Ocean Engineering 269 (2023) 113720

Contents lists available at ScienceDirect

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

Mechanical interference properties of two parallel strings in a deepwater

dual-derrick drilling system
Xiuquan Liu *, Yan Shi , Yanwei Li , Zhaowei Liu , Na Qiu , Yuanjiang Chang , Guoming Chen
Centre for Offshore Engineering and Safety Technology, China University of Petroleum, Qingdao, 266580, China

A R T I C L E I N F O A B S T R A C T

Handling Editor: Prof. A.I. Incecik Two parallel strings in a deepwater dual-derrick drilling system are likely to interfere with each other. The
mechanical interference properties of the two parallel strings have not been systematically studied. A mechanical
Keywords: interference model for parallel strings considering current, wave, platform motion, and the interference effect
Parallel strings between two strings is established on the basis of Bernoulli beam theory and wake models. An interference
Mechanical interference properties
analysis method is also proposed on the basis of the finite element method, Newton’s method, and Newmark-β
Drilling riser
method. Lastly, a mechanical interference analysis program is developed in MATLAB and applied in the static
Xmas tree installation
Wake model and dynamic interference property analysis of two parallel strings. It turns out that the wake effect of the riser is
significant and adsorbs the drill pipe around the riser. The adsorption effect may cause a collision between two
parallel strings. The collision risk of parallel strings exists at the end of the hanging string and increases with the
hanging water depth. It is also found that the collision between two strings in dynamic analysis decreases when
the drill pipe is in upstream due to combined wave and current loads but increases when the drill pipe is in
downstream due to the adsorption effect.

1. Introduction scholars have conducted mechanical analysis of strings during subsea

equipment installation (de Andrade et al., 2023). Wang et al. (2015a,
Dual derricks have been widely used in onshore and offshore drilling 2015b) studied the mechanical properties of drilling risers in installation
due to their high operation efficiency (Munch-Søegaard and Nergaard, operation and obtained the safe operating window that can guide riser
2001). Increasing deepwater drilling platforms have also been equipped installation. Wang et al. (2018) analyzed the effects of current velocity,
with dual derricks in recent years. The dual-derrick system is used in wave height, and manifold weight on the mechanical properties of
parallel making and breaking bottomhole assemblies and especially in subsea manifold installation by using OrcaFlex software. Hu et al.
running subsea equipment along with deepwater drilling. The running (2018) developed a novel analytical model of a variable-length marine
string and deepwater drilling riser are arranged in parallel in subsea string to investigate the effect of lowering speed on Xmas tree (XT)
equipment running operation. The two parallel strings are likely to installation. Qin et al. (2019) established a mechanical model to analyze
interfere with each other due to the limited distance between the two the stress and deformation of a drill pipe during the landing of a subsea
derricks and the lateral deflection of the strings under current and wave. XT. de Araújo Neto et al. (2019) presented a methodology for the design
The string interference may lead to a collision between the strings and and evaluation of an installation method based on the combination of
even the failure of the drilling system (Zhu et al., 2019). Therefore, the steel wire and polyester ropes. The goal was to avoid resonance effects
mechanical interference properties of two parallel strings during subsea along all phases of the lowering procedure. Oh et al. (2020) developed a
equipment installation in a deepwater dual-derrick system should be simulation based design framework to design the shape of steel lazy
investigated. wave riser in ultra deepwater depth. Tommasini et al. (2021) analyzed
The mechanical interference model of two parallel strings during dynamics of deepwater subsea lifting operations considering hydrody­
subsea equipment installation can be divided into two key parts, which namic coefficients that depend on the Keulegan-Carpenter (KC) number.
are the mechanical analysis model of the strings during the subsea Wang and Gao (2021) studied the tubular mechanical characteristics to
equipment installation and the wake model between the strings. Some ensure successful implementation of surface casing installation. Liu et al.

* Corresponding author.
E-mail address: [email protected] (X. Liu).

https://doi.org/10.1016/j.oceaneng.2023.113720
Received 3 October 2022; Received in revised form 9 January 2023; Accepted 13 January 2023
Available online 17 January 2023
0029-8018/© 2023 Elsevier Ltd. All rights reserved.
X. Liu et al. Ocean Engineering 269 (2023) 113720

interference. Gu et al. (2022) investigated the effects of initial spacing,

top tension factor, and reduced velocity on the collision probability of
parallel risers on the basis of Fu’s hypothesis. Previous studies focused
on the static interference properties and dynamic collision probability of
parallel strings. Studies on the dynamic interference properties of par­
allel strings are lacking. Moreover, the interference properties of parallel
strings in a dual-derrick system have not been systematically studied.
Knowledge gaps in the static and dynamic interference properties of two
parallel strings in a deepwater dual-derrick drilling system remain.
The remainder of this paper is organized as follows. Section 2 es­
tablishes a mechanical interference model of parallel strings. Section 3
proposes an interference analysis method. Section 4 carries out an
interference analysis of parallel strings during subsea installation oper­
ation in a dual-derrick drilling system. Section 5 provides conclusions.

