PETROLEUM EXPLORATION AND DEVELOPMENT

Volume 44, Issue 3, June 2017

Online English edition of the Chinese language journal

Cite this article as: PETROL. EXPLOR. DEVELOP., 2017, 44(3): 487–494. RESEARCH PAPER

A new type of anti-heave semi-submersible drilling platform

CHEN Bo, YU Zhiyong*, LYU Yong, LI Xiaojian, LI Chunfang
School of Mechanical and Electrical Engineering, Lanzhou University of Technology, Lanzhou 730050, China

Abstract: A new type of semi-submersible drilling platform is designed. The numerical simulation software is used to analyze the heave
response of the new platform in the frequency domain and time domain, and the semi-submersible drilling platforms with double-floating
four-column structure and heave plate structure are compared with the new platform. This paper introduces the design principles of the
new platform and the theoretical basis, mathematical model and boundary conditions during correlation analysis. The numerical simula-
tion results show that the maximum and mean values of the heave response of the new platform are significantly reduced in the frequency
domain analysis compared with the double-floating four-column and the heave-plate structure platform; and the new platform has a sig-
nificant increase in the natural heaving period of the new platform, which can effectively prevent the occurrence of resonance. In the
mooring time domain coupling analysis, the surge, sway and roll response of the new platform is small, and the heave response is greatly
reduced. In the spectral analysis, the new platform has a smaller peak response and better wave frequency characteristics. The new plat-
form has excellent anti-heave performance, reasonable structure and feasibility, and can provide reference for the design and selection of
new generation semi-submersible drilling platform.

Key words: drilling platform; semi-submersible drilling platform; heave prevention; frequency domain analysis; time domain analysis

Introduction compensation device must be installed on the platform, which

A drilling platform is important equipment for offshore oil
and gas exploitation, of which a semi-submersible drilling
platform with unique advantages becomes one of the plat-
forms with the most promising development prospect[13]. In
the marine environment, forces applied on a platform are in
various forms[45], so it is difficult to analyze the overall per-
formance of the platform with mathematical model, but hy-
drodynamic simulation is able to solve this problem[6]. The
motion of platform in waves is nonlinear, which is caused by
the coupling of different forms of motion (such as roll and
heave coupling), the platform large-amplitude rolling and
complex hydrodynamic loads[7]. The platform has six degrees
of freedom in the marine environment[8], due to the structural
characteristics of the semi-submersible drilling platform, the
roll and heave responses have strong impacts on the safety of
the platform[9] under the towing, operation and storm survival
conditions, but the increase of additional mass can inhibit the
roll and heave responses[1011]. Non-linearity of roll motion
can be simulated and studied by using chaos theory, differen-
tial dynamical system theory and bifurcation theory. There
existing large safety hazard during normal drilling if the heave
response is large, in order to reduce the impact of heave and
ensure the constant contact of bit with well bottom, heave
used telescopic drill pipe in the early stage, and now uses
traveling compensation, crane compensation and winch com-
pensation etc.[12]. At present, semi-submersible drilling plat-
form has already developed the sixth generation which com-
bines with the chain and the dynamic positioning system, can
reduce the heave motion to some extent, but can not adapt to
harsh sea conditions[1314]. Therefore, a new type of semi-sub-
mersible drilling platform has been designed (referred to as
the new platform), the performance of the new platform has
been investigated with hydrodynamic numerical simulation,
and compared with the original double floating four column
structure platform (referred to as the original platform) and
heave plate platform in this study.
1. Design of the new platform
1.1. Theoretical basis
When the ratio of the cross-sectional dimension of a struc-
ture to the wavelength is less than 0.2, the structure is called a
small structure. Wave forces on it are dominated by inertial
force and drag force, using a semi-theoretical and semi-em-
pirical Morrison formula[1516], the simplified calculation for-
mula is:
F  V  I 3  C a  vw  0.5  C D vw vw e (1)
When the ratio of cross-sectional dimension of a structure

Received date: 27 Apr. 2016; Revised date: 10 Mar. 2017.

