Journal of Coastal Research SI 94 475–479 Coconut Creek, Florida 2019

Optimal Operation of Subsea and Onshore Oil

Transmission Pipeline
Shanbi Peng†*, Yuan Liu‡, Li Xu†, Qiyong Zhu§, and Mohammadamin Azimi6
† ‡ §
Southwest Petroleum University Petro China Southwest Oil & Gasfield Petro China West Pipeline
Chengdu 610500, China Company Company
Chengdu 610066, China Chengdu 84100, China
6
Trenchless Technology Center
Louisiana Tech University
Ruston, LA 71270, USA

ABSTRACT
Peng, S.-B.; Liu, Y.; Xu, L.; Zhu, Q.-Y., and Amizi M., 2019. Optimal operation of subsea and onshore oil transmission
pipeline. In: Gong, D.; Zhu, H., and Liu, R. (eds.), Selected Topics in Coastal Research: Engineering, Industry, Economy,
and Sustainable Development. Journal of Coastal Research, Special Issue No. 94, pp. 475–479. Coconut Creek (Florida),
ISSN 0749-0208.

Pipeline is one of the most important means of oil and gas transportation, more and more subsea pipelines have been
built these years. As the subsea and onshore crude oil transmission pipelines were laid many years ago, the oil pump
combination schemes are complicated, which leads to high energy consumption and excessive cost in the pipeline
operation. Concerning this problem, based on parameters of the pipeline and crude oil, an economical operation
mathematical model for continuous transportation crude oil pipelines is set up, which taking the minimum energy
consumption as the objective function, the pump combination scheme and inlet temperature of heating station as the
decision variables. According to the characteristics of the model, Tabu search algorithm is used to solve the mathematical
model, and an optimization software is developed by C program. Based on actual operating data, the operation scheme of
a crude oil pipeline is optimized by the software. Comparison of energy consumption after optimization against those
from results before optimization is illustrated. From the optimal results, it can be seen that the total operation cost after
optimization is only 81.5% as before in winter, the effect is significant. Although the structure of subsea pipelines is
different from onshore pipelines, the hydraulic and thermodynamic calculation methods are the same, and the optimal
operation method can be used for both onshore and subsea pipelines.

ADDITIONAL INDEX WORDS: Crude oil pipeline, optimal operation, Tabu search algorithm, energy consumption.

INTRODUCTION The object function of this method was the minimum fuel cost,
Pipeline is one of the most important means of oil and gas and the mathematical programming algorithm was chosen to
transportation. There are more than 26 thousand kilometers of get the optimal pump combination scheme. In 2001, Jokic has
crude oil transmission pipeline in China (Lu et al., 2018). In the studied the optimization of pipeline network by using nonlin-
process of operation, the combination of pumping stations is ear programming method (Jokic, 2001). They assumed that the
complex, it is difficult to formulate a reasonable operation plan velocity was constant, and the inlet and outlet pressure were
(Peng et al., 2015). In order to reduce operating energy constant, the optimal object was the minimum operation cost of
consumption and reduce operating costs, it is necessary to the whole pipeline network. Hybrid differential evolution
optimize the crude oil pipelines (Liu, Li, and Cai, 2019). algorithm have been presented by Lin, Hwang, and Wang,
Since the early 1960s, many researchers have studied this (2004), with the objective of proposing a new optimal solution
problem and in current literature there is a significant amount strategy. Genetic algorithm has been used in the optimal
of research has been carried out (Hou et al., 2017; Reddy, operation research of natural gas pipeline networks in Canada
Karimi, and Srinivasan, 2004). Gefferson analyzed the hot oil (Boeros et al., 2004; Nguyen and Chan, 2005). Then, Ant colony
pipeline optimal operation, which using dynamic programming algorithm have been presented by Mora and Ulieru (2005),
method to solve the problem. The pressure of each pump which uses the minimum consumption cost as the optimal
station was calculated, the pumping number of each station object, the number of running compressor and outlet pressure
and the hydrodynamic head provided by the pump were as decision variables. Amount of companies also have devel-
obtained. After that, some scholars improved the optimal oped software to optimize the pipeline operation (Chebouba,
method based on Gefferson’s research, and the optimal pump
2009; Kurak, 1989; Linke, 1996). The system can obtain the
combination scheme of pipeline was determined by using
running parameters of the pipeline online, do real-time
implicit enumeration method. Another method for optimal
optimization, and transfer the optimized data in real time.
operation of each pump station was proposed (Gopal, 1980).
The experimental results showed that the pipeline cost is
reduced by 10% after the optimized operation. In 2014, Liu
DOI: 10.2112/SI94-094.1 received 2 February 2019; accepted in
revision 3 March 2019.
introduced production unit consumption indicators to establish
*Corresponding author: [email protected] an objective function for achieving the goal of lowering energy
Ó
Coastal Education and Research Foundation, Inc. 2019 consumption. By optimizing the boot program, the import
476 Peng et al.