2. Interference model

A typical structural sketch of two parallel strings in a deepwater

dual-derrick system is shown in Fig. 1. A drill pipe, which is the first
string, is attached to Derrick 1 for lowering the XT. The top of the drill
pipe is fixed to the platform, and the bottom of the drill pipe is attached
to the XT that hangs freely. The riser, which is the second string, is
attached to Derrick 2 for recovery of the LMRP/BOPs. The top of the
riser connects to the platform through the upper flex joint (UFJ), and the
bottom of the riser is connected to the LMRP/BOPs through the lower
flex joint (LFJ). The two parallel strings deflect and vibrate laterally
under the platform’s lateral motion and hydrodynamic loads. Moreover,
the hydrodynamic load on the downstream string (DS) is influenced by
the wake of the upstream string (US). A mechanical interference model
of the two parallel strings will be established to analyze their interfer­
ence properties in the following part.
Fig. 1. A Dual-derrick system with parallel strings. The two parallel strings are assumed to be a homogeneous, isotropic,
and linear elastic steel tube with a constant cross section. In consider­
(2022) developed a novel axial dynamic model for drilling riser that uses ation of the damping force, axial force, and external loads, the differ­
nonlinear distributed damping instead of conventional uniform damp­ ential control equations for the US and DS based on Euler–Bernoulli
ing. Adamiec-Wójcik et al. (2023) defined optimal base motion to beam theory are shown in Eqs. (1) and (2) (Zhou et al., 2020; Xu et al.,
compensate for the influence of sea currents during riser re-entry based 2022), respectively.
on segment method. The aforementioned studies did not consider the ( ) ( )
∂2 ∂2 x ∂2 x ∂ ∂x ∂x
interference between strings. In the aspect of interference models be­ EI1 (z) 2 + m1 2 − T1 (z) + c1 = f1 (z, t) (1)
∂z 2 ∂z ∂t ∂z ∂z ∂t
tween strings, some wake models have been proposed for interference
analysis of parallel strings. Schlichting (1961) proposed a wake analysis ( ) ( )
∂2 ∂2 x ∂2 x ∂ ∂x ∂x
model for cylinders, and the model was validated for far wake investi­ EI2 (z) 2 + m2 2 − T (z) + c2 = f2 (z, t) (2)
∂z 2 ∂z ∂t ∂z 2 ∂z ∂t
gation. Huse (1992) introduced the concept of “virtual source” based on
the Schlichting wake model and proposed the Huse wake model. Blevins
where subscripts 1 and 2 represent the US and DS, respectively; x is the
(2005) proposed the Blevins wake model to calculate the fluid forces on
lateral displacement of the string; z is the vertical coordinate whose
a cylinder in the wake of a parallel upstream cylinder. The Huse wake
positive direction coincides with the direction of gravity; E is the
model and the Blevins wake model remain the dominant wake models.
modulus of elasticity of the string; I is the area moment of inertia of the
In recent years, researchers have conducted parallel string analysis
string cross section; T(z) is the axial force in the string; m is the mass per
by combining mechanical analysis models of strings and wake models.
unit length of the string; f(z, t) is the lateral environmental load per unit
Sagatun et al. (2002) simulated the dynamic interaction of two adjacent
length of the string; c is the lateral damping coefficient of the string.
cylindrical risers moving relative to each other in an ambient steady
Two states of strings occur in the dual-derrick system, which are the
flow. Rustad et al. (2008) analyzed a pair of risers in a tension leg
hanging state of the drill pipe and the connected state of the riser, as
platform based on finite element method and controlled the top tension
shown in Fig. 1. The axial forces of the hanging string and connected
of risers to avoid collision. He and Low (2012, 2014) estimated the
string are described by Eqs. (3) and (4), respectively.
probability of collision between marine risers in consideration of hy­
drodynamic interference. Fu et al. (2017, 2018, 2021) focused on the T(z) = − wg(l − z) − MB g (3)
collision between flexible risers induced by the wake effect and pro­
posed a prediction model for string collision failure probability based on T(z) = Ttop − wg(l − z) − MB g (4)
the assumption that the wake model can be applied with wave. Zhu et al.
(2019) developed a static interference model for parallel risers in a where w is the linear density of strings, MB is the mass of the suspended
dual-derrick system and investigated the effect of current velocity on object, and Ttop is a force applied by the platform to the top of the string.
drilling operations. Tian et al. (2020a) used the modified Huse model to The lateral load on the US includes wave load and current load. The
study the interference between risers on a tension leg platform. Tian lateral load per unit length of the US is often calculated in accordance
et al. (2020b) proposed a new wake model and investigated the effects of with the modified Morrison equation (Fan et al., 2017; Yin et al., 2019),
riser spacing, top tension force, and current velocity on riser which can be expressed as

2
X. Liu et al. Ocean Engineering 269 (2023) 113720

Fig. 3. Blevins wake model.

