Reliability and Availability Assessment of Seabed Storage Tanks Using

Ocean Engineering 120 (2016) 1 – 14 14

Contents lists available at ScienceDirect at ScienceDirect

Ocean Engineering journal homepage: www.elsevier.com www.elsevier.com/locate/oceaneng /locate/oceaneng

Reliability and availability assessment of seabed storage tanks using fault tree analysis In-Hwan Choi, Daejun Chang n Division of Ocean Systems Engineering, Korea Advanced Institute of Science and Technology, Technology, 291 Daehak-ro, Yuseong-gu, Daejeon 305-701, 305-701, Republic of Korea

a r t i c l e

i n f o

Article history: Received 19 January 2015 Received in revised form 22 February 2016 Accepted 21 April 2016 Available online 7 May 2016 Keywords: Seabed storage tank New subsea production system Reliability Availability Maintainability

a b s t r a c t

This study proposes the concept of subsea production systems with a seabed storage tank to provide an alternative to conventional � oating facilities and performs the reliability, maintainability and availability study for the seabed storage tank. The reliability assessment of the seabed storage tank performs a fourstep procedure. procedure. A four-ste four-step p procedure procedure is to de�ne the system system boundary boundary,, collect collect the reliabil reliability ity data, data, construct construct a fault tree and estimate estimate the reliability reliability.. The failure and repair data are obtained obtained from the component of the lowest level of the seabed storage tank because the seabed storage tank is a new system unknown reliability data. Reliability of the seabed storage tank is estimated with a consideration of critical events. The total failure frequency of the SST is estimated approximately 2.27  10 4/hour. The maintainability analysis is estimated in accordance with MIL-HDBK 472 Procedure V. The elements of repair time is obtained from the subsea JIP 2000 and RAM study of � eld data. Active repair time from OREDA 2009 is regarded as the repair/replace of repair time elements. The total repair time is mainly effected by the preparation time of repair time elements. The system availability for the seabed storage tank under normal operation is calculated as approximately 91.8%. & 2016 Elsevier Ltd. All rights reserved.

1. Introduct Introduction ion Offshore installations for oil and gas production have gradually moved from shallow water to deeper water and ultra-deep water (Sandrea and Sandrea, 2007). 2007). Although the oil and gas production from offshore installations has appeared � at since 2002, there has been a gradual decrease in shallow water (0 – 400 400 m) production Abbreviations: AIR, Abbreviations: AIR, abnormal abnormal instrument instrument reading; reading; AU, accumulator accumulator unit; BOP, blowout preventer; BP, booster pump; CS,, coarse strainer; DSV, dive support vessel; ELP, external leakage process; ELU, external leakage utility; FMECA, failure mode, effects, and criticality analysis; FPSO, � oating production storage and offloading; FS, � ow sensor; Ft, � lter; FTA, fault tree analysis; HCU, hydraulic coupling unit; HLS, hydrocarbon leak sensor; hr, hours; HPFC, high pressure �uid carrier; HS, hydro-cyclone/skimmer; IGF, induced gas � oatation; ISCU, inlet seawater cleaning unit; MAROS, maintainability, availability, reliability and operability simulation; MTTF, mean time to failure, hrs; MTTR, mean time to repair, hrs; OREDA, offshore reliability data; OSCU, outlet seawater cleaning unit; PD, parameter deviation; PSCU, power/signal coupler unit; PSU, power supply unit; PTS, pressure and temperature sensor; PU, piping unit; r, repair time, hrs; R(t), reliability; RAM, reliability, availability and maintainability; RBD, reliability block diagrams; SBP, subsea booster pump; SCM, subsea control module; SD, structural de �ciency; SDS, sand detection sensor; SEU, subsea electronic unit; SM, subsea manifold; SR, subsea riser; SS, subsea separator; SSBU, seabed storage ballast unit; SST, seabed storage tank; SSTM, seabed storage tank module; SSTU, seabed storage tank unit; SU, sensor unit; TLPs, tension leg platforms; t , operation time, hrs; UF, ultra- �ltration; VD, vacuum de-aerator; λ , failure rate, failures per 10 6 h n Corresponding author. author. E-mail address: djchang@kaist [email protected] .edu (D. (D. Chang). http://dx.doi.org/10.1016/j.oceaneng.2016.04.021 0029-8018/ & 2016 Elsevier Ltd. All rights reserved.

and a rapid increase in deepwater (400 – 1500 1500 m) and ultra-deepwate waterr ( 4 1500 1500 m) productio production. n. Shallow-w Shallow-water ater productio production n accounted counted for almost all of offshore oil and gas production production until the 1990s. 1990s. Deepwate Deepwaterr and ultra-de ultra-deep-w ep-water ater oil and gas productio production n have displayed a remarkable increase over the last 15 years (Ad( Addison et al., 2010), 2010), and both deepwater and ultra-deepwater production are expected to be major contributors in the upcoming decades. A major economic factor in offshore oil and gas production is the feasibility of handling handling and transporting transporting of hydroca hydrocarbons rbons to market. market. Generally Generally,, the hydrocar hydrocarbons bons produced produced in offshore offshore �elds must be transported to onshore facilities for subsequent processing and distribution. In the case of shallow-water production, the prod produce uced d hydro hydrocar carbon bonss can be feasib feasibly ly transp transpor orted ted to shore shore through a pipeline system on the ocean � oor or seabed. A pipeline system is typically preferred because it permits continuous � ow to shore regardless of weather or other adverse conditions. However, this system may result in expensive installation costs and tariffs (Ageh et al., 2010). 2010). For For dee deepw pwate aterr and ultraultra-dee deepw pwate aterr produc productio tion, n, pipeli pipeline ne transportation is typically not practical due to the great distance from shore. In these cases, ship-based transportation via shuttle tankers is broadly used to transport the oil and gas to onshore facilities. Shuttle tankers are specially designed vessels that contain liquid hydrocarbon storage facilities. The operations of shuttle tankers tankers are largely dependent dependent on surface surface condition conditions, s, i.e., wind,