station pressure, and the temperature parameters, the pipeline Hydraulic constraint: Htotal  Hp
now has the potential to reduce energy consumption by 11% to Output temperature constraint: trmin  tri  trmax (4)
16% (Liu et al., 2013; Liu, Li, and Yi, 2014; Liu et al., 2015). Inlet temperature constraint: tzi  tzmin
The commonly used optimization algorithms are genetic Heat load constraint for reheating furnace: Qmin  Qri  Qmax
algorithm (GA), ant colony optimization algorithm (ACO), Pump power constraint: Pmin  P  Pmax
Tabu search method and so on (Chang et al., 2016). Considering Pump combination scheme constraint:

the problem of optimal operation for crude oil pipeline, a 0; pumps in parallel
Ci ¼
methodology is proposed that uses Tabu search algorithm to 1; pumps in series
optimize operation scheme of oil transmission pipeline.
In equation (4), Hini represents the inlet pressure head of the
According to the technical characteristics of crude oil trans-
mission pipelines, taking the minimum energy consumption ith pump station, m; Hinimin represents the minimum inlet
cost as the objective function, the oil pump combination scheme pressure head of the ith pump station, m; Hexi represents the
and the inlet temperature of intermediate heating station as outlet pressure head of the ith pump station, m; Hdmax
independent variables, a mathematical model of optimal represents the maximum operating pressure head, m; Htotal
operation was set up. represents the total pressure head provided by the pump
station, m; Hp represents the total pressure head loss along the
METHODS pipeline, m; tzi represents the inlet oil temperature of the ith
The Optimization Model heating station, 8C; tzmin represents the minimum inlet oil
(1) The objective function. temperature, 8C; trmin , trmax, tri represent the minimum,
The mathematical model of the objective function of maximum, actual outlet oil temperature, respectively, 8C. Qmin
operation optimization is as follow (Liu, Li, and Cai, 2019).: represents the minimum thermal load of the heating furnace,
kJ/kg; Qmax represents the maximum thermal load of the
min S ¼ Sp þ SR ð1Þ heating furnace, kJ/kg; Qri represents the actual heating
where, S represents the total energy consumption, CNY; Sp amount of the ith heating station, kJ/kg; P represents the
represents the total electricity cost, CNY; and SR represents power of pump station, W; Pmin represents the minimum power
the total fuel cost, CNY. of pump station, W; Pmax represents the maximum power of
The total electricity cost is pump station, W.
3) Characteristic curve equation of pump. The relation
XNp between the flow and the head can be written as follow:
GHgtp
Sp ¼ ed 3
ð2Þ
i¼1 10 gpi gei H ¼ A0 þ A1 Q þ A2 Q 2 ð5Þ
The total fuel cost is where, H represents the head of the pump, m; Q represents the
flow rate of the oil, m3/h; A0, A1, A2, are coefficients.
X
Nh
GcðTR  TZ Þth
S R ¼ ey ð3Þ As for pump efficiency, the relation between the flow and the
i¼1 103 gR q efficiency can be written as follow:
where, Np, Nh represent the total number of pumping stations
g ¼ C0 þ C1 Q þ C2 Q2 ð6Þ
and heating stations, respectively. ed represents the electricity
price, CNY/(kWh); ey represents the price of fuel, CNY/t; G where, g represents the efficiency the pump, C0, C1, C2, are
represents the mass flow of crude oil, kg/s; H represents the coefficients.
pressure head required between pumping stations, m; g is the
Algorithm Description
acceleration of gravity, tp represents the run time of the pump
The problem is a nonlinear optimization problem to find the
station, h; gpi represents the efficiency of the pump (the ratio of
optimal pump combination scheme and outlet temperature.
effective pump power to pump shaft power), gei represents the
Considering the characters of the model, Tabu search algo-
efficiency of the electric motor (the ratio of motor output power
rithm was chosen to solve this problem. Tabu search algorithm
to motor input power), c represents the specific heat capacity of
avoids circuitous search by introducing a flexible storage
crude oil, kJ/(kg8C); TR represent the oil temperature after
structure and corresponding Tabu criteria, and forgives some
heating, 8C; TZ represent the oil temperature before heating,
good taboo states by defying the criteria, thus guaranteeing the
8C; th represents the run time of the heating station, h;
gR represents the efficiency of the heating furnace, and q is the effective exploration of diversification and ultimately achieving
calorific value of fuel, kJ/kg. global optimization. The basic idea of the Tabu search
(2) Constraint conditions. The constraints include inlet algorithm is (Abido, 2002; Chen, Peng, and Yu, 2017; Cunha
pressure, outlet pressure, pipe strength, hydraulic constraint, and Ribeiro, 2004; Gallego, Romero, and Monticelli, 2000;
outlet and inlet oil temperature, thermal load of heating Montané and Galvao, 2006; Peng, Liao, and Tan, 2018; Su, Liu,
furnace, oil pump power, and pump combination scheme(Liu et and Xu, 2019): given an initial solution and its neighborhood,
al., 2017). then some neighbor solutions are selected. If the target value of
the neighbor solution is better than the ‘‘best so far’’, then
Inlet pressure constraint: Hini . Hinimin neglect its taboo characteristics; if there is no neighbor
Output pressure constraint: Hexi , Hdmax solution, then search other neighborhood until meet the