Fig. 2. Huse wake model for local current velocity calculation.
Huse wake model and the modified Morrison equation, which can be
[ ]
πρw D21 ∂2 x expressed as
f1 (z, t) = CM1 u̇w (z, t) − (CM1 − 1) 2
4 ∂t [ ] ⃒ ⃒( )
⃒ ⃒( ) πρ D2 ∂2 x ρ D2 CD2 ⃒⃒ ∂x⃒⃒ ∂x
ρw D1 CD1 ⃒⃒ ∂x⃒⃒ ∂x f2 (z, t) = w 2 CM2 u̇w (z, t) − (CM2 − 1) 2 + w ⃒ u − ⃒ u−
+ ⃒u + u − u + u − (5) 4 ∂t 2 ∂t ∂t
2
w c
∂t ⃒ w c
∂t
(10)
where CM is the inertia coefficient, CD is the drag coefficient, D is the Blevins (2005) stated that the resistance of the string in the wake is
hydrodynamic diameter, u̇w (z, t) is the water particle acceleration in the proportional to the dynamic pressure at its center, as shown in Fig. 3.
horizontal direction induced by wave, uw is the water particle velocity in The drag coefficient of the DS is expressed as
the horizontal direction induced by wave, and uc is the steady current
velocity. ĊD2 (x, y) = CD2 ⋅[u(x, y) / U0 ] (11)
The current velocity at different water depths (WDs) can be
when Eq. (7) is substituted into Eq. (11), the corrected drag force co­
computed using Eq. (6) if no measured data of current are available
efficient can be written as
(Fang and Duan, 2014).
{ (√̅̅̅̅̅̅̅̅̅̅̅̅̅) ( )}2
( )17 ( ) CD1 D1 − a2 y2
h− z h− z ĊD2 (x, y) = CD2 ⋅ 1 − a1 exp (12)
uc (z) = ut ⋅ + uw ⋅ (6) x CD1 D1 x
h h

where uc(z) is the current velocity at different WDs, ut is the tidal current where the coefficients a1 and a2 are 1 and 4.5, respectively.
velocity, uw is the wind current velocity, and h is the WD. The RMS of the corrected drag force coefficient is used as the
Unlike the hydrodynamic load of the US, that of the DS needs to equivalent drag force coefficient of the DS, which can be expressed as
consider the wake of the US. The Huse wake model and the Blevins wake √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
∫
model are commonly used to correct the hydrodynamic load of the DS. 1 R2 2
CD2 = Ċ (x, y)dy (13)
Huse (1992) introduced the concept of a “virtual source” to estimate D2 − R2 D2
the local current velocity in the wake field, as shown in Fig. 2. The The lateral load per unit length of the DS is derived by combining the
current velocity at any position in the two-dimensional wake field can be Blevins wake model and the modified Morrison equation, which can be
described as expressed as
⎧ √̅̅̅̅̅̅̅̅̅̅̅̅̅ )
⎪ CD1 D1 [ ]
⎪
⎪
⎪ u = u − u exp ( − 0.639(y/b)2 πρ D2 ∂2 x
⎨ c c
xs f2 (z, t) = w 2 CM2 u̇w (z, t) − (CM2 − 1) 2
(7)
4 ∂t
⎪ xs = xv + x ⃒ ⃒( )
⎪
⎪ ⃒
ρ D2 CD2 ⃒ ∂x⃒⃒ ∂x
⎪
⎩ + w ⃒ u + u − u + u − (14)
xv = 4D1 /CD1 2
w c
∂t ⃒ w c
∂t

where uc is the current velocity, xs is the distance from the wake source The boundary conditions at the top of the hanging string are
to the center of the circle of the DS, u is the local current velocity of the expressed as
DS, y is the longitudinal distance of the strings, xv is the distance from ⎧
⎪
⎨ x(0, t) = Sp (t)
the virtual source to the US, and b satisfies the following equation:
dx(0, t) (15)
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ ⎪
⎩ θ(0, t) = =0
b = 0.25 CD1 D1 xs (8) dz

The root mean square (RMS) of the local current velocity is used as The boundary conditions at the top of the connected string are shown
the equivalent current velocity of the DS, which can be written as in Eq. (16), and the boundary condition at the bottom is shown in Eq.
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ (17).
∫
1 R2 2 ⎧
u= u (xr , yr )dy (9) ⎪
⎪ x(0, t) = Sp (t)
D2 − R2 ⎪
⎪
⎪
⎨
dx(0, t)
(16)
where the coordinates (xr , yr ) describe the spatial location of the outer ⎪
⎪
M(0, t) = Kuf
dz
⎪
diameter semicircle of the DS, and R2 is the radius of the DS. ⎪
⎪
⎩ T(0) = Ttop
The lateral load per unit length of the DS is derived by combining the

3
X. Liu et al. Ocean Engineering 269 (2023) 113720

Fig. 4. Interference analysis process of the parallel strings.