2

I.-H. Choi, D. Chang / Ocean Engineering 120 (2016) 1 –14

waves, and current conditions. Thus, shuttle tankers are typically not operated under severe conditions, and their operations are periodically interrupted for maintenance. Consequently, these might result in long delay for shuttle tankers at offshore sites. In consequence, it is necessary to place storage facilities at the offshore sites to avoid delays in production due to the unavailability of the tankers. An offshore storage facility allows continuous production operations that are independent of shuttle tanker availability. In the case of offshore oil and gas production in deepwater and ultradeepwater, �oating storage facilities are needed to store and transfer hydrocarbons to the shuttle tankers. The � oating facilities depend on location, size, and capabilities as well as water depth and climate at that location. Conventional �oating facilities include semi-submersibles, tension leg platforms (TLPs), spars, and � oating production storage and of �oading (FPSO). Semi-submersibles comprise 21% of � oating facilities. This type of production system was a popular solution during the early years of �oating production. A large number of surplus drill rig hulls that could be �tted with the process plants were converted reliably into production units. TLPs are the third most common type of �oating facility and represents 10% ( James, 2006). This type is not suited for use on ultra-deepwater �elds. Tendon weight grows exponentially with increasing water depth and the potential use of full size TLPs is generally considered to be no more than 1800 m. The major disadvantage of TLPs is the lack of storage and the limitation of water-depth. Spars are relatively newcomers to �oating production. Spars have the advantage of enabling the placement of production trees on the platform deck, which is critical in areas in which the use of high paraf �n oil requires frequent intervention. Water depth does not seem to be a limitation. Spars have been utilized in 1710 m water depth and theoretically can be employed in water depths up to 3000 m and beyond. FPSOs are the most common type of �oating facilities. This type is particularly effective in locations that are far from existing pipeline infrastructure or in deepwater locations and accounts for 63% of all �oating facilities. However, an FPSO requires a notably large hull to store oil as well as heavy and complicated topside facilities and also contains a gas-�aring stack, which leads to wasted energy and release of carbon dioxide, a cause of global warming. Furthermore, this structure must be stationary during operation. Over the past several decades, �oating facilities and pipeline transportation systems have been successfully developed in response to challenges in offshore �eld development. However, �oating facilities and pipeline transportation face several disadvantages, i.e., large-scale storage of crude oil in the hull, heavy weight, and increase in the required pipeline installation cost. The concept of a subsea production system with a seabed storage tank (SST) is proposed to offer an alternative to conventional solutions, i.e., subsea tieback development, FPSO and offshore platforms with a subsea production system. Subsea crude oil storage facilities have previously been proposed for various applications. Mott (1968) proposed the submerged oil storage vessel contained within the device � oats on a layer of water which is in open communication with the sea. Burns et al. (1972) proposed an offshore storage tank, to be placed on a sea � oor, having a domeshaped rigid continuous enclosure. Edwards (1972) suggested a storage tank for offshore storage of crude oil and the method of constructing and installing same, wherein the storage tank has a side wall with the shape of a frustum of a hollow right circular cone and an enlarged diameter base or lower end and a reduced diameter top or upper end having a roof. Pogonowski (1972) proposed the deep water storage facility formed primarily of concrete, for holding a liquid such as crude oil having a lesser density than that of the surrounding water. The facility includes a

�oatable

base which is supportably fastened to a storage tank. Banzoli et al. (1976) proposed a � exible membrane between the oil and the water in a submersible oil-storage tank. A submersible oilstorage tank is constructed from a foundation plate which has a �at lower face and a concave upper face and contains a series of hollow cells. Choi et al. (2002) suggested a seabed oil storage tank to store hydrocarbons. At least one � uid channel is included which has a � rst end positioned inside of the tank proximal a bottom of the tank and a second end in � uid communication with seawater outside of the tank. Arild et al. (2004) proposed seabed located storage for crude oil or other �uid, distinguished in that it comprises a storage section in form of an oil and waterproof cloth formed as a �exible balloon that can be �lled with, store and emptied for a storage � uid. As a result it has been necessary in the past studies to build large undersea oil storage facilities of very heavy reinforced concrete to ensure that the oil containing compartments were isolated from the sea and at the same time to withstand the large forces produced by the sea on the walls of the oil compartments. These concrete or steel structures were expensive to manufacture; and, because of their great weight, their installation was also very dif �cult and expensive. Therefore, the proposed seabed storage tank in this study was patented to solve these problems as mentioned above. Proposed seabed storage tank includes a body having a storage space therein and formed of light weight concrete inner and outer sides of which are watertight coated or plated; a ballast placed on the body of the subsea tank; and a separation unit disposed inside the body and partitioning the storage space upper and lower, the separation unit being movable vertically in the storage space in accordance with the degree of storage �uid �lling (Chang and Bergan, 2014). The use of SST includes the following advantages. The �rst advantage is that it provides storage for offshore production such that large amounts of hydrocarbons can be continually produced irrespective of surface weather condition and also avoids the need for a shuttle tanker to be stationed at the production site at all times. The second advantage is the use of these systems in con junction with the subsea processing system and the production platform. The third advantage eliminates the need for costly deepwater pipelines to shore and in certain cases, the proposed system may be used to avoid expensive pipeline installations. The fourth advantage is greater storage capacity for offshore production sites in deepwater and ultra-deepwater that is less costly to operate and maintain than that of previous � oating facilities that are primarily dependent upon large-scale storage, and shuttle tankers. As a result, the size of the �oating facility is signi�cantly minimized. The minimized �oating facility contains the required utilities and hydrocarbon handling systems without the need for large storage and topside processes. The � fth advantage offers an eco-friendly system due to the absence of a gas- �aring stack on the � oating facility. The sixth advantage is to reduce the number of shuttle tankers required for hydrocarbon transportation. System reliability, availability and maintainability (RAM) have become the focus of many industries in recent times. Early RAM studies provided baseline values to compare alternative designs in the development stage of systems. During the conceptual development stage, numerous design concepts are generated and evaluated to determine whether a particular set of requirements (performance, costs, safety, etc.) is met and associated with the appropriate level of technology and risk. During the period of 1970s to the beginning of 1980s, the RAM analysis was mainly applied in the military � eld. After that period, the applications of RAM analysis have spread to other industries such as aerospace, electrical power plants, oil and chemical industries, and nuclear power plants. Carlier et al. (1996) evaluated the reliability, availability, maintainability and safety requirements


