Home Journals Articles Updates About
CN
image
收藏切换
A hindcast of the Bohai Bay oil spill during June to August 2011
收藏切换
PDF
Yiqiu YANG1, Yan LI1, *, Guimei LIU1, Qingqing PAN1, Zhaoyi WANG1
Acta Oceanologica Sinica | 2017, 36(11) : 21 - 26
Less
收藏切换
Acta Oceanologica Sinica | 2017, 36(11): 21-26
• •
A hindcast of the Bohai Bay oil spill during June to August 2011
Full
Yiqiu YANG1, Yan LI1, *, Guimei LIU1, Qingqing PAN1, Zhaoyi WANG1
Affiliations
  • 1 Marine Environmental Forecasting Division, National Marine Environmental Forecasting Center, State Oceanic Administration, Beijing 100081, China
Published: 2017-11-01 doi: 10.1007/s13131-017-1135-7
Outline
收藏切换

An operational three-dimensional oil spill model is developed by the National Marine Environmental Forecasting Center (NMEFC), State Oceanic Administration, China, and the model has been running for 9 a. On June 4 and 17, 2011, oil is spilled into the sea water from two separate oil platforms in the Bohai Bay, i.e., Platforms B and C of Penglai 19-3 oilfield. The spill causes pollution of thousands of square kilometres of sea area. The NMEFC’s oil spill model is employed to study the Penglai 19-3 oil-spill pollution during June to August 2011. The wind final analysis data of the NMEFC, which is based on a weather research and forecasting (WRF) model, are analyzed and corrected by comparing with the observation data. A corrected current filed is obtained by forcing the princeton ocean model (POM) with the corrected wind field. With the above marine environmental field forcing the oil spill model, the oil mass balance and oil distribution can be produced. The simulation is validated against the observation, and it is concluded that the oil spill model of the NMEFC is able to commendably simulate the oil spill distribution. Thus the NMEFC’s oil spill model can provide a tool in an environmental impact assessment after the event.

oil spill  /  hindcast  /  Lagrangian random walk  /  oil distribution  /  swept area
Yiqiu YANG, Yan LI, Guimei LIU, Qingqing PAN, Zhaoyi WANG. A hindcast of the Bohai Bay oil spill during June to August 2011[J]. Acta Oceanologica Sinica, 2017 , 36 (11) : 21 -26 . DOI: 10.1007/s13131-017-1135-7
Full Text
Less
1 Introduction
Less
With the increasing demand for offshore oil exploration, more petroleum has been exploited and drilled from a seabed and transported by tankers. Correspondingly the threat of possible marine oil spill accidents is increasing. Accidental marine oil-spill pollution would cause damage to marine environment, the health of mankind, fisheries and so on (Wang et al., 2008; Coppini et al., 2011). The prediction of an oil spill trajectory, a contamination area and a swept area is very important for an environmental impact assessment. And numerical modelling provides an important tool to study and predict the transport and fate of the oil spills, which would assist to plan response actions in advance and minimize the loss.
Nowadays, some groups and international community have developed oil spill models and played an important role in an oil spill emergency response. GNOME (general NOAA operational modeling environment) is a publicly available oil spill trajectory model that simulates oil movement and was developed by the National Oceanic and Atmospheric Administration (Beegle-Krause 2001). It was used to simulate the oil spill trajectory of oil released from a pipeline leaking in the Gulf of Mexico and in the North Sea (Cheng et al., 2011, 2014). OSCAR (oil spill contingency and response) is a state of the art model and simulation tool for predicting the fates and effects of oil released, which has been developed by SINTEF (Stiftelsen for industriell og teknisk forskning) (Reed et al., 1999). It has been involved in and is still in use for planning, hindcasting and forecasting of accidental releases in locations such as the North Sea and Baltic Sea, the Gulf of Mexico and the Mediterranean Sea. MEDSLIK simulates the transport of the surface slick governed by the water currents and wind. The operational implementation of the MEDSLIK oil spill model during the Lebanese oil pollution crisis assisted decision makers in drawing up action plans for the clean-up operations (Coppini et al., 2011).
