Flange analysis is performed by use of Brilliant, Petrell's proprietary numerical simulation platform. Brilliant is fully integrated CFD (Computational Fluid Dynamics) system and Finite Element Method (FEM) system. It includes a library for material properties of a number of metallic materials including insulation materials, and also a comprehensive thermodynamic library (ThermoProp).
Brilliant is tailored-made to serve multi-physic simulations, and it is used a platform for other applications including VessFire, Brilliant NWS (Near Wellbore Simulator) and SolidCO2Sim. Brilliant is built on fundamental conservation equations for transport of continuum properties as: • Momentum equation (Newton's second law F=ma) • Mass continuity • Energy equation • Transport of species • Chemical reactions Numerical Models Numerical models have been developed for ASME RF (Raced Face), RJT (Ring Joint Type), Compact and Hub flanges, and different materials in flange, gasket and bolts/nuts have been implemented (Figure 1, Figure 2, Figure 3). Geometrically, the flanges are symmetric and the boundary conditions for the symmetry planes are easy to implement. The model system is flexible and can be adapted to a ¼ of a half flange, a half flange and even the whole flange. A quarter of the flange is used for the calculations to save simulation time. All simulations are generated by running a script that catches required parameters from relevant ASME standards. A condition for applying a reduced model is that the heating is uniform around the flange, which is different from experiments. To see the effect of non-uniform heating, a full 360˚ model should be run. Uniform heat load Higher temperature leads to reduction of the strength of the steel, and consequently a reduction in the ability to take forces. When comparing with experiments, it makes sense to use temperature variation as seen in the experiments, but in calculations performed for design purpose, it makes no sense to apply variation in temperature as part of the scenario. Because a real fire has no general load profile, the scenario description for design purpose should assume a uniform heat load. It is therefore interesting to see how close a uniform heat load comes to the results of an experiment. In a non-uniform heat load situation, the weakest part of the flange might start a chain reaction that leads to rupture as a consequence of force transfer between bolts. The weakest part of the flange is the one with the highest temperature and the lowest strength, therefore the highest temperature for all bolts describes the situation where the flange is at its weakest. Applying the highest temperature as the case temperature make the calculations staying on the conservative side. The pretension force in the bolt is created by giving an initial displacement on the bottom side of the bolt. The material properties required in a Finite Element simulation are, in addition to the yield stress, Young's modulus and data for plasto-elastic nonlinear behaviour (hardening), Poisson coefficient, thermal linear expansion, specific heat, conductivity and density. Most of these are dependent on temperature, but few properties are available for temperatures above 400°C. The temperatures of interest for our calculations are above 400°C and especially in the range above 1000°C. Plastic domain Young's modulus describes the linear part of the relation between stress and strain. The linear domain is elastic. That means a point returns to its original position with unloading. When the load exceeds the yield stress, the material enters the non-linear domain, also called the plastic domain. The material in the plastic domain will get a permanent deformation and a point will not return to its original position with unloading. The material will return along an elastic path, but from a permanent deformed position. In the plastic domain the material goes through a process called hardening, the material becomes stronger caused by the permanent deformation. Figure 4 shows stress-strain curves for true stress and engineering stress (called normal stress in the figure) and Figure 5 shows the stress-strain relation during the linear (elastic) and non-linear (plastic) domain as temperature changes. Two effects on the Young's modulus can be seen. Increase of temperature shorten the length of the elastic domain and the Young's modulus is also reduced. The flange models allow for determination of time to leak and leak rate.
Petrell has performed an analysis of the SFP segment at Statoil's processing plant at Kårstø. Scope of work included to determine temperature profiles in the segment during cold blowdown, and to calculate time to rupture in case of fire exposure, and advise on PFP requirements.
Teekay Petrojarl has contracted to Petrell to review passive fire protection on the Knarr FPSO. Work includes review and update of fire scenarios and blowdown.
