A risk register records the discrete things that might happen to a project: a late utility connection, a failed inspection, an unknown site condition, a long-lead equipment slip. Each entry has an owner, a probability, an impact range for cost and time, and a planned response. In a schedule risk analysis, each open risk is simulated as an event that occurs in some iterations and adds delay to the activities it affects.
A register can also hold opportunities, such as a faster permit route or a cheaper equivalent product, which is why some teams call it a risk and opportunity register.
What each entry needs
| Field | Example |
|---|---|
| ID and title | R-07 Utility power energized late |
| Cause, event, effect | Utility’s backlog → permanent power after planned date → testing and commissioning can’t start |
| Category | Third party / utilities |
| Owner | Project manager |
| Probability | 30% |
| Schedule impact | 10 to 40 working days on permanent power and successors |
| Cost impact | Extended general conditions, temporary power rental |
| Response | Book a temporary generator for commissioning |
| Residual impact | 2 to 8 working days |
| Trigger | No energization date from the utility 90 days before need |
| Status and review date | Open, reviewed at the monthly risk meeting |
Writing each risk as cause, event and effect keeps entries specific. “Weather” is not a risk you can manage. “Rain during the October roofing window delays dry-in and pushes interior finishes” is.
How it works
- Identify. Run a workshop at bid or preconstruction with estimating, operations, the superintendent and key trades. Use the lessons learned from similar jobs.
- Assess. Score probability and impact. A qualitative register uses bands, such as 1 to 5 on a probability-impact matrix. A quantitative one uses percentages, days and dollars.
- Plan responses. For threats, the usual options are avoid, transfer (through the contract, insurance or a subcontract), mitigate or accept. Each response has a cost and an owner.
- Model it. For schedule risk, tie each risk to the activities it affects and run a Monte Carlo simulation alongside the three-point estimates for ordinary variability.
- Review and update. Close risks that have passed, add new ones, and move risks that occurred into the change or issue process.
Worked example
Illustrative: R-07, “Utility power energized late”, has a 30% probability and an impact of 10 to 40 working days on permanent power and everything after it. If the impact is spread evenly across that range, its average is 25 days, so the expected value is 0.3 × 25 = 7.5 working days.
Mitigation, a temporary generator for commissioning, cuts the impact to 2 to 8 days. The expected value falls to 0.3 × 5 = 1.5 days.
Running the schedule model before and after mitigation shows the P80 finish moving 12 days earlier. That is more than the 6-day drop in expected value, because the risk sits on the critical path and the model captures what happens in the bad iterations, not just the average. If extended general conditions on this job run $6,000 per working day (illustrative), and the generator costs $18,000 to book and run, the model gives the team a concrete basis for deciding whether to buy it.
Risk register vs related tools
| Tool | Holds | Example |
|---|---|---|
| Risk register | Events that may happen | Late utility power, 30% |
| Issue log | Problems that have happened | Utility has confirmed a 3-week slip |
| Contingency | Money or time set aside for risks | $250,000 cost contingency, 15 days schedule margin |
| Three-point estimates | Ordinary variability in planned work | Steel erection 36 / 42 / 55 days |
Keep events in the register and normal variability in the duration ranges, so the same delay isn’t counted twice.
Why it matters
Duration ranges capture everyday variability; a register captures the events that cause step changes. Comparing results before and after mitigation turns a risk workshop into a decision about where to spend money. Contract terms matter too: who carries differing site conditions, force majeure or material price movement under an escalation clause changes which risks are yours to price.
Common mistakes
- Vague entries such as “weather” or “subcontractor performance” with no cause, event or effect.
- No owner, so nobody acts on the trigger.
- Built once and never updated. A register written for the bid and never opened again doesn’t manage anything.
- Adding up expected values and calling the total a confidence level.
- Double counting a risk in both the register and an activity’s pessimistic duration.
- Ignoring opportunities, which are often cheaper to capture than threats are to mitigate.
- Scoring with a matrix only, when the decision really needs days and dollars.
How os.construction handles it
The Risk page keeps a register with owner, category, open or closed status, probability, impact and optional post-mitigation values. Open events are simulated as duration impacts or as fragnets ahead of the affected activities, and the result ranks events by their contribution. See Risk analysis.
FAQ
What is a risk register in construction?
A risk register is a living list of the specific things that could go wrong (or better than planned) on a project. Each entry records the risk, its cause, an owner, the probability it happens, its cost and schedule impact if it does, the planned response, and its current status. It is reviewed and updated through the project.
What should a risk register include?
Typical fields are an ID, a description written as cause, event and effect, category, owner, probability, cost impact, schedule impact, a score or ranking, the response and its cost, residual probability and impact after the response, trigger or warning signs, status and the date last reviewed.
How do you calculate the expected value of a risk?
Multiply the probability by the impact. A 30% chance of a 25-day delay has an expected value of 7.5 days. Expected values help rank risks and size contingency, but adding them up doesn't give a confidence level. A Monte Carlo simulation of the schedule shows how risks combine along the critical path.
What is the difference between a risk register and an issue log?
A risk is something that might happen. An issue is something that has happened and needs managing now. When a risk occurs, it moves from the register to the issue log or change process, and the register records that it materialized.
What is a risk register called under NEC contracts?
NEC4 uses an Early Warning Register, which replaced the Risk Register of NEC3. It lists matters notified as early warnings and the actions agreed to avoid or reduce their effect, and it is reviewed at early warning meetings.