* Corresponding author. E-mail: 779639085@qq.com
Copyright © 2017, Research Institute of Petroleum Exploration and Development, PetroChina. Published by Elsevier BV. All rights reserved.
 CHEN Bo et al. / Petroleum Exploration and Development, 2017, 44(3): 487–494

to the wavelength is greater than or equal to 0.2, three dimen-  Fjpq 

sional potential flow theory is taken[17] to do the analysis. The B jpq   Re   (10)
 v
 jp 
theory is an important basic theory for hydrodynamic analysis
of structures, in which the velocity potential of fluid is par- 1.3. Boundary conditions
ticularly crucial. The total velocity potential includes the in-
1.3.1. Platform displacement boundary conditions
cident potential caused by the incident wave, the diffraction
potential caused by the floating to the flow field, and the ra- To avoid rigid displacement of the model, displacement
diation potential caused by the perturbation of floating motion boundary conditions are required in the analytical model, that
to the flow field, namely: is to say, 3 non-collinear nodes with great strength far away
Φ  x, y , z , t   ΦI  x, y , z , t   ΦD  x, y , z , t   from platform structure evaluation area should be chosen to
fix the structure. Displacement boundary conditions are ap-
ΦR  x , y , z , t  (2)
plied at the selected nodes: node 1 restricts z direction dis-
1.2. Mathematical model placement; node 2 restricts y and z direction displacement;
node 3 restricts x, y, and z direction displacement.
The force of current on the platform structure is:
Fliu  0.5Cd  Avliu 2 (3) 1.3.2. Mooring boundary conditions

where the vliu is the velocity of ocean current. The current ① Statics boundary conditions. For the upper end of the
velocity of the South China Sea is about 0.2m/s, so relative to mooring anchor:
wave force on the platform, the force of the ocean current on v  l  La   vtop (11)
platform is negligible. For lower end of the mooring anchor:
The wave force and torque of platform are respectively: v  l  0   vbot (12)
π 2 r 2  HL
Fwave  Cm 1  e kh  sin  (4) ② Dynamic boundary conditions. For the upper end of the
T2 mooring anchor:
π 2 r 2  HLh  1  e  kh  v  l  La   vtop
M  Cm sin  1   (5)
T 2
 kh  
v  l  La   vtop
 (13)
Motion balance equation of platform under the wave load
For lower end of the mooring anchor:
is:
M m X   CX   D1 X   D2 f ( X )  KX  Fsum (6) v  l  0   vbot
 (14)
Determining the hydrodynamic parameter of the platform v  l  0   vbot

motion response in waves is the basis of the numerical calcu-
1.3.3. Flow field boundary conditions
lation. When the platform is subjected to a slight oscillatory
motion in a specific mode, an outward radiation flow field is ① Free surface boundary conditions. At each point of the
generated in a stable flow field to form radiated waves[18]. free surface:
Separating the velocity and acceleration of the corresponding  2Φ Φ
g 0
structure from radiation wave load, we can get the parameters t 2 z (15)
of the radiation wave load related to the velocity and accelera- In the local coordinate system, column surface immersed in
tion, namely, additional damping parameter and additional sea water should meet the requirements of:
quality parameter are collectively referred to as hydrodynamic  Φ
parameters[18].  z  0
 (16)
The radiation wave load force is:
 Φ  0
Fjpq   p jp  n j ds  r
Sj
②Surface boundary conditions. Instantaneous normal ac-
(j=1, 2, …, N; p=1, 2, 3; q=1, 2, 3) (7) celeration of fluid particle is zero for the area where surface
where the pjp is pressure vector at any point in flow field, de- total force is zero:
termined by Bernoulli's equation[18].  2Φ  0 (17)
The relation between radiation wave force and hydrody- ③ Seabed boundary conditions. Seabed as a rigid wall,
namic parameters is: normal velocity component of the fluid particle is zero.
Fjpq   Ajpq vjp  B jpq v jp (8) ④ Radiation conditions of infinite point. Wave forms
With further analysis, the calculation formulas of hydrody- spread from near to far, that is when r approaches infinity,
namic parameters are:  approaches zero.
 Fjpq  1.4. Design principle
Ajpq   Re   (9)
 v
 jp  Since the external excitation time cannot be controlled, in
 488 
 CHEN Bo et al. / Petroleum Exploration and Development, 2017, 44(3): 487–494

