A Comprehensive Approach to Safety Management
In the ever-evolving landscape of operational technology, the imperative to maintain safety within data centers cannot be overstated. These hubs of digital operations not only power our global economy but also present unique risks that demand meticulous management.
Following our recent webinar on Managing Safety Risks in Operational Data Centers, this article delves into the comprehensive strategies essential for upholding safety in these critical environments, encompassing procedures, training, assessments, inspections, and compliance.
This article provides:
- A comprehensive approach to safety management.
- Recording of the webinar
- List of audience questions and answers from the host
Procedures
At the heart of data center safety is the development and implementation of clear, detailed procedures. These guidelines are not solely for complex tasks like switching activities or working in confined spaces; even seemingly mundane tasks require attention to safety. Job Hazard Analysis (JHA) emerges as a pivotal initial step, distinguishing when a full procedure is necessary versus when simple instructions suffice. This meticulous approach, bolstered by the art of writing and managing procedures, ensures that every action taken within a data center is guided by a thorough hazard/risk assessment and implementation of safe work practices.
The journey of crafting these procedures involves everyone from in-house teams to external vendors and subcontractors. This collaboration is crucial, transforming individual knowledge into collective safety protocols. This collaborative spirit extends to managing Job Hazard Analyses, integrating them with standard operating procedures to streamline safety across all operations.
Training
Equipping employees with the knowledge and skills to execute procedures safely is paramount. A structured training process, leveraging everything from commissioning scripts to advanced Learning Management Systems (LMS), ensures that every team member—from technicians to managers—understands and adheres to established safety practices. This education extends to hands-on skills assessments, a blend of theoretical and practical learning, and innovative approaches like Virtual Reality (VR) to simulate high-risk scenarios in a risk-free environment.
Training is an ongoing journey, one that requires regular assessments and updates to keep pace with technological advancements and operational changes. It’s a multifaceted approach, designed to ensure that employees are not just familiar with the procedures but are proficient in executing them safely and effectively.
Assessments
Maintaining safety in data centers extends beyond initial training to include continuous assessments of skills, facility operations, and the competency of vendors and subcontractors. From pre-hire technical interviews to ongoing skills evaluations and facility inspections, these assessments ensure compliance with safety protocols and operational efficiency. They highlight the importance of understanding the human element in operational risk, emphasizing regular observation and engagement with all personnel to foster a culture of safety and responsibility.
Inspections
Leveraging technology like Computerized Maintenance Management Systems (CMMS) enhances the ability to manage inspections and ensure compliance within Environmental Health and Safety (EHS) programs. By documenting inspections, tracking corrective actions, and integrating safety checks into daily operations, data centers can maintain a high level of safety and operational integrity. This proactive approach, coupled with regular facility inspections and the incorporation of safety equipment checks, underscores the commitment to a safe working environment.
Documentation and Compliance
The role of comprehensive documentation and adherence to compliance standards cannot be understated in the context of data center operations. Whether opting for electronic or physical storage, the accessibility and regular review of documents are critical. Training employees on how to access this documentation ensures everyone is equipped with the knowledge needed to operate safely. Moreover, understanding and meeting regulatory requirements is essential for legal compliance and the overall safety of operations. This encompasses everything from trade licenses for personnel and vendors to regular audits of documentation and compliance with current laws.
The management of safety in operational data centers requires a holistic approach, integrating procedures, training, assessments, inspections, and compliance into a seamless safety management system. By prioritizing these elements, data centers can not only mitigate risks but also foster a culture of safety and responsibility. This ensures not just the operational uptime and efficiency of the data center but also the well-being of everyone involved in its operation. As technology evolves and operational demands shift, the commitment to safety must remain steadfast, underpinning every action and decision within the data center environment.
Managing Safety in Operational Data Centers in 2025
Our 2025 webinar provides health and safety professionals, facility managers, and operations leaders with the latest strategies and best practices for managing safety risks in data centers. Our panel shares their experiences and knowledge to help you navigate the dynamic landscape of data center safety.