for manned space vehicle with extended on-orbit stay time. Du Julio and Leet (1988) presented space station synergetic RAM-logistics analysis. This study emphasized to analyze the maintenance activities and processes that were accomplished on-orbit within the known design and support constraints of the space station. Rosin et al. (1999) performed the availability analysis of airport runway. Sandberg and Stromberg (1992) described how the terms system effectiveness from a logistics perspective and life cycle cost effectiveness were designed into the Gripen combat aircraft. Willard (2001) discussed the RAM cost control on air Traf �c. Cockerill (1990) studied RAM analysis of a turbine-generator system. Prince and Haire (1991) investigated some Markov models to evaluate the impact of maintenance system availability on the overall plant. Kumar et al. (1992) presented some results from an analytic study of reliability and availability of the crystallization system in sugar plants. Hajeeh and Chaudhuri (2000) worked on reliability and availability assessment of reverse osmosis. This research work assessed the performance of the reverse osmosis plants in Arabian Gulf region by analyzing its failure behavior and down time patterns. Barabady (2005) studied reliability and maintainability analysis of crushing plants. In this study crushing plants were divided into seven subsystems for each of which reliability analysis was done using failures data. Martorell et al. (2004) discussed the roll of technical speci�cation and maintenance activities at nuclear power plants, aiming to increase RAM of safety-related equipment, which in turn must yield to an improved level of plant safety. RAM analysis of subsea systems have been studied for a few years. Duell et al. (2001) identi�ed a number of key processes and tools needed to support British Petroleum's deepwater reliability vision and de�ne tasks necessary to move the vision forward. Robert and Laing (2002) studied methods available to increase the understanding of reliability, including understanding the causes and frequency of failure, through data acquired from �led experience and accelerated testing. Brandt and Eriksen (2001) demonstrated how RAM analysis can be used to quantify the costs associated with well interventions and subsea repairs. Brandt (2003) proposed a methodology for systematic evaluation of the system, applying risk and reliability techniques in combination with veri�cation and quali�cation procedures, to identify technical uncertainties and successfully manage these risks and uncertainties. Holand (2001) collected the blowout preventer (BOP) reliability data and analyzed to reveal BOP reliability problems.

3

Results focus on rig downtime caused by BOP failures, criticality of failures in terms of ability to control a well kick, and BOP subsea test-time consumption. Results have been compared with corresponding results from previous BOP reliability studies carried out for other areas and periods. Sunde (2003) proposed a new design guideline, based on a computerized tool for assessment of the reliability and cost of subsea process systems. Alhanati and Trevisan (2012) investigated reliability gaps in an electrical submersible pump technology for deepwater applications. The previous researches represent signi �cant efforts towards RAM study for subsea systems. However, RAM study for subsea systems have not yet been suf �ciently developed. The RAM study for subsea systems are mainly focused on the existing subsea system based on the historical and experienced data. This paper is an attempt to study the methodology of reliability estimation for new subsea system in the absence of historical and experience data and modeling of real-world collective maintenance policies for repair sources. The objectives of this study are to propose a new subsea production system that uses a SST, to estimate the reliability for a new system (such as the SST) using the FTA method, to analyze the maintainability in accordance with MIL-HDBK 472 Procedure V and to assess the availability of the SST based on the estimations of reliability and maintainability.

2. System description The new subsea production system with the SST consists of a wellhead/X-mas tree, subsea manifold (SM), subsea separator (SS), SST, subsea booster pump (SBP), subsea riser (SR), umbilical, �oater and high-pressure �uid carrier (HPFC). Fig. 1 provides a schematic of the new subsea production system with the SST. New subsea production system with the SST is located in an oil-dominant �eld. The produced well �uids in the reservoir are commingled to the SM and then �ow into the SS which is gravity based 3-phase separator. The well �uids consist of crude oil, gas, water and various contaminants. The SS separates gas, oil and sand/ water. Separated gas and water from the SS are re-injected in the reservoir to increase pressure of well reservoir that has been depleted by production. Also, re-injection helps to decrease unwanted waste, such as � aring. Most of the sand in produced water from separator is removed by a sand cyclone and discharged into

Fig. 1. Schematic of the subsea production system with the SST.

4


Fig. 2. Schematic of the breakdown of the SST.

the sea. Fig. 2 illustrates the hierarchical structure of the SST. The SST can be broken down into the seabed storage tank module (SSTM) and the subsea control module (SCM). The SSTM consists of several components, i.e., the seabed storage tank unit (SSTU), seabed storage ballast unit (SSBU), inlet seawater cleaning unit (ISCU), outlet seawater cleaning unit (OSCU), piping unit (PU) and sensor unit (SU). The SCM is composed of the subsea electronic unit (SEU), accumulator unit (AU), hydraulic coupling unit (HCU), power supply unit (PSU) and power/signal coupler unit (PSCU). Fig. 3(a) – (d) indicates the breakdown of the ISCU, OSCU, PU and SU, respectively, and these units consist of several components. 2.1. Seabed storage tank module The SSTM is an assembly of several units. The function of the SSTM is to store the well � uid until it is transported to the HPLC. Fig. 4 shows a schematic of the SSTM and the SSTM units are divided into SSTU, SSBU, ISCU, OSCU, PU and SU. The produced oil from the SS � ows through the feed PU and is stored in the SSTU. A tiny amount of gas (dissolved gas) is released from the oil due to the pressure difference between the SS and the SST and accumulates in the SSTU. The layers of oil and gas are separated due to a difference of oil and gas density. The average density of seawater is greater than that of oil; consequently, the well �uid may be separated into oil layer and seawater layer. When the SSTU is gradually � lled with gas and oil, the seawater from the storage space leaves the SSTU. The ISCU and OSCU remove the deposits, impurities, and hydrocarbons that follow the � ow of seawater. The cleaning system should include the functions of � ltering and disposal of the � ltered materials. Oil stored in the oil storage space is exported by the oil product PU to the SBPU. A small amount of gas is released due to the pressure difference between the SS and SSTU. As the inventory is cooled by the surrounding seawater at a lower temperature, this gas dissolves into the oil layer. If gas accumulates in the SST, it is re-injected into the well reservoir.

concrete are watertight coated or surface plated.

2.1.1. Seabed storage tank unit The SSTU is an assembly of a storage tank in which the well �uid is stored until they are transported to the SSBM. The SSTU contains three storage spaces, i.e., a gas storage space, oil storage space and seawater storage space. The body takes the form of a cylinder or polyprism and may be composed of lightweight concrete. In this case, the inner and outer sides of the lightweight

2.1.2. Seabed storage ballast unit The SSBU is �lled with the ballast body after towing and placement of the SST on the sea �oor. The purpose of the unit is to ensure that the SST will not depart from the sea � oor after the SST is �lled with hydrocarbons, which have a lower density than seawater. The SSBU can include a concrete body, a concrete block, a sandbag, loose sand, gravel or rock.

Fig. 3. Hierarchy of the ISCU, OSCU, and SU.


