Looking at 21 hazardous chemicals accidents in June: what often fails in reality is not the systems, but the hierarchy of values

Thread Content

Looking at the 21 hazardous materials accidents that occurred in June: what is truly failing is often not the systems in place, but rather the hierarchy of values. Note: This article conducts a management review analysis based on the accident data provided by users; it focuses not only on the \"immediate causes\" but also explores the underlying fundamental management issues such as technical defects, violations of operating procedures, risk identification, change management, contractor oversight, and emergency response for each accident. I. General assessment: In June, accidents involving hazardous chemicals exhibited distinct characteristics of the high-temperature season: increased volatilization of flammable gases, heightened risks of thermal stress and corrosion-induced equipment failure, a rise in inspection, maintenance, startup, and shutdown operations, as well as greater exposure to confined space and hot work activities. On the surface, accidents appear to be caused by leaks, explosions, fires, or poisoning and asphyxiation, but the underlying causes mainly fall into the following six categories: 1. Loss of control during special operations: Permits for operations such as hot work, work in confined spaces, inspection and maintenance, tank cleaning, and dismantling equipment under pressure are often treated as a formality. 2. Insufficient awareness of process risks: There is inadequate understanding of the hazardous properties of ethylene oxide, ethanol, sodium chlorate, organic substances, intermediates, peroxides, hydrogen sulfide, nitrogen, etc. 3. Insufficient equipment integrity: Issues such as pipe breaks, valve stem detachment, corrosion-induced thinning, weld cracks, missing safety valves, and pressure relief failures were not detected or addressed in advance. 4. Weak handling of startup/shutdown and abnormal operating conditions: Abnormal backflow, overpressure, continuous heating, and rising pressure were not identified and stopped in a timely manner. 5. Inadequate on-site isolation and energy control: Removing components while pressure is still present, failing to replace gases, failing to conduct inspections, failing to isolate the system, and proceeding with maintenance without shutting down the equipment – these are common factors that trigger multiple accidents. 6. Misconduct in emergency rescue: In multiple cases of poisoning and asphyxiation, it wasn’t just “one person in danger”; rather, blind rescue efforts led to multiple deaths. ★Insight Analysis: Accidents don’t occur randomly; they often result from the simultaneous occurrence of “hazardous materials + abnormal operating conditions + special operations + lapses in management”. · The root causes usually lie not in the final action, but in prior risk identification, permit approval, equipment integrity, and the crew’s execution culture. · For production management, what truly matters is not slogan-based safety education; rather, it’s establishing a manageable closed loop in which high-risk actions can be verified, made accountable for, and used as grounds for halting operations. II. Root cause analysis of each incident
No. Surface events of the incident Root cause analysis Management implications
1. The “6·18” explosion at the ethylene glycol plant of Sinopec Shanghai Petrochemical Company in 2022: A pipe related to the ethylene oxide purification tower broke and leaked; upon encountering a source of ignition, the ethylene oxide inside the tower decomposed and caused an explosion. There were deficiencies in equipment integrity management, as well as inadequate identification of risks related to clamp repairs, pipe stress, corrosion, or fatigue ； Underestimation of the consequences of decomposition and explosion following an ethylene oxide leak ； Inadequate capabilities in early leak isolation, interlock mechanisms, and emergency response; failure to maintain registers and conduct life-cycle management for temporary fittings, pipelines operating under defective conditions, and pipelines transporting critical hazardous substances ； For high-risk media such as ethylene oxide, higher standards must be established regarding leak monitoring, emergency shutoff, and emergency response. In the “6·8” leak and fire incident at Sinopec Maoming Branch in 2022, during the pressurized removal of the pneumatic motor fasteners from the ball valve at the outlet of an ethylene transfer pump, the valve stem came loose, resulting in a massive leak and subsequent explosion of ethylene. Energy isolation and pressure relief verification were not completed prior to the operation ； On-site personnel have insufficient understanding of the valve structure, anti-detachment functionality, and risks associated with pressure differences ； Ineffectiveness in approval, supervision, and technical briefings for maintenance operations; establishment of a “zero tolerance” policy regarding disassembly under pressure ； When disassembling pressure-bearing components such as valves, pumps, and compressors, it is necessary to carry out isolation, pressure relief, purging, inspection, and labeling for confirmation. In the “6·5” major explosion and fire incident at Shandong Linyi Jinyu Petrochemical Company, a quick-connect coupling used for unloading liquefied petroleum gas came loose in 2017, leading to a massive leak followed by an explosion. This incident indicates that the verification system for unloading connections had failed; moreover, there were insufficient measures to prevent the quick-connect coupling from coming loose, being misconnected, or causing leaks ； Inadequate inspections before and after tanker unloading, as well as insufficient on-site supervision ； Insufficient mechanisms for gas detection after LPG leaks, emergency shutdown, and personnel evacuation; double verification is required for connection, pressure testing, and tension confirmation during unloading operations ； Loading arms, quick connectors, breakaway valves, and emergency shut-off valves must be inspected and tested ； The unloading area must have its sources of fire and vehicles managed in accordance with the requirements for explosive gas environments. In the major fire that occurred on June 15 at Sinopec Shijiazhuang Refining & Chemical Co., slag from welding operations fell into a polypropylene mist remover, causing a fire; the workers suffered poisoning and suffocation. The identification of risks associated with welding operations focused only on the location where the work was being carried out, without taking into account the flammable components located below ； Inadequate vertical space separation, fire sealing, and fire containment measures ； There is insufficient integrated management of risks associated with working at heights, confined spaces, and toxic fumes. Hot work permits must specify the “potential locations where sparks may land” and the “flammable materials located below.”” ； Before hot work, devices containing plastic, rubber, anti-corrosion