Key topics include:
- Advancements in data center safety protocols since 2024
- Integrating existing technologies to improve worker safety
- Addressing emerging risks in modern data center operations
- Case studies highlighting successful safety implementations
- Cultivating a proactive safety culture within your organization
Too often, safety professionals are seen as enforcers rather than problem solvers. I believe in identifying common challenges and providing real solutions instead of just citing regulations. It’s about working with people to find the best path forward. The past year has seen increasing emphasis in how data center owners and operators approach risk management. Protection from electrical energy is paramount because shock can kill, and arc flash can kill. It’s not just about equipment damage - it’s about protecting human lives. This webinar is a great opportunity to educate those who may not fully understand what an electrical safety program should encompass.
Watch 2024 Managing Safety Risks in Operational Data Centers Video
Your Questions, Our Answers
How should procedures be authored to ensure safety in data centers?
To ensure safety in data centers, it’s crucial to have someone with quality assurance and quality control (QAQC) experience. Individuals with military backgrounds, particularly those with nuclear Navy experience, bring valuable procedural discipline, ensuring consistency and quality in the development of safety procedures.
Who should review safety procedures to ensure compliance and accuracy?
Subject matter experts (SMEs) should be involved in reviewing procedures within their areas of expertise. This ensures that the procedures are accurate, complete, and compliant with company standards and regulatory requirements.
How can we effectively train employees on safety procedures in an SOP-driven environment?
Training should go beyond computer-based methods and include hands-on training. Shadowing experienced engineers and conducting skills assessments are key ways to ensure that employees understand and can apply safety procedures effectively.
What is the impact of NFPA 70B on preventative maintenance?
NFPA 70B sets mandatory requirements for preventative maintenance, ensuring that electrical equipment operates safely and effectively. Proper maintenance helps prevent downtime and is critical for both worker safety and operational uptime.
How can we get management and supervisors to support safety initiatives?
Safety initiatives require visible leadership commitment. When leadership actively supports and participates in safety efforts, it sets the tone for the rest of the organization and encourages other employees to follow suit.
How should safety be managed in hot and cold aisles in data centers?
To manage safety in hot and cold aisles, it’s important to limit exposure time, use cooling vests, and have protocols in place to prevent heat stress. These measures help protect operators working in high-temperature areas.
How do we balance operational uptime with electrical safety?
While uptime is critical, safety must always be the top priority. Proper training, following procedures, and creating a safety culture can help balance operational needs with the protection of workers.
How can safety professionals influence leadership to prioritize safety?
Safety professionals can influence leadership by using data to show the long-term benefits of safety initiatives. Linking safety to performance metrics and demonstrating cost savings helps integrate safety into the organization’s core priorities.
What are the best alternatives to ladders for safe access?
The JLG EcoLift 50 and the JLG 830P are good alternatives to ladders. The EcoLift 50 is manually operated and doesn’t require batteries, while the JLG 830P is battery-operated, offering more convenience for frequent use.
How should last-minute scope changes be managed safely?
For last-minute scope changes, it’s essential to engage stakeholders, conduct a Job Hazard Analysis (JHA), identify necessary PPE, and review safety procedures to ensure a safe and efficient change process.
How can safety procedures be consistently implemented across multiple sites?
Ensuring consistency across multiple sites requires leadership buy-in, designating local responsibility, maintaining regular communication, and providing ongoing training to create a unified safety culture.
How do we address regulatory differences in European markets?
To address regulatory differences in Europe, it’s important to have Health, Safety, and Environmental (HSE) professionals in each region to meet local requirements. Adopting stricter standards as best practices where beneficial can further ensure safety.
How do we balance regulated safety with managed safety?
Data centers are moving toward a managed safety approach, where employees are trained not just in standard operating procedures (SOPs) but also in proactive safety measures, skill development, and qualification programs.
Why is incident reporting so important for safety?
Reporting all incidents, even minor ones, is crucial for identifying patterns and preventing future injuries. A proactive approach, including engaging workers in safety decisions, helps create a culture of continuous improvement.
What do operators expect from safety professionals in data centers and construction?
In construction, safety professionals need to oversee various trades with a focus on quality and efficiency. In data centers, safety professionals are expected to have experience in the data center environment or relevant trades to ensure safety in specialized operations.
Is live electrical work necessary in data centers?
While live electrical work may be necessary for certain tasks, it should be avoided as much as possible. Redundant systems in data centers typically allow for de-energization, reducing the need for energized electrical work.