2.1.3. Inlet seawater cleaning unit The ISCU is composed of several components: a coarse strainer (CS), � lter (Ft), vacuum de-aerator (VD) and booster pump (BP), as shown in Fig. 5. The CS removes large oceanic contaminants, i.e., silt, algae, and crustaceans, and is typically rated for 80 – 100 m m. The CS is self-cleaning, thus allowing for uninterrupted continuous �ow at full capacity during backwash. Filtered seawater from the coarse strainers is subsequently �ltrated by ultra-�ltration (UF) membranes. The UF membranes provide high-ef �ciency (98%) removal of particulates with sizes greater than 2 – 5 m m. The VD removes dissolved oxygen to control corrosion and/or aerobic bacteria activity and is conventionally used in vacuum-stripping or gas-stripping packed towers. Water from the VD tower is treated with further �ltration or polishing steps using cartridge �lters to provide absolute �ltration performance. 2.1.4. Outlet seawater cleaning unit Fig. 6 indicates the schematic of the OSCU. The OSCU is comprised of several components: a hydro-cyclone/skimmer (HS), induced gas �oatation (IGF), Ft and BP. The HS removes the dispersed oil containing the seawater output from the SST and provides a dispersed oil level of less than 500 mg/L. The IGF is treated with further �ltration below 20 ppm/L and further polishing is accomplished via �ltration by the UF membrane of the ISCU. The treated seawater is discharged by the BP. 2.1.5. Piping unit The feed PU and product PU are designed for the in �ow and

Fig. 4. Schematic of the SSTM.

5

out�ow piping that control the in�ow and out�ow of gas, oil and seawater into the SST. The PU is composed of the connector, sealine, safety joint and pipe spool. 2.1.6. Sensor unit The SU is divided into a pressure and temperature sensor (PTS), �ow sensor (FS), hydrocarbon leak sensor (HLS) and sand detection sensor (SDS), as shown in Fig. 3(c). 2.2. Subsea control module The SCM is composed of the SEU, AU, HCU, PSU and PSCU, as illustrated in Fig. 2. The typical subsea control module receives communication signals and electrical power from the topside control equipment and these modules use the signals to control the subsea system.

3. Modeling of RAM 3.1. Reliability Reliability is de�ned as the ability of an item to perform a required function under the given environmental and operational conditions for a stated period of time (International Standards Organization, 1986). The most common methods of reliability analysis include reliability block diagrams (RBD), fault tree analysis (FTA), and failure mode effects and criticality analysis (FMECA) (Bai and Bai, 2010; Quilici et al., 1998; Wabnitz and Netherland, 2001). This study uses FTA, which is a widely used systematic and deductive method for de �ning and determining all possible reasons that could cause a top event to occur ( Carter and Powell, 2006; Duell et al., 2001; Skeels et al., 2003). This method originated in the aerospace industry and was adapted by the nuclear power plant industry to qualify and quantify the hazards and risks involved in nuclear power generation. The reliability of a system may be viewed as a measure of its quality and performance. Mathematically, reliability (R(t )) and Unreliability (F (t ))is expressed as follows for a system with a constant failure rate: R (t )=e− λt

(1)

F (t )=1 − e− λt

(2)

Fig. 5. Schematic of the ISCU.

6


Fig. 6. Schematic of the OSCU.

The active repair time denotes the real time required to repair or replace the equipment. Due to the lack of reliability data for the ISCU and OSCU for the SST, the constant failure rate and active repair time are inversely obtained from the component of the lowest level of the systems. Next, the reliability data for the components are obtained from the relevant topside systems. The reliability data of the SSTU are estimated from the casing string obtained from OREDA because the cemented structure has a similar function for protection of the aquifers. The reliability data for the remainder of the subsea units are obtained from a subsea section of the OREDA.

Fig. 7. Procedure for reliability estimation.

where λ is the constant failure rate in failures per 10 6 h and t is the time period. The reliability assessment procedure for SST is composed of the four steps shown in Fig. 7. 3.1.1. Step 1: De �nition of the system boundary The �rst step involves de�nition of the system boundary and the critical event. The system boundary analysis is divided into subsystems of module, unit and component. Figs. 2 and 3 indicate the system breakdown of the SST. A critical event is de�ned by considering what type of critical event occurs and where it occurs. The critical event to be analyzed is commonly referred to as the top event for the fault tree. 3.1.2. Step 2: Collection of the reliability data The second step involves collection of reliability data from generic reliability data sources. In this step, the reliability and maintainability analyses require several types of input data such as constant failure data and active repair time, which provide information related to how often units and components fail. The information on the constant failure rate of the critical failures and the active repair time for the units and components is collected from OREDA (2009). Table 1 presents the constant failure rate and active repair time for the units and components. The constant failure rate is a mean value and may be constant or time-varying.

3.1.3. Step 3: Construction of the fault tree A fault tree is a logic diagram that indicates the relationship between a potential critical event in the system and the cause of this event. The potential critical event is referred to as the top event. The immediate causal events lead to the top event and connect to the top event via a logic gate. This procedure of fault tree construction is deductively continued until a suitable level of detail is reached. The events on the lowest level are known as the basic events of the fault tree. The AND gate indicates that both of the basic events must occur for the top event to occur, whereas the OR gate denotes that only one of the basic events may be the cause of a top event (Ayyub, 2003; Hong et al., 2009; Reniers et al., 2005; Yuhua and Datao, 2005). 3.1.4. Step 4: Estimation of the reliability The reliability of the SST is assessed relative to its intended and designed functions under normal production conditions. The SST can be used in other operational modes, such as work-over, intervention, shutdown and retrieval to surface. However, the reliability of the SST under normal operation represents the most important assessment. Table 2 indicates the basic formula of FTA. The critical events are estimated using the rules for the logic gates. 3.2. Maintainability estimation Maintenance is de�ned as the action that restores failed units to an operational condition or retains non-failed units in an operational condition (British Standards Institution, 1991). For repairable systems, maintenance plays a vital role in the life of a system and affects its reliability, availability, downtime and cost of operation, among other factors. Maintenance actions can be

Table 1 Failure and repair data for the SST. Facility

Module

Unit

Component

Seabed Storage Tank (SST)

Seabed Storage Tank Mod- Seabed StorageTank Unit (SSTU) – ule (SSTM) Seabed Storage Ballast Unit – (SSBU) Inlet Seawater Cleaning Unit Coarse strainer (ISCU) Filter Vacuum de-aerator

Outlet Seawater Cleaning Unit (OSCU)

Piping Unit (PU) Sensor Unit (SU)

Subsea Control Module (SCM)

Subsea Electronic Unit (SEU) Accumulator Unit (AU) Hydraulic Coupling Unit (HCU) Power Supply Unit (PSU) Power/Signal Coupler Unit (PSCU) Module Base Plate Unit (MBPU) Solenoid Control Valve Unit (SCVU)

Failure rate (per 106 h)

Active repair time (hours)

