The most in-depth and practical analysis to date of the “3·21” explosion accident in Xiangshui (5)

(1) Research and Development

Many companies have their own R&D departments. From a safety perspective, during the process‑development phase, two key priorities are: first, to identify the hazards of chemicals—such as whether raw materials, intermediates, and final products exhibit unusual toxicity or instability—and avoid selecting substances with particularly high risks; second, to thoroughly understand the hazards associated with chemical reactions, especially their exothermic characteristics. If, at the R&D stage, inherently safer design principles can be applied whenever possible and preventive measures are implemented early to mitigate risks, this will significantly enhance safety in subsequent industrial production.

All of the above are specific technical tasks. To ensure their proper implementation, companies can establish a process‑safety assessment framework during the R&D phase, covering how to screen chemicals and process routes, how to evaluate risks, and how to transfer the safety information gathered during R&D to downstream design and manufacturing teams.

(2) Design

From a safety perspective, the most critical tasks during the design phase are conducting an in-depth process hazard analysis and implementing the design‑improvement measures identified in that analysis.

Typically, a What‑if analysis can be employed during the preliminary design phase to conduct an initial process hazard analysis, followed by a detailed process hazard analysis in the detailed design phase. The latter should address at least two key aspects: (1) Conduct a high‑quality HAZOP analysis based on the Piping and Instrumentation Diagram (P&ID) with control points; for accident scenarios with unacceptably high risks, a semi‑quantitative Layer of Protection Analysis (LOPA) should also be performed to ensure that each potential scenario is safeguarded by adequate safety measures, reducing risk to an acceptable level. Throughout this process, it is necessary to determine whether a Safety Instrumented System (SIS) is required and, if so, to specify its SIL rating. (2) Complete a facility layout risk analysis, paying particular attention to high‑occupancy areas—such as the central control room and administrative offices. Consequence analyses can be carried out using software like ALOHA (a free tool) to ensure that these densely populated zones are not severely damaged in the event of an explosion, thereby providing occupants with viable escape routes. If such high‑occupancy areas are located too close to hazard sources—measured in terms of blast overpressure—it may be necessary to relocate them; if relocation is not feasible, appropriate blast‑resistant design measures must be implemented to mitigate consequences.

During the design phase, additional tasks can be undertaken, such as: reviewing the alarm and interlock systems of process units; conducting human factors analyses to optimize layout and design for ease of operation and reduction of operator errors; performing release‑capacity calculations for pressure‑relief devices like safety valves; carrying out thermal‑radiation assessments for flare‑equipped units and relocating heat‑exposed operating platforms to safe zones; reviewing ventilation requirements for areas handling toxic substances; and conducting specialized reviews of combustible‑dust‑related process systems to prevent dust explosions. Furthermore, if other facilities are located nearby, it is essential to identify their specific installations and assess the potential impacts on your facility in the event of an accident.

Enterprises should establish process hazard analysis management elements and clearly define the specific requirements for conducting and reviewing process hazard analyses of process systems at stages such as design and operation.

The foregoing are some industry best practices; during the design phase, it is also necessary to comply with relevant regulatory requirements, including preparing a dedicated safety‑design section, among other tasks, which will not be discussed in detail here.

(3) Construction

During the construction phase, in addition to ensuring operational safety, the primary focus is on rigorously managing the manufacturing and installation quality of critical equipment and pipelines—work that is essential to maintaining the integrity of mechanical systems. This includes supervising the fabrication process and implementing robust quality‑control measures, such as nondestructive testing and hydrostatic pressure testing. Furthermore, functional tests must be carried out on instrumentation and control systems, as well as on safety‑instrumented interlock loops.

Before transitioning from the construction phase to the production‑operation phase, a pre‑startup safety review shall be conducted to systematically verify that the process system’s construction and installation comply with the design specifications and that all conditions for safe startup have been met. Enterprises are required to establish pre‑startup safety review as a key management element.

During the construction phase, if design changes occur, a hazard analysis must also be conducted for the modified portions to prevent the introduction of new hazards.

In principle, during the design and construction phases, these safety‑related tasks are under the overall responsibility of the project leader—either the project director or the project manager.

(4) Operation

This is the factory’s routine management phase, which spans a long period. It requires extensive hazard identification and risk control efforts, as well as the establishment of mechanisms for continuous improvement. During the operational phase, the key elements involved are personnel, equipment, and work activities. By implementing and effectively applying a set of safety management components (or systems), the factory can ensure safe operations and sustainable performance. Whether it is referred to as a safety standardization system or a specific safety management system, the name itself is not critical; what truly matters is that there are clearly assigned responsibilities for managing newly emerging hazards—such as through change management—and for ensuring that measures to eliminate or control those hazards are properly implemented and enforced.

Many of our enterprises are working hard to manage safety and are very busy, yet their efforts remain inefficient. A key reason is that they fail to focus on the most critical priorities. At our current stage of development, the top priority is preventing catastrophic accidents—those that could result in one or more fatalities. Therefore, we must prioritize accordingly, with a particular emphasis on making process safety management the cornerstone of our safety efforts.

