EFFECTS OF HYDROSTATIC PRESSURE ON THE BURST AND COLLAPSE BEHAVIOUR OF MARINE PIPELINES 

ABSTRACT

When a pipeline is submerged through different levels of water depth for use in offshore applications, it experiences the effects of high pressure variations resulting in net external or internal loads. If such loads exceed a critical value, the marine pipeline may undergo permanent deformation or even complete failure, subsequently causing high safety risks and major loss in revenue to all involved. The mechanical behaviour under high net loads acting externally and internally must therefore be studied to understand which adjustments must be made in order to prevent these events from occurring. This work will carry out a numerical and computational study to a) produce theoretical solutions to find effective parameter changes which can be made to prevent undesirable behaviour, b) find the effects of pressure on the mechanical behaviour of pipelines with initial ovalities, validated by c) a simulation using various conditions of the marine pipeline in order to find the maximum values of stress present in the model.

1. Introduction

1.   Background and Literature Review

There is a rapidly ever-increasing demand for more energy around the globe, and due to the continual depletion of easily accessible oil reservoirs, engineers are tempted to expand the reach of ocean facilities into remote locations in far deeper waters than before (Salama et al., 2000). This increases concern to the safety of all personnel involved in operating such equipment and runs the risk of extreme environmental pollution should any failures occur.  Of course, in far deeper waters brings far harsher environments and increasing loads, putting extra strain on subsea equipment. However, there are not only high safety risks to personnel involved, but also detrimental impacts to the company’s revenue should the occurrence of a failure happen. It is therefore a necessity to consider the costs involved in maintaining and repairing damaged equipment and so, focus must be put on preventing these events happening in the first place.

Failures may occur to a pipeline for a number of reasons; primarily high mechanical loading as a result of extreme net pressures acting internally or externally, and environmental conditions such as corrosion, temperature, and water diffusion (for example, Assarar et al., 2011, Pauchard et al., 2002) – the latter being a fairly revised subject due to a relatively consistent spread of conditions throughout currently ongoing offshore operations. However, failure due to extreme loads at deep sea has been considered notably less until recent years where engineers have been making huge leaps in the design of offshore facilities in far deeper waters than before. Therefore most of the material produced in this field has been written many decades ago, before scientists even considered that it were possible to reach water depths in excess of 2500m.

Recent studies have shown that corrosion is the cause of 36% of subsea pipeline failure as opposed to only 13% being to material failure under high loads (Yang et al. 2017). It may be true in saying that corrosion plays a pivotal role in decreasing surface thickness however, high loads must also be simultaneously applied for the pipe surface to break and therefore, equally important to consider.

There have been many cases of previous research into the understanding of mechanical behaviour of pipelines due to hydrostatic pressure. However, many of these reports haven been specific to a defined material. More specifically, ‘The Mechanical Behaviour of Materials Under Pressure’ (Vodar and Kieffer, 1971) was amongst the earliest to perform tests on the effects on metallic materials, and similar tests being held more recently (Wu et al., 2009, Zhang et al., 2017). It is worth noting that the large time gap between studies shows that research into this field is still needed and of interest in the present day. A fair amount of research has also looked into the effects of mechanical loading on composite cylindrical structures (for example, Hine et al., 2005, Davies, 2016), but considerably less on titanium and aluminium – likely due to high economical expenses. Despite such a wide range of knowledge in this field, there have been very few direct comparisons between materials with respect to ocean applications, and this report intends to expose these knowledge gaps.

Studies similar to the one carried out in this work exist publicly but with some uncertainties in the accuracy of the work. One previous study in particular describes a collapse test with ovality in the pipe circumference by modelling a half circumference to represent symmetry on one axis (Kaviti et al., 2013). The geometry however, follows a biconvex shape as opposed to an oval, meaning if the shape were to be reflected it would appear to have two seems on either side, without stating this was intentional. The effect of this would mean a likely increase in maximum stress by a considerable amount and therefore, an improved geometry definition may be needed.

Selecting the correct material to use is one of the most fundamental disciplines in engineering. It determines the reliability of the design in terms of safety and economic viability. So of course, it is vital to have a vast range in knowledge of different materials and an open mind set into exploring new ones. The main considerations are focused on the properties of each material and its ability to withstand various loads. The specification and use of materials which combine high mechanical strength with corrosion resistance and resistance to wear is a fundamental requirement.

An example of insufficient material usage comes from the North Sea, where the Norwegian Petroleum Directorate (NPD) has expressed their concern about recent experiences with respect to material selection and safety. Gas leakages, albeit very small, have been reported to stand at over 220 during the period of 1996 to 2002 (Lange and Berge, 2004).  Many of these leakages come from failures in subsea pipelines through pipe wall fractures, which could possibly have been resolved had a more appropriate material been chosen. Luckily, none of these caused significant damage; however, it still holds the potential for danger to the lives of operating personnel and marine life. This is why material selection is vital. Furthermore, it is estimated that roughly half a billion pounds (including loss of income) was lost due to these failures. Such significant loss of revenue puts high demand on the correct selection of materials.