coatings, defoggers, etc., must be removed, isolated, or covered ； For hot work at heights, supervision must be provided below the work area. In the “June 2” major explosion and fire incident at the benzene/toluene/xylene tank farm of PetroChina Dalian Petrochemical Branch in 2013, a contractor conducted unauthorized gas cutting on top of a tank, igniting gases such as toluene and resulting in the tank’s explosion. This incident highlights a severe lack of control over contractor management and hot work permits ； Inadequate detection of combustible gases, cleaning and purging, and isolation verification in the storage tank area ； There is insufficient awareness of the catastrophic consequences of hot work in tank farms; hot work in such areas must be approved under the highest risk category ； The contractor shall not perform hot work independently without the employer’s supervision ； Before any hot work is carried out, it is necessary to complete six procedures: cleaning, displacement, testing, blind flange isolation, and continuous monitoring. In the “6·29” relatively serious explosion incident at PetroChina Liaoyang Petrochemical Branch, during the cleaning of crude oil tanks in 2010, the mixture of oil vapors and air created an explosive atmosphere; an explosion occurred upon contact with non-explosion-proof lighting or sparks from iron tools. This incident highlights inadequate risk control measures during tank-cleaning operations ； Improper management of explosion-proof electrical equipment and fire- and static electricity prevention tools ； Absence of continuous monitoring in confined spaces and for flammable gases ； The tank-cleaning plan lacks a comprehensive review regarding processes, safety, and emergency response. Tank cleaning must be managed as a combination of special operations: confined space work, fire hazards, explosion-proof electrical equipment, anti-static measures, ventilation, gas detection, and rescue plans all need to be incorporated into the plan. In the “June 11” asphyxiation incident at Gansu Honghui Energy & Chemical Company in 2024, no purging, displacement, or gas testing was carried out during maintenance on the moving particle bed. After opening the manhole, nitrogen-induced asphyxiation occurred; ill-advised rescue efforts further exacerbated the situation. This incident highlights the failure of the confined space management system, as well as the non-compliance with the principle of “ventilate first, then conduct gas testing, and only then proceed with operations.”” ； Hazards of nitrogen inerting systems are underestimated ； Insufficient rescue training and on-site emergency equipment; all nitrogen protection, inerting, and displacement equipment are treated as such for oxygen-deficient environments by default ； Entry without inspection, permission, or protection is prohibited ； Rescue operations must involve the use of air respirators and external rescue methods. In the major explosion that occurred on June 27 at the Wuhai Huazi Coal and Coke Company in Inner Mongolia, flammable gases in the desulfurization solution circulation tank combined with air to form an explosive mixture; cutting or welding was carried out without obtaining the necessary permission for such work ； Insufficient awareness of the risks of ammonia and volatile flammable gas accumulation in the desulfurization system ； For work carried out at the top of the equipment, gas testing inside the tank is not required; moreover, systems involving desulfurization, wastewater, tanks, and those containing volatile substances must all be treated as being in a combustible gas environment ； Welding and cutting tools must not be used without gas analysis and a fire permit. In the major explosion that occurred on June 16 at Gansu Lanzhou Binnong Technology Company, maintenance work was carried out on the discharge valve while the dryer was still in operation and heating; as a result, the mother liquor waste was exposed to heat for an extended period, which led to an explosion. The maintenance work was carried out without shutting down the equipment, isolating it, or reducing its temperature ； Insufficient understanding of the thermal decomposition, exothermic reactions, and risk of sympathetic detonation when sodium chlorate coexists with organic materials ； Insufficient temporary storage of solid waste and process safety assessments. Thermal risk assessments must be conducted for drying, concentration, and distillation processes involving oxidizers, organic substances, and heat-sensitive materials ； Before maintenance, the machine must be stopped, powered down, cooled down, and cleared of materials ； Solid waste must not be arbitrarily piled up in high-risk areas. In the “6·26” major fire and explosion incident at Xumei Biotechnology Co., Ltd. in Kaifeng, Henan Province, a valve failed to open during an ethanol extraction operation in 2019, resulting in overpressure. Consequently, the discharge cover blew off, and the ethanol ignited due to static electricity. The operating procedures were not properly verified; there was a lack of interlocks for critical valve positions, as well as pre-startup inspections ； Inadequate control of overpressure, static electricity, and flammable vapor risks during the heated extraction process with ethanol ； The storage of combustible materials on-site exacerbates the consequences; for processes involving ethanol, it is necessary to implement valve position verification, pressure protection, static grounding, ventilation, and explosion-proof electrical equipment ； The temporary storage volume of flammable solvents must be controlled; the production site should not be turned into a temporary warehouse. In the serious explosion that occurred on June 18 at Xingfa Furfural Company in Chai Gang, Nong’an, Jilin, the pressure relief valve in the hydrolysis reactor failed to function properly due to inherent safety defects in the equipment, and safety valves were not installed in the cylinder sections as designed ； The personnel failed to open the valves as required, causing the pressure to continue rising ； Illegal production leads to the failure of the safety management system; pressure vessels and safety accessories must operate in compliance with regulations ； Safety valves, pressure gauges, and interlocks must be present ； Suspension of production, resumption of production, and temporary production must all be subject to approval and safety assessments. In the major explosion incident that occurred at Linjiang Chemical Company in Shaoxing, Zhejiang on June 9, unstable intermediates decomposed due to heating in the reaction vessels during the pilot testing of new products in 2017, resulting in a sudden rise in temperature and pressure and an explosion; systematic safety evaluations were lacking throughout the process from research and development to pilot-scale production ； Conducting tests using industrial equipment without understanding the thermal stability and reaction risks of intermediates ； Insufficient change management and pilot production approval; for new products and new processes, as well as scale-up