In the past—and even today—we have tended to treat safety as a holistic concept. Within the industry, it is common practice to manage process safety using the same approaches and tools employed for operational safety. While this approach requires considerable effort, it often falls short of effectively controlling process safety, yet the consequences of process‑related incidents—such as fires, explosions, or toxic releases—are typically far more severe. Operational safety management focuses on regulating human behavior, whereas process safety relies primarily on engineering controls to eliminate hazards and mitigate risks. The two approaches differ significantly in their underlying philosophies and areas of focus. To fundamentally prevent catastrophic accidents, greater emphasis must be placed on process safety management.

During the operational phase, certain key elements must be rigorously addressed. Many companies invest heavily in a wide array of management components but fail to focus their efforts on the most critical processes, resulting in diminished returns for the effort expended. For some pivotal factors, even well‑established systems can leave room for accidents if their implementation lacks thoroughness. Below, I will highlight these critical elements and discuss them in detail:

Hazard analysis. During operations, the scope of hazard analysis is broad, encompassing both process hazard analyses and their periodic reviews every few years—conducted using methods such as HAZOP, What‑If analysis, and LOPA—as well as job‑safety‑related hazard analyses performed with techniques like Job Safety Analysis (JSA) or Job Hazard Analysis (JHA). It also includes general hazard identification for the work area, such as risks inherent in the working environment and hazards associated with chemicals. Process hazard analysis is of paramount importance, aiming to eliminate hidden hazards within the process system; superficial inspections will fail to uncover these concealed risks. Unfortunately, many organizations do not give this task the attention it deserves, with some even treating it merely as a compliance exercise to satisfy regulatory requirements. If you are a business leader, we urge you to take this matter seriously starting now—it is far more valuable than countless cursory walkthroughs of the plant.

Change management. Domestic enterprises are now placing greater emphasis on change management; pharmaceutical companies were among the first to adopt and implement it, driven by GMP requirements. However, in some organizations, change management has been reduced to merely following procedural steps, losing sight of its fundamental purpose. The original intent is not simply to obtain a signature or approval from a particular individual, but rather to use this process to prevent new hazards from arising due to changes—eliminating risks is at the heart of effective change management. In certain cases, while change documentation and formal sign-offs are in place, hazard analysis is often inadequately addressed. This situation is partly attributable to plant managers and technical personnel lacking proficiency in hazard‑analysis tools and methodologies; targeted training is needed to help them master these techniques. Given that enterprises may face numerous changes, they can prioritize and categorize them according to risk levels, with high‑risk changes requiring thorough hazard analyses.

Mechanical integrity. The mechanical integrity of equipment—including the reliability of critical instruments—forms the material foundation for safety in chemical plants. The recent accident at Shenghua Chemical was caused by a failure in mechanical integrity. In many of our plants, all equipment and instruments are treated equally, and we remain at a rudimentary stage of maintenance, addressing failures only after they occur. To improve this area, we can adopt risk‑based preventive or predictive maintenance approaches, implementing tiered management based on risk levels. (Many plants do have equipment classification systems, but these typically prioritize asset value rather than the potential risks posed by equipment failures.) By developing and consistently executing preventive maintenance plans for critical assets, we can achieve the goal of ensuring their reliable operation with limited resources. Mechanical integrity management is both crucial and highly challenging, requiring coordinated efforts across multiple functions, including equipment, maintenance, engineering, and production.

Operating Procedures and Training. Operators and maintenance personnel are critical to ensuring safety. Many companies today face labor shortages and high employee turnover. There are two approaches to addressing this issue: first, refine the design to enhance fault tolerance, so that even if an operator makes a mistake, additional engineering safeguards can prevent accidents or mitigate their consequences—this should be accomplished during the process hazard analysis phase; second, invest in robust employee training. While there is much complaint about difficulties in recruiting and low employee competence, such complaints do not solve the problem—this is the current reality. The most important step for companies is to organize effective training programs. Some enterprises excel in this area, but more commonly, new hires are brought on board, the safety department briefly reviews safety regulations, and operating procedures are handed out for self‑study, after which employees are expected to learn on the job. This type of training falls far short of what is needed. At a minimum, new employees should receive foundational safety training—covering permit‑to‑work, confined spaces, and emergency response—and comprehensive, detailed training on job‑specific operating procedures. The latter is especially crucial, as it must equip workers with the knowledge to identify potential issues during operations and the corresponding corrective measures.

Risk management for high‑risk operations. Hot work has been the cause of numerous accidents; in some enterprises, when issuing hot‑work permits, flammable‑gas detection is not even performed—some companies do not even have a flammable‑gas detector—and the personnel responsible for gas testing often lack adequate training. Moreover, most organizations have yet to establish a permit‑to‑work system for confined‑space entry, and insufficient attention is paid to energy isolation. The cornerstone of controlling risks in high‑risk operations is ensuring that critical hazard‑control measures—such as flammable‑gas detection, purging, and isolation—are properly implemented.