⎧ strings and is also introduced to conduct mechanical analysis of parallel

⎨ x(L, t) = 0
⎪
strings in this study (Liu et al., 2020; Qin et al., 2022; Kim et al., 2021).
(17)
⎪
⎩ θ(L, t) =
dx(L, t)
=0 The weighted residual method is used to discretize the established Eqs.
dz (1) and (2). The mechanical analysis model of the two parallel strings is
then expressed as
where Sp(t) is the displacement of the platform, M is the bending
moment of the string, θ is the deflection angle of the string, Kuf is the M1 δ̈1 (t) + C1 δ̇1 (t) + K 1 δ1 (t) = F1 (t) (18)
rotation stiffness of the UFJ.
M2 δ̈2 (t) + C2 δ̇2 (t) + K 2 δ2 (t) = F2 (t) (19)
3. Interference analysis method
where δ(t) represents the nodal displacement vectors of the string; F(t) is
The finite element method is often chosen for mechanical analysis of the nodal force vector of the string; M, C, and K are the mass matrix,

4
X. Liu et al. Ocean Engineering 269 (2023) 113720

Fig. 5. Correctness verification of the self-developed MATLAB program.

damping matrix, and stiffness matrix of the string, respectively. parallel strings and the parameters of waves and currents are first
A static interference analysis model of the parallel strings is deter­ defined. The static offset of the two parallel strings is then calculated in
mined by neglecting the velocity and acceleration terms of the dynamic accordance with the established static interference analysis model
interference analysis model, which can be expressed as shown in Eqs. (20) and (21). Finally, dynamic analysis of the parallel
strings in the dual-derrick system can be conducted on the basis of the
K 1 δ 1 = F1 (20)
established dynamic interference analysis model shown in Eqs. (18) and
(19). Moreover, the hydrodynamics of the DS needs to be updated timely
K 2 δ 2 = F2 (21)
in accordance with the wake model in the dynamic interference analysis.
The Newton–Raphson iterative method (Connaire et al., 2015; Wang The correctness of the self-developed MATLAB program should be
et al., 2021) is an accurate and fast one-dimensional root calculation verified before interference analysis of the two parallel strings. The
method that is often used in static numerical analysis. Therefore, this verification is done using the finite element analysis software ABAQUS.
method is selected to solve the static interference analysis model shown A 1000 m string with an outer diameter of 0.533 m and an inner
in Eqs. (20) and (21). The specific solution steps are shown in Appendix diameter of 0.483 m is chosen for analysis. The wind current velocity is
A. 0.7 m/s. The wave period and wave height are 12.1 s and 3.9 m,
The Newmark-β method is a stable time-stepping solution method respectively. The sinusoidal motion amplitude and period of platform in
that can accurately and efficiently carry out time-domain dynamic the horizontal direction are 2m and 12.1s, respectively. The static and
analysis (Liu et al., 2021; Han et al., 2022). Therefore, the Newmark-β dynamic responses of the 1000 m string are analyzed in the self-
method is selected to solve the dynamic interference analysis model developed MATLAB program and ABAQUS, respectively, and then
shown in Eqs. (18) and (19). The specific solution steps are shown in compared with each other. The contrast results show that the result is
Appendix B. basically consistent, as shown in Fig. 5. Besides, the computational ef­
A MATLAB program is self-developed in accordance with the inter­ ficiency of the self-developed program and ABAQUS is compared. The
ference analysis method shown in Fig. 4. The structural configuration of dynamic analysis time is set to 200s. The self-developed MATLAB

Fig. 6. Three XT installation modes.

5
X. Liu et al. Ocean Engineering 269 (2023) 113720

Table 1
Dual-derrick system parameters.
Parameters Values Units

Height of the derrick from the platform 70 m

Platform height from the water surface 20 m
Center distance of the derrick 15.7 m

Table 2
LMRP/BOP system parameters.
Parameters Values Units Parameters Values Units

Riser outer diameter 0.533 m Riser wet weight 571 kg

Riser inner diameter 0.4822 m Riser density 7850 kg/
m3
Riser wall thickness 0.0254 m LMRP/BOP 1 m
outside
diameter
Riser hydrodynamic 1.37 m LMRP/BOP dry 400,000 kg
outside diameter weight
Riser dry weight 20,442 kg LMRP/BOP wet 340,000 kg
weight

Table 3
XT system parameters.
Parameters Values Units Parameters Values Units

Drill pipe outer 0.1397 m XT outside 4 m

diameter diameter
Drill pipe inner 0.12136 m XT dry weight 45,000 kg
diameter Fig. 7. Lateral displacement of strings under different wake models when the
Drill pipe wall 0.00917 m XT wet weight 36,000 kg riser is in upstream.
thickness
Drill pipe density 7850 kg/
m3

Table 4
Environmental parameters.
Parameters Values Units Parameters Values Units

Seawater density 1025 kg/ Tidal current 0 m/s

m3 velocity
Drag force 1.2 Wind current 0.58 m/s
coefficient velocity
Inertia coefficient 2.0 Wave height 6 m
WD 1510 m Wave period 11.2 s

program takes 2.21s and 125.14s to solve static and dynamic models,
respectively. ABAQUS takes 24.44s and 895.16s to solve static and dy­
namic models, respectively. The computational efficiency of the self-
developed program is higher than that of ABAQUS.

