Quantitative Risk Analysis of Dangerous Goods Road Transport Vehicles and Networks

Abstract

The risk analysis of dangerous goods is the main basis for making all decision-making of dangerous goods transportation. The accuracy and practicability of risk analysis model directly determine the level of dangerous goods transportation management. On the basis of summarizing the research results of risk analysis of dangerous goods transportation at home and abroad, a quantitative analysis model of dangerous goods transportation suitable for China’s national conditions was proposed. The risks caused by non-traffic accidents and black spots were included in the model for the first time, and TransCAD was used to apply the model to the risk analysis of actual road networks.

13.1 Introduction

In recent years, with the rapid development of China’s social economy, the promotion role of transportation in social and economic development is more and more obvious, especially the rapid development of highway construction in the guarantee of China’s rapid and healthy economic development has played an important role. After more than a decade of development, China had 149,600 km of expressways open to traffic by the end of 2019, basically realizing inter-regional connectivity and forming a road network structure. With the formation of expressway network structure, its operation and management mode is gradually changing from single line relatively independent operation to multi-line comprehensive operation, forming a new situation of expressway network operation. Compared with the previous single operation, the characteristics of highway network operating mainly in the road network scale expands unceasingly, the obvious contradiction between road network traffic capacity and traffic demand, road traffic accidents and severe degree is high, total local congestion spread to the entire network operating system the linkage effect is more outstanding, operational risk increased. Therefore, how to improve the stability, reliability, and risk control ability of the highway network to ensure its safe and efficient operation has become a prominent issue in the development of the highway while constantly strengthening the construction of the highway network and improving the capacity and efficiency of traffic flow.

From the point of the present study, most of the existing highway safety analysis method without considering the operating characteristics of highway into mesh distribution after risk for road network and traffic flow dynamic effect is still confined to sections and intersections as evaluation objects [1, 2], with accident statistics as evaluation index [3, 4], such as principal component analysis and analytic hierarchy process (AHP) method as the evaluation method of after evaluation [5, 6], multi-scale and multi-level representation is not from network real-time risk profile evolution.

In view of this, risks caused by non-traffic accidents and black spots will be included in the model for the first time in this paper, and TransCAD will be used to apply the model to risk analysis of real road networks.

13.2 Risk Analysis Model

On the basis of studying the existing risk analysis models and drawing lessons from the more common theoretical models in the world at present, this paper puts forward a more suitable grading model for regional dangerous goods road transport, in order to adapt to the lack of basic data in China. The model consists of three parts: risk of death, risk of vehicle loss, and risk of road loss.

The risk analysis model is:

$$ D_{r} = \alpha D_{P,r} + \, + \beta D_{C,r} + \chi D_{R} , $$

(13.1)

$$ D_{P,r} = P_{r} \times C_{P,S} $$

(13.2)

$$ D_{C,r} = P_{r} \times C_{C,S} $$

(13.3)

$$ D_{R,r} = P_{r} \times C_{R,S} $$

(13.4)

where D_r: risk value of section r; D_{P, r}: risk value of death in section r; D_{C, r}: risk value of vehicle loss in section r; D_{R, r}: risk value of road loss in section r; P_r: probability of dangerous goods accident in section r; C_{P, S}: cost of death caused by dangerous goods accidents caused by accident type S; C_{C, S}: vehicle loss cost caused by dangerous goods accident caused by accident type S; C_{R, S}: road repair costs caused by dangerous goods accidents caused by accident type S; α, β, χ are 0 or 1.