Reference

0.35 –

288 –

Casing obtained from subsea part from OREDA (2009), p 138 Non-repairable unit

87.66

13

0.09 30.21

2 28

Booster pump

51.29

35

Hydro-cyclone/skimmer

6.15

1.7

Induced gas � oatation

10.69

2.1

Filter Booster pump

0.09 51.29

2 12

Contactor obtained from topside part from OREDA (2009), p 339 Subsea part from OREDA (2009), p 85 Coalescer obtained from topside part from OREDA (2009), p 331 Centrifugal seawater Lift pump obtained from topside part from OREDA (2009), p 165 Hydro-cyclone obtained from topside part from OREDA (2009), p 360 Separator (10m3) obtained from topside part from OREDA (2009), p 383 Subsea part from OREDA (2009), p 85 Centrifugal oilwater treatment pump obtained from topside part from OREDA (2009), p 163 Subsea part from OREDA (2009), p 95 Subsea part from OREDA, 2009, p 74

– Pressure, temperature sensor Flow Sensor Hydrocarbon Leak sensor Sand detection Sensor – – – – –

0.49 4.63

22.2 3.3

13.36 0.64 4.76 4.93 0.15 0.03 0.07 0.19

16.2 2 1 6.5 12 14 4 19.8

– –

0.28 0.76

7.3 15.4

Subsea part from OREDA (2009), pp 60 – 61

.I -H . C h o i , D . C h a n g / O c e a n E n g i n e e r i n g 1 2 0 ( 2 0 1 6 ) 1 – 1 4

7

8


divided into two types: corrective maintenance and preventive maintenance. Preventive maintenance is the practice of replacing components or subsystems before they fail to promote continuous system operation. The schedule for preventive maintenance is based on observation of past system behavior and knowledge of which components are vital to continued system operation. Cost is always a factor in scheduling of preventive maintenance. In many circumstances, it is �nancially more sensible to replace parts or components that have not yet failed at predetermined intervals rather than wait for a system failure that may result in a costly disruption in operations. Corrective maintenance describes the actions taken such that a failed system is restored to operational status. This process usually involves replacing or repairing a component or subsystem that is responsible for the failure of the overall system. Such maintenance is performed at unpredictable intervals because a component's failure time is unknown. The objective of corrective maintenance is to restore the system to satisfactory operation within the shortest possible time. Corrective maintenance is considered in this study and Fig. 8 illustrates the procedure for estimating repair time in accordance with the MIL-HDBK-472 Procedure V (U. S. Department of Defense, 1966). The elements for the repair time are de �ned as follows:

3.3. Availability Availability is de�ned as the ability of an item (under the combined aspects of reliability, maintainability and maintenance support) to perform its required function at a stated instant of time or over a stated period of time ( British Standards Institution, 1991). This measure takes into account the system reliability and maintainability. The average availability denotes the mean proportion of time that the item is functioning and is de �ned as follows: Aav =

MTTF

MTTF+MTTR

(3)

where Mean Time To Failure (MTTF) denotes the average amount of time until a system fails or its expected failure time. Please note that the MTTF can be assumed to be the inverse of failure rate ( λ) for a series of components, all of which have a constant failure rate and MTTR denotes the mean downtime after a failure. MTTR is a useful parameter that should be used early in planning and designing stages of a system. The parameter is used in assessing the accessibility/locations of system; for example, a system that often fails should be located where it can easily be removed and replaced. Eq. (4) indicates the equation for MTTR estimation. MTTR consists of the following elements: Preparation, Fault isolation, Disassembly, Repair/replacement, Reassembly, Alignment, Checkout and Start-up.

8


divided into two types: corrective maintenance and preventive maintenance. Preventive maintenance is the practice of replacing components or subsystems before they fail to promote continuous system operation. The schedule for preventive maintenance is based on observation of past system behavior and knowledge of which components are vital to continued system operation. Cost is always a factor in scheduling of preventive maintenance. In many circumstances, it is �nancially more sensible to replace parts or components that have not yet failed at predetermined intervals rather than wait for a system failure that may result in a costly disruption in operations. Corrective maintenance describes the actions taken such that a failed system is restored to operational status. This process usually involves replacing or repairing a component or subsystem that is responsible for the failure of the overall system. Such maintenance is performed at unpredictable intervals because a component's failure time is unknown. The objective of corrective maintenance is to restore the system to satisfactory operation within the shortest possible time. Corrective maintenance is considered in this study and Fig. 8 illustrates the procedure for estimating repair time in accordance with the MIL-HDBK-472 Procedure V (U. S. Department of Defense, 1966). The elements for the repair time are de �ned as follows:

3.3. Availability Availability is de�ned as the ability of an item (under the combined aspects of reliability, maintainability and maintenance support) to perform its required function at a stated instant of time or over a stated period of time ( British Standards Institution, 1991). This measure takes into account the system reliability and maintainability. The average availability denotes the mean proportion of time that the item is functioning and is de �ned as follows: Aav =

MTTF+MTTR

performed before fault isolation can be executed. isolate the fault to the level at which fault correction begins.  Disassembly: The time associated with gaining access to the

replaceable item or items identi�ed during the fault isolation process.  Repair/replacement: The time associated with the removal and replacement of a faulty replaceable item or suspected faulty item.  Reassembly: The time associated with closing up the equipment after the interchange is performed.  Alignment: The time associated with aligning or tuning the system or replaceable item after a fault has been corrected.  Checkout: The time associated with veri �cation that a fault has been corrected and that the system is operational.  Start-up: The time associated with bringing a system up to the operational state and verifying the system.

(3)

where Mean Time To Failure (MTTF) denotes the average amount of time until a system fails or its expected failure time. Please note that the MTTF can be assumed to be the inverse of failure rate ( λ) for a series of components, all of which have a constant failure rate and MTTR denotes the mean downtime after a failure. MTTR is a useful parameter that should be used early in planning and designing stages of a system. The parameter is used in assessing the accessibility/locations of system; for example, a system that often fails should be located where it can easily be removed and replaced. Eq. (4) indicates the equation for MTTR estimation. MTTR consists of the following elements: Preparation, Fault isolation, Disassembly, Repair/replacement, Reassembly, Alignment, Checkout and Start-up.

 Preparation: The time associated with those tasks that must be  Fault isolation: The time associated with those tasks required to

MTTF

n

MTTR =

∑i = 1 λ i r i ∑in= 1 λ i

(4)

where n is the number of sub-systems, λ i is the failure rate of the i-th subsystem, and r i is the repair time of the i-th unit.