As it stands, API 5L X65 steel linepipe is the most commonly used steel grade in the offshore industry however, extensive research is looking into the possibilities for using more advanced materials to meet industries demanding needs. In North America and China the X80 steel linepipe is on the rise with approximately 10,000km of offshore pipelines in each continent being used as of 2012, while North America are currently in the process of trialing the X100 linepipe (Sung, 2012). From this evidence it is clear that giving accurate relevant performance comparisons between materials should be given full consideration.

2.   Research Idea, Aim and Objectives

As ocean structures and the marine pipelines which link these to onshore facilities are being pushed to operate in far deeper waters, we must fully understand the effects of increasing hydrostatic pressures to effectively design pipelines in high-pressure situations. Following this, the report aims to focus on the effects of mechanical loading on a variety of selected materials by understanding the changes in mechanical behaviour of marine pipelines caused by extreme hydrostatic and internal pressures. The initial aim is to provide accurate relationships between the most influential parameters which affect the mechanical behaviour of pipelines in order to prevent failure. These results will then be validated by giving a finite element model solution to prove that the values will prevent the pipeline from ever exceeding the maximum allowable operating stress. Finally, the importance of the most influential parameters will be emphasised to show why careful control must be kept in retaining allowable values, such as diameter to thickness ratio and initial imperfections in pipe ovality.

Engineers wishing to design ocean structures and in need of relevant information of material performance  and adjustable parameters for effective design may use this piece of work as a general reference for values such as maximum operating pressure and required diameter to thickness ratio which should be used. Thereby, this work should provide a valuable stepping stone to engineers who wish to better their general understanding in the field and provide as a valuable teaching resource for the way a marine pipeline behaves subject to high mechanical loading whilst varying the most influential factors. This will also aid engineers to understand why consideration must be given in the design process to durability, safety and economic viability. By reducing these knowledge gaps we therefore hope to reduce the risk of incidents in marine pipelines. Furthermore, the model was designed with a single material cylindrical shell therefore, the basic principal of this project may be used for learning purposes when considering a variety of applications such as marine risers and casing and tubing of subsea wells.

3.   Overview of Contents

Collaboration of simplified databases will be given with information on materials which are most commonly used in ocean appliances, such as pipelines/flowlines and marine risers. This will comprise of a selection of 5 different steel grades most commonly found in offshore appliances, namely, API 5L X Grades. Conclusions will be made in the latter stages of the report of the found advantages over one another in terms of resistance to high net pressures, safety and economic viability.

This will be followed by calculations to estimate the magnitude of forces which pipelines undergo in an offshore environment; primarily being hydrostatic pressure. This will allow us to understand the pressure variations at each level of water depth and therefore, the variation of forces imposed on pipelines.

The next section will include wall thickness calculations subject to imposing pressures only at water level. This will be enough to calculate the maximum thickness for burst, however, to estimate thicknesses to prevent collapse, the pipe will be subject to a variation of high hydrostatic pressures as water depth increases. By initially keeping the thickness constant as it varies through water depth, we can observe when the collapse pressure safety check is not satisfied and the adjustments we need to make to the wall thickness. Furthermore, these results will be validated using the ANSYS Mechanical 16.0 software.

The adjustments needed to be made to prevent collapsing of the pipe with initial imperfections will then be examined, which will be done with imposing hydrostatic pressure through a variation of ovality percentages. This will allow conclusions to be made on the acceptability of allowable ovality in the pipe cross-section and the acceptable water depth at which a pipe with said ovality may operate. These results will also be validated by using the ANSYS Mechanical 16.0 Software.

2. Theoretical Solutions to Prevent Failure

2.1  Materials for Pipeline Design

The focus of material selection will be associated for use with offshore line pipes – a high strength high yield carbon steel pipe used for transporting crude oil, natural gas and water.

The main factors which influence the selection of materials for this type of application are: the physical and mechanical properties of the material; its resistance to corrosion against the environment and a wide range of operating conditions; its resistance to marine biofouling; and its ability to perform fabrication operations – all of which contribute in some form to the safety aspects of using such materials and to economic considerations. The main focus of the content in this report will focus on the physical and mechanical properties of the material with respect to how they change under high loading, the rest of these factors are considered to be outside the scope of work in this project.