An operational oil spill emergence forecasting system was established in 2007 at NMEFC (National Marine Environmental Forecasting Center, China), which is capable to predict the trajectory, the swept area, and beached oil distributions and so on. The current and wind fields forcing the oil spill model are acquired from a China’s global operational oceanography forecasting system (CGOFS) developed by the NMEFC. The Bohai Sea forecasting system that used in this event is a major component of CGOFS.
On June 4 and 17, 2011, the oil spill was observed from oil platforms B and C of the Penglai 19-3 oilfield, which was operated by Conoco Phillips-China Incorporation (COPC), located at 38.37°N, 120.08°E in the Bohai Sea. According to the COPC, a sum of approximate 723 barrels (115 m3) of oil and 2 620 barrels (416 m3) of mineral oil-based drilling mud seeped into the Bohai Sea. The average water depth of Bohai Sea is only 18 m, and water exchange capacity with the open sea is weak as a semi-closed sea. These aggravate the pollution if an oil spill occurred.
During the Penglai 19-3 oil spill accident, many researchers have forecasted the oil spill trajectories (Guo et al., 2010) and validated against satellite observations (Xu et al., 2013, 2015). Li et al. (2013a, b) analyzed the error source of spill transport modelling using the Penglai 19-3 case and identified that the wind is the major error source. The drift factor of wind and current appears to be the most relevant parameter to improve the quality of the result (Coppini et al., 2011). In this paper, a long-term simulation was carried out to analyze the swept area with the corrected wind and current, followed with validation study after the accident occurred.
2 Model and data
Less
The hindcasting scheme is illustrated in Fig. 1. We start with the correction of wind reanalysis filed by comparing with observations. And then we obtain the current filed by forcing the POM (Princeton ocean model) circulation model with the corrected wind filed. With the above marine environmental field forcing the oil spill model, the oil mass balance and the oil distribution could be produced. It will be introduced in detail in the following subsections which including the method to produce of the wind and current field, oil spill model and model parameters.
2.1 The correction of wind analysis data
The wind analysis data are obtained from a WRF (weather research and forecasting) model, where the initial filed and boundary condition obtained from National Centers for Environmental Prediction (NCEP) final (FNL) operational global analysis data (0.5°×0.5°). Meanwhile, the observation-nudging and the grid-nudging are used to assimilate the global telecommunication system (GTS) data.
Li et al. (2013a, b) found out that oil spill transport modelling is very sensitive to the accuracy of the wind filed. Thus we have studied this aspect carefully in this paper. Figure 2 shows the discrepancy between the model result (analysis wind) and the observational wind. The mean absolute error of a wind speed is 2 m/s and the mean absolute error of a wind direction is 47°. The analysis wind is corrected by the observational wind speed and direction received from station close to oil platform of Conoco Phillips (Fig. 3) during June to August. According to the linear fitting shown in Fig. 4, we set up a relationship, as shown by the following Eq. (1), between the corrected wind filed and the simulated one.
{ u=1.105  3 u a −0.383  6, v=0.816  9 v a +2.335  1,  
where ua is the simulated eastward component of the wind speed; va is the simulated northward component of the wind speed; u is the corrected eastward component of the wind speed; and v is the corrected northward component of the wind speed.
In order to present the improvement, the comparison between the wind final analysis data (corrected wind) and the observation wind is given in Fig. 5. The mean absolute error of the wind speed is 1.7 m/s and the mean absolute error of the wind direction is 43°. The improvement reduced the error of model wind outputs.