order to effectively reduce the vertical displacement of the m, 1.552 m respectively, all by more than 36%. At wave an-
platform, reducing the instantaneous acceleration is an effec- gles of 0°, 30°, 60°, 90°, compared with the original platform,
tive way. According to Newton's second law of motion, the the mean values of unit wave of the new platform heave re-
mass of the platform should be increased or the resultant force sponse decrease by 0.283 1 m, 0.253 1 m, 0.239 5 m, 0.269 6
on the platform should be reduced to reduce the instantaneous m respectively, all by more than 48%; compared with the
acceleration of the platform. The resultant force of the plat- heave plate platform, the mean values of unit wave of the new
form is the difference of wave force and damping force. platform heave response decrease by 0.241 5 m, 0.231 9 m,
Therefore, in the design of the new platform, we have tried to 0.243 3 m, 0.277 4 m respectively, all by more than 47%.
reduce the heave response of platform by increasing the plat- Therefore, under the action of unit regular waves, the new
form's overall quality and the damping force. Under normal platform has a good anti-heave characteristic. From Table 2, it
conditions, when the water depth is more than 30 m, the sea can be seen that with the increase of wave angle from 0° to
water is stable, almost unaffected by the waves. Full water 90°, the maximum values of platform heave increase, and the
tank with a large cross-sectional area and quality, can provide average values first decrease and then increase.
large vertical damping force, and increase the quality of plat-
2.2. Natural period of heave
form. In order to validate the rationality and feasibility of the
new platform structure, the platform structure has been com- Analysis shows that the natural periods of the original plat-
pared with double floating four-column structure semi-sub- form, heave plate platform and the new platform are 18.92,
mersible drilling platform and heaving plate semi-submersible 16.31 and 39.81 s respectively. The waves in the ocean have
drilling platform. Main sizes of 3 kinds of platforms, simula- uneven periods, but the waves with periods of 130 s occupy
tion environment, platform (columns, floating) working depth more than half of the total energy of the ocean waves, and the
and other parameters are the same (Table 1). corresponding energy peak period is about 10 s. In the selec-
2. Frequency domain analysis tion and design of platform, the heave natural period should
be kept away from 10 s, the new platform has a maximal
In this study, the platform model was established by AN-
SYS software, and the hydrodynamic analysis was carried out Table 1. Main dimensions and draft of the platform model
by classical hydrodynamic analysis software AQWA. AQWA Parameter Parameter
software is the software has the loosest restriction on the Parameter Parameter
value/m value/m
number of nodes among various hydrodynamic analysis soft- Floating body length 101.23 Heave plate width 20.00
ware, so the grid unit can be smaller and the accuracy of cal- Connecting column
culation and analysis can be improved. Fig. 1 shows the fre- Floating body width 18.60 1.50
diameter
quency domain analysis model for the platform. Connecting column
Floating body height 10.72 40.00
length
2.1. Heave response under unit wave
Column length 18.51 Full tank length 101.00
Fig. 2 shows the heave response of the platform under unit Column width 18.51 Full tank width 40.00
regular waves of different wave directions, and the statistics Transverse brace
are shown in Table 2. Table 2 shows that at the wave angles of 2.15 Full tank height 20.00
diameter
0°, 30°, 60°, 90°, compared with the original platform, the Transverse brace Tank transverse brace
60.31 2.00
maximum values of unit wave of the new platform heave re- length diameter
sponse decrease by 1.790 m, 1.880 m, 2.031 m, 2.094 m re- Platform main Tank transverse
20.00 20.00
spectively, all by more than 49%; compared with the platform body draft brace length
of heave plate, the maximum values of unit wave of the new Heave plate and
Heave plate length 40.00 60.00
platform heave response decrease by 1.041 m, 1.173 m, 1.428 tank draft

Fig. 1. Frequency domain analysis model of the platform.