Line pipes make up the vast majority of the market for offshore pipelines – most commonly fixed to the seabed. There are many pieline standard codes used around the world, with the most common ones being, American Natl. Standards Inst. (ANSI), American Soc. of Mechanical Enginers (ASME) and Det Norske Veritas, Norway, and Germanischer Lloyd, Germany (DNV GL) however, the majority of codes used in this report will be under regulations specified by the American Petroleum Institute (API) which defines the API Specification 5L standard used worldwide. Pipes can be either welded or seamless, however, the great advantage of the seamless pipe is its ability to withstand pressure as opposed to a welded pipe which have considerable weaknesses along the welded seam. For this prime reason, we will consider seamless pipes to give a superior performance and will be used in the following methods.

The thermophysical properties of the materials are neglected with the assumption being made that the ductile to brittle transition temperature will never be reached (roughly -80ºC) and that steels can withstand temperatures in excess of 300ºC without being affected, therefore, in this instance temperature is outwith the scope of this work. Table 2.1 presents mechanical properties of a selection of the most common materials used in line pipes as specified by API.

Table 2.1 – Steel grade and corresponding yield and tensile strengths.

The following methods will use values associated with steel grade API 5L X65, however, values can be easily interchangeable and are given in Table 2.1 for a fuller range of materials which will also be analysed theoretically.

2.   Internal and External Pressure

As the subsea pipeline is submerged in water to deep sea conditions, it naturally experiences continual variations in external pressure, due to the pressure within the liquid, the acceleration due to gravity and the depth within the liquid, as stated by Pascal’s Principle. If the internal pressure is kept at constant throughout, either being atmospheric pressure (empty) or operating pressure, then net pressures will vary linearly with water depth. The combination of high internal pressure and low external pressure will result in an externally acting force – which if large enough will cause the pipe to burst. Similarly, the combination of low internal pressure and high external pressure will result in an internally acting force – which if large enough will cause the pipe to collapse. This is the basis of the mechanical behaviour which we shall be looking at. It is worth noting that the internal pressures of a pipe can be controlled and set manually to an extent; whereas the prime source of external pressure is hydrostatic which occurs naturally as stated previously. For safety measures, we shall consider the unusual case where internal pressure cannot be carefully controlled and will be kept unpressurised in the following collapse tests to represent the most extreme case occurring.

3.   Hydrostatic Pressure

Hydrostatic pressures experienced at deep sea can have detrimental effects on the mechanical behaviour of pipes. The classical form of Pascal’s Principle may be used to give sufficient estimations of imposing pressures formed at extreme water depths, carrying forward the assumption that the fluid is at rest and is incompressible. Such assumptions may give inaccuracies in exact water pressures as these are rarely the case however, greater accuracy is not needed to obtain relevant conclusions in this study. Pascal’s Principle states that the pressure within a liquid depends on the density of the liquid, the acceleration due to gravity, and the depth within the liquid.

PExternal=PAtmoshpere+PFluid

(1)

PFluid=ρgh

(2)

Where, atmospheric pressure is

101kPa, the density of seawater is

1.025 g/cm3and the standard local gravity is

9.80665 m/s2. Water depth,

h, is varied from the water level to

3000mto accommodate applications ranging from shallow to the deepest of waters. The external pressure will be represented by

PO. The following external pressures were achieved in order to represent the linearly external loads experienced by subsea pipelines, as seen in Table 2.2.

Table 2.2 – External pressure variation with water depth (Atmospheric and Hydrostatic).

4.   Pipe Wall Thickness

The main parameters which strongly influence the behaviour of thick-walled pipes are material properties, net pressure, the outer diameter to thickness ratio,

DO/t, and imperfections, such as the initial ovality.  Each of the mentioned parameters will be considered as variables to combat the effect of excessive net pressures acting on a cylindrical shell as it is submerged through a range of water depths.

Initially, the outer diameter to thickness ratio,

DO/t, will be calculated suited for a pipe which lies above sea level and kept constant as the pipe experiences the effects of submergence in sea water. Thereby, allowing us to understand the adjustments needed to be made to repel deformation of the shell as the external pressure ranges through deep sea imposing high loads. Wall thickness calculations can be made in accordance with the ANSI/ASME Standard B31.3 code – an extension of Barlow’s Formula to allow for high safety margins, such as corrosion allowance,

te, given as

0.079in(2mm) as a general guideline for steel pipes carrying crude oil, and a value of

0.05in(1.27mm) is used for the thread or groove depth,

th, in accordance with the pipe nominal size. We will also use an outside diameter,

DO, of

10in(254mm) for the test pipe and therefore, the manufacturer’s allowable tolerance,

ttol, is

12.5%for API 5l steel pipes up to

20in(508mm) diameter, giving us the following expression for minimum design wall thickness

t=te+th+PIDO2SE+PIY100100-Ttol

(3)

Where, longitudinal weld-joint factor,

E, is set at a value of

1.0for a seamless pipe and a derating factor,

Y, is set at 0.4 for ferrous materials operating below

900oF. The maximum allowable internal pressure in the pipe,

PI, is set at a constant initial value of 2500psi (17,237kPa).