from pilot scale, reaction risk assessment, thermal stability testing, HAZOP, or equivalent reviews must be conducted ； It is not permissible to “conduct trial production using old equipment” as a makeshift solution. In the “June 20” incident at Shixing Pharmaceutical & Chemical Co., Ltd. in Huludao, Liaoning Province, a relatively serious asphyxiation accident occurred: after nitrogen purging of a crystallization kettle in 2018, it was cleaned without proper oxygen testing; this led to personnel suffocating. Additionally, rescue efforts were carried out recklessly, and risk management regarding confined spaces and nitrogen-induced asphyxiation proved ineffective ； No status confirmation is available before restarting a disabled device ； Before reactivating, sealing, or nitrogen-purging equipment for which operation was halted due to inadequate emergency response plans or personal protective measures, it is necessary to create an “Equipment Status Card”” ； Before entering, oxygen detection, toxicity testing, ventilation, monitoring, and rescue equipment are all essential. In the “June 5” suffocation incident at Shandong Weifang Huahao Agrochemical Company in 2016, workers entered a tank to investigate why materials couldn’t be extracted; they suffered from oxygen deficiency and suffocation. Subsequently, rescue efforts were carried out recklessly. The handling of this abnormal production situation was not elevated to the level of special operation management ； Employees treat “going in to take a look” as a routine procedure ； For any activity involving entry into tanks, vats, tanks, wells, or pits, gas detection, protective equipment, and monitoring are all required; a confined space permit must be obtained first ； When dealing with abnormalities such as blockages or inability to extract materials, it is prohibited for personnel to enter directly to investigate the cause. In the major explosion and fire incident that occurred on June 28 at Yidong Jiuding Chemical Company in Ordos, Inner Mongolia, aging cracks in the welds of the heat exchangers led to the sudden release of desulfurization gas, resulting in a hydrogen explosion and fire. Defects in the quality of equipment manufacturing as well as in welding work have persisted for a long time ； No adequate testing, evaluation, or replacement decision was made after multiple repairs and welds ； Insufficient equipment integrity management for hydrogen-containing flammable gases requires the establishment of mandatory assessment and retirement mechanisms for devices that undergo repeated repairs and welding, experience repeated leaks, or have defects in critical welds ； Hydrogen-containing systems require enhanced non-destructive testing, material management, and analysis of leakage consequences. In the “6·12” hydrogen sulfide poisoning incident at Qitian Fertilizer Company in Anning, Kunming, Yunnan Province, an excessive amount of sodium sulfide in a phosphoric acid tank led to the generation of large quantities of hydrogen sulfide in 2008. There was insufficient risk assessment regarding trial production; both on-site personnel and rescuers were poisoned. Moreover, there was a lack of awareness regarding the danger posed by the reaction between sodium sulfide and acids, which produces hydrogen sulfide ； Abnormal material feeding and valve failures were not addressed promptly and safely ； Processes involving open tanks that lack sealing, ventilation, and alarm systems for sulfides and acidic materials must be managed in accordance with the risks associated with severe hydrogen sulfide poisoning ； Set standards for fixed detection and alarm systems, local exhaust ventilation, sealed material feeding, and emergency evacuation. In the \"6·25\" fire accident at Bohua Chemical Company in the Binhai New Area of Tianjin in 2024, alkali solution spilled into the oxidation system during the preparation for starting up the hydrogen peroxide production plant, resulting in rapid decomposition of hydrogen peroxide and an explosion and fire in the extraction tower; insufficient verification of the system’s condition and inadequate isolation of valves and pipelines prior to startup ； Risk of abnormal backflow not identified ； Insufficient handling of hydrogen peroxide decomposition risks and abnormal operating conditions ； Insufficient monitoring of key parameters and interlock protection; before starting up, it is necessary to verify the system flow, check the positions of blind flanges and valves, and identify any backflow paths ； For facilities handling unstable materials such as hydrogen peroxide, it is essential to pay close attention to pollutants, pH levels, temperature, and the risk of decomposition. The severe hydrogen sulfide poisoning incident that occurred on June 12 at Zhejiang Taizhou Fengrun Biochemical Company was caused by high hydrogen sulfide concentrations in underground boreholes, which led to poisoning among workers; subsequently, many people attempted to carry out rescue operations without realizing that those underground areas were confined spaces ； Toxic gases were not detected prior to construction ； There is a lack of ventilation, protective gear, monitoring, and rescue equipment on site. Pits, wells, pile holes, and trenches must be included in the list of confined spaces ； Construction-related activities must also comply with the same gas detection and rescue requirements as those applied in chemical enterprises. The extremely serious explosion that occurred on June 30 at the Sigaichi Industrial Company in Telangana, India; in 2025, there was an abnormal increase in pressure in the microcrystalline cellulose drying unit, leading to a failure in pressure relief and ultimately to an explosion. There were defects in the design of the drying equipment, as well as in the pressure monitoring and safety release systems ； Abnormal pressure trends were not warned in a timely manner ； After the failure of the key protective layer, there is a lack of independent protection for drying, powder, and spraying equipment; thus, risks related to overpressure, dust, blockage, and thermal decomposition must be addressed ； Safety pressure relief devices must be regularly calibrated, and they should be equipped with independent alarm systems as well as interlock shutdown functions. The “6·21” explosion at the refinery operated by Energy Solutions Company in Philadelphia, USA; in 2019, corrosion and thinning of the pipe elbows in the hydrofluoric acid alkylation unit led to ruptures, resulting in propane leaks, fires, explosions – due to inadequate monitoring of corrosion patterns and wall thicknesses ； Insufficient pipeline integrity management for high-risk installations ； The high-consequence risks associated with hydrofluoric acid alkylation units necessitate stricter management of materials, inspections, and alternative solutions. It is necessary to maintain a corrosion monitoring register and designate thickness measurement points based on factors such as the medium, temperature, flow rate, and erosion caused by elbows ； High-consequence areas cannot be inspected on a fixed schedule