6.1. Research Specificities
As stated above, the research specificities’ scores of the civil engineering laboratories are relatively high. The highest score is 3.788, indicating that most of the teachers agree with the importance of the experimental partners. If an unidentified risk cannot be managed, it will tend to expand into a major problem. If the operator and the experimental partner cannot recognize the risk at this time, it is quite possible to ignore the occurrence of relevant risk events. If the experimental partner understands the risk, he/she will always reminds the operator, or points out the risk factors in time when the operator is not operating properly, so that the operator will pay attention to the risk [
56]. Therefore, civil engineering laboratory managers need to strengthen the popularization of the knowledge of laboratory risks to experimenters to reduce the possibility of the risks when conducting experiments or to find and correct mistakes in time to avoid potential risks when others carry out experiments [
34].
The frequent replacement of operators leads to incomplete records of relevant experiment of laboratory safety, which also has a high score of 3.737. Not only for teachers’ scientific research, civil engineering laboratories in universities are also used for scientific research experiments conducted by postgraduate and undergraduate students. However, as a teaching and scientific research laboratory, the operators change frequently in batches. At the same time, the relevant personnel did not record some possible risks encountered in their experiments [
56], which is not conducive to the improvement of the corresponding management measures. Therefore, for the experiment with high risk, special experimenters should be set up to assist in the experiment and record the relevant experiment safety experience [
62]. For the general experiment, scholars should be encouraged to record their own relevant experience—thus improving laboratory safety management.
Though the scores for other aspects are relatively low, they however cannot be ignored. Therefore, it is suggested that some incentive measures for the operators should be provided by the laboratory managers [
63]. The safety responsibility hierarchy for some complex and changeable projects can be set up and the safety responsibility can be implemented in place [
64].
6.2. Human Unsafe Behaviors
Among the risk factors of human unsafe behaviors, A2 (insufficient safety awareness of operators) has the highest LCI score of 7.097, which needs to be managed first. The scores of A1 (insufficient capacity of operator) and A3 (poor physical and psychological conditions) were relatively low (6.984 and 6.921 respectively). WF and S of these three risk factors are relatively high while the scores for the other parameters are relatively low, as shown on the left of
Figure 1.
(1) As shown on the right of
Figure 1, when A2 occurs, loose management will accelerate the occurrence of the risk. The operators of civil engineering laboratories in universities are mostly students who are highly mobile. Before the experiment, the teacher’s emphasis on the relevant experimental precautions will affect the operator’s cognition of the laboratory risks. However, most of the operators in industrial laboratories are highly qualified professionals who are more familiar with the experiments and experimental environment. As a result, they have a higher risk perception and a stronger safety awareness [
55]. Therefore, when students ignore these precautions and teachers do not remind students about them, they become likely to engage in unsafe operation behaviors, thus resulting in increased risks [
35]. Moreover, the students’ safety attitude will change after the intervention of laboratory safety education in a short time. Therefore, it is necessary to strengthen operators’ safety education and training in civil engineering laboratories to improve their safety awareness compared with industrial laboratories [
34,
65], which can strengthen the safety attitudes of students in the short term and improve their safety awareness during the experimental process.
Meanwhile, for A2, the score of S (severity) was higher (6.712). The reason is that man is the leader of the experiment. When man’s safety awareness is not strong, individual protection will be ignored [
25], and the compliance of individual protection equipment will generally decrease with the drop of safety awareness. If protective measures are not in place, the injury of operators will be worse [
55]. In addition, compared with chemical laboratories where operators are required to wear appropriate laboratory coats, operators in civil engineering laboratories do not have special clothing [
66,
67]. Therefore, the operators should wear appropriate individual protective equipment when entering the laboratory [
68].