How should noise levels be managed in data centers?
Noise levels in data halls should be managed by conducting noise surveys to determine where hearing protection is needed. In areas where noise levels exceed safe limits, mandatory hearing protection should be enforced.
How can safety be prioritized when leadership lacks awareness of injury costs?
When leadership is unaware of injury costs, safety professionals can engage peers in safety conversations and ask how they can assist. Building mutual respect and leveraging support from colleagues helps make the case for prioritizing safety.
How can leadership’s priorities shift after an incident?
After an incident, it’s important to ask leadership how they plan to maintain a focus on safety, ensuring that safety remains a top priority even as new leaders come into play or as attention shifts over time.
What technology can support the safety of lone workers?
The StaySafe app is a useful tool for lone workers. It offers a mobile solution to help keep workers safe when operating in isolated areas, such as mechanical rooms or rooftops.
How should procedures be authored to ensure safety in data centers?
To ensure safety in data centers, it’s crucial to have someone with quality assurance and quality control (QAQC) experience. Individuals with military backgrounds, particularly those with nuclear Navy experience, bring valuable procedural discipline, ensuring consistency and quality in the development of safety procedures.
Who should review safety procedures to ensure compliance and accuracy?
Subject matter experts (SMEs) should be involved in reviewing procedures within their areas of expertise. This ensures that the procedures are accurate, complete, and compliant with company standards and regulatory requirements.
How can we effectively train employees on safety procedures in an SOP-driven environment?
Training should go beyond computer-based methods and include hands-on training. Shadowing experienced engineers and conducting skills assessments are key ways to ensure that employees understand and can apply safety procedures effectively.
What is the impact of NFPA 70B on preventative maintenance?
NFPA 70B sets mandatory requirements for preventative maintenance, ensuring that electrical equipment operates safely and effectively. Proper maintenance helps prevent downtime and is critical for both worker safety and operational uptime.
How can we get management and supervisors to support safety initiatives?
Safety initiatives require visible leadership commitment. When leadership actively supports and participates in safety efforts, it sets the tone for the rest of the organization and encourages other employees to follow suit.
How should safety be managed in hot and cold aisles in data centers?
To manage safety in hot and cold aisles, it’s important to limit exposure time, use cooling vests, and have protocols in place to prevent heat stress. These measures help protect operators working in high-temperature areas.
How do we balance operational uptime with electrical safety?
While uptime is critical, safety must always be the top priority. Proper training, following procedures, and creating a safety culture can help balance operational needs with the protection of workers.
How can safety professionals influence leadership to prioritize safety?
Safety professionals can influence leadership by using data to show the long-term benefits of safety initiatives. Linking safety to performance metrics and demonstrating cost savings helps integrate safety into the organization’s core priorities.
What are the best alternatives to ladders for safe access?
The JLG EcoLift 50 and the JLG 830P are good alternatives to ladders. The EcoLift 50 is manually operated and doesn’t require batteries, while the JLG 830P is battery-operated, offering more convenience for frequent use.
How should last-minute scope changes be managed safely?
For last-minute scope changes, it’s essential to engage stakeholders, conduct a Job Hazard Analysis (JHA), identify necessary PPE, and review safety procedures to ensure a safe and efficient change process.
How can safety procedures be consistently implemented across multiple sites?
Ensuring consistency across multiple sites requires leadership buy-in, designating local responsibility, maintaining regular communication, and providing ongoing training to create a unified safety culture.
How do we address regulatory differences in European markets?
To address regulatory differences in Europe, it’s important to have Health, Safety, and Environmental (HSE) professionals in each region to meet local requirements. Adopting stricter standards as best practices where beneficial can further ensure safety.
How do we balance regulated safety with managed safety?
Data centers are moving toward a managed safety approach, where employees are trained not just in standard operating procedures (SOPs) but also in proactive safety measures, skill development, and qualification programs.
Why is incident reporting so important for safety?
Reporting all incidents, even minor ones, is crucial for identifying patterns and preventing future injuries. A proactive approach, including engaging workers in safety decisions, helps create a culture of continuous improvement.
What do operators expect from safety professionals in data centers and construction?