 489 
 CHEN Bo et al. / Petroleum Exploration and Development, 2017, 44(3): 487–494

Fig. 2. Platform heave response under the action of unit regular wave at different wave angles.

Table 2. Statistics of platform heave response under the action ertia force to the acceleration of the platform is the additional
of unit regular waves mass of the platform, and the larger the value is, the larger the
Wave Minimum Maximum Average resistance of the accelerated motion of the platform will be.
Platform type
angle/(°) value/103 m value/m value/m Due to space limitation, only the additional mass components
Original platform 4.662 3.637 0.549 8 curves of heaving and rolling are shown (Fig. 3), it can be
0 Heave plate platform 6.700 2.888 0.508 2 seen that vertical and rolling additional mass of the new plat-
New platform 1.430 1.847 0.266 7 form increase significantly compared with the original plat-
Original platform 16.580 3.731 0.512 0 form and heave plate platform, so the new platform can effec-
30 Heave plate platform 9.348 3.024 0.490 8 tively inhibit heave and roll.
New platform 2.097 1.851 0.258 9
3. Time domain coupling analysis
Original platform 7.998 3.890 0.490 4
60 Heave plate platform 9.400 3.287 0.494 2 In the time domain coupling analysis, JONSWAP spectrum,
New platform 1.555 1.859 0.250 9 which is similar to South China Sea spectrum was used, tak-
Original platform 30.610 3.956 0.524 6 ing 0.026 2 and 1.520 0 Hz as the lower limit and upper limit
90
of frequency, significant wave height was 3 m, peak fre-
Heave plate platform 14.010 3.414 0.532 4
quency was 0.698 1 Hz, and peak factor was 4; The platform
New platform 4.837 1.862 0.255 0
was fixed with 12 mooring anchor chains, and the angle be-
heave natural period of 39.81 s, far longer than the heave tween two adjacent anchor chains is 30° (Fig. 4). Considering
natural periods of original platform and heave plate platform, the key factors affecting platform motion response, surge,
so the new platform can prevent the occurrence of resonance sway, heave and roll are selected to be main responses for
effectively. analysis, especially for the heave response.

2.3. Additional mass 3.1. Time course of surge, sway and roll

The reaction force of the drilling platform is called the ad- From Figs. 5 and 6, it can be seen that the new platform has
ditional inertia force, which is opposite to the acceleration of much smaller surge and sway response compared with origi-
the platform. In accelerated motion of platform, additional nal platform and platform of heave plate, so the possibility of
inertia force acts as a resistance, but it acts as a dynamic force fatigue failure due to excessive longitudinal and lateral motion
in decelerated motion of platform. The ratio of additional in- is reduced; the new platform also has much smaller roll
 490 
 CHEN Bo et al. / Petroleum Exploration and Development, 2017, 44(3): 487–494

Fig. 3. Platform additional mass.

Fig. 4. Time domain coupling analysis model of the platform.

Fig. 5. Surge and sway response time course curves.

response, lowering the possibility of ship water on deck due to
excessive rolling, so the new platform meets the requirements
of safe operation in harsh sea conditions.
3.2. Time course of heave
Fig. 7 shows that, compared with the original platform and
heave plate platform, the new platform has much smaller time
domain coupling heave response, indicating that the new
platform has excellent anti-heave performance.
3.3. Time domain coupled heave response spectrum
The wave spectrum density function is used to express the
distribution of the wave energy to the platform relative to
 491 
frequency in the irregular wave, also called energy spectrum.
The response spectrum is equal to the product of wave spec-
tral density function and transfer function RAO squared. Time
domain coupling heaving response spectrum curve of plat-
form is shown in Fig. 8. It can be seen that, the responses
concentrate in the range of 0.30.4 Hz for original platform,
with peak of 6.50 m2s, showing wave frequency response; for
platform of heave plate, the responses concentrate in the range
of 0.30.7 Hz, with peak of 0.85 m2s, presenting wave fre-
quency response; for the new platform, the responses concen-
trate in the range of 0.30.6 Hz, with peak of 0.13 m2s, pre-
senting wave frequency response mainly with some low fre-
quency responses. Compared with the original platform and
 CHEN Bo et al. / Petroleum Exploration and Development, 2017, 44(3): 487–494

Fig. 6. Roll response time course curves. Fig. 7. Heave response time course curve.