2.2 The hindcast of ocean model
In this paper, the Bohai operational 3-D current forecasting system of the NMEFC (BOCFS-NMEFC) is used to supply the current velocity, which is based on POM2k. The BOCFS-NMEFC is forced by an open boundary water level condition with the harmonic constant of tide from Japan (Matsumoto et al., 2000). Sixteen main constituents of tide are added to the open boundaries. The current forecasting model uses a uniform grid (2.5′×2.5′) in the Cartesian coordinate system. It covers a region bounded by latitudes ranging from 37° to 41°N and longitudes ranging from 117.5° to 122.5°E. The further description can be found in the paper by Liu et al. (2005). The forecasting results are available on website of the NMEFC every day from 2003. The BOCFS-NMEFC has been running stably for 12 a, and the forecasted currents and sea surface temperatures were verified every month. Tide simulation has been verified by comparing the forecasting water level with the observational water level (Liu et al., 2005).
2.3 NMEFC oil spill model
During an oil spill accident, it is most concerned in the oil trajectory, concentration distribution, swept region and so on. The oil spill model would produce the above when the initial information of spilled oil is known, such as oil type, when and where the spill occurred, oil mass leaked and duration. The NMEFC have developed its own oil spill forecasting system since 2007.
An oil spill convection and diffusion model has been widely adopted to simulate the movement of spilled oil in the marine environment, which is based on the oil particle tracking in Lagrangian frame of reference. The turbulent diffusion is handled by a random walk model. An oil parcel concept is a general approach. It is assumed that spilled oil is composed by a large number of particles with equal mass (Johansen, 1984, 1987; Elliott et al., 1986) and subject to wind, current, waves and turbulent diffusion. The velocity of oil particle is calculated as
u 0 = u c +α( u 10 cosβ− v a sinβ )+ u w + u ′  ​,
v 0 = v c +α( u 10 sinβ+ v a cosβ )+ v w + v ′ ,
w 0 = w c + w ok + w ′ ,
where u0 and v0 are the horizontal velocities of the oil particles; w0 is the vertical velocity of the oil particles. The first terms on the right hand side of Eqs (2)–(4) represent the impacts of the ocean current on the oil particle, which derived from the ocean model. uc and vc are the horizontal components of current and wc is the vertical component of current. The second terms on the right hand side of Eqs (2) and (3) represent the wind drift effects on the oil particles u10 and v10 are the wind velocities at 10 m above the sea surface. α means the coefficient of the wind drift and β is the deviation angle. According to the suggestion of Li et al. (2013a, b), α=0.02 and β=0 was chosen in this paper. uw and vw are the velocities due to nonlinear waves. It can be ignored as the nonlinear effect can be weakened when oil slick covers the sea surface, which increases the surface tension and the smoothness. wok is the vertical velocity due to buoyancy of the oil particles, which is given by the Stokes law:
{ w ok = g d 2 ( 1− ρ 0 / ρ w ) 18ν    ( d≤ d c ), w ok = 8 3 gd( 1− ρ 0 / ρ w )         (d> d c ),
d c = 9.52 ν 2/3 g 1/3 ( 1− ρ 0 / ρ w ) 1/3 ,
where d is the diameter of oil particle and dc is the critical diameter; ν=1.31 mm3/s, is the seawater viscosity; g is the gravity; ρ0=830 kg/m3 and ρ0=1 025 kg/m3, are oil and seawater density, respectively.
The turbulent velocities ( u ′, v ′, w ′ ) can be calculated with a random walk technology:
u ′ =ξ c ′ A h /Δt cos( 2πξ ),
v ′ =ξ c ′ A h /Δt sin( 2πξ ),
w ′ =ξ c ′ K v /Δt ,
where ξ is the standard Gaussian white noise with unit intensity and null mean; Ah is the horizontal diffusion; Kv is the vertical diffusion; Δt is the time step; and c′ is a constant. The horizontal diffusion is regarded as Fick diffusion, and take kve (vertical eddy viscosity) as Kv. The vertical eddy viscosity is semi empirically coefficient derived from the Reynolds stresses due to surface waves (Ichiye, 1967):
k ve =0.028 H s 2 T e −2κz ,
where Hs is the significant wave height; κ is the wave number; T is the average period; and z is the depth where the oil particle is located.
