Distribution of oil spill response capability through considering probable incident, environmental sensitivity and geographical weather in Vietnamese waters

ABSTRACT

Various occurrences of marine incidental oil spills in the Vietnamese waters require effective high response system. The probability of oil spill, environmental sensitivity, and geographical weather conditions were simultaneously considered to calculate the distribution of the oil spill response capability (OSRC). In this contribution, 13 important factors (3 priority protected areas, marine traffic, max. amount spill, industrial installation, offshore activities, historical spill, distance to spill, sea state, tidal current, visibility and sea-water temperature) of 8 oil spill response (OSR) areas were investigated and calculated by using Fuzzy Analytic Hierarchy Process (Fuzzy AHP). The results were reflected the highest OSRC in Vung Tau area, peaking about 1,457 tons of the amount of Vietnam national marine OSRC. The study suggested the new scientific platform for the distribution of areal OSRCs to mitigate the damage of future marine incidental oil spills to marine ecosystems, coastal resources, humans, and socio-economy.

KEYWORDS:

• OSRC
• incidental oil spills
• fuzzy-AHP
• Vietnamese waters

Introduction

Oil as an important factor impacting on economic development and human’s life, global seaborne trade of oil by tankers, oil exploration and production industries has been increasing (EIA 2017; UNCTADStat n.d.2017; Reynolds 2014; Behmiri et al. 2013), contributing to great benefits to various localities, regions and nation (Alidi and Al-Faraj 1994; Garrett et al. 2017; PetroVietnam 2017). However, the risk of oil spill influence on marine ecosystems and coastal resources (Carla et al. 2015) gave rise to losses of the fisheries, aquaculture and tourism industries and consequences on social and environmental values that concerned by researchers and policymakers (ITOPF, The International Tanker Owners Pollution Federation Limited 2014a, 2014b; Alvarez et al. 2014; Garza-Gil et al. 2006). In order to minimize damage from oil spill incident, various contingency plans, policies, legislation, and effective countermeasure would be established.

Vietnam is located on the East Sea where more than half of the world’s annual merchant tonnage, one-third of the global trade in crude oil and over half of the LNG passed through. The oil and gas exploration and production activities in Vietnam are increasing sharply and playing a vital role in national economy, about 16–18% of the State budget (petroVietnam 2017). The demand for national energy consumption to support economic development has been sharply climbing (Le et al. 2016). It is well-known that characteristics of Vietnamese sea are various and complex such as Ha Long Bay, Cat Ba Island with many marine and coastal ecosystem (Hong et al. 2008; Nguyen, Jacobson, and Ross 2017).

The Neptune Aries pollution incident in 1994, the tanker Formosa One pollution incident in 2001 and the mystery oil spill in 2007 are important lessons in preparing and responding to large oil pollution accidents (ITOPF 2017), resulting in the establishment of Circular 2262/TT-MTG and a series of laws and policies. Vietnam’s Government has built three National Oil Spill Response Centers. There are 24/28 Oil spill contingency plans of Coastal Provinces and various Contingency plans in-house level that considering about historical spill have approved. There were some solutions that have been proposed to improve oil spill response capability such as response measures for oil spills in Da Nang City (Phung 2005), enhancing local capacities in oil spill preparedness and response (Nguyen 2009; Nguyen et al. 2014). Oil spill response capability (OSRC) is defined as the resources to deal with the oil spill incident, including three categories: equipment, response personnel, and additional support (IPIECA-IOGP 2015). However, oil spill response equipment, facilities, materials, and manpower resources and additional support in Vietnam still cannot be reasonably allocated and used, and there is a lack of scientific basis for allocating oil spill response resources to combat oil spills. Therefore, the scientific and systematic researches on factors impacting on the distribution of OSRC to comprehensively evaluate national OSRC are necessary to be performed. Hence, it provides scientific basis and guidance for the distribution OSRCs and gives rise to scientific suggestions in process of construction of national oil spill response capability.

In this study, 13 important factors were investigated that were divided into three categories: oil spill sensitive areas, the probability of oil spill incident, and geographical environment conditions. It is well known that fuzzy – AHP model is good decision-making method that was applied to solve the different managerial problem (Asghari et al. 2017; Isaai et al. 2011; Shaverdi, Heshmati, and Ramezani 2014). The fuzzy linguistic approach can take into account the optimum and pessimism rating attitude of decision makers. The membership functions are characterized by fuzzy triangle number and are assessed preference ratings instead of conventional numerical equivalence measures (Shaverdi, Heshmati, and Ramezani 2014). Therefore, Vietnam response capability (VRC) model was developed by using Fuzzy AHP technique to calculate performance weight of important factors. These weight and statistics are unified into one value to suggest new areal OSRC for each region.