(2) For A1 and A3, WF had higher scores. When the operator has low ability, the probability of risks caused by the unsafe state of equipment is more likely to increase than that caused by deficiencies in man and management, as shown on the right of
Figure 1. If the operator is inexperienced, she or he will not understand how to use the equipment and become unskilled in operation and insensitive to the related risks [
33]. When the physiological and psychological conditions are not positive, the effects of object, management, and environment are similar. It is difficult for the operator to concentrate on the experiment and keep paying attention to the experimental process due to excessive fatigue, depression, or poor physical condition. At such a time, if there is any influence of object, management, or the environment, the deterioration of the result will be accelerated [
69].
Based on the investigation, for A1 and A3, the scores of S are lower than that of A2, since the operator will be more cautious in the experimental process when he/she is inexperienced and may be assisted in the experiment. However, when the operator is excessively fatigued, with a low mood or poor physical condition, he/she usually only performs some auxiliary and risk-free tasks, such as recording data. Relatively speaking, in addition to the ability of laboratory personnel, industrial laboratories have standardized operating procedures, while the exploratory experiments in civil engineering laboratories are irregular, novel and complex [
11]. Meanwhile, doing experiments is not a student’s daily routine. Therefore, students are required to keep in mind the experimental precautions before the experiment. When the experiment is more complex, the senior students should help and students can improve their experimental ability. In addition, when the operator’s physiological or psychological state is poor, it is not suitable for them to carry out complex experiments by force.
(3) For A1, A2, A3, HD (including A, R, Se) and POA (including HPt, FD), scores are relatively low, indicating that human unsafe behaviors is easy to monitor and unlikely to occur. Because in the process of the experiment, people’s unsafe behavior can be corrected in time through peer’s reminders [
70], the person who can operate in the laboratory must have received certain knowledge training, and the person who does not have certain knowledge will not be allowed to experiment blindly [
35]. For undergraduate students, most civil engineering experiments have well-defined operation procedures. Therefore, the possibility of accidents caused by human unsafe behaviors is relatively low.
6.3. Unsafe State of Objects
Among the risk sources of unsafe state of objects, the LCI score of B2 (danger due to equipment failure) is the highest of 7.142, requiring priority management. The LCI score of B1 (danger in the test product or mechanical equipment) is relatively low at 7.006. For B1 and B2, the WF score is high and the S score is average, as shown on the left of
Figure 2.
(1) As shown on the right of
Figure 2, man and management have a great influence on the WF of B2. There is an interaction between man and object. Improper operation of the operator will aggravate the risk caused by equipment failure [
71]. We found that when people’s risk awareness or abilities are not enough to detect equipment failures, accident rates will be accelerated [
72]. Similarly, inadequate supervision of equipment safety and incomplete regular inspection will also increase the possibility of accidents [
73]. The equipment itself is responsible for most accidents which are relevant to equipment in civil engineering laboratories. Environmental changes do not have substantial impacts on the risk—leading to the low impact of the environment on B2 [
74]. Therefore, the management should regularly check the civil engineering laboratory equipment, strictly implement the instrument operation specifications, and control the risk point to the lowest level [
68,
73,
75].
In terms of B2, the score of severity is 6.314, which is average. The use of old machinery and equipment without basic protection components by operators will indirectly lead to equipment failure [
76]. Therefore, to some extent, regular equipment maintenance and supervision in civil engineering laboratories are also critical.
(2) For B1,
Figure 2 shows that the impacts of man and management are not very different. The impact of the environment is the lowest among these three factors. Most of the equipment in civil engineering laboratories is relatively large. For example, the size of the general shaking table is 8 m × 10 m [
77], and the size of a galvanized steel tank for model pile test is 975 mm × 695 mm × 680 mm [
78]. For other instrument-related or chemical laboratories, the equipment is rarely used and the volume is much smaller. When there is danger in the experimental materials or mechanical equipment, the operator’s improper operation or the absence of warning signs will undoubtedly increase the probability of accidents [
72,
79]. Therefore, it is necessary to standardize the operators’ behavior and place the danger warning signs within the visible range. In the aspect of severity, according to the on-site investigation, we find that most civil engineering laboratories are confirmatory experiments, most of which operate the equipment in the form of groups with specified operating procedures. To some extent, the risk of the equipment is thus controllable.