Journal of Coastal Research, Special Issue No. 94, 2019

 Optimal Operation of Subsea and Onshore Oil Transmission Pipeline 477

Figure 1. The flowchart of the programming.

stopping criteria. Figure 1 presents the flowchart of the increased. Figure 3 presents the viscosity-temperature
programming (Lu et al., 2019; Shen and Shi, 2016). curve of oil. Table 1 lists the pump combination of each
station, we can see that all the pumps in each station are
RESULTS AND DISCUSSION parallel, but the number and type of pump in each pump
Parameters of the Crude Oil Pipeline station are different.
The pipeline, starting at the No.1 pump station, ending at The pressure limit, maximum heat load and heating
terminal station. Pipe diameter is V52937, whose design efficiency of heating furnace at each station are shown in
pressure was 4.5 MPa and a total length of 294.9 km. As Table 2. According to the actual situation in the oil field, the
shown in Figure 2, the elevation of each station is gradually price of natural gas is 1.15 CNY/m3, the electricity price is

Figure 2. Summary of elevation and mileage of the pipeline. Note: No.1-No.6

represents the number of the pump station. Figure 3. The viscosity-temperature curve of the crude oil.

Journal of Coastal Research, Special Issue No. 94, 2019

478 Peng et al.

Table 1. The pump combination of each station.

Stations No.1 No.3 No.4 No.5 No.6

Pump Type 1 GDKS2 200D KDY1 200D 200D
Number of Pump 1 3 3 2 3
Pump Type 2 SNHG1 SNHG1 SNHG2 KDY2 GDKS3
Number of Pump 1 1 1 1 2
Pump Type 3 — SNHG2 GDKS2 — —
Number of Pumps — 1 1 — —
Connection Mode Parallel Parallel Parallel Parallel Parallel

Table 2. Pressure limit, maximum heat load and heating efficiency.

Maximum Outlet Minimum Inlet Maximum Heat

Stations Pressure (MPa) Pressure (MPa) Heat Load (kW) Efficiency (%)
No.1 4.5 0.05 10000 88
No.2 4.5 0.01 6726 85
No.3 4.5 0.05 7150 85
No.4 4.5 0.05 5000 82
No.5 4.5 0.05 4652 85
No.6 4.5 0.05 7150 82
Terminal Station — 0.05 — —

Table 3. Transportation scheme (Current and after optimization).

Current Transportation Scheme Optimized Operating Scheme

Inlet Outlet Inlet Outlet
Temperature Pressure Temperature Pressure Pump Temperature Pressure Temperature Pressure Pump
Station (8C) (MPa) (8C) (MPa) Combination (8C) (MPa) (8C) (MPa) Combination
No.1 29 0.4 45 3.97 1 GDKS2 29 0.4 46.259 2.027 1 GDKS2
No.2 33 2.93 38 2.92 — — — — — —
No.3 23 1.6 32 1.56 — 18 0.1 29.956 3.055 1 200D
No.4 19 0.13 33 3.94 1 GDKS2 18 1.67 26.241 1.67 —
No.5 21 2 35 2 — 18 0.182 27.042 3.146 1 200D
No.6 21 0.22 28 2.33 1 GDKS3 18 1.523 26.973 2.083 1 GDKS3
Terminal 18.4 0.33 18.4 — — 18 0.1 18 — —

Table 4. The power of pump stations.