Research methodology

Identifying factors and development of hieratical VRC model

According to the current situation of marine oil spill response capability in Vietnam and the experience of developed countries in marine oil spill response capability in-house and oversea, including Guidance on Oil Spill Risk Evaluation and Assessment of Response Preparedness of IMO, Sensitivity mapping for oil spill response of IPIECA, Guidance to State and Local Governments in oil spill response of USA, Geographic Response Plan Oil spill Sensitive Sites, the Regulations of Vietnam National Assembly on Environmental Protection, the Law on Petroleum, the Law on Resources and Environment of Sea and Islands, Vietnam Maritime Code, the Law on Inland waterway traffic, the Law of the Sea of Vietnam, Law on Water resources, the Decision No. 02/2013/QD-TTg of Prime Minister, dated 14 January 2013, to promulgate the regulation on oil spill response as well as other Laws, regulations, standards and guidelines of Vietnam on marine oil spill pollution, we set up an evaluation important factor system, which includes 3 factors in the upper level and 13 factors in the lower level (Figure 1 and Table 3).

Table 1. Linguistic responses and fuzzy scale transformation chart (Saaty 2008; Buyukozkan, 2004).

Linguistic scale	TFN scale	TFN reciprocal scale
Just equal	(1,1,1)	(1,1,1)
Equal Importance	(1/2,1,3/2)	(2/3,1,2)
Weakly more important	(1,3/2,2)	(1/2,2/3,1)
Strongly more important	(3/2,2,5/2)	(2/5,1/2,2/3)
Very strongly more important	(2,5/2,3)	(1/3,2/5,1/2)
Absolutely more important	(5/2,3,7/2)	(2/7,1/3,2/5)

Table 2. Values of the Random Index (RI) (Saaty and Vargas 2001).

n	1	2	3	4	5	6	7	8	9	10
RI	0	0	0.52	0.89	1.11	1.25	1.35	1.40	1.45	1.49

Table 3. Detailing important factors for evaluation of VRC Index.

Goal	Upper level	Lower level	Detail
VRC	Oil spill sensitive area protected level: OSS	High priority protected areas (HPA)	Marine protected areas (World heritage sites, national park, marine nature reserves, etc.) Widlife protected include marine bird (wetlands as waterfowl habitat in Ramsar site), marine mammals and sea turtles (Sea turtles spawning ground and habitat area). Coastal habitat (consists of mangrove and mudflat forest, coral reefs, seagrass beds)
			Aquaculture (fish stocks, fish and shellfish farming, seaweed cultivation, water intakes for hatchery, etc.)
			Oil port area
		Medium priority protected areas (MPA)	Beach, tourism and recreation areas
			Salt field
			Intensive land of rice and vegetables coastal area
			Subsistence, artisanal and commercial fishing, and fishing villages
			Main port and marinas
			Urban residental area, rocky beach, sandy beach for many users
		Low priority protected areas (LPA)	National park, nature reserves, habitat on land, industrial land
			Land for planting industrial trees, perennial trees, coastal area a little used
			Small port and marinas
	Probability of oil spill incident: POS	No. Vessel enter and depart from area (VED)	Marine traffic density
		Estimation of maximum amount of oil spill in the areas (EMO)	Type of oil
		Industrial installation (IND)	Oil storage tanks
			Type of oil import-export area
		Offshore operational oil port (OSP)	Oil rigs, platforms operations, SPM, SPSOs
			Oil and gas pipeline system
		Historical oil Spill incident (HOS)	Spill frequency
			Amount of oil spill
	Geographic condition (seasonal, meteorological condition); GEO	Distance (DIS)	Farthest point
		Sea state (Wind, wave) (SSW)	Direction
			Intensity
		Tidal current (TID)	Direction
			Intensity
		Visibility (VIS)	Daylight
			Darkness
		Temperature (TEM)	Average highest temperature
			Average lowest temperature

Figure 1. Hierarchical VRC model.