Fig. 8. Time domain coupling heave response spectrum curves of platforms.

the heave plate platform, the new platform has smaller peak of
response spectrum and better characteristics of heave wave
frequency.
4. Conclusions
In the frequency domain, compared with the original plat-
form and heave plate platform, the new platform has much
smaller maximum and mean values of heave response, indi-
cating that the new platform has excellent anti-heave per-
formance. Natural period of heave is an important parameter
for platform performance evaluation, natural heave period of
the new platform is about two times of the original platform
and heave plate platform, can effectively prevent the occur-
rence of heave resonance, and improve the security of the
platform. Additional mass value reflects how difficult it is to
change the platform motion state, the heaving additional mass
of the new platform is much larger than that of the original
platform and heave plate platform, and thus can reduce the
instantaneous vertical effective excitation of platform.
In waters of less than 1 500 m deep, the semi-submersible
drilling platform is positioned mainly by mooring, the new
platform has much smaller surge, sway, and roll in time do-
main coupling analysis of mooring, and the heave response is
much smaller than that of the original platform and heave
plate platform, so it has much better anti-heave performance
in the same conditions, better operation ability, better adapta-
bility and stability. Compared with the original platform and
heave plate platform, the new platform has smaller peak value
of response spectrum and better wave frequency characteris-
tics.
Nomenclature
A—projection area of structure in the direction perpendicular to
the current, m2;
Ajpq—additional mass parameter, kg;

 492 
 CHEN Bo et al. / Petroleum Exploration and Development, 2017, 44(3): 487–494

Bjpq—additional damping parameter, kg/s; θ—wave angle, (°);