2.4 The parameter of oil spill model
The first oil spill occurred at Penglai 19-3B oil platform (Fig. 3) on June 4, 2011 and was controlled on June 16. Another oil spill occurred at the other platform Penglai 19-3C (Fig. 3) on June 17, 2011 and was in control on June 19. From July 15, some spilled oil was observed near two oil platform successively, but it was not included in model because of quick response. The duration of crude oil released is 13 d (4–16) and 3 d (17–19) when simulating the Penglai 19-3B case and Penglai 19-3C case, respectively. The model was run from June 4 and June 17 to August 31, respectively (Table 1).
Considering the inaccuracy of spilled location, all oil particles were initially distributed randomly in two circles cantered at Penglai 19-3B and Penglai 19-3C with radius of 25 m. We assume the total mass of spilled oil to be 200 t and the density to be 830 kg/m3. Fifty oil particles were released every10 min.
3 The hindcast analysis
Less
The hindcast analysis, shown in Figs 6–8, indicated that the oil from two platforms B and C moves to the northwest of the oil platform, where the Bohai Bay, and the Liaodong Bay were affected successively. The influenced sea area is about 7 800 km2 including most of the Bohai Bay, the Laizhou Bay and the Bohai Strait. The affected area is located in the northwest of the Bohai Sea, including coastal cities (Qinhuangdao, Leting, Changli, Dongdaihe) of Hebei Province and Liaoning Province. The total beached length is about 160 km as shown in the red color along coast of Fig. 8.
The hindcast results indicated that the spilled oil first arrives shoreline on June 29 near Tangshan and then reachs to the Dongdaihe Coast on July 11. It was consistent with the findings from oil fingerprinting identification of spilled oil from the Penglai 19-3 platform. The place where the oil beached is consistent with the one reported, but the model gave an earlier date than actual one. The time when the two separated oil spill beached the shoreline is different in the hindcast analysis. Oil from 19-3B arrived Tangshan on June 29 and reached to Dongdaihe 12 d later (July 11). Oil from 19-3C arrived at Tangshan coast before July 9. At the end of the simulation period, i.e., on August 9, almost all oil reached to the coast. Although the convection direction of oil spilled from two platforms are same, the swept area and effected coastline are different. Oil spilled from 19-3B drifted further to the north and effected longer coastline and larger sea area than oil spilled from 19-3C (Figs 6 and 7).
Figure 9 shows the polluted region reported in the China Marine Environment Quality Bulletin from June to August in 2011. It is mapped using monitoring data which distributed in the Bohai Sea. The area in dark brown represents that the water is polluted heavily, and it mainly scatters around the oil platform and some appears in the northwest of the platform. The area in yellow represents the slightly polluted water, which is located in the northwest of the platform.
Comparing the simulated swept area in Fig. 8 with the observations in Fig. 9, it is found as follows: the pattern of two regions is very similar; the direction and extent of oil transport are the same; the direction of oil convection is northwest of the Penglai 19-3 the platforms in the simulation; but there is oil slick observed travelling to the south of platforms, which is not reflected in the hindcast results. The simulated polluted region is 7 800 km2, which is larger than observed 6 200 km2. Because there is not enough monitoring sites in the Bohai Sea to observe oil pollution, the model result is always larger than the observation.
Some of early predictions may be in error due to the lack of observational data (Proctor et al., 1994). We can conclude the hindcast result is in accord with the observation. And the long-term hindcast simulation could reproduce the transport of spilled oil in reality. Thus the NMEFC’s oil spill model could provide a tool in an environmental impact assessment after the event.