This hierarchical model is developed to analyse and estimate the OSRC with Fuzzy-AHP methodology. The VRC model can calculate separately probability of oil spill incident, oil spill sensitive area to marine and coastal ecosystems, and geographical weather conditions according to season. Then in the lower level of the model, the oil spill sensitive area has been divided into three main factors: high priority protected areas, medium priority protected areas, and low priority protected areas. The probability of oil spill incident has been divided five main factors: the number of vessels enter and depart from area, estimation of the maximum amount of oil spill in the areas, industrial installation, offshore operational ports and historical oil spill incidents. Moreover, the geographical weather conditions have been divided five main factors: distance to spill site, sea state (wind, wave), tidal current, visibility, and sea-water temperature.

The aim of this study is providing the best balance of important factors for the distribution of OSRC in Vietnamese waters that support policymaking, plan-making, response personnel to mitigate impacts of an oil spill.

Fuzzy AHP methodology

In order to avoid various failures when defining new precise of approximation and vague characteristics of decision-making issues Zadeh (1965) first proposed on the generalization of standard set theory to fuzzy sets, which oriented the rationality of uncertainty due to vagueness (Kahraman, Cebeci, and Ruan 2004). In this study, F-AHP method is found that will be used to find the final weight factors for distribution of areal OSRCs.

The analytic hierarchy process (AHP) is one of the most frequent techniques for decision-making, a multi-criteria decision-making tool (MCDM) was initially introduced by Saaty (1972, 1987), and has been used to solve multiple criteria decision issues in different fields such as political, socio-economic and management sciences, planning, resource allocations, optimization, etc. (Lee, Chen, and Chang 2008; Vaidya and Kumar 2006). Using a systematic hierarchy structure, complex estimation criteria can be represented clearly and definitely (Chou et al. 2001). However, this technique is inability to assign exact numerical values to the comparison judgments and effective when applied to unstructured and ambiguous problems. Generally, the conventional AHP still cannot provide full guidance on the highly vague world. Therefore, AHP has been integrated with various techniques such as linear program, quantity function deployment, fuzzy logic, etc. (Saaty and Vargas 2001). In order to achieve the desired goal in a better way, the benefits of all methods should be combined. In this case, some scholars have combined the Fuzzy theory with AHP method to handle fuzzy comparison matrices such as Chang’s extend analysis method (Chang 1996).

Fuzzy numbers and fuzzy synthetic extent

The fuzzy set is a class of objects with a continuum of grades of values ranging from 0 to 1. Complete non-membership is represented by 0, and complete membership as 1. The values lie between 0 and 1 belongs to the fuzzy sets (Zadeh 1965). Fuzzy number ɱ on R to be a triangular fuzzy number (TFN) with three points if its membership function of ɱ is equal to Eq. (1).

(1)

Considering two triangular fuzzy number ɱ₁ = (l₁, m₁, u₁) and ɱ₂ = (l₂, m₂, u₂) that can be operated according to their laws as below:

(2)

(3)

(4)

(5)

The value of fuzzy methods has considered as a linguistic value that the definite values are converted to fuzzy numbers according to the definitions in Table 1.

Algorithm of fuzzy AHP

The steps of FAHP that the extent of FAHP is utilized, which was originally introduced by Chang as below.

Let X = {x₁, x₂, …, x_n} be an object set and U = {u₁, u₂, …, u_n} be a goal set. And then, each object is taken and performed extent analysis for each goal, respectively. Hence, m extent analysis values for each goal can be gotten with the following signs:

(6)

where all the (j = 1, 2, …, m) are triangular fuzzy numbers and g_i is the corresponding to the goal.

Step 1. The value of fuzzy synthetic extent with respect to the ith object is defined as:

To obtain perform the fuzzy addition operation of m extent analysis values for a particular matrix such as

(8)

In order to obtain, the fuzzy addition operation of (j = 1, 2, …, m) value is performed as follows:

(9)

And the inverse of the vector in Equation. (9) is computed as

(10)

Step 2. As M₁, M₂ are two triangular fuzzy numbers, the degree of possibility of M₁ = (l₁, m₁, u₁) ≤ M₂ = (l₂, m₂, u₂) is defined as (Kahraman, Cebeci, and Ruan 2004)

(11)

(12)

where d is the ordinate of the highest intersection point between and , and hgt is height. The numbers M₁ and M₂ should be compared calculating both, and values.