alone; inspections should be adjusted according to evolving risks. The \"6·26\" explosion in the methanol distillation tower at the Chiba plant of Nippon Shishi Co., Ltd. on June 21st occurred because trace amounts of methyl peroxide were generated from methanol and hydrogen peroxide. During the shutdown process, local concentration and decomposition led to the explosion. There was insufficient awareness of the risks associated with the accumulation, concentration, and thermal decomposition of these trace by-products ； Inadequate risk analysis of the parking process ； Inadequate management of local concentration in distillation towers and residual liquids; systems containing peroxides, oxidizers, and organic solvents require an assessment of side reaction and concentration risks ； Parking, distillation, and residual liquid treatment must be incorporated into the management of process safety boundaries. III. Classification of common root causes of accidents 1. Special operations are the most common triggers for accidents. Accidents occurring during operations such as hot work, work in confined spaces, tank cleaning, maintenance and repair, disassembly, startup preparation, and trial production account for a very high proportion. The problem is not just that employees are “careless,” but that the management system does not turn these actions into mandatory checkpoints. Key issues: · No permits for hot work or work in confined spaces were obtained. · Removal under pressure, maintenance without shutting down the system, lack of isolation, no replacement, no testing. · The guardian failed to fulfill their duties properly and was only present in a nominal capacity. · The contractor lost control after entering a high-risk area. 2. Insufficient awareness of hazardous materials: Numerous accidents have revealed that companies do not have a thorough enough understanding of the hazards associated with these materials. For example: · Ethylene oxide: It is flammable in the event of a leak, and may decompose and explode when exposed to heat or fire. · Ethylene, liquefied petroleum gas, toluene, ethanol: highly flammable and explosive; can be ignited by static electricity, sparks, or hot surfaces. · Nitrogen: It is not toxic in itself, but it can cause asphyxiation due to lack of oxygen. · Hydrogen sulfide: Highly toxic; tends to accumulate in low-lying and confined spaces. · Sodium chlorate and organic materials: A strong oxidizing agent in contact with flammable organic materials can undergo a violent reaction when heated. · Peroxides: Generated in trace amounts and concentrated locally, they can also become a source of explosion. 3. No closed loop has been established for equipment integrity. Equipment issues include pipe ruptures, corrosion-induced thinning, valve structural failures, weld cracks, missing safety valves, and failed pressure relief mechanisms, etc. What they have in common is that there are often signs prior to an accident, but these signs are not translated into decisions to shut down operations, carry out maintenance, or replace components. Key weakness: · Long-term use of temporary fixtures. · It continues to operate even after multiple repair welds. · Critical safety accessories are missing or malfunctioning. · The corrosion thickness measurement points do not cover high-risk elbows, welds, and dead corners. · The equipment anomaly did not trigger a review by management. 4. The risks associated with starting up and shutting down, pilot production, and handling abnormalities are relatively high. These processes are more dangerous than regular production, as the process conditions are complex, there are many valves involved, and personnel tend to rely on experience when operating. Accidents such as those involving hydrogen peroxide in Tianjin, ethanol extraction in Henan, pilot tests in Zhejiang, and trial production in Yunnan all fall into this category. Management issues: · Lack of pre-driving safety checks. · The valve positions, blind flanges, and reverse flow paths were not confirmed item by item. · There are no standard handling cards for abnormal conditions. · The new processes and products have not undergone process safety assessments. 5. Blind rescue efforts repeatedly lead to an increase in casualties. In accidents involving nitrogen asphyxiation, oxygen deficiency, and hydrogen sulfide poisoning, many deaths occur when rescuers enter the scene without wearing self-contained breathing apparatus. The root cause is that companies fail to incorporate the understanding that \"rescue is also a high-risk task\" into their training, equipment, and drills. Management red line: · It is prohibited to enter spaces suspected of having low oxygen levels or being toxic to rescue people without wearing an air respirator. · Upon discovering someone fallen, the first reaction is not to rush in, but to call for help, ventilate the area, conduct tests, put on protective equipment, and summon external rescue. · The site must have rescue tripods, safety ropes, air respirators, gas detectors, and other equipment, which must also be used. IV. Guidelines for Production Management 1. Establish a special risk list for the hot months of June. It is recommended to conduct a specialized inspection of \"high-risk operations and high-risk equipment in June\" by the end of May each year, in order to maintain a systematic record of such risks. Key targets: · Flammable and explosive substances: ethylene, LPG, toluene, ethanol, hydrogen, ethylene oxide, etc. · Highly toxic asphyxiating agents: hydrogen sulfide, nitrogen, oxygen-deficient environments. · Thermosensitive and reactive materials: hydrogen peroxide, peroxides, chlorates, organic solid waste, unstable intermediates. · High-risk operations: hot work, confined space work, tank cleaning, inspection and maintenance, startup/shutdown, trial production, and contractor activities. · High-consequence equipment: storage tanks, towers, reactors, dryers, heat exchangers, pressure pipelines, loading and unloading facilities. 2. Change the special operation permit from an “approval form” to a “field verification form”. In managing special operations, one must not merely check whether the permit is fully filled out; rather, it’s necessary to verify whether the actual site conditions meet the requirements. It is recommended to set six mandatory verifications: 1. Whether the system has truly come to a stop, its power has been cut off, pressure has been released, and it has been isolated. 2. Has purging, replacement, and ventilation been completed? 3. Is oxygen content, flammable gases, and toxic gases detected? 4. Whether it can identify the landing site on Mars, the reverse flow path, and residual liquids and gases. 5. Are there qualified guardians and continuous monitoring? 6. Are emergency rescue equipment and evacuation routes available? Management action: · Sign-off confirmation by the site supervisor, safety officer, process engineer, and task leader. · For high-risk operations, photographs must be taken as documentation; key isolation points, gas detection values, and blind flange locations must be traceable. · After the operating conditions change, the original work order becomes invalid automatically and must be re-confirmed. 