4. Results and discussion 4.1. Assumptions The following assumptions were made in the reliability, maintainability, and availability analysis:  All of the modules, units and components comply with the

exponential failure model.  The constant failure rate and active repair time are obtained

from historical reliability data sources such as OREDA (2009).  The constant failure rate and active repair time of the unknown

Table 2 Basic formula of FTA. Gate Formula

Description

AND P(A)*P(B)

Output fault occurs if all of the input events occur Output fault occurs if at least one of the input events occur

OR

P(A) þ P(B)  P(A)*P (B)

subsea components for the SST are obtained from corresponding topside components from OREDA (2009).  The critical failures of the SST are considered under normal operation.  The lifecycle of the SST is assumed to be ten years.  Corrective maintenance for critical failures is considered.  The repair time of the modules, units and components are calculated using MIL-HDBK-472 Procedure V.  The elements of the repair time are obtained from the subsea JIP

Fig. 8. Procedure for repair time estimation.


Fig. 9. FTA for the abnormal instrument reading.

Fig. 10. FTA of the external leakage process.

Fig. 11. FTA of the external leakage utility.

9

10


Fig. 12. FTA of the parameter deviation.

Fig. 13. FTA of the structural de �ciencies.

2000 and the RAM study of � eld data.  The active repair time is treated as the repair/replacement of

repair time elements. 4.2. Reliability estimation of SST The reliability of the SST is estimated using the method proposed in Section 3.1. The critical events of the SST are obtained using a separator from the OREDA topside part due to similar functions. The constant failure times and active repair times of the components (basic events) are obtained from OREDA subsea part. For some components the data is gathered from the relevant equipment based on OREDA topside part. Table 1 gives failure data and repair data for selected key components of the SST. Critical events for the SST under normal conditions include the following:  Abnormal instrument reading (AIR)  External leakage process (ELP)  External leakage utility (ELU)  Parameter deviation (PD)  Structural de�ciency (SD)

These critical events under normal conditions are crucial to SST failures in which the system fails to store hydrocarbons and increases maintenance costs. Each of the critical events is analyzed in different fault trees in Figs. 9 – 13. Fig. 9 shows the fault tree for the AIR that indicates a false alarm or faulty instrument indications from the SSTU, SU and SCM. The SCM fails to send the correct values or signals in time to the production valves or sensors on the SST. Fig. 10 illustrates the fault

tree for the ELP, which means that the produced � uids leak to the environment and the production rate is decreased. This critical failure can occur in the SSTU, ISCU, OSCU and PU. The ELU, whose fault tree shown in Fig. 11, corresponds to the leakage of lubrication and seal oils for the booster pump, � lter, AU, HCU, etc. Fig. 12 illustrates the fault tree for the PD, which means that the CS, VD and IGF exceed the tolerances of the monitored parameter. The SD in Fig. 13 indicates the fault tree. The fault tree analysis is performed to estimate the probability � of ve critical events by using Eq. (2), which lead to the reliability of SST. Table 3 indicates the constant failure rate for basic events in �ve critical events. Fig. 14 presents the reliability of the top events for the SST and the reliabilities of the top events are calculated as described in Eq. (1). The time-dependent reliability calculations are carried out for operation time up to 87,600 h. The long operation time of 87,600 h is chosen to observe the differing time-varying reliability behavior between critical events. It must be noted that the reliability of critical events with time decreases without any maintenance actions. As shown in Fig. 14, the reliability of the SST is indicated as the product of the reliability of the top events for the SST. All of the reliabilities of the top events steadily decrease with passage of time. After 10,000 h, the reliabilities of the AIR, ELP, ELU, PD and SD are approximately 0.67, 0.4, 0.64, 0.77 and 0.83, respectively, which means that the reliability of the SD exhibits the highest value of all critical failures. The reliability of SST is approximately 0.1 after 10,000 h. For AIR and EPU, the reliabilities have similar values for the operation time of the SST.

I.-H. Choi, D. Chang / Ocean Engineering 120 (2016) 1 –14 Table 3 Constant failure rate for basic event. Critical failures (Top events)

Basic event

Abnormal instrument reading Failure of SSTU Failure of CS Failure of VD Failure of PTS Failure of FS Failure of HLS Failure of SDS Failure of IGF Failure of SEU Failure of AU Failure of PSU Failure of PSCU External leakage process Failure of SSTU Failure of VD Failure of BP Failure of Ft Failure of HS Failure of IGF Failure of BP Failure of Ft Failure of PU External leakage utility Failure of SSTU Failure of VD Failure of BP Failure of Ft Failure of IGF Failure of BP Failure of Ft Failure of AU Parameter deviation Failure of SSTU Failure of CS Failure of VD Failure of IGF Structural de�ciency Failure of SSTU Failure of VD Failure of BP Failure of IGF Failure of BP

11

Table 4 Mobilization time of repair sources. Constant failure rate (/hrs) 0.00000983 0.00001253 0.00000213 0.00000077 0.00000893 0.00000064 0.00000245 0.00000213 0.00000027 0.00000015 0.00000007 0.00000019 0.00000289 0.00000506 0.00001501 0.00000007 0.00000615 0.00000506 0.00005505 0.00000007 0.0000005 0.0000007 0.00000506 0.00000292 0.00000007 0.00000506 0.00003044 0.00000007 0.00000015 0.00000133 0.00001425 0.00000506 0.00000506 0.00000471 0.00000497 0.00000319 0.00000497 0.00001288

Resources

Rig

Description

 Heavy work-over of the SST  Seawater

Dive Support Vessel (DSV)

DSV/Weld Cable-lay Vessel

Mobilization times (days)

cleaning

14

unit

replacement c ont rol m od ul e 7 replacement  Pipeline and connector repair  Sensor unit repair Hyperbaric weld operations for 10 connector and pipeline repair. Replacement of pipeline and 14 umbilical  S ub se a

failure and are ready to begin the repair tasks. The mobilization time for different vessels depends on the availability of the vessels and assumes that these vessels are immediately available. The rig is used for repair and heavy work-over of the SST and replacement of the seawater cleaning unit. The dive support vessel (DSV) is used to replace the SCM, repair the PU and connector, and repair the SU. The DSV/weld is applied to repair pipeline welding, and the cable-lay vessel is to replace the pipeline and umbilical. 4.3.2. Repair time for SST Table 5 shows the repair time of each unit for the SST. The repair time is estimated in accordance with the MIL-HDBK-472 Procedure V. The maintenance times for the tasks are obtained from the subsea JIP 2000 and � eld data. It should be indicated that the preparation time is the largest part of the elements of the repair time. The preparation time of the SSTU is 356 h and occupies 51% of a total repair time while the repair/replacement time of SSTU is 288 h (41%). However, the preparation time of the rest units for the SST is 86% of the mean value and repair/replacement time accounts for approximately 5.8% of the mean value. Consequently, the preparation time occupies the dominant part of the total repair time due to the mobilization time of repair sources.