Step 3. The degree possibility of a convex fuzzy number to be greater than k convex fuzzy M_i (i = 1, 2,…, i) numbers can be defined by

(13)

Assume that

(14)

And then the weight vector is given by

(15)

Where A_i (i = 1, 2, …, n) are n elements.

Step 4. Via normalization, the normalized weight vectors are

(16)

where W is a non-fuzzy number.

Saaty (1977) has proposed a consistency index (CI), which is relatedto the eigenvalue method applied in matrix, for which:

(17)

where n: dimension of the matrix and λ_max = maximal eigenvalue

Saaty (1980) suggested that if CI/RI < 0.1, the pairwise comparison matrix is characterized by an acceptable level of consistency. RI is a random index (the average CI of 500 randomly filled matrices), the values of which have been predefined by Saaty (Saaty 2001) for problems with n 10 as indicated in Table 2.

From the theoretical basis above, a pair-wise comparison tool is set up by using the Visual C # tool programming (see Figure 3). In order to check the consistency the matrix, it is required to calculate the consistency ratio:

(18)

(19)

It means that FAHP sequencing result is of high consistency, the distribution of weight coefficient is very reasonable. Figure 2 shows the flowchart of this study. 

Figure 2. Flowchart of setting up OSRC.

Figure 3. Sample pair-wise comparison matrices.

Results and discussion

Calculating weight factor using F-AHP

In this study, we conducted questionnaires to “pool experts” as advisory group, consisting of individuals with expertise in the following areas: navigation, ship technology, naval architecture, chemical hazards, marine biology, physical and chemical oceanography, dispersion modeling, spill response, and firefighting who have at least over 5-years experiences. These experts are asked to compare two main criteria and 13 important factors in the range of this study as the table above. The results of the survey are analyzed by using FAHP measures that are presented as below. The first step of the analysis, set up the pair-wise comparison matrix to compare the relative importance between various factors that obtains the relative importance of criteria factor from the result of questionnaires, the definite values are converted to fuzzy numbers (see Table 1). Table 3 shows the fuzzy evaluation matrix, local weights and pair-wise comparison of factors, including 3 factors: OSS, POS, GEO in the upper level, 13 important factors: HPA, MPA, LPA, VED, EMO, IND, OSP, HOS, DIS, SSW, TID, VIS, and TEM in the lower level. These fuzzy values are compared (for example in Figure 3).

The weight factor was determined as W_P = (0.300, 0.273, 0.156, 0.151, 0.120). In the five factors, VED is the most preferred factor when compared with the other factor. The other main factors and sub-factors are calculated in a similar way. The weight of upper evaluation and local weight were multiplied to produce the global weights, as shown in Table 4.

Table 4. Weighted scores for each factor by Fuzzy-AHP.

Upper evaluation		Lower evaluation	Local weight	Global weight
POS	0.619	VED	0.300	0.186
		EMO	0.273	0.169
		IND	0.156	0.097
		OSP	0.151	0.093
		HOS	0.120	0.074
OSP	0.236	HPA	0.521	0.123
		MPA	0.312	0.074
		LPA	0.167	0.039
GEO	0.145	DIS	0.325	0.047
		SSW	0.263	0.038
		TID	0.231	0.033
		VIS	0.082	0.012
		TEM	0.059	0.009

Regional status and normalization

Probability of oil spill incident

The five factors used to estimate probability of oil spill incident in Table 5 are the number of vessel enter and depart from area (2015), estimation of maximum amount of oil spill in the area (2017), industrial installation (2017), offshore operational oil port (2017) and historical oil spill incident (1995–2016). Generally, the five factors showed the highest point in S1 (Binh Thuan – Ba Ria Vung Tau – Tien Giang – Ben Tre – Tra Vinh – Soc Trang), which is the central area in which oil and gas exploration and largest sea-ports system of Vietnam are concentrated. The summed data values of area and region are shown for each factor through average normalization step (see Table 8).

Table 5. Present status of areal probability factors.

Region	Area	VED (No.)	EMO (Tons)	IND (Tons)	OSP (No.)	HOS (Case)
NOSRCEN	N1- Hai Phong	26810	6000	2362100	5	10
	N2- Thanh Hoa*	7643	45000	1125400	3	3
	N3- Ha Tinh+	2202	2700	158000	4	2
SOSRCEM	C1 – Da Nang	7160	4500	409500	8	12
	C2 – Quang Ngai+	8030	22500	1281900	5	15
	C3 – Khanh Hoa	4894	22500	1124900	6	2
NASOS	S1 – Vung Tau	41405	45000	3513860	116	66
	S2 – Ca Mau+	591	18000	65000	4	7

Table 6. Present status of areal oil spill sensitivity factors.