3. Zero tolerance is applied to the act of \"going inside the equipment to take a look.\" Many suffocation accidents start with the idea of \"going in to take a look,\" \"going down to deal with it,\" or \"opening the manhole to check it.\" Such actions must be recognized by the system as work in a confined space, rather than ordinary operations. It is recommended to establish a team password: · Whenever entering a tank, reactor, tank, well, pit, tower, or any low-lying area in a pipeline corridor, one must stop first to obtain the necessary permit. · Any equipment that has been nitrogen-sealed, inerted, taken out of service, or used for storing materials must first be treated as a hypoxic or toxic atmosphere. · Whenever someone falls to the ground, no one shall enter to rescue them without protection. 4. Establish a hierarchical management system for the integrity of critical equipment. For media with high risks, as well as equipment under high pressure, high temperature, subject to corrosion, or requiring frequent repairs, more stringent equipment records should be maintained. It is recommended to implement a three-level management system: · Level A: Equipment handling high-consequence media such as ethylene oxide, hydrogen, LPG, ethylene, hydrogen sulfide, and hydrogen peroxide, which requires close monitoring. · Grade B: For equipment showing corrosion, erosion, thermal fatigue, repeated repair welding, or use of temporary fixtures, special inspections shall be conducted. · Class C: General equipment, managed according to regular inspection and maintenance cycles. Key requirements: · Temporary fixtures must not be used on a permanent basis; there must be a set deadline, risk assessment, and alternative plans. · For equipment that requires repeated repair and welding, a technical review must be organized, and replacement must be enforced if necessary. · Safety valves, rupture disks, interlocks, alarms, and emergency shut-off valves must be subject to regular calibration. · Add thickness measurement points at corroded elbows, welds, dead corners, low points, and changes in diameter. 5. Start-up, shutdown, and abnormal operating conditions should be managed using checklists; start-up and shutdown are not matters to be handled based on experience, but rather involve high-risk operations. It is recommended to establish a start-up and shutdown confirmation form as well as an abnormality handling card for each set of equipment. Before driving, make sure to check: · Whether the valve position is correct. · Were the blind plates installed and removed as per the list? · Whether the temporary pipeline has been removed. · Check whether there are backflow paths in the sampling line, drain line, and bypass line. · Are the safety interlocks, alarms, and pressure relief systems in service? · Is there a risk of contamination, mixing, excess, or local concentration of key materials? Principles for handling abnormalities: · In the event of abnormally high pressure, prioritize cooling down, stopping feed, and releasing pressure to the safety system. · If the materials cannot be extracted or the valves cannot be opened, do not disassemble the equipment or enter it without permission. · In the event of leakage, unusual odors, or alarms, evacuate personnel first, isolate the area, conduct inspections, and then take action. 6. New products, pilot tests, and process changes must go through change management. Incident reports related to processes such as pilot testing in Zhejiang, peroxide distillation, and solid waste drying show that new processes and pilot-scale up are areas where accidents occur frequently. Suggested requirements: · All new raw materials, new formulations, new equipment, new operating conditions, new suppliers, new catalysts, and new cleaning methods should be included in change management. · Before pilot testing, assessments of the thermal stability of the materials, reaction exothermicity, gas release, pressure buildup, and the hazards associated with by-products must be completed. · Temporary testing using discontinued old equipment or mismatched equipment is prohibited. · During the trial production phase, reduce the amount of materials on site and implement stricter controls over temperature, pressure, mixing, cooling, and pressure relief. 7. Strengthening the management of contractors and external construction management contractors: A common issue is that \"people come in, but the risks are not brought into the system.\" Recommended implementation: · Conduct unit-level safety briefings for contractors prior to their entry, rather than just plant-level training. · Operations such as hot work, lifting, working at heights, and work in confined spaces must be confirmed on-site by the owner. · The owner’s supervisor must understand process risks and have the authority to halt operations. · If a contractor violates the red lines, they must be dismissed immediately and added to the blacklist. 8. Establish a \"protect first in rescue\" mechanism for hydrogen sulfide, nitrogen, and oxygen deficiency incidents; it is recommended to conduct confined space rescue drills on a quarterly basis. It must be made clear on site: · Who called the police. · Who ensures air circulation and ventilation. · Who detects gases? · Who enters wearing an air respirator? · Who is responsible for external traction rescue? · Who is responsible for on-site security to prevent unauthorized personnel from entering? Mandatory configuration: · Portable four-in-one gas detector. · Hydrogen sulfide detector. · Positive-pressure air respirator. · Safety rope, tripod, rescue stretcher. · Explosion-proof lighting and explosion-proof ventilation equipment. V. It is recommended that you implement the following checklists immediately at the production site.
A. Daily inspections by work teams
Inspection items and judgment criteria:
- Whether there are hot work, confined space operations, maintenance activities, loading/unloading operations, or startup/shutdown procedures; if so, these should be included in that day’s list of high-risk operations.
- Whether there are any abnormal odors, leaks, alarms, or pressure fluctuations; if present, immediate escalation and corrective actions must be taken—operations must not proceed under such conditions.
- Whether gas detectors, explosion-proof tools, and protective gear are functional; if not, related operations must be halted.
- Whether contractors are operating under proper supervision; if not, operations must be stopped immediately.
- Whether there are temporary pipelines, temporary clamps, or temporary electrical connections; if so, verify their approval status, expiration dates, and associated risk mitigation measures.

B. Pre-start verification for special operations
Categories requiring confirmation:
- Hot work: cleaning and purging, gas detection, identification of potential spark locations, presence of combustible materials below the work area, fire prevention measures, continuous monitoring.