There is an increasing interest for VessFire in Malaysia but also in other countries in the region. To make sure both current, new and prospective users are up to date on the possibilities of VessFire 1.4 we are now organizing a course in Kuala Lumpur. Dr. Geir Berge of Petrell will lecture on topics such as the physics and process of blowdown, including application of ISO 23251/API521 with a special focus on the issues related to simulation. Topics also include heat transfer, material properties, rupture criteria and the use of VessFire 1.4 including modelling and presentation of results. Program – Day 1 (9 am – 5 pm)
- Registration (from 0830)
- Introduction to the subject
- Blowdown calculation procedure
- Heat load and duration
- Acceptance criteria
- Theoretical basis
- Example on calculation results
- VessFire user interface
- Familiarization with the case
- Clarify the limiting conditions
- Get hold of the technical information
- Set up a matrix of required simulations
- Run the simulations
- Analysis and reporting
There are numerous modifications to offshore oil & gas installations due to e.g. tie-in of new fields and wells. When an existing production system is considered host for a tie-in there are a number of issues that need to be considered; the composition is different, and there is probably also different pressure, temperature and flow rates. A design that once was deemed fit may be severely challenged. Fire loads may also have changed, and corrosion and wear & tear may have taken away capacity. So what do you do if the flare capacity is too low? Doing modifications to the flare system is normally not an option. More or less feasible alternatives include changing the size of the segments by i.e. introducing another segment, replacing weaker piping and flowlines with better materials or larger wall thickness, or adding passive fire protection (PFP). Example of planned modification A modification to a platform located in the North Sea was planned. The modification had impact on several process segments, including 53 flowlines and pipe spools, whereof 19 were exposed to fire. These belonged to four process segments, with operating pressures ranging from 3 to 33 barg. The inventory was a multiphase fluid. Cost saving alternatives for PFP The peak fire load varied from 250 to 350 kW/m2 with a background fire load of 100 kW/m2. The initial analysis, using VessFire, revealed that 9 flowlines and piping spools with Nominal Pipe Size (NPS) (as shown in the table) would rupture. In order to possibly avoid costly PFP, we applied different pipe schedules and wall thicknesses; the heat-stress curves for the steel qualities were already available in VessFire. The results are shown below. With the revised schedule and wall thicknesses there were no unacceptable ruptures and hence no need for PFP. The fire insulation in this example is 40-50 mm cellular glass with a 15 mm alkaline earth silicate (AES) wool and 0.7 mm steel mantling. Huge cost saving potential With an estimated effort of 10 man-hours/meter pipe (including bends and support) for offshore application at 180 USD/mhr, the saving in PFP work was about 475.000 USD. Painting of pipes (if necessary) and scaffolding are not included. Then there is the saving in insulation costs; the results showed an estimated saving in PFP of 4.65 tons at the expense of an additional net weight increase of 3.52 tons of steel due to increased wall thickness. The piping and flowline surface area that would otherwise have been provided with fire insulation was about 263 m2. The additional 8.2 tons of steel represents a minor cost increase. In addition to savings in installation cost and reduced offshore work, benefits include reduced risk of corrosion under the PFP on piping and flowline, reduced cost of inspection and re-application, as well as reduced cost of repair of PFP on piping and flowlines. Multi-discipline tool In addition to documenting blowdown capacity and fire integrity, the analysis also contributed to reduced installation cost and reduced inspection, maintenance and repair costs. This is actually the beauty of VessFire: It is truly a multi-discipline tool where Piping & Layout, Process, Safety and Material disciplines can meet, test and explore consequences of options before they conclude on mitigations. Learn more about the possibilites of VessFire now
Petrell launches version 2.0 of VessFire. Version 2.0 makes dynamic calculation possible for all processes involved in depressurisation. In most situations, this means more precise results and thus even better safety for process systems, with better utilisation of the flare system as a result.
VessFire 2.0 calculates multiphase flow in pipe networks and treats different vessel types. A new and improved interface provides a better description of the process systems. It also allows data stored in Excel to be pasted into the user interface. It is easy and it is fast.
New thermodynamic library
Certain constraints in version 1.0 have been eliminated and new features added. For example: there is no limit to the number of vessels, blowdown valves and pressure relief valves in the process segment. Petrell has also developed a new thermodynamic library (ThermoProp), implemented in version 2.0. ThermoProp handles water/oil mixtures and CO2.
Version 2.0 is more stable and precise. Users familiar with version 1.0 will find major improvements to the user interface in terms of modelling and result presentation.
During the development of the 2.0 version, Petrell identified calculations that took the most time. Improved algorithms, the new thermodynamic library and smarter implementation have resulted in an increased calculation speed.