In construction, safety professionals need to oversee various trades with a focus on quality and efficiency. In data centers, safety professionals are expected to have experience in the data center environment or relevant trades to ensure safety in specialized operations.
Is live electrical work necessary in data centers?
While live electrical work may be necessary for certain tasks, it should be avoided as much as possible. Redundant systems in data centers typically allow for de-energization, reducing the need for energized electrical work.
How should noise levels be managed in data centers?
Noise levels in data halls should be managed by conducting noise surveys to determine where hearing protection is needed. In areas where noise levels exceed safe limits, mandatory hearing protection should be enforced.
How can safety be prioritized when leadership lacks awareness of injury costs?
When leadership is unaware of injury costs, safety professionals can engage peers in safety conversations and ask how they can assist. Building mutual respect and leveraging support from colleagues helps make the case for prioritizing safety.
How can leadership’s priorities shift after an incident?
After an incident, it’s important to ask leadership how they plan to maintain a focus on safety, ensuring that safety remains a top priority even as new leaders come into play or as attention shifts over time.
What technology can support the safety of lone workers?
The StaySafe app is a useful tool for lone workers. It offers a mobile solution to help keep workers safe when operating in isolated areas, such as mechanical rooms or rooftops.
What training / qualification have you found to be exceptional to A- Manage EHS in the real world DC environment and; B - is particularly desireable of HSE Proffesionals in DC Project Team?
• A: Any and all training related to electrical safety and the control of hazardous energy (Lockout/Tagout) would be at the top of the list when working in the real world DC environment. The National Fire Protection Association has two certifications related to electrical safety that require a certain number of years working in electrical safety and require the passing of an exam, and is focused on the NFPA 70E – Standard for Electrical Safety in the Workplace.
B: On the project side, OSHA 30 hour training for construction would be a very good start, but much of what I have learned has come from years’ of experience on construction sites conducting risk assessments and hazard observations. Additional training in working at heights and excavation are a definite plus.
Have you run into situations where due to privacy and security concerns, server rooms are deemed restricted areas per customer request. How do you address the need to access to safety audits, inspections, etc?
• Some of our data center clients have customers who have server rooms that are considered restricted areas. However, OSHA (1903.7(c)) allows Compliance Safety and Health Officers (CSHOs) to access and inspect all areas of a facility necessary to conduct a thorough inspection provided they follow the required safety precautions. This ensures that all potential hazards at a facility are evaluated, and compliance with safety standards is verified. No area is exempt from safety inspections. Many clients have a provision in their customer agreements that allow them to access the server to perform required maintenance, fire inspections, and safety inspections to ensure all parties are aware of, and in agreement, with this regulatory requiremenet. (RS)
• Prior to visiting the site, I would recommend asking if there are any secure areas that have additional access requirements or require special permission. If access cannot be obtained, it would be appropriate to identify the specific room or area on the assessment form being utilized. (DH)
• Yes, pre-arranging security clearances in advance may be necessary, or ability to enter under escort. This includes the use of any pictures to document corrective actions. (DG)
Have you performed risk assessments in the newer AI data centers where electrical requirements are greatly increased, and liquid cooling? If so, can you speak to that a bit?
Risk Assessments that have been performed in newer AI data centers have typically identified the following key risk elements:
1. Increased Electrical Load and Power Demand
AI data centers require significantly higher electrical power due to the intensive computational tasks performed by AI systems. This increased demand can strain existing power infrastructure and necessitate upgrades to electrical systems to ensure reliability and safety. Key risks include: Overloading circuits: Higher power requirements can lead to overloaded circuits, increasing the risk of electrical fires. Power outages: Insufficient power supply can cause outages, disrupting operations and potentially damaging sensitive equipment.
2. Enhanced Cooling Requirements
Liquid cooling systems are increasingly used in AI data centers to manage the heat generated by high-performance computing equipment. While effective, these systems introduce new risks, potentially including: Leakage: Liquid cooling systems can leak, potentially causing electrical shorts and equipment damage. Chemical hazards: The coolants used may pose chemical hazards if not handled properly, requiring stringent safety protocols.