(3) For B1 and B2, HD (including A, R, Se) and POA (including HPt and FD) have relatively low scores, as shown in
Figure 2. In comparison, the HD of B2 has higher R and Se scores which means a lower ability can be identified and the risk is higher. It may be related to the failure of safety valve, leakage caused by aging and insulation of equipment, instability, unreliable power supply, etc., which are difficult to observe with the naked eye [
80].
6.4. Management Factors
Among the risk sources of management factors, the LCI score for C1 (imperfect management policies) was found to be the highest (7.047), thus requiring priority management. The LCI scores of C2 (with capital risk) and C3 (with weak execution of safety management) are relatively low, which are 6.907 and 6.855 respectively. The WF scores of C1 and C3 are higher while the scores for all parameters of C2 are not high, as shown on the left of
Figure 3.
(1) As shown on the right of
Figure 3, for C1, the analysis of WF shows that man and object have significant impact on C1, and their scores are relatively close. When the management policy is incomplete, the management will not consider the quality of the equipment. If the equipment is damaged at this time, it will inevitably worsen the risk. Similarly, the lack of management of students can accelerate the occurrence rate of hazards. Therefore, the management should improve the laboratory safety management system, pay attention to the status of the equipment, and strengthen the supervision of the equipment and the management of students [
18,
72].
(2) On the WF of C3, the impact levels of man, object, and environment were similar. If the management does not pay attention to the implementation of rules and regulations, the operators will relax the daily maintenance of machinery and equipment, as well as the pursuit of a clean environment. This will aggravate the impact of risks caused by poor management [
81]. Therefore, the management should strengthen its supervision while improving the laboratory safety management system [
82].
(3) Compared to C1 and C3, the scores of all parameters of C2 are not high. It can also be found that the scores of C1, C2, and C3′s S, HD, and POA are not high, which may be due to the systematic and perfect management policy formulated after various considerations, and the special department has been set up before the implementation. Therefore, it is easy to detect risk before the risk occurs, and the risk is not likely to occur due to the control of management personnel at all levels [
83]. However, the management still needs to reserve sufficient funds for the construction of the laboratories and ensure that the funds are used in an appropriate manner [
28].
6.5. Laboratory Environment
Among the risk sources of the laboratory environment, D5 (complex floor conditions) has the highest LCI scores of 7.182, and thus needs to be prioritized. The scores of D1 (strong noise vibration), D2 (adverse temperature and humidity), D3 (bad ventilation), and D4 (poor lighting) are relatively low, which are 6.826, 6.642, 6.864, and 6.739 respectively. For D5, the scores for WF, S, and HPt are relatively high. For D1, D2, D3 and D4, the scores for WF are the highest. Meanwhile, the scores for S of D3 and D1 and the scores for HPt of D3 are relatively high, as shown on the left of
Figure 4.