4. Case study

The XT installation, including three operation modes, which are

hanging drill pipe & connected riser (Operation model 1), connected
drill pipe & hanging riser (Operation model 2), hanging drill pipe &
hanging riser (Operation model 3), is a typical subsea operation with
parallel strings in deepwater dual-derrick systems, as shown in Fig. 6.
The XT installation operation is chosen as the object to study the parallel
string interference properties in the deepwater dual-derrick system. The
structural parameters of the dual-derrick system, LMRP/BOP system,
and XT system are listed in Tables 1–3, respectively. The marine envi­
ronmental parameters are listed in Table 4, where wave is simulated Fig. 8. Lateral displacement of strings under different wake models when the
according to Airy wave theory. drill pipe is in upstream.

6
X. Liu et al. Ocean Engineering 269 (2023) 113720

Table 5 riser is located upstream, as shown in Fig. 7. Moreover, the effect of the
Working conditions of Operation modes 1 and 2. Blevins wake on the lateral deformation of the DS is greater than that of
Influencing factors the Huse wake, which is consistent with the findings of Tian et al.
(2020b), as shown in Figs. 7 and 8. The main reason is related to the
Operation mode 1-1 Drill pipe located upstream
Operation mode 1–2 Drill pipe located downstream “virtual source” in the Huse wake model, which increases the calculated
Operation mode 2-1 Drill pipe located upstream distance of the wake between the two strings and reduces the wake flow
Operation mode 2-2 Drill pipe located downstream decay. Conservatively, the Blevins wake model is chosen for interference
analysis of parallel strings in this study.
4.1. Wake model analysis
4.2. Static analysis
The Huse wake model and the Blevins wake model are the dominant
wake models to simulate the interference effect between parallel strings. The interference properties of the parallel strings in Operation modes
A static interference analysis of parallel strings is performed with no 1 and 2 are similar because they both have hanging string and connected
wake, the Huse wake model, and the Blevins wake model to select a string. Static interference analysis of parallel strings at different hanging
suitable wake model for the parallel string interference analysis. The WDs is performed for Operation modes 1 and 2, as listed in Table 5. The
lateral deformation of the US and DS in different current directions state of the US has a significant effect on the interference properties of
under Operation mode 3 is shown in Figs. 7 and 8, respectively. It turns the two parallel strings. There is no risk of collision between the two
out that collision between the strings is difficult when the wake is not strings when the US is a connected string, as shown in Figs. 9(b) and 10
considered. However, the lateral deformation of the DS decreases under (a). On the contrary, there is a risk of collision between the two strings at
the wake effect of the US, while the lateral deformation of the US is not the end of the US when the US is a hanging string. The collision risk
affected when the wake is considered. The distance between the two increases with hanging WD, as shown in Figs. 9(a) and 10(b). The main
strings decreases due to the wake effect. The decreased distance leads to reason is that the lateral stiffness of the connected string is large
an increase in collision risk between the two strings, especially when the compared with that of the hanging string due to different bottom

Fig. 9. Static lateral displacement of parallel strings in Operation mode 1.

Fig. 10. Static lateral displacement of parallel strings in Operation mode 2.

7
X. Liu et al. Ocean Engineering 269 (2023) 113720

Table 6 have an adsorption effect on the downstream drill pipe. The adsorption
Working conditions of Operation mode 3. effect of the riser causes the drill pipe to move close to the upstream riser
Influencing factors gradually as the hanging depth increases, which changes the response
pattern of the downstream drill pipe and increases the risk of collision
Operation mode 3-1 Riser suspension length
Operation mode 3-2 Drill pipe suspension length between the parallel strings. Moreover, the collision of the parallel
Operation mode 3-3 Two strings with the same suspension depth strings is likely to occur at the end of the hanging riser, which is the same
as that in Operation modes 1 and 2.

boundary conditions. The distance between the two parallel strings is

large and the collision probability is small when the connected string is 4.3. Dynamic analysis
in upstream.
In Operation mode 3, both parallel strings are hanging strings, and The parallel strings in the deepwater dual-derrick system are subject
the strings are more flexible and unstable than those in the other oper­ not only to static current load but also to dynamic wave load and plat­
ation modes (Fan et al., 2017), resulting in more complex interference form motion. Dynamic interference analysis must be carried out for
properties of the parallel strings. Static interference analysis of parallel parallel strings. The working conditions for the dynamic interference
strings at different hanging WDs is performed for Operation mode 3, as
listed in Table 6. The wake effect of the riser is extremely pronounced in Table 7
Operation mode 3 compared with that in Operation modes 1 and 2. An Working conditions for dynamic interference analysis.
approximately 86% reduction in lateral displacement of the downstream
Operation modes Working condition
drill pipe is observed, as shown in Fig. 11(a). The reason is that the wake
Operation mode 1 Drill pipe suspension depth 900 m
effect is related to both the initial distance between the parallel strings
Operation mode 2 Riser suspension depth 900 m
and the hydrodynamic outer diameter of the US according to the wake Operation mode 3-3 Two strings suspended simultaneously 900 m
model. The significant wake effect of the riser also causes the riser to

Fig. 11. Static lateral displacement of parallel strings in Operation mode 3.