Region	Area	High priority area (HPA)			Medium priority area (MPA)						Low priority area (LPA)
		Marine protected area (km^2)	Aquaculture (km^2)	Oil port (No.)	Beach, tourism and recreation areas (ha.)	Salt field (ha.)	Intensive land of rice and vegetables coastal area (ha.)	Subsistence, artisanal and commercial fishing, and fishing villages (tons)	Main port and marinas (No.)	Urban residental area, rocky beach, sandy beach for many users (People)	National park, nature reserves, habitat on land, industrial land (km^2)	Land for planting industrial trees, perennial trees (km^2)	Small port and marinas (No.)
NOSRCEN	N1- Hai Phong	3567.91	60.5	26	24130	196	429.5	241359	3	6848	18096	25.5	3
	N2-Thanh Hoa*	67.00	48.5	6	23750	600	818.7	211912	2	7586.9	263118	149.7	2
	N3- Ha Tinh+	0.00	12.5	7	4938	0	158.4	76369	1	2144.4	200003	35.1	2
SOSRCEM	C1 – Da Nang	587.44	19.2	11	18635	0	196.4	163160	2	4307.2	543747	70.9	3
	C2 – Quang Ngai+	243.85	9.4	9	6346	315	235.8	438,098	2	3675.5	35937	69.2	2
	C3 – Khanh Hoa	738.26	5.9	5	13450	2371	81.3	176056	3	1815.2	68843.5	35.5	1
NASOS	S1 – Vung Tau	631.54	177.2	32	51572	3300	1021.9	952670	4	15970.1	45974	163.3	7
	S2 – Ca Mau+	15451.63	576.7	5	57007	2300	1063.5	802270	0	3885.5	17806.7	5.9	3

Table 7. Present status of areal geographic conditions.

Region	Area	Distance to spill site (Nm)	Sea state (average wind intensity)	Tidal current (max velocity – cm/s)	Visibility	Temperature (water surface – °C)
NOSRCEN	N1- Hai Phong	110	2.475	10	1700	19
	N2-Thanh Hoa*	62	1.800	5	1659	20
	N3- Ha Tinh+	96	2.000	2	1761	21
SOSRCEM	C1 – Da Nang	86	1.775	1	2061	24
	C2 – Quang Ngai+	170	1.800	1	2359	25
	C3 – Khanh Hoa	55	2.400	10	2606	25
NASOS	S1 – Vung Tau	150	2.400	30	2667	25
	S2 – Ca Mau+	180	1.933	10	2321	27

Table 8. Resulting average normalization of present status data.

Region	Area	HPA	MPA	LPA	VED	EMO	IND	OSP	HOS	DIS	SSW	TID	VIS	TEM
NOSRCEN (0.2932)	N1	0.1638	0.1088	0.0638	0.2715	0.0361	0.2353	0.0331	0.0893	0.121	0.1492	0.1449	0.0992	0.1180
	N2	0.0386	0.1234	0.1924	0.0774	0.2708	0.1121	0.0199	0.0000	0.068	0.1085	0.0725	0.0968	0.1242
	N3	0.0277	0.0324	0.1059	0.0223	0.0162	0.0157	0.0265	0.0000	0.106	0.1206	0.0290	0.1027	0.1304
SOSRCEM (0.2847)	C1	0.0525	0.0677	0.2379	0.0725	0.0271	0.0408	0.0530	0.1071	0.095	0.1070	0.0145	0.1203	0.1491
	C2	0.0370	0.0776	0.0806	0.0813	0.1354	0.1277	0.0331	0.1339	0.187	0.1085	0.0145	0.1377	0.1553
	C3	0.0302	0.1037	0.0550	0.0496	0.1354	0.1120	0.0397	0.0179	0.061	0.1447	0.1449	0.1521	0.0000
NASOS (0.4221)	S1	0.1804	0.2947	0.2123	0.4194	0.2708	0.3500	0.7682	0.5893	0.165	0.1447	0.4348	0.1556	0.1553
	S2	0.4697	0.1917	0.0520	0.0060	0.1083	0.0065	0.0265	0.0625	0.198	0.1166	0.1449	0.1355	0.1677

