the leak and drift of the leaking medium. When leaking medium comes in contact with the sensor, the sensitivity is high. A limitation to this product is that direct contact with the leaking medium is required. Seawater currents or buoyancy effects may lead the leaking medium away from the sensor. This can be solved by installing a collector for hydrocarbons over the monitored structure. Where template protective covers are used as protection to fishery, these covers can be modified to serve as a hydrocarbon collector. Laboratory tests have shown challenges with collecting oil, since the oil flow does not stop in the collector. This effect may be less significant in a realistic subsea environment, due to the much larger size of a template roof collector compared to the experimental collector. Additionally, live crude oil always contains some natural gas which may be easier to collect by the template roof. The product maturity of capacitance sensors is high. These sensors have been on the market since the 1990s. Operators have experienced some false alarms from these sensors. However, by number of installed sensors, this type is the most common. 5.3. Passive Acoustic These sensors contain hydrophones (under water microphones) picking up the pressure wave, or sound, generated by a rupture or leak and transmitted through a structure or water. As long as there is a sufficiently strong pressure wave, passive acoustic sensors are not dependent on the chemical compound of the leaking medium. Passive acoustic detectors come in variants designed for spatial coverage as well as variants for monitoring of specific critical components. Positioning is possible by using more than two sensors for spatial coverage. Arrival time of a sound at each sensor can be used to locate the origin of the sound. These sensors are little affected by seawater currents and turbidity. Passive acoustic sensors are available on the market and have been commercially delivered. A limitation to this technology is that the sound from small leaks might not reach the hydrophones. Background noise may disturb the measurements and shadowing of the acoustic waves may be a problem. Considerable sound engineering and calibration is needed and a sufficient pressure drop over the leak path is a requirement for detection. 5.4. Active Acoustic Active acoustic sensors are sonar detectors. The detector emits pulses of sound that are reflected by boundaries between different media, i.e. the boundaries of impedance change. This material characteristic depends on sound velocity, density, salinity and the temperature of the medium. Fluids of different density will have different acoustic impedance. This means that as the sound pulse travels through water and hits a bubble of gas or droplet of oil, sound will be reflected back. This technology does not depend on the leaking medium being of a specific composition; however, the acoustic impedance must be different to that of water. Active acoustic methods give area coverage and leak positioning is possible. This technology has high sensitivity for gas, due to the high impedance contrast to water. Larger droplets or plumes of a leaking medium will give a stronger backscattered acoustic signal and are easier to detect. A limitation to the active acoustic method is that it can be sensitive to shadowing of the acoustic signal by the subsea structures. This may, however, be solved by using more than one detector. Also, some active acoustic detectors currently generate a lot of data. Suppliers are currently working on new solutions that will make subsea interfacing easier and allow more efficient data transfer. Experience has shown that the performance will depend on water depth, as gas bubble size will change with water depth. Active acoustic sensors have been commercially delivered for use on ROV and have been used to find leaks in the North Sea. Solutions for permanent monitoring are under development. 5.5. Optical Cameras Optical camera methods are based on a video camera for surveillance of the subsea system. This technology provides spatial coverage and determining the direction from the camera to the leak can be possible. The capabilities of such methods is typically to record and send 3 -30 min of video and 1 -10 still pictures per hour. Cameras for subsea leak detection are sensitive to water turbidity. Another limitation is the need for contrast background to detect oil (the camera must be directed towards the yellow structure). It has been shown in laboratory testing that cameras need additional light for detection beyond 1 m. The limitation with extra light is 3-4 meters. Marine growth may also be a problem, but this can be solved by regular maintenance. The optical camera technology is in use with ROVs and pilots have been installed subsea. 5.6. Bio Sensor Methods Bio sensors utilize the response of organisms to pollution in the surroundings. Suitable organisms are placed on the structure to be monitored. One example of an organism used as sensors are mussels. Sensors register the heart rhythm and the degree and frequency of opening and closing of the clam. This is a point sensing method. Positioning the leakage relative to the sensor will not be possible, but area coverage may be achieved by using several sensors. The sensitivity compared to size of leak will be dependent on distance to the leak 4