8
X. Liu et al. Ocean Engineering 269 (2023) 113720

upstream due to its small hydrodynamic diameter. The combined wave

and current loads on the upstream drill pipe and the downstream riser in
the dynamic analysis increase compared with static current loads,
especially on the downstream riser due to its large hydrodynamic
diameter. The dynamic lateral deformation of the downstream riser is
larger than that of the upstream drill pipe. Therefore, the dynamic DBS is
larger than the static DBS when the drill pipe is in upstream, and the
collision between strings decreases, as shown in Fig. 13. However, the
wake effect is obvious when the riser is in upstream due to its large
hydrodynamic diameter. In this condition, the lateral deformation of the
upstream riser increases due to the combined wave and current load,
while the lateral deformation of the downstream drill pipe almost does
not change due to the adsorption effect. Thus, the dynamic DBS is
smaller than the static DBS when the riser is in upstream, and the
collision between strings increases, as shown in Fig. 14.

5. Conclusions

This study addresses a challenging problem: the mechanical inter­

ference properties of two parallel strings in a dual-derrick system. A
mechanical model for parallel strings considering current, wave, and
platform motion is established on the basis of Bernoulli beam theory.
Fig. 12. RAOs of the platform. The interference effect between two strings is simulated on the basis of
the Blevins wake model. An interference analysis method is then pro­
analysis are listed in Table 7. The platform motion can be calculated in posed on the basis of the finite element method, Newton’s method, and
accordance with RAO and wave data. The RAOs of the platform is shown Newmark-β method. Lastly, an interference analysis program for the
in Fig. 12. The pitch and surge motions of the platform are set as the top interference of two parallel strings in a dual-derrick system is developed.
displacement boundary of parallel strings while the heave motion is Static interference analysis of parallel strings subject to steady cur­
neglected since it does not have direct influence on the lateral defor­ rents in the dual-derrick system is performed on the basis of the pro­
mation of parallel strings. posed analysis model. The wake effect of the riser is significant and
The Blevins wake model is also applied to each current-dominated adsorbs the drill pipe around the riser. The adsorption effect may cause a
time step in the dynamic analysis to ensure that the wake shielding ef­ collision between the two parallel strings. The two parallel strings are
fect of the upstream cylinder always acts on the downstream cylinder mostly likely to collide with each other when the upstream pipe is
when subjected to current and wave (Fu et al., 2017). Figs. 13 and 14 hanging and the downstream one is connected. The collision risk of
show the results of the dynamic interference analysis of the parallel parallel strings exists at the end of the hanging string and increases with
strings. The results depict the distance between two strings (DBS), the the hanging WD.
lateral displacement of the US, and the lateral displacement of the DS, Dynamic interference analysis of parallel strings in the dual-derrick
where the dashed line is the static response, and the solid line is the system under dynamic excitation of waves and platform motions is
dynamic response. The dynamic DBS is smaller than the static DBS when also conducted on the basis of the proposed analysis model. The collision
the drill pipe is in upstream, whereas the dynamic DBS is larger than the between the two strings decreases when the drill pipe is in upstream due
static DBS when the drill pipe is in downstream. In addition, the collision to the combined dynamic wave and current loads. On the contrary, the
between strings decreases in the dynamic analysis when the drill pipe is collision between the two strings increases when the riser is in upstream
in upstream but increases when the drill pipe is in downstream. The due to the adsorption effect.
main reason is that the wake effect is minimal when the drill pipe is in

Fig. 13. Lateral displacement envelope of parallel strings when drill pipe is in upstream.

9
X. Liu et al. Ocean Engineering 269 (2023) 113720

Fig. 14. Lateral displacement envelope of parallel strings when drill pipe is in downstream.

CRediT authorship contribution statement Data availability

Xiuquan Liu: Conceptualization, Methodology, Writing – review & No data was used for the research described in the article.
editing. Yan Shi: Investigation, Writing – original draft, Formal analysis.
Yanwei Li: Investigation, Modeling. Zhaowei Liu: Writing – review & Acknowledgements
editing. Na Qiu: Writing – review & editing. Yuanjiang Chang: Inves­
tigation. Guoming Chen: Formal analysis. This work was supported by National Natural Science Foundation of
China (Grant No: 52271300, 51809279, 52071337), High-tech Ship
Declaration of competing interest Research Projects Sponsored by MIIT (CBG2N21-4-2-5), Program for
Changjiang Scholars and Innovative Research Team in University (Grant
The authors declare that they have no known competing financial No. IRT14R58).
interests or personal relationships that could have appeared to influence
the work reported in this paper.

Appendix A

The solution steps for the static interference analysis model are as follows.

(1) Establish the stiffness matrix K and force vectors F. The nodal force of the DS needs to be corrected in accordance with the wake model in the
interference analysis.
(2) Define the initial deformation increment Δδ0 and the initial deformation value δ0.
(3) Define the deformation convergence condition DP and the load convergence condition FP.
(4) Calculate the residual force vector ΔF:
ΔF = F − Kδ0 (A-1)

(5) Update the deformation increment Δδ0:

ΔF
Δδ0 = (A-2)
K

(6) Update the deformation value δ0:

δ0 = δ0 + Δδ0 (A-3)

(7) Judge the result convergence. If the following equation is more than zero, then converge; otherwise, repeat step (4).
|Δδ0 | |ΔF|
Dp − , Fp − (A-4)
|δ0 | |F|

10
X. Liu et al. Ocean Engineering 269 (2023) 113720

Appendix B

The solution steps for the dynamic interference analysis model are as follows.