According to the quantitative risk analysis model of dangerous goods transport, the dangerous goods transport accident rate P_r includes the dangerous goods accident rate P_{F, S} caused by accident type S. P_{N, S}, the accident rate of dangerous goods caused by non-traffic accidents; the accident rate of dangerous goods caused by accident black spots is composed of P_{P, S} three parts.

$$ P_{r} = P_{F,S} + P_{N,S} + P_{P,S} $$

(13.5)

where P_{F, S} represents the probability of dangerous goods accidents caused by traffic accidents of accident type S in this section. It is obtained by multiplying the national accident frequency and the average running time of section r with the length of section r and the accident probability of accident type S.

$$ P_{F,S} = f_{F,S} \times P_{S} \times t_{r} $$

(13.6)

$$ f_{F,f} = r_{r} \times L_{r} \times n_{r} $$

(13.7)

$$ r_{r} = r_{0,r} \prod\limits_{j = 1}^{8} {h_{j} } $$

(13.8)

where t_r: the average running time of section r; $f_{F,f}$: accident frequency of section r; r_r: the expected frequency of road section r accidents; L_r: length of road section r; n_r: the number of vehicles in section r; r_{0, r}: the national accident frequency; h_j: coefficient of regional enhancement or weakening.

The dangerous goods accidents caused by non-traffic accidents refer to the transportation accidents of dangerous goods caused by reasons other than traffic accidents. The typical dangerous goods accidents caused by non-traffic accidents include leakage of safety valve and leakage in the device. Overload of storage tank or failure of storage tank may lead to material leakage, thus leading to dangerous goods accidents. The failure of pressure relief valve and bursting disk will also lead to the occurrence of dangerous goods accidents under general operating conditions.

Generally speaking, the type of non-accidental leakage is largely related to the duration of use, so the failure rate is expressed by the time of use or the number of operations rather than the distance traveled. The accident rate of dangerous goods caused by non-traffic accidents P_{N, S} was:

$$ P_{N,s} = \eta_{tr} $$

(13.9)

η: the failure rate of single vehicle caused by non-accident; tr: the running time of dangerous goods transport vehicle in section r.

Section accident black spot in a long period of time, the frequency or number or characteristics of road traffic accidents are significantly more prominent than other normal road sections (or locations). It is mainly caused by the horizontal and vertical combination of roads, cross-sectional form, pavement structure form, traffic control mode, natural and climatic conditions, landscape performance, etc. The number of black spots and accident frequency was obtained by statistics. The model of dangerous goods accident rate of accident black spot is as follows:

$$ P_{P,S} = \sum\limits_{k = 1} {f_{k,r} } $$

(13.10)

$f_{k,r}$: statistical accident frequency of the kth accident black spot in section r.

The death caused by road transportation accidents of dangerous goods can be divided into three categories: (1) on road death $N_{r,s}^{{{\text{in}}}}$, mainly passengers, is the direct victim of the accident consequences; (2) out of road deaths $N_{r,s}^{{{\text{off}}}}$, mainly residents and floating people along the line, the size of the impact range is related to the type of dangerous goods and transportation volume, and the probability of people appearing outdoors $P_{r}^{{{\text{out}}}}$ and the risk mitigation effect of indoor a_r are also considered; (3) Personnel n of population gathering centers along the line $N_{r,s}^{{{\text{cent}}}}$, such as schools, hospitals, and shopping center. In the process of establishing the model, the human risk of population concentration center is included in the risk of off-road personnel. The establishment of death risk model is based on the number of deaths and the damage model of roads and vehicles. The model is as follows:

$$ D_{P,r} = D_{P,r}^{{{\text{in}}}} + D_{P,r}^{{{\text{off}}}} $$

(13.11)

Among them: $D_{P,r}^{{{\text{in}}}}$: risk value of on road death in section r; $D_{P,r}^{{{\text{off}}}}$: risk value of death outside road in section r.

The risk model of on-the-road personnel death caused by dangerous goods accidents is calculated by multiplying the number of dead people in the road by the death cost of a single person as follows:

$$ D_{P,r}^{{{\text{in}}}} = V_{P} \times N_{r,s}^{{{\text{in}}}} $$

(13.12)

Among them, V_P is the death cost of a single person; $N_{r,s}^{{{\text{in}}}}$: the number of on road deaths caused by dangerous goods accidents caused by accident S in section r;

When the dangerous goods accident occurs, all passengers within the radius of death will die. The number of passengers is determined by the number of vehicles within the radius of death and the average number of passengers. The model is as follows:

$$ N_{r,s}^{{{\text{in}}}} = N_{r,C} \times K $$

(13.13)

$$ N_{r,C} = 2Q_{r} \times R \times V $$

(13.14)

Among them: N_{r, C}: the number of vehicles within the death radius of dangerous goods accidents in section r; K: the average riding factor; Q_r: the total amount of dangerous goods transportation in section r; R: the death radius caused by the accidents of dangerous goods of unit mass; V: the traffic density of section r.