- Confined space operations: ventilation, oxygen level measurement, toxic gas detection, explosion risk assessment, supervision, availability of rescue equipment, and unobstructed entry/exit routes.
- Maintenance activities: shutdown procedures, power disconnection, tagging, pressure relief, emptying of systems, use of blind flanges, and trial disassembly checks.
- Loading/unloading operations: securement of connectors, static grounding, pressure and leak testing, emergency shutoff mechanisms, and on-site supervision.
- Startup/shutdown procedures: valve positions, use of blind flanges, interlocks, alarms, pressure relief mechanisms, reverse flow pathways, and protocols for handling abnormal situations.

C. Weekly inspections by management
Key issues to be examined:
- Whether high-risk operation permits have been signed by unauthorized persons, backdated, or reissued after changes in operating conditions.
- Equipment integrity: whether there are recurring leaks, repeated welding repairs, or prolonged use of temporary clamps.
- Process abnormalities: whether there are frequent alarms, bypassed interlocks, or instances where operations continue despite parameter limits being exceeded.
- On-site storage: whether quantities of flammable solvents, solid waste, or oxidizing agents exceed permitted levels or are stored inappropriately.
- Training and drills: whether employees understand that they must not attempt rescues blindly in cases of poisoning or asphyxiation.

VI. Key recommendations for your production management
If you can only focus on five critical aspects, it is advised to prioritize them in the following order:
1. Ensure authenticity of special operations: permits, isolation measures, detection procedures, supervision, and rescue readiness must all be verified on-site. 2. Address confined space hazards and avoid reckless rescue attempts: Make \"ventilate first, then test, and only afterward proceed with work\" as well as \"no rescue without protection\" absolute rules for the team. 3. Ensure equipment integrity: A closed-loop checklist should be established for temporary fixtures, repeated repairs and welding, corrosion-induced thinning, as well as failed safety valves and interlocks. 4. Address parking and abnormal condition handling: All valve positions, blind flanges, backflow paths, and interlock statuses must be confirmed in a list format. 5. Enhancing awareness of hazardous materials: Specialized training is provided on ethylene oxide, ethylene, LPG, ethanol, hydrogen, hydrogen sulfide, nitrogen, hydrogen peroxide, peroxides, and organic oxidant mixtures. VII. Underlying root causes related to safety culture and values behind the problems: If we delve deeper, these accidents are not merely failures at the institutional, equipment, training, or operational levels. What lies behind this is a deviation in the corporate security culture and values: what an organization prioritizes in its actual decision-making, what it considers acceptable to compromise on, which voices are silenced, and which behaviors are rewarded. Accidents often aren’t the result of one person suddenly making a mistake; rather, they stem from a set of values that have long been tacitly accepted and that come to fruition in high-risk situations. 1. “Production priority” overrides “risk priority”; many accidents occur during inspection and maintenance, startup and shutdown, unloading, pilot production, and abnormal condition handling. These steps should require greater caution than in normal production, but on site, there is often a tendency to say \"finish the work first,\" \"don’t delay progress,\" and \"get production back up immediately.\" When schedule, output, cost, and performance pressures are constantly given priority, actions such as isolation, replacement, testing, approval, and review are likely to be seen as burdens that slow down efficiency. The issue of underlying values is not that companies verbally downplay the importance of safety; rather, when safety comes into conflict with production volume, deadlines, and costs, the organization actually rewards those who can resolve problems quickly, those who can keep production going, and those who can spend less money. Over time, employees learn an unwritten rule: safety can be emphasized all you want, but progress must not actually be delayed. Management reflection: What companies need to check is not only whether safety protocols exist, but also whether they were recognized positively over the past year for insisting on shutdowns, delays, reduced operations, or retesting ； Has there been any instance where rushing things in order to meet deadlines was tolerated or even praised? Without that, the so-called “safety first” has not yet become an organizational value. 2. “Empiricism” replaces “reverence for processes”. The hazardous materials involved in many accidents are not unfamiliar to us; ethanol, LPG, nitrogen, hydrogen sulfide, hydrogen peroxide, peroxides, chlorates, organic solid waste, hydrogen, and ethylene oxide all fall into the category of typical high-risk substances. Yet accidents still occur, indicating that there is no genuine respect for the hazards associated with materials at the workplace. Instead, scientific judgment is replaced by statements like “We’ve always done it this way,” “This equipment has always functioned like this,” “A small amount shouldn’t be a problem,” and “I’ll just take a quick look inside.” Experience itself is valuable; however, in chemical production, it must be subordinate to material properties, energy states, equipment limitations, and process safety analysis. The risks associated with hazardous chemicals do not decrease just because the operator is familiar with them. The more familiar a position is, the more prone one is to complacency ； The more repetitive the operation, the easier it is to mistake anomalies for normal background. Management reflection: In corporate culture, there should be less reliance on the absolute authority of those with extensive experience who say something is acceptable, and more discussion about what the test data indicate, what the boundaries of the manufacturing process are, and what the worst possible outcomes could be. A truly mature safety culture does not rely on courage and experience to suppress risks, but rather on evidence, boundaries, and review to manage them. 3. “Heroic rescue attempts” conceal the lack of “professional rescue”; in many cases of poisoning and suffocation, casualties increase as a result of reckless attempts to provide help. On the surface, it seems that the employees are eager to save lives ； Deeper down, it’s that enterprises have long failed to make it a part of their culture that “rescue operations must also adhere to professional rules.” In many incidents, rushing in immediately is still regarded as brave, while waiting for testing, wearing air respirators, and organizing external rescue efforts are misinterpreted as being slow to act. This culture is very dangerous. At a hazardous chemicals accident scene, bravery is not rewarded; what is rewarded are control, discipline, and professional equipment. Hypoxia, hydrogen sulfide, toxic fumes, and flammable gases do not reduce their harm just because the motives behind rescue efforts are good. Unprotected rescue is essentially not rescue at all; it turns one accident scene into multiple victim scenes. Management reflection: Companies must redefine \"courage.