(1) As shown in
Figure 4, for the WF of D5, when the floor conditions are complex, problems in man and management will accelerate the occurrence of risks. For civil engineering laboratories, there will be many different students doing experiments according to the needs of the course which is different in industrial laboratories where the number of operators are fixed. Meanwhile, some civil engineering laboratories do not have much space. If the operator cannot well clean the laboratory, the environment of the civil engineering laboratories will become very messy, which will affect the quality of the operator’s experiments and easily cause accidents [
84,
85]. Similarly, management is also an invisible driving force for a clean laboratory environment. Because the laboratory space itself is small, if there are many items on the ground but the management is poor, the ground environment will become more complex [
86]. The unsafe state of the objects has slightly lower impact on it, though it cannot be ignored. Because there are many large-scale mechanical devices in civil engineering laboratories. If the ground is smooth, then the operator is prone to mechanical injury and collision. That’s why the scores for D5′s S is higher. Therefore, this kind of accident cannot be ignored, but it can be effectively reduced by maintaining a clean and tidy working ground [
35]. As a result, the operator should ensure the suitability of the laboratory environment before and pay attention to the comfort of the experimenter before and after the experiment. On the one hand, operators should try their best to avoid disorderly behaviors in the laboratory, keep their laboratories clean and tidy after the completion of their experiments, and dispose of experimental waste in time [
35]. On the other hand, the waste management of the laboratories should be mandatory [
87]. The management of the laboratory should be regularly checked. Management should implement a reward and punishment system and punish people for disorderly placing materials. It is also necessary to publicize the adverse effects of this aspect to the teachers, take keeping the laboratories clean as a significant index for the assessment of the person in charge of each civil engineering laboratory, and let them carry out the requirements of this aspect when students do the experiment. The rewards and punishments of industrial laboratories are linked to their individual wages. Similarly, the specific reward and punishment measures of civil engineering laboratories can be considered to link with students’ usual performance. The person in charge of an industrial laboratory has absolute leadership over the staff, and it is easier to achieve job management, but this is very different from that of a university [
88]. The person in charge of each civil engineering laboratory is fixed who have no direct relationship with the students, while the lab teachers are mobile and they have a direct teaching relationship with the students. Thus, it is necessary for laboratory leaders to participate in the students’ usual performance evaluation. Moreover, when the laboratory is constructed, the floor needs to be antiskid [
41]. Compared to the impact of shoe type on antiskid performance, the rough floor has a better anti-sliding effect [
89].
For the HPt of D5, the various civil engineering experiments with complex process increase the probability of risk occurrence. For example, when conducting the relevant experiments of clayey soil, the operator’s hands will be stained with clayey soil. The operator might also ignore whether the experimental environment was made to be messy.
(2) The WF scores for D1, D2, D3, and D4 are the highest, as shown in
Figure 4. In the WF, the impact of man, object, and management on the environment is relatively low because the environment is a relatively low-risk source compared to the others, which also coincides with the low LCI value of most environment calculated in
Table 7. The score of D1′s S is higher, which may be due to the psychological pressure, fidgety, fatigue, and the decrease of experimental efficiency caused by strong noise vibration [
90]. The noise produced by some experiments in civil engineering laboratory (such as grinding aggregate by ball grinder, crushing aggregate by jaw crusher and cutting concrete by saw) exceeds the limit value specified in the standard compared with chemical and biological experiments, etc. [
91]. Hence, it is essential for civil engineering laboratories to strengthen noise management. This requires the laboratory to place sound-proof boards or sound-absorbing or sound-proof cotton and other materials in partition walls. The operators should wear hearing protection equipment when necessary [
92], such as earplugs (such equipment should also be kept in the laboratory). For D3, the S and HPt are higher, which may be related to the toxic and harmful gases often produced by some civil engineering specialties (such as building environment specialty) when conducting the experiments. If the ventilation conditions are not good, the poisoning can easily occur. Ventilation is a key priority in laboratory design [
50], but if the exhaust ventilation equipment is placed on the wall near the pollution source, it will effectively reduce the risk [
48]. Other risk factors should be emphasized, e.g., for laboratories with poor temperature and humidity, air conditioning should be set in the laboratory. This can effectively improve the indoor temperature and humidity conditions, as well as the comfort of the operator [
93]. Furthermore, improving the lighting in the learning environment can also inspire students to learn. Therefore, it is recommended to use the combination of sunlight and artificial light to provide lighting in the laboratory [
51].
(3) From D1 to D5, the HD (including A, R, Se) and FD are relatively low. The relatively low level of HD maybe due to that the poor environment can be found by the operator (which can be identified), however whether the operator can take measures to change this dangerous environment depends on the operator himself.