(1) Establish the mass matrix M, damping matrix C, and stiffness matrix K.
(2) Define the initial values of δ0 , δ̇0 , and δ̈0 .
(3) Determine the time step Δt and parameters α and β.
(4) Calculate the integral parameters:
1 β 1
a1 = , a2 = , a3 = ,
αΔt2 αΔt αΔt
( ) (B-1)
1 β Δt β
a4 = − 1, a5 = − 1, a6 = − 2
2α α 2 α

(5) Calculate the effective stiffness matrix K:

̃

̃ = K + a1 M + a2 C
K (B-2)

(6) Calculate the vector of the effective load ̃ Ft+Δt for the time of t+Δt:
[ ] [ ]
̃ t+Δt = Ft+Δt + a1 δt + a3 δ̇t + a4 δ̈t M + a2 δt + a5 δ̇t + a6 δ̈t C
F (B-3)

(7) Apply the boundary conditions to the effective stiffness matrix K

̃ and the effective load ̃
Ft+Δt by using the zero-setting method.
(8) Calculate the displacement δt+Δt for the time of t+Δt:

̃− 1F
δt+Δt = K ̃ t+Δt (B-4)

(9) Calculate the velocity vector δ̇t+Δt and the acceleration vector δ̈t+Δt at the next time step t+Δt:
( )
1 1 1
δ̈t+Δt = (δ t+Δt − δ t ) − δ̇ t − − 1 δ̈t
αΔt2 αΔt 2α (B-5)
δ̇t+Δt = δ̇t + (1 − β)Δtδ̈t + βΔtδ̈t+Δt

References He, J.W., Low, Y.M., 2012. An approach for estimating the probability of collision
between marine risers. Appl. Ocean Res. 35, 68–76.
He, J.W., Low, Y.M., 2014. A collision-allowed approach for probabilistic assessment of
Adamiec-Wójcik, I., Brzozowska, L., Drąg, Ł., Wojciech, S., 2023. Optimal base motion to
riser interference. Ocean Eng. 76, 21–31.
compensate for the influence of sea currents during riser re-entry. Mar. Struct. 88,
Hu, Y., Cao, J., Yao, B., Zeng, Z., Lian, L., 2018. Dynamic behaviors of a marine riser with
103357.
variable length during the installation of a subsea production tree. J. Mar. Sci.
Blevins, R.D., 2005. Forces on and stability of a cylinder in a wake. J. Offshore Mech.
Technol. 23 (2), 378–388.
Arctic Eng. 127 (1), 39–45.
Huse, E., 1992. Current Force on Individual Elements of Riser Arrays. The Second
Connaire, A., O’Brien, P., Harte, A., O’Connor, A., 2015. Advancements in subsea riser
International Offshore and Polar Engineering Conference, San Francisco, California,
analysis using quasi-rotations and the Newton–Raphson method. Int. J. Non Lin.
USA. June 1992, ISOPE-I-92-135.
Mech. 70, 48–62.
Kim, J.D., Jang, B.S., Kim, H.J., 2021. Application of analytical model in the prediction
de Andrade, E.M., de Oliveira Costa, D., Fernandes, A.C., Sales Junior, J.S., 2023.
of dynamic responses and fatigue damage of flexible risers: Part II – dynamic analysis
A review on the modeling of subsea lifting operations. Ocean Eng. 268, 113293.
of flexible risers in large-scale domain using a direct moment correction method.
de Araújo Neto, E.D., Jacovazzo, B.M., Correa, F.N., Jacob, B.P., 2019. Numerical
Mar. Struct. 79, 103051.
evaluation of a subsea equipment installation method designed to avoid resonant
Liu, X., Li, Y., Zhang, N., Sun, H., Chang, Y., Chen, G., et al., 2020. Improved axial
responses. Appl. Ocean Res. 88, 288–305.
dynamic analysis of risers based on finite element method and data-driven models.
Fan, H., Li, C., Wang, Z., Xu, L., Wang, Y., Feng, X., 2017. Dynamic analysis of a hang-off
Ocean Eng. 214, 107782.
drilling riser considering internal solitary wave and vessel motion. J. Nat. Gas Sci.
Liu, X., Sun, H., Yu, M., Qiu, N., Li, Y., Liu, F., et al., 2021. Mechanical analysis of
Eng. 37, 512–522.
deepwater drilling riser system based on multibody system dynamics. Petrol. Sci. 18,
Fang, H., Duan, M., 2014. Offshore Operation Facilities: Equipment and Procedures. Gulf
603–617.
Professional Publishing, pp. 19–22.
Liu, J., Ma, X., Zhang, X., Zhang, X., Zhang, Z., Liu, G., 2022. Axial vibration of deep-
Fu, P., Leira, B.J., Myrhaug, D., 2017. Reliability analysis of wake-induced collision of
water drilling risers under lifting conditions. J. Pet. Sci. Eng. 209, 109903.
flexible risers. Appl. Ocean Res. 62, 49–56.
Munch-Søegaard, L., Nergaard, A., 2001. Offshore Drilling Experience with Dual Derrick
Fu, P., Leira, B.J., Myrhaug, D., 2018. The effect of local wake model on collision
Operations. SPE/IADC Drilling Conference, Amsterdam, Netherlands. February
probability for risers subjected to current and random waves. Ocean Eng. 162,
2001, SPE-67706-MS.
331–340.
Oh, J., Jung, D., Kim, H., Min, C., Cho, S., 2020. A study on the simulation-based
Fu, P., Leira, B.J., Myrhaug, D., Chai, W., 2021. Assessment of methods for short-term
installation shape design method of steel lazy wave riser (SLWR) in ultra deepwater
analysis of riser collision probability. Ocean Eng. 238, 109221.
depth. Ocean Eng. 197, 106902.
Gu, H., Guo, H., Bai, F., Li, X., Li, F., 2022. Study on the collision probability of vertical
Qin, H., Liu, J., Xiao, W., Wang, B., 2019. Quasistatic nonlinear analysis of a drill pipe in
risers considering the wake interference effect. Ocean Eng. 245, 110583.
subsea Xmas tree installation. Math. Probl Eng., 4241363, 2019.
Han, X., Tang, Y., Meng, Z., Ruan, W., Qiu, A., Gu, J., et al., 2022. Surface roughness
Qin, K., Di, Q., Zhou, X., He, Y., Wang, W., Chen, F., Zhang, H., 2022. Nonlinear dynamic
effects on a tensioned riser vortex-induced vibration in the uniform current. Appl.
characteristics of the drill-string for deep-water and ultra-deep water drilling. J. Pet.
Ocean Res. 118, 102970.
Sci. Eng. 209, 109905.