\" What is truly praiseworthy is not charging in without protection, but rather the ability of some people to overcome emotional pressures when their colleagues fall, to set up barriers, shout to stop the action, call for help, ensure proper ventilation, conduct inspections, put on air respirators, and carry out rescue operations according to established procedures. 4. “Compliance culture regarding permits” has replaced the concept of “genuine on-site safety.” Hot work permits, confined space permits, maintenance permits, and start-up/shutdown confirmation forms were originally tools for risk control; however, in some enterprises, they have become mere documents used to meet inspection requirements. As long as all the permits are in order, the organization considers the management process to be complete ； As for whether proper isolation was in place at the site, whether there were still residual liquids and gases, whether the testing points reflected the actual risks, and whether the supervisors understood the manufacturing process, these issues were not properly investigated. The essence of this culture is downgrading safety management from “controlling risks” to “proving that one has done something”. Once managers prioritize the appearance of records over on-site safety, employees at the lower levels will naturally resort to practices such as filling in documents later, having others sign on their behalf, using pre-made templates, and taking photos as proof without verifying the actual situation. Management reflection: Tickets only prove that a process has been initiated, but not that risks are under control. Companies should shift the focus of evaluation from \"whether there are tickets\" to \"whether the key conditions on the tickets have been verified on-site\". The safety management team should regularly ask counterquestions: If this ticket is fake, which control point on site could detect that it is fake. 5. The risky mindset that \"as long as nothing happens, it’s safe\" – issues such as the use of temporary fixtures over extended periods, repeated repairs and welding, operation despite defects, bypassing interlocks, missing safety valves, and continued production despite frequent alarms – often exist before an accident occurs. They were not dealt with in a timely manner because the organization mistook \"no problems in the past\" for \"acceptable risk.\" Chemical accidents have a cruel characteristic: the system can continue to operate despite being faulty for a long time, appearing to be functioning normally, but the risk does not disappear – it only keeps accumulating. Once exposed to high temperatures, fluctuations, misoperations, leaks, ignition sources, abnormal backflow, or rescue errors, problems that have accumulated over time will all become apparent at the same time. Management reflection: Companies cannot replace process safety with outcome safety. The absence of an accident is merely a result, not evidence. Truly reliable evidence is verifiable device status, effective protection layers, reported anomalies, resolved risks, and adherence to shutdown standards. 6. \"Hierarchical compliance\" weakens the \"on-site shutdown right.\" In many accident sites, there are actually opportunities to stop operations: abnormal odors, rising pressure, connection issues, unknown valve positions, lack of monitoring, failure to perform purging, equipment that has not been shut down, and personnel attempting to enter confined spaces. But frontline employees often do not have a real right to stop, or dare not use that right. The reason might be concern about being seen as uncooperative in production, delaying progress, causing trouble for managers, or even facing accountability. This indicates that an excessive emphasis on hierarchical obedience and result orientation still exists in the corporate safety culture. Frontline workers are closest to the risks, yet they may have the least say ； Management is located far from the site, yet it controls the final pace. This structure dilutes danger signals at each level. Management reflection: A key indicator of the maturity of a safety culture is whether frontline employees can say “no” without hesitation. Companies must be clear: anyone who discovers that there is no isolation, no testing, no replacement of personnel, no supervision, and no rescue measures in place has the right to stop work ； No accountability if stopped in the wrong place; accountability applies only if not stopped at all. 7. “Outsourcing risks” are often mistaken for “outsourcing responsibilities”. A common value discrepancy behind contractor accidents is that when companies delegate high-risk tasks to external teams, they also mentally transfer the responsibility for those risks. On-site, it may be assumed that the contractor is more professional, or that the construction team is responsible for everything, as a result of which the owner’s responsibilities regarding process instructions, on-site supervision, isolation verification, and validation of working conditions are weakened. However, the process risks of hazardous chemical enterprises cannot truly be outsourced. Contractors are familiar with welding, cutting, lifting, and cleaning, but they may not understand the residual media within the equipment, backflow paths, inerting conditions, areas where flammable gases can accumulate, and the consequences of abnormal operating conditions. If the owner does not incorporate process risks into the construction process, the contractor will carry out actions without being aware of those risks. Management reflection: Outsourcing can only be applied to labor and professional services; safety responsibilities cannot be outsourced. Those who possess the equipment and know the relevant techniques are responsible for clearly explaining the risks, verifying the conditions, and ensuring ongoing supervision at the site. 8. A “culture of accountability” is stronger than a “culture of learning”. Holding people accountable after accidents is necessary; however, if companies only enforce accountability after incidents occur, without encouraging the proactive reporting of near-misses, abnormal signals, warning signs, and factors that may lead to violations, this will give rise to a culture of silence. Employees tend to hide minor problems, while work teams prefer to handle abnormalities internally; as a result, the situation on site appears increasingly “normal” to management. A true safety culture isn’t about having no problems; rather, it’s about problems being detected, exposed, discussed, and corrected early on. An organization that reports minor anomalies every day might be safer than one that never reports any problems. Because the former still has the ability to learn, while the latter may simply be at risk of experiencing a crisis yet. Management reflection: Companies should regard those who expose problems as part of the risk prevention mechanism, rather than as troublemakers. Positive incentives should be established for those who proactively report potential hazards, halt unsafe operations, detect fraudulent documents, or point out equipment abnormalities. 