C—damping matrix, kg/s; ρ—sea water density, kg/m3;
Ca—3 × 3 order additional mass coefficient diagonal matrix; Φ—total velocity potential;
Cd—drag coefficients obtained from experience and experiment; ΦD—diffraction potential;
CD—3 × 3 order drag force coefficient matrix; ΦI—incident potential;
Cm—inertial force coefficient; ΦR—radiation potential.
D1—linear damping matrix, kg/s;
D2—secondary damping matrix, kg/s; References
2
e—morrison unit projection area, m ;
f—velocity matrix function; [1] HALKYARD J. Large spar drilling and production platforms
F—resultant force matrix on the small structure calculated ac- for deep water oil and gas. In: WANG C M, WANG B T.
cording to the Morrison theory, N; Large floating structures. Singapore: Springer Singapore,
Fjpq—radiation wave load force in q mode direction caused by 2015: 221–260.
slight oscillatory movement of jth cylinder in p mode, N; [2] WANG S, CAO Y, FU Q, et al. Hydrodynamic performance of
Fliu—current force on platform, N; a novel semi-submersible platform with nonsymmetrical pon-
Fsum—resultant force matrix, N; toons. Ocean Engineering, 2015, 110: 106–115.
Fwave—the wave force applied on the platform, N; [3] GUO Y, SONG L, ZENG H, et al. A field trial of a new system
g—gravitational acceleration, 9.8 m/s2; using both free-standing drilling riser and mid-depth BOP in
h—distance from column bottom surface to the waterplane, m; the South China Sea. In: The Twenty-fifth International Off-
H—design wave height, m; shore and Polar Engineering Conference.Kona, Hawaii, USA:
I3—3 × 3 order unit matrix; International Society of Offshore and Polar Engineers, 2015.
k—equivalent coefficient; [4] MENG Q, ZHANG K, CHENG H, et al. An analytical method
K—stiffness matrix, N/m; for predicting the fluctuation of thrust force during drilling of
l—lagrange coordinate, m; unidirectional carbon fiber reinforced plastics. Journal of
L—design wavelength corresponding to intermediate water depth, m; Composite Materials, 2015, 49(6): 699–711.
La—length of the mooring anchor chain, m; [5] HOLMES S, BHAT S, BEYNET P, et al. Heave plate design
M—wave force moment applied on the platform, N·m; with computational fluid dynamics. Journal of Offshore Me-
Mm—platform mass matrix, kg; chanics & Arctic Engineering, 2001, 123(1): 2477–2485.
N—number of cylinders; [6] WU Linjian, WANG Yuanzhan, XIAO Zhong, et al. Hydro-
nj—normal vector in jth cylinder unit; dynamic response for flexible connectors of mobile offshore
pjp—pressure vector at any point in flow field of jth cylinder in p base at rough sea states. Petroleum Exploration and Develop-
mode, Pa; ment, 2016, 43(6): 997–1004.
r—column radius, m; [7] KALIŃSKI K J, BUCHHOLZ C. Mechatronic design of
Re—taking the real part of a complex number; strongly nonlinear systems on a basis of three wheeled mobile
s—area, m2; platform. Mechanical Systems and Signal Processing, 2015,
Sj—wetted surface area of the jth cylinder, m2; 52: 700–721.
t—time, s; [8] VAN DER TEMPEL J, SALZMANN D J C, KOCH J, et al.
T—wave period, s; Vessel, motion platform, method for compensating motions of
vliu—ocean current velocity, m/s; a vessel and use of a Stewart platform: U.S. Patent 8672288.
v—speed of a point on the anchor chain, m/s; 2014-03-18.
3
V—displacement volume of Morrison unit, m ; [9] YIN Haitao. Hydrodynamic performance analysis of the heave
vjp—speed of jthcylinder in p mode, m/s; plate of Cell Sparplatform.Dalian:Dalian University of Tech-
vjp —acceleration of the jth column in p mode, m/s2; nology, 2010.
vtop, vbot—speed of the upper and lower ends of the anchor chain, [10] SHI Qiqi. Motion and dynamic characteristics research of the
m/s; deepwater mooring semi-submersible drilling platform.
 —acceleration of the upper and lower ends of the anchor
vtop , vbot Shanghai: Shanghai Jiao Tong University, 2011.
chain, m/s2; [11] XU H, HAN S, BI X, et al. Atmospheric metallic and arsenic
vw—speed of wave perpendicular to the bracing, m/s; pollution at an offshore drilling platform in the Bo Sea: A
vw —acceleration of wave perpendicular to the bracing, m/s2; health risk assessment for the workers. Journal of Hazardous
x, y, z—rectangular coordinate system, m; Materials, 2016, 304: 93-102.
x, y, z—speeds in x, y, z direction, m/s; [12] MARTIN T. High definition drilling rate of penetration for
X—displacement matrix, m; marine drilling: U.S. Patent 9217290. 2015-12-22.
X—velocity matrix, m/s; [13] KJERSTAD Ø K, METRIKIN I, LØSET S, et al. Experimen-
X—acceleration matrix, m/s2; tal and phenomenological investigation of dynamic position-

 493 
 CHEN Bo et al. / Petroleum Exploration and Development, 2017, 44(3): 487–494
ing in managed ice. Cold Regions Science and Technology,
2015, 111: 67–79.
[14] DENG S, FAN H, SHEN W, et al. An optimization method of
top tension in drilling riser-conductor system. Arabian Journal
for Science and Engineering, 2016, 41(7): 2707–2714.
[15] FANG Huacan, CHEN Shiyi.Analysis of heave motion of
semi-submersible drilling platform. ActaPetroleiSinica, 1986,
7(1): 83–91.
[16] ZHU Yazhou, SUNChengmeng, ZHANGXiaoyu, et al. Sensi-
 494 
tivity of self-elevating unit leg strength to different chord
space. Petroleum Exploration and Development, 2015, 42(5):
656–661.
[17] GuoHaiqiang, ZHU Renchuan, MIAO Guoping, et al.Test and
analysis technology of hydrodynamic coefficient of ship of
numerical wave in the pool. Shipbuilding of China, 2008,
49(S1): 58–65.
[18] DONG Yanqiu.Wave load and response of deep-sea platform.
Tianjin: Tianjin University Press, 2005: 225–246.