4.3. Maintainability estimation of SST 4.3.1. Repair sources Table 4 shows the mobilization time for the repair vessels and the relevant repair tasks. The mobilization time indicates that repair sources are located at the site after the occurrence of the

4.4. Availability estimation of SST The availability of the SST is estimated using Eq. (3) in Section 3.3. The MTTF of the SST is approximately 4405 h, considering the

Fig. 14. Reliability for the SST.

1 2

Table 5 Repair time of SST. Procedure step

Repair times (hr) SSTU (Rig)

Preparation

Fault isolation Disassembly Repair/replacement Reassembly Alignment Checkout Total

Mobilization Positioning Install guide-wires Run the running tools Release the pipeline /umbilical connection Repair /replacement Make up the pipeline/connection Remove the running tools Retrieve guidewires Test

336 12 8 16 4 288 8 8 3 12 695

ISCU(Rig)

OSCU (Rig)

PU (DSV)

CS

VD

Ft

BP

HS

IGS

Ft

BP

336 12 8 8 2 13 5 4 3 5 396

336 12 8 8 2 2 5 4 3 3 383

336 12 8 8 2 28 5 4 3 6 412

336 12 8 8 2 35 5 4 3 8 421

336 12 8 8 2 1.7 5 4 3 5 384.7

336 12 8 8 2 2.1 5 4 3 6 386.1

336 12 8 8 2 2 5 4 3 3 383

336 12 8 8 2 12 5 4 3 8 398

168 6 8 6 4 22.2 5 4 3 12 238.2

SU (DSV)

SCM (DSV)

PTS

HLS

FS

SDS

SEU

AU

HCU

PSU

PSCU

MBPU

168 6 8 2 4 3.3 4 1 2 2 200.3

168 6 8 2 4 2 4 1 2 2 199

168 6 8 2 4 16 4 1 2 4 215

168 6 8 2 4 1 4 1 2 8 204

168 6 8 2 4 6.5 4 1 2 2 203.5

168 6 8 2 4 12 4 1 2 4 211

168 6 8 2 6 14 4 2 2 4 216

168 6 8 2 4 4 4 1 2 2 201

168 168 6 6 8 8 2 2 4 6 19.8 7.3 4 4 1 2 2 2 8 4 222.8 209.3

SCVU 168 6 8 2 3 15.4 4 2 2 4 214.4

.I -H . C h o i , D . C h a n g / O c e a n E n g i n e e r i n g 1 2 0 ( 2 0 1 6 ) 1 – 1 4


13

Fig. 15. Availability for the SST.

critical failure rate of the SST. The MTTR of the SST is approximately 391.3 h and is calculated using the weighted average method in Eq. (4). The availability from MTTF and MTTR is 91.8%. It means that an unavailability of the SST is 9.2% during the operation time of the SST. The down time of the SST is approximately 4900 h. Fig. 15 indicates the availability with the number of re-

Acknowledgment This work was supported by the Future Industry's Leading Technology Development Program of MOTIE/KEIT. (10042430, Development of 500 MPa URF & SIL 3 Manifold and Subsea System Engineering for Deepsea Field).


13

Fig. 15. Availability for the SST.

critical failure rate of the SST. The MTTR of the SST is approximately 391.3 h and is calculated using the weighted average method in Eq. (4). The availability from MTTF and MTTR is 91.8%. It means that an unavailability of the SST is 9.2% during the operation time of the SST. The down time of the SST is approximately 4900 h. Fig. 15 indicates the availability with the number of replications. The availability of the SST is calculated by a MAROS version 8.04. It should be noted that the results reveal similar values against the number of replications.

5. Conclusions The concept of the SST in a subsea production system was proposed and sequentially studied via RAM analysis. An important step in the RAM analysis is the collection of appropriate data, and collection of high-quality failure and repair data which is necessary for obtaining reliable and accurate results. In this study, due to unknown reliability data for the SST, the failure and repair data are obtained from the component of the lowest level of the systems. The reliability assessment for the SST was estimated using a four-step procedure. A four-step procedure of reliability estimation was presented to de �ne the system boundary, collect the reliability data, construct a fault tree and estimate the reliability. Reliability of the SST was estimated with a consideration of critical events such as AIR, ELP, ELU, PD and SD under normal operations. ELP failure is the highest probability in critical events. The repair time for SST was estimated with a re �ection of the following elements: preparation, fault isolation, disassembly, repair/replacement, reassembly, alignment, check-out. The repair/ replacement time was from the active repair time in OREDA (2009) and the time for the rest of the elements was obtained by the subsea JIP 2000 and �eld data. The preparation time is the dominant element for the total repair time with 86% of the mean value and repair/replacement time occupied approximately 5.8% of the mean value. MTBF and MTTR of the SST were 4405 h and 391.3 h, respectively. Consequently, the availability of the SST was approximately 91.8%. Further study is recommended for a more detailed RAM assessment. The RAM analysis of the current study was constrained by selected factors, which were taken into account as certain assumptions. The detailed RAM analysis should consider such additional factors as a system con �guration, operation conditions, equipment redundancy and preventive maintenance for the real world.

Acknowledgment This work was supported by the Future Industry's Leading Technology Development Program of MOTIE/KEIT. (10042430, Development of 500 MPa URF & SIL 3 Manifold and Subsea System Engineering for Deepsea Field).