Emergency response planning. Emergency response is not only about addressing immediate issues when an incident occurs—such as containing a chemical spill—but also about preventing the situation from escalating into something far more serious. To maintain effective preparedness for potential accident scenarios, it is essential to have well‑developed plans, adequate emergency supplies, and qualified personnel, with regular, thorough training to enhance response capabilities. While some of our plants do have emergency plans, they tend to be overly generic—for example, relying on a single hazardous‑chemical spill‑response procedure to handle diverse incidents, such as a toluene leak in the tank farm and a liquid chlorine cylinder rupture on the shop floor. The required response measures for these two scenarios are vastly different, calling into question the practical utility of such a one‑size‑fits‑all approach. A better practice is to identify credible accident scenarios within the process system that could lead to severe consequences, then develop tailored emergency response procedures as part of the overall emergency plan. Based on these procedures, we can clearly define the plant’s major emergency scenarios, determine the necessary tools, equipment, and personal protective gear, and specify their locations, thereby enabling targeted emergency drills and preparing for potential incidents.

Contractors. Here, contractors primarily refer to those who provide manufacturing, maintenance, and construction services to the plant; they often undertake tasks that are particularly challenging or pose significant hazards. When fires or explosions occur during contractor operations, the ultimate victims are the enterprise itself. Managing contractors is complex, spanning stages such as contractor selection, site access, on-site work, and departure, and it presents a significant challenge for many organizations. Plants should restrict contractors’ access to non‑essential process areas and can also leverage permit‑to‑work systems to effectively control contractors’ activities and associated risks.

There are many aspects involved in day-to-day operations. The elements listed above represent the management priorities for eliminating and controlling hazards, and it is particularly important to explore how to effectively implement them.

Enterprises can also promptly identify management deficiencies and achieve continuous improvement by conducting root cause analyses of adverse incidents—including accidents and near-misses—and by carrying out internal safety audits.

Root cause analysis of adverse incidents. Before many catastrophic accidents occur, a series of near-misses often precede them. Factories can encourage employees to report these near-misses and then conduct root‑cause analyses to identify managerial shortcomings and address them promptly. This approach is highly effective in preventing other accidents that might arise from those same deficiencies. Unfortunately, most organizations do not place sufficient emphasis on investigating near-misses, nor do they conduct thorough accident investigations; in some cases, the focus is narrowly on imposing personnel‑level penalties under the “four no‑放过” principle, making it difficult to uncover the true causes of accidents. Furthermore, management personnel generally lack formal training in root‑cause analysis, and their investigations typically rely on rudimentary methods such as discussions and brainstorming. We recommend improving the reporting and investigation systems for both accidents and potential hazards, and adopting advanced tools like fault‑tree analysis to drill down into the underlying causes. Engaging all employees in reporting accidents, near‑misses, and hazards also serves as a powerful lever for fostering a robust safety culture within the organization.

Internal safety management audits. At present, relatively few small and medium-sized chemical enterprises have established formal internal safety‑management audit mechanisms. Such audits are systematic, in‑depth review processes—distinct from ad hoc inspections—and play a critical role in ensuring that safety management systems are implemented and enforced as required. It is recommended to establish an internal audit framework for safety management and to put in place a mechanism for tracking the implementation of corrective actions identified during these audits. In particular, internal audits should be risk‑based, rather than merely focusing on document reviews.

(5) Demolition

This phase involves significant operational safety hazards, such as the need to perform thorough cleaning and purging prior to dismantling, as well as measures to prevent the collapse of equipment and structures that have deteriorated over time; further details will not be elaborated here.

All of the aforementioned aspects are critical to ensuring the safe operation of chemical enterprises. Some of our companies have relatively weak foundations, making it difficult to achieve full compliance overnight; however, we must act promptly and strive to implement these measures as soon as possible. Otherwise, looking back in a few years, we may find that our management standards have remained stagnant.

How to effectively manage safety is a complex issue that cannot be fully addressed in a brief article. While each organization faces unique circumstances, there is always room for improvement—requiring down-to-earth, practical actions. Without concrete steps, the workplace will remain unchanged, and hazards will continue to lurk, waiting for us to make mistakes.

I had no intention of writing such a lengthy and rather dry piece—partly because I’ve been extremely busy, but mainly because it’s hard to strike the right balance in my tone, and I worried it might lack objectivity or give rise to misunderstandings. A year ago, I spent three days working at Zhijiang Chemical, which is located right next door to the Tianjiayi Chemical plant where this accident occurred. In that tragedy, several people from Zhijiang lost their lives; they had once shared conversations with me, inspected the scene, and attended meetings to discuss matters. Their warmth and kindness remain etched in my memory. They were supposed to retire like most people, enjoying a happy life surrounded by their children and grandchildren—but the accident abruptly shattered their futures. Perhaps it was their passing, unconsciously stirring something within me, that enabled me to write this three‑part series.

Throughout this writing, I have consistently approached the subject with sorrow, reverence, and a commitment to objectivity and prudence. The person is gone; we must learn from this tragedy and strive to do better, lest such misfortunes recur. At the same time, there is no need to be unduly alarmed by accidents—let us maintain confidence. With sound scientific management and well‑implemented measures, the risks associated with chemical enterprises can be fully contained; indeed, working in a chemical plant can even be safer than staying at home. (End)

This article is compiled from Wenran Academy; author: Su Zhenyu