Current Operating Scheme Optimized Operating Scheme

Electric Power of Power of Power of
Pump Current Voltage Pumps Heating Furnace Pump Power Heating Furnace
Station Type (A) (kV) (kW) (kW) Type (kW) (kW)
No.1 GDKS2 88 6.4 780 3911.40 GDKS2 311.586 4228.971
No.2 1249.13
No.3 2223.25 200D 660.903 2942.063
No.4 GDKS2 95 6.4 842 3591.68 2087.287
No.5 3477.97 200D 661.575 2212.714
No.6 GDKS3 70 6.3 611.1 1778.92 GDKS3 96.413 2275.864
Total 2233.1 16232.4 1730.48 13746.9

0.6564 CNY/(kW. H), and the low calorific value of natural gas After optimization, due to the drop of outlet oil temperature
is 35880 kJ/m3. The maximum outlet pressure of all the station at each station, the total hydraulic frictional resistance
except the terminal station were 4.5 MPa, and the heat increased and the unit gas cost decreased. It is to be observed
efficiency varied between 82%~88%. that the electricity cost also decreased after optimization, the
reason is that there was a throttle out of the pump, so the
Optimal Operation Results pressure after throttling is far less than the design pressure.
One service data of the pipeline was collected on February 23, The monthly total cost decreased from 2,404,277 CNY to
which mass flow was 9500 t/d, the average ground temperature 1,959,898 CNY.The total cost after optimization was 81.5%
was 78C, the inlet oil temperature at initial station was 298C. before optimization, saving 444379 CNY.
Table 3 plots the operating parameters of each station in winter
before and after optimization. Based on this current transpor- CONCLUSIONS
tation scheme, the total power of pumps at each station was A mathematical model of optimal operation for crude oil
calculated, which shown in Table 4. pipeline was set up, which using the minimum energy