4 Conclusions
Less
In this paper, a long-term hindcast simulation for the Bohai oil spill accident in 2011 is carried out to study the behaviour and fate of the spilled oil in the marine environment. The wind field plays an important role in trajectory modeling, thus linear fitting is adopted to correct the wind data. It is a key factor to force the oil spill mode with the corrected wind and current in order to reproduce the oil distribution. The long-term hindcast simulation produced a polluted sea area of about 7 800 km2 and polluted coast of about 160 km. Owing to the lack of monitoring, the polluted area is larger than the observation and the contaminated coastline where concentrated more than 50% poulation of province (Yang et al., 2012) is also not reflected in the observation. But the long-term hindcast simulation provied the polluted sea area, the contaminated coastline, the time of oil beached, especialy for serious oil spill accident.
Long time predictions may be in error due to the accumlation error of a numerical integration during a model calculation and it can not represent the real scene of oil spill. The long-term hindcast which assimilated observation data can reduce the modeling error.
In general, the long-term hindcast simulation proposed in this paper could provide scientific bases for government in the environmental impact assessment after the event. It is concluded that the hindcast result is consistent with the observation.
Funding
Less
  • The Open Project Fund of the Key Laboratory of Shangdong Province for Marine Ecological Environment and Disaster Prevention and Mitigation of China under contract No. 201402; the Project of the Key Laboratory of Marine Spill Oil Identification and Damage Assessment Technology, State Oceanic Administration of China under contract No. 201604; the National Marine Public Welfare Research Project of China under contract No. 201305031; the Cooperation on the Development of Basic Technologies for the Yellow Sea and East China Sea Operational Oceanographic System funded by the China-Korea Joint Ocean Research and the National Natural Science Foundation of China under contract Nos 41206106 and 41406042.
References
Less
Beegle-Krause J. 2001. General NOAA oil modeling environment (GNOME): a new spill trajectory model. Krause, 2001(2): 865–871
Cheng Yongcun, Li Xiaofeng, Xu Qing, et al. 2011. SAR observation and model tracking of an oil spill event in coastal waters. Marine Pollution Bulletin, 62(2): 350–363
Cheng Yongcun, Liu Bingqing, Li Xiaofeng, et al. 2014. Monitoring of oil spill trajectories with COSMO-SkyMed X-band SAR images and model simulation. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 7(7): 2895–2901
Coppini G, De Dominicis M, Zodiatis G, et al. 2011. Hindcast of oil-spill pollution during the Lebanon crisis in the Eastern Mediterranean, July–August 2006. Marine Pollution Bulletin, 62(1): 140–153
Elliott A J, Hurford N, Penn C J. 1986. Shear diffusion and the spreading of oil slicks. Marine Pollution Bulletin, 17(7): 308–313
Guo Mingxian, Shi Ping, Liu Xin. 2010. A three-dimensional transport model for oil spills in coastal waters based on particle approach. ICNC, 213(2): 381–411
Ichiye T. 1967. Upper ocean boundary-layer flow determined by dye diffusion. Physics of Fluids, 10(9): S270–S277
Johansen O. 1984. The Halten Bank experiment observations and model studies of drift and fate of oil in the marine environment. In: Proceedings of the 11th Arctic Marine Oil Spill Program (AMOP) Technology Seminar. Canada: Environment Canada, 18–36
Johansen Ø. 1987. DOOSIM-A new simulation model for oil spill management. In: Proceedings of 1987 International Oil Spill Conference. Washington DC: API Publication, 529–532
Li Yan, Zhu Jiang, Wang Hui. 2013a. The impact of different vertical diffusion schemes in a three-dimensional oil spill model in the Bohai Sea. Advances in Atmospheric Sciences, 30(6): 1569–1586