11
X. Liu et al. Ocean Engineering 269 (2023) 113720

Rustad, A.M., Larsen, C.M., Sørensen, A.J., 2008. FEM modelling and automatic control Wang, Y., Gao, D., Fang, J., 2015b. Study on lateral vibration analysis of marine riser in
for collision prevention of top tensioned risers. Mar. Struct. 21, 80–112. installation-via variational approach. J. Nat. Gas Sci. Eng. 22, 523–529.
Sagatun, S.I., Herfjord, K., Holmås, T., 2002. Dynamic simulation of marine risers Wang, Y., Tuo, H., Li, L., Zhao, Y., Qin, H., An, C., 2018. Dynamic simulation of
moving relative to each other due to vortex and wake effects. J. Fluid Struct. 16 (3), installation of the subsea cluster manifold by drilling pipe in deep water based on
375–390. OrcaFlex. J. Pet. Sci. Eng. 163, 67–78.
Schlichting, H., Kestin, J., 1961. Boundary Layer Theory, vol. 121. McGraw-Hill, New Wang, W., Hao, Y., Li, H., Wu, Z., Chen, Y., Gu, J., 2021. Mechanical model and
York. experimental study of marine coiled tubing-in-riser system considering string
Tian, D., Fan, H., Leira, B.J., Sævik, S., 2020a. Study on the static behavior of installing a coupling effects. Appl. Ocean Res. 111, 102632.
deep-water drilling riser on a production platform. J. Pet. Sci. Eng. 185, 106652. Xu, L., Wang, J., Li, Y., Sheng, L., Li, C., Liu, J., 2022. Development and sea trial
Tian, D., Fan, H., Leira, B.J., Sævik, S., Fu, P., 2020b. Static analysis of interaction investigation for deepwater drilling riser specialized soft hang-off system during
between two adjacent top tensioned risers with consideration of wake effects. Ocean transit. Ocean Eng. 243, 110310.
Eng. 195, 106662. Yin, D., Passano, E., Lie, H., Grytøyr, G., Aronsenb, K., Tognarellic, M., Kebadze, E.,
Tommasini, R.B., Hill, T.L., Macdonald, J.H.G., Pavanello, R., de Oliveira Carvalho, L., 2019. Experimental and numerical study of a top tensioned riser subjected to vessel
2021. Nonlinear dynamics of deep water subsea lifting operations considering KC- motion. Ocean Eng. 171, 565–574.
dependent hydrodynamic coefficients. Ocean Eng. 233, 109172. Zhou, X., Sun, Y., Duan, M., Li, W., 2020. Numerical investigation on crack identification
Wang, Y., Gao, D., 2021. On the static mechanics of the tubular system during using natural frequencies and mode shapes of a drilling riser during deployment and
installation of the surface casing in deepwater drilling. Appl. Ocean Res. 110, retrieval. J. Petrol. Sci. Eng. 195, 107721.
102599. Zhu, J., Liu, K., Chen, G., Zhu, G., 2019. Interference analysis and operability envelopes
Wang, Y., Gao, D., Fang, J., 2015a. Mechanical behavior analysis for the determination design for deepwater parallel strings. Oil Gas Sci. Technol. 74, 71.
of riser installation window in offshore drilling. J. Nat. Gas Sci. Eng. 24, 317–323.

12