Dynamic analysis of flare system
Design of process systems for oil and gas has traditionally been based on a stationary approach. This applies to flare system design as well. When gas is discharged into the flare header, the pressure rises and then gradually decreases as the segments are emptied. In flare analyses, a fixed backpressure is traditionally used, because the dynamic pressure profile is not known. VessFire 2.0 calculates backpressure dynamically based on the properties and condition of the gas released into the flare system from various segments. A process system should be dimensioned for actual conditions and what is dynamic should be dealt with dynamically. That's what VessFire 2.0 is designed to do.
Pipes and vessels
Previous versions of VessFire is based on the condition that the vessel is the dominating object in the segment. If the segment contained more than one vessel, the additional vessels were modelled as pipes. In version 2.0, all vessels are recognised as vessels and pipes as pipes. All components constitute a network interconnected by the flow between the components. Segments without vessels (or pipes) are of course permitted. As in version 1.0, users can select steel quality and materials for passive fire protection, and if the material is not listed, new types can be added.
Figure 1 Screen shot showing how pipes and vessels etc. are defined. The list can be export as an Excel file or the list can be prepared in excel and inserted in VessFire using copy/paste.
Figure 2 Screen shot showing a graphical representation of a process segment. Heat load and low temperatures
Heat load and low temperatures
There is no limitation on defining heat loads – peak and background loads can vary over time as before, and if required, a fire load can be allocated for each vessel or pipe.
Version 2.0 provides better support for analysis of low temperatures in connection with e.g. cold depressurisation, through a much broader spectrum of reporting options. For example: temperature fluctuations during a given period for a given pipe section can be extracted.
Licenses available 1Q 2017
Licenses will be made available to existing and new customers in 1Q2017. Until then is VessFire 2.0 available through consultancy services from Petrell.
Fact box on process systems and depressurisation:
The blowdown and flare system is the most important process safety system as their purpose during an accidental event or emergency situation is to reduce the pressure in process segments in order to reduce the risk of rupture and subsequent release of flammable substances that cause fire and explosion. Once overpressure occurs in a process system, getting the pressure down as quickly as possible is important, by diverting the contents to the flare header. Blowdown is a challenging process, even more so in a situation where parts or all of the system are exposed to fire. The blowdown and flare system must handle both scenarios.
Fact box, VessFire:
VessFire is a multi-physics simulation system for blowdown analysis, treating the thermo-mechanical processes interacting with each other. VessFire meets the requirements for advanced depressurisation analysis as outlined in the 2014 edition of API 521 (ISO 23251). Blowdown analysis in VessFire takes into account external heat sources, heat transfer from the outside and through vessel and pipe walls, and through the fluid and gas phases. Evaporation and condensation take place inside vessels and pipes causing the composition of phases to change over time. When a blowdown valve (and/or pressure release valve) opens, mass is removed from the segment, reducing pressure. If the pressure is reduced too slowly in a fire case, rupture can result. VessFire calculates time to rupture. Parameters such as wall thickness, flow orifice diameter, material quality etc. can be varied to see the effect on time to rupture.
Teekay Petrojarl Production is executing modifications of the FPSO Petrojarl 1 for deployment at the Atlanta field offshore Brasil. The Atlanta field is characterised by heavy oil. Petrell has performed a depressurisation and fire integrity analysis of the process systems, by use of VessFire. Work included optimization of depressurization with regards to flare capacity. T
We wish to congratulate Nazanin Jahani with her PhD degree on Coupled Fluid Flow and Elasto-plastic Damage Analysis of Acid Stimulated Chalk Reservoirs. The doctoral thesis has been carried out at Petrell AS and the Department of Engineering Design and Materials at the Norwegian University of Science and Technology (NTNU), under the supervision of Dr. Geir Berge (Petrell AS), Associate Professor Bjørn Haugen (NTNU) and co-supervision of Professor Rune Martin Holt( NTNU and SINTEF Petroleum). An abstract of the thesis: Coupled Fluid Flow and Elasto-plastic Damage Analysis of Acid Stimulated Chalk Reservoirs
BP Norge AS has awarded a contract to Petrell for modelling and simulation of unignited flare (flame out scenarios) at Valhall.