Li Yan, Zhu Jiang, Wang Hui, et al. 2013b. The error source analysis of oil spill transport modeling: a case study. Acta Oceanologica Sinica, 32(10): 41–47
Liu Qinzheng, Zhang Jianhua, Liu Yu, et al. 2005. Numerical forecast for the 3-dimensional sea temperature and current in the Bohai Sea. Marine Forecasts, 22(suppl): 27–34
Matsumoto K, Takanezawa T, Ooe M. 2000. Ocean tide models developed by assimilating TOPEX/POSEIDON altimeter data into hydrodynamical model: a global model and a regional model around Japan. Journal of Oceanography, 56(5): 567–581
Proctor R, Elliot A J, Flather R A. 1994. Forecast and hindcast simulations of the Braer oil spill. Marine Pollution Bulletin, 28(4): 219–229
Reed M, Ekrol N, Rye H, et al. 1999. Oil spill contingency and response (OSCAR) analysis in support of environmental impact assessment offshore Namibia. Spill Science & Technology Bulletin, 5(1): 29–38
Wang Shoudong, Shen Yongming, Guo Yakun, et al. 2008. Three-dimensional numerical simulation for transport of oil spills in seas. Ocean Engineering, 35(5–6): 503–510
Xu Qing, Cheng Yongcun, Liu Bingqing, et al. 2015. Modeling of oil spill beaching along the coast of the Bohai Sea, China. Frontiers of Earth Science, 9(4): 637–641
Xu Qing, Li Xiaofeng, Wei Yongliang, et al. 2013. Satellite observations and modeling of oil spill trajectories in the Bohai Sea. Marine Pollution Bulletin, 71(1–2): 107–116
Yang Yiqing, Zuo Juncheng, Du Lin, et al. 2012. Cost-benefit analysis of adaptation to sea level rise in major vulnerable regions along the coast of China. In: Proceedings of the Twenty-second (2012) International Offshore and Polar Engineering Conference. Rhodes, Greece: National Bureau of Economic Research, 1522–1528
Appendix
Less
Year 2017 volume 36 Issue 11
PDF
34
19
Cite this Article
BibTeX
Article Info
doi: 10.1007/s13131-017-1135-7
  • Receive Date：2016-12-21
  • Online Date：2026-04-16
  • Published：2017-11-01
Article Data
Affiliations
History
  • Received：2016-12-21
  • Accepted：2017-06-20
Funding
The Open Project Fund of the Key Laboratory of Shangdong Province for Marine Ecological Environment and Disaster Prevention and Mitigation of China under contract No. 201402; the Project of the Key Laboratory of Marine Spill Oil Identification and Damage Assessment Technology, State Oceanic Administration of China under contract No. 201604; the National Marine Public Welfare Research Project of China under contract No. 201305031; the Cooperation on the Development of Basic Technologies for the Yellow Sea and East China Sea Operational Oceanographic System funded by the China-Korea Joint Ocean Research and the National Natural Science Foundation of China under contract Nos 41206106 and 41406042.
Affiliations
    1 Marine Environmental Forecasting Division, National Marine Environmental Forecasting Center, State Oceanic Administration, Beijing 100081, China

Corresponding:

References
Share
https://castjournals.cast.org.cn/joweb/aos/EN/10.1007/s13131-017-1135-7
Share to
QR

Scan QR to access full text

Cite this article
BibTeX
Citations
表12种不同金属材料的力学参数
科
Family		 属数
Number of
genus		 种数
Number of
species		 占总种数比例
Percentage of
total species (%)		 属
Genus		 种数
Number of
species		 占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏
  • BibTeX
  • EndNote
  • RefWorks
  • TxT
Links
Articles
Latest Articles
Most Read
Collections
Updates
Events
News
Multimedia
About
About Us
Contact
No. 86 Xueyuan South Road, Haidian District, Beijing
100081
010-62199257
qkjq@cast.org.cn

Copyright © 2025 China Association for Science and Technology. All rights reserved. For all open access content, the relevant licensing terms apply.
Sponsored by the Office of the Leading Group for Cybersecurity and Informatization of CAST, and supported by Science and Technology Review Publishing House