Rio Oil & Gas Expo and Conference 2012 and drift of the leaking medium. Direct contact with the leaking medium is required and seawater currents may lead the leaking medium away from the sensor. 5.7. Fiber Optic Methods Fiber optic methods are used for locating and measuring mechanical disturbances at acoustic frequencies along a continuous optical fiber. Disturbances can be caused by e.g. vibrations, seismic waves or acoustic signals from for example leaking gas or liquids. Simultaneous disturbances may be detected and positioned to approximately one meter accuracy along the fiber. By performing comparison with a data library the likely cause of the disturbance can be proposed. There is a trade-off between spatial resolution along the sensing fiber and detection sensitivity. For example, if it is not necessary to isolate the detection of vibrations between every adjacent meter of fiber, but ten meters would be acceptable, then the detection sensitivity can be increased by roughly a factor of ten. A benefit with fiber optic methods is that no power or electronics is required along the length of the cable and it is immune to electrical interference. As far as we know, this technology has not yet been taken beyond the conceptual stage for monitoring of subsea structures, but the concept has been tested for pipelines onshore. 5.8. Fluorescent Methods Fluorescent detectors use a light source of a certain wavelength for exciting molecules in the target material to a higher energy level. The molecules then relax to a lower state and light is emitted at a different wavelength which can be picked up by a detector. To use fluorescent methods, the medium to be detected must naturally fluoresce or a fluorescent marker must be added into the fluid. This is the reason why this technology has traditionally been adopted for subsea inspection and pressure testing. However, many hydraulic fluids have fluorescent markers added as standard. For hydrocarbon leak detection, crude oil has significant natural fluorescence. Fluorescent technology has been proven for use with ROVs and is currently under development for permanent installation on subsea structures. However at the time of writing there is no information about prototype installations available. Fluorescence based detectors can potentially differentiate between hydraulic fluid and oil leaks due to the different fluorescence spectrum of these fluids and also provide an indication of the leak size from the relative signal intensity. These detectors can be point sensors or can provide coverage over 3-5 meters line of sight. As with optical cameras marine growth over optics could potentially be an issue but may be solved with maintenance and optimal choice of surface coatings. 5.9. Methane Sniffer Methods Two measurement principles exist on the market for measuring methane dissolved in water. Both sensor types are based on dissolved methane diffusing over a membrane and into a sensor chamber. Methane sniffers are point sensors. Positioning the leakage relative to the sensor will not be possible. The sensitivity of these sensor types compared to the size of a leak will be dependent on the distance to the leak and the drift of the leaking medium. Both sensor types can detect very small concentrations of dissolved gas in water. Limitations to these technologies are that quantification of leaks is difficult. Also, identifying a leak is dependent on diffusion towards the sensor and seawater currents may lead the leaking medium away from the sensor. 5.10. Developing Methods There are also a number of methods that are on the developing stage, including - Semi-conductor sensors are relying on dissolved methane changing the resistance of an internal component in the sensor chamber. This generates an electrical signal from the detector. Ongoing developments are aiming at achieving long term stability of 5-10 years. - Optical non-dispersive infrared spectrometry (NDIR) is determining the methane concentration by measuring the absorption degree of infrared light at a certain wavelength. When the infrared light is hitting the detector in the sensor chamber, the intensity of the light will measure the methane concentration. While sensors lasting 3+ years are available, the goal is to achieve longer durations.

6. Design of Subsea Leak Detection Systems

While the performances of the various techniques are important for the detection capabilities, it is also vital to take into consideration how suited the technology is for the possible spill scenarios (including field design and spill fluids), how a complete system can be created with complementing techniques and how well the system is complying with the various requirements. 5

Rio Oil & Gas Expo and Conference 2012 The leak detection performances and needs for several fields in the Norwegian sector have been evaluated and the analyses have mainly been for new installations. Experience shows that the cost are lower and opportunities higher if the operators are considering the implementation of the leak detection system in the early planning stage of the installation/field. Retrofitting detection systems is costly and might limit the selection of applicable and available technologies. It is also important to ensure that the data transfer capabilities in the umbilical are sufficient for the usage of subsea leak detection technology. 6.1. System Performance Requirements A methodology for choosing leak detection systems is partly based on the decision process described in NORSOK Standard Environmental Care (Standards Norway, 2005) and has been further developed through various recent projects. Figure 1 indicates the original process that the methodology has been founded on. In the first phase, the initial requirements are mapped through conducting an environmental risk analysis in order to regard potential consequences of spills. This phase also includes mapping the relevant regulations, the legal framework and company requirements, in addition to considering the current practices and guidelines in the sector. It is also important to get a thorough understanding of the planned installations and the limitations and opportunities that exists on the field. Quantitative risk assessments are beneficial and can be used to map areas or zones that are more exposed to risk than others. For example are templates, manifolds and valves significantly more likely to leak than a trenched pipeline made of corrosion resistant alloys. Requirements should also be based on field specific conditions that can affect the performance of the system, including weather conditions, process pressure and the relevant fluids. Other factors include interfaces to the control system and the integration into the overall operational management philosophy of the subsea equipment.

Figure 1. An example of a decision process (Standards Norway, 2005) 6.2. Best Available Techniques (BAT) The relevant requirements provide criteria that can be used when the best available techniques are assessed. For this part of an analysis, it is beneficial to divide the installation into various zones. If possible, this should be linked to the quantitative risk assessment in order to map the exposed zones and leak points. Areas that are associated with higher risk should preferably be equipped with point sensors and there should be area sensors or detection techniques that can provide coverage for all the installations. In this phase of the evaluation, it is important to compare the various techniques with the criteria and also take into consideration the strengths and weaknesses of the technologies. The BAT analysis should provide recommendations regarding available technologies and preferred solutions for the various zones. 6