References Addison, F., Kennelley, K., Botros, F., 2010. Thunder horse and atlantis deepwater frontier developments in the gulf of mexico: future challenges for deepwater developments. In: Proceedings of the Offshore Technology Conference. OTC 20404, 3 – 6 May. Houston (TX). Ageh, EA., Uzoh, OJ., Ituah, I., 2010. Production technology challenges in deepwater subsea tie-back developments. In: Proceedings of the 34th Annual SPE International Conference and Exhibition. SPE 140620, 31 July – 7 August. Tinapam, Nigeria. Alhanati, FJS., Trevisan, F., 2012. Reliability Gaps in ESP technology for deepwater applications. In: Proceedings of the Offshore Technology Conference. OTC 23341, 30 April – 3 May. Houston (TX). Arild, Samuelsen, Kaare, Breivik, Ola, Ravndal, 2004. Seabed located storage. WO 2004037681 A1. Ayyub, B.M., 2003. Risk Analysis in Engineering and Economics. CRC Press, Boca Raton (FL). Bai, Y., Bai, Q., 2010. Subsea Engineering Handbook. Gulf Professional Publishing, Boston (MA). Barabady, J., 2005. Reliability and maintainability analysis of crushing plants in Jajarm bauxite mine of Iran. In: Proceedings of the Reliability and Maintainability Symposium. 24 – 27 January. New York (NY). Brandt, H., 2003. Reliability management of deepwater subsea � eld developments. In: Proceedings of the Offshore Technology Conference. OTC 15343, 5 May. Houston (TX). Brandt, H., Eriksen, R., 2001. RAM analysis for deepwater subsea developments. In: Proceedings of the Offshore Technology Conference. OTC 13003, 30 April. Houston (TX). British Standards Institution, 1991. BS 4778: Glossary of Terms Used in Quality Assurance Including Reliability and Maintainability Terms. British Standard Institution, London. Burns, G., Mc Kenna, K., Stafford, D., 1972. Offshore Underwater Storage Tank. US3824942A. Carlier, S., Garbellini, L., Altavilaa, A., 1996. Evaluation of reliability, availability, maintainability and safety requirements for manned space vehicles on-orbit stay time. Acta Astronaut. 38 (2), 115 – 123. Carter, M., Powell, K., 2006. Increasing reliability in subsea systems. E &P Magazine. Hart Energy Publishing, Houston (TX) . Cockerill, A.W., 1990. RAM analysis helps cut turbine-generator systems costs. Power Eng. 94 (7), 27 – 29. Dae Jun, Chang, Pal G., Bergan, 2014. Large Scale Subsea Storage Tank and Method for Constructing and Installing the Same. WO2014061837 A1. Duell, C., Fleming, R., Strutt, J., 2001. Implementing deepwater subsea reliability strategy. In: Proceedings of the Offshore Technology Conference. OTC 12998, 30 April. Houston (TX). DuJulio, ET., Leet, JH., 1988. Space station synergetic RAM – logistic analysis. In: Proceedings of the Annual Reliability and Maintainability Symposium. 26 – 28 January. New York (NY).

14


Edwards, N., 1972. Storage Tank for Offshore storage of liquid and method of constructing and installing same. US 3777497A. George, E Mott., 1968. Submerged Oil Storage Vessel and Oil Loading Facility for Offshore Wells. US 3408971A. Hajeeh, M., Chaudhuri, D., 2000. Reliability and availability assessment of reverse osmosis. Desalition 130, 185 – 192. Holand, P., 2001. Reliability of deepwater subsea blowout preventers. Soc. Pet. Eng. J. 16 (1), 12 – 18. Hong, E., Lee, I., Shin, H., Nam, S., Kong, J., 2009. Quantitative risk evaluation based on event tree analysis technique: application to the design of shield TBM. Tunn. Undergr. Space Technol. 24 (3), 269 – 277. International Standards Organization, 1986. ISO 8402: Quality Vocabulary. International Standards Organization, Geneva . James, R. Mc.Caul, 2006. Growing Requirement for Floating Production Systems. Maritime Reportter and Engineering News, April, New York (NY), pp. 24 – 26 http://www.marinelink.com. Kumar, D., Singh, J., Pandey, P.C., 1992. Availability of the crystallization system in the sugar industry under common-cause failure. IEEE Trans. Reliab. 41 (1), 85 – 91. Martorell, S., Villanueva, J.F., Carlos, S., Nebot, Y., Sanchez, A., Pitarch, J.L., Serrandell, 2004. RAMS þ C informed decision-making with application of multi-objective optimization of technical speci �cations and maintenance using genetic algorithms. Reliab. Eng. Syst Safe. 87 (1), 65 – 75. Michael, Choi, Jack, Chan, David, Tuturea, 2002. Seabed Oil Storage and Tanker Offtake System. US 20020141829 A1. OREDA, 2009. Offshore Reliability Data Handbook, 5th ed. Det Norsk Veritas, DNV, Høvik, Norway. Pogonowski, Ivo, C, 1972. Underwater Liquid Storage Facility. US 3695047 A. Prince, B.E., Haire, MJ., 1991. The in �uence of maintenance system reliability and maintainability characteristics on plant availability. In: Proceedings of the Annual Reliability and Maintainability Symposium. 29 – 31 January. New York (NY). Quilici, M., Roche, T., Fougere, P., Juda, D., 1998. Risk assessment of a BOP and control system for 10,000 m water depth. In: Proceedings of the Offshore Technology Conference. OTC 8791, 4 May. Houston (TX).

Reniers, G.L.L., Dullaert, W., Ale, B.J.M., Soudan, K., 2005. The use of current risk analysis tools evaluated towards preventing external domino accidents. J Loss Prev. Process Ind. 18 (3), 119 – 126. Robert, C., Laing, T., 2002. Achieving reliability improvement for subsea challenges. In: Proceedings of the Society for Underwater Technology: Subsea Controls and Data Acquisition. 13 – 14 June. Paris (FR). Rosin, A., Hecht, M., Handal, J., 1999. Analysis of airport-runway availability. In Proceedings of the Annual Reliability and Maintainability Symposium. 18 – 21 January. Washington (DC). Sandberg, A., Stromberg, U., 1992. Gripen: with focus on availability performance and life support cost over the product life cycle. J. Qual. Maint. Eng. 5 (4), 325 – 334. Sandrea, I., Sandrea, R., 2007. Global offshore oil-2: growth expected in global offshore crude oil supplies. Oil Gas J. 105 (10), Retrieved from 〈 http://www.ogj. com/articles/print/volume-105/issue-10/exploration-development/global-off shore-oil-2-growth-expected-in-global-offshore-crude-oil-supply.html 〉. Skeels, HB., Taylor, M., Wabnitz, F., 2003. Subsea � eld architecture selection based on reliability considerations. In: Proceedings of the Deep Offshore Technology Conference. Marseille (Fr). Sunde, LT., 2003. Subsea process design guideline for reliability. In: Proceedings of the Offshore Technology Conference. OTC 15171, 5 May. Houston (TX). U. S. Department of Defense, 1966. MIL-HDBK-472: Military Standard Handbook, Maintainability Prediction. Washington (DC): U. S. Department of Defense. Valeriano, Banzoli, Giovanni, De Nora, Vincenzo, Di Tella, Domenico, Lalli, Gianfranco, Tempo. 1976. Underwater Stationary Tank for Storing Large Amounts of Crude Oil. US 3943724 A. Wabnitz, F., Netherland, H., 2001. Use of reliability engineering tools to enhance subsea system reliability. In: Proceedings of the Offshore Technology Conference. OTC 12944, 30 April. Houston (TX). Willard, PJ., 2001. Air traf �c control RAMS costs. In: Proceedings of the Annual Reliability and Maintainability Symposium. 22 – 25 January. Philadelphia (PA). Yuhua, D., Datao, Y., 2005. Estimation of failure probability of oil and gas transmission pipelines by fuzzy fault tree analysis. J. Loss Prev. Process Ind. 18 (2), 83 – 88.

Reliability and Availability Assessment of Seabed Storage Tanks Using

Recommend Documents