What a plant business continuity plan must contain
A business continuity plan (BCP) answers one question: how will the plant maintain or quickly restore critical processes when power, the control system, communications, a supplier or people fail. ISO 22301:2019 treats the plan as part of a business continuity management system (BCMS), not as a stand-alone document. It covers business impact analysis, strategies, response plans, exercises and management review.
In an industrial plant the plan has three features that office plans lack. First, physical processes have no "restore from backup" button: a furnace, a reactor or a continuous line needs a safe shutdown and a controlled restart. Second, continuity depends on OT, meaning controllers, SCADA and industrial networks, which have a different life cycle than IT. Third, people work in manual mode during a failure, so the plan must also include health and safety requirements for that mode.
- scope and purpose of the plan, and a list of critical processes with RTO and MTPD values
- roles: plan owner, incident manager, communications coordinator, health and safety coordinator
- event scenarios with trigger thresholds and a decision tree
- manual mode procedures and safety rules for that mode
- resources: backup power, spare parts, configuration backups, contact details in paper form
- exercise schedule, pass criteria and update procedure
Steps 1 and 2: scope and business impact analysis
The scope of the plan is set after talking to the heads of production, maintenance, IT/OT and health and safety. Not everything has to be covered by one plan: it makes sense to start with the processes whose stoppage causes the greatest loss or hazard, and widen the scope later. An entity that is essential or important under the Polish National Cybersecurity System Act (KSC) should include every service covered by its entry in the register.
Business impact analysis (BIA) assigns two numbers to each process. RTO (Recovery Time Objective) is the time within which the process must be restored. MTPD (Maximum Tolerable Period of Disruption) is the limit beyond which the damage becomes unacceptable. RTO is always shorter than MTPD, and the gap is a buffer for whatever goes wrong. For data assets there is also RPO, the acceptable data loss measured in time since the last backup.
Questions for the BIA interview
Ask the same questions for every process: what happens after 1 hour, 8 hours and a full day of downtime, which resources are essential (people, utilities, systems, suppliers), what the upstream and downstream dependencies are, and whether a manual workaround exists and how long it can be sustained safely. The answers go into a single table, from which the scenarios are later built.
Line recovery time after loss of control: without and with a plan (sample calculation)
- Without a plan (32 h)
- With a plan (8 h)
Step 3: scenarios and plan activation thresholds
A scenario describes an event from the point of view of its effect, not its cause. Loss of power to a section, unavailability of the control system, loss of communications, a ransomware attack on servers, failure of a key supplier and unavailability of the crew are six effects that cover most threats. The same "loss of the control system" scenario will handle a hardware failure, an error after an update and a cybersecurity incident.
An activation threshold is an unambiguous criterion after which the shift supervisor declares the plan active without asking for approval. A good threshold is measurable: no system response for more than 15 minutes, a parameter falling below its limit value, no contact with a section for longer than 10 minutes. If the threshold requires judging "is this serious", the decision will come too late. The plan should state who is authorised to declare activation and who deputises at night and at weekends.
- a threshold written as a number and a unit, not as an adjective
- two people authorised to declare the plan on every shift
- a template activation message, ready to be read out over radio or telephone
Step 4: roles, communications and the paper version
In the plan, name functions rather than people, and assign two people to each function. The plan owner is responsible for keeping it current and for exercises. The incident manager decides on switching to emergency mode. The communications coordinator runs the information channels to the crew, to management and, where regulations require it, to the authorities. The health and safety coordinator approves how work is done in manual mode and knows who is physically present on site.
The plan must be available when the network and the file server are down. A paper version in the control room, at the command post and with shift supervisors is a practical requirement, not a formal one. The same applies to the contact list: telephone numbers of key people and suppliers written on a sheet that resists moisture and dust. The paper version is updated together with the electronic one and carries an issue date.
Step 5: sample recovery time calculation
The sample calculation shows how the recovery time of a process is built up after the loss of a line control system. The values are illustrative assumptions, not data from any specific plant. Each plant substitutes its own figures from the BIA and from exercises.
Assumptions for the variant with a plan: detection 0.5 h, decision to activate the plan 0.5 h, switch to manual mode 1 h, system restoration from backup 4 h, verification 1 h, controlled line restart 1 h. The total is 8 hours, which is the RTO adopted in the plan. For the variant without a plan: detection 1 h, establishing who decides 4 h, switch to manual mode 6 h, restoration without a tested backup 16 h, verification 3 h, restart 2 h. The total is 32 hours.
The 24-hour difference comes mainly from three places: the decision (3.5 h), manual mode (5 h) and restoration (12 h). The plan and exercises eliminate the first two; a tested configuration backup with proven restoration eliminates the third. This shows where to invest first.
Step 6: exercises, review and audit evidence
ISO 22301:2019 requires plans to be exercised and tested at planned intervals, with the results evaluated by management. A practical cycle is a tabletop exercise once a quarter, a live test of one scenario every six months, and a full plan review after every significant event and technology change. Every exercise leaves a record: what worked, how long it took, what to improve, who is responsible and by when.
Article 21 of the NIS2 Directive lists business continuity, including backup management and disaster recovery, among the risk management measures. The plan together with exercise records provides audit evidence that the entity has implemented and maintains this measure. The plan alone does not mean compliance: it supports meeting the requirement as long as it is current and exercised.
Software helps maintain continuity at the operational level. An evacuation list built from gate scans shows who is on site, and the InnerWeb instance runs in containers on the customer's own server, independent of the cloud. This complements the plan and does not replace it.
Process recovery time after loss of a line control system
| Parameter | Value |
|---|---|
| Detection and decision with a plan | 1 hour |
| Switch to manual mode with a plan | 1 hour |
| System restoration from a tested backup | 4 hours |
| Verification and restart with a plan | 2 hours |
| Restoration without a proven backup | 16 hours |
Result: About 8 hours with a plan, about 32 hours without one, a difference of about 24 hours per event.
Note: Sample calculation: illustrative values, not data from a specific plant. Substitute your own figures from the BIA and from exercises.
Video
Frequently asked questions
How does a business continuity plan differ from an emergency plan?
An emergency plan describes the response to a specific event, for example a fire or a leak. A continuity plan covers maintaining or restoring critical processes after any event, based on BIA, RTO and resources. Emergency plans are its components.
Where to start if the plant has no plan at all?
Start with a BIA for the three to five most important processes and one scenario, for example loss of the control system. A first version of a dozen or so pages, exercised at the table, adds more value than a long document that has never been tested.
Does a continuity plan require ISO 22301 certification?
No. ISO 22301 is a voluntary standard. However, the Polish National Cybersecurity System Act (KSC) requires implementing, documenting, testing and maintaining business continuity plans, so the standard is a good methodological framework, but a certificate alone does not prove compliance.
How often should the plan be updated?
At least once a year, and after every significant change in technology, organisational structure or after an incident. Contact details are worth verifying every quarter.
Sources
Let us check what in your continuity plan needs evidence
Let us talk about how the on-site personnel list, permits and access control support emergency mode in your plant. We will show what runs on the customer's server, independent of the external network.
Let's talk