When a traffic accident occurs on the road, the accident point will affect the upstream, downstream and object traffic flow density. In order to more accurately express the change of traffic density in this process, the traffic density V of section r in the model is expressed as the sum of upstream, downstream and object traffic flow density after dangerous goods accident:

$$ V = V_{{{\text{up}}}} + V_{{{\text{down}}}} + V_{{{\text{opp}}}} $$

(13.15)

Among them, V_up is the upstream traffic density of the accident lane; V_down is the downstream traffic flow density of the accident lane; V_opp is the opposite traffic flow density of the accident lane.

There are two types of off-road deaths caused by dangerous goods accidents, one is the nearby residents and floating population, the other is the people in the population gathering center with relatively concentrated personnel such as schools, hospitals, and shopping centers. The risk model of off-road death caused by dangerous goods accident is as follows:

$$ D_{P,S}^{{{\text{off}}}} = \left( {N_{r,s}^{{{\text{off}}}} + N_{r,s}^{{{\text{cent}}}} } \right) \times V_{P} $$

(13.16)

Among them, $D_{P,S}^{{{\text{off}}}}$ is the risk value of off-road death caused by dangerous goods accidents caused by accident S in section r; $N_{r,s}^{{{\text{off}}}}$ is the number of off-road deaths caused by dangerous goods accidents caused by accident s in section r; $N_{r,s}^{{{\text{cent}}}}$ is the number of deaths in the off-road population gathering area caused by dangerous goods accidents caused by accident S in section r.

The number of out of road deaths caused by dangerous goods accidents caused by the accident S in section r refers to the death toll within the death radius of dangerous goods accidents except the area outside the road, that is, the death toll within the influence area outside the road. The specific model is as follows:

$$ N_{r,s}^{{{\text{off}}}} = \delta \times d_{p,r} \times \left( {P_{r}^{{{\text{out}}}} + (1 - P_{r}^{{{\text{out}}}} ) \times \alpha_{r} } \right) $$

(13.17)

$$ \delta = \left( {\pi (Q_{r} \times R)^{2} - 2Q_{r} \times R \times W_{r} } \right) $$

(13.18)

where δ is the impact area outside the road; d_{p, r}: the population distribution density outside the road of section r; $P_{r}^{{{\text{out}}}}$: the probability of the affected personnel appearing outdoors in section r; α_r: the risk mitigation coefficient of indoor personnel in section r; W_r: the road width of section r.

The risk model of vehicle damage caused by road transportation accident of dangerous goods is the product of the total number of vehicles within the death radius caused by the total amount of dangerous goods transportation accident and the average cost of each vehicle:

$$ D_{C,S} = V_{c} \times N_{r,C} $$

(13.19)

The risk model of road damage caused by road transportation accidents of dangerous goods is the product of the road cost per unit length and the length of death diameter caused by the total amount of dangerous goods transported:

$$ D_{R,S} = V_{R} \times 2Q_{r} \times R $$

(13.20)

Among them: V_c: unit vehicle cost; V_R: unit length road repair cost.

13.3 Conclusion

The risk value of a road section mainly depends on the accident rate of the road section and the personnel, vehicles, and roads in the consequence of the accident. Due to the different attributes of the road sections, the accident rates of the road sections calculated by the national average road accident rate often differ greatly. The distribution of the off-road population in the same district is the same, and the consequence of the risk of accident mainly depends on the population distribution on the road. That is, it depends on the value of traffic flow of each section.