Create a procedure for a construction site that is exposed to public riots

A construction site procedure for public riots or civil unrest should be built into the written emergency response plan and integrated with the site safety and health plan. At minimum, the plan should define pre-emergency coordination with police, fire, EMS, client representatives, and neighboring properties; personnel roles, lines of authority, training, and communication; emergency recognition and prevention triggers; safe distances and places of refuge; site security and control; evacuation routes and procedures; emergency medical treatment and first aid; emergency alerting and response procedures; PPE and emergency equipment; and post-incident critique and follow-up. The plan should be reviewed periodically and amended for changing site conditions, and it should be compatible with local and state emergency plans. [2] [3] [7]

Risk assessment and activation criteria should be specific to civil unrest.

• Monitor credible intelligence sources, police advisories, protest routes, curfews, transport disruptions, and threats to the project, workers, deliveries, fuel storage, cranes, and public-facing access points.
• Use a documented dynamic risk assessment that considers likelihood and severity of violence, arson, vandalism, forced entry, blocked egress, vehicle intrusion, utility interruption, and exposure of workers to projectiles, smoke, or crowd movement.
• Define clear escalation triggers such as unrest within a set radius, police request, road closures, attempted trespass, fire nearby, or loss of communications; each trigger should correspond to actions such as suspend work, shelter, partial lockdown, or full evacuation.
• Reassess conditions continuously during the event and after any major change in site conditions.

[1] [10] [3] Use an Incident Command System for any riot or civil unrest event affecting the site. The senior responsible emergency official should assume command until relieved under the established chain of authority. All communications, decisions, and coordination with police, fire, medical services, security contractors, and employer representatives should flow through that command structure. For larger events, assign functional leads for safety, evacuation/accountability, medical/first aid, site control, logistics, and media/client communications. [3] [12] [11]

Site lockdown and security controls should focus on preventing worker exposure and unauthorized entry, not confrontation.

• Stop non-essential work, secure cranes and suspended loads, shut down hot work, isolate hazardous energy where safe to do so, and secure fuel, gas cylinders, flammables, explosives, keys, and mobile plant.
• Lock or guard gates, close perimeter openings, remove or immobilize unattended equipment, and restrict access to one controlled entry point if sheltering in place is safer than evacuation.
• Establish site control zones, internal refuge areas, and exclusion areas near perimeter fencing, glazing, and street frontage. Keep workers away from roofs, scaffold edges, and exposed perimeter positions unless needed for safe shutdown.
• Do not direct workers to physically engage rioters or protect property. Security personnel, if used, should operate under a defined scope, communication protocol, and liaison with police.

[9] [5] [1] Worker protection measures should include training, accountability, communications, first aid readiness, and hazard-appropriate PPE. Train workers on alarm signals, lockdown versus evacuation criteria, refuge locations, buddy accountability, prohibited actions, and how to report threats. Maintain a reliable employee alarm or mass-notification system capable of stopping work and initiating emergency procedures. PPE for civil unrest is normally task- and hazard-based: high-visibility clothing may need to be removed if it increases targeting risk off-site; hard hats, eye protection, gloves, sturdy footwear, and hearing/respiratory protection may be needed depending on debris, smoke, or chemical irritants. PPE selection should be based on hazard assessment, recognizing that PPE itself can impair vision, mobility, communication, and heat tolerance. [6] [3] [12]

Public interface controls are critical where the site borders streets, sidewalks, transit stops, or occupied buildings. Pre-plan barricades, hoarding integrity, signage, traffic diversions, and methods to suspend deliveries and visitor access. Designate one management spokesperson for police, regulators, neighbors, and media; all other workers should be instructed not to argue with crowds or provide unofficial statements. If unrest affects adjacent public areas, coordinate crowd and traffic control with law enforcement and local authorities rather than improvising site-led control measures. [7] [12] [2]

Evacuation and shelter procedures should be predetermined and simple.

• Identify primary and alternate evacuation routes that avoid likely crowd routes, chokepoints, and glass-fronted streets. Include pedestrian and vehicle evacuation options, transport assembly points, and a method to assist injured or mobility-impaired workers.
• If external conditions are more dangerous than remaining on site, move workers to hardened refuge areas away from the perimeter, maintain accountability, and wait for police or incident command direction.
• Use roll calls or electronic accountability to confirm all workers, visitors, and subcontractors are accounted for. Maintain current attendance and contractor sign-in records to support this.
• Do not release workers individually into an active riot zone without route verification, transport coordination, and communication capability.

[4] [8] [1] Incident reporting and post-incident management should be formalized. Require immediate reporting of threats, trespass, assaults, vandalism, arson attempts, blocked exits, suspicious packages, and police interactions. Preserve evidence where safe, including CCTV footage, photos, witness statements, access logs, and damage records. Notify required internal leaders, insurers, clients, and government agencies as applicable. After the event, conduct a documented critique covering what happened, what controls worked, injuries or near misses, corrective actions, and whether the emergency plan, training, staffing, or physical security measures need revision. [7] [3] [2]

Business continuity planning should address how to protect life first while preserving the project’s ability to recover. Before unrest occurs, identify critical operations that must be safely shut down, critical records that must be backed up off-site, alternate suppliers and delivery windows, remote management capability, emergency payroll/contact processes, and criteria for phased restart. After the incident, inspect structural stability, utilities, fire protection, hazardous materials storage, access control, and plant/equipment before reoccupancy. Resume work only after hazards are reassessed, damaged controls are restored, and workers are briefed on changed conditions. [7] [10] [1]

For compliance, the employer should ensure the riot/civil unrest procedure is documented, site-specific, trained, rehearsed, and accessible to workers; integrated with the broader emergency response plan and site safety plan; supported by hazard assessments, PPE provisions, first aid arrangements, alarm/communication systems, and evacuation procedures; and coordinated with external emergency agencies. In Alberta, the OHS Code index specifically identifies emergency response plans, evacuation procedures, hazard assessment, first aid, PPE, training, and unauthorized entry as required control topics. In OSHA/MIOSHA-based frameworks, the same core elements recur: written emergency planning, site security and control, evacuation, alerting, first aid, PPE, ICS, training before real emergencies, periodic review, and follow-up critique. [1] [5] [6]

A practical site procedure can therefore be summarized as: assess threat, activate ICS, notify workers, stop and secure operations, decide lockdown versus evacuation, coordinate with police and emergency services, protect and account for workers, control public interfaces, document and report the incident, then reassess and recover before restarting work.