Oil spill sensitive area

The three factors used to estimate oil spill sensitivity area are the high priority area, medium priority area, and low priority area (see Table 6). The high priority is estimated through the marine protected area (2016–2017), aquaculture (2016), and oil port (2017). The medium priority includes beach, tourism and recreation areas (2016), salt field (2017), intensive land of rice and vegetables coastal area (2016), subsistence, artisanal and commercial fishing, and fishing villages (2016), main port and marinas (2017), urban residential area, rocky beach, sandy beach for many users (2016). The low priority area was estimated through the national park, nature reserves, habitat on land, industrial land (2016), land for planting industrial trees, perennial trees, coastal area a little used (2017), small port and marinas (2017). These factors reflect the sensitive characteristics of each area and used in the normalization step (see Table 8).

Geographical weather conditions

The five factors used to estimate geographical weather conditions according to tropical monsoon weather area distance from OSR center to spill site (2015–2017), sea state, tidal current, visibility (hours of daylight/year), and temperature of surface seawater (<10 m) according to QCVN 02: 2009/BXD. The average normalization of these data is shown in Table 7).

Figure 4. A scenario of an oil spill response in Thanh Hoa area.

The result of each distribution was determined by a total of multiplying the normalization valuation and global weight of each factor of each area by fuzzy AHP (formula 17). The final weight of each area to oil spill response capability in Vietnam are shown in Table 9 and Figure 4. Practically, new OSR bases (N2, N3, C2, S2) reach the important percentage of national OSRC. The highest of OSR base in national OSRC is available OSR in S1, accounting for 31.67%. It means that Vung Tau area has the highest level of oil spill response capability in the country. This reflects the fact that the highest number of vessels entering and departing Vung Tau – Ho Chi Minh City area as the biggest economic center, concentrating most of the oil and gas exploitation activities of the country. In addition, it is also an area of high environmental sensitivity with priority protected areas such as Con Dao Island, Can Gio Mangrove, Hon Cau – Vinh Hai, Binh Chau – Phuoc Buu, etc.

(20)

where:

• A_i: Result of Area i

• N_ij: Normalization evaluation of factor j of area i

• GW_ij: Global weight of factor j of area i.

The statistics of recovery capacity of oil skimmer in three OSR centers (VINASARCOM 2017; PCDN 2016; PCVT 2016) combined with applying Korea’s formula (Lee 2001; Min-Jae 2014) to calculate the existing capabilities in Vietnam: , 1764 tons by NOSRCEN, 961 tons by SOSRCEM and 1875 tons by NASOS.

Looking on Table 10, the results of OSRC are absolutely different from the existing capabilities. Because of the areal OSRC are more sensitive and effective to areal protected priority compare with the existing standard. The existing excessive OSRC of some areas was adjusted to areal protected priority and to the central areas. Therefore, the results of OSRC reflect the impacts of an oil spill considering cause of incident, impacts of geographical weather condition and are more balanced than the existing capability.

Table 9. Final weight for setting up OSRC for areas.

Region	Area	HPA	MPA	LPA	VED	EMO	IND	OSP	HOS	DIS	SSW	TID	VIS	TEM
NOSRCEN (0.2837)	N1	0.0305	0.0184	0.0062	0.0252	0.0027	0.0289	0.0024	0.0035	0.0057	0.0057	0.0048	0.0012	0.0011
	N2	0.0072	0.0209	0.0187	0.0072	0.0200	0.0138	0.0015	0.0000	0.0032	0.0041	0.0024	0.0012	0.0011
	N3	0.0052	0.0055	0.0103	0.0021	0.0012	0.0019	0.0020	0.0000	0.0050	0.0046	0.0010	0.0012	0.0012
SOSRCEM (0.2394)	C1	0.0098	0.0114	0.0231	0.0067	0.0020	0.0050	0.0039	0.0042	0.0045	0.0041	0.0005	0.0014	0.0013
	C2	0.0069	0.0131	0.0078	0.0076	0.0100	0.0157	0.0024	0.0052	0.0088	0.0041	0.0005	0.0017	0.0014
	C3	0.0056	0.0175	0.0053	0.0046	0.0100	0.0138	0.0029	0.0007	0.0029	0.0055	0.0048	0.0018	0.0000
NASOS (0.4769)	S1	0.0336	0.0498	0.0206	0.0390	0.0200	0.0431	0.0568	0.0230	0.0078	0.0055	0.0143	0.0019	0.0014
	S2	0.0874	0.0324	0.0050	0.0006	0.0080	0.0008	0.0020	0.0024	0.0093	0.0044	0.0048	0.0016	0.0015

Table 10. Final weight and OSRC (4 days).