9. Comparison of Cultural Roots and On-site Performance: Cultural and value discrepancies can lead to a chain of accidents; management should correct these issues. Prioritizing production over risk means skipping isolation, replacement, testing, and rechecking procedures in order to meet deadlines. When special operating conditions are not met, leaks, explosions, and poisonings can occur. Shutdowns, delays, and reduced operations should be considered part of positive safety performance. Empiricism replaces respect for processes; people rely on experience, assuming that small amounts, short durations, or past successes mean safety. The danger of materials and the consequences of abnormal conditions are misjudged; instead, testing data, process boundaries, and worst-case analysis should be used instead of relying on verbal experience. Heroic attempts at rescue replace professional responders; people enter the scene without protection when someone is in danger, leading to more casualties. Rescue protocols need to be revised to emphasize that no rescue should take place without proper protection. Compliance with paperwork replaces actual on-site safety measures; even if all paperwork is in order, if isolation, testing, and supervision are inadequate, work permits lose their function in risk control. Approval forms should be replaced by on-site verification forms, with key points being traceable. Assuming that no accidents occur means continuing to operate equipment despite defects, using temporary measures indefinitely, and treating alarms as routine. Equipment defects and abnormalities accumulate until they cause problems; risks should be measured using process safety indicators, rather than just focusing on accident outcomes. Hierarchy-based decision-making weakens the right to stop operations; frontline workers detect abnormalities but dare not halt operations, resulting in delayed or suppressed risk signals. Everyone should be authorized to stop operations; no penalties should be imposed for wrongful stops, but strict actions should be taken if operations are not stopped. The risks associated with outsourcing are often mistaken for the contractor’s responsibility; contractors may not be aware of the risks associated with the equipment, and owner oversight weakens, leading to uncontrolled activities such as welding, tank cleaning, and maintenance. The owner must assume responsibility for providing process risk information and verifying on-site conditions. A culture of accountability is more important than a culture of learning; warning signs and minor abnormalities are not reported, preventing the organization from identifying potential accidents early. Encouragement should be given for the proactive reporting of near-misses, abnormalities, and potential hazards. 10. Recommendations for Management to Examine Their Culture: Managers should regularly assess their culture using the following questions. They reveal the true culture better than typical safety slogans: 1. When was the last time operations were voluntarily halted, postponed, or had their load reduced due to safety reasons? Was the person later praised, or was it considered that they affected production? 2. Do front-line employees believe they have the authority to stop tasks assigned by managers, contractor work, and emergency repairs? Are there any real-world examples to prove this? 3. Has the work order approver personally visited the site to verify the key isolation points, measurement values, and rescue conditions? Did you only look at the written materials? 4. In the face of repeated alarms, leaks, repairs and welding, as well as the prolonged use of temporary fixtures, do companies take the initiative to stop operations to resolve these issues, or do they habitually continue operating? 5. When employees report warning signs, near-misses, and management shortcomings, do they receive protection and rewards, or scrutiny and pressure? 6. After the contractor enters the site, does the owner still consider the risks as their own rather than those of the contractor? 7. Does the team truly understand the principle of \"no rescue without protection,\" or do they still consider rushing in to save people as a praiseworthy instinctive reaction? 8. When conducting on-site inspections, do leaders pay more attention to the completeness of record-keeping, or are they more inclined to ask whether the working conditions actually exist? 11. The foundation of cultural improvement: Culture is not slogans posted on walls, but the choices an organization makes repeatedly day after day. To address the underlying value issues contributing to these accidents, it’s not enough to simply hold another meeting or sign another commitment letter; rather, the correct values must be integrated into evaluations, delegation of authority, rewards and punishments, and on-site decision-making. It is recommended to start with four actions: 1. Incorporate \"proactive shutdown, proactive reporting, and proactive escalation of anomalies\" into positive incentives to create visible examples. 2. Designate “risking deadlines, forging permits, conducting rescues without protection, dismantling under pressure, and entering without inspection” as non-negotiable red lines. 3. Shift management’s security responsibilities from \"holding meetings to give instructions\" to \"on-site verification\", with leaders must regularly conduct random checks to ensure the authenticity of key operations. 4. Shift accident learning from merely reporting incidents that happened to others to examining where in one’s own system such incidents could occur, thereby creating a closed loop for identifying potential risks. In one sentence: Technical issues determine how accidents occur, management issues determine why such accidents are not prevented, and cultural and value-related issues determine why similar accidents keep happening. VIII. Safety reminder phrases applicable to pre-shift meetings
During the hot month of June, the greatest risks on site do not stem from normal production activities; rather, they arise from hot work, working in confined spaces, inspection and maintenance, loading/unloading operations, startup/shutdown procedures, and handling of abnormal situations. Today, whenever it is necessary to turn on equipment, disconnect pipelines, enter tanks or pits, perform welding or heating operations, or deal with blockages and leaks, it is essential to first stop and verify whether isolation has been carried out, whether pressure has been relieved, whether the area has been purged, whether inspections have been done, whether supervision is in place, and whether rescue equipment is available. Upon discovering someone fallen, no one should enter to carry out rescue efforts without proper protection; first call the police, ensure ventilation, conduct inspections, wear air respirators, and then organize the rescue operation. The biggest lesson these June incidents imparted to production management is this: Accidents usually occur not because employees fail to understand that “safety is important,” but because key risky operations are not properly managed on the ground, and “safety first” is not truly made the organizational choice when conflicts arise. Production management should shift from post-event accountability to pre-event identification, in-process verification, immediate suspension in case of abnormalities, and closed-loop corrective actions. Truly effective safety management does not consist of creating more rules and regulations; rather, it ensures that for every instance of working with fire, entering a tank, performing disassembly tasks, starting up equipment, or dealing with abnormal situations, there are clear designated responsible persons, inspection points, shutdown conditions, and rescue preparations in place ； It also ensures that everyone knows that stopping when a risk is identified, speaking up about it, and standing up to pressure are behaviors that companies truly encourage.