3. Electrical Safety Protocols
With the increased electrical load and the introduction of liquid cooling, adherence to NFPA 70e standards becomes even more critical. Key areas of focus include: Arc flash protection: Enhanced measures to protect against arc flash incidents, which can be more likely with higher electrical loads. Personal protective equipment (PPE): Ensuring that all personnel are equipped with appropriate PPE to mitigate risks associated with electrical hazards.
Raised access flooring is being phased out in some data centers due to safety incidents during construction. What’s your perspective on the use of RAF in modern facilities?
• Any time you design an access system that creates a ‘drop’ you introduce a new hazard and risk, therefore I believe that taking out that hazard and risk by designing a better access system can only be a good thing in terms of safety. The key to new design however, will be that during routine servicing and maintenance workers are not required to adopt awkward postures or use excessive forces (ergonimics issues) in order to complete their scope of works – NLF
* The other concern with raised flooring is that it has the potential to create a confined space due to equipment configurations where you have limited means of entry and egress, particularly if you have an entry point in one area but what needs to be maintained or serviced is in another area, requiring workers to crawl under the raised floor.
* The weight of the server racks themselves can create stability issues and there can be a tendency to exceed the capacity of the raised flooring system.
JSAs often focus narrowly on tasks, overlooking upstream/downstream impacts across teams or systems. How are organizations evolving JSAs to better address these interdependencies, especially in multi-contractor environments?
Job Safety Analyses (JSA) are an invaluable tool to help workers identify potential hazards associated with specific job tasks and develop strategies to mitigate these risks. The process involves breaking down a job into individual steps, analyzing each step for potential hazards, and recommending safety measures to prevent accidents and injuries minimize the risks as they perform the tasks. While JSA can be extended to consider upstream and downstream impacts, they are generally more effective for task-specific hazard identification and may not comprehensively address broader system interactions. For identifying broader impacts beyond the task level, risk assessment tools such as Hazard and Operability Study (HAZOP) can be used to identify and evaluate risks in complex systems, considering interactions between different components and teams. Where chemical safety is a primary concern, Process Hazard Analyses (PSM)/Process Safety Managemenet (PSM) are also valuable tools for identifying and managing upstream and downstream impacts, as well as interactions between different systems and teams. (RS)
* JSAs should focus on identifying the hazards associated with the work tasks and mitigation strategies for those hazards. A Method of Procedure (MOP) would then be developed to identify the steps of the work activity. Potential risks or points of failure, and impacts to upstream/downstream systems should be identified during the development process. This may require identifying SMEs (IT, Security, Mechanical/Electrical/Plumbing, etc.) during the development of the MOP to assist in identifying risks and operational impacts, including rollback or recovery steps in the event of an emergency or failure during the execution of the MOP. The steps of the MOP should be clear, concise, and actionable, and include who is to perform each step, who supervises and approves or verifies the work, and points for escalation if something goes wrong. Review and validate the MOP by condcuting a peer review or walkthrough with stakeholders (engineers, operations, project managers) and make any revisions based on feedback. If it’s possible, and highly recommended, pilot test the MOP in a controlled environment to ensure that upstream and downstream impacts have been identified. (DH)
• Another alternate practice can be the use of Task hazard Assessments (THAs) where the focus can include upstream/downstream impacts to a particular activity and how this affects multiple stakeholders. (DG)
ISO :42001:2023 AIIMS is also essential part of data centres nows Along woth ISO 27001
ISO/IEC 42001:2023 is the international standard for managing artificial intelligence (AI) systems. It provides a framework for establishing, implementing, maintaining, and continually improving an Artificial Intelligence Management System (AIMS) within organizations. As data centers increasingly integrate AI to optimize operations, enhance, security, and improve efficiency, implementing ISO 42001:2023 may help a company improve their risk management protocols against AI-related security threats, improve operational efficiency, reduce energy consumptions, and provide a documented management system to ensure compliance with evolving AI regulations.
Are these standards also applicable to BTC mining centers, or is there a need for additional guidance - such as a more holistic approach like NFPA 70E?
Both BTC mining centers and data centers require significant power demands and robust cooling systems and both must comply with the OSHA standards (including OSHA Subpart S and OSHA 1910.147) and with the required provisions of NFPA 70e. BTC mining centers typically have higher energy consumption, due to the higher need for computational power, which can lead to more stringent environmental regulations around noise and and use of fossil fuels that are used to power these centers.