Region	Area	Final weight	OSRC (tons)	Current capability (tons)	Difference
NOSRCEN	N1	0.1376	633	1764	(‒)1131
	N2	0.1035	476	0	476
	N3	0.0426	196	0	196
	Sum.	0.2837	1305	1764	(‒)459
SOSRCEM	C1	0.0786	362	961	(‒)599
	C2	0.0852	392	0	392
	C3	0.0756	348	0	348
	Sum.	0.2394	1101	961	140
NASOS	S1	0.3167	1457	1875	(-)419
	S2	0.1602	737	0	737
	Sum.	0.4769	2194	1875	318

Table 11. Regional OSRC for the worst case discharge in North Region.

Region	Area	Distance (NM)	Time dispatching (hour)	Time operating (hour)	OSRC Simulation(tons)	Existing capability
NOSRCEN	N1- Hai Phong	100	9	28	554	1764
	N2-Thanh Hoa*	0	0	32	476	0
	N3- Ha Tinh+	60	5	28	171	0
SOSRCEM	C1 – Da Nang	241	22	21	237	961
	C2 – Quang Ngai+	295	27	17	208	0
	C3 – Khanh Hoa	506	46	13	141	0
NASOS	S1 – Vung Tau	690	63	8	364	1875
	S2 – Ca Mau+	850	94	2	46	0

Table 12. Gathering OSRCs in Thanh Hoa area.

Hour working		1	2	3	4	5	6	7	8
Cumulative amount of oil recovery (ton)	Day 1	15	30	45	60	101	142	183	235
	Day 2	287	351	415	479	543	607	671	735
	Day 3	799	863	938	1013	1088	1163	1238	1313
	Day 4	1434	1555	1676	1797	1918	2039	2183	2327

Validity of OSRC

According the results of areal OSRC on water, we verified by simulating the distribution of oil spill response equipment. It was assumed 2000 tons of oil, the target recovery amount of oil spill in Thanh Hoa at 13:00. In Thanh Hoa, there is a high risk of oil spill and most disadvantage in term of mobilization of response resources (Figure 4). Because of the fact that the OSR equipment is not available at Thanh Hoa area. In this simulation, marine OSR equipment was dispatched to spill site with the speed vessel was assumed to be 11 knots, weather conditions were assumed to be good. The daytime working period was assumed to be 8 h (07:00 to 17:00). The response operation was started immediately upon OSR equipment entering at spill site. In cases of equipment arrived on spill site at night, operation was assumed to start at 7:00 the following day.

In this scenario, the actual OSRC is estimated about 2,327 tons/h when considering the dispatching OSRC of other areas to Thanh Hoa area (Table 12). It is really satisfy with the existing OSRC of NORCEN, about 1764 tons (Table 11).

Conclusion

This paper proposed the new criteria for the distribution of OSRC that is a platform in order to establish Vietnam national OSRC. The present status of 8 areal OSR according to 13 important factors was investigated and normalized. The global weight was calculated by using fuzzy AHP approach that helps to choose the best decision – making strategy using a weighting process through pair-wise comparison matrices. The global weight for each factor was combined with the results of normalization step to produce the final weight for each OSR area. Interestingly, Vung Tau area (S1) of southern region is estimated to be around 1457 tons, ranking the first of national OSRC. In the northern region, Hai Phong area (N1) is made up of 633 tons, followed by Thanh Hoa area (N2) and Ha Tinh area (N3), about 476 tons and 196 tons, respectively. In Quang Ngai (C2) and Da Nang (C1) account for 392 tons and 362 tons, respectively. The result of this research is the balanced probability of oil spill, sensitive environment of OSR area, and geographical weather conditions according to tropical monsoon to the central high risk of marine pollution areas.

It is expected that this research would make any contribution to improving the efficiency of OSRC, supporting to OSR responders, policy-making, contingency plan–making, and the damage of marine pollution incidents petering out.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Ministry of Oceans and Fisheries.