Journal of Coastal Research, Special Issue No. 94, 2019

 Optimal Operation of Subsea and Onshore Oil Transmission Pipeline 479

consumption as the objective function and the pump combina- Kurak, S.K., 1989. Power optimization yields savings for products
tion scheme, outlet pressure, and outlet oil temperature as pipeline. Oil & Gas Journal, 87, 65-66.
Lin, Y.C.; Hwang, K.S., and Wang, F.S., 2004. A mixed-coding scheme
design variables.
of evolutionary algorithms to solve mixed-integer nonlinear
When Tabu search algorithm is used to solve the optimal programming problems. Computers & Mathematics with Applica-
mathematical model, there is no unified solution. According to tions, 47, 1295-1307.
the characteristics of Tabu search method, combined with the Linke, A., 1996. Computer aided computer minimization of crude oil
and product pipeline transport costs. Oil Gas European Magazine,
actual problem of crude oil pipeline operation optimization, the
15, 56-58.
key factors that affect the performance of the Tabu search Liu, C.; Ye, Y.; Sun, S.; Chen, Q., and Meng, Q., 2013. Experimental
algorithm are neighborhood, Tabu table, Tabu length, neigh- studies on the P-T stability conditions and influencing factors of
borhood solution, and aspiration criterion. These key factors gas hydrate in different systems. Science China Earth Sciences, 56,
can help us to find the optimal solution of the mathematical 594-600.
Liu, E.; Li, C., and Yi, Y., 2014. Optimal energy consumption analysis
model of oil pipeline optimization very quickly. of natural gas pipeline. Scientific World Journal, No.506138.
The operation cost before optimization and after optimization Liu, E.; Li, C.; Yang, L.; Wu, M., and Wang, D., 2015. Research on the
were compared. The total operation cost after optimization is optimal energy consumption of oil pipeline. Journal of environ-
only 81.5% as before in winter, the effect is significant. mental biology, 36, 703-711.
Liu, E.; Li, W.S., and Cai, H.J., 2019. Formation mechanism of
trailing oil in product oil pipeline. Processes, 7, 7.
ACKNOWLEDGMENTS Liu, E.; Lv, L.; Ma, Q.; Kuang, J., and Zhang, L., 2019. Steady-state
This work is supported by the Applied basic research optimization operation of the West-East Gas Pipeline. Advances in
project of Sichuan Province under grant number 2019YJ0352. Mechanical Engineering, 11, 1-14.
Liu, E.; Wen, D.; Peng, S.; Sun, H.Y., and Yang, Y., 2017. A study of
the numerical simulation of water hammer with column separation
LITERATURE CITED and cavity collapse in pipelines. Advances in Mechanical Engi-
Abido, M.A., 2002. Optimal power flow using Tabu search algorithm. neering, 9, 1-13.
Electric power components and systems, 30, 469-483. Lu, H.; Wu, X., and Huang, K., 2018. Study on the effect of
Botros, K.K., 2004. Multi-objective optimization of large pipeline reciprocating pump pipeline system vibration on oil transportation
networks using genetic algorithm. Proceedings of the Biennial stations. Energies, 11, 132.
International Pipeline Conference, 2005-2015. Lu, H.; Huang,K., and Azimi, M., 2019. Blockchain technology in the
Botros, K.K.; Sennhauser, D.; Jungowski, K.J.; Poissant, G.; Golshan, oil and gas industry: A review of applications, opportunities,
H., and Stoffregen, J., 2004. Multi-objective optimization of large challenges, and risks. IEEE Access,7, 41426-41444.
pipeline networks using genetic algorithm. International Pipeline Montané, F.A.T. and Galvao, R.D., 2006. A Tabu search algorithm for
Conference, 2005-2015. the vehicle routing problem with simultaneous pick-up and
Chang, X.; Wa, B.; Gua, X.; Cui, J., and Wang, T., 2016. A case study delivery service. Computers & Operations Research, 33, 595-619.
of crude oil alteration in a clastic reservoir by waterflooding. Mora, T. and Ulieru, M., 2005. Minimization of energy use in pipeline
Journal of Petroleum Science and Engineering, 146, 380-391. operations an application to natural gas transmission systems.
Chebouba, A., 2009. Optimization of natural gas pipeline transpor- IECON Proceedings (Industrial Electronics Conference), 2190-
tation using ant colony optimization. Computers and Operations 2197.
Research, 1916-1923. Nguyen, H.H. and Chan, C.W., 2005. Optimal scheduling of gas
Chen, Y.; Peng, X., and Yu, H., 2017. Mechanical performance pipeline operation using genetic algorithms. Canadian Conference
experiments on rock and cement, casing residual stress evaluation on Electrical and Computer Engineering, 2195-2198.
in the thermal recovery well based on thermal-structure coupling. Peng, S.; Liu, E.; Xian, W.; Wang, D., and Zhang, H., 2015. Dynamic
Energy Exploration & Exploitation, 35, 591-608. simulation of an underground gas storage injection-production
Cunha, M. and Ribeiro, L., 2004. Tabu search algorithms for water network. Journal of Environmental Biology, 36, 799-806.
network optimization. European Journal of Operational Research, Peng, S.B.; Liao, W., and Tan, H., 2018. Performance optimization of
157, 746-758. ultrasonic flow meter based on computational fluid dynamics.
Gallego, R.A.; Romero, R., and Monticelli, A.J., 2000. Tabu search Advances in Mechanical Engineering, 10, 1-9
algorithm for network synthesis. IEEE Transactions on Power Reddy, P.; Karimi, I.A., and Srinivasan, R., 2004. Novel solution
Systems, 15, 490-495. approach for optimizing crude oil operations. AIChE Journal, 50,
Gopal, V.N., 1980. Optimization pipeline Operation. Journal of 1177-1197.
Petroleum Technology, 23, 2063-2067. Su, Z.Y.; Liu, E.B., and Xu, Y.W., 2019. Flow field and noise
Hou, Y.; Wu, N.Q.; Zhou, M.C., and Li, Z.W., 2017. Pareto- characteristics of manifold in natural gas transportation station.
optimization for scheduling of crude oil operations in refinery via Oil & Gas Science and Technology-Revue D IFP Energies
genetic algorithm. IEEE Transactions on Systems, Man, and Nouvelles, 74, 1-12.
Cybernetics: Systems, 47, 517-530. Yong, C.; Tao, L., and Xu, P., 2017. Calculation and analysis of the
Jokic, A., 2001. Optimization of pipeline network for oil transport. first interface micro-gaps of the thermal production wells. Advanc-
Hungarian Journal of Industrial Chemistry, 29, 113-117. es in Mechanical Engineering, 9, 1-9.

Journal of Coastal Research, Special Issue No. 94, 2019

Copyright of Journal of Coastal Research is the property of Allen Press Publishing Services
Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv
without the copyright holder's express written permission. However, users may print,
download, or email articles for individual use.