Rio Oil & Gas Expo and Conference 2012 If possible, it should also include expected detection times and sensitivity. DNV (2010) also suggest that all commercially available leak detection technologies should have a specification sheet that can enable operators to select the right sensor types for their system. The sheet should be specifying parameters like: - Bandwidth and power requirements - Mechanical, electrical and communication interfaces - Requirements to location and environment - Sensitivity to noise or other effects from components - Design pressure and temperature - Test conditions - Sensitivity and drift - Ability to determine location of leak 6.3. Combining Technologies and Evaluation of Performance When analyzing the strengths, weaknesses, opportunities and limitations of the different technologies and comparing it with the authorities and companys requirements, it is possible to identify gaps. Both topside detection systems and subsea equipment should be evaluated and it is imperative to ensure that the employed technologies are completing and complementing each other. Often a combination of technologies and systems need to be utilized in order to cover the prerequisites. It is beneficial to have technology able to cover detection over an area and point sensors installed over critical leak points. By using technologies that also work independently, it is possible to confirm leaks before acting and it may provide the operators with more confidence in the overall leak detection system. The limitations of the complete system can also provide the crew operating the facilities with valuable information and demonstrate under which conditions special attention is required. Figure 2 indicates how a combination of techniques is utilized to provide leak detection on the Yme installations in the Norwegian part of the North Sea. The complete system is using mass balance to detect larger leaks from the pipelines and components further away from the platform. It is also incorporating area sensors and using closed circuit television (CCTV) to cover exposed areas.

Figure 2. Illustration showing the leak detection coverage at the Yme installations 7

Rio Oil & Gas Expo and Conference 2012

7. Further Optimizing at System Level, Operating Philosophy and Maintenance

After the most favorable combination of technologies has been identified, it is important to optimize further at the system-level. A satisfactory, complete detection system also comprises organizational measures, predetermined procedures and preparedness plans, and it is important to integrate leak detection into facility management systems. This means that the technologies and methods used are integrated in procedures and that it is defined how data will be used. Documented procedures concerning operation of the equipment, alarm actions and maintenance should be produced. It is also essential that the feedback from the leak detection equipment is evaluated regularly in order to detect drifting of parameters as this might indicate a progressive leak, a need for calibration or necessary maintenance. 7.1. Warning Management The main principle of handling a warning is to Detect Confirm Act. If a leak is detected the emergency response officer should be contacted and the central control room should be responsible for the communication. Further investigation of a leakage warning is generally needed in order to confirm and quantify the leak before a shut down and evacuation is done. For major leakages the response time should be a matter of minutes. If a second detection system or a further check (e.g. ROV, visual observation) confirm the leak, the procedure of the Emergency Preparedness Plan must be followed and the emergency response officer shall be notified immediately. In critical situations a shut down and evacuation may be the actions. If none of the other detection techniques or the further checks confirm a leak, the warning is considered false. The warning may then be cancelled and an investigation shall be initiated. Any alarm, real or false, should be recorded according to a defined procedure for recording. Documentation can provide the operator or the supplier of the technology with vital information regarding the performance and use of the techniques. Figure 3 indicates a flow-diagram illustrating the relevant actions when a detector signals a leak.

DETECT
Warning from detector 1

CONFIRM
Check detector 2

ACT
Initiate predefined actions

YES

Record

NO

Further check (e.g. ROV)

YES

Initiate predefined actions

Record

NO

Cancel warning

Record

Figure 3. Flow chart depicting the suggested actions following an warning

Rio Oil & Gas Expo and Conference 2012 7.2. Audits and Reviews Regular audits should be carried out to ensure that the personnel knows and follows the relevant procedures and to ensure that the procedures are correct and up to date. It is recommended that real life tests/drills based on a leak detection alarm is run regularly. The system should be regularly reviewed, and opportunities for improvement should be identified and implemented. Particular attention should be given to the management of change (MOC), since future modifications on the facilities might impact the efficiency of the leak detection system.

8. Conclusions and Further Developments

Leak detection is becoming increasingly important and the technologies used have improved significantly over the last decades. Detectors should ideally be able to perform identification, localization, quantification and classification of a leakage for the organization to be able to take the appropriate actions and thereby reduce spill volumes, environmental impacts and possible damage to company reputation. However, there are no single technologies today that can perform all of the above mentioned operations and it might even be challenging to accomplish this by using a combination of sensors. There is a need for a systematic process to evaluate the best suited offshore leak detection system for each installation or field and it is vital that the optimal solutions are chosen. Other challenges related to further developments within the field of leak detection technologies include robustness of sensors, interfaces and integration of the systems and appropriate qualification and testing. As more systems are installed, it is also important to document the practical experience using the systems. With limited real testing possibilities, such as oil on water exercises, well-documented field use can provide developers with vital information regarding the limitations and capabilities of the technologies.

9. Acknowledgements
The authors would like to thank colleagues participating in DNVs leak detection projects. Their research and inputs have been invaluable for the projects within this field.

10. References
DNV, Selection and Use of Subsea Leak Detection Systems, DNV-RP-F302, 2010 Standards Norway, Environmental Care, NORSOK Standard, S-003, Rev. 3, 2005