And

Sis the allowable stress for the pipe, taken as the Specified Minimum Yield Strength (SMYS) of the material which is the limit of elastic deformation.

5.   Burst and Collapse Pressure

Theoretical solutions are proposed for the elastic burst and collapse pressures to prevent yielding of a cylindrical shell exposed to uniform internal and external pressures. It is noted that the main cause for failures to occur be that the internal pressure exceeds external pressure, or vice versa, by a substantial amount given by the product of some factor,

f0, giving way for a safety allowance, and the critical pressure to cause collapsing or bursting of the pipe.

It should be noted that in this report, the burst and collapse term is used to differentiate between the yielding pressure in that relative motion. For example, the pipe will not completely burst or collapse until the ultimate tensile stress has been reached, however, values of maximum stress used in the following methods will relate directly to the value of yielding stress. Beyond this point the model will begin to plastically deform and so we will deem it to have failed and therefore the maximum allowable stress is the point of yielding. Should we achieve resultant values greater than the point of yield, previous studies (Toscano, 2009) state that disregarding the steel work-hardening of a material in the plastic transition zone has a neglible effect on the burst and collapse pressures and therefore, the same equation used to calculate stress values in the perfectly elastic zone will be assumed to also give accurate results when determining failure.

And so, in accordance with API-1111 the critical failure pressure of the pipe must exceed the net pressure everywhere along the pipe and is expressed as follows

PI-PO≤f0 Pbr

(4)

PO-PI≤f0 Pc

(5)

Where,

PIis the internal pressure of the pipe which will range from atmospheric pressure (depressurised) to a maximum operating pressure of

2500psi. For safety measures, collapse pressure checks will be made with respect to the condition where the pipe is empty (unpressurised), and burst pressure checks to the maximum operating condition.

Pbrand

Pcare burst pressure and collapse pressure of the pipe, respectively. All values of pressure used must therefore be given as positive to comply with the previous expressions.

We can calculate the ultimate burst pressure,

Pbr, by using the following formula in accordance with API-1111 as follows

Pbr=0.45 S+Uln⁡DODI

(6)

Where,

Sis the specified minimum yield strength and

Uis the specified minimum ultimate tensile strength which is given in Table 2.1, and

DI=DO-2t. This relationship is needed as pipe burst does not occur precisely at the yield or ultimate tensile strength.

Figure 2.1 shows an offshore pipeline pipe wall fracture due to bursting motion. This was achieved with extreme internal pressures causing the pipe material to reach the ultimate tensile stress, causing complete failure, as opposed to yield stress which would only result in permanent deformation.

Figure 2.1: Pipe Burst due to Large Internal Pressure.

Source: Gajdoš and Šperl (2012) Figure 16 ‘Evaluating the Integrity of Pressure Pipelines by Fracture Mechanics’ Academy of Sciences of the Czech Republic, Czech Republic.

It must be noted, bursting of a pipeline is entirely a problem of an internal pressure which creates a stress in exceedance of the maximum stresses allowed. Once calculating the wall thickness to diameter ratio which will achieve 72% SMYS with respect to the applied internal pressure, it can be assumed that the pipeline will never veer close to bursting. Instead, this work shall aim to optomise the maximum allowed internal pressure by achieving equilibrium with the opposing hydrostatic pressure as the water depth is increased.

Relating back to equation (4), we can substitute in the burst pressure value while varying the external pressure accordingly with Table 2.2 and keeping the internal pressure constant as the maximum operating pressure throughout. We are reminded that equation (4) states the following:

PI-PO≤f0 Pbr

Where,

f0represents a series of factors for example, the temperature derating factor,

ftas specified by API-1111 as 1.0, the longitudinal weld joint factor,

fealso being 1.0 for a seamless pipe, and the internal pressure (burst) design factor,

fdbeing 0.9 for pipelines. According to most safety standards around the world, this safety factor of design stress should stand at around 72% of the specified minimum yield strength; however, since the burst formulation is given with respect to the ultimate yield strength,

f0will stand at a value of around 65%.

Since now knowing all factors in equation (4), we can observe if this is satisfied for all water depths and variations in external loads by checking that the hydrostatic test pressure (internal minus external pressure) never exceeds the product of the safety factor and the burst pressure. The values in Table 2.3 represent the required wall thickness at the maximum value of

PI-PO,being 2,500psi (17,237kPa), and if the following ANSYS validation proves equation (4) to be satisfied, then we shall assume the equation will also be satisfied for all values of water depth.